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

Pull drm updates from Dave Airlie:
 "Been a bit busy, first week of kids school, and waiting on other trees
  to go in before I could send this, so its a bit later than I'd
  normally like.

  Highlights:
   - core:
      timestamp fixes, lots of misc cleanups
   - new drivers:
      bochs virtual vga
   - vmwgfx:
      major overhaul for their nextgen virt gpu.
   - i915:
      runtime D3 on HSW, watermark fixes, power well work, fbc fixes,
      bdw is no longer prelim.
   - nouveau:
      gk110/208 acceleration, more pm groundwork, old overlay support
   - radeon:
      dpm rework and clockgating for CIK, pci config reset, big endian
      fixes
   - tegra:
      panel support and DSI support, build as module, prime.
   - armada, omap, gma500, rcar, exynos, mgag200, cirrus, ast:
      fixes
   - msm:
      hdmi support for mdp5"

* 'drm-next' of git://people.freedesktop.org/~airlied/linux: (595 commits)
  drm/nouveau: resume display if any later suspend bits fail
  drm/nouveau: fix lock unbalance in nouveau_crtc_page_flip
  drm/nouveau: implement hooks for needed for drm vblank timestamping support
  drm/nouveau/disp: add a method to fetch info needed by drm vblank timestamping
  drm/nv50: fill in crtc mode struct members from crtc_mode_fixup
  drm/radeon/dce8: workaround for atom BlankCrtc table
  drm/radeon/DCE4+: clear bios scratch dpms bit (v2)
  drm/radeon: set si_notify_smc_display_change properly
  drm/radeon: fix DAC interrupt handling on DCE5+
  drm/radeon: clean up active vram sizing
  drm/radeon: skip async dma init on r6xx
  drm/radeon/runpm: don't runtime suspend non-PX cards
  drm/radeon: add ring to fence trace functions
  drm/radeon: add missing trace point
  drm/radeon: fix VMID use tracking
  drm: ast,cirrus,mgag200: use drm_can_sleep
  drm/gma500: Lock struct_mutex around cursor updates
  drm/i915: Fix the offset issue for the stolen GEM objects
  DRM: armada: fix missing DRM_KMS_FB_HELPER select
  drm/i915: Decouple GPU error reporting from ring initialisation
  ...

480 files changed, 34609 insertions, 11749 deletions

diff --git a/drivers/gpu/drm/Kconfig b/drivers/gpu/drm/Kconfig
index f86427591167..8e7fa4dbaed8 100644
--- a/drivers/gpu/drm/Kconfig
+++ b/drivers/gpu/drm/Kconfig
@@ -20,6 +20,10 @@ menuconfig DRM
 	  details.  You should also select and configure AGP
 	  (/dev/agpgart) support if it is available for your platform.
 
+config DRM_MIPI_DSI
+	bool
+	depends on DRM
+
 config DRM_USB
 	tristate
 	depends on DRM
@@ -188,6 +192,10 @@ source "drivers/gpu/drm/tilcdc/Kconfig"
 
 source "drivers/gpu/drm/qxl/Kconfig"
 
+source "drivers/gpu/drm/bochs/Kconfig"
+
 source "drivers/gpu/drm/msm/Kconfig"
 
 source "drivers/gpu/drm/tegra/Kconfig"
+
+source "drivers/gpu/drm/panel/Kconfig"
diff --git a/drivers/gpu/drm/Makefile b/drivers/gpu/drm/Makefile
index cc08b845f965..292a79d64146 100644
--- a/drivers/gpu/drm/Makefile
+++ b/drivers/gpu/drm/Makefile
@@ -18,6 +18,7 @@ drm-y       :=	drm_auth.o drm_buffer.o drm_bufs.o drm_cache.o \
 drm-$(CONFIG_COMPAT) += drm_ioc32.o
 drm-$(CONFIG_DRM_GEM_CMA_HELPER) += drm_gem_cma_helper.o
 drm-$(CONFIG_PCI) += ati_pcigart.o
+drm-$(CONFIG_DRM_PANEL) += drm_panel.o
 
 drm-usb-y   := drm_usb.o
 
@@ -31,6 +32,7 @@ obj-$(CONFIG_DRM_KMS_HELPER) += drm_kms_helper.o
 CFLAGS_drm_trace_points.o := -I$(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-$(CONFIG_DRM_TDFX)	+= tdfx/
@@ -56,6 +58,8 @@ obj-$(CONFIG_DRM_SHMOBILE) +=shmobile/
 obj-$(CONFIG_DRM_OMAP)	+= omapdrm/
 obj-$(CONFIG_DRM_TILCDC)	+= tilcdc/
 obj-$(CONFIG_DRM_QXL) += qxl/
+obj-$(CONFIG_DRM_BOCHS) += bochs/
 obj-$(CONFIG_DRM_MSM) += msm/
 obj-$(CONFIG_DRM_TEGRA) += tegra/
 obj-y			+= i2c/
+obj-y			+= panel/
diff --git a/drivers/gpu/drm/armada/Kconfig b/drivers/gpu/drm/armada/Kconfig
index 40d371521fe1..50ae88ad4d76 100644
--- a/drivers/gpu/drm/armada/Kconfig
+++ b/drivers/gpu/drm/armada/Kconfig
@@ -5,6 +5,7 @@ config DRM_ARMADA
 	select FB_CFB_COPYAREA
 	select FB_CFB_IMAGEBLIT
 	select DRM_KMS_HELPER
+	select DRM_KMS_FB_HELPER
 	help
 	  Support the "LCD" controllers found on the Marvell Armada 510
 	  devices.  There are two controllers on the device, each controller
diff --git a/drivers/gpu/drm/armada/armada_drv.c b/drivers/gpu/drm/armada/armada_drv.c
index 62d0ff3efddf..acf3a36c9ebc 100644
--- a/drivers/gpu/drm/armada/armada_drv.c
+++ b/drivers/gpu/drm/armada/armada_drv.c
@@ -128,6 +128,7 @@ static int armada_drm_load(struct drm_device *dev, unsigned long flags)
 		return -ENOMEM;
 	}
 
+	platform_set_drvdata(dev->platformdev, dev);
 	dev->dev_private = priv;
 
 	/* Get the implementation specific driver data. */
@@ -381,7 +382,7 @@ static int armada_drm_probe(struct platform_device *pdev)
 
 static int armada_drm_remove(struct platform_device *pdev)
 {
-	drm_platform_exit(&armada_drm_driver, pdev);
+	drm_put_dev(platform_get_drvdata(pdev));
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/ast/ast_fb.c b/drivers/gpu/drm/ast/ast_fb.c
index 7b33e14e44aa..3f65dd6676b2 100644
--- a/drivers/gpu/drm/ast/ast_fb.c
+++ b/drivers/gpu/drm/ast/ast_fb.c
@@ -65,7 +65,7 @@ static void ast_dirty_update(struct ast_fbdev *afbdev,
 	 * then the BO is being moved and we should
 	 * store up the damage until later.
 	 */
-	if (!in_interrupt())
+	if (!drm_can_sleep())
 		ret = ast_bo_reserve(bo, true);
 	if (ret) {
 		if (ret != -EBUSY)
diff --git a/drivers/gpu/drm/ast/ast_main.c b/drivers/gpu/drm/ast/ast_main.c
index af0b868a9dfd..50535fd5a88d 100644
--- a/drivers/gpu/drm/ast/ast_main.c
+++ b/drivers/gpu/drm/ast/ast_main.c
@@ -189,53 +189,6 @@ static int ast_get_dram_info(struct drm_device *dev)
 	return 0;
 }
 
-uint32_t ast_get_max_dclk(struct drm_device *dev, int bpp)
-{
-	struct ast_private *ast = dev->dev_private;
-	uint32_t dclk, jreg;
-	uint32_t dram_bus_width, mclk, dram_bandwidth, actual_dram_bandwidth, dram_efficency = 500;
-
-	dram_bus_width = ast->dram_bus_width;
-	mclk = ast->mclk;
-
-	if (ast->chip == AST2100 ||
-	    ast->chip == AST1100 ||
-	    ast->chip == AST2200 ||
-	    ast->chip == AST2150 ||
-	    ast->dram_bus_width == 16)
-		dram_efficency = 600;
-	else if (ast->chip == AST2300)
-		dram_efficency = 400;
-
-	dram_bandwidth = mclk * dram_bus_width * 2 / 8;
-	actual_dram_bandwidth = dram_bandwidth * dram_efficency / 1000;
-
-	if (ast->chip == AST1180)
-		dclk = actual_dram_bandwidth / ((bpp + 1) / 8);
-	else {
-		jreg = ast_get_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xd0, 0xff);
-		if ((jreg & 0x08) && (ast->chip == AST2000))
-			dclk = actual_dram_bandwidth / ((bpp + 1 + 16) / 8);
-		else if ((jreg & 0x08) && (bpp == 8))
-			dclk = actual_dram_bandwidth / ((bpp + 1 + 24) / 8);
-		else
-			dclk = actual_dram_bandwidth / ((bpp + 1) / 8);
-	}
-
-	if (ast->chip == AST2100 ||
-	    ast->chip == AST2200 ||
-	    ast->chip == AST2300 ||
-	    ast->chip == AST1180) {
-		if (dclk > 200)
-			dclk = 200;
-	} else {
-		if (dclk > 165)
-			dclk = 165;
-	}
-
-	return dclk;
-}
-
 static void ast_user_framebuffer_destroy(struct drm_framebuffer *fb)
 {
 	struct ast_framebuffer *ast_fb = to_ast_framebuffer(fb);
@@ -449,7 +402,7 @@ int ast_dumb_create(struct drm_file *file,
 	return 0;
 }
 
-void ast_bo_unref(struct ast_bo **bo)
+static void ast_bo_unref(struct ast_bo **bo)
 {
 	struct ttm_buffer_object *tbo;
 
diff --git a/drivers/gpu/drm/ast/ast_mode.c b/drivers/gpu/drm/ast/ast_mode.c
index 7fc9f7272b56..cca063b11083 100644
--- a/drivers/gpu/drm/ast/ast_mode.c
+++ b/drivers/gpu/drm/ast/ast_mode.c
@@ -404,7 +404,7 @@ static void ast_set_ext_reg(struct drm_crtc *crtc, struct drm_display_mode *mode
 	}
 }
 
-void ast_set_sync_reg(struct drm_device *dev, struct drm_display_mode *mode,
+static void ast_set_sync_reg(struct drm_device *dev, struct drm_display_mode *mode,
 		      struct ast_vbios_mode_info *vbios_mode)
 {
 	struct ast_private *ast = dev->dev_private;
@@ -415,7 +415,7 @@ void ast_set_sync_reg(struct drm_device *dev, struct drm_display_mode *mode,
 	ast_io_write8(ast, AST_IO_MISC_PORT_WRITE, jreg);
 }
 
-bool ast_set_dac_reg(struct drm_crtc *crtc, struct drm_display_mode *mode,
+static bool ast_set_dac_reg(struct drm_crtc *crtc, struct drm_display_mode *mode,
 		     struct ast_vbios_mode_info *vbios_mode)
 {
 	switch (crtc->fb->bits_per_pixel) {
@@ -427,7 +427,7 @@ bool ast_set_dac_reg(struct drm_crtc *crtc, struct drm_display_mode *mode,
 	return true;
 }
 
-void ast_set_start_address_crt1(struct drm_crtc *crtc, unsigned offset)
+static void ast_set_start_address_crt1(struct drm_crtc *crtc, unsigned offset)
 {
 	struct ast_private *ast = crtc->dev->dev_private;
 	u32 addr;
@@ -623,7 +623,7 @@ static const struct drm_crtc_funcs ast_crtc_funcs = {
 	.destroy = ast_crtc_destroy,
 };
 
-int ast_crtc_init(struct drm_device *dev)
+static int ast_crtc_init(struct drm_device *dev)
 {
 	struct ast_crtc *crtc;
 	int i;
@@ -710,7 +710,7 @@ static const struct drm_encoder_helper_funcs ast_enc_helper_funcs = {
 	.mode_set = ast_encoder_mode_set,
 };
 
-int ast_encoder_init(struct drm_device *dev)
+static int ast_encoder_init(struct drm_device *dev)
 {
 	struct ast_encoder *ast_encoder;
 
@@ -777,7 +777,7 @@ static const struct drm_connector_funcs ast_connector_funcs = {
 	.destroy = ast_connector_destroy,
 };
 
-int ast_connector_init(struct drm_device *dev)
+static int ast_connector_init(struct drm_device *dev)
 {
 	struct ast_connector *ast_connector;
 	struct drm_connector *connector;
@@ -810,7 +810,7 @@ int ast_connector_init(struct drm_device *dev)
 }
 
 /* allocate cursor cache and pin at start of VRAM */
-int ast_cursor_init(struct drm_device *dev)
+static int ast_cursor_init(struct drm_device *dev)
 {
 	struct ast_private *ast = dev->dev_private;
 	int size;
@@ -847,7 +847,7 @@ fail:
 	return ret;
 }
 
-void ast_cursor_fini(struct drm_device *dev)
+static void ast_cursor_fini(struct drm_device *dev)
 {
 	struct ast_private *ast = dev->dev_private;
 	ttm_bo_kunmap(&ast->cache_kmap);
@@ -965,7 +965,7 @@ static void ast_i2c_destroy(struct ast_i2c_chan *i2c)
 	kfree(i2c);
 }
 
-void ast_show_cursor(struct drm_crtc *crtc)
+static void ast_show_cursor(struct drm_crtc *crtc)
 {
 	struct ast_private *ast = crtc->dev->dev_private;
 	u8 jreg;
@@ -976,7 +976,7 @@ void ast_show_cursor(struct drm_crtc *crtc)
 	ast_set_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xcb, 0xfc, jreg);
 }
 
-void ast_hide_cursor(struct drm_crtc *crtc)
+static void ast_hide_cursor(struct drm_crtc *crtc)
 {
 	struct ast_private *ast = crtc->dev->dev_private;
 	ast_set_index_reg_mask(ast, AST_IO_CRTC_PORT, 0xcb, 0xfc, 0x00);
diff --git a/drivers/gpu/drm/ast/ast_ttm.c b/drivers/gpu/drm/ast/ast_ttm.c
index 32aecb34dbce..4ea9b17ac17a 100644
--- a/drivers/gpu/drm/ast/ast_ttm.c
+++ b/drivers/gpu/drm/ast/ast_ttm.c
@@ -80,7 +80,7 @@ static int ast_ttm_global_init(struct ast_private *ast)
 	return 0;
 }
 
-void
+static void
 ast_ttm_global_release(struct ast_private *ast)
 {
 	if (ast->ttm.mem_global_ref.release == NULL)
@@ -102,7 +102,7 @@ static void ast_bo_ttm_destroy(struct ttm_buffer_object *tbo)
 	kfree(bo);
 }
 
-bool ast_ttm_bo_is_ast_bo(struct ttm_buffer_object *bo)
+static bool ast_ttm_bo_is_ast_bo(struct ttm_buffer_object *bo)
 {
 	if (bo->destroy == &ast_bo_ttm_destroy)
 		return true;
@@ -208,7 +208,7 @@ static struct ttm_backend_func ast_tt_backend_func = {
 };
 
 
-struct ttm_tt *ast_ttm_tt_create(struct ttm_bo_device *bdev,
+static struct ttm_tt *ast_ttm_tt_create(struct ttm_bo_device *bdev,
 				 unsigned long size, uint32_t page_flags,
 				 struct page *dummy_read_page)
 {
diff --git a/drivers/gpu/drm/bochs/Kconfig b/drivers/gpu/drm/bochs/Kconfig
new file mode 100644
index 000000000000..c8fcf12019f0
--- /dev/null
+++ b/drivers/gpu/drm/bochs/Kconfig
@@ -0,0 +1,11 @@
+config DRM_BOCHS
+	tristate "DRM Support for bochs dispi vga interface (qemu stdvga)"
+	depends on DRM && PCI
+	select DRM_KMS_HELPER
+	select FB_SYS_FILLRECT
+	select FB_SYS_COPYAREA
+	select FB_SYS_IMAGEBLIT
+	select DRM_TTM
+	help
+	  Choose this option for qemu.
+	  If M is selected the module will be called bochs-drm.
diff --git a/drivers/gpu/drm/bochs/Makefile b/drivers/gpu/drm/bochs/Makefile
new file mode 100644
index 000000000000..844a55614920
--- /dev/null
+++ b/drivers/gpu/drm/bochs/Makefile
@@ -0,0 +1,4 @@
+ccflags-y := -Iinclude/drm
+bochs-drm-y := bochs_drv.o bochs_mm.o bochs_kms.o bochs_fbdev.o bochs_hw.o
+
+obj-$(CONFIG_DRM_BOCHS)	+= bochs-drm.o
diff --git a/drivers/gpu/drm/bochs/bochs.h b/drivers/gpu/drm/bochs/bochs.h
new file mode 100644
index 000000000000..741965c001a6
--- /dev/null
+++ b/drivers/gpu/drm/bochs/bochs.h
@@ -0,0 +1,164 @@
+#include <linux/io.h>
+#include <linux/fb.h>
+
+#include <drm/drmP.h>
+#include <drm/drm_crtc.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_fb_helper.h>
+
+#include <ttm/ttm_bo_driver.h>
+#include <ttm/ttm_page_alloc.h>
+
+/* ---------------------------------------------------------------------- */
+
+#define VBE_DISPI_IOPORT_INDEX           0x01CE
+#define VBE_DISPI_IOPORT_DATA            0x01CF
+
+#define VBE_DISPI_INDEX_ID               0x0
+#define VBE_DISPI_INDEX_XRES             0x1
+#define VBE_DISPI_INDEX_YRES             0x2
+#define VBE_DISPI_INDEX_BPP              0x3
+#define VBE_DISPI_INDEX_ENABLE           0x4
+#define VBE_DISPI_INDEX_BANK             0x5
+#define VBE_DISPI_INDEX_VIRT_WIDTH       0x6
+#define VBE_DISPI_INDEX_VIRT_HEIGHT      0x7
+#define VBE_DISPI_INDEX_X_OFFSET         0x8
+#define VBE_DISPI_INDEX_Y_OFFSET         0x9
+#define VBE_DISPI_INDEX_VIDEO_MEMORY_64K 0xa
+
+#define VBE_DISPI_ID0                    0xB0C0
+#define VBE_DISPI_ID1                    0xB0C1
+#define VBE_DISPI_ID2                    0xB0C2
+#define VBE_DISPI_ID3                    0xB0C3
+#define VBE_DISPI_ID4                    0xB0C4
+#define VBE_DISPI_ID5                    0xB0C5
+
+#define VBE_DISPI_DISABLED               0x00
+#define VBE_DISPI_ENABLED                0x01
+#define VBE_DISPI_GETCAPS                0x02
+#define VBE_DISPI_8BIT_DAC               0x20
+#define VBE_DISPI_LFB_ENABLED            0x40
+#define VBE_DISPI_NOCLEARMEM             0x80
+
+/* ---------------------------------------------------------------------- */
+
+enum bochs_types {
+	BOCHS_QEMU_STDVGA,
+	BOCHS_UNKNOWN,
+};
+
+struct bochs_framebuffer {
+	struct drm_framebuffer base;
+	struct drm_gem_object *obj;
+};
+
+struct bochs_device {
+	/* hw */
+	void __iomem   *mmio;
+	int            ioports;
+	void __iomem   *fb_map;
+	unsigned long  fb_base;
+	unsigned long  fb_size;
+
+	/* mode */
+	u16 xres;
+	u16 yres;
+	u16 yres_virtual;
+	u32 stride;
+	u32 bpp;
+
+	/* drm */
+	struct drm_device  *dev;
+	struct drm_crtc crtc;
+	struct drm_encoder encoder;
+	struct drm_connector connector;
+	bool mode_config_initialized;
+
+	/* ttm */
+	struct {
+		struct drm_global_reference mem_global_ref;
+		struct ttm_bo_global_ref bo_global_ref;
+		struct ttm_bo_device bdev;
+		bool initialized;
+	} ttm;
+
+	/* fbdev */
+	struct {
+		struct bochs_framebuffer gfb;
+		struct drm_fb_helper helper;
+		int size;
+		int x1, y1, x2, y2; /* dirty rect */
+		spinlock_t dirty_lock;
+		bool initialized;
+	} fb;
+};
+
+#define to_bochs_framebuffer(x) container_of(x, struct bochs_framebuffer, base)
+
+struct bochs_bo {
+	struct ttm_buffer_object bo;
+	struct ttm_placement placement;
+	struct ttm_bo_kmap_obj kmap;
+	struct drm_gem_object gem;
+	u32 placements[3];
+	int pin_count;
+};
+
+static inline struct bochs_bo *bochs_bo(struct ttm_buffer_object *bo)
+{
+	return container_of(bo, struct bochs_bo, bo);
+}
+
+static inline struct bochs_bo *gem_to_bochs_bo(struct drm_gem_object *gem)
+{
+	return container_of(gem, struct bochs_bo, gem);
+}
+
+#define DRM_FILE_PAGE_OFFSET (0x100000000ULL >> PAGE_SHIFT)
+
+static inline u64 bochs_bo_mmap_offset(struct bochs_bo *bo)
+{
+	return drm_vma_node_offset_addr(&bo->bo.vma_node);
+}
+
+/* ---------------------------------------------------------------------- */
+
+/* bochs_hw.c */
+int bochs_hw_init(struct drm_device *dev, uint32_t flags);
+void bochs_hw_fini(struct drm_device *dev);
+
+void bochs_hw_setmode(struct bochs_device *bochs,
+		      struct drm_display_mode *mode);
+void bochs_hw_setbase(struct bochs_device *bochs,
+		      int x, int y, u64 addr);
+
+/* bochs_mm.c */
+int bochs_mm_init(struct bochs_device *bochs);
+void bochs_mm_fini(struct bochs_device *bochs);
+int bochs_mmap(struct file *filp, struct vm_area_struct *vma);
+
+int bochs_gem_create(struct drm_device *dev, u32 size, bool iskernel,
+		     struct drm_gem_object **obj);
+int bochs_gem_init_object(struct drm_gem_object *obj);
+void bochs_gem_free_object(struct drm_gem_object *obj);
+int bochs_dumb_create(struct drm_file *file, struct drm_device *dev,
+		      struct drm_mode_create_dumb *args);
+int bochs_dumb_mmap_offset(struct drm_file *file, struct drm_device *dev,
+			   uint32_t handle, uint64_t *offset);
+
+int bochs_framebuffer_init(struct drm_device *dev,
+			   struct bochs_framebuffer *gfb,
+			   struct drm_mode_fb_cmd2 *mode_cmd,
+			   struct drm_gem_object *obj);
+int bochs_bo_pin(struct bochs_bo *bo, u32 pl_flag, u64 *gpu_addr);
+int bochs_bo_unpin(struct bochs_bo *bo);
+
+extern const struct drm_mode_config_funcs bochs_mode_funcs;
+
+/* bochs_kms.c */
+int bochs_kms_init(struct bochs_device *bochs);
+void bochs_kms_fini(struct bochs_device *bochs);
+
+/* bochs_fbdev.c */
+int bochs_fbdev_init(struct bochs_device *bochs);
+void bochs_fbdev_fini(struct bochs_device *bochs);
diff --git a/drivers/gpu/drm/bochs/bochs_drv.c b/drivers/gpu/drm/bochs/bochs_drv.c
new file mode 100644
index 000000000000..395bba261c9a
--- /dev/null
+++ b/drivers/gpu/drm/bochs/bochs_drv.c
@@ -0,0 +1,178 @@
+/*
+ * 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.
+ */
+
+#include <linux/mm.h>
+#include <linux/module.h>
+#include <linux/slab.h>
+
+#include "bochs.h"
+
+static bool enable_fbdev = true;
+module_param_named(fbdev, enable_fbdev, bool, 0444);
+MODULE_PARM_DESC(fbdev, "register fbdev device");
+
+/* ---------------------------------------------------------------------- */
+/* drm interface                                                          */
+
+static int bochs_unload(struct drm_device *dev)
+{
+	struct bochs_device *bochs = dev->dev_private;
+
+	bochs_fbdev_fini(bochs);
+	bochs_kms_fini(bochs);
+	bochs_mm_fini(bochs);
+	bochs_hw_fini(dev);
+	kfree(bochs);
+	dev->dev_private = NULL;
+	return 0;
+}
+
+static int bochs_load(struct drm_device *dev, unsigned long flags)
+{
+	struct bochs_device *bochs;
+	int ret;
+
+	bochs = kzalloc(sizeof(*bochs), GFP_KERNEL);
+	if (bochs == NULL)
+		return -ENOMEM;
+	dev->dev_private = bochs;
+	bochs->dev = dev;
+
+	ret = bochs_hw_init(dev, flags);
+	if (ret)
+		goto err;
+
+	ret = bochs_mm_init(bochs);
+	if (ret)
+		goto err;
+
+	ret = bochs_kms_init(bochs);
+	if (ret)
+		goto err;
+
+	if (enable_fbdev)
+		bochs_fbdev_init(bochs);
+
+	return 0;
+
+err:
+	bochs_unload(dev);
+	return ret;
+}
+
+static const struct file_operations bochs_fops = {
+	.owner		= THIS_MODULE,
+	.open		= drm_open,
+	.release	= drm_release,
+	.unlocked_ioctl	= drm_ioctl,
+#ifdef CONFIG_COMPAT
+	.compat_ioctl	= drm_compat_ioctl,
+#endif
+	.poll		= drm_poll,
+	.read		= drm_read,
+	.llseek		= no_llseek,
+	.mmap           = bochs_mmap,
+};
+
+static struct drm_driver bochs_driver = {
+	.driver_features	= DRIVER_GEM | DRIVER_MODESET,
+	.load			= bochs_load,
+	.unload			= bochs_unload,
+	.fops			= &bochs_fops,
+	.name			= "bochs-drm",
+	.desc			= "bochs dispi vga interface (qemu stdvga)",
+	.date			= "20130925",
+	.major			= 1,
+	.minor			= 0,
+	.gem_free_object        = bochs_gem_free_object,
+	.dumb_create            = bochs_dumb_create,
+	.dumb_map_offset        = bochs_dumb_mmap_offset,
+	.dumb_destroy           = drm_gem_dumb_destroy,
+};
+
+/* ---------------------------------------------------------------------- */
+/* pci interface                                                          */
+
+static int bochs_kick_out_firmware_fb(struct pci_dev *pdev)
+{
+	struct apertures_struct *ap;
+
+	ap = alloc_apertures(1);
+	if (!ap)
+		return -ENOMEM;
+
+	ap->ranges[0].base = pci_resource_start(pdev, 0);
+	ap->ranges[0].size = pci_resource_len(pdev, 0);
+	remove_conflicting_framebuffers(ap, "bochsdrmfb", false);
+	kfree(ap);
+
+	return 0;
+}
+
+static int bochs_pci_probe(struct pci_dev *pdev,
+			   const struct pci_device_id *ent)
+{
+	int ret;
+
+	ret = bochs_kick_out_firmware_fb(pdev);
+	if (ret)
+		return ret;
+
+	return drm_get_pci_dev(pdev, ent, &bochs_driver);
+}
+
+static void bochs_pci_remove(struct pci_dev *pdev)
+{
+	struct drm_device *dev = pci_get_drvdata(pdev);
+
+	drm_put_dev(dev);
+}
+
+static DEFINE_PCI_DEVICE_TABLE(bochs_pci_tbl) = {
+	{
+		.vendor      = 0x1234,
+		.device      = 0x1111,
+		.subvendor   = 0x1af4,
+		.subdevice   = 0x1100,
+		.driver_data = BOCHS_QEMU_STDVGA,
+	},
+	{
+		.vendor      = 0x1234,
+		.device      = 0x1111,
+		.subvendor   = PCI_ANY_ID,
+		.subdevice   = PCI_ANY_ID,
+		.driver_data = BOCHS_UNKNOWN,
+	},
+	{ /* end of list */ }
+};
+
+static struct pci_driver bochs_pci_driver = {
+	.name =		"bochs-drm",
+	.id_table =	bochs_pci_tbl,
+	.probe =	bochs_pci_probe,
+	.remove =	bochs_pci_remove,
+};
+
+/* ---------------------------------------------------------------------- */
+/* module init/exit                                                       */
+
+static int __init bochs_init(void)
+{
+	return drm_pci_init(&bochs_driver, &bochs_pci_driver);
+}
+
+static void __exit bochs_exit(void)
+{
+	drm_pci_exit(&bochs_driver, &bochs_pci_driver);
+}
+
+module_init(bochs_init);
+module_exit(bochs_exit);
+
+MODULE_DEVICE_TABLE(pci, bochs_pci_tbl);
+MODULE_AUTHOR("Gerd Hoffmann <kraxel@redhat.com>");
+MODULE_LICENSE("GPL");
diff --git a/drivers/gpu/drm/bochs/bochs_fbdev.c b/drivers/gpu/drm/bochs/bochs_fbdev.c
new file mode 100644
index 000000000000..4da5206b7cc9
--- /dev/null
+++ b/drivers/gpu/drm/bochs/bochs_fbdev.c
@@ -0,0 +1,215 @@
+/*
+ * 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.
+ */
+
+#include "bochs.h"
+
+/* ---------------------------------------------------------------------- */
+
+static struct fb_ops bochsfb_ops = {
+	.owner = THIS_MODULE,
+	.fb_check_var = drm_fb_helper_check_var,
+	.fb_set_par = drm_fb_helper_set_par,
+	.fb_fillrect = sys_fillrect,
+	.fb_copyarea = sys_copyarea,
+	.fb_imageblit = sys_imageblit,
+	.fb_pan_display = drm_fb_helper_pan_display,
+	.fb_blank = drm_fb_helper_blank,
+	.fb_setcmap = drm_fb_helper_setcmap,
+};
+
+static int bochsfb_create_object(struct bochs_device *bochs,
+				 struct drm_mode_fb_cmd2 *mode_cmd,
+				 struct drm_gem_object **gobj_p)
+{
+	struct drm_device *dev = bochs->dev;
+	struct drm_gem_object *gobj;
+	u32 size;
+	int ret = 0;
+
+	size = mode_cmd->pitches[0] * mode_cmd->height;
+	ret = bochs_gem_create(dev, size, true, &gobj);
+	if (ret)
+		return ret;
+
+	*gobj_p = gobj;
+	return ret;
+}
+
+static int bochsfb_create(struct drm_fb_helper *helper,
+			  struct drm_fb_helper_surface_size *sizes)
+{
+	struct bochs_device *bochs =
+		container_of(helper, struct bochs_device, fb.helper);
+	struct drm_device *dev = bochs->dev;
+	struct fb_info *info;
+	struct drm_framebuffer *fb;
+	struct drm_mode_fb_cmd2 mode_cmd;
+	struct device *device = &dev->pdev->dev;
+	struct drm_gem_object *gobj = NULL;
+	struct bochs_bo *bo = NULL;
+	int size, ret;
+
+	if (sizes->surface_bpp != 32)
+		return -EINVAL;
+
+	mode_cmd.width = sizes->surface_width;
+	mode_cmd.height = sizes->surface_height;
+	mode_cmd.pitches[0] = mode_cmd.width * ((sizes->surface_bpp + 7) / 8);
+	mode_cmd.pixel_format = drm_mode_legacy_fb_format(sizes->surface_bpp,
+							  sizes->surface_depth);
+	size = mode_cmd.pitches[0] * mode_cmd.height;
+
+	/* alloc, pin & map bo */
+	ret = bochsfb_create_object(bochs, &mode_cmd, &gobj);
+	if (ret) {
+		DRM_ERROR("failed to create fbcon backing object %d\n", ret);
+		return ret;
+	}
+
+	bo = gem_to_bochs_bo(gobj);
+
+	ret = ttm_bo_reserve(&bo->bo, true, false, false, 0);
+	if (ret)
+		return ret;
+
+	ret = bochs_bo_pin(bo, TTM_PL_FLAG_VRAM, NULL);
+	if (ret) {
+		DRM_ERROR("failed to pin fbcon\n");
+		ttm_bo_unreserve(&bo->bo);
+		return ret;
+	}
+
+	ret = ttm_bo_kmap(&bo->bo, 0, bo->bo.num_pages,
+			  &bo->kmap);
+	if (ret) {
+		DRM_ERROR("failed to kmap fbcon\n");
+		ttm_bo_unreserve(&bo->bo);
+		return ret;
+	}
+
+	ttm_bo_unreserve(&bo->bo);
+
+	/* init fb device */
+	info = framebuffer_alloc(0, device);
+	if (info == NULL)
+		return -ENOMEM;
+
+	info->par = &bochs->fb.helper;
+
+	ret = bochs_framebuffer_init(bochs->dev, &bochs->fb.gfb, &mode_cmd, gobj);
+	if (ret)
+		return ret;
+
+	bochs->fb.size = size;
+
+	/* setup helper */
+	fb = &bochs->fb.gfb.base;
+	bochs->fb.helper.fb = fb;
+	bochs->fb.helper.fbdev = info;
+
+	strcpy(info->fix.id, "bochsdrmfb");
+
+	info->flags = FBINFO_DEFAULT;
+	info->fbops = &bochsfb_ops;
+
+	drm_fb_helper_fill_fix(info, fb->pitches[0], fb->depth);
+	drm_fb_helper_fill_var(info, &bochs->fb.helper, sizes->fb_width,
+			       sizes->fb_height);
+
+	info->screen_base = bo->kmap.virtual;
+	info->screen_size = size;
+
+#if 0
+	/* FIXME: get this right for mmap(/dev/fb0) */
+	info->fix.smem_start = bochs_bo_mmap_offset(bo);
+	info->fix.smem_len = size;
+#endif
+
+	ret = fb_alloc_cmap(&info->cmap, 256, 0);
+	if (ret) {
+		DRM_ERROR("%s: can't allocate color map\n", info->fix.id);
+		return -ENOMEM;
+	}
+
+	return 0;
+}
+
+static int bochs_fbdev_destroy(struct bochs_device *bochs)
+{
+	struct bochs_framebuffer *gfb = &bochs->fb.gfb;
+	struct fb_info *info;
+
+	DRM_DEBUG_DRIVER("\n");
+
+	if (bochs->fb.helper.fbdev) {
+		info = bochs->fb.helper.fbdev;
+
+		unregister_framebuffer(info);
+		if (info->cmap.len)
+			fb_dealloc_cmap(&info->cmap);
+		framebuffer_release(info);
+	}
+
+	if (gfb->obj) {
+		drm_gem_object_unreference_unlocked(gfb->obj);
+		gfb->obj = NULL;
+	}
+
+	drm_fb_helper_fini(&bochs->fb.helper);
+	drm_framebuffer_unregister_private(&gfb->base);
+	drm_framebuffer_cleanup(&gfb->base);
+
+	return 0;
+}
+
+void bochs_fb_gamma_set(struct drm_crtc *crtc, u16 red, u16 green,
+			u16 blue, int regno)
+{
+}
+
+void bochs_fb_gamma_get(struct drm_crtc *crtc, u16 *red, u16 *green,
+			u16 *blue, int regno)
+{
+	*red   = regno;
+	*green = regno;
+	*blue  = regno;
+}
+
+static struct drm_fb_helper_funcs bochs_fb_helper_funcs = {
+	.gamma_set = bochs_fb_gamma_set,
+	.gamma_get = bochs_fb_gamma_get,
+	.fb_probe = bochsfb_create,
+};
+
+int bochs_fbdev_init(struct bochs_device *bochs)
+{
+	int ret;
+
+	bochs->fb.helper.funcs = &bochs_fb_helper_funcs;
+	spin_lock_init(&bochs->fb.dirty_lock);
+
+	ret = drm_fb_helper_init(bochs->dev, &bochs->fb.helper,
+				 1, 1);
+	if (ret)
+		return ret;
+
+	drm_fb_helper_single_add_all_connectors(&bochs->fb.helper);
+	drm_helper_disable_unused_functions(bochs->dev);
+	drm_fb_helper_initial_config(&bochs->fb.helper, 32);
+
+	bochs->fb.initialized = true;
+	return 0;
+}
+
+void bochs_fbdev_fini(struct bochs_device *bochs)
+{
+	if (!bochs->fb.initialized)
+		return;
+
+	bochs_fbdev_destroy(bochs);
+	bochs->fb.initialized = false;
+}
diff --git a/drivers/gpu/drm/bochs/bochs_hw.c b/drivers/gpu/drm/bochs/bochs_hw.c
new file mode 100644
index 000000000000..dbe619e6aab4
--- /dev/null
+++ b/drivers/gpu/drm/bochs/bochs_hw.c
@@ -0,0 +1,177 @@
+/*
+ * 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.
+ */
+
+#include "bochs.h"
+
+/* ---------------------------------------------------------------------- */
+
+static void bochs_vga_writeb(struct bochs_device *bochs, u16 ioport, u8 val)
+{
+	if (WARN_ON(ioport < 0x3c0 || ioport > 0x3df))
+		return;
+
+	if (bochs->mmio) {
+		int offset = ioport - 0x3c0 + 0x400;
+		writeb(val, bochs->mmio + offset);
+	} else {
+		outb(val, ioport);
+	}
+}
+
+static u16 bochs_dispi_read(struct bochs_device *bochs, u16 reg)
+{
+	u16 ret = 0;
+
+	if (bochs->mmio) {
+		int offset = 0x500 + (reg << 1);
+		ret = readw(bochs->mmio + offset);
+	} else {
+		outw(reg, VBE_DISPI_IOPORT_INDEX);
+		ret = inw(VBE_DISPI_IOPORT_DATA);
+	}
+	return ret;
+}
+
+static void bochs_dispi_write(struct bochs_device *bochs, u16 reg, u16 val)
+{
+	if (bochs->mmio) {
+		int offset = 0x500 + (reg << 1);
+		writew(val, bochs->mmio + offset);
+	} else {
+		outw(reg, VBE_DISPI_IOPORT_INDEX);
+		outw(val, VBE_DISPI_IOPORT_DATA);
+	}
+}
+
+int bochs_hw_init(struct drm_device *dev, uint32_t flags)
+{
+	struct bochs_device *bochs = dev->dev_private;
+	struct pci_dev *pdev = dev->pdev;
+	unsigned long addr, size, mem, ioaddr, iosize;
+	u16 id;
+
+	if (/* (ent->driver_data == BOCHS_QEMU_STDVGA) && */
+	    (pdev->resource[2].flags & IORESOURCE_MEM)) {
+		/* mmio bar with vga and bochs registers present */
+		if (pci_request_region(pdev, 2, "bochs-drm") != 0) {
+			DRM_ERROR("Cannot request mmio region\n");
+			return -EBUSY;
+		}
+		ioaddr = pci_resource_start(pdev, 2);
+		iosize = pci_resource_len(pdev, 2);
+		bochs->mmio = ioremap(ioaddr, iosize);
+		if (bochs->mmio == NULL) {
+			DRM_ERROR("Cannot map mmio region\n");
+			return -ENOMEM;
+		}
+	} else {
+		ioaddr = VBE_DISPI_IOPORT_INDEX;
+		iosize = 2;
+		if (!request_region(ioaddr, iosize, "bochs-drm")) {
+			DRM_ERROR("Cannot request ioports\n");
+			return -EBUSY;
+		}
+		bochs->ioports = 1;
+	}
+
+	id = bochs_dispi_read(bochs, VBE_DISPI_INDEX_ID);
+	mem = bochs_dispi_read(bochs, VBE_DISPI_INDEX_VIDEO_MEMORY_64K)
+		* 64 * 1024;
+	if ((id & 0xfff0) != VBE_DISPI_ID0) {
+		DRM_ERROR("ID mismatch\n");
+		return -ENODEV;
+	}
+
+	if ((pdev->resource[0].flags & IORESOURCE_MEM) == 0)
+		return -ENODEV;
+	addr = pci_resource_start(pdev, 0);
+	size = pci_resource_len(pdev, 0);
+	if (addr == 0)
+		return -ENODEV;
+	if (size != mem) {
+		DRM_ERROR("Size mismatch: pci=%ld, bochs=%ld\n",
+			size, mem);
+		size = min(size, mem);
+	}
+
+	if (pci_request_region(pdev, 0, "bochs-drm") != 0) {
+		DRM_ERROR("Cannot request framebuffer\n");
+		return -EBUSY;
+	}
+
+	bochs->fb_map = ioremap(addr, size);
+	if (bochs->fb_map == NULL) {
+		DRM_ERROR("Cannot map framebuffer\n");
+		return -ENOMEM;
+	}
+	bochs->fb_base = addr;
+	bochs->fb_size = size;
+
+	DRM_INFO("Found bochs VGA, ID 0x%x.\n", id);
+	DRM_INFO("Framebuffer size %ld kB @ 0x%lx, %s @ 0x%lx.\n",
+		 size / 1024, addr,
+		 bochs->ioports ? "ioports" : "mmio",
+		 ioaddr);
+	return 0;
+}
+
+void bochs_hw_fini(struct drm_device *dev)
+{
+	struct bochs_device *bochs = dev->dev_private;
+
+	if (bochs->mmio)
+		iounmap(bochs->mmio);
+	if (bochs->ioports)
+		release_region(VBE_DISPI_IOPORT_INDEX, 2);
+	if (bochs->fb_map)
+		iounmap(bochs->fb_map);
+	pci_release_regions(dev->pdev);
+}
+
+void bochs_hw_setmode(struct bochs_device *bochs,
+		      struct drm_display_mode *mode)
+{
+	bochs->xres = mode->hdisplay;
+	bochs->yres = mode->vdisplay;
+	bochs->bpp = 32;
+	bochs->stride = mode->hdisplay * (bochs->bpp / 8);
+	bochs->yres_virtual = bochs->fb_size / bochs->stride;
+
+	DRM_DEBUG_DRIVER("%dx%d @ %d bpp, vy %d\n",
+			 bochs->xres, bochs->yres, bochs->bpp,
+			 bochs->yres_virtual);
+
+	bochs_vga_writeb(bochs, 0x3c0, 0x20); /* unblank */
+
+	bochs_dispi_write(bochs, VBE_DISPI_INDEX_BPP,         bochs->bpp);
+	bochs_dispi_write(bochs, VBE_DISPI_INDEX_XRES,        bochs->xres);
+	bochs_dispi_write(bochs, VBE_DISPI_INDEX_YRES,        bochs->yres);
+	bochs_dispi_write(bochs, VBE_DISPI_INDEX_BANK,        0);
+	bochs_dispi_write(bochs, VBE_DISPI_INDEX_VIRT_WIDTH,  bochs->xres);
+	bochs_dispi_write(bochs, VBE_DISPI_INDEX_VIRT_HEIGHT,
+			  bochs->yres_virtual);
+	bochs_dispi_write(bochs, VBE_DISPI_INDEX_X_OFFSET,    0);
+	bochs_dispi_write(bochs, VBE_DISPI_INDEX_Y_OFFSET,    0);
+
+	bochs_dispi_write(bochs, VBE_DISPI_INDEX_ENABLE,
+			  VBE_DISPI_ENABLED | VBE_DISPI_LFB_ENABLED);
+}
+
+void bochs_hw_setbase(struct bochs_device *bochs,
+		      int x, int y, u64 addr)
+{
+	unsigned long offset = (unsigned long)addr +
+		y * bochs->stride +
+		x * (bochs->bpp / 8);
+	int vy = offset / bochs->stride;
+	int vx = (offset % bochs->stride) * 8 / bochs->bpp;
+
+	DRM_DEBUG_DRIVER("x %d, y %d, addr %llx -> offset %lx, vx %d, vy %d\n",
+			 x, y, addr, offset, vx, vy);
+	bochs_dispi_write(bochs, VBE_DISPI_INDEX_X_OFFSET, vx);
+	bochs_dispi_write(bochs, VBE_DISPI_INDEX_Y_OFFSET, vy);
+}
diff --git a/drivers/gpu/drm/bochs/bochs_kms.c b/drivers/gpu/drm/bochs/bochs_kms.c
new file mode 100644
index 000000000000..62ec7d4b3816
--- /dev/null
+++ b/drivers/gpu/drm/bochs/bochs_kms.c
@@ -0,0 +1,294 @@
+/*
+ * 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.
+ */
+
+#include "bochs.h"
+
+static int defx = 1024;
+static int defy = 768;
+
+module_param(defx, int, 0444);
+module_param(defy, int, 0444);
+MODULE_PARM_DESC(defx, "default x resolution");
+MODULE_PARM_DESC(defy, "default y resolution");
+
+/* ---------------------------------------------------------------------- */
+
+static void bochs_crtc_load_lut(struct drm_crtc *crtc)
+{
+}
+
+static void bochs_crtc_dpms(struct drm_crtc *crtc, int mode)
+{
+	switch (mode) {
+	case DRM_MODE_DPMS_ON:
+	case DRM_MODE_DPMS_STANDBY:
+	case DRM_MODE_DPMS_SUSPEND:
+	case DRM_MODE_DPMS_OFF:
+	default:
+		return;
+	}
+}
+
+static bool bochs_crtc_mode_fixup(struct drm_crtc *crtc,
+				  const struct drm_display_mode *mode,
+				  struct drm_display_mode *adjusted_mode)
+{
+	return true;
+}
+
+static int bochs_crtc_mode_set_base(struct drm_crtc *crtc, int x, int y,
+				    struct drm_framebuffer *old_fb)
+{
+	struct bochs_device *bochs =
+		container_of(crtc, struct bochs_device, crtc);
+	struct bochs_framebuffer *bochs_fb;
+	struct bochs_bo *bo;
+	u64 gpu_addr = 0;
+	int ret;
+
+	if (old_fb) {
+		bochs_fb = to_bochs_framebuffer(old_fb);
+		bo = gem_to_bochs_bo(bochs_fb->obj);
+		ret = ttm_bo_reserve(&bo->bo, true, false, false, 0);
+		if (ret) {
+			DRM_ERROR("failed to reserve old_fb bo\n");
+		} else {
+			bochs_bo_unpin(bo);
+			ttm_bo_unreserve(&bo->bo);
+		}
+	}
+
+	if (WARN_ON(crtc->fb == NULL))
+		return -EINVAL;
+
+	bochs_fb = to_bochs_framebuffer(crtc->fb);
+	bo = gem_to_bochs_bo(bochs_fb->obj);
+	ret = ttm_bo_reserve(&bo->bo, true, false, false, 0);
+	if (ret)
+		return ret;
+
+	ret = bochs_bo_pin(bo, TTM_PL_FLAG_VRAM, &gpu_addr);
+	if (ret) {
+		ttm_bo_unreserve(&bo->bo);
+		return ret;
+	}
+
+	ttm_bo_unreserve(&bo->bo);
+	bochs_hw_setbase(bochs, x, y, gpu_addr);
+	return 0;
+}
+
+static int bochs_crtc_mode_set(struct drm_crtc *crtc,
+			       struct drm_display_mode *mode,
+			       struct drm_display_mode *adjusted_mode,
+			       int x, int y, struct drm_framebuffer *old_fb)
+{
+	struct bochs_device *bochs =
+		container_of(crtc, struct bochs_device, crtc);
+
+	bochs_hw_setmode(bochs, mode);
+	bochs_crtc_mode_set_base(crtc, x, y, old_fb);
+	return 0;
+}
+
+static void bochs_crtc_prepare(struct drm_crtc *crtc)
+{
+}
+
+static void bochs_crtc_commit(struct drm_crtc *crtc)
+{
+}
+
+static void bochs_crtc_gamma_set(struct drm_crtc *crtc, u16 *red, u16 *green,
+				 u16 *blue, uint32_t start, uint32_t size)
+{
+}
+
+/* These provide the minimum set of functions required to handle a CRTC */
+static const struct drm_crtc_funcs bochs_crtc_funcs = {
+	.gamma_set = bochs_crtc_gamma_set,
+	.set_config = drm_crtc_helper_set_config,
+	.destroy = drm_crtc_cleanup,
+};
+
+static const struct drm_crtc_helper_funcs bochs_helper_funcs = {
+	.dpms = bochs_crtc_dpms,
+	.mode_fixup = bochs_crtc_mode_fixup,
+	.mode_set = bochs_crtc_mode_set,
+	.mode_set_base = bochs_crtc_mode_set_base,
+	.prepare = bochs_crtc_prepare,
+	.commit = bochs_crtc_commit,
+	.load_lut = bochs_crtc_load_lut,
+};
+
+static void bochs_crtc_init(struct drm_device *dev)
+{
+	struct bochs_device *bochs = dev->dev_private;
+	struct drm_crtc *crtc = &bochs->crtc;
+
+	drm_crtc_init(dev, crtc, &bochs_crtc_funcs);
+	drm_mode_crtc_set_gamma_size(crtc, 256);
+	drm_crtc_helper_add(crtc, &bochs_helper_funcs);
+}
+
+static bool bochs_encoder_mode_fixup(struct drm_encoder *encoder,
+				     const struct drm_display_mode *mode,
+				     struct drm_display_mode *adjusted_mode)
+{
+	return true;
+}
+
+static void bochs_encoder_mode_set(struct drm_encoder *encoder,
+				   struct drm_display_mode *mode,
+				   struct drm_display_mode *adjusted_mode)
+{
+}
+
+static void bochs_encoder_dpms(struct drm_encoder *encoder, int state)
+{
+}
+
+static void bochs_encoder_prepare(struct drm_encoder *encoder)
+{
+}
+
+static void bochs_encoder_commit(struct drm_encoder *encoder)
+{
+}
+
+static const struct drm_encoder_helper_funcs bochs_encoder_helper_funcs = {
+	.dpms = bochs_encoder_dpms,
+	.mode_fixup = bochs_encoder_mode_fixup,
+	.mode_set = bochs_encoder_mode_set,
+	.prepare = bochs_encoder_prepare,
+	.commit = bochs_encoder_commit,
+};
+
+static const struct drm_encoder_funcs bochs_encoder_encoder_funcs = {
+	.destroy = drm_encoder_cleanup,
+};
+
+static void bochs_encoder_init(struct drm_device *dev)
+{
+	struct bochs_device *bochs = dev->dev_private;
+	struct drm_encoder *encoder = &bochs->encoder;
+
+	encoder->possible_crtcs = 0x1;
+	drm_encoder_init(dev, encoder, &bochs_encoder_encoder_funcs,
+			 DRM_MODE_ENCODER_DAC);
+	drm_encoder_helper_add(encoder, &bochs_encoder_helper_funcs);
+}
+
+
+int bochs_connector_get_modes(struct drm_connector *connector)
+{
+	int count;
+
+	count = drm_add_modes_noedid(connector, 8192, 8192);
+	drm_set_preferred_mode(connector, defx, defy);
+	return count;
+}
+
+static int bochs_connector_mode_valid(struct drm_connector *connector,
+				      struct drm_display_mode *mode)
+{
+	struct bochs_device *bochs =
+		container_of(connector, struct bochs_device, connector);
+	unsigned long size = mode->hdisplay * mode->vdisplay * 4;
+
+	/*
+	 * Make sure we can fit two framebuffers into video memory.
+	 * This allows up to 1600x1200 with 16 MB (default size).
+	 * If you want more try this:
+	 *     'qemu -vga std -global VGA.vgamem_mb=32 $otherargs'
+	 */
+	if (size * 2 > bochs->fb_size)
+		return MODE_BAD;
+
+	return MODE_OK;
+}
+
+static struct drm_encoder *
+bochs_connector_best_encoder(struct drm_connector *connector)
+{
+	int enc_id = connector->encoder_ids[0];
+	struct drm_mode_object *obj;
+	struct drm_encoder *encoder;
+
+	/* pick the encoder ids */
+	if (enc_id) {
+		obj = drm_mode_object_find(connector->dev, enc_id,
+					   DRM_MODE_OBJECT_ENCODER);
+		if (!obj)
+			return NULL;
+		encoder = obj_to_encoder(obj);
+		return encoder;
+	}
+	return NULL;
+}
+
+static enum drm_connector_status bochs_connector_detect(struct drm_connector
+							*connector, bool force)
+{
+	return connector_status_connected;
+}
+
+struct drm_connector_helper_funcs bochs_connector_connector_helper_funcs = {
+	.get_modes = bochs_connector_get_modes,
+	.mode_valid = bochs_connector_mode_valid,
+	.best_encoder = bochs_connector_best_encoder,
+};
+
+struct drm_connector_funcs bochs_connector_connector_funcs = {
+	.dpms = drm_helper_connector_dpms,
+	.detect = bochs_connector_detect,
+	.fill_modes = drm_helper_probe_single_connector_modes,
+	.destroy = drm_connector_cleanup,
+};
+
+static void bochs_connector_init(struct drm_device *dev)
+{
+	struct bochs_device *bochs = dev->dev_private;
+	struct drm_connector *connector = &bochs->connector;
+
+	drm_connector_init(dev, connector, &bochs_connector_connector_funcs,
+			   DRM_MODE_CONNECTOR_VIRTUAL);
+	drm_connector_helper_add(connector,
+				 &bochs_connector_connector_helper_funcs);
+}
+
+
+int bochs_kms_init(struct bochs_device *bochs)
+{
+	drm_mode_config_init(bochs->dev);
+	bochs->mode_config_initialized = true;
+
+	bochs->dev->mode_config.max_width = 8192;
+	bochs->dev->mode_config.max_height = 8192;
+
+	bochs->dev->mode_config.fb_base = bochs->fb_base;
+	bochs->dev->mode_config.preferred_depth = 24;
+	bochs->dev->mode_config.prefer_shadow = 0;
+
+	bochs->dev->mode_config.funcs = (void *)&bochs_mode_funcs;
+
+	bochs_crtc_init(bochs->dev);
+	bochs_encoder_init(bochs->dev);
+	bochs_connector_init(bochs->dev);
+	drm_mode_connector_attach_encoder(&bochs->connector,
+					  &bochs->encoder);
+
+	return 0;
+}
+
+void bochs_kms_fini(struct bochs_device *bochs)
+{
+	if (bochs->mode_config_initialized) {
+		drm_mode_config_cleanup(bochs->dev);
+		bochs->mode_config_initialized = false;
+	}
+}
diff --git a/drivers/gpu/drm/bochs/bochs_mm.c b/drivers/gpu/drm/bochs/bochs_mm.c
new file mode 100644
index 000000000000..ce6858765b37
--- /dev/null
+++ b/drivers/gpu/drm/bochs/bochs_mm.c
@@ -0,0 +1,546 @@
+/*
+ * 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.
+ */
+
+#include "bochs.h"
+
+static void bochs_ttm_placement(struct bochs_bo *bo, int domain);
+
+/* ---------------------------------------------------------------------- */
+
+static inline struct bochs_device *bochs_bdev(struct ttm_bo_device *bd)
+{
+	return container_of(bd, struct bochs_device, ttm.bdev);
+}
+
+static int bochs_ttm_mem_global_init(struct drm_global_reference *ref)
+{
+	return ttm_mem_global_init(ref->object);
+}
+
+static void bochs_ttm_mem_global_release(struct drm_global_reference *ref)
+{
+	ttm_mem_global_release(ref->object);
+}
+
+static int bochs_ttm_global_init(struct bochs_device *bochs)
+{
+	struct drm_global_reference *global_ref;
+	int r;
+
+	global_ref = &bochs->ttm.mem_global_ref;
+	global_ref->global_type = DRM_GLOBAL_TTM_MEM;
+	global_ref->size = sizeof(struct ttm_mem_global);
+	global_ref->init = &bochs_ttm_mem_global_init;
+	global_ref->release = &bochs_ttm_mem_global_release;
+	r = drm_global_item_ref(global_ref);
+	if (r != 0) {
+		DRM_ERROR("Failed setting up TTM memory accounting "
+			  "subsystem.\n");
+		return r;
+	}
+
+	bochs->ttm.bo_global_ref.mem_glob =
+		bochs->ttm.mem_global_ref.object;
+	global_ref = &bochs->ttm.bo_global_ref.ref;
+	global_ref->global_type = DRM_GLOBAL_TTM_BO;
+	global_ref->size = sizeof(struct ttm_bo_global);
+	global_ref->init = &ttm_bo_global_init;
+	global_ref->release = &ttm_bo_global_release;
+	r = drm_global_item_ref(global_ref);
+	if (r != 0) {
+		DRM_ERROR("Failed setting up TTM BO subsystem.\n");
+		drm_global_item_unref(&bochs->ttm.mem_global_ref);
+		return r;
+	}
+
+	return 0;
+}
+
+static void bochs_ttm_global_release(struct bochs_device *bochs)
+{
+	if (bochs->ttm.mem_global_ref.release == NULL)
+		return;
+
+	drm_global_item_unref(&bochs->ttm.bo_global_ref.ref);
+	drm_global_item_unref(&bochs->ttm.mem_global_ref);
+	bochs->ttm.mem_global_ref.release = NULL;
+}
+
+
+static void bochs_bo_ttm_destroy(struct ttm_buffer_object *tbo)
+{
+	struct bochs_bo *bo;
+
+	bo = container_of(tbo, struct bochs_bo, bo);
+	drm_gem_object_release(&bo->gem);
+	kfree(bo);
+}
+
+static bool bochs_ttm_bo_is_bochs_bo(struct ttm_buffer_object *bo)
+{
+	if (bo->destroy == &bochs_bo_ttm_destroy)
+		return true;
+	return false;
+}
+
+static int bochs_bo_init_mem_type(struct ttm_bo_device *bdev, uint32_t type,
+				  struct ttm_mem_type_manager *man)
+{
+	switch (type) {
+	case TTM_PL_SYSTEM:
+		man->flags = TTM_MEMTYPE_FLAG_MAPPABLE;
+		man->available_caching = TTM_PL_MASK_CACHING;
+		man->default_caching = TTM_PL_FLAG_CACHED;
+		break;
+	case TTM_PL_VRAM:
+		man->func = &ttm_bo_manager_func;
+		man->flags = TTM_MEMTYPE_FLAG_FIXED |
+			TTM_MEMTYPE_FLAG_MAPPABLE;
+		man->available_caching = TTM_PL_FLAG_UNCACHED |
+			TTM_PL_FLAG_WC;
+		man->default_caching = TTM_PL_FLAG_WC;
+		break;
+	default:
+		DRM_ERROR("Unsupported memory type %u\n", (unsigned)type);
+		return -EINVAL;
+	}
+	return 0;
+}
+
+static void
+bochs_bo_evict_flags(struct ttm_buffer_object *bo, struct ttm_placement *pl)
+{
+	struct bochs_bo *bochsbo = bochs_bo(bo);
+
+	if (!bochs_ttm_bo_is_bochs_bo(bo))
+		return;
+
+	bochs_ttm_placement(bochsbo, TTM_PL_FLAG_SYSTEM);
+	*pl = bochsbo->placement;
+}
+
+static int bochs_bo_verify_access(struct ttm_buffer_object *bo,
+				  struct file *filp)
+{
+	struct bochs_bo *bochsbo = bochs_bo(bo);
+
+	return drm_vma_node_verify_access(&bochsbo->gem.vma_node, filp);
+}
+
+static int bochs_ttm_io_mem_reserve(struct ttm_bo_device *bdev,
+				    struct ttm_mem_reg *mem)
+{
+	struct ttm_mem_type_manager *man = &bdev->man[mem->mem_type];
+	struct bochs_device *bochs = bochs_bdev(bdev);
+
+	mem->bus.addr = NULL;
+	mem->bus.offset = 0;
+	mem->bus.size = mem->num_pages << PAGE_SHIFT;
+	mem->bus.base = 0;
+	mem->bus.is_iomem = false;
+	if (!(man->flags & TTM_MEMTYPE_FLAG_MAPPABLE))
+		return -EINVAL;
+	switch (mem->mem_type) {
+	case TTM_PL_SYSTEM:
+		/* system memory */
+		return 0;
+	case TTM_PL_VRAM:
+		mem->bus.offset = mem->start << PAGE_SHIFT;
+		mem->bus.base = bochs->fb_base;
+		mem->bus.is_iomem = true;
+		break;
+	default:
+		return -EINVAL;
+		break;
+	}
+	return 0;
+}
+
+static void bochs_ttm_io_mem_free(struct ttm_bo_device *bdev,
+				  struct ttm_mem_reg *mem)
+{
+}
+
+static int bochs_bo_move(struct ttm_buffer_object *bo,
+			 bool evict, bool interruptible,
+			 bool no_wait_gpu,
+			 struct ttm_mem_reg *new_mem)
+{
+	return ttm_bo_move_memcpy(bo, evict, no_wait_gpu, new_mem);
+}
+
+
+static void bochs_ttm_backend_destroy(struct ttm_tt *tt)
+{
+	ttm_tt_fini(tt);
+	kfree(tt);
+}
+
+static struct ttm_backend_func bochs_tt_backend_func = {
+	.destroy = &bochs_ttm_backend_destroy,
+};
+
+static struct ttm_tt *bochs_ttm_tt_create(struct ttm_bo_device *bdev,
+					  unsigned long size,
+					  uint32_t page_flags,
+					  struct page *dummy_read_page)
+{
+	struct ttm_tt *tt;
+
+	tt = kzalloc(sizeof(struct ttm_tt), GFP_KERNEL);
+	if (tt == NULL)
+		return NULL;
+	tt->func = &bochs_tt_backend_func;
+	if (ttm_tt_init(tt, bdev, size, page_flags, dummy_read_page)) {
+		kfree(tt);
+		return NULL;
+	}
+	return tt;
+}
+
+struct ttm_bo_driver bochs_bo_driver = {
+	.ttm_tt_create = bochs_ttm_tt_create,
+	.ttm_tt_populate = ttm_pool_populate,
+	.ttm_tt_unpopulate = ttm_pool_unpopulate,
+	.init_mem_type = bochs_bo_init_mem_type,
+	.evict_flags = bochs_bo_evict_flags,
+	.move = bochs_bo_move,
+	.verify_access = bochs_bo_verify_access,
+	.io_mem_reserve = &bochs_ttm_io_mem_reserve,
+	.io_mem_free = &bochs_ttm_io_mem_free,
+};
+
+int bochs_mm_init(struct bochs_device *bochs)
+{
+	struct ttm_bo_device *bdev = &bochs->ttm.bdev;
+	int ret;
+
+	ret = bochs_ttm_global_init(bochs);
+	if (ret)
+		return ret;
+
+	ret = ttm_bo_device_init(&bochs->ttm.bdev,
+				 bochs->ttm.bo_global_ref.ref.object,
+				 &bochs_bo_driver, DRM_FILE_PAGE_OFFSET,
+				 true);
+	if (ret) {
+		DRM_ERROR("Error initialising bo driver; %d\n", ret);
+		return ret;
+	}
+
+	ret = ttm_bo_init_mm(bdev, TTM_PL_VRAM,
+			     bochs->fb_size >> PAGE_SHIFT);
+	if (ret) {
+		DRM_ERROR("Failed ttm VRAM init: %d\n", ret);
+		return ret;
+	}
+
+	bochs->ttm.initialized = true;
+	return 0;
+}
+
+void bochs_mm_fini(struct bochs_device *bochs)
+{
+	if (!bochs->ttm.initialized)
+		return;
+
+	ttm_bo_device_release(&bochs->ttm.bdev);
+	bochs_ttm_global_release(bochs);
+	bochs->ttm.initialized = false;
+}
+
+static void bochs_ttm_placement(struct bochs_bo *bo, int domain)
+{
+	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
+			| TTM_PL_FLAG_VRAM;
+	}
+	if (domain & TTM_PL_FLAG_SYSTEM) {
+		bo->placements[c++] = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
+	}
+	if (!c) {
+		bo->placements[c++] = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
+	}
+	bo->placement.num_placement = c;
+	bo->placement.num_busy_placement = c;
+}
+
+static inline u64 bochs_bo_gpu_offset(struct bochs_bo *bo)
+{
+	return bo->bo.offset;
+}
+
+int bochs_bo_pin(struct bochs_bo *bo, u32 pl_flag, u64 *gpu_addr)
+{
+	int i, ret;
+
+	if (bo->pin_count) {
+		bo->pin_count++;
+		if (gpu_addr)
+			*gpu_addr = bochs_bo_gpu_offset(bo);
+		return 0;
+	}
+
+	bochs_ttm_placement(bo, pl_flag);
+	for (i = 0; i < bo->placement.num_placement; i++)
+		bo->placements[i] |= TTM_PL_FLAG_NO_EVICT;
+	ret = ttm_bo_validate(&bo->bo, &bo->placement, false, false);
+	if (ret)
+		return ret;
+
+	bo->pin_count = 1;
+	if (gpu_addr)
+		*gpu_addr = bochs_bo_gpu_offset(bo);
+	return 0;
+}
+
+int bochs_bo_unpin(struct bochs_bo *bo)
+{
+	int i, ret;
+
+	if (!bo->pin_count) {
+		DRM_ERROR("unpin bad %p\n", bo);
+		return 0;
+	}
+	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;
+	ret = ttm_bo_validate(&bo->bo, &bo->placement, false, false);
+	if (ret)
+		return ret;
+
+	return 0;
+}
+
+int bochs_mmap(struct file *filp, struct vm_area_struct *vma)
+{
+	struct drm_file *file_priv;
+	struct bochs_device *bochs;
+
+	if (unlikely(vma->vm_pgoff < DRM_FILE_PAGE_OFFSET))
+		return drm_mmap(filp, vma);
+
+	file_priv = filp->private_data;
+	bochs = file_priv->minor->dev->dev_private;
+	return ttm_bo_mmap(filp, vma, &bochs->ttm.bdev);
+}
+
+/* ---------------------------------------------------------------------- */
+
+static int bochs_bo_create(struct drm_device *dev, int size, int align,
+			   uint32_t flags, struct bochs_bo **pbochsbo)
+{
+	struct bochs_device *bochs = dev->dev_private;
+	struct bochs_bo *bochsbo;
+	size_t acc_size;
+	int ret;
+
+	bochsbo = kzalloc(sizeof(struct bochs_bo), GFP_KERNEL);
+	if (!bochsbo)
+		return -ENOMEM;
+
+	ret = drm_gem_object_init(dev, &bochsbo->gem, size);
+	if (ret) {
+		kfree(bochsbo);
+		return ret;
+	}
+
+	bochsbo->bo.bdev = &bochs->ttm.bdev;
+	bochsbo->bo.bdev->dev_mapping = dev->dev_mapping;
+
+	bochs_ttm_placement(bochsbo, TTM_PL_FLAG_VRAM | TTM_PL_FLAG_SYSTEM);
+
+	acc_size = ttm_bo_dma_acc_size(&bochs->ttm.bdev, size,
+				       sizeof(struct bochs_bo));
+
+	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);
+	if (ret)
+		return ret;
+
+	*pbochsbo = bochsbo;
+	return 0;
+}
+
+int bochs_gem_create(struct drm_device *dev, u32 size, bool iskernel,
+		     struct drm_gem_object **obj)
+{
+	struct bochs_bo *bochsbo;
+	int ret;
+
+	*obj = NULL;
+
+	size = ALIGN(size, PAGE_SIZE);
+	if (size == 0)
+		return -EINVAL;
+
+	ret = bochs_bo_create(dev, size, 0, 0, &bochsbo);
+	if (ret) {
+		if (ret != -ERESTARTSYS)
+			DRM_ERROR("failed to allocate GEM object\n");
+		return ret;
+	}
+	*obj = &bochsbo->gem;
+	return 0;
+}
+
+int bochs_dumb_create(struct drm_file *file, struct drm_device *dev,
+		      struct drm_mode_create_dumb *args)
+{
+	struct drm_gem_object *gobj;
+	u32 handle;
+	int ret;
+
+	args->pitch = args->width * ((args->bpp + 7) / 8);
+	args->size = args->pitch * args->height;
+
+	ret = bochs_gem_create(dev, args->size, false,
+			       &gobj);
+	if (ret)
+		return ret;
+
+	ret = drm_gem_handle_create(file, gobj, &handle);
+	drm_gem_object_unreference_unlocked(gobj);
+	if (ret)
+		return ret;
+
+	args->handle = handle;
+	return 0;
+}
+
+static void bochs_bo_unref(struct bochs_bo **bo)
+{
+	struct ttm_buffer_object *tbo;
+
+	if ((*bo) == NULL)
+		return;
+
+	tbo = &((*bo)->bo);
+	ttm_bo_unref(&tbo);
+	if (tbo == NULL)
+		*bo = NULL;
+
+}
+
+void bochs_gem_free_object(struct drm_gem_object *obj)
+{
+	struct bochs_bo *bochs_bo = gem_to_bochs_bo(obj);
+
+	if (!bochs_bo)
+		return;
+	bochs_bo_unref(&bochs_bo);
+}
+
+int bochs_dumb_mmap_offset(struct drm_file *file, struct drm_device *dev,
+			   uint32_t handle, uint64_t *offset)
+{
+	struct drm_gem_object *obj;
+	int ret;
+	struct bochs_bo *bo;
+
+	mutex_lock(&dev->struct_mutex);
+	obj = drm_gem_object_lookup(dev, file, handle);
+	if (obj == NULL) {
+		ret = -ENOENT;
+		goto out_unlock;
+	}
+
+	bo = gem_to_bochs_bo(obj);
+	*offset = bochs_bo_mmap_offset(bo);
+
+	drm_gem_object_unreference(obj);
+	ret = 0;
+out_unlock:
+	mutex_unlock(&dev->struct_mutex);
+	return ret;
+
+}
+
+/* ---------------------------------------------------------------------- */
+
+static void bochs_user_framebuffer_destroy(struct drm_framebuffer *fb)
+{
+	struct bochs_framebuffer *bochs_fb = to_bochs_framebuffer(fb);
+	if (bochs_fb->obj)
+		drm_gem_object_unreference_unlocked(bochs_fb->obj);
+	drm_framebuffer_cleanup(fb);
+	kfree(fb);
+}
+
+static const struct drm_framebuffer_funcs bochs_fb_funcs = {
+	.destroy = bochs_user_framebuffer_destroy,
+};
+
+int bochs_framebuffer_init(struct drm_device *dev,
+			   struct bochs_framebuffer *gfb,
+			   struct drm_mode_fb_cmd2 *mode_cmd,
+			   struct drm_gem_object *obj)
+{
+	int ret;
+
+	drm_helper_mode_fill_fb_struct(&gfb->base, mode_cmd);
+	gfb->obj = obj;
+	ret = drm_framebuffer_init(dev, &gfb->base, &bochs_fb_funcs);
+	if (ret) {
+		DRM_ERROR("drm_framebuffer_init failed: %d\n", ret);
+		return ret;
+	}
+	return 0;
+}
+
+static struct drm_framebuffer *
+bochs_user_framebuffer_create(struct drm_device *dev,
+			      struct drm_file *filp,
+			      struct drm_mode_fb_cmd2 *mode_cmd)
+{
+	struct drm_gem_object *obj;
+	struct bochs_framebuffer *bochs_fb;
+	int ret;
+
+	DRM_DEBUG_DRIVER("%dx%d, format %c%c%c%c\n",
+	       mode_cmd->width, mode_cmd->height,
+	       (mode_cmd->pixel_format)       & 0xff,
+	       (mode_cmd->pixel_format >> 8)  & 0xff,
+	       (mode_cmd->pixel_format >> 16) & 0xff,
+	       (mode_cmd->pixel_format >> 24) & 0xff);
+
+	if (mode_cmd->pixel_format != DRM_FORMAT_XRGB8888)
+		return ERR_PTR(-ENOENT);
+
+	obj = drm_gem_object_lookup(dev, filp, mode_cmd->handles[0]);
+	if (obj == NULL)
+		return ERR_PTR(-ENOENT);
+
+	bochs_fb = kzalloc(sizeof(*bochs_fb), GFP_KERNEL);
+	if (!bochs_fb) {
+		drm_gem_object_unreference_unlocked(obj);
+		return ERR_PTR(-ENOMEM);
+	}
+
+	ret = bochs_framebuffer_init(dev, bochs_fb, mode_cmd, obj);
+	if (ret) {
+		drm_gem_object_unreference_unlocked(obj);
+		kfree(bochs_fb);
+		return ERR_PTR(ret);
+	}
+	return &bochs_fb->base;
+}
+
+const struct drm_mode_config_funcs bochs_mode_funcs = {
+	.fb_create = bochs_user_framebuffer_create,
+};
diff --git a/drivers/gpu/drm/cirrus/cirrus_drv.h b/drivers/gpu/drm/cirrus/cirrus_drv.h
index b6aded73838b..117d3eca5e37 100644
--- a/drivers/gpu/drm/cirrus/cirrus_drv.h
+++ b/drivers/gpu/drm/cirrus/cirrus_drv.h
@@ -222,7 +222,7 @@ void cirrus_fbdev_fini(struct cirrus_device *cdev);
 void cirrus_driver_irq_preinstall(struct drm_device *dev);
 int cirrus_driver_irq_postinstall(struct drm_device *dev);
 void cirrus_driver_irq_uninstall(struct drm_device *dev);
-irqreturn_t cirrus_driver_irq_handler(DRM_IRQ_ARGS);
+irqreturn_t cirrus_driver_irq_handler(int irq, void *arg);
 
 				/* cirrus_kms.c */
 int cirrus_driver_load(struct drm_device *dev, unsigned long flags);
diff --git a/drivers/gpu/drm/cirrus/cirrus_fbdev.c b/drivers/gpu/drm/cirrus/cirrus_fbdev.c
index b27e95666fab..2fd4a92162cb 100644
--- a/drivers/gpu/drm/cirrus/cirrus_fbdev.c
+++ b/drivers/gpu/drm/cirrus/cirrus_fbdev.c
@@ -39,7 +39,7 @@ static void cirrus_dirty_update(struct cirrus_fbdev *afbdev,
 	 * then the BO is being moved and we should
 	 * store up the damage until later.
 	 */
-	if (!in_interrupt())
+	if (!drm_can_sleep())
 		ret = cirrus_bo_reserve(bo, true);
 	if (ret) {
 		if (ret != -EBUSY)
@@ -233,6 +233,9 @@ static int cirrusfb_create(struct drm_fb_helper *helper,
 	info->apertures->ranges[0].base = cdev->dev->mode_config.fb_base;
 	info->apertures->ranges[0].size = cdev->mc.vram_size;
 
+	info->fix.smem_start = cdev->dev->mode_config.fb_base;
+	info->fix.smem_len = cdev->mc.vram_size;
+
 	info->screen_base = sysram;
 	info->screen_size = size;
 
diff --git a/drivers/gpu/drm/cirrus/cirrus_main.c b/drivers/gpu/drm/cirrus/cirrus_main.c
index 78e76f24343d..4b0170cf53fd 100644
--- a/drivers/gpu/drm/cirrus/cirrus_main.c
+++ b/drivers/gpu/drm/cirrus/cirrus_main.c
@@ -255,7 +255,7 @@ int cirrus_dumb_create(struct drm_file *file,
 	return 0;
 }
 
-void cirrus_bo_unref(struct cirrus_bo **bo)
+static void cirrus_bo_unref(struct cirrus_bo **bo)
 {
 	struct ttm_buffer_object *tbo;
 
diff --git a/drivers/gpu/drm/cirrus/cirrus_mode.c b/drivers/gpu/drm/cirrus/cirrus_mode.c
index adabc3daaa5b..530f78f84dee 100644
--- a/drivers/gpu/drm/cirrus/cirrus_mode.c
+++ b/drivers/gpu/drm/cirrus/cirrus_mode.c
@@ -102,7 +102,7 @@ static bool cirrus_crtc_mode_fixup(struct drm_crtc *crtc,
 	return true;
 }
 
-void cirrus_set_start_address(struct drm_crtc *crtc, unsigned offset)
+static void cirrus_set_start_address(struct drm_crtc *crtc, unsigned offset)
 {
 	struct cirrus_device *cdev = crtc->dev->dev_private;
 	u32 addr;
@@ -273,8 +273,8 @@ static int cirrus_crtc_mode_set(struct drm_crtc *crtc,
 		sr07 |= 0x11;
 		break;
 	case 16:
-		sr07 |= 0xc1;
-		hdr = 0xc0;
+		sr07 |= 0x17;
+		hdr = 0xc1;
 		break;
 	case 24:
 		sr07 |= 0x15;
@@ -453,7 +453,7 @@ static void cirrus_encoder_commit(struct drm_encoder *encoder)
 {
 }
 
-void cirrus_encoder_destroy(struct drm_encoder *encoder)
+static void cirrus_encoder_destroy(struct drm_encoder *encoder)
 {
 	struct cirrus_encoder *cirrus_encoder = to_cirrus_encoder(encoder);
 	drm_encoder_cleanup(encoder);
@@ -492,7 +492,7 @@ static struct drm_encoder *cirrus_encoder_init(struct drm_device *dev)
 }
 
 
-int cirrus_vga_get_modes(struct drm_connector *connector)
+static int cirrus_vga_get_modes(struct drm_connector *connector)
 {
 	int count;
 
@@ -509,7 +509,7 @@ static int cirrus_vga_mode_valid(struct drm_connector *connector,
 	return MODE_OK;
 }
 
-struct drm_encoder *cirrus_connector_best_encoder(struct drm_connector
+static struct drm_encoder *cirrus_connector_best_encoder(struct drm_connector
 						  *connector)
 {
 	int enc_id = connector->encoder_ids[0];
diff --git a/drivers/gpu/drm/cirrus/cirrus_ttm.c b/drivers/gpu/drm/cirrus/cirrus_ttm.c
index 75becdeac07d..8b37c25ff9bd 100644
--- a/drivers/gpu/drm/cirrus/cirrus_ttm.c
+++ b/drivers/gpu/drm/cirrus/cirrus_ttm.c
@@ -80,7 +80,7 @@ static int cirrus_ttm_global_init(struct cirrus_device *cirrus)
 	return 0;
 }
 
-void
+static void
 cirrus_ttm_global_release(struct cirrus_device *cirrus)
 {
 	if (cirrus->ttm.mem_global_ref.release == NULL)
@@ -102,7 +102,7 @@ static void cirrus_bo_ttm_destroy(struct ttm_buffer_object *tbo)
 	kfree(bo);
 }
 
-bool cirrus_ttm_bo_is_cirrus_bo(struct ttm_buffer_object *bo)
+static bool cirrus_ttm_bo_is_cirrus_bo(struct ttm_buffer_object *bo)
 {
 	if (bo->destroy == &cirrus_bo_ttm_destroy)
 		return true;
@@ -208,7 +208,7 @@ static struct ttm_backend_func cirrus_tt_backend_func = {
 };
 
 
-struct ttm_tt *cirrus_ttm_tt_create(struct ttm_bo_device *bdev,
+static struct ttm_tt *cirrus_ttm_tt_create(struct ttm_bo_device *bdev,
 				 unsigned long size, uint32_t page_flags,
 				 struct page *dummy_read_page)
 {
@@ -375,26 +375,6 @@ int cirrus_bo_pin(struct cirrus_bo *bo, u32 pl_flag, u64 *gpu_addr)
 	return 0;
 }
 
-int cirrus_bo_unpin(struct cirrus_bo *bo)
-{
-	int i, ret;
-	if (!bo->pin_count) {
-		DRM_ERROR("unpin bad %p\n", bo);
-		return 0;
-	}
-	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;
-	ret = ttm_bo_validate(&bo->bo, &bo->placement, false, false);
-	if (ret)
-		return ret;
-
-	return 0;
-}
-
 int cirrus_bo_push_sysram(struct cirrus_bo *bo)
 {
 	int i, ret;
diff --git a/drivers/gpu/drm/drm_agpsupport.c b/drivers/gpu/drm/drm_agpsupport.c
index e301d653d97e..dde205cef384 100644
--- a/drivers/gpu/drm/drm_agpsupport.c
+++ b/drivers/gpu/drm/drm_agpsupport.c
@@ -53,7 +53,7 @@
  */
 int drm_agp_info(struct drm_device *dev, struct drm_agp_info *info)
 {
-	DRM_AGP_KERN *kern;
+	struct agp_kern_info *kern;
 
 	if (!dev->agp || !dev->agp->acquired)
 		return -EINVAL;
@@ -198,17 +198,15 @@ int drm_agp_enable_ioctl(struct drm_device *dev, void *data,
 int drm_agp_alloc(struct drm_device *dev, struct drm_agp_buffer *request)
 {
 	struct drm_agp_mem *entry;
-	DRM_AGP_MEM *memory;
+	struct agp_memory *memory;
 	unsigned long pages;
 	u32 type;
 
 	if (!dev->agp || !dev->agp->acquired)
 		return -EINVAL;
-	if (!(entry = kmalloc(sizeof(*entry), GFP_KERNEL)))
+	if (!(entry = kzalloc(sizeof(*entry), GFP_KERNEL)))
 		return -ENOMEM;
 
-	memset(entry, 0, sizeof(*entry));
-
 	pages = (request->size + PAGE_SIZE - 1) / PAGE_SIZE;
 	type = (u32) request->type;
 	if (!(memory = agp_allocate_memory(dev->agp->bridge, pages, type))) {
@@ -393,14 +391,16 @@ int drm_agp_free_ioctl(struct drm_device *dev, void *data,
  * Gets the drm_agp_t structure which is made available by the agpgart module
  * via the inter_module_* functions. Creates and initializes a drm_agp_head
  * structure.
+ *
+ * Note that final cleanup of the kmalloced structure is directly done in
+ * drm_pci_agp_destroy.
  */
 struct drm_agp_head *drm_agp_init(struct drm_device *dev)
 {
 	struct drm_agp_head *head = NULL;
 
-	if (!(head = kmalloc(sizeof(*head), GFP_KERNEL)))
+	if (!(head = kzalloc(sizeof(*head), GFP_KERNEL)))
 		return NULL;
-	memset((void *)head, 0, sizeof(*head));
 	head->bridge = agp_find_bridge(dev->pdev);
 	if (!head->bridge) {
 		if (!(head->bridge = agp_backend_acquire(dev->pdev))) {
@@ -439,7 +439,7 @@ void drm_agp_clear(struct drm_device *dev)
 {
 	struct drm_agp_mem *entry, *tempe;
 
-	if (!drm_core_has_AGP(dev) || !dev->agp)
+	if (!dev->agp)
 		return;
 	if (drm_core_check_feature(dev, DRIVER_MODESET))
 		return;
@@ -460,35 +460,20 @@ void drm_agp_clear(struct drm_device *dev)
 }
 
 /**
- * drm_agp_destroy - Destroy AGP head
- * @dev: DRM device
- *
- * Destroy resources that were previously allocated via drm_agp_initp. Caller
- * must ensure to clean up all AGP resources before calling this. See
- * drm_agp_clear().
- *
- * Call this to destroy AGP heads allocated via drm_agp_init().
- */
-void drm_agp_destroy(struct drm_agp_head *agp)
-{
-	kfree(agp);
-}
-
-/**
  * Binds a collection of pages into AGP memory at the given offset, returning
  * the AGP memory structure containing them.
  *
  * No reference is held on the pages during this time -- it is up to the
  * caller to handle that.
  */
-DRM_AGP_MEM *
+struct agp_memory *
 drm_agp_bind_pages(struct drm_device *dev,
 		   struct page **pages,
 		   unsigned long num_pages,
 		   uint32_t gtt_offset,
 		   u32 type)
 {
-	DRM_AGP_MEM *mem;
+	struct agp_memory *mem;
 	int ret, i;
 
 	DRM_DEBUG("\n");
diff --git a/drivers/gpu/drm/drm_buffer.c b/drivers/gpu/drm/drm_buffer.c
index 39a718340319..0406110f83ed 100644
--- a/drivers/gpu/drm/drm_buffer.c
+++ b/drivers/gpu/drm/drm_buffer.c
@@ -114,7 +114,7 @@ int drm_buffer_copy_from_user(struct drm_buffer *buf,
 
 	for (idx = 0; idx < nr_pages; ++idx) {
 
-		if (DRM_COPY_FROM_USER(buf->data[idx],
+		if (copy_from_user(buf->data[idx],
 			user_data + idx * PAGE_SIZE,
 			min(PAGE_SIZE, size - idx * PAGE_SIZE))) {
 			DRM_ERROR("Failed to copy user data (%p) to drm buffer"
diff --git a/drivers/gpu/drm/drm_bufs.c b/drivers/gpu/drm/drm_bufs.c
index 471e051d295e..edec31fe3fed 100644
--- a/drivers/gpu/drm/drm_bufs.c
+++ b/drivers/gpu/drm/drm_bufs.c
@@ -261,7 +261,7 @@ static int drm_addmap_core(struct drm_device * dev, resource_size_t offset,
 		struct drm_agp_mem *entry;
 		int valid = 0;
 
-		if (!drm_core_has_AGP(dev)) {
+		if (!dev->agp) {
 			kfree(map);
 			return -EINVAL;
 		}
@@ -303,9 +303,6 @@ static int drm_addmap_core(struct drm_device * dev, resource_size_t offset,
 
 		break;
 	}
-	case _DRM_GEM:
-		DRM_ERROR("tried to addmap GEM object\n");
-		break;
 	case _DRM_SCATTER_GATHER:
 		if (!dev->sg) {
 			kfree(map);
@@ -483,9 +480,6 @@ int drm_rmmap_locked(struct drm_device *dev, struct drm_local_map *map)
 		dmah.size = map->size;
 		__drm_pci_free(dev, &dmah);
 		break;
-	case _DRM_GEM:
-		DRM_ERROR("tried to rmmap GEM object\n");
-		break;
 	}
 	kfree(map);
 
@@ -1396,7 +1390,7 @@ int drm_mapbufs(struct drm_device *dev, void *data,
 	spin_unlock(&dev->count_lock);
 
 	if (request->count >= dma->buf_count) {
-		if ((drm_core_has_AGP(dev) && (dma->flags & _DRM_DMA_USE_AGP))
+		if ((dev->agp && (dma->flags & _DRM_DMA_USE_AGP))
 		    || (drm_core_check_feature(dev, DRIVER_SG)
 			&& (dma->flags & _DRM_DMA_USE_SG))) {
 			struct drm_local_map *map = dev->agp_buffer_map;
diff --git a/drivers/gpu/drm/drm_crtc.c b/drivers/gpu/drm/drm_crtc.c
index d6cf77c472e7..3b7d32da1604 100644
--- a/drivers/gpu/drm/drm_crtc.c
+++ b/drivers/gpu/drm/drm_crtc.c
@@ -675,6 +675,29 @@ void drm_crtc_cleanup(struct drm_crtc *crtc)
 EXPORT_SYMBOL(drm_crtc_cleanup);
 
 /**
+ * drm_crtc_index - find the index of a registered CRTC
+ * @crtc: CRTC to find index for
+ *
+ * Given a registered CRTC, return the index of that CRTC within a DRM
+ * device's list of CRTCs.
+ */
+unsigned int drm_crtc_index(struct drm_crtc *crtc)
+{
+	unsigned int index = 0;
+	struct drm_crtc *tmp;
+
+	list_for_each_entry(tmp, &crtc->dev->mode_config.crtc_list, head) {
+		if (tmp == crtc)
+			return index;
+
+		index++;
+	}
+
+	BUG();
+}
+EXPORT_SYMBOL(drm_crtc_index);
+
+/**
  * drm_mode_probed_add - add a mode to a connector's probed mode list
  * @connector: connector the new mode
  * @mode: mode data
@@ -2767,10 +2790,8 @@ int drm_mode_dirtyfb_ioctl(struct drm_device *dev,
 	}
 
 	if (fb->funcs->dirty) {
-		drm_modeset_lock_all(dev);
 		ret = fb->funcs->dirty(fb, file_priv, flags, r->color,
 				       clips, num_clips);
-		drm_modeset_unlock_all(dev);
 	} else {
 		ret = -ENOSYS;
 	}
diff --git a/drivers/gpu/drm/drm_crtc_helper.c b/drivers/gpu/drm/drm_crtc_helper.c
index 01361aba033b..ea92b827e787 100644
--- a/drivers/gpu/drm/drm_crtc_helper.c
+++ b/drivers/gpu/drm/drm_crtc_helper.c
@@ -324,35 +324,6 @@ void drm_helper_disable_unused_functions(struct drm_device *dev)
 }
 EXPORT_SYMBOL(drm_helper_disable_unused_functions);
 
-/**
- * drm_encoder_crtc_ok - can a given crtc drive a given encoder?
- * @encoder: encoder to test
- * @crtc: crtc to test
- *
- * Return false if @encoder can't be driven by @crtc, true otherwise.
- */
-static bool drm_encoder_crtc_ok(struct drm_encoder *encoder,
-				struct drm_crtc *crtc)
-{
-	struct drm_device *dev;
-	struct drm_crtc *tmp;
-	int crtc_mask = 1;
-
-	WARN(!crtc, "checking null crtc?\n");
-
-	dev = crtc->dev;
-
-	list_for_each_entry(tmp, &dev->mode_config.crtc_list, head) {
-		if (tmp == crtc)
-			break;
-		crtc_mask <<= 1;
-	}
-
-	if (encoder->possible_crtcs & crtc_mask)
-		return true;
-	return false;
-}
-
 /*
  * Check the CRTC we're going to map each output to vs. its current
  * CRTC.  If they don't match, we have to disable the output and the CRTC
@@ -536,7 +507,7 @@ bool drm_crtc_helper_set_mode(struct drm_crtc *crtc,
 	 * are later needed by vblank and swap-completion
 	 * timestamping. They are derived from true hwmode.
 	 */
-	drm_calc_timestamping_constants(crtc);
+	drm_calc_timestamping_constants(crtc, &crtc->hwmode);
 
 	/* FIXME: add subpixel order */
 done:
diff --git a/drivers/gpu/drm/drm_drv.c b/drivers/gpu/drm/drm_drv.c
index d9137e49c4e8..345be03c23db 100644
--- a/drivers/gpu/drm/drm_drv.c
+++ b/drivers/gpu/drm/drm_drv.c
@@ -315,9 +315,6 @@ long drm_ioctl(struct file *filp,
 	if (drm_device_is_unplugged(dev))
 		return -ENODEV;
 
-	atomic_inc(&dev->ioctl_count);
-	++file_priv->ioctl_count;
-
 	if ((nr >= DRM_CORE_IOCTL_COUNT) &&
 	    ((nr < DRM_COMMAND_BASE) || (nr >= DRM_COMMAND_END)))
 		goto err_i1;
@@ -410,7 +407,6 @@ long drm_ioctl(struct file *filp,
 
 	if (kdata != stack_kdata)
 		kfree(kdata);
-	atomic_dec(&dev->ioctl_count);
 	if (retcode)
 		DRM_DEBUG("ret = %d\n", retcode);
 	return retcode;
diff --git a/drivers/gpu/drm/drm_edid.c b/drivers/gpu/drm/drm_edid.c
index 8835dcddfac3..b924306b8477 100644
--- a/drivers/gpu/drm/drm_edid.c
+++ b/drivers/gpu/drm/drm_edid.c
@@ -605,347 +605,347 @@ static const struct drm_display_mode edid_cea_modes[] = {
 	{ DRM_MODE("640x480", DRM_MODE_TYPE_DRIVER, 25175, 640, 656,
 		   752, 800, 0, 480, 490, 492, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 2 - 720x480@60Hz */
 	{ DRM_MODE("720x480", DRM_MODE_TYPE_DRIVER, 27000, 720, 736,
 		   798, 858, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 3 - 720x480@60Hz */
 	{ DRM_MODE("720x480", DRM_MODE_TYPE_DRIVER, 27000, 720, 736,
 		   798, 858, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 4 - 1280x720@60Hz */
 	{ DRM_MODE("1280x720", DRM_MODE_TYPE_DRIVER, 74250, 1280, 1390,
 		   1430, 1650, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 5 - 1920x1080i@60Hz */
 	{ DRM_MODE("1920x1080i", DRM_MODE_TYPE_DRIVER, 74250, 1920, 2008,
 		   2052, 2200, 0, 1080, 1084, 1094, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC |
 			DRM_MODE_FLAG_INTERLACE),
-	  .vrefresh = 60, },
+	  .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,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 60, },
+	  .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,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 60, },
+	  .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,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 60, },
+	  .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,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 10 - 2880x480i@60Hz */
 	{ DRM_MODE("2880x480i", DRM_MODE_TYPE_DRIVER, 54000, 2880, 2956,
 		   3204, 3432, 0, 480, 488, 494, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 11 - 2880x480i@60Hz */
 	{ DRM_MODE("2880x480i", DRM_MODE_TYPE_DRIVER, 54000, 2880, 2956,
 		   3204, 3432, 0, 480, 488, 494, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 12 - 2880x240@60Hz */
 	{ DRM_MODE("2880x240", DRM_MODE_TYPE_DRIVER, 54000, 2880, 2956,
 		   3204, 3432, 0, 240, 244, 247, 262, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 13 - 2880x240@60Hz */
 	{ DRM_MODE("2880x240", DRM_MODE_TYPE_DRIVER, 54000, 2880, 2956,
 		   3204, 3432, 0, 240, 244, 247, 262, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 14 - 1440x480@60Hz */
 	{ DRM_MODE("1440x480", DRM_MODE_TYPE_DRIVER, 54000, 1440, 1472,
 		   1596, 1716, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 15 - 1440x480@60Hz */
 	{ DRM_MODE("1440x480", DRM_MODE_TYPE_DRIVER, 54000, 1440, 1472,
 		   1596, 1716, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 16 - 1920x1080@60Hz */
 	{ DRM_MODE("1920x1080", DRM_MODE_TYPE_DRIVER, 148500, 1920, 2008,
 		   2052, 2200, 0, 1080, 1084, 1089, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 17 - 720x576@50Hz */
 	{ DRM_MODE("720x576", DRM_MODE_TYPE_DRIVER, 27000, 720, 732,
 		   796, 864, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 18 - 720x576@50Hz */
 	{ DRM_MODE("720x576", DRM_MODE_TYPE_DRIVER, 27000, 720, 732,
 		   796, 864, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 19 - 1280x720@50Hz */
 	{ DRM_MODE("1280x720", DRM_MODE_TYPE_DRIVER, 74250, 1280, 1720,
 		   1760, 1980, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 20 - 1920x1080i@50Hz */
 	{ DRM_MODE("1920x1080i", DRM_MODE_TYPE_DRIVER, 74250, 1920, 2448,
 		   2492, 2640, 0, 1080, 1084, 1094, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC |
 			DRM_MODE_FLAG_INTERLACE),
-	  .vrefresh = 50, },
+	  .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,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 50, },
+	  .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,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 50, },
+	  .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,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 50, },
+	  .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,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 25 - 2880x576i@50Hz */
 	{ DRM_MODE("2880x576i", DRM_MODE_TYPE_DRIVER, 54000, 2880, 2928,
 		   3180, 3456, 0, 576, 580, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 26 - 2880x576i@50Hz */
 	{ DRM_MODE("2880x576i", DRM_MODE_TYPE_DRIVER, 54000, 2880, 2928,
 		   3180, 3456, 0, 576, 580, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 27 - 2880x288@50Hz */
 	{ DRM_MODE("2880x288", DRM_MODE_TYPE_DRIVER, 54000, 2880, 2928,
 		   3180, 3456, 0, 288, 290, 293, 312, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 28 - 2880x288@50Hz */
 	{ DRM_MODE("2880x288", DRM_MODE_TYPE_DRIVER, 54000, 2880, 2928,
 		   3180, 3456, 0, 288, 290, 293, 312, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 29 - 1440x576@50Hz */
 	{ DRM_MODE("1440x576", DRM_MODE_TYPE_DRIVER, 54000, 1440, 1464,
 		   1592, 1728, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 30 - 1440x576@50Hz */
 	{ DRM_MODE("1440x576", DRM_MODE_TYPE_DRIVER, 54000, 1440, 1464,
 		   1592, 1728, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 31 - 1920x1080@50Hz */
 	{ DRM_MODE("1920x1080", DRM_MODE_TYPE_DRIVER, 148500, 1920, 2448,
 		   2492, 2640, 0, 1080, 1084, 1089, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 32 - 1920x1080@24Hz */
 	{ DRM_MODE("1920x1080", DRM_MODE_TYPE_DRIVER, 74250, 1920, 2558,
 		   2602, 2750, 0, 1080, 1084, 1089, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
-	  .vrefresh = 24, },
+	  .vrefresh = 24, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 33 - 1920x1080@25Hz */
 	{ DRM_MODE("1920x1080", DRM_MODE_TYPE_DRIVER, 74250, 1920, 2448,
 		   2492, 2640, 0, 1080, 1084, 1089, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
-	  .vrefresh = 25, },
+	  .vrefresh = 25, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 34 - 1920x1080@30Hz */
 	{ DRM_MODE("1920x1080", DRM_MODE_TYPE_DRIVER, 74250, 1920, 2008,
 		   2052, 2200, 0, 1080, 1084, 1089, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
-	  .vrefresh = 30, },
+	  .vrefresh = 30, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 35 - 2880x480@60Hz */
 	{ DRM_MODE("2880x480", DRM_MODE_TYPE_DRIVER, 108000, 2880, 2944,
 		   3192, 3432, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 36 - 2880x480@60Hz */
 	{ DRM_MODE("2880x480", DRM_MODE_TYPE_DRIVER, 108000, 2880, 2944,
 		   3192, 3432, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 60, },
+	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 37 - 2880x576@50Hz */
 	{ DRM_MODE("2880x576", DRM_MODE_TYPE_DRIVER, 108000, 2880, 2928,
 		   3184, 3456, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 38 - 2880x576@50Hz */
 	{ DRM_MODE("2880x576", DRM_MODE_TYPE_DRIVER, 108000, 2880, 2928,
 		   3184, 3456, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 39 - 1920x1080i@50Hz */
 	{ DRM_MODE("1920x1080i", DRM_MODE_TYPE_DRIVER, 72000, 1920, 1952,
 		   2120, 2304, 0, 1080, 1126, 1136, 1250, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE),
-	  .vrefresh = 50, },
+	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 40 - 1920x1080i@100Hz */
 	{ DRM_MODE("1920x1080i", DRM_MODE_TYPE_DRIVER, 148500, 1920, 2448,
 		   2492, 2640, 0, 1080, 1084, 1094, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC |
 			DRM_MODE_FLAG_INTERLACE),
-	  .vrefresh = 100, },
+	  .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 41 - 1280x720@100Hz */
 	{ DRM_MODE("1280x720", DRM_MODE_TYPE_DRIVER, 148500, 1280, 1720,
 		   1760, 1980, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
-	  .vrefresh = 100, },
+	  .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 42 - 720x576@100Hz */
 	{ DRM_MODE("720x576", DRM_MODE_TYPE_DRIVER, 54000, 720, 732,
 		   796, 864, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 100, },
+	  .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 43 - 720x576@100Hz */
 	{ DRM_MODE("720x576", DRM_MODE_TYPE_DRIVER, 54000, 720, 732,
 		   796, 864, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 100, },
+	  .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,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 100, },
+	  .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,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 100, },
+	  .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 46 - 1920x1080i@120Hz */
 	{ DRM_MODE("1920x1080i", DRM_MODE_TYPE_DRIVER, 148500, 1920, 2008,
 		   2052, 2200, 0, 1080, 1084, 1094, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC |
 			DRM_MODE_FLAG_INTERLACE),
-	  .vrefresh = 120, },
+	  .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 47 - 1280x720@120Hz */
 	{ DRM_MODE("1280x720", DRM_MODE_TYPE_DRIVER, 148500, 1280, 1390,
 		   1430, 1650, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
-	  .vrefresh = 120, },
+	  .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 48 - 720x480@120Hz */
 	{ DRM_MODE("720x480", DRM_MODE_TYPE_DRIVER, 54000, 720, 736,
 		   798, 858, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 120, },
+	  .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 49 - 720x480@120Hz */
 	{ DRM_MODE("720x480", DRM_MODE_TYPE_DRIVER, 54000, 720, 736,
 		   798, 858, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 120, },
+	  .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,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 120, },
+	  .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,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 120, },
+	  .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 52 - 720x576@200Hz */
 	{ DRM_MODE("720x576", DRM_MODE_TYPE_DRIVER, 108000, 720, 732,
 		   796, 864, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 200, },
+	  .vrefresh = 200, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 53 - 720x576@200Hz */
 	{ DRM_MODE("720x576", DRM_MODE_TYPE_DRIVER, 108000, 720, 732,
 		   796, 864, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 200, },
+	  .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,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 200, },
+	  .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,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 200, },
+	  .vrefresh = 200, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 56 - 720x480@240Hz */
 	{ DRM_MODE("720x480", DRM_MODE_TYPE_DRIVER, 108000, 720, 736,
 		   798, 858, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 240, },
+	  .vrefresh = 240, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
 	/* 57 - 720x480@240Hz */
 	{ DRM_MODE("720x480", DRM_MODE_TYPE_DRIVER, 108000, 720, 736,
 		   798, 858, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
-	  .vrefresh = 240, },
+	  .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,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 240, },
+	  .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,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
 			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
-	  .vrefresh = 240, },
+	  .vrefresh = 240, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 60 - 1280x720@24Hz */
 	{ DRM_MODE("1280x720", DRM_MODE_TYPE_DRIVER, 59400, 1280, 3040,
 		   3080, 3300, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
-	  .vrefresh = 24, },
+	  .vrefresh = 24, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 61 - 1280x720@25Hz */
 	{ DRM_MODE("1280x720", DRM_MODE_TYPE_DRIVER, 74250, 1280, 3700,
 		   3740, 3960, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
-	  .vrefresh = 25, },
+	  .vrefresh = 25, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 62 - 1280x720@30Hz */
 	{ DRM_MODE("1280x720", DRM_MODE_TYPE_DRIVER, 74250, 1280, 3040,
 		   3080, 3300, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
-	  .vrefresh = 30, },
+	  .vrefresh = 30, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 63 - 1920x1080@120Hz */
 	{ DRM_MODE("1920x1080", DRM_MODE_TYPE_DRIVER, 297000, 1920, 2008,
 		   2052, 2200, 0, 1080, 1084, 1089, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
-	 .vrefresh = 120, },
+	 .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 	/* 64 - 1920x1080@100Hz */
 	{ DRM_MODE("1920x1080", DRM_MODE_TYPE_DRIVER, 297000, 1920, 2448,
 		   2492, 2640, 0, 1080, 1084, 1094, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
-	 .vrefresh = 100, },
+	 .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
 };
 
 /*
@@ -2562,25 +2562,40 @@ add_alternate_cea_modes(struct drm_connector *connector, struct edid *edid)
 	return modes;
 }
 
-static int
-do_cea_modes(struct drm_connector *connector, const u8 *db, u8 len)
+static struct drm_display_mode *
+drm_display_mode_from_vic_index(struct drm_connector *connector,
+				const u8 *video_db, u8 video_len,
+				u8 video_index)
 {
 	struct drm_device *dev = connector->dev;
-	const u8 *mode;
+	struct drm_display_mode *newmode;
 	u8 cea_mode;
-	int modes = 0;
 
-	for (mode = db; mode < db + len; mode++) {
-		cea_mode = (*mode & 127) - 1; /* CEA modes are numbered 1..127 */
-		if (cea_mode < ARRAY_SIZE(edid_cea_modes)) {
-			struct drm_display_mode *newmode;
-			newmode = drm_mode_duplicate(dev,
-						     &edid_cea_modes[cea_mode]);
-			if (newmode) {
-				newmode->vrefresh = 0;
-				drm_mode_probed_add(connector, newmode);
-				modes++;
-			}
+	if (video_db == NULL || video_index >= video_len)
+		return NULL;
+
+	/* CEA modes are numbered 1..127 */
+	cea_mode = (video_db[video_index] & 127) - 1;
+	if (cea_mode >= ARRAY_SIZE(edid_cea_modes))
+		return NULL;
+
+	newmode = drm_mode_duplicate(dev, &edid_cea_modes[cea_mode]);
+	newmode->vrefresh = 0;
+
+	return newmode;
+}
+
+static int
+do_cea_modes(struct drm_connector *connector, const u8 *db, u8 len)
+{
+	int i, modes = 0;
+
+	for (i = 0; i < len; i++) {
+		struct drm_display_mode *mode;
+		mode = drm_display_mode_from_vic_index(connector, db, len, i);
+		if (mode) {
+			drm_mode_probed_add(connector, mode);
+			modes++;
 		}
 	}
 
@@ -2674,21 +2689,13 @@ static int add_hdmi_mode(struct drm_connector *connector, u8 vic)
 static int add_3d_struct_modes(struct drm_connector *connector, u16 structure,
 			       const u8 *video_db, u8 video_len, u8 video_index)
 {
-	struct drm_device *dev = connector->dev;
 	struct drm_display_mode *newmode;
 	int modes = 0;
-	u8 cea_mode;
-
-	if (video_db == NULL || video_index >= video_len)
-		return 0;
-
-	/* CEA modes are numbered 1..127 */
-	cea_mode = (video_db[video_index] & 127) - 1;
-	if (cea_mode >= ARRAY_SIZE(edid_cea_modes))
-		return 0;
 
 	if (structure & (1 << 0)) {
-		newmode = drm_mode_duplicate(dev, &edid_cea_modes[cea_mode]);
+		newmode = drm_display_mode_from_vic_index(connector, video_db,
+							  video_len,
+							  video_index);
 		if (newmode) {
 			newmode->flags |= DRM_MODE_FLAG_3D_FRAME_PACKING;
 			drm_mode_probed_add(connector, newmode);
@@ -2696,7 +2703,9 @@ static int add_3d_struct_modes(struct drm_connector *connector, u16 structure,
 		}
 	}
 	if (structure & (1 << 6)) {
-		newmode = drm_mode_duplicate(dev, &edid_cea_modes[cea_mode]);
+		newmode = drm_display_mode_from_vic_index(connector, video_db,
+							  video_len,
+							  video_index);
 		if (newmode) {
 			newmode->flags |= DRM_MODE_FLAG_3D_TOP_AND_BOTTOM;
 			drm_mode_probed_add(connector, newmode);
@@ -2704,7 +2713,9 @@ static int add_3d_struct_modes(struct drm_connector *connector, u16 structure,
 		}
 	}
 	if (structure & (1 << 8)) {
-		newmode = drm_mode_duplicate(dev, &edid_cea_modes[cea_mode]);
+		newmode = drm_display_mode_from_vic_index(connector, video_db,
+							  video_len,
+							  video_index);
 		if (newmode) {
 			newmode->flags |= DRM_MODE_FLAG_3D_SIDE_BY_SIDE_HALF;
 			drm_mode_probed_add(connector, newmode);
@@ -2728,7 +2739,7 @@ static int
 do_hdmi_vsdb_modes(struct drm_connector *connector, const u8 *db, u8 len,
 		   const u8 *video_db, u8 video_len)
 {
-	int modes = 0, offset = 0, i, multi_present = 0;
+	int modes = 0, offset = 0, i, multi_present = 0, multi_len;
 	u8 vic_len, hdmi_3d_len = 0;
 	u16 mask;
 	u16 structure_all;
@@ -2774,32 +2785,84 @@ do_hdmi_vsdb_modes(struct drm_connector *connector, const u8 *db, u8 len,
 	}
 	offset += 1 + vic_len;
 
-	if (!(multi_present == 1 || multi_present == 2))
-		goto out;
+	if (multi_present == 1)
+		multi_len = 2;
+	else if (multi_present == 2)
+		multi_len = 4;
+	else
+		multi_len = 0;
 
-	if ((multi_present == 1 && len < (9 + offset)) ||
-	    (multi_present == 2 && len < (11 + offset)))
+	if (len < (8 + offset + hdmi_3d_len - 1))
 		goto out;
 
-	if ((multi_present == 1 && hdmi_3d_len < 2) ||
-	    (multi_present == 2 && hdmi_3d_len < 4))
+	if (hdmi_3d_len < multi_len)
 		goto out;
 
-	/* 3D_Structure_ALL */
-	structure_all = (db[8 + offset] << 8) | db[9 + offset];
+	if (multi_present == 1 || multi_present == 2) {
+		/* 3D_Structure_ALL */
+		structure_all = (db[8 + offset] << 8) | db[9 + offset];
 
-	/* check if 3D_MASK is present */
-	if (multi_present == 2)
-		mask = (db[10 + offset] << 8) | db[11 + offset];
-	else
-		mask = 0xffff;
-
-	for (i = 0; i < 16; i++) {
-		if (mask & (1 << i))
-			modes += add_3d_struct_modes(connector,
-						     structure_all,
-						     video_db,
-						     video_len, i);
+		/* check if 3D_MASK is present */
+		if (multi_present == 2)
+			mask = (db[10 + offset] << 8) | db[11 + offset];
+		else
+			mask = 0xffff;
+
+		for (i = 0; i < 16; i++) {
+			if (mask & (1 << i))
+				modes += add_3d_struct_modes(connector,
+						structure_all,
+						video_db,
+						video_len, i);
+		}
+	}
+
+	offset += multi_len;
+
+	for (i = 0; i < (hdmi_3d_len - multi_len); i++) {
+		int vic_index;
+		struct drm_display_mode *newmode = NULL;
+		unsigned int newflag = 0;
+		bool detail_present;
+
+		detail_present = ((db[8 + offset + i] & 0x0f) > 7);
+
+		if (detail_present && (i + 1 == hdmi_3d_len - multi_len))
+			break;
+
+		/* 2D_VIC_order_X */
+		vic_index = db[8 + offset + i] >> 4;
+
+		/* 3D_Structure_X */
+		switch (db[8 + offset + i] & 0x0f) {
+		case 0:
+			newflag = DRM_MODE_FLAG_3D_FRAME_PACKING;
+			break;
+		case 6:
+			newflag = DRM_MODE_FLAG_3D_TOP_AND_BOTTOM;
+			break;
+		case 8:
+			/* 3D_Detail_X */
+			if ((db[9 + offset + i] >> 4) == 1)
+				newflag = DRM_MODE_FLAG_3D_SIDE_BY_SIDE_HALF;
+			break;
+		}
+
+		if (newflag != 0) {
+			newmode = drm_display_mode_from_vic_index(connector,
+								  video_db,
+								  video_len,
+								  vic_index);
+
+			if (newmode) {
+				newmode->flags |= newflag;
+				drm_mode_probed_add(connector, newmode);
+				modes++;
+			}
+		}
+
+		if (detail_present)
+			i++;
 	}
 
 out:
diff --git a/drivers/gpu/drm/drm_edid_load.c b/drivers/gpu/drm/drm_edid_load.c
index 9081172ef057..1b4c7a5442c5 100644
--- a/drivers/gpu/drm/drm_edid_load.c
+++ b/drivers/gpu/drm/drm_edid_load.c
@@ -141,7 +141,7 @@ static int edid_size(const u8 *edid, int data_size)
 	return (edid[0x7e] + 1) * EDID_LENGTH;
 }
 
-static u8 *edid_load(struct drm_connector *connector, const char *name,
+static void *edid_load(struct drm_connector *connector, const char *name,
 			const char *connector_name)
 {
 	const struct firmware *fw = NULL;
@@ -263,7 +263,7 @@ int drm_load_edid_firmware(struct drm_connector *connector)
 	if (*last == '\n')
 		*last = '\0';
 
-	edid = (struct edid *) edid_load(connector, edidname, connector_name);
+	edid = edid_load(connector, edidname, connector_name);
 	if (IS_ERR_OR_NULL(edid))
 		return 0;
 
diff --git a/drivers/gpu/drm/drm_fb_helper.c b/drivers/gpu/drm/drm_fb_helper.c
index 0a19401aff80..98a03639b413 100644
--- a/drivers/gpu/drm/drm_fb_helper.c
+++ b/drivers/gpu/drm/drm_fb_helper.c
@@ -359,6 +359,11 @@ static bool drm_fb_helper_is_bound(struct drm_fb_helper *fb_helper)
 	struct drm_crtc *crtc;
 	int bound = 0, crtcs_bound = 0;
 
+	/* Sometimes user space wants everything disabled, so don't steal the
+	 * display if there's a master. */
+	if (dev->primary->master)
+		return false;
+
 	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
 		if (crtc->fb)
 			crtcs_bound++;
@@ -368,6 +373,7 @@ static bool drm_fb_helper_is_bound(struct drm_fb_helper *fb_helper)
 
 	if (bound < crtcs_bound)
 		return false;
+
 	return true;
 }
 
diff --git a/drivers/gpu/drm/drm_fops.c b/drivers/gpu/drm/drm_fops.c
index c5b929c3f77a..7f2af9aca038 100644
--- a/drivers/gpu/drm/drm_fops.c
+++ b/drivers/gpu/drm/drm_fops.c
@@ -232,7 +232,6 @@ static int drm_open_helper(struct inode *inode, struct file *filp,
 		goto out_put_pid;
 	}
 
-	priv->ioctl_count = 0;
 	/* for compatibility root is always authenticated */
 	priv->always_authenticated = capable(CAP_SYS_ADMIN);
 	priv->authenticated = priv->always_authenticated;
@@ -392,9 +391,6 @@ static void drm_legacy_dev_reinit(struct drm_device *dev)
 	if (drm_core_check_feature(dev, DRIVER_MODESET))
 		return;
 
-	atomic_set(&dev->ioctl_count, 0);
-	atomic_set(&dev->vma_count, 0);
-
 	dev->sigdata.lock = NULL;
 
 	dev->context_flag = 0;
@@ -578,12 +574,7 @@ int drm_release(struct inode *inode, struct file *filp)
 	 */
 
 	if (!--dev->open_count) {
-		if (atomic_read(&dev->ioctl_count)) {
-			DRM_ERROR("Device busy: %d\n",
-				  atomic_read(&dev->ioctl_count));
-			retcode = -EBUSY;
-		} else
-			retcode = drm_lastclose(dev);
+		retcode = drm_lastclose(dev);
 		if (drm_device_is_unplugged(dev))
 			drm_put_dev(dev);
 	}
diff --git a/drivers/gpu/drm/drm_gem.c b/drivers/gpu/drm/drm_gem.c
index 4761adedad2a..5bbad873c798 100644
--- a/drivers/gpu/drm/drm_gem.c
+++ b/drivers/gpu/drm/drm_gem.c
@@ -91,19 +91,19 @@
 int
 drm_gem_init(struct drm_device *dev)
 {
-	struct drm_gem_mm *mm;
+	struct drm_vma_offset_manager *vma_offset_manager;
 
 	mutex_init(&dev->object_name_lock);
 	idr_init(&dev->object_name_idr);
 
-	mm = kzalloc(sizeof(struct drm_gem_mm), GFP_KERNEL);
-	if (!mm) {
+	vma_offset_manager = kzalloc(sizeof(*vma_offset_manager), GFP_KERNEL);
+	if (!vma_offset_manager) {
 		DRM_ERROR("out of memory\n");
 		return -ENOMEM;
 	}
 
-	dev->mm_private = mm;
-	drm_vma_offset_manager_init(&mm->vma_manager,
+	dev->vma_offset_manager = vma_offset_manager;
+	drm_vma_offset_manager_init(vma_offset_manager,
 				    DRM_FILE_PAGE_OFFSET_START,
 				    DRM_FILE_PAGE_OFFSET_SIZE);
 
@@ -113,11 +113,10 @@ drm_gem_init(struct drm_device *dev)
 void
 drm_gem_destroy(struct drm_device *dev)
 {
-	struct drm_gem_mm *mm = dev->mm_private;
 
-	drm_vma_offset_manager_destroy(&mm->vma_manager);
-	kfree(mm);
-	dev->mm_private = NULL;
+	drm_vma_offset_manager_destroy(dev->vma_offset_manager);
+	kfree(dev->vma_offset_manager);
+	dev->vma_offset_manager = NULL;
 }
 
 /**
@@ -129,11 +128,12 @@ int drm_gem_object_init(struct drm_device *dev,
 {
 	struct file *filp;
 
+	drm_gem_private_object_init(dev, obj, size);
+
 	filp = shmem_file_setup("drm mm object", size, VM_NORESERVE);
 	if (IS_ERR(filp))
 		return PTR_ERR(filp);
 
-	drm_gem_private_object_init(dev, obj, size);
 	obj->filp = filp;
 
 	return 0;
@@ -175,11 +175,6 @@ drm_gem_remove_prime_handles(struct drm_gem_object *obj, struct drm_file *filp)
 	mutex_unlock(&filp->prime.lock);
 }
 
-static void drm_gem_object_ref_bug(struct kref *list_kref)
-{
-	BUG();
-}
-
 /**
  * Called after the last handle to the object has been closed
  *
@@ -195,13 +190,6 @@ static void drm_gem_object_handle_free(struct drm_gem_object *obj)
 	if (obj->name) {
 		idr_remove(&dev->object_name_idr, obj->name);
 		obj->name = 0;
-		/*
-		 * The object name held a reference to this object, drop
-		 * that now.
-		*
-		* This cannot be the last reference, since the handle holds one too.
-		 */
-		kref_put(&obj->refcount, drm_gem_object_ref_bug);
 	}
 }
 
@@ -374,9 +362,8 @@ void
 drm_gem_free_mmap_offset(struct drm_gem_object *obj)
 {
 	struct drm_device *dev = obj->dev;
-	struct drm_gem_mm *mm = dev->mm_private;
 
-	drm_vma_offset_remove(&mm->vma_manager, &obj->vma_node);
+	drm_vma_offset_remove(dev->vma_offset_manager, &obj->vma_node);
 }
 EXPORT_SYMBOL(drm_gem_free_mmap_offset);
 
@@ -398,9 +385,8 @@ int
 drm_gem_create_mmap_offset_size(struct drm_gem_object *obj, size_t size)
 {
 	struct drm_device *dev = obj->dev;
-	struct drm_gem_mm *mm = dev->mm_private;
 
-	return drm_vma_offset_add(&mm->vma_manager, &obj->vma_node,
+	return drm_vma_offset_add(dev->vma_offset_manager, &obj->vma_node,
 				  size / PAGE_SIZE);
 }
 EXPORT_SYMBOL(drm_gem_create_mmap_offset_size);
@@ -602,9 +588,6 @@ drm_gem_flink_ioctl(struct drm_device *dev, void *data,
 			goto err;
 
 		obj->name = ret;
-
-		/* Allocate a reference for the name table.  */
-		drm_gem_object_reference(obj);
 	}
 
 	args->name = (uint64_t) obj->name;
@@ -833,7 +816,6 @@ int drm_gem_mmap(struct file *filp, struct vm_area_struct *vma)
 {
 	struct drm_file *priv = filp->private_data;
 	struct drm_device *dev = priv->minor->dev;
-	struct drm_gem_mm *mm = dev->mm_private;
 	struct drm_gem_object *obj;
 	struct drm_vma_offset_node *node;
 	int ret = 0;
@@ -843,7 +825,8 @@ int drm_gem_mmap(struct file *filp, struct vm_area_struct *vma)
 
 	mutex_lock(&dev->struct_mutex);
 
-	node = drm_vma_offset_exact_lookup(&mm->vma_manager, vma->vm_pgoff,
+	node = drm_vma_offset_exact_lookup(dev->vma_offset_manager,
+					   vma->vm_pgoff,
 					   vma_pages(vma));
 	if (!node) {
 		mutex_unlock(&dev->struct_mutex);
diff --git a/drivers/gpu/drm/drm_info.c b/drivers/gpu/drm/drm_info.c
index 7d5a152eeb02..7473035dd28b 100644
--- a/drivers/gpu/drm/drm_info.c
+++ b/drivers/gpu/drm/drm_info.c
@@ -186,14 +186,14 @@ int drm_clients_info(struct seq_file *m, void *data)
 	struct drm_file *priv;
 
 	mutex_lock(&dev->struct_mutex);
-	seq_printf(m, "a dev	pid    uid	magic	  ioctls\n\n");
+	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 %10lu\n",
+		seq_printf(m, "%c %3d %5d %5d %10u\n",
 			   priv->authenticated ? 'y' : 'n',
 			   priv->minor->index,
 			   pid_vnr(priv->pid),
 			   from_kuid_munged(seq_user_ns(m), priv->uid),
-			   priv->magic, priv->ioctl_count);
+			   priv->magic);
 	}
 	mutex_unlock(&dev->struct_mutex);
 	return 0;
@@ -234,14 +234,18 @@ int drm_vma_info(struct seq_file *m, void *data)
 	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);
-	seq_printf(m, "vma use count: %d, high_memory = %pK, 0x%pK\n",
-		   atomic_read(&dev->vma_count),
-		   high_memory, (void *)(unsigned long)virt_to_phys(high_memory));
+	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;
diff --git a/drivers/gpu/drm/drm_irq.c b/drivers/gpu/drm/drm_irq.c
index 64c34d5876ff..c2676b5908d9 100644
--- a/drivers/gpu/drm/drm_irq.c
+++ b/drivers/gpu/drm/drm_irq.c
@@ -368,7 +368,7 @@ 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++) {
-			DRM_WAKEUP(&dev->vblank[i].queue);
+			wake_up(&dev->vblank[i].queue);
 			dev->vblank[i].enabled = false;
 			dev->vblank[i].last =
 				dev->driver->get_vblank_counter(dev, i);
@@ -436,45 +436,41 @@ int drm_control(struct drm_device *dev, void *data,
 }
 
 /**
- * drm_calc_timestamping_constants - Calculate and
- * store various constants which are later needed by
- * vblank and swap-completion timestamping, e.g, by
- * drm_calc_vbltimestamp_from_scanoutpos().
- * They are derived from crtc's true scanout timing,
- * so they take things like panel scaling or other
- * adjustments into account.
+ * drm_calc_timestamping_constants - Calculate vblank timestamp constants
  *
  * @crtc drm_crtc whose timestamp constants should be updated.
+ * @mode display mode containing the scanout timings
  *
+ * Calculate and store various constants which are later
+ * needed by vblank and swap-completion timestamping, e.g,
+ * by drm_calc_vbltimestamp_from_scanoutpos(). They are
+ * derived from crtc's true scanout timing, so they take
+ * things like panel scaling or other adjustments into account.
  */
-void drm_calc_timestamping_constants(struct drm_crtc *crtc)
+void drm_calc_timestamping_constants(struct drm_crtc *crtc,
+				     const struct drm_display_mode *mode)
 {
-	s64 linedur_ns = 0, pixeldur_ns = 0, framedur_ns = 0;
-	u64 dotclock;
-
-	/* Dot clock in Hz: */
-	dotclock = (u64) crtc->hwmode.clock * 1000;
-
-	/* Fields of interlaced scanout modes are only half a frame duration.
-	 * Double the dotclock to get half the frame-/line-/pixelduration.
-	 */
-	if (crtc->hwmode.flags & DRM_MODE_FLAG_INTERLACE)
-		dotclock *= 2;
+	int linedur_ns = 0, pixeldur_ns = 0, framedur_ns = 0;
+	int dotclock = mode->crtc_clock;
 
 	/* Valid dotclock? */
 	if (dotclock > 0) {
-		int frame_size;
-		/* Convert scanline length in pixels and video dot clock to
-		 * line duration, frame duration and pixel duration in
-		 * nanoseconds:
+		int frame_size = mode->crtc_htotal * mode->crtc_vtotal;
+
+		/*
+		 * Convert scanline length in pixels and video
+		 * dot clock to line duration, frame duration
+		 * and pixel duration in nanoseconds:
 		 */
-		pixeldur_ns = (s64) div64_u64(1000000000, dotclock);
-		linedur_ns  = (s64) div64_u64(((u64) crtc->hwmode.crtc_htotal *
-					      1000000000), dotclock);
-		frame_size = crtc->hwmode.crtc_htotal *
-				crtc->hwmode.crtc_vtotal;
-		framedur_ns = (s64) div64_u64((u64) frame_size * 1000000000,
-					      dotclock);
+		pixeldur_ns = 1000000 / dotclock;
+		linedur_ns  = div_u64((u64) mode->crtc_htotal * 1000000, dotclock);
+		framedur_ns = div_u64((u64) frame_size * 1000000, dotclock);
+
+		/*
+		 * Fields of interlaced scanout modes are only half a frame duration.
+		 */
+		if (mode->flags & DRM_MODE_FLAG_INTERLACE)
+			framedur_ns /= 2;
 	} else
 		DRM_ERROR("crtc %d: Can't calculate constants, dotclock = 0!\n",
 			  crtc->base.id);
@@ -484,11 +480,11 @@ void drm_calc_timestamping_constants(struct drm_crtc *crtc)
 	crtc->framedur_ns = framedur_ns;
 
 	DRM_DEBUG("crtc %d: hwmode: htotal %d, vtotal %d, vdisplay %d\n",
-		  crtc->base.id, crtc->hwmode.crtc_htotal,
-		  crtc->hwmode.crtc_vtotal, crtc->hwmode.crtc_vdisplay);
+		  crtc->base.id, mode->crtc_htotal,
+		  mode->crtc_vtotal, mode->crtc_vdisplay);
 	DRM_DEBUG("crtc %d: clock %d kHz framedur %d linedur %d, pixeldur %d\n",
-		  crtc->base.id, (int) dotclock/1000, (int) framedur_ns,
-		  (int) linedur_ns, (int) pixeldur_ns);
+		  crtc->base.id, dotclock, framedur_ns,
+		  linedur_ns, pixeldur_ns);
 }
 EXPORT_SYMBOL(drm_calc_timestamping_constants);
 
@@ -521,6 +517,7 @@ EXPORT_SYMBOL(drm_calc_timestamping_constants);
  *         0 = Default.
  *         DRM_CALLED_FROM_VBLIRQ = If function is called from vbl irq handler.
  * @refcrtc: drm_crtc* of crtc which defines scanout timing.
+ * @mode: mode which defines the scanout timings
  *
  * Returns negative value on error, failure or if not supported in current
  * video mode:
@@ -540,14 +537,14 @@ int drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev, int crtc,
 					  int *max_error,
 					  struct timeval *vblank_time,
 					  unsigned flags,
-					  struct drm_crtc *refcrtc)
+					  const struct drm_crtc *refcrtc,
+					  const struct drm_display_mode *mode)
 {
 	ktime_t stime, etime, mono_time_offset;
 	struct timeval tv_etime;
-	struct drm_display_mode *mode;
-	int vbl_status, vtotal, vdisplay;
+	int vbl_status;
 	int vpos, hpos, i;
-	s64 framedur_ns, linedur_ns, pixeldur_ns, delta_ns, duration_ns;
+	int framedur_ns, linedur_ns, pixeldur_ns, delta_ns, duration_ns;
 	bool invbl;
 
 	if (crtc < 0 || crtc >= dev->num_crtcs) {
@@ -561,10 +558,6 @@ int drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev, int crtc,
 		return -EIO;
 	}
 
-	mode = &refcrtc->hwmode;
-	vtotal = mode->crtc_vtotal;
-	vdisplay = mode->crtc_vdisplay;
-
 	/* Durations of frames, lines, pixels in nanoseconds. */
 	framedur_ns = refcrtc->framedur_ns;
 	linedur_ns  = refcrtc->linedur_ns;
@@ -573,7 +566,7 @@ int drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev, int crtc,
 	/* If mode timing undefined, just return as no-op:
 	 * Happens during initial modesetting of a crtc.
 	 */
-	if (vtotal <= 0 || vdisplay <= 0 || framedur_ns == 0) {
+	if (framedur_ns == 0) {
 		DRM_DEBUG("crtc %d: Noop due to uninitialized mode.\n", crtc);
 		return -EAGAIN;
 	}
@@ -590,7 +583,7 @@ int drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev, int crtc,
 		 * Get vertical and horizontal scanout position vpos, hpos,
 		 * and bounding timestamps stime, etime, pre/post query.
 		 */
-		vbl_status = dev->driver->get_scanout_position(dev, crtc, &vpos,
+		vbl_status = dev->driver->get_scanout_position(dev, crtc, flags, &vpos,
 							       &hpos, &stime, &etime);
 
 		/*
@@ -611,18 +604,18 @@ int drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev, int crtc,
 		duration_ns = ktime_to_ns(etime) - ktime_to_ns(stime);
 
 		/* Accept result with <  max_error nsecs timing uncertainty. */
-		if (duration_ns <= (s64) *max_error)
+		if (duration_ns <= *max_error)
 			break;
 	}
 
 	/* Noisy system timing? */
 	if (i == DRM_TIMESTAMP_MAXRETRIES) {
 		DRM_DEBUG("crtc %d: Noisy timestamp %d us > %d us [%d reps].\n",
-			  crtc, (int) duration_ns/1000, *max_error/1000, i);
+			  crtc, duration_ns/1000, *max_error/1000, i);
 	}
 
 	/* Return upper bound of timestamp precision error. */
-	*max_error = (int) duration_ns;
+	*max_error = duration_ns;
 
 	/* Check if in vblank area:
 	 * vpos is >=0 in video scanout area, but negative
@@ -635,25 +628,7 @@ int drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev, int crtc,
 	 * since start of scanout at first display scanline. delta_ns
 	 * can be negative if start of scanout hasn't happened yet.
 	 */
-	delta_ns = (s64) vpos * linedur_ns + (s64) hpos * pixeldur_ns;
-
-	/* Is vpos outside nominal vblank area, but less than
-	 * 1/100 of a frame height away from start of vblank?
-	 * If so, assume this isn't a massively delayed vblank
-	 * interrupt, but a vblank interrupt that fired a few
-	 * microseconds before true start of vblank. Compensate
-	 * by adding a full frame duration to the final timestamp.
-	 * Happens, e.g., on ATI R500, R600.
-	 *
-	 * We only do this if DRM_CALLED_FROM_VBLIRQ.
-	 */
-	if ((flags & DRM_CALLED_FROM_VBLIRQ) && !invbl &&
-	    ((vdisplay - vpos) < vtotal / 100)) {
-		delta_ns = delta_ns - framedur_ns;
-
-		/* Signal this correction as "applied". */
-		vbl_status |= 0x8;
-	}
+	delta_ns = vpos * linedur_ns + hpos * pixeldur_ns;
 
 	if (!drm_timestamp_monotonic)
 		etime = ktime_sub(etime, mono_time_offset);
@@ -673,7 +648,7 @@ int drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev, int crtc,
 		  crtc, (int)vbl_status, hpos, vpos,
 		  (long)tv_etime.tv_sec, (long)tv_etime.tv_usec,
 		  (long)vblank_time->tv_sec, (long)vblank_time->tv_usec,
-		  (int)duration_ns/1000, i);
+		  duration_ns/1000, i);
 
 	vbl_status = DRM_VBLANKTIME_SCANOUTPOS_METHOD;
 	if (invbl)
@@ -960,7 +935,7 @@ void drm_vblank_put(struct drm_device *dev, int crtc)
 	if (atomic_dec_and_test(&dev->vblank[crtc].refcount) &&
 	    (drm_vblank_offdelay > 0))
 		mod_timer(&dev->vblank_disable_timer,
-			  jiffies + ((drm_vblank_offdelay * DRM_HZ)/1000));
+			  jiffies + ((drm_vblank_offdelay * HZ)/1000));
 }
 EXPORT_SYMBOL(drm_vblank_put);
 
@@ -980,7 +955,7 @@ void drm_vblank_off(struct drm_device *dev, int crtc)
 
 	spin_lock_irqsave(&dev->vbl_lock, irqflags);
 	vblank_disable_and_save(dev, crtc);
-	DRM_WAKEUP(&dev->vblank[crtc].queue);
+	wake_up(&dev->vblank[crtc].queue);
 
 	/* Send any queued vblank events, lest the natives grow disquiet */
 	seq = drm_vblank_count_and_time(dev, crtc, &now);
@@ -1244,7 +1219,7 @@ 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 * DRM_HZ,
+	DRM_WAIT_ON(ret, dev->vblank[crtc].queue, 3 * HZ,
 		    (((drm_vblank_count(dev, crtc) -
 		       vblwait->request.sequence) <= (1 << 23)) ||
 		     !dev->irq_enabled));
@@ -1363,7 +1338,7 @@ bool drm_handle_vblank(struct drm_device *dev, int crtc)
 			  crtc, (int) diff_ns);
 	}
 
-	DRM_WAKEUP(&dev->vblank[crtc].queue);
+	wake_up(&dev->vblank[crtc].queue);
 	drm_handle_vblank_events(dev, crtc);
 
 	spin_unlock_irqrestore(&dev->vblank_time_lock, irqflags);
diff --git a/drivers/gpu/drm/drm_memory.c b/drivers/gpu/drm/drm_memory.c
index 64e44fad8ae8..00c67c0f2381 100644
--- a/drivers/gpu/drm/drm_memory.c
+++ b/drivers/gpu/drm/drm_memory.c
@@ -82,19 +82,19 @@ static void *agp_remap(unsigned long offset, unsigned long size,
 }
 
 /** Wrapper around agp_free_memory() */
-void drm_free_agp(DRM_AGP_MEM * handle, int pages)
+void drm_free_agp(struct agp_memory * handle, int pages)
 {
 	agp_free_memory(handle);
 }
 
 /** Wrapper around agp_bind_memory() */
-int drm_bind_agp(DRM_AGP_MEM * handle, unsigned int start)
+int drm_bind_agp(struct agp_memory * handle, unsigned int start)
 {
 	return agp_bind_memory(handle, start);
 }
 
 /** Wrapper around agp_unbind_memory() */
-int drm_unbind_agp(DRM_AGP_MEM * handle)
+int drm_unbind_agp(struct agp_memory * handle)
 {
 	return agp_unbind_memory(handle);
 }
@@ -110,8 +110,7 @@ static inline void *agp_remap(unsigned long offset, unsigned long size,
 
 void drm_core_ioremap(struct drm_local_map *map, struct drm_device *dev)
 {
-	if (drm_core_has_AGP(dev) &&
-	    dev->agp && dev->agp->cant_use_aperture && map->type == _DRM_AGP)
+	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);
@@ -120,8 +119,7 @@ EXPORT_SYMBOL(drm_core_ioremap);
 
 void drm_core_ioremap_wc(struct drm_local_map *map, struct drm_device *dev)
 {
-	if (drm_core_has_AGP(dev) &&
-	    dev->agp && dev->agp->cant_use_aperture && map->type == _DRM_AGP)
+	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);
@@ -133,8 +131,7 @@ void drm_core_ioremapfree(struct drm_local_map *map, struct drm_device *dev)
 	if (!map->handle || !map->size)
 		return;
 
-	if (drm_core_has_AGP(dev) &&
-	    dev->agp && dev->agp->cant_use_aperture && map->type == _DRM_AGP)
+	if (dev->agp && dev->agp->cant_use_aperture && map->type == _DRM_AGP)
 		vunmap(map->handle);
 	else
 		iounmap(map->handle);
diff --git a/drivers/gpu/drm/drm_mipi_dsi.c b/drivers/gpu/drm/drm_mipi_dsi.c
new file mode 100644
index 000000000000..b155ee2ffa17
--- /dev/null
+++ b/drivers/gpu/drm/drm_mipi_dsi.c
@@ -0,0 +1,315 @@
+/*
+ * MIPI DSI Bus
+ *
+ * Copyright (C) 2012-2013, Samsung Electronics, Co., Ltd.
+ * Andrzej Hajda <a.hajda@samsung.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 the rights to use, copy, modify, merge, publish,
+ * distribute, sub license, and/or sell copies of the Software, and to
+ * permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the
+ * next paragraph) shall be included in all copies or substantial portions
+ * of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
+ * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
+ * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
+ * USE OR OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+#include <drm/drm_mipi_dsi.h>
+
+#include <linux/device.h>
+#include <linux/module.h>
+#include <linux/of_device.h>
+#include <linux/pm_runtime.h>
+#include <linux/slab.h>
+
+#include <video/mipi_display.h>
+
+static int mipi_dsi_device_match(struct device *dev, struct device_driver *drv)
+{
+	return of_driver_match_device(dev, drv);
+}
+
+static const struct dev_pm_ops mipi_dsi_device_pm_ops = {
+	.runtime_suspend = pm_generic_runtime_suspend,
+	.runtime_resume = pm_generic_runtime_resume,
+	.suspend = pm_generic_suspend,
+	.resume = pm_generic_resume,
+	.freeze = pm_generic_freeze,
+	.thaw = pm_generic_thaw,
+	.poweroff = pm_generic_poweroff,
+	.restore = pm_generic_restore,
+};
+
+static struct bus_type mipi_dsi_bus_type = {
+	.name = "mipi-dsi",
+	.match = mipi_dsi_device_match,
+	.pm = &mipi_dsi_device_pm_ops,
+};
+
+static void mipi_dsi_dev_release(struct device *dev)
+{
+	struct mipi_dsi_device *dsi = to_mipi_dsi_device(dev);
+
+	of_node_put(dev->of_node);
+	kfree(dsi);
+}
+
+static const struct device_type mipi_dsi_device_type = {
+	.release = mipi_dsi_dev_release,
+};
+
+static struct mipi_dsi_device *mipi_dsi_device_alloc(struct mipi_dsi_host *host)
+{
+	struct mipi_dsi_device *dsi;
+
+	dsi = kzalloc(sizeof(*dsi), GFP_KERNEL);
+	if (!dsi)
+		return ERR_PTR(-ENOMEM);
+
+	dsi->host = host;
+	dsi->dev.bus = &mipi_dsi_bus_type;
+	dsi->dev.parent = host->dev;
+	dsi->dev.type = &mipi_dsi_device_type;
+
+	device_initialize(&dsi->dev);
+
+	return dsi;
+}
+
+static int mipi_dsi_device_add(struct mipi_dsi_device *dsi)
+{
+	struct mipi_dsi_host *host = dsi->host;
+
+	dev_set_name(&dsi->dev, "%s.%d", dev_name(host->dev),  dsi->channel);
+
+	return device_add(&dsi->dev);
+}
+
+static struct mipi_dsi_device *
+of_mipi_dsi_device_add(struct mipi_dsi_host *host, struct device_node *node)
+{
+	struct mipi_dsi_device *dsi;
+	struct device *dev = host->dev;
+	int ret;
+	u32 reg;
+
+	ret = of_property_read_u32(node, "reg", &reg);
+	if (ret) {
+		dev_err(dev, "device node %s has no valid reg property: %d\n",
+			node->full_name, ret);
+		return ERR_PTR(-EINVAL);
+	}
+
+	if (reg > 3) {
+		dev_err(dev, "device node %s has invalid reg property: %u\n",
+			node->full_name, reg);
+		return ERR_PTR(-EINVAL);
+	}
+
+	dsi = mipi_dsi_device_alloc(host);
+	if (IS_ERR(dsi)) {
+		dev_err(dev, "failed to allocate DSI device %s: %ld\n",
+			node->full_name, PTR_ERR(dsi));
+		return dsi;
+	}
+
+	dsi->dev.of_node = of_node_get(node);
+	dsi->channel = reg;
+
+	ret = mipi_dsi_device_add(dsi);
+	if (ret) {
+		dev_err(dev, "failed to add DSI device %s: %d\n",
+			node->full_name, ret);
+		kfree(dsi);
+		return ERR_PTR(ret);
+	}
+
+	return dsi;
+}
+
+int mipi_dsi_host_register(struct mipi_dsi_host *host)
+{
+	struct device_node *node;
+
+	for_each_available_child_of_node(host->dev->of_node, node)
+		of_mipi_dsi_device_add(host, node);
+
+	return 0;
+}
+EXPORT_SYMBOL(mipi_dsi_host_register);
+
+static int mipi_dsi_remove_device_fn(struct device *dev, void *priv)
+{
+	struct mipi_dsi_device *dsi = to_mipi_dsi_device(dev);
+
+	device_unregister(&dsi->dev);
+
+	return 0;
+}
+
+void mipi_dsi_host_unregister(struct mipi_dsi_host *host)
+{
+	device_for_each_child(host->dev, NULL, mipi_dsi_remove_device_fn);
+}
+EXPORT_SYMBOL(mipi_dsi_host_unregister);
+
+/**
+ * mipi_dsi_attach - attach a DSI device to its DSI host
+ * @dsi: DSI peripheral
+ */
+int mipi_dsi_attach(struct mipi_dsi_device *dsi)
+{
+	const struct mipi_dsi_host_ops *ops = dsi->host->ops;
+
+	if (!ops || !ops->attach)
+		return -ENOSYS;
+
+	return ops->attach(dsi->host, dsi);
+}
+EXPORT_SYMBOL(mipi_dsi_attach);
+
+/**
+ * mipi_dsi_detach - detach a DSI device from its DSI host
+ * @dsi: DSI peripheral
+ */
+int mipi_dsi_detach(struct mipi_dsi_device *dsi)
+{
+	const struct mipi_dsi_host_ops *ops = dsi->host->ops;
+
+	if (!ops || !ops->detach)
+		return -ENOSYS;
+
+	return ops->detach(dsi->host, dsi);
+}
+EXPORT_SYMBOL(mipi_dsi_detach);
+
+/**
+ * mipi_dsi_dcs_write - send DCS write command
+ * @dsi: DSI device
+ * @channel: virtual channel
+ * @data: pointer to the command followed by parameters
+ * @len: length of @data
+ */
+int mipi_dsi_dcs_write(struct mipi_dsi_device *dsi, unsigned int channel,
+		       const void *data, size_t len)
+{
+	const struct mipi_dsi_host_ops *ops = dsi->host->ops;
+	struct mipi_dsi_msg msg = {
+		.channel = channel,
+		.tx_buf = data,
+		.tx_len = len
+	};
+
+	if (!ops || !ops->transfer)
+		return -ENOSYS;
+
+	switch (len) {
+	case 0:
+		return -EINVAL;
+	case 1:
+		msg.type = MIPI_DSI_DCS_SHORT_WRITE;
+		break;
+	case 2:
+		msg.type = MIPI_DSI_DCS_SHORT_WRITE_PARAM;
+		break;
+	default:
+		msg.type = MIPI_DSI_DCS_LONG_WRITE;
+		break;
+	}
+
+	return ops->transfer(dsi->host, &msg);
+}
+EXPORT_SYMBOL(mipi_dsi_dcs_write);
+
+/**
+ * mipi_dsi_dcs_read - send DCS read request command
+ * @dsi: DSI device
+ * @channel: virtual channel
+ * @cmd: DCS read command
+ * @data: pointer to read buffer
+ * @len: length of @data
+ *
+ * Function returns number of read bytes or error code.
+ */
+ssize_t mipi_dsi_dcs_read(struct mipi_dsi_device *dsi, unsigned int channel,
+			  u8 cmd, void *data, size_t len)
+{
+	const struct mipi_dsi_host_ops *ops = dsi->host->ops;
+	struct mipi_dsi_msg msg = {
+		.channel = channel,
+		.type = MIPI_DSI_DCS_READ,
+		.tx_buf = &cmd,
+		.tx_len = 1,
+		.rx_buf = data,
+		.rx_len = len
+	};
+
+	if (!ops || !ops->transfer)
+		return -ENOSYS;
+
+	return ops->transfer(dsi->host, &msg);
+}
+EXPORT_SYMBOL(mipi_dsi_dcs_read);
+
+static int mipi_dsi_drv_probe(struct device *dev)
+{
+	struct mipi_dsi_driver *drv = to_mipi_dsi_driver(dev->driver);
+	struct mipi_dsi_device *dsi = to_mipi_dsi_device(dev);
+
+	return drv->probe(dsi);
+}
+
+static int mipi_dsi_drv_remove(struct device *dev)
+{
+	struct mipi_dsi_driver *drv = to_mipi_dsi_driver(dev->driver);
+	struct mipi_dsi_device *dsi = to_mipi_dsi_device(dev);
+
+	return drv->remove(dsi);
+}
+
+/**
+ * mipi_dsi_driver_register - register a driver for DSI devices
+ * @drv: DSI driver structure
+ */
+int mipi_dsi_driver_register(struct mipi_dsi_driver *drv)
+{
+	drv->driver.bus = &mipi_dsi_bus_type;
+	if (drv->probe)
+		drv->driver.probe = mipi_dsi_drv_probe;
+	if (drv->remove)
+		drv->driver.remove = mipi_dsi_drv_remove;
+
+	return driver_register(&drv->driver);
+}
+EXPORT_SYMBOL(mipi_dsi_driver_register);
+
+/**
+ * mipi_dsi_driver_unregister - unregister a driver for DSI devices
+ * @drv: DSI driver structure
+ */
+void mipi_dsi_driver_unregister(struct mipi_dsi_driver *drv)
+{
+	driver_unregister(&drv->driver);
+}
+EXPORT_SYMBOL(mipi_dsi_driver_unregister);
+
+static int __init mipi_dsi_bus_init(void)
+{
+	return bus_register(&mipi_dsi_bus_type);
+}
+postcore_initcall(mipi_dsi_bus_init);
+
+MODULE_AUTHOR("Andrzej Hajda <a.hajda@samsung.com>");
+MODULE_DESCRIPTION("MIPI DSI Bus");
+MODULE_LICENSE("GPL and additional rights");
diff --git a/drivers/gpu/drm/drm_panel.c b/drivers/gpu/drm/drm_panel.c
new file mode 100644
index 000000000000..2ef988e037b7
--- /dev/null
+++ b/drivers/gpu/drm/drm_panel.c
@@ -0,0 +1,100 @@
+/*
+ * Copyright (C) 2013, NVIDIA Corporation.  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 AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ */
+
+#include <linux/err.h>
+#include <linux/module.h>
+
+#include <drm/drm_crtc.h>
+#include <drm/drm_panel.h>
+
+static DEFINE_MUTEX(panel_lock);
+static LIST_HEAD(panel_list);
+
+void drm_panel_init(struct drm_panel *panel)
+{
+	INIT_LIST_HEAD(&panel->list);
+}
+EXPORT_SYMBOL(drm_panel_init);
+
+int drm_panel_add(struct drm_panel *panel)
+{
+	mutex_lock(&panel_lock);
+	list_add_tail(&panel->list, &panel_list);
+	mutex_unlock(&panel_lock);
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_panel_add);
+
+void drm_panel_remove(struct drm_panel *panel)
+{
+	mutex_lock(&panel_lock);
+	list_del_init(&panel->list);
+	mutex_unlock(&panel_lock);
+}
+EXPORT_SYMBOL(drm_panel_remove);
+
+int drm_panel_attach(struct drm_panel *panel, struct drm_connector *connector)
+{
+	if (panel->connector)
+		return -EBUSY;
+
+	panel->connector = connector;
+	panel->drm = connector->dev;
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_panel_attach);
+
+int drm_panel_detach(struct drm_panel *panel)
+{
+	panel->connector = NULL;
+	panel->drm = NULL;
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_panel_detach);
+
+#ifdef CONFIG_OF
+struct drm_panel *of_drm_find_panel(struct device_node *np)
+{
+	struct drm_panel *panel;
+
+	mutex_lock(&panel_lock);
+
+	list_for_each_entry(panel, &panel_list, list) {
+		if (panel->dev->of_node == np) {
+			mutex_unlock(&panel_lock);
+			return panel;
+		}
+	}
+
+	mutex_unlock(&panel_lock);
+	return NULL;
+}
+EXPORT_SYMBOL(of_drm_find_panel);
+#endif
+
+MODULE_AUTHOR("Thierry Reding <treding@nvidia.com>");
+MODULE_DESCRIPTION("DRM panel infrastructure");
+MODULE_LICENSE("GPL and additional rights");
diff --git a/drivers/gpu/drm/drm_pci.c b/drivers/gpu/drm/drm_pci.c
index 02679793c9e2..5736aaa7e86c 100644
--- a/drivers/gpu/drm/drm_pci.c
+++ b/drivers/gpu/drm/drm_pci.c
@@ -262,16 +262,11 @@ static int drm_pci_irq_by_busid(struct drm_device *dev, struct drm_irq_busid *p)
 	return 0;
 }
 
-static int drm_pci_agp_init(struct drm_device *dev)
+static void drm_pci_agp_init(struct drm_device *dev)
 {
-	if (drm_core_has_AGP(dev)) {
+	if (drm_core_check_feature(dev, DRIVER_USE_AGP)) {
 		if (drm_pci_device_is_agp(dev))
 			dev->agp = drm_agp_init(dev);
-		if (drm_core_check_feature(dev, DRIVER_REQUIRE_AGP)
-		    && (dev->agp == NULL)) {
-			DRM_ERROR("Cannot initialize the agpgart module.\n");
-			return -EINVAL;
-		}
 		if (dev->agp) {
 			dev->agp->agp_mtrr = arch_phys_wc_add(
 				dev->agp->agp_info.aper_base,
@@ -279,15 +274,14 @@ static int drm_pci_agp_init(struct drm_device *dev)
 				1024 * 1024);
 		}
 	}
-	return 0;
 }
 
-static void drm_pci_agp_destroy(struct drm_device *dev)
+void drm_pci_agp_destroy(struct drm_device *dev)
 {
-	if (drm_core_has_AGP(dev) && dev->agp) {
+	if (dev->agp) {
 		arch_phys_wc_del(dev->agp->agp_mtrr);
 		drm_agp_clear(dev);
-		drm_agp_destroy(dev->agp);
+		kfree(dev->agp);
 		dev->agp = NULL;
 	}
 }
@@ -299,8 +293,6 @@ static struct drm_bus drm_pci_bus = {
 	.set_busid = drm_pci_set_busid,
 	.set_unique = drm_pci_set_unique,
 	.irq_by_busid = drm_pci_irq_by_busid,
-	.agp_init = drm_pci_agp_init,
-	.agp_destroy = drm_pci_agp_destroy,
 };
 
 /**
@@ -338,17 +330,25 @@ int drm_get_pci_dev(struct pci_dev *pdev, const struct pci_device_id *ent,
 	if (drm_core_check_feature(dev, DRIVER_MODESET))
 		pci_set_drvdata(pdev, dev);
 
+	drm_pci_agp_init(dev);
+
 	ret = drm_dev_register(dev, ent->driver_data);
 	if (ret)
-		goto err_pci;
+		goto err_agp;
 
 	DRM_INFO("Initialized %s %d.%d.%d %s for %s on minor %d\n",
 		 driver->name, driver->major, driver->minor, driver->patchlevel,
 		 driver->date, pci_name(pdev), dev->primary->index);
 
+	/* No locking needed since shadow-attach is single-threaded since it may
+	 * only be called from the per-driver module init hook. */
+	if (!drm_core_check_feature(dev, DRIVER_MODESET))
+		list_add_tail(&dev->legacy_dev_list, &driver->legacy_dev_list);
+
 	return 0;
 
-err_pci:
+err_agp:
+	drm_pci_agp_destroy(dev);
 	pci_disable_device(pdev);
 err_free:
 	drm_dev_free(dev);
@@ -375,7 +375,6 @@ int drm_pci_init(struct drm_driver *driver, struct pci_driver *pdriver)
 
 	DRM_DEBUG("\n");
 
-	INIT_LIST_HEAD(&driver->device_list);
 	driver->kdriver.pci = pdriver;
 	driver->bus = &drm_pci_bus;
 
@@ -383,6 +382,7 @@ int drm_pci_init(struct drm_driver *driver, struct pci_driver *pdriver)
 		return pci_register_driver(pdriver);
 
 	/* If not using KMS, fall back to stealth mode manual scanning. */
+	INIT_LIST_HEAD(&driver->legacy_dev_list);
 	for (i = 0; pdriver->id_table[i].vendor != 0; i++) {
 		pid = &pdriver->id_table[i];
 
@@ -452,6 +452,7 @@ int drm_pci_init(struct drm_driver *driver, struct pci_driver *pdriver)
 	return -1;
 }
 
+void drm_pci_agp_destroy(struct drm_device *dev) {}
 #endif
 
 EXPORT_SYMBOL(drm_pci_init);
@@ -465,8 +466,11 @@ void drm_pci_exit(struct drm_driver *driver, struct pci_driver *pdriver)
 	if (driver->driver_features & DRIVER_MODESET) {
 		pci_unregister_driver(pdriver);
 	} else {
-		list_for_each_entry_safe(dev, tmp, &driver->device_list, driver_item)
+		list_for_each_entry_safe(dev, tmp, &driver->legacy_dev_list,
+					 legacy_dev_list) {
 			drm_put_dev(dev);
+			list_del(&dev->legacy_dev_list);
+		}
 	}
 	DRM_INFO("Module unloaded\n");
 }
diff --git a/drivers/gpu/drm/drm_platform.c b/drivers/gpu/drm/drm_platform.c
index fc24fee8ec83..21fc82006b78 100644
--- a/drivers/gpu/drm/drm_platform.c
+++ b/drivers/gpu/drm/drm_platform.c
@@ -147,18 +147,6 @@ int drm_platform_init(struct drm_driver *driver, struct platform_device *platfor
 
 	driver->kdriver.platform_device = platform_device;
 	driver->bus = &drm_platform_bus;
-	INIT_LIST_HEAD(&driver->device_list);
 	return drm_get_platform_dev(platform_device, driver);
 }
 EXPORT_SYMBOL(drm_platform_init);
-
-void drm_platform_exit(struct drm_driver *driver, struct platform_device *platform_device)
-{
-	struct drm_device *dev, *tmp;
-	DRM_DEBUG("\n");
-
-	list_for_each_entry_safe(dev, tmp, &driver->device_list, driver_item)
-		drm_put_dev(dev);
-	DRM_INFO("Module unloaded\n");
-}
-EXPORT_SYMBOL(drm_platform_exit);
diff --git a/drivers/gpu/drm/drm_stub.c b/drivers/gpu/drm/drm_stub.c
index 66dd3a001cf1..98a33c580ca1 100644
--- a/drivers/gpu/drm/drm_stub.c
+++ b/drivers/gpu/drm/drm_stub.c
@@ -99,13 +99,19 @@ void drm_ut_debug_printk(unsigned int request_level,
 			 const char *function_name,
 			 const char *format, ...)
 {
+	struct va_format vaf;
 	va_list args;
 
 	if (drm_debug & request_level) {
-		if (function_name)
-			printk(KERN_DEBUG "[%s:%s], ", prefix, function_name);
 		va_start(args, format);
-		vprintk(format, args);
+		vaf.fmt = format;
+		vaf.va = &args;
+
+		if (function_name)
+			printk(KERN_DEBUG "[%s:%s], %pV", prefix,
+			       function_name, &vaf);
+		else
+			printk(KERN_DEBUG "%pV", &vaf);
 		va_end(args);
 	}
 }
@@ -521,16 +527,10 @@ int drm_dev_register(struct drm_device *dev, unsigned long flags)
 
 	mutex_lock(&drm_global_mutex);
 
-	if (dev->driver->bus->agp_init) {
-		ret = dev->driver->bus->agp_init(dev);
-		if (ret)
-			goto out_unlock;
-	}
-
 	if (drm_core_check_feature(dev, DRIVER_MODESET)) {
 		ret = drm_get_minor(dev, &dev->control, DRM_MINOR_CONTROL);
 		if (ret)
-			goto err_agp;
+			goto out_unlock;
 	}
 
 	if (drm_core_check_feature(dev, DRIVER_RENDER) && drm_rnodes) {
@@ -557,8 +557,6 @@ int drm_dev_register(struct drm_device *dev, unsigned long flags)
 			goto err_unload;
 	}
 
-	list_add_tail(&dev->driver_item, &dev->driver->device_list);
-
 	ret = 0;
 	goto out_unlock;
 
@@ -571,9 +569,6 @@ err_render_node:
 	drm_unplug_minor(dev->render);
 err_control_node:
 	drm_unplug_minor(dev->control);
-err_agp:
-	if (dev->driver->bus->agp_destroy)
-		dev->driver->bus->agp_destroy(dev);
 out_unlock:
 	mutex_unlock(&drm_global_mutex);
 	return ret;
@@ -597,8 +592,8 @@ void drm_dev_unregister(struct drm_device *dev)
 	if (dev->driver->unload)
 		dev->driver->unload(dev);
 
-	if (dev->driver->bus->agp_destroy)
-		dev->driver->bus->agp_destroy(dev);
+	if (dev->agp)
+		drm_pci_agp_destroy(dev);
 
 	drm_vblank_cleanup(dev);
 
@@ -608,7 +603,5 @@ void drm_dev_unregister(struct drm_device *dev)
 	drm_unplug_minor(dev->control);
 	drm_unplug_minor(dev->render);
 	drm_unplug_minor(dev->primary);
-
-	list_del(&dev->driver_item);
 }
 EXPORT_SYMBOL(drm_dev_unregister);
diff --git a/drivers/gpu/drm/drm_usb.c b/drivers/gpu/drm/drm_usb.c
index b179b70e7853..0f8cb1ae7607 100644
--- a/drivers/gpu/drm/drm_usb.c
+++ b/drivers/gpu/drm/drm_usb.c
@@ -1,4 +1,5 @@
 #include <drm/drmP.h>
+#include <drm/drm_usb.h>
 #include <linux/usb.h>
 #include <linux/module.h>
 
@@ -63,7 +64,6 @@ int drm_usb_init(struct drm_driver *driver, struct usb_driver *udriver)
 	int res;
 	DRM_DEBUG("\n");
 
-	INIT_LIST_HEAD(&driver->device_list);
 	driver->kdriver.usb = udriver;
 	driver->bus = &drm_usb_bus;
 
diff --git a/drivers/gpu/drm/drm_vm.c b/drivers/gpu/drm/drm_vm.c
index 93e95d7efd57..24e045c4f531 100644
--- a/drivers/gpu/drm/drm_vm.c
+++ b/drivers/gpu/drm/drm_vm.c
@@ -101,7 +101,7 @@ static int drm_do_vm_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 	/*
 	 * Find the right map
 	 */
-	if (!drm_core_has_AGP(dev))
+	if (!dev->agp)
 		goto vm_fault_error;
 
 	if (!dev->agp || !dev->agp->cant_use_aperture)
@@ -220,7 +220,6 @@ static void drm_vm_shm_close(struct vm_area_struct *vma)
 
 	DRM_DEBUG("0x%08lx,0x%08lx\n",
 		  vma->vm_start, vma->vm_end - vma->vm_start);
-	atomic_dec(&dev->vma_count);
 
 	map = vma->vm_private_data;
 
@@ -266,9 +265,6 @@ static void drm_vm_shm_close(struct vm_area_struct *vma)
 				dmah.size = map->size;
 				__drm_pci_free(dev, &dmah);
 				break;
-			case _DRM_GEM:
-				DRM_ERROR("tried to rmmap GEM object\n");
-				break;
 			}
 			kfree(map);
 		}
@@ -408,7 +404,6 @@ void drm_vm_open_locked(struct drm_device *dev,
 
 	DRM_DEBUG("0x%08lx,0x%08lx\n",
 		  vma->vm_start, vma->vm_end - vma->vm_start);
-	atomic_inc(&dev->vma_count);
 
 	vma_entry = kmalloc(sizeof(*vma_entry), GFP_KERNEL);
 	if (vma_entry) {
@@ -436,7 +431,6 @@ void drm_vm_close_locked(struct drm_device *dev,
 
 	DRM_DEBUG("0x%08lx,0x%08lx\n",
 		  vma->vm_start, vma->vm_end - vma->vm_start);
-	atomic_dec(&dev->vma_count);
 
 	list_for_each_entry_safe(pt, temp, &dev->vmalist, head) {
 		if (pt->vma == vma) {
@@ -595,7 +589,7 @@ int drm_mmap_locked(struct file *filp, struct vm_area_struct *vma)
 	switch (map->type) {
 #if !defined(__arm__)
 	case _DRM_AGP:
-		if (drm_core_has_AGP(dev) && dev->agp->cant_use_aperture) {
+		if (dev->agp && dev->agp->cant_use_aperture) {
 			/*
 			 * On some platforms we can't talk to bus dma address from the CPU, so for
 			 * memory of type DRM_AGP, we'll deal with sorting out the real physical
diff --git a/drivers/gpu/drm/exynos/exynos_drm_drv.c b/drivers/gpu/drm/exynos/exynos_drm_drv.c
index 22b8f5eced80..9d096a0c5f8d 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_drv.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_drv.c
@@ -14,6 +14,8 @@
 #include <drm/drmP.h>
 #include <drm/drm_crtc_helper.h>
 
+#include <linux/anon_inodes.h>
+
 #include <drm/exynos_drm.h>
 
 #include "exynos_drm_drv.h"
@@ -119,6 +121,8 @@ static int exynos_drm_load(struct drm_device *dev, unsigned long flags)
 
 	drm_vblank_offdelay = VBLANK_OFF_DELAY;
 
+	platform_set_drvdata(dev->platformdev, dev);
+
 	return 0;
 
 err_drm_device:
@@ -150,9 +154,14 @@ static int exynos_drm_unload(struct drm_device *dev)
 	return 0;
 }
 
+static const struct file_operations exynos_drm_gem_fops = {
+	.mmap = exynos_drm_gem_mmap_buffer,
+};
+
 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);
@@ -167,6 +176,16 @@ static int exynos_drm_open(struct drm_device *dev, struct drm_file *file)
 		file->driver_priv = NULL;
 	}
 
+	anon_filp = anon_inode_getfile("exynos_gem", &exynos_drm_gem_fops,
+					NULL, 0);
+	if (IS_ERR(anon_filp)) {
+		kfree(file_priv);
+		return PTR_ERR(anon_filp);
+	}
+
+	anon_filp->f_mode = FMODE_READ | FMODE_WRITE;
+	file_priv->anon_filp = anon_filp;
+
 	return ret;
 }
 
@@ -179,6 +198,7 @@ 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;
@@ -204,6 +224,9 @@ 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;
@@ -305,7 +328,7 @@ static int exynos_drm_platform_probe(struct platform_device *pdev)
 
 static int exynos_drm_platform_remove(struct platform_device *pdev)
 {
-	drm_platform_exit(&exynos_drm_driver, pdev);
+	drm_put_dev(platform_get_drvdata(pdev));
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/exynos/exynos_drm_drv.h b/drivers/gpu/drm/exynos/exynos_drm_drv.h
index eaa19668bf00..0eaf5a27e120 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_drv.h
+++ b/drivers/gpu/drm/exynos/exynos_drm_drv.h
@@ -226,6 +226,7 @@ struct exynos_drm_ipp_private {
 struct drm_exynos_file_private {
 	struct exynos_drm_g2d_private	*g2d_priv;
 	struct exynos_drm_ipp_private	*ipp_priv;
+	struct file			*anon_filp;
 };
 
 /*
diff --git a/drivers/gpu/drm/exynos/exynos_drm_fimd.c b/drivers/gpu/drm/exynos/exynos_drm_fimd.c
index a61878bf5dcd..a20440ce32e6 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_fimd.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_fimd.c
@@ -347,7 +347,7 @@ static void fimd_wait_for_vblank(struct device *dev)
 	 */
 	if (!wait_event_timeout(ctx->wait_vsync_queue,
 				!atomic_read(&ctx->wait_vsync_event),
-				DRM_HZ/20))
+				HZ/20))
 		DRM_DEBUG_KMS("vblank wait timed out.\n");
 }
 
@@ -706,7 +706,7 @@ static irqreturn_t fimd_irq_handler(int irq, void *dev_id)
 	/* set wait vsync event to zero and wake up queue. */
 	if (atomic_read(&ctx->wait_vsync_event)) {
 		atomic_set(&ctx->wait_vsync_event, 0);
-		DRM_WAKEUP(&ctx->wait_vsync_queue);
+		wake_up(&ctx->wait_vsync_queue);
 	}
 out:
 	return IRQ_HANDLED;
@@ -954,7 +954,7 @@ static int fimd_probe(struct platform_device *pdev)
 	}
 
 	ctx->driver_data = drm_fimd_get_driver_data(pdev);
-	DRM_INIT_WAITQUEUE(&ctx->wait_vsync_queue);
+	init_waitqueue_head(&ctx->wait_vsync_queue);
 	atomic_set(&ctx->wait_vsync_event, 0);
 
 	subdrv = &ctx->subdrv;
diff --git a/drivers/gpu/drm/exynos/exynos_drm_gem.c b/drivers/gpu/drm/exynos/exynos_drm_gem.c
index be59d50d8b16..42d2904d88c7 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_gem.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_gem.c
@@ -338,46 +338,22 @@ int exynos_drm_gem_map_offset_ioctl(struct drm_device *dev, void *data,
 			&args->offset);
 }
 
-static struct drm_file *exynos_drm_find_drm_file(struct drm_device *drm_dev,
-							struct file *filp)
-{
-	struct drm_file *file_priv;
-
-	/* find current process's drm_file from filelist. */
-	list_for_each_entry(file_priv, &drm_dev->filelist, lhead)
-		if (file_priv->filp == filp)
-			return file_priv;
-
-	WARN_ON(1);
-
-	return ERR_PTR(-EFAULT);
-}
-
-static int exynos_drm_gem_mmap_buffer(struct file *filp,
+int exynos_drm_gem_mmap_buffer(struct file *filp,
 				      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 exynos_drm_gem_buf *buffer;
-	struct drm_file *file_priv;
 	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;
 
-	/* restore it to driver's fops. */
-	filp->f_op = fops_get(drm_dev->driver->fops);
-
-	file_priv = exynos_drm_find_drm_file(drm_dev, filp);
-	if (IS_ERR(file_priv))
-		return PTR_ERR(file_priv);
-
-	/* restore it to drm_file. */
-	filp->private_data = file_priv;
-
 	update_vm_cache_attr(exynos_gem_obj, vma);
 
 	vm_size = vma->vm_end - vma->vm_start;
@@ -411,15 +387,13 @@ static int exynos_drm_gem_mmap_buffer(struct file *filp,
 	return 0;
 }
 
-static const struct file_operations exynos_drm_gem_fops = {
-	.mmap = exynos_drm_gem_mmap_buffer,
-};
-
 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)) {
@@ -427,47 +401,25 @@ int exynos_drm_gem_mmap_ioctl(struct drm_device *dev, void *data,
 		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;
 	}
 
-	/*
-	 * We have to use gem object and its fops for specific mmaper,
-	 * but vm_mmap() can deliver only filp. So we have to change
-	 * filp->f_op and filp->private_data temporarily, then restore
-	 * again. So it is important to keep lock until restoration the
-	 * settings to prevent others from misuse of filp->f_op or
-	 * filp->private_data.
-	 */
-	mutex_lock(&dev->struct_mutex);
-
-	/*
-	 * Set specific mmper's fops. And it will be restored by
-	 * exynos_drm_gem_mmap_buffer to dev->driver->fops.
-	 * This is used to call specific mapper temporarily.
-	 */
-	file_priv->filp->f_op = &exynos_drm_gem_fops;
-
-	/*
-	 * Set gem object to private_data so that specific mmaper
-	 * can get the gem object. And it will be restored by
-	 * exynos_drm_gem_mmap_buffer to drm_file.
-	 */
-	file_priv->filp->private_data = obj;
+	exynos_file_priv = file_priv->driver_priv;
+	anon_filp = exynos_file_priv->anon_filp;
+	anon_filp->private_data = obj;
 
-	addr = vm_mmap(file_priv->filp, 0, args->size,
-			PROT_READ | PROT_WRITE, MAP_SHARED, 0);
+	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)) {
-		/* check filp->f_op, filp->private_data are restored */
-		if (file_priv->filp->f_op == &exynos_drm_gem_fops) {
-			file_priv->filp->f_op = fops_get(dev->driver->fops);
-			file_priv->filp->private_data = file_priv;
-		}
 		mutex_unlock(&dev->struct_mutex);
 		return (int)addr;
 	}
diff --git a/drivers/gpu/drm/exynos/exynos_drm_gem.h b/drivers/gpu/drm/exynos/exynos_drm_gem.h
index b8c818ba2ff4..1592c0ba7de8 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_gem.h
+++ b/drivers/gpu/drm/exynos/exynos_drm_gem.h
@@ -122,6 +122,9 @@ int exynos_drm_gem_map_offset_ioctl(struct drm_device *dev, void *data,
 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_mixer.c b/drivers/gpu/drm/exynos/exynos_mixer.c
index 63bc5f92fbb3..2dfa48c76f54 100644
--- a/drivers/gpu/drm/exynos/exynos_mixer.c
+++ b/drivers/gpu/drm/exynos/exynos_mixer.c
@@ -868,7 +868,7 @@ static void mixer_wait_for_vblank(void *ctx)
 	 */
 	if (!wait_event_timeout(mixer_ctx->wait_vsync_queue,
 				!atomic_read(&mixer_ctx->wait_vsync_event),
-				DRM_HZ/20))
+				HZ/20))
 		DRM_DEBUG_KMS("vblank wait timed out.\n");
 }
 
@@ -1019,7 +1019,7 @@ static irqreturn_t mixer_irq_handler(int irq, void *arg)
 		/* set wait vsync event to zero and wake up queue. */
 		if (atomic_read(&ctx->wait_vsync_event)) {
 			atomic_set(&ctx->wait_vsync_event, 0);
-			DRM_WAKEUP(&ctx->wait_vsync_queue);
+			wake_up(&ctx->wait_vsync_queue);
 		}
 	}
 
@@ -1209,7 +1209,7 @@ static int mixer_probe(struct platform_device *pdev)
 	drm_hdmi_ctx->ctx = (void *)ctx;
 	ctx->vp_enabled = drv->is_vp_enabled;
 	ctx->mxr_ver = drv->version;
-	DRM_INIT_WAITQUEUE(&ctx->wait_vsync_queue);
+	init_waitqueue_head(&ctx->wait_vsync_queue);
 	atomic_set(&ctx->wait_vsync_event, 0);
 
 	platform_set_drvdata(pdev, drm_hdmi_ctx);
diff --git a/drivers/gpu/drm/gma500/accel_2d.c b/drivers/gpu/drm/gma500/accel_2d.c
index d5ef1a5793c8..de6f62a6ceb7 100644
--- a/drivers/gpu/drm/gma500/accel_2d.c
+++ b/drivers/gpu/drm/gma500/accel_2d.c
@@ -326,7 +326,7 @@ int psbfb_sync(struct fb_info *info)
 	struct psb_framebuffer *psbfb = &fbdev->pfb;
 	struct drm_device *dev = psbfb->base.dev;
 	struct drm_psb_private *dev_priv = dev->dev_private;
-	unsigned long _end = jiffies + DRM_HZ;
+	unsigned long _end = jiffies + HZ;
 	int busy = 0;
 	unsigned long flags;
 
diff --git a/drivers/gpu/drm/gma500/cdv_intel_dp.c b/drivers/gpu/drm/gma500/cdv_intel_dp.c
index f88a1815d87c..0490ce36b53f 100644
--- a/drivers/gpu/drm/gma500/cdv_intel_dp.c
+++ b/drivers/gpu/drm/gma500/cdv_intel_dp.c
@@ -483,7 +483,7 @@ cdv_intel_dp_aux_native_write(struct gma_encoder *encoder,
 
 	if (send_bytes > 16)
 		return -1;
-	msg[0] = AUX_NATIVE_WRITE << 4;
+	msg[0] = DP_AUX_NATIVE_WRITE << 4;
 	msg[1] = address >> 8;
 	msg[2] = address & 0xff;
 	msg[3] = send_bytes - 1;
@@ -493,9 +493,10 @@ cdv_intel_dp_aux_native_write(struct gma_encoder *encoder,
 		ret = cdv_intel_dp_aux_ch(encoder, msg, msg_bytes, &ack, 1);
 		if (ret < 0)
 			return ret;
-		if ((ack & AUX_NATIVE_REPLY_MASK) == AUX_NATIVE_REPLY_ACK)
+		ack >>= 4;
+		if ((ack & DP_AUX_NATIVE_REPLY_MASK) == DP_AUX_NATIVE_REPLY_ACK)
 			break;
-		else if ((ack & AUX_NATIVE_REPLY_MASK) == AUX_NATIVE_REPLY_DEFER)
+		else if ((ack & DP_AUX_NATIVE_REPLY_MASK) == DP_AUX_NATIVE_REPLY_DEFER)
 			udelay(100);
 		else
 			return -EIO;
@@ -523,7 +524,7 @@ cdv_intel_dp_aux_native_read(struct gma_encoder *encoder,
 	uint8_t ack;
 	int ret;
 
-	msg[0] = AUX_NATIVE_READ << 4;
+	msg[0] = DP_AUX_NATIVE_READ << 4;
 	msg[1] = address >> 8;
 	msg[2] = address & 0xff;
 	msg[3] = recv_bytes - 1;
@@ -538,12 +539,12 @@ cdv_intel_dp_aux_native_read(struct gma_encoder *encoder,
 			return -EPROTO;
 		if (ret < 0)
 			return ret;
-		ack = reply[0];
-		if ((ack & AUX_NATIVE_REPLY_MASK) == AUX_NATIVE_REPLY_ACK) {
+		ack = reply[0] >> 4;
+		if ((ack & DP_AUX_NATIVE_REPLY_MASK) == DP_AUX_NATIVE_REPLY_ACK) {
 			memcpy(recv, reply + 1, ret - 1);
 			return ret - 1;
 		}
-		else if ((ack & AUX_NATIVE_REPLY_MASK) == AUX_NATIVE_REPLY_DEFER)
+		else if ((ack & DP_AUX_NATIVE_REPLY_MASK) == DP_AUX_NATIVE_REPLY_DEFER)
 			udelay(100);
 		else
 			return -EIO;
@@ -569,12 +570,12 @@ cdv_intel_dp_i2c_aux_ch(struct i2c_adapter *adapter, int mode,
 
 	/* Set up the command byte */
 	if (mode & MODE_I2C_READ)
-		msg[0] = AUX_I2C_READ << 4;
+		msg[0] = DP_AUX_I2C_READ << 4;
 	else
-		msg[0] = AUX_I2C_WRITE << 4;
+		msg[0] = DP_AUX_I2C_WRITE << 4;
 
 	if (!(mode & MODE_I2C_STOP))
-		msg[0] |= AUX_I2C_MOT << 4;
+		msg[0] |= DP_AUX_I2C_MOT << 4;
 
 	msg[1] = address >> 8;
 	msg[2] = address;
@@ -606,16 +607,16 @@ cdv_intel_dp_i2c_aux_ch(struct i2c_adapter *adapter, int mode,
 			return ret;
 		}
 
-		switch (reply[0] & AUX_NATIVE_REPLY_MASK) {
-		case AUX_NATIVE_REPLY_ACK:
+		switch ((reply[0] >> 4) & DP_AUX_NATIVE_REPLY_MASK) {
+		case DP_AUX_NATIVE_REPLY_ACK:
 			/* I2C-over-AUX Reply field is only valid
 			 * when paired with AUX ACK.
 			 */
 			break;
-		case AUX_NATIVE_REPLY_NACK:
+		case DP_AUX_NATIVE_REPLY_NACK:
 			DRM_DEBUG_KMS("aux_ch native nack\n");
 			return -EREMOTEIO;
-		case AUX_NATIVE_REPLY_DEFER:
+		case DP_AUX_NATIVE_REPLY_DEFER:
 			udelay(100);
 			continue;
 		default:
@@ -624,16 +625,16 @@ cdv_intel_dp_i2c_aux_ch(struct i2c_adapter *adapter, int mode,
 			return -EREMOTEIO;
 		}
 
-		switch (reply[0] & AUX_I2C_REPLY_MASK) {
-		case AUX_I2C_REPLY_ACK:
+		switch ((reply[0] >> 4) & DP_AUX_I2C_REPLY_MASK) {
+		case DP_AUX_I2C_REPLY_ACK:
 			if (mode == MODE_I2C_READ) {
 				*read_byte = reply[1];
 			}
 			return reply_bytes - 1;
-		case AUX_I2C_REPLY_NACK:
+		case DP_AUX_I2C_REPLY_NACK:
 			DRM_DEBUG_KMS("aux_i2c nack\n");
 			return -EREMOTEIO;
-		case AUX_I2C_REPLY_DEFER:
+		case DP_AUX_I2C_REPLY_DEFER:
 			DRM_DEBUG_KMS("aux_i2c defer\n");
 			udelay(100);
 			break;
@@ -677,7 +678,7 @@ cdv_intel_dp_i2c_init(struct gma_connector *connector,
 	return ret;
 }
 
-void cdv_intel_fixed_panel_mode(struct drm_display_mode *fixed_mode,
+static void cdv_intel_fixed_panel_mode(struct drm_display_mode *fixed_mode,
 	struct drm_display_mode *adjusted_mode)
 {
 	adjusted_mode->hdisplay = fixed_mode->hdisplay;
diff --git a/drivers/gpu/drm/gma500/gma_display.c b/drivers/gpu/drm/gma500/gma_display.c
index 24e8af3d22bf..386de2c9dc86 100644
--- a/drivers/gpu/drm/gma500/gma_display.c
+++ b/drivers/gpu/drm/gma500/gma_display.c
@@ -349,6 +349,7 @@ int gma_crtc_cursor_set(struct drm_crtc *crtc,
 	/* If we didn't get a handle then turn the cursor off */
 	if (!handle) {
 		temp = CURSOR_MODE_DISABLE;
+		mutex_lock(&dev->struct_mutex);
 
 		if (gma_power_begin(dev, false)) {
 			REG_WRITE(control, temp);
@@ -365,6 +366,7 @@ int gma_crtc_cursor_set(struct drm_crtc *crtc,
 			gma_crtc->cursor_obj = NULL;
 		}
 
+		mutex_unlock(&dev->struct_mutex);
 		return 0;
 	}
 
@@ -374,9 +376,12 @@ int gma_crtc_cursor_set(struct drm_crtc *crtc,
 		return -EINVAL;
 	}
 
+	mutex_lock(&dev->struct_mutex);
 	obj = drm_gem_object_lookup(dev, file_priv, handle);
-	if (!obj)
-		return -ENOENT;
+	if (!obj) {
+		ret = -ENOENT;
+		goto unlock;
+	}
 
 	if (obj->size < width * height * 4) {
 		dev_dbg(dev->dev, "Buffer is too small\n");
@@ -440,10 +445,13 @@ int gma_crtc_cursor_set(struct drm_crtc *crtc,
 	}
 
 	gma_crtc->cursor_obj = obj;
+unlock:
+	mutex_unlock(&dev->struct_mutex);
 	return ret;
 
 unref_cursor:
 	drm_gem_object_unreference(obj);
+	mutex_unlock(&dev->struct_mutex);
 	return ret;
 }
 
diff --git a/drivers/gpu/drm/gma500/psb_drv.h b/drivers/gpu/drm/gma500/psb_drv.h
index b59e6588c343..5ad6a03e477e 100644
--- a/drivers/gpu/drm/gma500/psb_drv.h
+++ b/drivers/gpu/drm/gma500/psb_drv.h
@@ -212,8 +212,8 @@ enum {
 #define PSB_HIGH_REG_OFFS 0x0600
 
 #define PSB_NUM_VBLANKS 2
-#define PSB_WATCHDOG_DELAY (DRM_HZ * 2)
-#define PSB_LID_DELAY (DRM_HZ / 10)
+#define PSB_WATCHDOG_DELAY (HZ * 2)
+#define PSB_LID_DELAY (HZ / 10)
 
 #define MDFLD_PNW_B0 0x04
 #define MDFLD_PNW_C0 0x08
@@ -232,7 +232,7 @@ enum {
 #define MDFLD_DSR_RR		45
 #define MDFLD_DPU_ENABLE 	(1 << 31)
 #define MDFLD_DSR_FULLSCREEN 	(1 << 30)
-#define MDFLD_DSR_DELAY		(DRM_HZ / MDFLD_DSR_RR)
+#define MDFLD_DSR_DELAY		(HZ / MDFLD_DSR_RR)
 
 #define PSB_PWR_STATE_ON		1
 #define PSB_PWR_STATE_OFF		2
@@ -769,7 +769,7 @@ extern void psb_mmu_remove_pages(struct psb_mmu_pd *pd,
  *psb_irq.c
  */
 
-extern irqreturn_t psb_irq_handler(DRM_IRQ_ARGS);
+extern irqreturn_t psb_irq_handler(int irq, void *arg);
 extern int psb_irq_enable_dpst(struct drm_device *dev);
 extern int psb_irq_disable_dpst(struct drm_device *dev);
 extern void psb_irq_preinstall(struct drm_device *dev);
diff --git a/drivers/gpu/drm/gma500/psb_intel_drv.h b/drivers/gpu/drm/gma500/psb_intel_drv.h
index bde27fdb41bf..dc2c8eb030fa 100644
--- a/drivers/gpu/drm/gma500/psb_intel_drv.h
+++ b/drivers/gpu/drm/gma500/psb_intel_drv.h
@@ -250,11 +250,6 @@ extern void psb_intel_sdvo_set_hotplug(struct drm_connector *connector,
 extern int intelfb_probe(struct drm_device *dev);
 extern int intelfb_remove(struct drm_device *dev,
 			  struct drm_framebuffer *fb);
-extern struct drm_framebuffer *psb_intel_framebuffer_create(struct drm_device
-							*dev, struct
-							drm_mode_fb_cmd
-							*mode_cmd,
-							void *mm_private);
 extern bool psb_intel_lvds_mode_fixup(struct drm_encoder *encoder,
 				      const struct drm_display_mode *mode,
 				      struct drm_display_mode *adjusted_mode);
diff --git a/drivers/gpu/drm/gma500/psb_irq.c b/drivers/gpu/drm/gma500/psb_irq.c
index ba4830342d34..f883f9e4c524 100644
--- a/drivers/gpu/drm/gma500/psb_irq.c
+++ b/drivers/gpu/drm/gma500/psb_irq.c
@@ -200,7 +200,7 @@ static void psb_vdc_interrupt(struct drm_device *dev, uint32_t vdc_stat)
 		mid_pipe_event_handler(dev, 1);
 }
 
-irqreturn_t psb_irq_handler(DRM_IRQ_ARGS)
+irqreturn_t psb_irq_handler(int irq, void *arg)
 {
 	struct drm_device *dev = arg;
 	struct drm_psb_private *dev_priv = dev->dev_private;
@@ -253,7 +253,7 @@ irqreturn_t psb_irq_handler(DRM_IRQ_ARGS)
 
 	PSB_WVDC32(vdc_stat, PSB_INT_IDENTITY_R);
 	(void) PSB_RVDC32(PSB_INT_IDENTITY_R);
-	DRM_READMEMORYBARRIER();
+	rmb();
 
 	if (!handled)
 		return IRQ_NONE;
@@ -450,21 +450,6 @@ int psb_irq_disable_dpst(struct drm_device *dev)
 	return 0;
 }
 
-#ifdef PSB_FIXME
-static int psb_vblank_do_wait(struct drm_device *dev,
-			      unsigned int *sequence, atomic_t *counter)
-{
-	unsigned int cur_vblank;
-	int ret = 0;
-	DRM_WAIT_ON(ret, dev->vblank.queue, 3 * DRM_HZ,
-		    (((cur_vblank = atomic_read(counter))
-		      - *sequence) <= (1 << 23)));
-	*sequence = cur_vblank;
-
-	return ret;
-}
-#endif
-
 /*
  * It is used to enable VBLANK interrupt
  */
diff --git a/drivers/gpu/drm/gma500/psb_irq.h b/drivers/gpu/drm/gma500/psb_irq.h
index debb7f190c06..d0b45ffa1126 100644
--- a/drivers/gpu/drm/gma500/psb_irq.h
+++ b/drivers/gpu/drm/gma500/psb_irq.h
@@ -32,7 +32,7 @@ void sysirq_uninit(struct drm_device *dev);
 void psb_irq_preinstall(struct drm_device *dev);
 int  psb_irq_postinstall(struct drm_device *dev);
 void psb_irq_uninstall(struct drm_device *dev);
-irqreturn_t psb_irq_handler(DRM_IRQ_ARGS);
+irqreturn_t psb_irq_handler(int irq, void *arg);
 
 int psb_irq_enable_dpst(struct drm_device *dev);
 int psb_irq_disable_dpst(struct drm_device *dev);
diff --git a/drivers/gpu/drm/i810/i810_dma.c b/drivers/gpu/drm/i810/i810_dma.c
index 249fdff305c6..aeace37415aa 100644
--- a/drivers/gpu/drm/i810/i810_dma.c
+++ b/drivers/gpu/drm/i810/i810_dma.c
@@ -1193,6 +1193,10 @@ static int i810_flip_bufs(struct drm_device *dev, void *data,
 
 int i810_driver_load(struct drm_device *dev, unsigned long flags)
 {
+	/* Our userspace depends upon the agp mapping support. */
+	if (!dev->agp)
+		return -EINVAL;
+
 	pci_set_master(dev->pdev);
 
 	return 0;
diff --git a/drivers/gpu/drm/i810/i810_drv.c b/drivers/gpu/drm/i810/i810_drv.c
index d8180d22cedd..441ccf8f5bdc 100644
--- a/drivers/gpu/drm/i810/i810_drv.c
+++ b/drivers/gpu/drm/i810/i810_drv.c
@@ -57,7 +57,7 @@ static const struct file_operations i810_driver_fops = {
 
 static struct drm_driver driver = {
 	.driver_features =
-	    DRIVER_USE_AGP | DRIVER_REQUIRE_AGP |
+	    DRIVER_USE_AGP |
 	    DRIVER_HAVE_DMA,
 	.dev_priv_size = sizeof(drm_i810_buf_priv_t),
 	.load = i810_driver_load,
diff --git a/drivers/gpu/drm/i915/Kconfig b/drivers/gpu/drm/i915/Kconfig
index 6199d0b5b958..73ed59eff139 100644
--- a/drivers/gpu/drm/i915/Kconfig
+++ b/drivers/gpu/drm/i915/Kconfig
@@ -1,8 +1,10 @@
 config DRM_I915
 	tristate "Intel 8xx/9xx/G3x/G4x/HD Graphics"
 	depends on DRM
-	depends on AGP
-	depends on AGP_INTEL
+	depends on X86 && PCI
+	depends on (AGP || AGP=n)
+	select INTEL_GTT
+	select AGP_INTEL if AGP
 	# we need shmfs for the swappable backing store, and in particular
 	# the shmem_readpage() which depends upon tmpfs
 	select SHMEM
@@ -35,15 +37,14 @@ config DRM_I915
 config DRM_I915_KMS
 	bool "Enable modesetting on intel by default"
 	depends on DRM_I915
+	default y
 	help
-	  Choose this option if you want kernel modesetting enabled by default,
-	  and you have a new enough userspace to support this. Running old
-	  userspaces with this enabled will cause pain.  Note that this causes
-	  the driver to bind to PCI devices, which precludes loading things
-	  like intelfb.
+	  Choose this option if you want kernel modesetting enabled by default.
+
+	  If in doubt, say "Y".
 
 config DRM_I915_FBDEV
-	bool "Enable legacy fbdev support for the modesettting intel driver"
+	bool "Enable legacy fbdev support for the modesetting intel driver"
 	depends on DRM_I915
 	select DRM_KMS_FB_HELPER
 	select FB_CFB_FILLRECT
@@ -55,9 +56,12 @@ config DRM_I915_FBDEV
 	  support. Note that this support also provide the linux console
 	  support on top of the intel modesetting driver.
 
+	  If in doubt, say "Y".
+
 config DRM_I915_PRELIMINARY_HW_SUPPORT
 	bool "Enable preliminary support for prerelease Intel hardware by default"
 	depends on DRM_I915
+	default n
 	help
 	  Choose this option if you have prerelease Intel hardware and want the
 	  i915 driver to support it by default.  You can enable such support at
@@ -65,3 +69,15 @@ config DRM_I915_PRELIMINARY_HW_SUPPORT
 	  option changes the default for that module option.
 
 	  If in doubt, say "N".
+
+config DRM_I915_UMS
+	bool "Enable userspace modesetting on Intel hardware (DEPRECATED)"
+	depends on DRM_I915
+	default n
+	help
+	  Choose this option if you still need userspace modesetting.
+
+	  Userspace modesetting is deprecated for quite some time now, so
+	  enable this only if you have ancient versions of the DDX drivers.
+
+	  If in doubt, say "N".
diff --git a/drivers/gpu/drm/i915/Makefile b/drivers/gpu/drm/i915/Makefile
index d4ae48b04cf2..9fd44f5f3b3b 100644
--- a/drivers/gpu/drm/i915/Makefile
+++ b/drivers/gpu/drm/i915/Makefile
@@ -4,7 +4,6 @@
 
 ccflags-y := -Iinclude/drm
 i915-y := i915_drv.o i915_dma.o i915_irq.o \
-	  i915_debugfs.o \
 	  i915_gpu_error.o \
           i915_suspend.o \
 	  i915_gem.o \
@@ -54,6 +53,8 @@ i915-$(CONFIG_ACPI)	+= intel_acpi.o intel_opregion.o
 
 i915-$(CONFIG_DRM_I915_FBDEV) += intel_fbdev.o
 
+i915-$(CONFIG_DEBUG_FS) += i915_debugfs.o
+
 obj-$(CONFIG_DRM_I915)  += i915.o
 
 CFLAGS_i915_trace_points.o := -I$(src)
diff --git a/drivers/gpu/drm/i915/dvo_ns2501.c b/drivers/gpu/drm/i915/dvo_ns2501.c
index c4a255be6979..954acb2c7021 100644
--- a/drivers/gpu/drm/i915/dvo_ns2501.c
+++ b/drivers/gpu/drm/i915/dvo_ns2501.c
@@ -87,49 +87,6 @@ struct ns2501_priv {
  * when switching the resolution.
  */
 
-static void enable_dvo(struct intel_dvo_device *dvo)
-{
-	struct ns2501_priv *ns = (struct ns2501_priv *)(dvo->dev_priv);
-	struct i2c_adapter *adapter = dvo->i2c_bus;
-	struct intel_gmbus *bus = container_of(adapter,
-					       struct intel_gmbus,
-					       adapter);
-	struct drm_i915_private *dev_priv = bus->dev_priv;
-
-	DRM_DEBUG_KMS("%s: Trying to re-enable the DVO\n", __FUNCTION__);
-
-	ns->dvoc = I915_READ(DVO_C);
-	ns->pll_a = I915_READ(_DPLL_A);
-	ns->srcdim = I915_READ(DVOC_SRCDIM);
-	ns->fw_blc = I915_READ(FW_BLC);
-
-	I915_WRITE(DVOC, 0x10004084);
-	I915_WRITE(_DPLL_A, 0xd0820000);
-	I915_WRITE(DVOC_SRCDIM, 0x400300);	// 1024x768
-	I915_WRITE(FW_BLC, 0x1080304);
-
-	I915_WRITE(DVOC, 0x90004084);
-}
-
-/*
- * Restore the I915 registers modified by the above
- * trigger function.
- */
-static void restore_dvo(struct intel_dvo_device *dvo)
-{
-	struct i2c_adapter *adapter = dvo->i2c_bus;
-	struct intel_gmbus *bus = container_of(adapter,
-					       struct intel_gmbus,
-					       adapter);
-	struct drm_i915_private *dev_priv = bus->dev_priv;
-	struct ns2501_priv *ns = (struct ns2501_priv *)(dvo->dev_priv);
-
-	I915_WRITE(DVOC, ns->dvoc);
-	I915_WRITE(_DPLL_A, ns->pll_a);
-	I915_WRITE(DVOC_SRCDIM, ns->srcdim);
-	I915_WRITE(FW_BLC, ns->fw_blc);
-}
-
 /*
 ** Read a register from the ns2501.
 ** Returns true if successful, false otherwise.
@@ -300,7 +257,7 @@ static void ns2501_mode_set(struct intel_dvo_device *dvo,
 			    struct drm_display_mode *adjusted_mode)
 {
 	bool ok;
-	bool restore = false;
+	int retries = 10;
 	struct ns2501_priv *ns = (struct ns2501_priv *)(dvo->dev_priv);
 
 	DRM_DEBUG_KMS
@@ -476,20 +433,7 @@ static void ns2501_mode_set(struct intel_dvo_device *dvo,
 			ns->reg_8_shadow |= NS2501_8_BPAS;
 		}
 		ok &= ns2501_writeb(dvo, NS2501_REG8, ns->reg_8_shadow);
-
-		if (!ok) {
-			if (restore)
-				restore_dvo(dvo);
-			enable_dvo(dvo);
-			restore = true;
-		}
-	} while (!ok);
-	/*
-	 * Restore the old i915 registers before
-	 * forcing the ns2501 on.
-	 */
-	if (restore)
-		restore_dvo(dvo);
+	} while (!ok && retries--);
 }
 
 /* set the NS2501 power state */
@@ -510,7 +454,7 @@ static bool ns2501_get_hw_state(struct intel_dvo_device *dvo)
 static void ns2501_dpms(struct intel_dvo_device *dvo, bool enable)
 {
 	bool ok;
-	bool restore = false;
+	int retries = 10;
 	struct ns2501_priv *ns = (struct ns2501_priv *)(dvo->dev_priv);
 	unsigned char ch;
 
@@ -537,16 +481,7 @@ static void ns2501_dpms(struct intel_dvo_device *dvo, bool enable)
 			ok &=
 			    ns2501_writeb(dvo, 0x35,
 					  enable ? 0xff : 0x00);
-			if (!ok) {
-				if (restore)
-					restore_dvo(dvo);
-				enable_dvo(dvo);
-				restore = true;
-			}
-		} while (!ok);
-
-		if (restore)
-			restore_dvo(dvo);
+		} while (!ok && retries--);
 	}
 }
 
diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c
index 6ed45a984230..b2b46c52294c 100644
--- a/drivers/gpu/drm/i915/i915_debugfs.c
+++ b/drivers/gpu/drm/i915/i915_debugfs.c
@@ -40,8 +40,6 @@
 #include <drm/i915_drm.h>
 #include "i915_drv.h"
 
-#if defined(CONFIG_DEBUG_FS)
-
 enum {
 	ACTIVE_LIST,
 	INACTIVE_LIST,
@@ -406,16 +404,26 @@ static int i915_gem_object_info(struct seq_file *m, void* data)
 	seq_putc(m, '\n');
 	list_for_each_entry_reverse(file, &dev->filelist, lhead) {
 		struct file_stats stats;
+		struct task_struct *task;
 
 		memset(&stats, 0, sizeof(stats));
 		idr_for_each(&file->object_idr, per_file_stats, &stats);
+		/*
+		 * Although we have a valid reference on file->pid, that does
+		 * not guarantee that the task_struct who called get_pid() is
+		 * still alive (e.g. get_pid(current) => fork() => exit()).
+		 * Therefore, we need to protect this ->comm access using RCU.
+		 */
+		rcu_read_lock();
+		task = pid_task(file->pid, PIDTYPE_PID);
 		seq_printf(m, "%s: %u objects, %zu bytes (%zu active, %zu inactive, %zu unbound)\n",
-			   get_pid_task(file->pid, PIDTYPE_PID)->comm,
+			   task ? task->comm : "<unknown>",
 			   stats.count,
 			   stats.total,
 			   stats.active,
 			   stats.inactive,
 			   stats.unbound);
+		rcu_read_unlock();
 	}
 
 	mutex_unlock(&dev->struct_mutex);
@@ -564,10 +572,12 @@ static int i915_gem_seqno_info(struct seq_file *m, void *data)
 	ret = mutex_lock_interruptible(&dev->struct_mutex);
 	if (ret)
 		return ret;
+	intel_runtime_pm_get(dev_priv);
 
 	for_each_ring(ring, dev_priv, i)
 		i915_ring_seqno_info(m, ring);
 
+	intel_runtime_pm_put(dev_priv);
 	mutex_unlock(&dev->struct_mutex);
 
 	return 0;
@@ -585,6 +595,7 @@ static int i915_interrupt_info(struct seq_file *m, void *data)
 	ret = mutex_lock_interruptible(&dev->struct_mutex);
 	if (ret)
 		return ret;
+	intel_runtime_pm_get(dev_priv);
 
 	if (INTEL_INFO(dev)->gen >= 8) {
 		int i;
@@ -711,6 +722,7 @@ static int i915_interrupt_info(struct seq_file *m, void *data)
 		}
 		i915_ring_seqno_info(m, ring);
 	}
+	intel_runtime_pm_put(dev_priv);
 	mutex_unlock(&dev->struct_mutex);
 
 	return 0;
@@ -904,9 +916,11 @@ static int i915_rstdby_delays(struct seq_file *m, void *unused)
 	ret = mutex_lock_interruptible(&dev->struct_mutex);
 	if (ret)
 		return ret;
+	intel_runtime_pm_get(dev_priv);
 
 	crstanddelay = I915_READ16(CRSTANDVID);
 
+	intel_runtime_pm_put(dev_priv);
 	mutex_unlock(&dev->struct_mutex);
 
 	seq_printf(m, "w/ctx: %d, w/o ctx: %d\n", (crstanddelay >> 8) & 0x3f, (crstanddelay & 0x3f));
@@ -919,7 +933,9 @@ static int i915_cur_delayinfo(struct seq_file *m, void *unused)
 	struct drm_info_node *node = (struct drm_info_node *) m->private;
 	struct drm_device *dev = node->minor->dev;
 	drm_i915_private_t *dev_priv = dev->dev_private;
-	int ret;
+	int ret = 0;
+
+	intel_runtime_pm_get(dev_priv);
 
 	flush_delayed_work(&dev_priv->rps.delayed_resume_work);
 
@@ -945,9 +961,9 @@ static int i915_cur_delayinfo(struct seq_file *m, void *unused)
 		/* RPSTAT1 is in the GT power well */
 		ret = mutex_lock_interruptible(&dev->struct_mutex);
 		if (ret)
-			return ret;
+			goto out;
 
-		gen6_gt_force_wake_get(dev_priv);
+		gen6_gt_force_wake_get(dev_priv, FORCEWAKE_ALL);
 
 		reqf = I915_READ(GEN6_RPNSWREQ);
 		reqf &= ~GEN6_TURBO_DISABLE;
@@ -970,7 +986,7 @@ static int i915_cur_delayinfo(struct seq_file *m, void *unused)
 			cagf = (rpstat & GEN6_CAGF_MASK) >> GEN6_CAGF_SHIFT;
 		cagf *= GT_FREQUENCY_MULTIPLIER;
 
-		gen6_gt_force_wake_put(dev_priv);
+		gen6_gt_force_wake_put(dev_priv, FORCEWAKE_ALL);
 		mutex_unlock(&dev->struct_mutex);
 
 		seq_printf(m, "GT_PERF_STATUS: 0x%08x\n", gt_perf_status);
@@ -1018,23 +1034,24 @@ static int i915_cur_delayinfo(struct seq_file *m, void *unused)
 		seq_printf(m, "PUNIT_REG_GPU_FREQ_STS: 0x%08x\n", freq_sts);
 		seq_printf(m, "DDR freq: %d MHz\n", dev_priv->mem_freq);
 
-		val = vlv_punit_read(dev_priv, PUNIT_FUSE_BUS1);
+		val = valleyview_rps_max_freq(dev_priv);
 		seq_printf(m, "max GPU freq: %d MHz\n",
-			   vlv_gpu_freq(dev_priv->mem_freq, val));
+			   vlv_gpu_freq(dev_priv, val));
 
-		val = vlv_punit_read(dev_priv, PUNIT_REG_GPU_LFM);
+		val = valleyview_rps_min_freq(dev_priv);
 		seq_printf(m, "min GPU freq: %d MHz\n",
-			   vlv_gpu_freq(dev_priv->mem_freq, val));
+			   vlv_gpu_freq(dev_priv, val));
 
 		seq_printf(m, "current GPU freq: %d MHz\n",
-			   vlv_gpu_freq(dev_priv->mem_freq,
-					(freq_sts >> 8) & 0xff));
+			   vlv_gpu_freq(dev_priv, (freq_sts >> 8) & 0xff));
 		mutex_unlock(&dev_priv->rps.hw_lock);
 	} else {
 		seq_puts(m, "no P-state info available\n");
 	}
 
-	return 0;
+out:
+	intel_runtime_pm_put(dev_priv);
+	return ret;
 }
 
 static int i915_delayfreq_table(struct seq_file *m, void *unused)
@@ -1048,6 +1065,7 @@ static int i915_delayfreq_table(struct seq_file *m, void *unused)
 	ret = mutex_lock_interruptible(&dev->struct_mutex);
 	if (ret)
 		return ret;
+	intel_runtime_pm_get(dev_priv);
 
 	for (i = 0; i < 16; i++) {
 		delayfreq = I915_READ(PXVFREQ_BASE + i * 4);
@@ -1055,6 +1073,8 @@ static int i915_delayfreq_table(struct seq_file *m, void *unused)
 			   (delayfreq & PXVFREQ_PX_MASK) >> PXVFREQ_PX_SHIFT);
 	}
 
+	intel_runtime_pm_put(dev_priv);
+
 	mutex_unlock(&dev->struct_mutex);
 
 	return 0;
@@ -1076,12 +1096,14 @@ static int i915_inttoext_table(struct seq_file *m, void *unused)
 	ret = mutex_lock_interruptible(&dev->struct_mutex);
 	if (ret)
 		return ret;
+	intel_runtime_pm_get(dev_priv);
 
 	for (i = 1; i <= 32; i++) {
 		inttoext = I915_READ(INTTOEXT_BASE_ILK + i * 4);
 		seq_printf(m, "INTTOEXT%02d: 0x%08x\n", i, inttoext);
 	}
 
+	intel_runtime_pm_put(dev_priv);
 	mutex_unlock(&dev->struct_mutex);
 
 	return 0;
@@ -1099,11 +1121,13 @@ static int ironlake_drpc_info(struct seq_file *m)
 	ret = mutex_lock_interruptible(&dev->struct_mutex);
 	if (ret)
 		return ret;
+	intel_runtime_pm_get(dev_priv);
 
 	rgvmodectl = I915_READ(MEMMODECTL);
 	rstdbyctl = I915_READ(RSTDBYCTL);
 	crstandvid = I915_READ16(CRSTANDVID);
 
+	intel_runtime_pm_put(dev_priv);
 	mutex_unlock(&dev->struct_mutex);
 
 	seq_printf(m, "HD boost: %s\n", (rgvmodectl & MEMMODE_BOOST_EN) ?
@@ -1154,6 +1178,50 @@ static int ironlake_drpc_info(struct seq_file *m)
 	return 0;
 }
 
+static int vlv_drpc_info(struct seq_file *m)
+{
+
+	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;
+	u32 rpmodectl1, rcctl1;
+	unsigned fw_rendercount = 0, fw_mediacount = 0;
+
+	rpmodectl1 = I915_READ(GEN6_RP_CONTROL);
+	rcctl1 = I915_READ(GEN6_RC_CONTROL);
+
+	seq_printf(m, "Video Turbo Mode: %s\n",
+		   yesno(rpmodectl1 & GEN6_RP_MEDIA_TURBO));
+	seq_printf(m, "Turbo enabled: %s\n",
+		   yesno(rpmodectl1 & GEN6_RP_ENABLE));
+	seq_printf(m, "HW control enabled: %s\n",
+		   yesno(rpmodectl1 & GEN6_RP_ENABLE));
+	seq_printf(m, "SW control enabled: %s\n",
+		   yesno((rpmodectl1 & GEN6_RP_MEDIA_MODE_MASK) ==
+			  GEN6_RP_MEDIA_SW_MODE));
+	seq_printf(m, "RC6 Enabled: %s\n",
+		   yesno(rcctl1 & (GEN7_RC_CTL_TO_MODE |
+					GEN6_RC_CTL_EI_MODE(1))));
+	seq_printf(m, "Render Power Well: %s\n",
+			(I915_READ(VLV_GTLC_PW_STATUS) &
+				VLV_GTLC_PW_RENDER_STATUS_MASK) ? "Up" : "Down");
+	seq_printf(m, "Media Power Well: %s\n",
+			(I915_READ(VLV_GTLC_PW_STATUS) &
+				VLV_GTLC_PW_MEDIA_STATUS_MASK) ? "Up" : "Down");
+
+	spin_lock_irq(&dev_priv->uncore.lock);
+	fw_rendercount = dev_priv->uncore.fw_rendercount;
+	fw_mediacount = dev_priv->uncore.fw_mediacount;
+	spin_unlock_irq(&dev_priv->uncore.lock);
+
+	seq_printf(m, "Forcewake Render Count = %u\n", fw_rendercount);
+	seq_printf(m, "Forcewake Media Count = %u\n", fw_mediacount);
+
+
+	return 0;
+}
+
+
 static int gen6_drpc_info(struct seq_file *m)
 {
 
@@ -1167,6 +1235,7 @@ static int gen6_drpc_info(struct seq_file *m)
 	ret = mutex_lock_interruptible(&dev->struct_mutex);
 	if (ret)
 		return ret;
+	intel_runtime_pm_get(dev_priv);
 
 	spin_lock_irq(&dev_priv->uncore.lock);
 	forcewake_count = dev_priv->uncore.forcewake_count;
@@ -1192,6 +1261,8 @@ static int gen6_drpc_info(struct seq_file *m)
 	sandybridge_pcode_read(dev_priv, GEN6_PCODE_READ_RC6VIDS, &rc6vids);
 	mutex_unlock(&dev_priv->rps.hw_lock);
 
+	intel_runtime_pm_put(dev_priv);
+
 	seq_printf(m, "Video Turbo Mode: %s\n",
 		   yesno(rpmodectl1 & GEN6_RP_MEDIA_TURBO));
 	seq_printf(m, "HW control enabled: %s\n",
@@ -1256,7 +1327,9 @@ static int i915_drpc_info(struct seq_file *m, void *unused)
 	struct drm_info_node *node = (struct drm_info_node *) m->private;
 	struct drm_device *dev = node->minor->dev;
 
-	if (IS_GEN6(dev) || IS_GEN7(dev))
+	if (IS_VALLEYVIEW(dev))
+		return vlv_drpc_info(m);
+	else if (IS_GEN6(dev) || IS_GEN7(dev))
 		return gen6_drpc_info(m);
 	else
 		return ironlake_drpc_info(m);
@@ -1268,7 +1341,7 @@ static int i915_fbc_status(struct seq_file *m, void *unused)
 	struct drm_device *dev = node->minor->dev;
 	drm_i915_private_t *dev_priv = dev->dev_private;
 
-	if (!I915_HAS_FBC(dev)) {
+	if (!HAS_FBC(dev)) {
 		seq_puts(m, "FBC unsupported on this chipset\n");
 		return 0;
 	}
@@ -1330,7 +1403,7 @@ static int i915_ips_status(struct seq_file *m, void *unused)
 		return 0;
 	}
 
-	if (I915_READ(IPS_CTL) & IPS_ENABLE)
+	if (IS_BROADWELL(dev) || I915_READ(IPS_CTL) & IPS_ENABLE)
 		seq_puts(m, "enabled\n");
 	else
 		seq_puts(m, "disabled\n");
@@ -1406,6 +1479,7 @@ static int i915_ring_freq_table(struct seq_file *m, void *unused)
 	ret = mutex_lock_interruptible(&dev_priv->rps.hw_lock);
 	if (ret)
 		return ret;
+	intel_runtime_pm_get(dev_priv);
 
 	seq_puts(m, "GPU freq (MHz)\tEffective CPU freq (MHz)\tEffective Ring freq (MHz)\n");
 
@@ -1422,6 +1496,7 @@ static int i915_ring_freq_table(struct seq_file *m, void *unused)
 			   ((ia_freq >> 8) & 0xff) * 100);
 	}
 
+	intel_runtime_pm_put(dev_priv);
 	mutex_unlock(&dev_priv->rps.hw_lock);
 
 	return 0;
@@ -1437,8 +1512,10 @@ static int i915_gfxec(struct seq_file *m, void *unused)
 	ret = mutex_lock_interruptible(&dev->struct_mutex);
 	if (ret)
 		return ret;
+	intel_runtime_pm_get(dev_priv);
 
 	seq_printf(m, "GFXEC: %ld\n", (unsigned long)I915_READ(0x112f4));
+	intel_runtime_pm_put(dev_priv);
 
 	mutex_unlock(&dev->struct_mutex);
 
@@ -1565,13 +1642,21 @@ static int i915_gen6_forcewake_count_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_i915_private *dev_priv = dev->dev_private;
-	unsigned forcewake_count;
+	unsigned forcewake_count = 0, fw_rendercount = 0, fw_mediacount = 0;
 
 	spin_lock_irq(&dev_priv->uncore.lock);
-	forcewake_count = dev_priv->uncore.forcewake_count;
+	if (IS_VALLEYVIEW(dev)) {
+		fw_rendercount = dev_priv->uncore.fw_rendercount;
+		fw_mediacount = dev_priv->uncore.fw_mediacount;
+	} else
+		forcewake_count = dev_priv->uncore.forcewake_count;
 	spin_unlock_irq(&dev_priv->uncore.lock);
 
-	seq_printf(m, "forcewake count = %u\n", forcewake_count);
+	if (IS_VALLEYVIEW(dev)) {
+		seq_printf(m, "fw_rendercount = %u\n", fw_rendercount);
+		seq_printf(m, "fw_mediacount = %u\n", fw_mediacount);
+	} else
+		seq_printf(m, "forcewake count = %u\n", forcewake_count);
 
 	return 0;
 }
@@ -1610,6 +1695,7 @@ static int i915_swizzle_info(struct seq_file *m, void *data)
 	ret = mutex_lock_interruptible(&dev->struct_mutex);
 	if (ret)
 		return ret;
+	intel_runtime_pm_get(dev_priv);
 
 	seq_printf(m, "bit6 swizzle for X-tiling = %s\n",
 		   swizzle_string(dev_priv->mm.bit_6_swizzle_x));
@@ -1641,6 +1727,7 @@ static int i915_swizzle_info(struct seq_file *m, void *data)
 		seq_printf(m, "DISP_ARB_CTL = 0x%08x\n",
 			   I915_READ(DISP_ARB_CTL));
 	}
+	intel_runtime_pm_put(dev_priv);
 	mutex_unlock(&dev->struct_mutex);
 
 	return 0;
@@ -1701,16 +1788,19 @@ static int i915_ppgtt_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_i915_private *dev_priv = dev->dev_private;
 
 	int ret = mutex_lock_interruptible(&dev->struct_mutex);
 	if (ret)
 		return ret;
+	intel_runtime_pm_get(dev_priv);
 
 	if (INTEL_INFO(dev)->gen >= 8)
 		gen8_ppgtt_info(m, dev);
 	else if (INTEL_INFO(dev)->gen >= 6)
 		gen6_ppgtt_info(m, dev);
 
+	intel_runtime_pm_put(dev_priv);
 	mutex_unlock(&dev->struct_mutex);
 
 	return 0;
@@ -1735,28 +1825,28 @@ static int i915_dpio_info(struct seq_file *m, void *data)
 
 	seq_printf(m, "DPIO_CTL: 0x%08x\n", I915_READ(DPIO_CTL));
 
-	seq_printf(m, "DPIO_DIV_A: 0x%08x\n",
-		   vlv_dpio_read(dev_priv, PIPE_A, _DPIO_DIV_A));
-	seq_printf(m, "DPIO_DIV_B: 0x%08x\n",
-		   vlv_dpio_read(dev_priv, PIPE_A, _DPIO_DIV_B));
+	seq_printf(m, "DPIO PLL DW3 CH0 : 0x%08x\n",
+		   vlv_dpio_read(dev_priv, PIPE_A, VLV_PLL_DW3(0)));
+	seq_printf(m, "DPIO PLL DW3 CH1: 0x%08x\n",
+		   vlv_dpio_read(dev_priv, PIPE_A, VLV_PLL_DW3(1)));
 
-	seq_printf(m, "DPIO_REFSFR_A: 0x%08x\n",
-		   vlv_dpio_read(dev_priv, PIPE_A, _DPIO_REFSFR_A));
-	seq_printf(m, "DPIO_REFSFR_B: 0x%08x\n",
-		   vlv_dpio_read(dev_priv, PIPE_A, _DPIO_REFSFR_B));
+	seq_printf(m, "DPIO PLL DW5 CH0: 0x%08x\n",
+		   vlv_dpio_read(dev_priv, PIPE_A, VLV_PLL_DW5(0)));
+	seq_printf(m, "DPIO PLL DW5 CH1: 0x%08x\n",
+		   vlv_dpio_read(dev_priv, PIPE_A, VLV_PLL_DW5(1)));
 
-	seq_printf(m, "DPIO_CORE_CLK_A: 0x%08x\n",
-		   vlv_dpio_read(dev_priv, PIPE_A, _DPIO_CORE_CLK_A));
-	seq_printf(m, "DPIO_CORE_CLK_B: 0x%08x\n",
-		   vlv_dpio_read(dev_priv, PIPE_A, _DPIO_CORE_CLK_B));
+	seq_printf(m, "DPIO PLL DW7 CH0: 0x%08x\n",
+		   vlv_dpio_read(dev_priv, PIPE_A, VLV_PLL_DW7(0)));
+	seq_printf(m, "DPIO PLL DW7 CH1: 0x%08x\n",
+		   vlv_dpio_read(dev_priv, PIPE_A, VLV_PLL_DW7(1)));
 
-	seq_printf(m, "DPIO_LPF_COEFF_A: 0x%08x\n",
-		   vlv_dpio_read(dev_priv, PIPE_A, _DPIO_LPF_COEFF_A));
-	seq_printf(m, "DPIO_LPF_COEFF_B: 0x%08x\n",
-		   vlv_dpio_read(dev_priv, PIPE_A, _DPIO_LPF_COEFF_B));
+	seq_printf(m, "DPIO PLL DW10 CH0: 0x%08x\n",
+		   vlv_dpio_read(dev_priv, PIPE_A, VLV_PLL_DW10(0)));
+	seq_printf(m, "DPIO PLL DW10 CH1: 0x%08x\n",
+		   vlv_dpio_read(dev_priv, PIPE_A, VLV_PLL_DW10(1)));
 
 	seq_printf(m, "DPIO_FASTCLK_DISABLE: 0x%08x\n",
-		   vlv_dpio_read(dev_priv, PIPE_A, DPIO_FASTCLK_DISABLE));
+		   vlv_dpio_read(dev_priv, PIPE_A, VLV_CMN_DW0));
 
 	mutex_unlock(&dev_priv->dpio_lock);
 
@@ -1784,6 +1874,8 @@ static int i915_edp_psr_status(struct seq_file *m, void *data)
 	u32 psrperf = 0;
 	bool enabled = false;
 
+	intel_runtime_pm_get(dev_priv);
+
 	seq_printf(m, "Sink_Support: %s\n", yesno(dev_priv->psr.sink_support));
 	seq_printf(m, "Source_OK: %s\n", yesno(dev_priv->psr.source_ok));
 
@@ -1796,6 +1888,7 @@ static int i915_edp_psr_status(struct seq_file *m, void *data)
 			EDP_PSR_PERF_CNT_MASK;
 	seq_printf(m, "Performance_Counter: %u\n", psrperf);
 
+	intel_runtime_pm_put(dev_priv);
 	return 0;
 }
 
@@ -1845,6 +1938,76 @@ static int i915_pc8_status(struct seq_file *m, void *unused)
 	return 0;
 }
 
+static const char *power_domain_str(enum intel_display_power_domain domain)
+{
+	switch (domain) {
+	case POWER_DOMAIN_PIPE_A:
+		return "PIPE_A";
+	case POWER_DOMAIN_PIPE_B:
+		return "PIPE_B";
+	case POWER_DOMAIN_PIPE_C:
+		return "PIPE_C";
+	case POWER_DOMAIN_PIPE_A_PANEL_FITTER:
+		return "PIPE_A_PANEL_FITTER";
+	case POWER_DOMAIN_PIPE_B_PANEL_FITTER:
+		return "PIPE_B_PANEL_FITTER";
+	case POWER_DOMAIN_PIPE_C_PANEL_FITTER:
+		return "PIPE_C_PANEL_FITTER";
+	case POWER_DOMAIN_TRANSCODER_A:
+		return "TRANSCODER_A";
+	case POWER_DOMAIN_TRANSCODER_B:
+		return "TRANSCODER_B";
+	case POWER_DOMAIN_TRANSCODER_C:
+		return "TRANSCODER_C";
+	case POWER_DOMAIN_TRANSCODER_EDP:
+		return "TRANSCODER_EDP";
+	case POWER_DOMAIN_VGA:
+		return "VGA";
+	case POWER_DOMAIN_AUDIO:
+		return "AUDIO";
+	case POWER_DOMAIN_INIT:
+		return "INIT";
+	default:
+		WARN_ON(1);
+		return "?";
+	}
+}
+
+static int i915_power_domain_info(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 i915_power_domains *power_domains = &dev_priv->power_domains;
+	int i;
+
+	mutex_lock(&power_domains->lock);
+
+	seq_printf(m, "%-25s %s\n", "Power well/domain", "Use count");
+	for (i = 0; i < power_domains->power_well_count; i++) {
+		struct i915_power_well *power_well;
+		enum intel_display_power_domain power_domain;
+
+		power_well = &power_domains->power_wells[i];
+		seq_printf(m, "%-25s %d\n", power_well->name,
+			   power_well->count);
+
+		for (power_domain = 0; power_domain < POWER_DOMAIN_NUM;
+		     power_domain++) {
+			if (!(BIT(power_domain) & power_well->domains))
+				continue;
+
+			seq_printf(m, "  %-23s %d\n",
+				 power_domain_str(power_domain),
+				 power_domains->domain_use_count[power_domain]);
+		}
+	}
+
+	mutex_unlock(&power_domains->lock);
+
+	return 0;
+}
+
 struct pipe_crc_info {
 	const char *name;
 	struct drm_device *dev;
@@ -1857,6 +2020,9 @@ static int i915_pipe_crc_open(struct inode *inode, struct file *filep)
 	struct drm_i915_private *dev_priv = info->dev->dev_private;
 	struct intel_pipe_crc *pipe_crc = &dev_priv->pipe_crc[info->pipe];
 
+	if (info->pipe >= INTEL_INFO(info->dev)->num_pipes)
+		return -ENODEV;
+
 	spin_lock_irq(&pipe_crc->lock);
 
 	if (pipe_crc->opened) {
@@ -2005,8 +2171,8 @@ static int i915_pipe_crc_create(struct dentry *root, struct drm_minor *minor,
 	info->dev = dev;
 	ent = debugfs_create_file(info->name, S_IRUGO, root, info,
 				  &i915_pipe_crc_fops);
-	if (IS_ERR(ent))
-		return PTR_ERR(ent);
+	if (!ent)
+		return -ENOMEM;
 
 	return drm_add_fake_info_node(minor, ent, info);
 }
@@ -2347,7 +2513,7 @@ static int pipe_crc_set_source(struct drm_device *dev, enum pipe pipe,
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_pipe_crc *pipe_crc = &dev_priv->pipe_crc[pipe];
-	u32 val;
+	u32 val = 0; /* shut up gcc */
 	int ret;
 
 	if (pipe_crc->source == source)
@@ -2742,7 +2908,7 @@ i915_drop_caches_set(void *data, u64 val)
 	struct i915_vma *vma, *x;
 	int ret;
 
-	DRM_DEBUG_DRIVER("Dropping caches: 0x%08llx\n", val);
+	DRM_DEBUG("Dropping caches: 0x%08llx\n", val);
 
 	/* No need to check and wait for gpu resets, only libdrm auto-restarts
 	 * on ioctls on -EAGAIN. */
@@ -2810,8 +2976,7 @@ i915_max_freq_get(void *data, u64 *val)
 		return ret;
 
 	if (IS_VALLEYVIEW(dev))
-		*val = vlv_gpu_freq(dev_priv->mem_freq,
-				    dev_priv->rps.max_delay);
+		*val = vlv_gpu_freq(dev_priv, dev_priv->rps.max_delay);
 	else
 		*val = dev_priv->rps.max_delay * GT_FREQUENCY_MULTIPLIER;
 	mutex_unlock(&dev_priv->rps.hw_lock);
@@ -2841,9 +3006,9 @@ i915_max_freq_set(void *data, u64 val)
 	 * Turbo will still be enabled, but won't go above the set value.
 	 */
 	if (IS_VALLEYVIEW(dev)) {
-		val = vlv_freq_opcode(dev_priv->mem_freq, val);
+		val = vlv_freq_opcode(dev_priv, val);
 		dev_priv->rps.max_delay = val;
-		gen6_set_rps(dev, val);
+		valleyview_set_rps(dev, val);
 	} else {
 		do_div(val, GT_FREQUENCY_MULTIPLIER);
 		dev_priv->rps.max_delay = val;
@@ -2876,8 +3041,7 @@ i915_min_freq_get(void *data, u64 *val)
 		return ret;
 
 	if (IS_VALLEYVIEW(dev))
-		*val = vlv_gpu_freq(dev_priv->mem_freq,
-				    dev_priv->rps.min_delay);
+		*val = vlv_gpu_freq(dev_priv, dev_priv->rps.min_delay);
 	else
 		*val = dev_priv->rps.min_delay * GT_FREQUENCY_MULTIPLIER;
 	mutex_unlock(&dev_priv->rps.hw_lock);
@@ -2907,7 +3071,7 @@ i915_min_freq_set(void *data, u64 val)
 	 * Turbo will still be enabled, but won't go below the set value.
 	 */
 	if (IS_VALLEYVIEW(dev)) {
-		val = vlv_freq_opcode(dev_priv->mem_freq, val);
+		val = vlv_freq_opcode(dev_priv, val);
 		dev_priv->rps.min_delay = val;
 		valleyview_set_rps(dev, val);
 	} else {
@@ -2938,8 +3102,11 @@ i915_cache_sharing_get(void *data, u64 *val)
 	ret = mutex_lock_interruptible(&dev->struct_mutex);
 	if (ret)
 		return ret;
+	intel_runtime_pm_get(dev_priv);
 
 	snpcr = I915_READ(GEN6_MBCUNIT_SNPCR);
+
+	intel_runtime_pm_put(dev_priv);
 	mutex_unlock(&dev_priv->dev->struct_mutex);
 
 	*val = (snpcr & GEN6_MBC_SNPCR_MASK) >> GEN6_MBC_SNPCR_SHIFT;
@@ -2960,6 +3127,7 @@ i915_cache_sharing_set(void *data, u64 val)
 	if (val > 3)
 		return -EINVAL;
 
+	intel_runtime_pm_get(dev_priv);
 	DRM_DEBUG_DRIVER("Manually setting uncore sharing to %llu\n", val);
 
 	/* Update the cache sharing policy here as well */
@@ -2968,6 +3136,7 @@ i915_cache_sharing_set(void *data, u64 val)
 	snpcr |= (val << GEN6_MBC_SNPCR_SHIFT);
 	I915_WRITE(GEN6_MBCUNIT_SNPCR, snpcr);
 
+	intel_runtime_pm_put(dev_priv);
 	return 0;
 }
 
@@ -2983,7 +3152,8 @@ static int i915_forcewake_open(struct inode *inode, struct file *file)
 	if (INTEL_INFO(dev)->gen < 6)
 		return 0;
 
-	gen6_gt_force_wake_get(dev_priv);
+	intel_runtime_pm_get(dev_priv);
+	gen6_gt_force_wake_get(dev_priv, FORCEWAKE_ALL);
 
 	return 0;
 }
@@ -2996,7 +3166,8 @@ static int i915_forcewake_release(struct inode *inode, struct file *file)
 	if (INTEL_INFO(dev)->gen < 6)
 		return 0;
 
-	gen6_gt_force_wake_put(dev_priv);
+	gen6_gt_force_wake_put(dev_priv, FORCEWAKE_ALL);
+	intel_runtime_pm_put(dev_priv);
 
 	return 0;
 }
@@ -3016,8 +3187,8 @@ static int i915_forcewake_create(struct dentry *root, struct drm_minor *minor)
 				  S_IRUSR,
 				  root, dev,
 				  &i915_forcewake_fops);
-	if (IS_ERR(ent))
-		return PTR_ERR(ent);
+	if (!ent)
+		return -ENOMEM;
 
 	return drm_add_fake_info_node(minor, ent, &i915_forcewake_fops);
 }
@@ -3034,8 +3205,8 @@ static int i915_debugfs_create(struct dentry *root,
 				  S_IRUGO | S_IWUSR,
 				  root, dev,
 				  fops);
-	if (IS_ERR(ent))
-		return PTR_ERR(ent);
+	if (!ent)
+		return -ENOMEM;
 
 	return drm_add_fake_info_node(minor, ent, fops);
 }
@@ -3079,6 +3250,7 @@ static const struct drm_info_list i915_debugfs_list[] = {
 	{"i915_edp_psr_status", i915_edp_psr_status, 0},
 	{"i915_energy_uJ", i915_energy_uJ, 0},
 	{"i915_pc8_status", i915_pc8_status, 0},
+	{"i915_power_domain_info", i915_power_domain_info, 0},
 };
 #define I915_DEBUGFS_ENTRIES ARRAY_SIZE(i915_debugfs_list)
 
@@ -3102,10 +3274,10 @@ static const struct i915_debugfs_files {
 void intel_display_crc_init(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	int i;
+	enum pipe pipe;
 
-	for (i = 0; i < INTEL_INFO(dev)->num_pipes; i++) {
-		struct intel_pipe_crc *pipe_crc = &dev_priv->pipe_crc[i];
+	for_each_pipe(pipe) {
+		struct intel_pipe_crc *pipe_crc = &dev_priv->pipe_crc[pipe];
 
 		pipe_crc->opened = false;
 		spin_lock_init(&pipe_crc->lock);
@@ -3164,5 +3336,3 @@ void i915_debugfs_cleanup(struct drm_minor *minor)
 		drm_debugfs_remove_files(info_list, 1, minor);
 	}
 }
-
-#endif /* CONFIG_DEBUG_FS */
diff --git a/drivers/gpu/drm/i915/i915_dma.c b/drivers/gpu/drm/i915/i915_dma.c
index 5c648425c1e0..15a74f979b4b 100644
--- a/drivers/gpu/drm/i915/i915_dma.c
+++ b/drivers/gpu/drm/i915/i915_dma.c
@@ -42,6 +42,8 @@
 #include <linux/vga_switcheroo.h>
 #include <linux/slab.h>
 #include <acpi/video.h>
+#include <linux/pm.h>
+#include <linux/pm_runtime.h>
 
 #define LP_RING(d) (&((struct drm_i915_private *)(d))->ring[RCS])
 
@@ -791,7 +793,7 @@ static int i915_wait_irq(struct drm_device * dev, int irq_nr)
 		master_priv->sarea_priv->perf_boxes |= I915_BOX_WAIT;
 
 	if (ring->irq_get(ring)) {
-		DRM_WAIT_ON(ret, ring->irq_queue, 3 * DRM_HZ,
+		DRM_WAIT_ON(ret, ring->irq_queue, 3 * HZ,
 			    READ_BREADCRUMB(dev_priv) >= irq_nr);
 		ring->irq_put(ring);
 	} else if (wait_for(READ_BREADCRUMB(dev_priv) >= irq_nr, 3000))
@@ -828,7 +830,7 @@ static int i915_irq_emit(struct drm_device *dev, void *data,
 	result = i915_emit_irq(dev);
 	mutex_unlock(&dev->struct_mutex);
 
-	if (DRM_COPY_TO_USER(emit->irq_seq, &result, sizeof(int))) {
+	if (copy_to_user(emit->irq_seq, &result, sizeof(int))) {
 		DRM_ERROR("copy_to_user\n");
 		return -EFAULT;
 	}
@@ -1016,8 +1018,8 @@ static int i915_getparam(struct drm_device *dev, void *data,
 		return -EINVAL;
 	}
 
-	if (DRM_COPY_TO_USER(param->value, &value, sizeof(int))) {
-		DRM_ERROR("DRM_COPY_TO_USER failed\n");
+	if (copy_to_user(param->value, &value, sizeof(int))) {
+		DRM_ERROR("copy_to_user failed\n");
 		return -EFAULT;
 	}
 
@@ -1411,7 +1413,7 @@ void i915_master_destroy(struct drm_device *dev, struct drm_master *master)
 	master->driver_priv = NULL;
 }
 
-#ifdef CONFIG_DRM_I915_FBDEV
+#if IS_ENABLED(CONFIG_FB)
 static void i915_kick_out_firmware_fb(struct drm_i915_private *dev_priv)
 {
 	struct apertures_struct *ap;
@@ -1484,6 +1486,10 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags)
 		return -ENODEV;
 	}
 
+	/* UMS needs agp support. */
+	if (!drm_core_check_feature(dev, DRIVER_MODESET) && !dev->agp)
+		return -EINVAL;
+
 	dev_priv = kzalloc(sizeof(*dev_priv), GFP_KERNEL);
 	if (dev_priv == NULL)
 		return -ENOMEM;
@@ -1494,7 +1500,7 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags)
 
 	spin_lock_init(&dev_priv->irq_lock);
 	spin_lock_init(&dev_priv->gpu_error.lock);
-	spin_lock_init(&dev_priv->backlight.lock);
+	spin_lock_init(&dev_priv->backlight_lock);
 	spin_lock_init(&dev_priv->uncore.lock);
 	spin_lock_init(&dev_priv->mm.object_stat_lock);
 	mutex_init(&dev_priv->dpio_lock);
@@ -1639,8 +1645,7 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags)
 			goto out_gem_unload;
 	}
 
-	if (HAS_POWER_WELL(dev))
-		intel_power_domains_init(dev);
+	intel_power_domains_init(dev);
 
 	if (drm_core_check_feature(dev, DRIVER_MODESET)) {
 		ret = i915_load_modeset_init(dev);
@@ -1664,11 +1669,12 @@ int i915_driver_load(struct drm_device *dev, unsigned long flags)
 	if (IS_GEN5(dev))
 		intel_gpu_ips_init(dev_priv);
 
+	intel_init_runtime_pm(dev_priv);
+
 	return 0;
 
 out_power_well:
-	if (HAS_POWER_WELL(dev))
-		intel_power_domains_remove(dev);
+	intel_power_domains_remove(dev);
 	drm_vblank_cleanup(dev);
 out_gem_unload:
 	if (dev_priv->mm.inactive_shrinker.scan_objects)
@@ -1679,6 +1685,7 @@ out_gem_unload:
 
 	intel_teardown_gmbus(dev);
 	intel_teardown_mchbar(dev);
+	pm_qos_remove_request(&dev_priv->pm_qos);
 	destroy_workqueue(dev_priv->wq);
 out_mtrrfree:
 	arch_phys_wc_del(dev_priv->gtt.mtrr);
@@ -1704,25 +1711,27 @@ int i915_driver_unload(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int ret;
 
+	ret = i915_gem_suspend(dev);
+	if (ret) {
+		DRM_ERROR("failed to idle hardware: %d\n", ret);
+		return ret;
+	}
+
+	intel_fini_runtime_pm(dev_priv);
+
 	intel_gpu_ips_teardown();
 
-	if (HAS_POWER_WELL(dev)) {
-		/* The i915.ko module is still not prepared to be loaded when
-		 * the power well is not enabled, so just enable it in case
-		 * we're going to unload/reload. */
-		intel_display_set_init_power(dev, true);
-		intel_power_domains_remove(dev);
-	}
+	/* The i915.ko module is still not prepared to be loaded when
+	 * the power well is not enabled, so just enable it in case
+	 * we're going to unload/reload. */
+	intel_display_set_init_power(dev, true);
+	intel_power_domains_remove(dev);
 
 	i915_teardown_sysfs(dev);
 
 	if (dev_priv->mm.inactive_shrinker.scan_objects)
 		unregister_shrinker(&dev_priv->mm.inactive_shrinker);
 
-	ret = i915_gem_suspend(dev);
-	if (ret)
-		DRM_ERROR("failed to idle hardware: %d\n", ret);
-
 	io_mapping_free(dev_priv->gtt.mappable);
 	arch_phys_wc_del(dev_priv->gtt.mtrr);
 
@@ -1777,7 +1786,6 @@ int i915_driver_unload(struct drm_device *dev)
 
 	list_del(&dev_priv->gtt.base.global_link);
 	WARN_ON(!list_empty(&dev_priv->vm_list));
-	drm_mm_takedown(&dev_priv->gtt.base.mm);
 
 	drm_vblank_cleanup(dev);
 
@@ -1910,6 +1918,7 @@ const struct drm_ioctl_desc i915_ioctls[] = {
 	DRM_IOCTL_DEF_DRV(I915_GEM_CONTEXT_CREATE, i915_gem_context_create_ioctl, DRM_UNLOCKED|DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(I915_GEM_CONTEXT_DESTROY, i915_gem_context_destroy_ioctl, DRM_UNLOCKED|DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(I915_REG_READ, i915_reg_read_ioctl, DRM_UNLOCKED|DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF_DRV(I915_GET_RESET_STATS, i915_get_reset_stats_ioctl, DRM_UNLOCKED|DRM_RENDER_ALLOW),
 };
 
 int i915_max_ioctl = DRM_ARRAY_SIZE(i915_ioctls);
diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c
index 5b7b7e06cb3a..04f1f02c4019 100644
--- a/drivers/gpu/drm/i915/i915_drv.c
+++ b/drivers/gpu/drm/i915/i915_drv.c
@@ -59,7 +59,7 @@ MODULE_PARM_DESC(powersave,
 		"Enable powersavings, fbc, downclocking, etc. (default: true)");
 
 int i915_semaphores __read_mostly = -1;
-module_param_named(semaphores, i915_semaphores, int, 0600);
+module_param_named(semaphores, i915_semaphores, int, 0400);
 MODULE_PARM_DESC(semaphores,
 		"Use semaphores for inter-ring sync (default: -1 (use per-chip defaults))");
 
@@ -114,7 +114,7 @@ MODULE_PARM_DESC(enable_hangcheck,
 		"(default: true)");
 
 int i915_enable_ppgtt __read_mostly = -1;
-module_param_named(i915_enable_ppgtt, i915_enable_ppgtt, int, 0600);
+module_param_named(i915_enable_ppgtt, i915_enable_ppgtt, int, 0400);
 MODULE_PARM_DESC(i915_enable_ppgtt,
 		"Enable PPGTT (default: true)");
 
@@ -155,7 +155,6 @@ MODULE_PARM_DESC(prefault_disable,
 		"Disable page prefaulting for pread/pwrite/reloc (default:false). For developers only.");
 
 static struct drm_driver driver;
-extern int intel_agp_enabled;
 
 static const struct intel_device_info intel_i830_info = {
 	.gen = 2, .is_mobile = 1, .cursor_needs_physical = 1, .num_pipes = 2,
@@ -173,6 +172,7 @@ static const struct intel_device_info intel_i85x_info = {
 	.gen = 2, .is_i85x = 1, .is_mobile = 1, .num_pipes = 2,
 	.cursor_needs_physical = 1,
 	.has_overlay = 1, .overlay_needs_physical = 1,
+	.has_fbc = 1,
 	.ring_mask = RENDER_RING,
 };
 
@@ -192,6 +192,7 @@ static const struct intel_device_info intel_i915gm_info = {
 	.cursor_needs_physical = 1,
 	.has_overlay = 1, .overlay_needs_physical = 1,
 	.supports_tv = 1,
+	.has_fbc = 1,
 	.ring_mask = RENDER_RING,
 };
 static const struct intel_device_info intel_i945g_info = {
@@ -204,6 +205,7 @@ static const struct intel_device_info intel_i945gm_info = {
 	.has_hotplug = 1, .cursor_needs_physical = 1,
 	.has_overlay = 1, .overlay_needs_physical = 1,
 	.supports_tv = 1,
+	.has_fbc = 1,
 	.ring_mask = RENDER_RING,
 };
 
@@ -265,6 +267,7 @@ static const struct intel_device_info intel_ironlake_m_info = {
 static const struct intel_device_info intel_sandybridge_d_info = {
 	.gen = 6, .num_pipes = 2,
 	.need_gfx_hws = 1, .has_hotplug = 1,
+	.has_fbc = 1,
 	.ring_mask = RENDER_RING | BSD_RING | BLT_RING,
 	.has_llc = 1,
 };
@@ -280,6 +283,7 @@ static const struct intel_device_info intel_sandybridge_m_info = {
 #define GEN7_FEATURES  \
 	.gen = 7, .num_pipes = 3, \
 	.need_gfx_hws = 1, .has_hotplug = 1, \
+	.has_fbc = 1, \
 	.ring_mask = RENDER_RING | BSD_RING | BLT_RING, \
 	.has_llc = 1
 
@@ -292,7 +296,6 @@ static const struct intel_device_info intel_ivybridge_m_info = {
 	GEN7_FEATURES,
 	.is_ivybridge = 1,
 	.is_mobile = 1,
-	.has_fbc = 1,
 };
 
 static const struct intel_device_info intel_ivybridge_q_info = {
@@ -307,6 +310,7 @@ static const struct intel_device_info intel_valleyview_m_info = {
 	.num_pipes = 2,
 	.is_valleyview = 1,
 	.display_mmio_offset = VLV_DISPLAY_BASE,
+	.has_fbc = 0, /* legal, last one wins */
 	.has_llc = 0, /* legal, last one wins */
 };
 
@@ -315,6 +319,7 @@ static const struct intel_device_info intel_valleyview_d_info = {
 	.num_pipes = 2,
 	.is_valleyview = 1,
 	.display_mmio_offset = VLV_DISPLAY_BASE,
+	.has_fbc = 0, /* legal, last one wins */
 	.has_llc = 0, /* legal, last one wins */
 };
 
@@ -332,12 +337,10 @@ static const struct intel_device_info intel_haswell_m_info = {
 	.is_mobile = 1,
 	.has_ddi = 1,
 	.has_fpga_dbg = 1,
-	.has_fbc = 1,
 	.ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING,
 };
 
 static const struct intel_device_info intel_broadwell_d_info = {
-	.is_preliminary = 1,
 	.gen = 8, .num_pipes = 3,
 	.need_gfx_hws = 1, .has_hotplug = 1,
 	.ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING,
@@ -346,7 +349,6 @@ static const struct intel_device_info intel_broadwell_d_info = {
 };
 
 static const struct intel_device_info intel_broadwell_m_info = {
-	.is_preliminary = 1,
 	.gen = 8, .is_mobile = 1, .num_pipes = 3,
 	.need_gfx_hws = 1, .has_hotplug = 1,
 	.ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING,
@@ -476,12 +478,12 @@ check_next:
 bool i915_semaphore_is_enabled(struct drm_device *dev)
 {
 	if (INTEL_INFO(dev)->gen < 6)
-		return 0;
+		return false;
 
 	/* Until we get further testing... */
 	if (IS_GEN8(dev)) {
 		WARN_ON(!i915_preliminary_hw_support);
-		return 0;
+		return false;
 	}
 
 	if (i915_semaphores >= 0)
@@ -493,7 +495,7 @@ bool i915_semaphore_is_enabled(struct drm_device *dev)
 		return false;
 #endif
 
-	return 1;
+	return true;
 }
 
 static int i915_drm_freeze(struct drm_device *dev)
@@ -501,6 +503,8 @@ static int i915_drm_freeze(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_crtc *crtc;
 
+	intel_runtime_pm_get(dev_priv);
+
 	/* ignore lid events during suspend */
 	mutex_lock(&dev_priv->modeset_restore_lock);
 	dev_priv->modeset_restore = MODESET_SUSPENDED;
@@ -688,6 +692,8 @@ static int __i915_drm_thaw(struct drm_device *dev, bool restore_gtt_mappings)
 	mutex_lock(&dev_priv->modeset_restore_lock);
 	dev_priv->modeset_restore = MODESET_DONE;
 	mutex_unlock(&dev_priv->modeset_restore_lock);
+
+	intel_runtime_pm_put(dev_priv);
 	return error;
 }
 
@@ -762,14 +768,14 @@ int i915_reset(struct drm_device *dev)
 		DRM_INFO("Simulated gpu hang, resetting stop_rings\n");
 		dev_priv->gpu_error.stop_rings = 0;
 		if (ret == -ENODEV) {
-			DRM_ERROR("Reset not implemented, but ignoring "
-				  "error for simulated gpu hangs\n");
+			DRM_INFO("Reset not implemented, but ignoring "
+				 "error for simulated gpu hangs\n");
 			ret = 0;
 		}
 	}
 
 	if (ret) {
-		DRM_ERROR("Failed to reset chip.\n");
+		DRM_ERROR("Failed to reset chip: %i\n", ret);
 		mutex_unlock(&dev->struct_mutex);
 		return ret;
 	}
@@ -790,12 +796,9 @@ int i915_reset(struct drm_device *dev)
 	 */
 	if (drm_core_check_feature(dev, DRIVER_MODESET) ||
 			!dev_priv->ums.mm_suspended) {
-		bool hw_contexts_disabled = dev_priv->hw_contexts_disabled;
 		dev_priv->ums.mm_suspended = 0;
 
 		ret = i915_gem_init_hw(dev);
-		if (!hw_contexts_disabled && dev_priv->hw_contexts_disabled)
-			DRM_ERROR("HW contexts didn't survive reset\n");
 		mutex_unlock(&dev->struct_mutex);
 		if (ret) {
 			DRM_ERROR("Failed hw init on reset %d\n", ret);
@@ -831,17 +834,7 @@ static int i915_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	if (PCI_FUNC(pdev->devfn))
 		return -ENODEV;
 
-	/* We've managed to ship a kms-enabled ddx that shipped with an XvMC
-	 * implementation for gen3 (and only gen3) that used legacy drm maps
-	 * (gasp!) to share buffers between X and the client. Hence we need to
-	 * keep around the fake agp stuff for gen3, even when kms is enabled. */
-	if (intel_info->gen != 3) {
-		driver.driver_features &=
-			~(DRIVER_USE_AGP | DRIVER_REQUIRE_AGP);
-	} else if (!intel_agp_enabled) {
-		DRM_ERROR("drm/i915 can't work without intel_agp module!\n");
-		return -ENODEV;
-	}
+	driver.driver_features &= ~(DRIVER_USE_AGP);
 
 	return drm_get_pci_dev(pdev, ent, &driver);
 }
@@ -915,6 +908,49 @@ static int i915_pm_poweroff(struct device *dev)
 	return i915_drm_freeze(drm_dev);
 }
 
+static int i915_runtime_suspend(struct device *device)
+{
+	struct pci_dev *pdev = to_pci_dev(device);
+	struct drm_device *dev = pci_get_drvdata(pdev);
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	WARN_ON(!HAS_RUNTIME_PM(dev));
+
+	DRM_DEBUG_KMS("Suspending device\n");
+
+	i915_gem_release_all_mmaps(dev_priv);
+
+	del_timer_sync(&dev_priv->gpu_error.hangcheck_timer);
+	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.
+	 */
+	intel_opregion_notify_adapter(dev, PCI_D1);
+
+	return 0;
+}
+
+static int i915_runtime_resume(struct device *device)
+{
+	struct pci_dev *pdev = to_pci_dev(device);
+	struct drm_device *dev = pci_get_drvdata(pdev);
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	WARN_ON(!HAS_RUNTIME_PM(dev));
+
+	DRM_DEBUG_KMS("Resuming device\n");
+
+	intel_opregion_notify_adapter(dev, PCI_D0);
+	dev_priv->pm.suspended = false;
+
+	return 0;
+}
+
 static const struct dev_pm_ops i915_pm_ops = {
 	.suspend = i915_pm_suspend,
 	.resume = i915_pm_resume,
@@ -922,6 +958,8 @@ static const struct dev_pm_ops i915_pm_ops = {
 	.thaw = i915_pm_thaw,
 	.poweroff = i915_pm_poweroff,
 	.restore = i915_pm_resume,
+	.runtime_suspend = i915_runtime_suspend,
+	.runtime_resume = i915_runtime_resume,
 };
 
 static const struct vm_operations_struct i915_gem_vm_ops = {
@@ -949,7 +987,7 @@ static struct drm_driver driver = {
 	 * deal with them for Intel hardware.
 	 */
 	.driver_features =
-	    DRIVER_USE_AGP | DRIVER_REQUIRE_AGP |
+	    DRIVER_USE_AGP |
 	    DRIVER_HAVE_IRQ | DRIVER_IRQ_SHARED | DRIVER_GEM | DRIVER_PRIME |
 	    DRIVER_RENDER,
 	.load = i915_driver_load,
@@ -1024,14 +1062,24 @@ static int __init i915_init(void)
 		driver.driver_features &= ~DRIVER_MODESET;
 #endif
 
-	if (!(driver.driver_features & DRIVER_MODESET))
+	if (!(driver.driver_features & DRIVER_MODESET)) {
 		driver.get_vblank_timestamp = NULL;
+#ifndef CONFIG_DRM_I915_UMS
+		/* Silently fail loading to not upset userspace. */
+		return 0;
+#endif
+	}
 
 	return drm_pci_init(&driver, &i915_pci_driver);
 }
 
 static void __exit i915_exit(void)
 {
+#ifndef CONFIG_DRM_I915_UMS
+	if (!(driver.driver_features & DRIVER_MODESET))
+		return; /* Never loaded a driver. */
+#endif
+
 	drm_pci_exit(&driver, &i915_pci_driver);
 }
 
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 1caa5e34fbe3..4a2bf8e3f739 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -89,6 +89,18 @@ enum port {
 };
 #define port_name(p) ((p) + 'A')
 
+#define I915_NUM_PHYS_VLV 1
+
+enum dpio_channel {
+	DPIO_CH0,
+	DPIO_CH1
+};
+
+enum dpio_phy {
+	DPIO_PHY0,
+	DPIO_PHY1
+};
+
 enum intel_display_power_domain {
 	POWER_DOMAIN_PIPE_A,
 	POWER_DOMAIN_PIPE_B,
@@ -101,6 +113,7 @@ enum intel_display_power_domain {
 	POWER_DOMAIN_TRANSCODER_C,
 	POWER_DOMAIN_TRANSCODER_EDP,
 	POWER_DOMAIN_VGA,
+	POWER_DOMAIN_AUDIO,
 	POWER_DOMAIN_INIT,
 
 	POWER_DOMAIN_NUM,
@@ -310,13 +323,14 @@ struct drm_i915_error_state {
 	u32 instps[I915_NUM_RINGS];
 	u32 extra_instdone[I915_NUM_INSTDONE_REG];
 	u32 seqno[I915_NUM_RINGS];
-	u64 bbaddr;
+	u64 bbaddr[I915_NUM_RINGS];
 	u32 fault_reg[I915_NUM_RINGS];
 	u32 done_reg;
 	u32 faddr[I915_NUM_RINGS];
 	u64 fence[I915_MAX_NUM_FENCES];
 	struct timeval time;
 	struct drm_i915_error_ring {
+		bool valid;
 		struct drm_i915_error_object {
 			int page_count;
 			u32 gtt_offset;
@@ -351,6 +365,7 @@ struct drm_i915_error_state {
 	enum intel_ring_hangcheck_action hangcheck_action[I915_NUM_RINGS];
 };
 
+struct intel_connector;
 struct intel_crtc_config;
 struct intel_crtc;
 struct intel_limit;
@@ -358,7 +373,7 @@ struct dpll;
 
 struct drm_i915_display_funcs {
 	bool (*fbc_enabled)(struct drm_device *dev);
-	void (*enable_fbc)(struct drm_crtc *crtc, unsigned long interval);
+	void (*enable_fbc)(struct drm_crtc *crtc);
 	void (*disable_fbc)(struct drm_device *dev);
 	int (*get_display_clock_speed)(struct drm_device *dev);
 	int (*get_fifo_size)(struct drm_device *dev, int plane);
@@ -413,11 +428,20 @@ struct drm_i915_display_funcs {
 	/* render clock increase/decrease */
 	/* display clock increase/decrease */
 	/* pll clock increase/decrease */
+
+	int (*setup_backlight)(struct intel_connector *connector);
+	uint32_t (*get_backlight)(struct intel_connector *connector);
+	void (*set_backlight)(struct intel_connector *connector,
+			      uint32_t level);
+	void (*disable_backlight)(struct intel_connector *connector);
+	void (*enable_backlight)(struct intel_connector *connector);
 };
 
 struct intel_uncore_funcs {
-	void (*force_wake_get)(struct drm_i915_private *dev_priv);
-	void (*force_wake_put)(struct drm_i915_private *dev_priv);
+	void (*force_wake_get)(struct drm_i915_private *dev_priv,
+							int fw_engine);
+	void (*force_wake_put)(struct drm_i915_private *dev_priv,
+							int fw_engine);
 
 	uint8_t  (*mmio_readb)(struct drm_i915_private *dev_priv, off_t offset, bool trace);
 	uint16_t (*mmio_readw)(struct drm_i915_private *dev_priv, off_t offset, bool trace);
@@ -442,6 +466,9 @@ struct intel_uncore {
 	unsigned fifo_count;
 	unsigned forcewake_count;
 
+	unsigned fw_rendercount;
+	unsigned fw_mediacount;
+
 	struct delayed_work force_wake_work;
 };
 
@@ -669,7 +696,6 @@ struct i915_fbc {
 		struct delayed_work work;
 		struct drm_crtc *crtc;
 		struct drm_framebuffer *fb;
-		int interval;
 	} *fbc_work;
 
 	enum no_fbc_reason {
@@ -708,7 +734,6 @@ enum intel_sbi_destination {
 #define QUIRK_PIPEA_FORCE (1<<0)
 #define QUIRK_LVDS_SSC_DISABLE (1<<1)
 #define QUIRK_INVERT_BRIGHTNESS (1<<2)
-#define QUIRK_NO_PCH_PWM_ENABLE (1<<3)
 
 struct intel_fbdev;
 struct intel_fbc_work;
@@ -761,8 +786,6 @@ struct i915_suspend_saved_registers {
 	u32 saveBLC_PWM_CTL;
 	u32 saveBLC_PWM_CTL2;
 	u32 saveBLC_HIST_CTL_B;
-	u32 saveBLC_PWM_CTL_B;
-	u32 saveBLC_PWM_CTL2_B;
 	u32 saveBLC_CPU_PWM_CTL;
 	u32 saveBLC_CPU_PWM_CTL2;
 	u32 saveFPB0;
@@ -932,21 +955,29 @@ struct intel_ilk_power_mgmt {
 
 /* Power well structure for haswell */
 struct i915_power_well {
+	const char *name;
+	bool always_on;
 	/* power well enable/disable usage count */
 	int count;
+	unsigned long domains;
+	void *data;
+	void (*set)(struct drm_device *dev, struct i915_power_well *power_well,
+		    bool enable);
+	bool (*is_enabled)(struct drm_device *dev,
+			   struct i915_power_well *power_well);
 };
 
-#define I915_MAX_POWER_WELLS 1
-
 struct i915_power_domains {
 	/*
 	 * Power wells needed for initialization at driver init and suspend
 	 * time are on. They are kept on until after the first modeset.
 	 */
 	bool init_power_on;
+	int power_well_count;
 
 	struct mutex lock;
-	struct i915_power_well power_wells[I915_MAX_POWER_WELLS];
+	int domain_use_count[POWER_DOMAIN_NUM];
+	struct i915_power_well *power_wells;
 };
 
 struct i915_dri1_state {
@@ -1077,34 +1108,30 @@ struct i915_gpu_error {
 	unsigned long missed_irq_rings;
 
 	/**
-	 * State variable and reset counter controlling the reset flow
+	 * State variable controlling the reset flow and count
 	 *
-	 * Upper bits are for the reset counter.  This counter is used by the
-	 * wait_seqno code to race-free noticed that a reset event happened and
-	 * that it needs to restart the entire ioctl (since most likely the
-	 * seqno it waited for won't ever signal anytime soon).
+	 * This is a counter which gets incremented when reset is triggered,
+	 * and again when reset has been handled. So odd values (lowest bit set)
+	 * means that reset is in progress and even values that
+	 * (reset_counter >> 1):th reset was successfully completed.
+	 *
+	 * If reset is not completed succesfully, the I915_WEDGE bit is
+	 * set meaning that hardware is terminally sour and there is no
+	 * recovery. All waiters on the reset_queue will be woken when
+	 * that happens.
+	 *
+	 * This counter is used by the wait_seqno code to notice that reset
+	 * event happened and it needs to restart the entire ioctl (since most
+	 * likely the seqno it waited for won't ever signal anytime soon).
 	 *
 	 * This is important for lock-free wait paths, where no contended lock
 	 * naturally enforces the correct ordering between the bail-out of the
 	 * waiter and the gpu reset work code.
-	 *
-	 * Lowest bit controls the reset state machine: Set means a reset is in
-	 * progress. This state will (presuming we don't have any bugs) decay
-	 * into either unset (successful reset) or the special WEDGED value (hw
-	 * terminally sour). All waiters on the reset_queue will be woken when
-	 * that happens.
 	 */
 	atomic_t reset_counter;
 
-	/**
-	 * Special values/flags for reset_counter
-	 *
-	 * Note that the code relies on
-	 * 	I915_WEDGED & I915_RESET_IN_PROGRESS_FLAG
-	 * being true.
-	 */
 #define I915_RESET_IN_PROGRESS_FLAG	1
-#define I915_WEDGED			0xffffffff
+#define I915_WEDGED			(1 << 31)
 
 	/**
 	 * Waitqueue to signal when the reset has completed. Used by clients
@@ -1158,6 +1185,11 @@ struct intel_vbt_data {
 	int edp_bpp;
 	struct edp_power_seq edp_pps;
 
+	struct {
+		u16 pwm_freq_hz;
+		bool active_low_pwm;
+	} backlight;
+
 	/* MIPI DSI */
 	struct {
 		u16 panel_id;
@@ -1184,7 +1216,7 @@ struct intel_wm_level {
 	uint32_t fbc_val;
 };
 
-struct hsw_wm_values {
+struct ilk_wm_values {
 	uint32_t wm_pipe[3];
 	uint32_t wm_lp[3];
 	uint32_t wm_lp_spr[3];
@@ -1262,6 +1294,10 @@ struct i915_package_c8 {
 	} regsave;
 };
 
+struct i915_runtime_pm {
+	bool suspended;
+};
+
 enum intel_pipe_crc_source {
 	INTEL_PIPE_CRC_SOURCE_NONE,
 	INTEL_PIPE_CRC_SOURCE_PLANE1,
@@ -1366,15 +1402,9 @@ typedef struct drm_i915_private {
 
 	/* overlay */
 	struct intel_overlay *overlay;
-	unsigned int sprite_scaling_enabled;
 
-	/* backlight */
-	struct {
-		int level;
-		bool enabled;
-		spinlock_t lock; /* bl registers and the above bl fields */
-		struct backlight_device *device;
-	} backlight;
+	/* backlight registers and fields in struct intel_panel */
+	spinlock_t backlight_lock;
 
 	/* LVDS info */
 	bool no_aux_handshake;
@@ -1426,6 +1456,7 @@ typedef struct drm_i915_private {
 	int num_shared_dpll;
 	struct intel_shared_dpll shared_dplls[I915_NUM_PLLS];
 	struct intel_ddi_plls ddi_plls;
+	int dpio_phy_iosf_port[I915_NUM_PHYS_VLV];
 
 	/* Reclocking support */
 	bool render_reclock_avail;
@@ -1470,7 +1501,6 @@ typedef struct drm_i915_private {
 	struct drm_property *broadcast_rgb_property;
 	struct drm_property *force_audio_property;
 
-	bool hw_contexts_disabled;
 	uint32_t hw_context_size;
 	struct list_head context_list;
 
@@ -1492,11 +1522,13 @@ typedef struct drm_i915_private {
 		uint16_t cur_latency[5];
 
 		/* current hardware state */
-		struct hsw_wm_values hw;
+		struct ilk_wm_values hw;
 	} wm;
 
 	struct i915_package_c8 pc8;
 
+	struct i915_runtime_pm pm;
+
 	/* Old dri1 support infrastructure, beware the dragons ya fools entering
 	 * here! */
 	struct i915_dri1_state dri1;
@@ -1813,15 +1845,15 @@ struct drm_i915_file_private {
 
 #define HAS_FW_BLC(dev) (INTEL_INFO(dev)->gen > 2)
 #define HAS_PIPE_CXSR(dev) (INTEL_INFO(dev)->has_pipe_cxsr)
-#define I915_HAS_FBC(dev) (INTEL_INFO(dev)->has_fbc)
+#define HAS_FBC(dev) (INTEL_INFO(dev)->has_fbc)
 
 #define HAS_IPS(dev)		(IS_ULT(dev) || IS_BROADWELL(dev))
 
 #define HAS_DDI(dev)		(INTEL_INFO(dev)->has_ddi)
-#define HAS_POWER_WELL(dev)	(IS_HASWELL(dev) || IS_BROADWELL(dev))
 #define HAS_FPGA_DBG_UNCLAIMED(dev)	(INTEL_INFO(dev)->has_fpga_dbg)
 #define HAS_PSR(dev)		(IS_HASWELL(dev) || IS_BROADWELL(dev))
 #define HAS_PC8(dev)		(IS_HASWELL(dev)) /* XXX HSW:ULX */
+#define HAS_RUNTIME_PM(dev)	(IS_HASWELL(dev))
 
 #define INTEL_PCH_DEVICE_ID_MASK		0xff00
 #define INTEL_PCH_IBX_DEVICE_ID_TYPE		0x3b00
@@ -1911,7 +1943,6 @@ extern void intel_hpd_init(struct drm_device *dev);
 extern void intel_uncore_sanitize(struct drm_device *dev);
 extern void intel_uncore_early_sanitize(struct drm_device *dev);
 extern void intel_uncore_init(struct drm_device *dev);
-extern void intel_uncore_clear_errors(struct drm_device *dev);
 extern void intel_uncore_check_errors(struct drm_device *dev);
 extern void intel_uncore_fini(struct drm_device *dev);
 
@@ -1987,6 +2018,7 @@ void i915_gem_object_unpin(struct drm_i915_gem_object *obj);
 int __must_check i915_vma_unbind(struct i915_vma *vma);
 int __must_check i915_gem_object_ggtt_unbind(struct drm_i915_gem_object *obj);
 int i915_gem_object_put_pages(struct drm_i915_gem_object *obj);
+void i915_gem_release_all_mmaps(struct drm_i915_private *dev_priv);
 void i915_gem_release_mmap(struct drm_i915_gem_object *obj);
 void i915_gem_lastclose(struct drm_device *dev);
 
@@ -2063,12 +2095,17 @@ int __must_check i915_gem_check_wedge(struct i915_gpu_error *error,
 static inline bool i915_reset_in_progress(struct i915_gpu_error *error)
 {
 	return unlikely(atomic_read(&error->reset_counter)
-			& I915_RESET_IN_PROGRESS_FLAG);
+			& (I915_RESET_IN_PROGRESS_FLAG | I915_WEDGED));
 }
 
 static inline bool i915_terminally_wedged(struct i915_gpu_error *error)
 {
-	return atomic_read(&error->reset_counter) == I915_WEDGED;
+	return atomic_read(&error->reset_counter) & I915_WEDGED;
+}
+
+static inline u32 i915_reset_count(struct i915_gpu_error *error)
+{
+	return ((atomic_read(&error->reset_counter) & ~I915_WEDGED) + 1) / 2;
 }
 
 void i915_gem_reset(struct drm_device *dev);
@@ -2180,7 +2217,7 @@ i915_gem_obj_ggtt_pin(struct drm_i915_gem_object *obj,
 }
 
 /* i915_gem_context.c */
-void i915_gem_context_init(struct drm_device *dev);
+int __must_check i915_gem_context_init(struct drm_device *dev);
 void i915_gem_context_fini(struct drm_device *dev);
 void i915_gem_context_close(struct drm_device *dev, struct drm_file *file);
 int i915_switch_context(struct intel_ring_buffer *ring,
@@ -2399,6 +2436,8 @@ extern int intel_enable_rc6(const struct drm_device *dev);
 extern bool i915_semaphore_is_enabled(struct drm_device *dev);
 int i915_reg_read_ioctl(struct drm_device *dev, void *data,
 			struct drm_file *file);
+int i915_get_reset_stats_ioctl(struct drm_device *dev, void *data,
+			       struct drm_file *file);
 
 /* overlay */
 extern struct intel_overlay_error_state *intel_overlay_capture_error_state(struct drm_device *dev);
@@ -2414,8 +2453,8 @@ extern void intel_display_print_error_state(struct drm_i915_error_state_buf *e,
  * must be set to prevent GT core from power down and stale values being
  * returned.
  */
-void gen6_gt_force_wake_get(struct drm_i915_private *dev_priv);
-void gen6_gt_force_wake_put(struct drm_i915_private *dev_priv);
+void gen6_gt_force_wake_get(struct drm_i915_private *dev_priv, int fw_engine);
+void gen6_gt_force_wake_put(struct drm_i915_private *dev_priv, int fw_engine);
 
 int sandybridge_pcode_read(struct drm_i915_private *dev_priv, u8 mbox, u32 *val);
 int sandybridge_pcode_write(struct drm_i915_private *dev_priv, u8 mbox, u32 val);
@@ -2430,6 +2469,8 @@ u32 vlv_cck_read(struct drm_i915_private *dev_priv, u32 reg);
 void vlv_cck_write(struct drm_i915_private *dev_priv, u32 reg, u32 val);
 u32 vlv_ccu_read(struct drm_i915_private *dev_priv, u32 reg);
 void vlv_ccu_write(struct drm_i915_private *dev_priv, u32 reg, u32 val);
+u32 vlv_bunit_read(struct drm_i915_private *dev_priv, u32 reg);
+void vlv_bunit_write(struct drm_i915_private *dev_priv, u32 reg, u32 val);
 u32 vlv_gps_core_read(struct drm_i915_private *dev_priv, u32 reg);
 void vlv_gps_core_write(struct drm_i915_private *dev_priv, u32 reg, u32 val);
 u32 vlv_dpio_read(struct drm_i915_private *dev_priv, enum pipe pipe, int reg);
@@ -2438,9 +2479,30 @@ u32 intel_sbi_read(struct drm_i915_private *dev_priv, u16 reg,
 		   enum intel_sbi_destination destination);
 void intel_sbi_write(struct drm_i915_private *dev_priv, u16 reg, u32 value,
 		     enum intel_sbi_destination destination);
+u32 vlv_flisdsi_read(struct drm_i915_private *dev_priv, u32 reg);
+void vlv_flisdsi_write(struct drm_i915_private *dev_priv, u32 reg, u32 val);
+
+int vlv_gpu_freq(struct drm_i915_private *dev_priv, int val);
+int vlv_freq_opcode(struct drm_i915_private *dev_priv, int val);
+
+void vlv_force_wake_get(struct drm_i915_private *dev_priv, int fw_engine);
+void vlv_force_wake_put(struct drm_i915_private *dev_priv, int fw_engine);
+
+#define FORCEWAKE_VLV_RENDER_RANGE_OFFSET(reg) \
+	(((reg) >= 0x2000 && (reg) < 0x4000) ||\
+	((reg) >= 0x5000 && (reg) < 0x8000) ||\
+	((reg) >= 0xB000 && (reg) < 0x12000) ||\
+	((reg) >= 0x2E000 && (reg) < 0x30000))
+
+#define FORCEWAKE_VLV_MEDIA_RANGE_OFFSET(reg)\
+	(((reg) >= 0x12000 && (reg) < 0x14000) ||\
+	((reg) >= 0x22000 && (reg) < 0x24000) ||\
+	((reg) >= 0x30000 && (reg) < 0x40000))
+
+#define FORCEWAKE_RENDER	(1 << 0)
+#define FORCEWAKE_MEDIA		(1 << 1)
+#define FORCEWAKE_ALL		(FORCEWAKE_RENDER | FORCEWAKE_MEDIA)
 
-int vlv_gpu_freq(int ddr_freq, int val);
-int vlv_freq_opcode(int ddr_freq, int val);
 
 #define I915_READ8(reg)		dev_priv->uncore.funcs.mmio_readb(dev_priv, (reg), true)
 #define I915_WRITE8(reg, val)	dev_priv->uncore.funcs.mmio_writeb(dev_priv, (reg), (val), true)
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 76d3d1ab73c6..00c836154725 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -1015,9 +1015,11 @@ static int __wait_seqno(struct intel_ring_buffer *ring, u32 seqno,
 			struct drm_i915_file_private *file_priv)
 {
 	drm_i915_private_t *dev_priv = ring->dev->dev_private;
+	const bool irq_test_in_progress =
+		ACCESS_ONCE(dev_priv->gpu_error.test_irq_rings) & intel_ring_flag(ring);
 	struct timespec before, now;
 	DEFINE_WAIT(wait);
-	long timeout_jiffies;
+	unsigned long timeout_expire;
 	int ret;
 
 	WARN(dev_priv->pc8.irqs_disabled, "IRQs disabled\n");
@@ -1025,7 +1027,7 @@ static int __wait_seqno(struct intel_ring_buffer *ring, u32 seqno,
 	if (i915_seqno_passed(ring->get_seqno(ring, true), seqno))
 		return 0;
 
-	timeout_jiffies = timeout ? timespec_to_jiffies_timeout(timeout) : 1;
+	timeout_expire = timeout ? jiffies + timespec_to_jiffies_timeout(timeout) : 0;
 
 	if (dev_priv->info->gen >= 6 && can_wait_boost(file_priv)) {
 		gen6_rps_boost(dev_priv);
@@ -1035,8 +1037,7 @@ static int __wait_seqno(struct intel_ring_buffer *ring, u32 seqno,
 					 msecs_to_jiffies(100));
 	}
 
-	if (!(dev_priv->gpu_error.test_irq_rings & intel_ring_flag(ring)) &&
-	    WARN_ON(!ring->irq_get(ring)))
+	if (!irq_test_in_progress && WARN_ON(!ring->irq_get(ring)))
 		return -ENODEV;
 
 	/* Record current time in case interrupted by signal, or wedged */
@@ -1044,7 +1045,6 @@ static int __wait_seqno(struct intel_ring_buffer *ring, u32 seqno,
 	getrawmonotonic(&before);
 	for (;;) {
 		struct timer_list timer;
-		unsigned long expire;
 
 		prepare_to_wait(&ring->irq_queue, &wait,
 				interruptible ? TASK_INTERRUPTIBLE : TASK_UNINTERRUPTIBLE);
@@ -1070,23 +1070,22 @@ static int __wait_seqno(struct intel_ring_buffer *ring, u32 seqno,
 			break;
 		}
 
-		if (timeout_jiffies <= 0) {
+		if (timeout && time_after_eq(jiffies, timeout_expire)) {
 			ret = -ETIME;
 			break;
 		}
 
 		timer.function = NULL;
 		if (timeout || missed_irq(dev_priv, ring)) {
+			unsigned long expire;
+
 			setup_timer_on_stack(&timer, fake_irq, (unsigned long)current);
-			expire = jiffies + (missed_irq(dev_priv, ring) ? 1: timeout_jiffies);
+			expire = missed_irq(dev_priv, ring) ? jiffies + 1 : timeout_expire;
 			mod_timer(&timer, expire);
 		}
 
 		io_schedule();
 
-		if (timeout)
-			timeout_jiffies = expire - jiffies;
-
 		if (timer.function) {
 			del_singleshot_timer_sync(&timer);
 			destroy_timer_on_stack(&timer);
@@ -1095,7 +1094,8 @@ static int __wait_seqno(struct intel_ring_buffer *ring, u32 seqno,
 	getrawmonotonic(&now);
 	trace_i915_gem_request_wait_end(ring, seqno);
 
-	ring->irq_put(ring);
+	if (!irq_test_in_progress)
+		ring->irq_put(ring);
 
 	finish_wait(&ring->irq_queue, &wait);
 
@@ -1380,6 +1380,8 @@ int i915_gem_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 	int ret = 0;
 	bool write = !!(vmf->flags & FAULT_FLAG_WRITE);
 
+	intel_runtime_pm_get(dev_priv);
+
 	/* We don't use vmf->pgoff since that has the fake offset */
 	page_offset = ((unsigned long)vmf->virtual_address - vma->vm_start) >>
 		PAGE_SHIFT;
@@ -1427,8 +1429,10 @@ out:
 		/* If this -EIO is due to a gpu hang, give the reset code a
 		 * chance to clean up the mess. Otherwise return the proper
 		 * SIGBUS. */
-		if (i915_terminally_wedged(&dev_priv->gpu_error))
-			return VM_FAULT_SIGBUS;
+		if (i915_terminally_wedged(&dev_priv->gpu_error)) {
+			ret = VM_FAULT_SIGBUS;
+			break;
+		}
 	case -EAGAIN:
 		/*
 		 * EAGAIN means the gpu is hung and we'll wait for the error
@@ -1443,15 +1447,38 @@ out:
 		 * EBUSY is ok: this just means that another thread
 		 * already did the job.
 		 */
-		return VM_FAULT_NOPAGE;
+		ret = VM_FAULT_NOPAGE;
+		break;
 	case -ENOMEM:
-		return VM_FAULT_OOM;
+		ret = VM_FAULT_OOM;
+		break;
 	case -ENOSPC:
-		return VM_FAULT_SIGBUS;
+		ret = VM_FAULT_SIGBUS;
+		break;
 	default:
 		WARN_ONCE(ret, "unhandled error in i915_gem_fault: %i\n", ret);
-		return VM_FAULT_SIGBUS;
+		ret = VM_FAULT_SIGBUS;
+		break;
 	}
+
+	intel_runtime_pm_put(dev_priv);
+	return ret;
+}
+
+void i915_gem_release_all_mmaps(struct drm_i915_private *dev_priv)
+{
+	struct i915_vma *vma;
+
+	/*
+	 * Only the global gtt is relevant for gtt memory mappings, so restrict
+	 * list traversal to objects bound into the global address space. Note
+	 * that the active list should be empty, but better safe than sorry.
+	 */
+	WARN_ON(!list_empty(&dev_priv->gtt.base.active_list));
+	list_for_each_entry(vma, &dev_priv->gtt.base.active_list, mm_list)
+		i915_gem_release_mmap(vma->obj);
+	list_for_each_entry(vma, &dev_priv->gtt.base.inactive_list, mm_list)
+		i915_gem_release_mmap(vma->obj);
 }
 
 /**
@@ -2303,7 +2330,7 @@ static void i915_set_reset_status(struct intel_ring_buffer *ring,
 
 	if (ring->hangcheck.action != HANGCHECK_WAIT &&
 	    i915_request_guilty(request, acthd, &inside)) {
-		DRM_ERROR("%s hung %s bo (0x%lx ctx %d) at 0x%x\n",
+		DRM_DEBUG("%s hung %s bo (0x%lx ctx %d) at 0x%x\n",
 			  ring->name,
 			  inside ? "inside" : "flushing",
 			  offset,
@@ -2361,16 +2388,6 @@ static void i915_gem_reset_ring_status(struct drm_i915_private *dev_priv,
 static void i915_gem_reset_ring_cleanup(struct drm_i915_private *dev_priv,
 					struct intel_ring_buffer *ring)
 {
-	while (!list_empty(&ring->request_list)) {
-		struct drm_i915_gem_request *request;
-
-		request = list_first_entry(&ring->request_list,
-					   struct drm_i915_gem_request,
-					   list);
-
-		i915_gem_free_request(request);
-	}
-
 	while (!list_empty(&ring->active_list)) {
 		struct drm_i915_gem_object *obj;
 
@@ -2380,6 +2397,23 @@ static void i915_gem_reset_ring_cleanup(struct drm_i915_private *dev_priv,
 
 		i915_gem_object_move_to_inactive(obj);
 	}
+
+	/*
+	 * We must free the requests after all the corresponding objects have
+	 * been moved off active lists. Which is the same order as the normal
+	 * retire_requests function does. This is important if object hold
+	 * implicit references on things like e.g. ppgtt address spaces through
+	 * the request.
+	 */
+	while (!list_empty(&ring->request_list)) {
+		struct drm_i915_gem_request *request;
+
+		request = list_first_entry(&ring->request_list,
+					   struct drm_i915_gem_request,
+					   list);
+
+		i915_gem_free_request(request);
+	}
 }
 
 void i915_gem_restore_fences(struct drm_device *dev)
@@ -2760,7 +2794,6 @@ int i915_vma_unbind(struct i915_vma *vma)
 		obj->has_aliasing_ppgtt_mapping = 0;
 	}
 	i915_gem_gtt_finish_object(obj);
-	i915_gem_object_unpin_pages(obj);
 
 	list_del(&vma->mm_list);
 	/* Avoid an unnecessary call to unbind on rebind. */
@@ -2768,7 +2801,6 @@ int i915_vma_unbind(struct i915_vma *vma)
 		obj->map_and_fenceable = true;
 
 	drm_mm_remove_node(&vma->node);
-
 	i915_gem_vma_destroy(vma);
 
 	/* Since the unbound list is global, only move to that list if
@@ -2776,6 +2808,12 @@ int i915_vma_unbind(struct i915_vma *vma)
 	if (list_empty(&obj->vma_list))
 		list_move_tail(&obj->global_list, &dev_priv->mm.unbound_list);
 
+	/* And finally now the object is completely decoupled from this vma,
+	 * we can drop its hold on the backing storage and allow it to be
+	 * reaped by the shrinker.
+	 */
+	i915_gem_object_unpin_pages(obj);
+
 	return 0;
 }
 
@@ -3068,7 +3106,7 @@ i915_find_fence_reg(struct drm_device *dev)
 	}
 
 	if (avail == NULL)
-		return NULL;
+		goto deadlock;
 
 	/* None available, try to steal one or wait for a user to finish */
 	list_for_each_entry(reg, &dev_priv->mm.fence_list, lru_list) {
@@ -3078,7 +3116,12 @@ i915_find_fence_reg(struct drm_device *dev)
 		return reg;
 	}
 
-	return NULL;
+deadlock:
+	/* Wait for completion of pending flips which consume fences */
+	if (intel_has_pending_fb_unpin(dev))
+		return ERR_PTR(-EAGAIN);
+
+	return ERR_PTR(-EDEADLK);
 }
 
 /**
@@ -3123,8 +3166,8 @@ i915_gem_object_get_fence(struct drm_i915_gem_object *obj)
 		}
 	} else if (enable) {
 		reg = i915_find_fence_reg(dev);
-		if (reg == NULL)
-			return -EDEADLK;
+		if (IS_ERR(reg))
+			return PTR_ERR(reg);
 
 		if (reg->obj) {
 			struct drm_i915_gem_object *old = reg->obj;
@@ -4179,6 +4222,8 @@ void i915_gem_free_object(struct drm_gem_object *gem_obj)
 	drm_i915_private_t *dev_priv = dev->dev_private;
 	struct i915_vma *vma, *next;
 
+	intel_runtime_pm_get(dev_priv);
+
 	trace_i915_gem_object_destroy(obj);
 
 	if (obj->phys_obj)
@@ -4223,6 +4268,8 @@ void i915_gem_free_object(struct drm_gem_object *gem_obj)
 
 	kfree(obj->bit_17);
 	i915_gem_object_free(obj);
+
+	intel_runtime_pm_put(dev_priv);
 }
 
 struct i915_vma *i915_gem_obj_to_vma(struct drm_i915_gem_object *obj,
@@ -4479,7 +4526,13 @@ i915_gem_init_hw(struct drm_device *dev)
 	 * XXX: There was some w/a described somewhere suggesting loading
 	 * contexts before PPGTT.
 	 */
-	i915_gem_context_init(dev);
+	ret = i915_gem_context_init(dev);
+	if (ret) {
+		i915_gem_cleanup_ringbuffer(dev);
+		DRM_ERROR("Context initialization failed %d\n", ret);
+		return ret;
+	}
+
 	if (dev_priv->mm.aliasing_ppgtt) {
 		ret = dev_priv->mm.aliasing_ppgtt->enable(dev);
 		if (ret) {
diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c
index b0f42b9ca037..e08acaba5402 100644
--- a/drivers/gpu/drm/i915/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/i915_gem_context.c
@@ -247,36 +247,34 @@ err_destroy:
 	return ret;
 }
 
-void i915_gem_context_init(struct drm_device *dev)
+int i915_gem_context_init(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	int ret;
 
-	if (!HAS_HW_CONTEXTS(dev)) {
-		dev_priv->hw_contexts_disabled = true;
-		DRM_DEBUG_DRIVER("Disabling HW Contexts; old hardware\n");
-		return;
-	}
+	if (!HAS_HW_CONTEXTS(dev))
+		return 0;
 
 	/* If called from reset, or thaw... we've been here already */
-	if (dev_priv->hw_contexts_disabled ||
-	    dev_priv->ring[RCS].default_context)
-		return;
+	if (dev_priv->ring[RCS].default_context)
+		return 0;
 
 	dev_priv->hw_context_size = round_up(get_context_size(dev), 4096);
 
 	if (dev_priv->hw_context_size > (1<<20)) {
-		dev_priv->hw_contexts_disabled = true;
 		DRM_DEBUG_DRIVER("Disabling HW Contexts; invalid size\n");
-		return;
+		return -E2BIG;
 	}
 
-	if (create_default_context(dev_priv)) {
-		dev_priv->hw_contexts_disabled = true;
-		DRM_DEBUG_DRIVER("Disabling HW Contexts; create failed\n");
-		return;
+	ret = create_default_context(dev_priv);
+	if (ret) {
+		DRM_DEBUG_DRIVER("Disabling HW Contexts; create failed %d\n",
+				 ret);
+		return ret;
 	}
 
 	DRM_DEBUG_DRIVER("HW context support initialized\n");
+	return 0;
 }
 
 void i915_gem_context_fini(struct drm_device *dev)
@@ -284,7 +282,7 @@ void i915_gem_context_fini(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct i915_hw_context *dctx = dev_priv->ring[RCS].default_context;
 
-	if (dev_priv->hw_contexts_disabled)
+	if (!HAS_HW_CONTEXTS(dev))
 		return;
 
 	/* The only known way to stop the gpu from accessing the hw context is
@@ -327,16 +325,16 @@ i915_gem_context_get_hang_stats(struct drm_device *dev,
 				struct drm_file *file,
 				u32 id)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_i915_file_private *file_priv = file->driver_priv;
 	struct i915_hw_context *ctx;
 
 	if (id == DEFAULT_CONTEXT_ID)
 		return &file_priv->hang_stats;
 
-	ctx = NULL;
-	if (!dev_priv->hw_contexts_disabled)
-		ctx = i915_gem_context_get(file->driver_priv, id);
+	if (!HAS_HW_CONTEXTS(dev))
+		return ERR_PTR(-ENOENT);
+
+	ctx = i915_gem_context_get(file->driver_priv, id);
 	if (ctx == NULL)
 		return ERR_PTR(-ENOENT);
 
@@ -502,8 +500,6 @@ static int do_switch(struct i915_hw_context *to)
  * @ring: ring for which we'll execute the context switch
  * @file_priv: file_priv associated with the context, may be NULL
  * @id: context id number
- * @seqno: sequence number by which the new context will be switched to
- * @flags:
  *
  * 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,
@@ -517,7 +513,7 @@ int i915_switch_context(struct intel_ring_buffer *ring,
 	struct drm_i915_private *dev_priv = ring->dev->dev_private;
 	struct i915_hw_context *to;
 
-	if (dev_priv->hw_contexts_disabled)
+	if (!HAS_HW_CONTEXTS(ring->dev))
 		return 0;
 
 	WARN_ON(!mutex_is_locked(&dev_priv->dev->struct_mutex));
@@ -542,7 +538,6 @@ int i915_switch_context(struct intel_ring_buffer *ring,
 int i915_gem_context_create_ioctl(struct drm_device *dev, void *data,
 				  struct drm_file *file)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_i915_gem_context_create *args = data;
 	struct drm_i915_file_private *file_priv = file->driver_priv;
 	struct i915_hw_context *ctx;
@@ -551,7 +546,7 @@ int i915_gem_context_create_ioctl(struct drm_device *dev, void *data,
 	if (!(dev->driver->driver_features & DRIVER_GEM))
 		return -ENODEV;
 
-	if (dev_priv->hw_contexts_disabled)
+	if (!HAS_HW_CONTEXTS(dev))
 		return -ENODEV;
 
 	ret = i915_mutex_lock_interruptible(dev);
diff --git a/drivers/gpu/drm/i915/i915_gem_evict.c b/drivers/gpu/drm/i915/i915_gem_evict.c
index 8f3adc7d0dc8..2ca280f9ee53 100644
--- a/drivers/gpu/drm/i915/i915_gem_evict.c
+++ b/drivers/gpu/drm/i915/i915_gem_evict.c
@@ -27,8 +27,10 @@
  */
 
 #include <drm/drmP.h>
-#include "i915_drv.h"
 #include <drm/i915_drm.h>
+
+#include "i915_drv.h"
+#include "intel_drv.h"
 #include "i915_trace.h"
 
 static bool
@@ -53,6 +55,7 @@ i915_gem_evict_something(struct drm_device *dev, struct i915_address_space *vm,
 	struct list_head eviction_list, unwind_list;
 	struct i915_vma *vma;
 	int ret = 0;
+	int pass = 0;
 
 	trace_i915_gem_evict(dev, min_size, alignment, mappable);
 
@@ -119,14 +122,24 @@ none:
 	/* Can we unpin some objects such as idle hw contents,
 	 * or pending flips?
 	 */
-	ret = nonblocking ? -ENOSPC : i915_gpu_idle(dev);
-	if (ret)
-		return ret;
+	if (nonblocking)
+		return -ENOSPC;
 
 	/* Only idle the GPU and repeat the search once */
-	i915_gem_retire_requests(dev);
-	nonblocking = true;
-	goto search_again;
+	if (pass++ == 0) {
+		ret = i915_gpu_idle(dev);
+		if (ret)
+			return ret;
+
+		i915_gem_retire_requests(dev);
+		goto search_again;
+	}
+
+	/* If we still have pending pageflip completions, drop
+	 * back to userspace to give our workqueues time to
+	 * acquire our locks and unpin the old scanouts.
+	 */
+	return intel_has_pending_fb_unpin(dev) ? -EAGAIN : -ENOSPC;
 
 found:
 	/* drm_mm doesn't allow any other other operations while
diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
index a3ba9a8cd687..d269ecf46e26 100644
--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
@@ -46,7 +46,7 @@ struct eb_vmas {
 };
 
 static struct eb_vmas *
-eb_create(struct drm_i915_gem_execbuffer2 *args, struct i915_address_space *vm)
+eb_create(struct drm_i915_gem_execbuffer2 *args)
 {
 	struct eb_vmas *eb = NULL;
 
@@ -252,7 +252,7 @@ relocate_entry_cpu(struct drm_i915_gem_object *obj,
 	struct drm_device *dev = obj->base.dev;
 	uint32_t page_offset = offset_in_page(reloc->offset);
 	char *vaddr;
-	int ret = -EINVAL;
+	int ret;
 
 	ret = i915_gem_object_set_to_cpu_domain(obj, true);
 	if (ret)
@@ -287,7 +287,7 @@ relocate_entry_gtt(struct drm_i915_gem_object *obj,
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	uint32_t __iomem *reloc_entry;
 	void __iomem *reloc_page;
-	int ret = -EINVAL;
+	int ret;
 
 	ret = i915_gem_object_set_to_gtt_domain(obj, true);
 	if (ret)
@@ -335,7 +335,7 @@ i915_gem_execbuffer_relocate_entry(struct drm_i915_gem_object *obj,
 	struct drm_i915_gem_object *target_i915_obj;
 	struct i915_vma *target_vma;
 	uint32_t target_offset;
-	int ret = -EINVAL;
+	int ret;
 
 	/* we've already hold a reference to all valid objects */
 	target_vma = eb_get_vma(eb, reloc->target_handle);
@@ -344,7 +344,7 @@ i915_gem_execbuffer_relocate_entry(struct drm_i915_gem_object *obj,
 	target_i915_obj = target_vma->obj;
 	target_obj = &target_vma->obj->base;
 
-	target_offset = i915_gem_obj_ggtt_offset(target_i915_obj);
+	target_offset = target_vma->node.start;
 
 	/* Sandybridge PPGTT errata: We need a global gtt mapping for MI and
 	 * pipe_control writes because the gpu doesn't properly redirect them
@@ -365,7 +365,7 @@ i915_gem_execbuffer_relocate_entry(struct drm_i915_gem_object *obj,
 			  (int) reloc->offset,
 			  reloc->read_domains,
 			  reloc->write_domain);
-		return ret;
+		return -EINVAL;
 	}
 	if (unlikely((reloc->write_domain | reloc->read_domains)
 		     & ~I915_GEM_GPU_DOMAINS)) {
@@ -376,7 +376,7 @@ i915_gem_execbuffer_relocate_entry(struct drm_i915_gem_object *obj,
 			  (int) reloc->offset,
 			  reloc->read_domains,
 			  reloc->write_domain);
-		return ret;
+		return -EINVAL;
 	}
 
 	target_obj->pending_read_domains |= reloc->read_domains;
@@ -396,14 +396,14 @@ i915_gem_execbuffer_relocate_entry(struct drm_i915_gem_object *obj,
 			  obj, reloc->target_handle,
 			  (int) reloc->offset,
 			  (int) obj->base.size);
-		return ret;
+		return -EINVAL;
 	}
 	if (unlikely(reloc->offset & 3)) {
 		DRM_DEBUG("Relocation not 4-byte aligned: "
 			  "obj %p target %d offset %d.\n",
 			  obj, reloc->target_handle,
 			  (int) reloc->offset);
-		return ret;
+		return -EINVAL;
 	}
 
 	/* We can't wait for rendering with pagefaults disabled */
@@ -491,8 +491,7 @@ i915_gem_execbuffer_relocate_vma_slow(struct i915_vma *vma,
 }
 
 static int
-i915_gem_execbuffer_relocate(struct eb_vmas *eb,
-			     struct i915_address_space *vm)
+i915_gem_execbuffer_relocate(struct eb_vmas *eb)
 {
 	struct i915_vma *vma;
 	int ret = 0;
@@ -901,6 +900,24 @@ validate_exec_list(struct drm_i915_gem_exec_object2 *exec,
 	return 0;
 }
 
+static int
+i915_gem_validate_context(struct drm_device *dev, struct drm_file *file,
+			  const u32 ctx_id)
+{
+	struct i915_ctx_hang_stats *hs;
+
+	hs = i915_gem_context_get_hang_stats(dev, file, ctx_id);
+	if (IS_ERR(hs))
+		return PTR_ERR(hs);
+
+	if (hs->banned) {
+		DRM_DEBUG("Context %u tried to submit while banned\n", ctx_id);
+		return -EIO;
+	}
+
+	return 0;
+}
+
 static void
 i915_gem_execbuffer_move_to_active(struct list_head *vmas,
 				   struct intel_ring_buffer *ring)
@@ -980,8 +997,7 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 	struct drm_i915_gem_object *batch_obj;
 	struct drm_clip_rect *cliprects = NULL;
 	struct intel_ring_buffer *ring;
-	struct i915_ctx_hang_stats *hs;
-	u32 ctx_id = i915_execbuffer2_get_context_id(*args);
+	const u32 ctx_id = i915_execbuffer2_get_context_id(*args);
 	u32 exec_start, exec_len;
 	u32 mask, flags;
 	int ret, mode, i;
@@ -1108,6 +1124,8 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 		}
 	}
 
+	intel_runtime_pm_get(dev_priv);
+
 	ret = i915_mutex_lock_interruptible(dev);
 	if (ret)
 		goto pre_mutex_err;
@@ -1118,7 +1136,13 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 		goto pre_mutex_err;
 	}
 
-	eb = eb_create(args, vm);
+	ret = i915_gem_validate_context(dev, file, ctx_id);
+	if (ret) {
+		mutex_unlock(&dev->struct_mutex);
+		goto pre_mutex_err;
+	}
+
+	eb = eb_create(args);
 	if (eb == NULL) {
 		mutex_unlock(&dev->struct_mutex);
 		ret = -ENOMEM;
@@ -1141,7 +1165,7 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 
 	/* The objects are in their final locations, apply the relocations. */
 	if (need_relocs)
-		ret = i915_gem_execbuffer_relocate(eb, vm);
+		ret = i915_gem_execbuffer_relocate(eb);
 	if (ret) {
 		if (ret == -EFAULT) {
 			ret = i915_gem_execbuffer_relocate_slow(dev, args, file, ring,
@@ -1170,17 +1194,6 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 	if (ret)
 		goto err;
 
-	hs = i915_gem_context_get_hang_stats(dev, file, ctx_id);
-	if (IS_ERR(hs)) {
-		ret = PTR_ERR(hs);
-		goto err;
-	}
-
-	if (hs->banned) {
-		ret = -EIO;
-		goto err;
-	}
-
 	ret = i915_switch_context(ring, file, ctx_id);
 	if (ret)
 		goto err;
@@ -1242,6 +1255,10 @@ err:
 
 pre_mutex_err:
 	kfree(cliprects);
+
+	/* intel_gpu_busy should also get a ref, so it will free when the device
+	 * is really idle. */
+	intel_runtime_pm_put(dev_priv);
 	return ret;
 }
 
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index 3540569948db..40a2b36b276b 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -240,10 +240,16 @@ static int gen8_ppgtt_enable(struct drm_device *dev)
 		for_each_ring(ring, dev_priv, j) {
 			ret = gen8_write_pdp(ring, i, addr);
 			if (ret)
-				return 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 void gen8_ppgtt_clear_range(struct i915_address_space *vm,
@@ -293,23 +299,23 @@ static void gen8_ppgtt_insert_entries(struct i915_address_space *vm,
 	unsigned act_pte = first_entry % GEN8_PTES_PER_PAGE;
 	struct sg_page_iter sg_iter;
 
-	pt_vaddr = kmap_atomic(&ppgtt->gen8_pt_pages[act_pt]);
+	pt_vaddr = NULL;
 	for_each_sg_page(pages->sgl, &sg_iter, pages->nents, 0) {
-		dma_addr_t page_addr;
+		if (pt_vaddr == NULL)
+			pt_vaddr = kmap_atomic(&ppgtt->gen8_pt_pages[act_pt]);
 
-		page_addr = sg_dma_address(sg_iter.sg) +
-				(sg_iter.sg_pgoffset << PAGE_SHIFT);
-		pt_vaddr[act_pte] = gen8_pte_encode(page_addr, cache_level,
-						    true);
+		pt_vaddr[act_pte] =
+			gen8_pte_encode(sg_page_iter_dma_address(&sg_iter),
+					cache_level, true);
 		if (++act_pte == GEN8_PTES_PER_PAGE) {
 			kunmap_atomic(pt_vaddr);
+			pt_vaddr = NULL;
 			act_pt++;
-			pt_vaddr = kmap_atomic(&ppgtt->gen8_pt_pages[act_pt]);
 			act_pte = 0;
-
 		}
 	}
-	kunmap_atomic(pt_vaddr);
+	if (pt_vaddr)
+		kunmap_atomic(pt_vaddr);
 }
 
 static void gen8_ppgtt_cleanup(struct i915_address_space *vm)
@@ -318,6 +324,8 @@ static void gen8_ppgtt_cleanup(struct i915_address_space *vm)
 		container_of(vm, struct i915_hw_ppgtt, base);
 	int i, j;
 
+	drm_mm_takedown(&vm->mm);
+
 	for (i = 0; i < ppgtt->num_pd_pages ; i++) {
 		if (ppgtt->pd_dma_addr[i]) {
 			pci_unmap_page(ppgtt->base.dev->pdev,
@@ -381,6 +389,8 @@ static int gen8_ppgtt_init(struct i915_hw_ppgtt *ppgtt, uint64_t size)
 	ppgtt->base.clear_range = gen8_ppgtt_clear_range;
 	ppgtt->base.insert_entries = gen8_ppgtt_insert_entries;
 	ppgtt->base.cleanup = gen8_ppgtt_cleanup;
+	ppgtt->base.start = 0;
+	ppgtt->base.total = ppgtt->num_pt_pages * GEN8_PTES_PER_PAGE * PAGE_SIZE;
 
 	BUG_ON(ppgtt->num_pd_pages > GEN8_LEGACY_PDPS);
 
@@ -573,21 +583,23 @@ static void gen6_ppgtt_insert_entries(struct i915_address_space *vm,
 	unsigned act_pte = first_entry % I915_PPGTT_PT_ENTRIES;
 	struct sg_page_iter sg_iter;
 
-	pt_vaddr = kmap_atomic(ppgtt->pt_pages[act_pt]);
+	pt_vaddr = NULL;
 	for_each_sg_page(pages->sgl, &sg_iter, pages->nents, 0) {
-		dma_addr_t page_addr;
+		if (pt_vaddr == NULL)
+			pt_vaddr = kmap_atomic(ppgtt->pt_pages[act_pt]);
 
-		page_addr = sg_page_iter_dma_address(&sg_iter);
-		pt_vaddr[act_pte] = vm->pte_encode(page_addr, cache_level, true);
+		pt_vaddr[act_pte] =
+			vm->pte_encode(sg_page_iter_dma_address(&sg_iter),
+				       cache_level, true);
 		if (++act_pte == I915_PPGTT_PT_ENTRIES) {
 			kunmap_atomic(pt_vaddr);
+			pt_vaddr = NULL;
 			act_pt++;
-			pt_vaddr = kmap_atomic(ppgtt->pt_pages[act_pt]);
 			act_pte = 0;
-
 		}
 	}
-	kunmap_atomic(pt_vaddr);
+	if (pt_vaddr)
+		kunmap_atomic(pt_vaddr);
 }
 
 static void gen6_ppgtt_cleanup(struct i915_address_space *vm)
@@ -632,6 +644,8 @@ static int gen6_ppgtt_init(struct i915_hw_ppgtt *ppgtt)
 	ppgtt->base.insert_entries = gen6_ppgtt_insert_entries;
 	ppgtt->base.cleanup = gen6_ppgtt_cleanup;
 	ppgtt->base.scratch = dev_priv->gtt.base.scratch;
+	ppgtt->base.start = 0;
+	ppgtt->base.total = GEN6_PPGTT_PD_ENTRIES * I915_PPGTT_PT_ENTRIES * PAGE_SIZE;
 	ppgtt->pt_pages = kcalloc(ppgtt->num_pd_entries, sizeof(struct page *),
 				  GFP_KERNEL);
 	if (!ppgtt->pt_pages)
@@ -1124,7 +1138,6 @@ void i915_gem_setup_global_gtt(struct drm_device *dev,
 		if (ret)
 			DRM_DEBUG_KMS("Reservation failed\n");
 		obj->has_global_gtt_mapping = 1;
-		list_add(&vma->vma_link, &obj->vma_list);
 	}
 
 	dev_priv->gtt.base.start = start;
@@ -1400,6 +1413,8 @@ static void gen6_gmch_remove(struct i915_address_space *vm)
 {
 
 	struct i915_gtt *gtt = container_of(vm, struct i915_gtt, base);
+
+	drm_mm_takedown(&vm->mm);
 	iounmap(gtt->gsm);
 	teardown_scratch_page(vm->dev);
 }
@@ -1425,6 +1440,9 @@ static int i915_gmch_probe(struct drm_device *dev,
 	dev_priv->gtt.base.clear_range = i915_ggtt_clear_range;
 	dev_priv->gtt.base.insert_entries = i915_ggtt_insert_entries;
 
+	if (unlikely(dev_priv->gtt.do_idle_maps))
+		DRM_INFO("applying Ironlake quirks for intel_iommu\n");
+
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/i915/i915_gem_stolen.c b/drivers/gpu/drm/i915/i915_gem_stolen.c
index d284d892ed94..1a24e84f2315 100644
--- a/drivers/gpu/drm/i915/i915_gem_stolen.c
+++ b/drivers/gpu/drm/i915/i915_gem_stolen.c
@@ -250,7 +250,7 @@ i915_pages_create_for_stolen(struct drm_device *dev,
 	}
 
 	sg = st->sgl;
-	sg->offset = offset;
+	sg->offset = 0;
 	sg->length = size;
 
 	sg_dma_address(sg) = (dma_addr_t)dev_priv->mm.stolen_base + offset;
@@ -420,6 +420,7 @@ i915_gem_object_create_stolen_for_preallocated(struct drm_device *dev,
 
 	list_add_tail(&obj->global_list, &dev_priv->mm.bound_list);
 	list_add_tail(&vma->mm_list, &ggtt->inactive_list);
+	i915_gem_object_pin_pages(obj);
 
 	return obj;
 
diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c
index 79dcb8f896c6..d7fd2fd2f0a5 100644
--- a/drivers/gpu/drm/i915/i915_gpu_error.c
+++ b/drivers/gpu/drm/i915/i915_gpu_error.c
@@ -239,6 +239,9 @@ static void i915_ring_error_state(struct drm_i915_error_state_buf *m,
 				  unsigned ring)
 {
 	BUG_ON(ring >= I915_NUM_RINGS); /* shut up confused gcc */
+	if (!error->ring[ring].valid)
+		return;
+
 	err_printf(m, "%s command stream:\n", ring_str(ring));
 	err_printf(m, "  HEAD: 0x%08x\n", error->head[ring]);
 	err_printf(m, "  TAIL: 0x%08x\n", error->tail[ring]);
@@ -247,12 +250,11 @@ static void i915_ring_error_state(struct drm_i915_error_state_buf *m,
 	err_printf(m, "  IPEIR: 0x%08x\n", error->ipeir[ring]);
 	err_printf(m, "  IPEHR: 0x%08x\n", error->ipehr[ring]);
 	err_printf(m, "  INSTDONE: 0x%08x\n", error->instdone[ring]);
-	if (ring == RCS && INTEL_INFO(dev)->gen >= 4)
-		err_printf(m, "  BBADDR: 0x%08llx\n", error->bbaddr);
-	if (INTEL_INFO(dev)->gen >= 4)
+	if (INTEL_INFO(dev)->gen >= 4) {
+		err_printf(m, "  BBADDR: 0x%08llx\n", error->bbaddr[ring]);
 		err_printf(m, "  BB_STATE: 0x%08x\n", error->bbstate[ring]);
-	if (INTEL_INFO(dev)->gen >= 4)
 		err_printf(m, "  INSTPS: 0x%08x\n", error->instps[ring]);
+	}
 	err_printf(m, "  INSTPM: 0x%08x\n", error->instpm[ring]);
 	err_printf(m, "  FADDR: 0x%08x\n", error->faddr[ring]);
 	if (INTEL_INFO(dev)->gen >= 6) {
@@ -294,7 +296,6 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m,
 	struct drm_device *dev = error_priv->dev;
 	drm_i915_private_t *dev_priv = dev->dev_private;
 	struct drm_i915_error_state *error = error_priv->error;
-	struct intel_ring_buffer *ring;
 	int i, j, page, offset, elt;
 
 	if (!error) {
@@ -329,7 +330,7 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m,
 	if (INTEL_INFO(dev)->gen == 7)
 		err_printf(m, "ERR_INT: 0x%08x\n", error->err_int);
 
-	for_each_ring(ring, dev_priv, i)
+	for (i = 0; i < ARRAY_SIZE(error->ring); i++)
 		i915_ring_error_state(m, dev, error, i);
 
 	if (error->active_bo)
@@ -386,8 +387,7 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m,
 			}
 		}
 
-		obj = error->ring[i].ctx;
-		if (obj) {
+		if ((obj = error->ring[i].ctx)) {
 			err_printf(m, "%s --- HW Context = 0x%08x\n",
 				   dev_priv->ring[i].name,
 				   obj->gtt_offset);
@@ -668,7 +668,8 @@ i915_error_first_batchbuffer(struct drm_i915_private *dev_priv,
 			return NULL;
 
 		obj = ring->scratch.obj;
-		if (acthd >= i915_gem_obj_ggtt_offset(obj) &&
+		if (obj != NULL &&
+		    acthd >= i915_gem_obj_ggtt_offset(obj) &&
 		    acthd < i915_gem_obj_ggtt_offset(obj) + obj->base.size)
 			return i915_error_object_create(dev_priv, obj);
 	}
@@ -725,8 +726,9 @@ static void i915_record_ring_state(struct drm_device *dev,
 		error->ipehr[ring->id] = I915_READ(RING_IPEHR(ring->mmio_base));
 		error->instdone[ring->id] = I915_READ(RING_INSTDONE(ring->mmio_base));
 		error->instps[ring->id] = I915_READ(RING_INSTPS(ring->mmio_base));
-		if (ring->id == RCS)
-			error->bbaddr = I915_READ64(BB_ADDR);
+		error->bbaddr[ring->id] = I915_READ(RING_BBADDR(ring->mmio_base));
+		if (INTEL_INFO(dev)->gen >= 8)
+			error->bbaddr[ring->id] |= (u64) I915_READ(RING_BBADDR_UDW(ring->mmio_base)) << 32;
 		error->bbstate[ring->id] = I915_READ(RING_BBSTATE(ring->mmio_base));
 	} else {
 		error->faddr[ring->id] = I915_READ(DMA_FADD_I8XX);
@@ -775,11 +777,17 @@ static void i915_gem_record_rings(struct drm_device *dev,
 				  struct drm_i915_error_state *error)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_ring_buffer *ring;
 	struct drm_i915_gem_request *request;
 	int i, count;
 
-	for_each_ring(ring, dev_priv, i) {
+	for (i = 0; i < I915_NUM_RINGS; i++) {
+		struct intel_ring_buffer *ring = &dev_priv->ring[i];
+
+		if (ring->dev == NULL)
+			continue;
+
+		error->ring[i].valid = true;
+
 		i915_record_ring_state(dev, error, ring);
 
 		error->ring[i].batchbuffer =
diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c
index f13d5edc39d5..17d8fcb1b6f7 100644
--- a/drivers/gpu/drm/i915/i915_irq.c
+++ b/drivers/gpu/drm/i915/i915_irq.c
@@ -62,7 +62,7 @@ static const u32 hpd_mask_i915[] = {
 	[HPD_PORT_D] = PORTD_HOTPLUG_INT_EN
 };
 
-static const u32 hpd_status_gen4[] = {
+static const u32 hpd_status_g4x[] = {
 	[HPD_CRT] = CRT_HOTPLUG_INT_STATUS,
 	[HPD_SDVO_B] = SDVOB_HOTPLUG_INT_STATUS_G4X,
 	[HPD_SDVO_C] = SDVOC_HOTPLUG_INT_STATUS_G4X,
@@ -600,7 +600,7 @@ static u32 i915_get_vblank_counter(struct drm_device *dev, int pipe)
 	 * Cook up a vblank counter by also checking the pixel
 	 * counter against vblank start.
 	 */
-	return ((high1 << 8) | low) + (pixel >= vbl_start);
+	return (((high1 << 8) | low) + (pixel >= vbl_start)) & 0xffffff;
 }
 
 static u32 gm45_get_vblank_counter(struct drm_device *dev, int pipe)
@@ -621,36 +621,15 @@ static u32 gm45_get_vblank_counter(struct drm_device *dev, int pipe)
 #define __raw_i915_read32(dev_priv__, reg__) readl((dev_priv__)->regs + (reg__))
 #define __raw_i915_read16(dev_priv__, reg__) readw((dev_priv__)->regs + (reg__))
 
-static bool intel_pipe_in_vblank_locked(struct drm_device *dev, enum pipe pipe)
+static bool ilk_pipe_in_vblank_locked(struct drm_device *dev, enum pipe pipe)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	uint32_t status;
-	int reg;
 
-	if (IS_VALLEYVIEW(dev)) {
-		status = pipe == PIPE_A ?
-			I915_DISPLAY_PIPE_A_VBLANK_INTERRUPT :
-			I915_DISPLAY_PIPE_B_VBLANK_INTERRUPT;
-
-		reg = VLV_ISR;
-	} else if (IS_GEN2(dev)) {
-		status = pipe == PIPE_A ?
-			I915_DISPLAY_PIPE_A_VBLANK_INTERRUPT :
-			I915_DISPLAY_PIPE_B_VBLANK_INTERRUPT;
-
-		reg = ISR;
-	} else if (INTEL_INFO(dev)->gen < 5) {
-		status = pipe == PIPE_A ?
-			I915_DISPLAY_PIPE_A_VBLANK_INTERRUPT :
-			I915_DISPLAY_PIPE_B_VBLANK_INTERRUPT;
-
-		reg = ISR;
-	} else if (INTEL_INFO(dev)->gen < 7) {
+	if (INTEL_INFO(dev)->gen < 7) {
 		status = pipe == PIPE_A ?
 			DE_PIPEA_VBLANK :
 			DE_PIPEB_VBLANK;
-
-		reg = DEISR;
 	} else {
 		switch (pipe) {
 		default:
@@ -664,18 +643,14 @@ static bool intel_pipe_in_vblank_locked(struct drm_device *dev, enum pipe pipe)
 			status = DE_PIPEC_VBLANK_IVB;
 			break;
 		}
-
-		reg = DEISR;
 	}
 
-	if (IS_GEN2(dev))
-		return __raw_i915_read16(dev_priv, reg) & status;
-	else
-		return __raw_i915_read32(dev_priv, reg) & status;
+	return __raw_i915_read32(dev_priv, DEISR) & status;
 }
 
 static int i915_get_crtc_scanoutpos(struct drm_device *dev, int pipe,
-			     int *vpos, int *hpos, ktime_t *stime, ktime_t *etime)
+				    unsigned int flags, int *vpos, int *hpos,
+				    ktime_t *stime, ktime_t *etime)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_crtc *crtc = dev_priv->pipe_to_crtc_mapping[pipe];
@@ -698,6 +673,12 @@ static int i915_get_crtc_scanoutpos(struct drm_device *dev, int pipe,
 	vbl_start = mode->crtc_vblank_start;
 	vbl_end = mode->crtc_vblank_end;
 
+	if (mode->flags & DRM_MODE_FLAG_INTERLACE) {
+		vbl_start = DIV_ROUND_UP(vbl_start, 2);
+		vbl_end /= 2;
+		vtotal /= 2;
+	}
+
 	ret |= DRM_SCANOUTPOS_VALID | DRM_SCANOUTPOS_ACCURATE;
 
 	/*
@@ -722,17 +703,42 @@ static int i915_get_crtc_scanoutpos(struct drm_device *dev, int pipe,
 		else
 			position = __raw_i915_read32(dev_priv, PIPEDSL(pipe)) & DSL_LINEMASK_GEN3;
 
-		/*
-		 * The scanline counter increments at the leading edge
-		 * of hsync, ie. it completely misses the active portion
-		 * of the line. Fix up the counter at both edges of vblank
-		 * to get a more accurate picture whether we're in vblank
-		 * or not.
-		 */
-		in_vbl = intel_pipe_in_vblank_locked(dev, pipe);
-		if ((in_vbl && position == vbl_start - 1) ||
-		    (!in_vbl && position == vbl_end - 1))
-			position = (position + 1) % vtotal;
+		if (HAS_PCH_SPLIT(dev)) {
+			/*
+			 * The scanline counter increments at the leading edge
+			 * of hsync, ie. it completely misses the active portion
+			 * of the line. Fix up the counter at both edges of vblank
+			 * to get a more accurate picture whether we're in vblank
+			 * or not.
+			 */
+			in_vbl = ilk_pipe_in_vblank_locked(dev, pipe);
+			if ((in_vbl && position == vbl_start - 1) ||
+			    (!in_vbl && position == vbl_end - 1))
+				position = (position + 1) % vtotal;
+		} else {
+			/*
+			 * ISR vblank status bits don't work the way we'd want
+			 * them to work on non-PCH platforms (for
+			 * ilk_pipe_in_vblank_locked()), and there doesn't
+			 * appear any other way to determine if we're currently
+			 * in vblank.
+			 *
+			 * Instead let's assume that we're already in vblank if
+			 * we got called from the vblank interrupt and the
+			 * scanline counter value indicates that we're on the
+			 * line just prior to vblank start. This should result
+			 * in the correct answer, unless the vblank interrupt
+			 * delivery really got delayed for almost exactly one
+			 * full frame/field.
+			 */
+			if (flags & DRM_CALLED_FROM_VBLIRQ &&
+			    position == vbl_start - 1) {
+				position = (position + 1) % vtotal;
+
+				/* Signal this correction as "applied". */
+				ret |= 0x8;
+			}
+		}
 	} else {
 		/* Have access to pixelcount since start of frame.
 		 * We can split this into vertical and horizontal
@@ -809,7 +815,8 @@ static int i915_get_vblank_timestamp(struct drm_device *dev, int pipe,
 	/* Helper routine in DRM core does all the work: */
 	return drm_calc_vbltimestamp_from_scanoutpos(dev, pipe, max_error,
 						     vblank_time, flags,
-						     crtc);
+						     crtc,
+						     &to_intel_crtc(crtc)->config.adjusted_mode);
 }
 
 static bool intel_hpd_irq_event(struct drm_device *dev,
@@ -1015,10 +1022,8 @@ static void gen6_pm_rps_work(struct work_struct *work)
 	/* sysfs frequency interfaces may have snuck in while servicing the
 	 * interrupt
 	 */
-	if (new_delay < (int)dev_priv->rps.min_delay)
-		new_delay = dev_priv->rps.min_delay;
-	if (new_delay > (int)dev_priv->rps.max_delay)
-		new_delay = dev_priv->rps.max_delay;
+	new_delay = clamp_t(int, new_delay,
+			    dev_priv->rps.min_delay, dev_priv->rps.max_delay);
 	dev_priv->rps.last_adj = new_delay - dev_priv->rps.cur_delay;
 
 	if (IS_VALLEYVIEW(dev_priv->dev))
@@ -1235,9 +1240,10 @@ static inline void intel_hpd_irq_handler(struct drm_device *dev,
 	spin_lock(&dev_priv->irq_lock);
 	for (i = 1; i < HPD_NUM_PINS; i++) {
 
-		WARN(((hpd[i] & hotplug_trigger) &&
-		      dev_priv->hpd_stats[i].hpd_mark != HPD_ENABLED),
-		     "Received HPD interrupt although disabled\n");
+		WARN_ONCE(hpd[i] & hotplug_trigger &&
+			  dev_priv->hpd_stats[i].hpd_mark == HPD_DISABLED,
+			  "Received HPD interrupt (0x%08x) on pin %d (0x%08x) although disabled\n",
+			  hotplug_trigger, i, hpd[i]);
 
 		if (!(hpd[i] & hotplug_trigger) ||
 		    dev_priv->hpd_stats[i].hpd_mark != HPD_ENABLED)
@@ -1474,6 +1480,9 @@ static irqreturn_t valleyview_irq_handler(int irq, void *arg)
 
 			intel_hpd_irq_handler(dev, hotplug_trigger, hpd_status_i915);
 
+			if (hotplug_status & DP_AUX_CHANNEL_MASK_INT_STATUS_G4X)
+				dp_aux_irq_handler(dev);
+
 			I915_WRITE(PORT_HOTPLUG_STAT, hotplug_status);
 			I915_READ(PORT_HOTPLUG_STAT);
 		}
@@ -1993,7 +2002,7 @@ static void i915_error_work_func(struct work_struct *work)
 			kobject_uevent_env(&dev->primary->kdev->kobj,
 					   KOBJ_CHANGE, reset_done_event);
 		} else {
-			atomic_set(&error->reset_counter, I915_WEDGED);
+			atomic_set_mask(I915_WEDGED, &error->reset_counter);
 		}
 
 		/*
@@ -3140,10 +3149,10 @@ static int i8xx_irq_postinstall(struct drm_device *dev)
  * Returns true when a page flip has completed.
  */
 static bool i8xx_handle_vblank(struct drm_device *dev,
-			       int pipe, u16 iir)
+			       int plane, int pipe, u32 iir)
 {
 	drm_i915_private_t *dev_priv = dev->dev_private;
-	u16 flip_pending = DISPLAY_PLANE_FLIP_PENDING(pipe);
+	u16 flip_pending = DISPLAY_PLANE_FLIP_PENDING(plane);
 
 	if (!drm_handle_vblank(dev, pipe))
 		return false;
@@ -3151,7 +3160,7 @@ static bool i8xx_handle_vblank(struct drm_device *dev,
 	if ((iir & flip_pending) == 0)
 		return false;
 
-	intel_prepare_page_flip(dev, pipe);
+	intel_prepare_page_flip(dev, plane);
 
 	/* We detect FlipDone by looking for the change in PendingFlip from '1'
 	 * to '0' on the following vblank, i.e. IIR has the Pendingflip
@@ -3220,9 +3229,13 @@ static irqreturn_t i8xx_irq_handler(int irq, void *arg)
 			notify_ring(dev, &dev_priv->ring[RCS]);
 
 		for_each_pipe(pipe) {
+			int plane = pipe;
+			if (HAS_FBC(dev))
+				plane = !plane;
+
 			if (pipe_stats[pipe] & PIPE_VBLANK_INTERRUPT_STATUS &&
-			    i8xx_handle_vblank(dev, pipe, iir))
-				flip_mask &= ~DISPLAY_PLANE_FLIP_PENDING(pipe);
+			    i8xx_handle_vblank(dev, plane, pipe, iir))
+				flip_mask &= ~DISPLAY_PLANE_FLIP_PENDING(plane);
 
 			if (pipe_stats[pipe] & PIPE_CRC_DONE_INTERRUPT_STATUS)
 				i9xx_pipe_crc_irq_handler(dev, pipe);
@@ -3418,7 +3431,7 @@ static irqreturn_t i915_irq_handler(int irq, void *arg)
 
 		for_each_pipe(pipe) {
 			int plane = pipe;
-			if (IS_MOBILE(dev))
+			if (HAS_FBC(dev))
 				plane = !plane;
 
 			if (pipe_stats[pipe] & PIPE_VBLANK_INTERRUPT_STATUS &&
@@ -3655,7 +3668,11 @@ static irqreturn_t i965_irq_handler(int irq, void *arg)
 				  hotplug_status);
 
 			intel_hpd_irq_handler(dev, hotplug_trigger,
-					      IS_G4X(dev) ? hpd_status_gen4 : hpd_status_i915);
+					      IS_G4X(dev) ? hpd_status_g4x : hpd_status_i915);
+
+			if (IS_G4X(dev) &&
+			    (hotplug_status & DP_AUX_CHANNEL_MASK_INT_STATUS_G4X))
+				dp_aux_irq_handler(dev);
 
 			I915_WRITE(PORT_HOTPLUG_STAT, hotplug_status);
 			I915_READ(PORT_HOTPLUG_STAT);
@@ -3893,8 +3910,8 @@ void hsw_pc8_disable_interrupts(struct drm_device *dev)
 	dev_priv->pc8.regsave.gtier = I915_READ(GTIER);
 	dev_priv->pc8.regsave.gen6_pmimr = I915_READ(GEN6_PMIMR);
 
-	ironlake_disable_display_irq(dev_priv, ~DE_PCH_EVENT_IVB);
-	ibx_disable_display_interrupt(dev_priv, ~SDE_HOTPLUG_MASK_CPT);
+	ironlake_disable_display_irq(dev_priv, 0xffffffff);
+	ibx_disable_display_interrupt(dev_priv, 0xffffffff);
 	ilk_disable_gt_irq(dev_priv, 0xffffffff);
 	snb_disable_pm_irq(dev_priv, 0xffffffff);
 
@@ -3908,34 +3925,26 @@ void hsw_pc8_restore_interrupts(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	unsigned long irqflags;
-	uint32_t val, expected;
+	uint32_t val;
 
 	spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
 
 	val = I915_READ(DEIMR);
-	expected = ~DE_PCH_EVENT_IVB;
-	WARN(val != expected, "DEIMR is 0x%08x, not 0x%08x\n", val, expected);
+	WARN(val != 0xffffffff, "DEIMR is 0x%08x\n", val);
 
-	val = I915_READ(SDEIMR) & ~SDE_HOTPLUG_MASK_CPT;
-	expected = ~SDE_HOTPLUG_MASK_CPT;
-	WARN(val != expected, "SDEIMR non-HPD bits are 0x%08x, not 0x%08x\n",
-	     val, expected);
+	val = I915_READ(SDEIMR);
+	WARN(val != 0xffffffff, "SDEIMR is 0x%08x\n", val);
 
 	val = I915_READ(GTIMR);
-	expected = 0xffffffff;
-	WARN(val != expected, "GTIMR is 0x%08x, not 0x%08x\n", val, expected);
+	WARN(val != 0xffffffff, "GTIMR is 0x%08x\n", val);
 
 	val = I915_READ(GEN6_PMIMR);
-	expected = 0xffffffff;
-	WARN(val != expected, "GEN6_PMIMR is 0x%08x, not 0x%08x\n", val,
-	     expected);
+	WARN(val != 0xffffffff, "GEN6_PMIMR is 0x%08x\n", val);
 
 	dev_priv->pc8.irqs_disabled = false;
 
 	ironlake_enable_display_irq(dev_priv, ~dev_priv->pc8.regsave.deimr);
-	ibx_enable_display_interrupt(dev_priv,
-				     ~dev_priv->pc8.regsave.sdeimr &
-				     ~SDE_HOTPLUG_MASK_CPT);
+	ibx_enable_display_interrupt(dev_priv, ~dev_priv->pc8.regsave.sdeimr);
 	ilk_enable_gt_irq(dev_priv, ~dev_priv->pc8.regsave.gtimr);
 	snb_enable_pm_irq(dev_priv, ~dev_priv->pc8.regsave.gen6_pmimr);
 	I915_WRITE(GTIER, dev_priv->pc8.regsave.gtier);
diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index ee2742122a02..a48b7cad6f11 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -193,10 +193,13 @@
 #define   MI_SCENE_COUNT	(1 << 3) /* just increment scene count */
 #define   MI_END_SCENE		(1 << 4) /* flush binner and incr scene count */
 #define   MI_INVALIDATE_ISP	(1 << 5) /* invalidate indirect state pointers */
+#define MI_REPORT_HEAD		MI_INSTR(0x07, 0)
+#define MI_ARB_ON_OFF		MI_INSTR(0x08, 0)
+#define   MI_ARB_ENABLE			(1<<0)
+#define   MI_ARB_DISABLE		(0<<0)
 #define MI_BATCH_BUFFER_END	MI_INSTR(0x0a, 0)
 #define MI_SUSPEND_FLUSH	MI_INSTR(0x0b, 0)
 #define   MI_SUSPEND_FLUSH_EN	(1<<0)
-#define MI_REPORT_HEAD		MI_INSTR(0x07, 0)
 #define MI_OVERLAY_FLIP		MI_INSTR(0x11, 0)
 #define   MI_OVERLAY_CONTINUE	(0x0<<21)
 #define   MI_OVERLAY_ON		(0x1<<21)
@@ -212,10 +215,24 @@
 #define   MI_DISPLAY_FLIP_IVB_SPRITE_B (3 << 19)
 #define   MI_DISPLAY_FLIP_IVB_PLANE_C  (4 << 19)
 #define   MI_DISPLAY_FLIP_IVB_SPRITE_C (5 << 19)
-#define MI_ARB_ON_OFF		MI_INSTR(0x08, 0)
-#define   MI_ARB_ENABLE			(1<<0)
-#define   MI_ARB_DISABLE		(0<<0)
-
+#define MI_SEMAPHORE_MBOX	MI_INSTR(0x16, 1) /* gen6+ */
+#define   MI_SEMAPHORE_GLOBAL_GTT    (1<<22)
+#define   MI_SEMAPHORE_UPDATE	    (1<<21)
+#define   MI_SEMAPHORE_COMPARE	    (1<<20)
+#define   MI_SEMAPHORE_REGISTER	    (1<<18)
+#define   MI_SEMAPHORE_SYNC_VR	    (0<<16) /* RCS  wait for VCS  (RVSYNC) */
+#define   MI_SEMAPHORE_SYNC_VER	    (1<<16) /* RCS  wait for VECS (RVESYNC) */
+#define   MI_SEMAPHORE_SYNC_BR	    (2<<16) /* RCS  wait for BCS  (RBSYNC) */
+#define   MI_SEMAPHORE_SYNC_BV	    (0<<16) /* VCS  wait for BCS  (VBSYNC) */
+#define   MI_SEMAPHORE_SYNC_VEV	    (1<<16) /* VCS  wait for VECS (VVESYNC) */
+#define   MI_SEMAPHORE_SYNC_RV	    (2<<16) /* VCS  wait for RCS  (VRSYNC) */
+#define   MI_SEMAPHORE_SYNC_RB	    (0<<16) /* BCS  wait for RCS  (BRSYNC) */
+#define   MI_SEMAPHORE_SYNC_VEB	    (1<<16) /* BCS  wait for VECS (BVESYNC) */
+#define   MI_SEMAPHORE_SYNC_VB	    (2<<16) /* BCS  wait for VCS  (BVSYNC) */
+#define   MI_SEMAPHORE_SYNC_BVE	    (0<<16) /* VECS wait for BCS  (VEBSYNC) */
+#define   MI_SEMAPHORE_SYNC_VVE	    (1<<16) /* VECS wait for VCS  (VEVSYNC) */
+#define   MI_SEMAPHORE_SYNC_RVE	    (2<<16) /* VECS wait for RCS  (VERSYNC) */
+#define   MI_SEMAPHORE_SYNC_INVALID  (3<<16)
 #define MI_SET_CONTEXT		MI_INSTR(0x18, 0)
 #define   MI_MM_SPACE_GTT		(1<<8)
 #define   MI_MM_SPACE_PHYSICAL		(0<<8)
@@ -235,7 +252,7 @@
  */
 #define MI_LOAD_REGISTER_IMM(x)	MI_INSTR(0x22, 2*x-1)
 #define MI_STORE_REGISTER_MEM(x) MI_INSTR(0x24, 2*x-1)
-#define  MI_SRM_LRM_GLOBAL_GTT		(1<<22)
+#define   MI_SRM_LRM_GLOBAL_GTT		(1<<22)
 #define MI_FLUSH_DW		MI_INSTR(0x26, 1) /* for GEN6 */
 #define   MI_FLUSH_DW_STORE_INDEX	(1<<21)
 #define   MI_INVALIDATE_TLB		(1<<18)
@@ -246,30 +263,13 @@
 #define MI_BATCH_BUFFER		MI_INSTR(0x30, 1)
 #define   MI_BATCH_NON_SECURE		(1)
 /* for snb/ivb/vlv this also means "batch in ppgtt" when ppgtt is enabled. */
-#define   MI_BATCH_NON_SECURE_I965 	(1<<8)
+#define   MI_BATCH_NON_SECURE_I965	(1<<8)
 #define   MI_BATCH_PPGTT_HSW		(1<<8)
-#define   MI_BATCH_NON_SECURE_HSW 	(1<<13)
+#define   MI_BATCH_NON_SECURE_HSW	(1<<13)
 #define MI_BATCH_BUFFER_START	MI_INSTR(0x31, 0)
 #define   MI_BATCH_GTT		    (2<<6) /* aliased with (1<<7) on gen4 */
 #define MI_BATCH_BUFFER_START_GEN8	MI_INSTR(0x31, 1)
-#define MI_SEMAPHORE_MBOX	MI_INSTR(0x16, 1) /* gen6+ */
-#define  MI_SEMAPHORE_GLOBAL_GTT    (1<<22)
-#define  MI_SEMAPHORE_UPDATE	    (1<<21)
-#define  MI_SEMAPHORE_COMPARE	    (1<<20)
-#define  MI_SEMAPHORE_REGISTER	    (1<<18)
-#define  MI_SEMAPHORE_SYNC_VR	    (0<<16) /* RCS  wait for VCS  (RVSYNC) */
-#define  MI_SEMAPHORE_SYNC_VER	    (1<<16) /* RCS  wait for VECS (RVESYNC) */
-#define  MI_SEMAPHORE_SYNC_BR	    (2<<16) /* RCS  wait for BCS  (RBSYNC) */
-#define  MI_SEMAPHORE_SYNC_BV	    (0<<16) /* VCS  wait for BCS  (VBSYNC) */
-#define  MI_SEMAPHORE_SYNC_VEV	    (1<<16) /* VCS  wait for VECS (VVESYNC) */
-#define  MI_SEMAPHORE_SYNC_RV	    (2<<16) /* VCS  wait for RCS  (VRSYNC) */
-#define  MI_SEMAPHORE_SYNC_RB	    (0<<16) /* BCS  wait for RCS  (BRSYNC) */
-#define  MI_SEMAPHORE_SYNC_VEB	    (1<<16) /* BCS  wait for VECS (BVESYNC) */
-#define  MI_SEMAPHORE_SYNC_VB	    (2<<16) /* BCS  wait for VCS  (BVSYNC) */
-#define  MI_SEMAPHORE_SYNC_BVE	    (0<<16) /* VECS wait for BCS  (VEBSYNC) */
-#define  MI_SEMAPHORE_SYNC_VVE	    (1<<16) /* VECS wait for VCS  (VEVSYNC) */
-#define  MI_SEMAPHORE_SYNC_RVE	    (2<<16) /* VECS wait for RCS  (VERSYNC) */
-#define  MI_SEMAPHORE_SYNC_INVALID  (3<<16)
+
 
 #define MI_PREDICATE_RESULT_2	(0x2214)
 #define  LOWER_SLICE_ENABLED	(1<<0)
@@ -354,6 +354,7 @@
 #define   IOSF_BYTE_ENABLES_SHIFT		4
 #define   IOSF_BAR_SHIFT			1
 #define   IOSF_SB_BUSY				(1<<0)
+#define   IOSF_PORT_BUNIT			0x3
 #define   IOSF_PORT_PUNIT			0x4
 #define   IOSF_PORT_NC				0x11
 #define   IOSF_PORT_DPIO			0x12
@@ -361,12 +362,21 @@
 #define   IOSF_PORT_CCK				0x14
 #define   IOSF_PORT_CCU				0xA9
 #define   IOSF_PORT_GPS_CORE			0x48
+#define   IOSF_PORT_FLISDSI			0x1B
 #define VLV_IOSF_DATA				(VLV_DISPLAY_BASE + 0x2104)
 #define VLV_IOSF_ADDR				(VLV_DISPLAY_BASE + 0x2108)
 
+/* See configdb bunit SB addr map */
+#define BUNIT_REG_BISOC				0x11
+
 #define PUNIT_OPCODE_REG_READ			6
 #define PUNIT_OPCODE_REG_WRITE			7
 
+#define PUNIT_REG_DSPFREQ			0x36
+#define   DSPFREQSTAT_SHIFT			30
+#define   DSPFREQSTAT_MASK			(0x3 << DSPFREQSTAT_SHIFT)
+#define   DSPFREQGUAR_SHIFT			14
+#define   DSPFREQGUAR_MASK			(0x3 << DSPFREQGUAR_SHIFT)
 #define PUNIT_REG_PWRGT_CTRL			0x60
 #define PUNIT_REG_PWRGT_STATUS			0x61
 #define	  PUNIT_CLK_GATE			1
@@ -429,6 +439,7 @@
 #define  DSI_PLL_N1_DIV_MASK			(3 << 16)
 #define  DSI_PLL_M1_DIV_SHIFT			0
 #define  DSI_PLL_M1_DIV_MASK			(0x1ff << 0)
+#define CCK_DISPLAY_CLOCK_CONTROL		0x6b
 
 /*
  * DPIO - a special bus for various display related registers to hide behind
@@ -447,15 +458,13 @@
 #define  DPIO_SFR_BYPASS		(1<<1)
 #define  DPIO_CMNRST			(1<<0)
 
-#define _DPIO_TX3_SWING_CTL4_A		0x690
-#define _DPIO_TX3_SWING_CTL4_B		0x2a90
-#define DPIO_TX3_SWING_CTL4(pipe) _PIPE(pipe, _DPIO_TX3_SWING_CTL4_A, \
-					_DPIO_TX3_SWING_CTL4_B)
+#define DPIO_PHY(pipe)			((pipe) >> 1)
+#define DPIO_PHY_IOSF_PORT(phy)		(dev_priv->dpio_phy_iosf_port[phy])
 
 /*
  * Per pipe/PLL DPIO regs
  */
-#define _DPIO_DIV_A			0x800c
+#define _VLV_PLL_DW3_CH0		0x800c
 #define   DPIO_POST_DIV_SHIFT		(28) /* 3 bits */
 #define   DPIO_POST_DIV_DAC		0
 #define   DPIO_POST_DIV_HDMIDP		1 /* DAC 225-400M rate */
@@ -468,10 +477,10 @@
 #define   DPIO_ENABLE_CALIBRATION	(1<<11)
 #define   DPIO_M1DIV_SHIFT		(8) /* 3 bits */
 #define   DPIO_M2DIV_MASK		0xff
-#define _DPIO_DIV_B			0x802c
-#define DPIO_DIV(pipe) _PIPE(pipe, _DPIO_DIV_A, _DPIO_DIV_B)
+#define _VLV_PLL_DW3_CH1		0x802c
+#define VLV_PLL_DW3(ch) _PIPE(ch, _VLV_PLL_DW3_CH0, _VLV_PLL_DW3_CH1)
 
-#define _DPIO_REFSFR_A			0x8014
+#define _VLV_PLL_DW5_CH0		0x8014
 #define   DPIO_REFSEL_OVERRIDE		27
 #define   DPIO_PLL_MODESEL_SHIFT	24 /* 3 bits */
 #define   DPIO_BIAS_CURRENT_CTL_SHIFT	21 /* 3 bits, always 0x7 */
@@ -479,118 +488,112 @@
 #define   DPIO_PLL_REFCLK_SEL_MASK	3
 #define   DPIO_DRIVER_CTL_SHIFT		12 /* always set to 0x8 */
 #define   DPIO_CLK_BIAS_CTL_SHIFT	8 /* always set to 0x5 */
-#define _DPIO_REFSFR_B			0x8034
-#define DPIO_REFSFR(pipe) _PIPE(pipe, _DPIO_REFSFR_A, _DPIO_REFSFR_B)
+#define _VLV_PLL_DW5_CH1		0x8034
+#define VLV_PLL_DW5(ch) _PIPE(ch, _VLV_PLL_DW5_CH0, _VLV_PLL_DW5_CH1)
 
-#define _DPIO_CORE_CLK_A		0x801c
-#define _DPIO_CORE_CLK_B		0x803c
-#define DPIO_CORE_CLK(pipe) _PIPE(pipe, _DPIO_CORE_CLK_A, _DPIO_CORE_CLK_B)
+#define _VLV_PLL_DW7_CH0		0x801c
+#define _VLV_PLL_DW7_CH1		0x803c
+#define VLV_PLL_DW7(ch) _PIPE(ch, _VLV_PLL_DW7_CH0, _VLV_PLL_DW7_CH1)
 
-#define _DPIO_IREF_CTL_A		0x8040
-#define _DPIO_IREF_CTL_B		0x8060
-#define DPIO_IREF_CTL(pipe) _PIPE(pipe, _DPIO_IREF_CTL_A, _DPIO_IREF_CTL_B)
+#define _VLV_PLL_DW8_CH0		0x8040
+#define _VLV_PLL_DW8_CH1		0x8060
+#define VLV_PLL_DW8(ch) _PIPE(ch, _VLV_PLL_DW8_CH0, _VLV_PLL_DW8_CH1)
 
-#define DPIO_IREF_BCAST			0xc044
-#define _DPIO_IREF_A			0x8044
-#define _DPIO_IREF_B			0x8064
-#define DPIO_IREF(pipe) _PIPE(pipe, _DPIO_IREF_A, _DPIO_IREF_B)
+#define VLV_PLL_DW9_BCAST		0xc044
+#define _VLV_PLL_DW9_CH0		0x8044
+#define _VLV_PLL_DW9_CH1		0x8064
+#define VLV_PLL_DW9(ch) _PIPE(ch, _VLV_PLL_DW9_CH0, _VLV_PLL_DW9_CH1)
 
-#define _DPIO_PLL_CML_A			0x804c
-#define _DPIO_PLL_CML_B			0x806c
-#define DPIO_PLL_CML(pipe) _PIPE(pipe, _DPIO_PLL_CML_A, _DPIO_PLL_CML_B)
+#define _VLV_PLL_DW10_CH0		0x8048
+#define _VLV_PLL_DW10_CH1		0x8068
+#define VLV_PLL_DW10(ch) _PIPE(ch, _VLV_PLL_DW10_CH0, _VLV_PLL_DW10_CH1)
 
-#define _DPIO_LPF_COEFF_A		0x8048
-#define _DPIO_LPF_COEFF_B		0x8068
-#define DPIO_LPF_COEFF(pipe) _PIPE(pipe, _DPIO_LPF_COEFF_A, _DPIO_LPF_COEFF_B)
+#define _VLV_PLL_DW11_CH0		0x804c
+#define _VLV_PLL_DW11_CH1		0x806c
+#define VLV_PLL_DW11(ch) _PIPE(ch, _VLV_PLL_DW11_CH0, _VLV_PLL_DW11_CH1)
 
-#define DPIO_CALIBRATION		0x80ac
+/* Spec for ref block start counts at DW10 */
+#define VLV_REF_DW13			0x80ac
 
-#define DPIO_FASTCLK_DISABLE		0x8100
+#define VLV_CMN_DW0			0x8100
 
 /*
  * Per DDI channel DPIO regs
  */
 
-#define _DPIO_PCS_TX_0			0x8200
-#define _DPIO_PCS_TX_1			0x8400
+#define _VLV_PCS_DW0_CH0		0x8200
+#define _VLV_PCS_DW0_CH1		0x8400
 #define   DPIO_PCS_TX_LANE2_RESET	(1<<16)
 #define   DPIO_PCS_TX_LANE1_RESET	(1<<7)
-#define DPIO_PCS_TX(port) _PORT(port, _DPIO_PCS_TX_0, _DPIO_PCS_TX_1)
+#define VLV_PCS_DW0(ch) _PORT(ch, _VLV_PCS_DW0_CH0, _VLV_PCS_DW0_CH1)
 
-#define _DPIO_PCS_CLK_0			0x8204
-#define _DPIO_PCS_CLK_1			0x8404
+#define _VLV_PCS_DW1_CH0		0x8204
+#define _VLV_PCS_DW1_CH1		0x8404
 #define   DPIO_PCS_CLK_CRI_RXEB_EIOS_EN	(1<<22)
 #define   DPIO_PCS_CLK_CRI_RXDIGFILTSG_EN (1<<21)
 #define   DPIO_PCS_CLK_DATAWIDTH_SHIFT	(6)
 #define   DPIO_PCS_CLK_SOFT_RESET	(1<<5)
-#define DPIO_PCS_CLK(port) _PORT(port, _DPIO_PCS_CLK_0, _DPIO_PCS_CLK_1)
-
-#define _DPIO_PCS_CTL_OVR1_A		0x8224
-#define _DPIO_PCS_CTL_OVR1_B		0x8424
-#define DPIO_PCS_CTL_OVER1(port) _PORT(port, _DPIO_PCS_CTL_OVR1_A, \
-				       _DPIO_PCS_CTL_OVR1_B)
-
-#define _DPIO_PCS_STAGGER0_A		0x822c
-#define _DPIO_PCS_STAGGER0_B		0x842c
-#define DPIO_PCS_STAGGER0(port) _PORT(port, _DPIO_PCS_STAGGER0_A, \
-				      _DPIO_PCS_STAGGER0_B)
-
-#define _DPIO_PCS_STAGGER1_A		0x8230
-#define _DPIO_PCS_STAGGER1_B		0x8430
-#define DPIO_PCS_STAGGER1(port) _PORT(port, _DPIO_PCS_STAGGER1_A, \
-				      _DPIO_PCS_STAGGER1_B)
-
-#define _DPIO_PCS_CLOCKBUF0_A		0x8238
-#define _DPIO_PCS_CLOCKBUF0_B		0x8438
-#define DPIO_PCS_CLOCKBUF0(port) _PORT(port, _DPIO_PCS_CLOCKBUF0_A, \
-				       _DPIO_PCS_CLOCKBUF0_B)
-
-#define _DPIO_PCS_CLOCKBUF8_A		0x825c
-#define _DPIO_PCS_CLOCKBUF8_B		0x845c
-#define DPIO_PCS_CLOCKBUF8(port) _PORT(port, _DPIO_PCS_CLOCKBUF8_A, \
-				       _DPIO_PCS_CLOCKBUF8_B)
-
-#define _DPIO_TX_SWING_CTL2_A		0x8288
-#define _DPIO_TX_SWING_CTL2_B		0x8488
-#define DPIO_TX_SWING_CTL2(port) _PORT(port, _DPIO_TX_SWING_CTL2_A, \
-				       _DPIO_TX_SWING_CTL2_B)
-
-#define _DPIO_TX_SWING_CTL3_A		0x828c
-#define _DPIO_TX_SWING_CTL3_B		0x848c
-#define DPIO_TX_SWING_CTL3(port) _PORT(port, _DPIO_TX_SWING_CTL3_A, \
-				       _DPIO_TX_SWING_CTL3_B)
-
-#define _DPIO_TX_SWING_CTL4_A		0x8290
-#define _DPIO_TX_SWING_CTL4_B		0x8490
-#define DPIO_TX_SWING_CTL4(port) _PORT(port, _DPIO_TX_SWING_CTL4_A, \
-				       _DPIO_TX_SWING_CTL4_B)
-
-#define _DPIO_TX_OCALINIT_0		0x8294
-#define _DPIO_TX_OCALINIT_1		0x8494
+#define VLV_PCS_DW1(ch) _PORT(ch, _VLV_PCS_DW1_CH0, _VLV_PCS_DW1_CH1)
+
+#define _VLV_PCS_DW8_CH0		0x8220
+#define _VLV_PCS_DW8_CH1		0x8420
+#define VLV_PCS_DW8(ch) _PORT(ch, _VLV_PCS_DW8_CH0, _VLV_PCS_DW8_CH1)
+
+#define _VLV_PCS01_DW8_CH0		0x0220
+#define _VLV_PCS23_DW8_CH0		0x0420
+#define _VLV_PCS01_DW8_CH1		0x2620
+#define _VLV_PCS23_DW8_CH1		0x2820
+#define VLV_PCS01_DW8(port) _PORT(port, _VLV_PCS01_DW8_CH0, _VLV_PCS01_DW8_CH1)
+#define VLV_PCS23_DW8(port) _PORT(port, _VLV_PCS23_DW8_CH0, _VLV_PCS23_DW8_CH1)
+
+#define _VLV_PCS_DW9_CH0		0x8224
+#define _VLV_PCS_DW9_CH1		0x8424
+#define	VLV_PCS_DW9(ch) _PORT(ch, _VLV_PCS_DW9_CH0, _VLV_PCS_DW9_CH1)
+
+#define _VLV_PCS_DW11_CH0		0x822c
+#define _VLV_PCS_DW11_CH1		0x842c
+#define VLV_PCS_DW11(ch) _PORT(ch, _VLV_PCS_DW11_CH0, _VLV_PCS_DW11_CH1)
+
+#define _VLV_PCS_DW12_CH0		0x8230
+#define _VLV_PCS_DW12_CH1		0x8430
+#define VLV_PCS_DW12(ch) _PORT(ch, _VLV_PCS_DW12_CH0, _VLV_PCS_DW12_CH1)
+
+#define _VLV_PCS_DW14_CH0		0x8238
+#define _VLV_PCS_DW14_CH1		0x8438
+#define	VLV_PCS_DW14(ch) _PORT(ch, _VLV_PCS_DW14_CH0, _VLV_PCS_DW14_CH1)
+
+#define _VLV_PCS_DW23_CH0		0x825c
+#define _VLV_PCS_DW23_CH1		0x845c
+#define VLV_PCS_DW23(ch) _PORT(ch, _VLV_PCS_DW23_CH0, _VLV_PCS_DW23_CH1)
+
+#define _VLV_TX_DW2_CH0			0x8288
+#define _VLV_TX_DW2_CH1			0x8488
+#define VLV_TX_DW2(ch) _PORT(ch, _VLV_TX_DW2_CH0, _VLV_TX_DW2_CH1)
+
+#define _VLV_TX_DW3_CH0			0x828c
+#define _VLV_TX_DW3_CH1			0x848c
+#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 VLV_TX_DW4(ch) _PORT(ch, _VLV_TX_DW4_CH0, _VLV_TX_DW4_CH1)
+
+#define _VLV_TX3_DW4_CH0		0x690
+#define _VLV_TX3_DW4_CH1		0x2a90
+#define VLV_TX3_DW4(ch) _PORT(ch, _VLV_TX3_DW4_CH0, _VLV_TX3_DW4_CH1)
+
+#define _VLV_TX_DW5_CH0			0x8294
+#define _VLV_TX_DW5_CH1			0x8494
 #define   DPIO_TX_OCALINIT_EN		(1<<31)
-#define DPIO_TX_OCALINIT(port) _PORT(port, _DPIO_TX_OCALINIT_0, \
-				     _DPIO_TX_OCALINIT_1)
-
-#define _DPIO_TX_CTL_0			0x82ac
-#define _DPIO_TX_CTL_1			0x84ac
-#define DPIO_TX_CTL(port) _PORT(port, _DPIO_TX_CTL_0, _DPIO_TX_CTL_1)
-
-#define _DPIO_TX_LANE_0			0x82b8
-#define _DPIO_TX_LANE_1			0x84b8
-#define DPIO_TX_LANE(port) _PORT(port, _DPIO_TX_LANE_0, _DPIO_TX_LANE_1)
-
-#define _DPIO_DATA_CHANNEL1		0x8220
-#define _DPIO_DATA_CHANNEL2		0x8420
-#define DPIO_DATA_CHANNEL(port) _PORT(port, _DPIO_DATA_CHANNEL1, _DPIO_DATA_CHANNEL2)
-
-#define _DPIO_PORT0_PCS0		0x0220
-#define _DPIO_PORT0_PCS1		0x0420
-#define _DPIO_PORT1_PCS2		0x2620
-#define _DPIO_PORT1_PCS3		0x2820
-#define DPIO_DATA_LANE_A(port) _PORT(port, _DPIO_PORT0_PCS0, _DPIO_PORT1_PCS2)
-#define DPIO_DATA_LANE_B(port) _PORT(port, _DPIO_PORT0_PCS1, _DPIO_PORT1_PCS3)
-#define DPIO_DATA_CHANNEL1              0x8220
-#define DPIO_DATA_CHANNEL2              0x8420
+#define VLV_TX_DW5(ch) _PORT(ch, _VLV_TX_DW5_CH0, _VLV_TX_DW5_CH1)
+
+#define _VLV_TX_DW11_CH0		0x82ac
+#define _VLV_TX_DW11_CH1		0x84ac
+#define VLV_TX_DW11(ch) _PORT(ch, _VLV_TX_DW11_CH0, _VLV_TX_DW11_CH1)
+
+#define _VLV_TX_DW14_CH0		0x82b8
+#define _VLV_TX_DW14_CH1		0x84b8
+#define VLV_TX_DW14(ch) _PORT(ch, _VLV_TX_DW14_CH0, _VLV_TX_DW14_CH1)
 
 /*
  * Fence registers
@@ -732,6 +735,8 @@
 #define HWSTAM		0x02098
 #define DMA_FADD_I8XX	0x020d0
 #define RING_BBSTATE(base)	((base)+0x110)
+#define RING_BBADDR(base)	((base)+0x140)
+#define RING_BBADDR_UDW(base)	((base)+0x168) /* gen8+ */
 
 #define ERROR_GEN6	0x040a0
 #define GEN7_ERR_INT	0x44040
@@ -922,7 +927,6 @@
 #define   CM0_COLOR_EVICT_DISABLE (1<<3)
 #define   CM0_DEPTH_WRITE_DISABLE (1<<1)
 #define   CM0_RC_OP_FLUSH_DISABLE (1<<0)
-#define BB_ADDR		0x02140 /* 8 bytes */
 #define GFX_FLSH_CNTL	0x02170 /* 915+ only */
 #define GFX_FLSH_CNTL_GEN6	0x101008
 #define   GFX_FLSH_CNTL_EN	(1<<0)
@@ -999,6 +1003,7 @@
 
 #define GEN7_FF_THREAD_MODE		0x20a0
 #define   GEN7_FF_SCHED_MASK		0x0077070
+#define   GEN8_FF_DS_REF_CNT_FFME	(1 << 19)
 #define   GEN7_FF_TS_SCHED_HS1		(0x5<<16)
 #define   GEN7_FF_TS_SCHED_HS0		(0x3<<16)
 #define   GEN7_FF_TS_SCHED_LOAD_BALANCE	(0x1<<16)
@@ -1026,14 +1031,14 @@
 #define   FBC_CTL_UNCOMPRESSIBLE (1<<14)
 #define   FBC_CTL_C3_IDLE	(1<<13)
 #define   FBC_CTL_STRIDE_SHIFT	(5)
-#define   FBC_CTL_FENCENO	(1<<0)
+#define   FBC_CTL_FENCENO_SHIFT	(0)
 #define FBC_COMMAND		0x0320c
 #define   FBC_CMD_COMPRESS	(1<<0)
 #define FBC_STATUS		0x03210
 #define   FBC_STAT_COMPRESSING	(1<<31)
 #define   FBC_STAT_COMPRESSED	(1<<30)
 #define   FBC_STAT_MODIFIED	(1<<29)
-#define   FBC_STAT_CURRENT_LINE	(1<<0)
+#define   FBC_STAT_CURRENT_LINE_SHIFT	(0)
 #define FBC_CONTROL2		0x03214
 #define   FBC_CTL_FENCE_DBL	(0<<4)
 #define   FBC_CTL_IDLE_IMM	(0<<2)
@@ -2117,9 +2122,13 @@
  * Please check the detailed lore in the commit message for for experimental
  * evidence.
  */
-#define   PORTD_HOTPLUG_LIVE_STATUS               (1 << 29)
-#define   PORTC_HOTPLUG_LIVE_STATUS               (1 << 28)
-#define   PORTB_HOTPLUG_LIVE_STATUS               (1 << 27)
+#define   PORTD_HOTPLUG_LIVE_STATUS_G4X		(1 << 29)
+#define   PORTC_HOTPLUG_LIVE_STATUS_G4X		(1 << 28)
+#define   PORTB_HOTPLUG_LIVE_STATUS_G4X		(1 << 27)
+/* VLV DP/HDMI bits again match Bspec */
+#define   PORTD_HOTPLUG_LIVE_STATUS_VLV		(1 << 27)
+#define   PORTC_HOTPLUG_LIVE_STATUS_VLV		(1 << 28)
+#define   PORTB_HOTPLUG_LIVE_STATUS_VLV		(1 << 29)
 #define   PORTD_HOTPLUG_INT_STATUS		(3 << 21)
 #define   PORTC_HOTPLUG_INT_STATUS		(3 << 19)
 #define   PORTB_HOTPLUG_INT_STATUS		(3 << 17)
@@ -2130,6 +2139,11 @@
 #define   CRT_HOTPLUG_MONITOR_COLOR		(3 << 8)
 #define   CRT_HOTPLUG_MONITOR_MONO		(2 << 8)
 #define   CRT_HOTPLUG_MONITOR_NONE		(0 << 8)
+#define   DP_AUX_CHANNEL_D_INT_STATUS_G4X	(1 << 6)
+#define   DP_AUX_CHANNEL_C_INT_STATUS_G4X	(1 << 5)
+#define   DP_AUX_CHANNEL_B_INT_STATUS_G4X	(1 << 4)
+#define   DP_AUX_CHANNEL_MASK_INT_STATUS_G4X	(7 << 4)
+
 /* SDVO is different across gen3/4 */
 #define   SDVOC_HOTPLUG_INT_STATUS_G4X		(1 << 3)
 #define   SDVOB_HOTPLUG_INT_STATUS_G4X		(1 << 2)
@@ -3421,42 +3435,6 @@
 /* the unit of memory self-refresh latency time is 0.5us */
 #define  ILK_SRLT_MASK		0x3f
 
-/* define the fifo size on Ironlake */
-#define ILK_DISPLAY_FIFO	128
-#define ILK_DISPLAY_MAXWM	64
-#define ILK_DISPLAY_DFTWM	8
-#define ILK_CURSOR_FIFO		32
-#define ILK_CURSOR_MAXWM	16
-#define ILK_CURSOR_DFTWM	8
-
-#define ILK_DISPLAY_SR_FIFO	512
-#define ILK_DISPLAY_MAX_SRWM	0x1ff
-#define ILK_DISPLAY_DFT_SRWM	0x3f
-#define ILK_CURSOR_SR_FIFO	64
-#define ILK_CURSOR_MAX_SRWM	0x3f
-#define ILK_CURSOR_DFT_SRWM	8
-
-#define ILK_FIFO_LINE_SIZE	64
-
-/* define the WM info on Sandybridge */
-#define SNB_DISPLAY_FIFO	128
-#define SNB_DISPLAY_MAXWM	0x7f	/* bit 16:22 */
-#define SNB_DISPLAY_DFTWM	8
-#define SNB_CURSOR_FIFO		32
-#define SNB_CURSOR_MAXWM	0x1f	/* bit 4:0 */
-#define SNB_CURSOR_DFTWM	8
-
-#define SNB_DISPLAY_SR_FIFO	512
-#define SNB_DISPLAY_MAX_SRWM	0x1ff	/* bit 16:8 */
-#define SNB_DISPLAY_DFT_SRWM	0x3f
-#define SNB_CURSOR_SR_FIFO	64
-#define SNB_CURSOR_MAX_SRWM	0x3f	/* bit 5:0 */
-#define SNB_CURSOR_DFT_SRWM	8
-
-#define SNB_FBC_MAX_SRWM	0xf	/* bit 23:20 */
-
-#define SNB_FIFO_LINE_SIZE	64
-
 
 /* the address where we get all kinds of latency value */
 #define SSKPD			0x5d10
@@ -3600,8 +3578,6 @@
 #define DISP_BASEADDR_MASK	(0xfffff000)
 #define I915_LO_DISPBASE(val)	(val & ~DISP_BASEADDR_MASK)
 #define I915_HI_DISPBASE(val)	(val & DISP_BASEADDR_MASK)
-#define I915_MODIFY_DISPBASE(reg, gfx_addr) \
-		(I915_WRITE((reg), (gfx_addr) | I915_LO_DISPBASE(I915_READ(reg))))
 
 /* VBIOS flags */
 #define SWF00			(dev_priv->info->display_mmio_offset + 0x71410)
@@ -3787,7 +3763,7 @@
 
 #define _SPACNTR		(VLV_DISPLAY_BASE + 0x72180)
 #define   SP_ENABLE			(1<<31)
-#define   SP_GEAMMA_ENABLE		(1<<30)
+#define   SP_GAMMA_ENABLE		(1<<30)
 #define   SP_PIXFORMAT_MASK		(0xf<<26)
 #define   SP_FORMAT_YUV422		(0<<26)
 #define   SP_FORMAT_BGR565		(5<<26)
@@ -4139,6 +4115,8 @@
 #define DISP_ARB_CTL	0x45000
 #define  DISP_TILE_SURFACE_SWIZZLING	(1<<13)
 #define  DISP_FBC_WM_DIS		(1<<15)
+#define DISP_ARB_CTL2	0x45004
+#define  DISP_DATA_PARTITION_5_6	(1<<6)
 #define GEN7_MSG_CTL	0x45010
 #define  WAIT_FOR_PCH_RESET_ACK		(1<<1)
 #define  WAIT_FOR_PCH_FLR_ACK		(1<<0)
@@ -4159,6 +4137,10 @@
 #define GEN7_L3SQCREG4				0xb034
 #define  L3SQ_URB_READ_CAM_MATCH_DISABLE	(1<<27)
 
+/* GEN8 chicken */
+#define HDC_CHICKEN0				0x7300
+#define  HDC_FORCE_NON_COHERENT			(1<<4)
+
 /* WaCatErrorRejectionIssue */
 #define GEN7_SQ_CHICKEN_MBCUNIT_CONFIG		0x9030
 #define  GEN7_SQ_CHICKEN_MBCUNIT_SQINTMOB	(1<<11)
@@ -4843,6 +4825,8 @@
 #define  FORCEWAKE_ACK				0x130090
 #define  VLV_GTLC_WAKE_CTRL			0x130090
 #define  VLV_GTLC_PW_STATUS			0x130094
+#define VLV_GTLC_PW_RENDER_STATUS_MASK		0x80
+#define VLV_GTLC_PW_MEDIA_STATUS_MASK		0x20
 #define  FORCEWAKE_MT				0xa188 /* multi-threaded */
 #define   FORCEWAKE_KERNEL			0x1
 #define   FORCEWAKE_USER			0x2
@@ -4851,12 +4835,16 @@
 #define    FORCEWAKE_MT_ENABLE			(1<<5)
 
 #define  GTFIFODBG				0x120000
-#define    GT_FIFO_CPU_ERROR_MASK		7
+#define    GT_FIFO_SBDROPERR			(1<<6)
+#define    GT_FIFO_BLOBDROPERR			(1<<5)
+#define    GT_FIFO_SB_READ_ABORTERR		(1<<4)
+#define    GT_FIFO_DROPERR			(1<<3)
 #define    GT_FIFO_OVFERR			(1<<2)
 #define    GT_FIFO_IAWRERR			(1<<1)
 #define    GT_FIFO_IARDERR			(1<<0)
 
-#define  GT_FIFO_FREE_ENTRIES			0x120008
+#define  GTFIFOCTL				0x120008
+#define    GT_FIFO_FREE_ENTRIES_MASK		0x7f
 #define    GT_FIFO_NUM_RESERVED_ENTRIES		20
 
 #define  HSW_IDICR				0x9008
@@ -4890,6 +4878,7 @@
 #define   GEN6_RC_CTL_RC6_ENABLE		(1<<18)
 #define   GEN6_RC_CTL_RC1e_ENABLE		(1<<20)
 #define   GEN6_RC_CTL_RC7_ENABLE		(1<<22)
+#define   VLV_RC_CTL_CTX_RST_PARALLEL		(1<<24)
 #define   GEN7_RC_CTL_TO_MODE			(1<<28)
 #define   GEN6_RC_CTL_EI_MODE(x)		((x)<<27)
 #define   GEN6_RC_CTL_HW_ENABLE			(1<<31)
diff --git a/drivers/gpu/drm/i915/i915_suspend.c b/drivers/gpu/drm/i915/i915_suspend.c
index 98790c7cccb1..8150fdc08d49 100644
--- a/drivers/gpu/drm/i915/i915_suspend.c
+++ b/drivers/gpu/drm/i915/i915_suspend.c
@@ -192,7 +192,6 @@ static void i915_restore_vga(struct drm_device *dev)
 static void i915_save_display(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	unsigned long flags;
 
 	/* Display arbitration control */
 	if (INTEL_INFO(dev)->gen <= 4)
@@ -203,46 +202,27 @@ static void i915_save_display(struct drm_device *dev)
 	if (!drm_core_check_feature(dev, DRIVER_MODESET))
 		i915_save_display_reg(dev);
 
-	spin_lock_irqsave(&dev_priv->backlight.lock, flags);
-
 	/* LVDS state */
 	if (HAS_PCH_SPLIT(dev)) {
 		dev_priv->regfile.savePP_CONTROL = I915_READ(PCH_PP_CONTROL);
-		dev_priv->regfile.saveBLC_PWM_CTL = I915_READ(BLC_PWM_PCH_CTL1);
-		dev_priv->regfile.saveBLC_PWM_CTL2 = I915_READ(BLC_PWM_PCH_CTL2);
-		dev_priv->regfile.saveBLC_CPU_PWM_CTL = I915_READ(BLC_PWM_CPU_CTL);
-		dev_priv->regfile.saveBLC_CPU_PWM_CTL2 = I915_READ(BLC_PWM_CPU_CTL2);
 		if (HAS_PCH_IBX(dev) || HAS_PCH_CPT(dev))
 			dev_priv->regfile.saveLVDS = I915_READ(PCH_LVDS);
 	} else if (IS_VALLEYVIEW(dev)) {
 		dev_priv->regfile.savePP_CONTROL = I915_READ(PP_CONTROL);
 		dev_priv->regfile.savePFIT_PGM_RATIOS = I915_READ(PFIT_PGM_RATIOS);
 
-		dev_priv->regfile.saveBLC_PWM_CTL =
-			I915_READ(VLV_BLC_PWM_CTL(PIPE_A));
 		dev_priv->regfile.saveBLC_HIST_CTL =
 			I915_READ(VLV_BLC_HIST_CTL(PIPE_A));
-		dev_priv->regfile.saveBLC_PWM_CTL2 =
-			I915_READ(VLV_BLC_PWM_CTL2(PIPE_A));
-		dev_priv->regfile.saveBLC_PWM_CTL_B =
-			I915_READ(VLV_BLC_PWM_CTL(PIPE_B));
 		dev_priv->regfile.saveBLC_HIST_CTL_B =
 			I915_READ(VLV_BLC_HIST_CTL(PIPE_B));
-		dev_priv->regfile.saveBLC_PWM_CTL2_B =
-			I915_READ(VLV_BLC_PWM_CTL2(PIPE_B));
 	} else {
 		dev_priv->regfile.savePP_CONTROL = I915_READ(PP_CONTROL);
 		dev_priv->regfile.savePFIT_PGM_RATIOS = I915_READ(PFIT_PGM_RATIOS);
-		dev_priv->regfile.saveBLC_PWM_CTL = I915_READ(BLC_PWM_CTL);
 		dev_priv->regfile.saveBLC_HIST_CTL = I915_READ(BLC_HIST_CTL);
-		if (INTEL_INFO(dev)->gen >= 4)
-			dev_priv->regfile.saveBLC_PWM_CTL2 = I915_READ(BLC_PWM_CTL2);
 		if (IS_MOBILE(dev) && !IS_I830(dev))
 			dev_priv->regfile.saveLVDS = I915_READ(LVDS);
 	}
 
-	spin_unlock_irqrestore(&dev_priv->backlight.lock, flags);
-
 	if (!IS_I830(dev) && !IS_845G(dev) && !HAS_PCH_SPLIT(dev))
 		dev_priv->regfile.savePFIT_CONTROL = I915_READ(PFIT_CONTROL);
 
@@ -257,7 +237,7 @@ static void i915_save_display(struct drm_device *dev)
 	}
 
 	/* Only regfile.save FBC state on the platform that supports FBC */
-	if (I915_HAS_FBC(dev)) {
+	if (HAS_FBC(dev)) {
 		if (HAS_PCH_SPLIT(dev)) {
 			dev_priv->regfile.saveDPFC_CB_BASE = I915_READ(ILK_DPFC_CB_BASE);
 		} else if (IS_GM45(dev)) {
@@ -278,7 +258,6 @@ static void i915_restore_display(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 mask = 0xffffffff;
-	unsigned long flags;
 
 	/* Display arbitration */
 	if (INTEL_INFO(dev)->gen <= 4)
@@ -287,12 +266,6 @@ static void i915_restore_display(struct drm_device *dev)
 	if (!drm_core_check_feature(dev, DRIVER_MODESET))
 		i915_restore_display_reg(dev);
 
-	spin_lock_irqsave(&dev_priv->backlight.lock, flags);
-
-	/* LVDS state */
-	if (INTEL_INFO(dev)->gen >= 4 && !HAS_PCH_SPLIT(dev))
-		I915_WRITE(BLC_PWM_CTL2, dev_priv->regfile.saveBLC_PWM_CTL2);
-
 	if (drm_core_check_feature(dev, DRIVER_MODESET))
 		mask = ~LVDS_PORT_EN;
 
@@ -305,13 +278,6 @@ static void i915_restore_display(struct drm_device *dev)
 		I915_WRITE(PFIT_CONTROL, dev_priv->regfile.savePFIT_CONTROL);
 
 	if (HAS_PCH_SPLIT(dev)) {
-		I915_WRITE(BLC_PWM_PCH_CTL1, dev_priv->regfile.saveBLC_PWM_CTL);
-		I915_WRITE(BLC_PWM_PCH_CTL2, dev_priv->regfile.saveBLC_PWM_CTL2);
-		/* NOTE: BLC_PWM_CPU_CTL must be written after BLC_PWM_CPU_CTL2;
-		 * otherwise we get blank eDP screen after S3 on some machines
-		 */
-		I915_WRITE(BLC_PWM_CPU_CTL2, dev_priv->regfile.saveBLC_CPU_PWM_CTL2);
-		I915_WRITE(BLC_PWM_CPU_CTL, dev_priv->regfile.saveBLC_CPU_PWM_CTL);
 		I915_WRITE(PCH_PP_ON_DELAYS, dev_priv->regfile.savePP_ON_DELAYS);
 		I915_WRITE(PCH_PP_OFF_DELAYS, dev_priv->regfile.savePP_OFF_DELAYS);
 		I915_WRITE(PCH_PP_DIVISOR, dev_priv->regfile.savePP_DIVISOR);
@@ -319,21 +285,12 @@ static void i915_restore_display(struct drm_device *dev)
 		I915_WRITE(RSTDBYCTL,
 			   dev_priv->regfile.saveMCHBAR_RENDER_STANDBY);
 	} else if (IS_VALLEYVIEW(dev)) {
-		I915_WRITE(VLV_BLC_PWM_CTL(PIPE_A),
-			   dev_priv->regfile.saveBLC_PWM_CTL);
 		I915_WRITE(VLV_BLC_HIST_CTL(PIPE_A),
 			   dev_priv->regfile.saveBLC_HIST_CTL);
-		I915_WRITE(VLV_BLC_PWM_CTL2(PIPE_A),
-			   dev_priv->regfile.saveBLC_PWM_CTL2);
-		I915_WRITE(VLV_BLC_PWM_CTL(PIPE_B),
-			   dev_priv->regfile.saveBLC_PWM_CTL);
 		I915_WRITE(VLV_BLC_HIST_CTL(PIPE_B),
 			   dev_priv->regfile.saveBLC_HIST_CTL);
-		I915_WRITE(VLV_BLC_PWM_CTL2(PIPE_B),
-			   dev_priv->regfile.saveBLC_PWM_CTL2);
 	} else {
 		I915_WRITE(PFIT_PGM_RATIOS, dev_priv->regfile.savePFIT_PGM_RATIOS);
-		I915_WRITE(BLC_PWM_CTL, dev_priv->regfile.saveBLC_PWM_CTL);
 		I915_WRITE(BLC_HIST_CTL, dev_priv->regfile.saveBLC_HIST_CTL);
 		I915_WRITE(PP_ON_DELAYS, dev_priv->regfile.savePP_ON_DELAYS);
 		I915_WRITE(PP_OFF_DELAYS, dev_priv->regfile.savePP_OFF_DELAYS);
@@ -341,11 +298,9 @@ static void i915_restore_display(struct drm_device *dev)
 		I915_WRITE(PP_CONTROL, dev_priv->regfile.savePP_CONTROL);
 	}
 
-	spin_unlock_irqrestore(&dev_priv->backlight.lock, flags);
-
 	/* only restore FBC info on the platform that supports FBC*/
 	intel_disable_fbc(dev);
-	if (I915_HAS_FBC(dev)) {
+	if (HAS_FBC(dev)) {
 		if (HAS_PCH_SPLIT(dev)) {
 			I915_WRITE(ILK_DPFC_CB_BASE, dev_priv->regfile.saveDPFC_CB_BASE);
 		} else if (IS_GM45(dev)) {
diff --git a/drivers/gpu/drm/i915/i915_sysfs.c b/drivers/gpu/drm/i915/i915_sysfs.c
index cef38fd320a7..33bcae314bf8 100644
--- a/drivers/gpu/drm/i915/i915_sysfs.c
+++ b/drivers/gpu/drm/i915/i915_sysfs.c
@@ -40,10 +40,13 @@ static u32 calc_residency(struct drm_device *dev, const u32 reg)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u64 raw_time; /* 32b value may overflow during fixed point math */
 	u64 units = 128ULL, div = 100000ULL, bias = 100ULL;
+	u32 ret;
 
 	if (!intel_enable_rc6(dev))
 		return 0;
 
+	intel_runtime_pm_get(dev_priv);
+
 	/* On VLV, residency time is in CZ units rather than 1.28us */
 	if (IS_VALLEYVIEW(dev)) {
 		u32 clkctl2;
@@ -52,7 +55,8 @@ static u32 calc_residency(struct drm_device *dev, const u32 reg)
 			CLK_CTL2_CZCOUNT_30NS_SHIFT;
 		if (!clkctl2) {
 			WARN(!clkctl2, "bogus CZ count value");
-			return 0;
+			ret = 0;
+			goto out;
 		}
 		units = DIV_ROUND_UP_ULL(30ULL * bias, (u64)clkctl2);
 		if (I915_READ(VLV_COUNTER_CONTROL) & VLV_COUNT_RANGE_HIGH)
@@ -62,7 +66,11 @@ static u32 calc_residency(struct drm_device *dev, const u32 reg)
 	}
 
 	raw_time = I915_READ(reg) * units;
-	return DIV_ROUND_UP_ULL(raw_time, div);
+	ret = DIV_ROUND_UP_ULL(raw_time, div);
+
+out:
+	intel_runtime_pm_put(dev_priv);
+	return ret;
 }
 
 static ssize_t
@@ -183,13 +191,13 @@ i915_l3_write(struct file *filp, struct kobject *kobj,
 	int slice = (int)(uintptr_t)attr->private;
 	int ret;
 
+	if (!HAS_HW_CONTEXTS(drm_dev))
+		return -ENXIO;
+
 	ret = l3_access_valid(drm_dev, offset);
 	if (ret)
 		return ret;
 
-	if (dev_priv->hw_contexts_disabled)
-		return -ENXIO;
-
 	ret = i915_mutex_lock_interruptible(drm_dev);
 	if (ret)
 		return ret;
@@ -259,7 +267,7 @@ static ssize_t gt_cur_freq_mhz_show(struct device *kdev,
 	if (IS_VALLEYVIEW(dev_priv->dev)) {
 		u32 freq;
 		freq = vlv_punit_read(dev_priv, PUNIT_REG_GPU_FREQ_STS);
-		ret = vlv_gpu_freq(dev_priv->mem_freq, (freq >> 8) & 0xff);
+		ret = vlv_gpu_freq(dev_priv, (freq >> 8) & 0xff);
 	} else {
 		ret = dev_priv->rps.cur_delay * GT_FREQUENCY_MULTIPLIER;
 	}
@@ -276,8 +284,7 @@ static ssize_t vlv_rpe_freq_mhz_show(struct device *kdev,
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
 	return snprintf(buf, PAGE_SIZE, "%d\n",
-			vlv_gpu_freq(dev_priv->mem_freq,
-				     dev_priv->rps.rpe_delay));
+			vlv_gpu_freq(dev_priv, dev_priv->rps.rpe_delay));
 }
 
 static ssize_t gt_max_freq_mhz_show(struct device *kdev, struct device_attribute *attr, char *buf)
@@ -291,7 +298,7 @@ static ssize_t gt_max_freq_mhz_show(struct device *kdev, struct device_attribute
 
 	mutex_lock(&dev_priv->rps.hw_lock);
 	if (IS_VALLEYVIEW(dev_priv->dev))
-		ret = vlv_gpu_freq(dev_priv->mem_freq, dev_priv->rps.max_delay);
+		ret = vlv_gpu_freq(dev_priv, dev_priv->rps.max_delay);
 	else
 		ret = dev_priv->rps.max_delay * GT_FREQUENCY_MULTIPLIER;
 	mutex_unlock(&dev_priv->rps.hw_lock);
@@ -318,7 +325,7 @@ static ssize_t gt_max_freq_mhz_store(struct device *kdev,
 	mutex_lock(&dev_priv->rps.hw_lock);
 
 	if (IS_VALLEYVIEW(dev_priv->dev)) {
-		val = vlv_freq_opcode(dev_priv->mem_freq, val);
+		val = vlv_freq_opcode(dev_priv, val);
 
 		hw_max = valleyview_rps_max_freq(dev_priv);
 		hw_min = valleyview_rps_min_freq(dev_priv);
@@ -342,15 +349,15 @@ static ssize_t gt_max_freq_mhz_store(struct device *kdev,
 		DRM_DEBUG("User requested overclocking to %d\n",
 			  val * GT_FREQUENCY_MULTIPLIER);
 
+	dev_priv->rps.max_delay = val;
+
 	if (dev_priv->rps.cur_delay > val) {
-		if (IS_VALLEYVIEW(dev_priv->dev))
-			valleyview_set_rps(dev_priv->dev, val);
+		if (IS_VALLEYVIEW(dev))
+			valleyview_set_rps(dev, val);
 		else
-			gen6_set_rps(dev_priv->dev, val);
+			gen6_set_rps(dev, val);
 	}
 
-	dev_priv->rps.max_delay = val;
-
 	mutex_unlock(&dev_priv->rps.hw_lock);
 
 	return count;
@@ -367,7 +374,7 @@ static ssize_t gt_min_freq_mhz_show(struct device *kdev, struct device_attribute
 
 	mutex_lock(&dev_priv->rps.hw_lock);
 	if (IS_VALLEYVIEW(dev_priv->dev))
-		ret = vlv_gpu_freq(dev_priv->mem_freq, dev_priv->rps.min_delay);
+		ret = vlv_gpu_freq(dev_priv, dev_priv->rps.min_delay);
 	else
 		ret = dev_priv->rps.min_delay * GT_FREQUENCY_MULTIPLIER;
 	mutex_unlock(&dev_priv->rps.hw_lock);
@@ -394,7 +401,7 @@ static ssize_t gt_min_freq_mhz_store(struct device *kdev,
 	mutex_lock(&dev_priv->rps.hw_lock);
 
 	if (IS_VALLEYVIEW(dev)) {
-		val = vlv_freq_opcode(dev_priv->mem_freq, val);
+		val = vlv_freq_opcode(dev_priv, val);
 
 		hw_max = valleyview_rps_max_freq(dev_priv);
 		hw_min = valleyview_rps_min_freq(dev_priv);
@@ -411,15 +418,15 @@ static ssize_t gt_min_freq_mhz_store(struct device *kdev,
 		return -EINVAL;
 	}
 
+	dev_priv->rps.min_delay = val;
+
 	if (dev_priv->rps.cur_delay < val) {
 		if (IS_VALLEYVIEW(dev))
 			valleyview_set_rps(dev, val);
 		else
-			gen6_set_rps(dev_priv->dev, val);
+			gen6_set_rps(dev, val);
 	}
 
-	dev_priv->rps.min_delay = val;
-
 	mutex_unlock(&dev_priv->rps.hw_lock);
 
 	return count;
@@ -449,7 +456,9 @@ static ssize_t gt_rp_mhz_show(struct device *kdev, struct device_attribute *attr
 	ret = mutex_lock_interruptible(&dev->struct_mutex);
 	if (ret)
 		return ret;
+	intel_runtime_pm_get(dev_priv);
 	rp_state_cap = I915_READ(GEN6_RP_STATE_CAP);
+	intel_runtime_pm_put(dev_priv);
 	mutex_unlock(&dev->struct_mutex);
 
 	if (attr == &dev_attr_gt_RP0_freq_mhz) {
diff --git a/drivers/gpu/drm/i915/i915_ums.c b/drivers/gpu/drm/i915/i915_ums.c
index 967da4772c44..caa18e855815 100644
--- a/drivers/gpu/drm/i915/i915_ums.c
+++ b/drivers/gpu/drm/i915/i915_ums.c
@@ -270,6 +270,18 @@ void i915_save_display_reg(struct drm_device *dev)
 	}
 	/* FIXME: regfile.save TV & SDVO state */
 
+	/* Backlight */
+	if (HAS_PCH_SPLIT(dev)) {
+		dev_priv->regfile.saveBLC_PWM_CTL = I915_READ(BLC_PWM_PCH_CTL1);
+		dev_priv->regfile.saveBLC_PWM_CTL2 = I915_READ(BLC_PWM_PCH_CTL2);
+		dev_priv->regfile.saveBLC_CPU_PWM_CTL = I915_READ(BLC_PWM_CPU_CTL);
+		dev_priv->regfile.saveBLC_CPU_PWM_CTL2 = I915_READ(BLC_PWM_CPU_CTL2);
+	} else {
+		dev_priv->regfile.saveBLC_PWM_CTL = I915_READ(BLC_PWM_CTL);
+		if (INTEL_INFO(dev)->gen >= 4)
+			dev_priv->regfile.saveBLC_PWM_CTL2 = I915_READ(BLC_PWM_CTL2);
+	}
+
 	return;
 }
 
@@ -280,6 +292,21 @@ void i915_restore_display_reg(struct drm_device *dev)
 	int dpll_b_reg, fpb0_reg, fpb1_reg;
 	int i;
 
+	/* Backlight */
+	if (HAS_PCH_SPLIT(dev)) {
+		I915_WRITE(BLC_PWM_PCH_CTL1, dev_priv->regfile.saveBLC_PWM_CTL);
+		I915_WRITE(BLC_PWM_PCH_CTL2, dev_priv->regfile.saveBLC_PWM_CTL2);
+		/* NOTE: BLC_PWM_CPU_CTL must be written after BLC_PWM_CPU_CTL2;
+		 * otherwise we get blank eDP screen after S3 on some machines
+		 */
+		I915_WRITE(BLC_PWM_CPU_CTL2, dev_priv->regfile.saveBLC_CPU_PWM_CTL2);
+		I915_WRITE(BLC_PWM_CPU_CTL, dev_priv->regfile.saveBLC_CPU_PWM_CTL);
+	} else {
+		if (INTEL_INFO(dev)->gen >= 4)
+			I915_WRITE(BLC_PWM_CTL2, dev_priv->regfile.saveBLC_PWM_CTL2);
+		I915_WRITE(BLC_PWM_CTL, dev_priv->regfile.saveBLC_PWM_CTL);
+	}
+
 	/* Display port ratios (must be done before clock is set) */
 	if (SUPPORTS_INTEGRATED_DP(dev)) {
 		I915_WRITE(_PIPEA_DATA_M_G4X, dev_priv->regfile.savePIPEA_GMCH_DATA_M);
diff --git a/drivers/gpu/drm/i915/intel_bios.c b/drivers/gpu/drm/i915/intel_bios.c
index e4fba39631a5..f22041973f3a 100644
--- a/drivers/gpu/drm/i915/intel_bios.c
+++ b/drivers/gpu/drm/i915/intel_bios.c
@@ -281,6 +281,34 @@ parse_lfp_panel_data(struct drm_i915_private *dev_priv,
 	}
 }
 
+static void
+parse_lfp_backlight(struct drm_i915_private *dev_priv, struct bdb_header *bdb)
+{
+	const struct bdb_lfp_backlight_data *backlight_data;
+	const struct bdb_lfp_backlight_data_entry *entry;
+
+	backlight_data = find_section(bdb, BDB_LVDS_BACKLIGHT);
+	if (!backlight_data)
+		return;
+
+	if (backlight_data->entry_size != sizeof(backlight_data->data[0])) {
+		DRM_DEBUG_KMS("Unsupported backlight data entry size %u\n",
+			      backlight_data->entry_size);
+		return;
+	}
+
+	entry = &backlight_data->data[panel_type];
+
+	dev_priv->vbt.backlight.pwm_freq_hz = entry->pwm_freq_hz;
+	dev_priv->vbt.backlight.active_low_pwm = entry->active_low_pwm;
+	DRM_DEBUG_KMS("VBT backlight PWM modulation frequency %u Hz, "
+		      "active %s, min brightness %u, level %u\n",
+		      dev_priv->vbt.backlight.pwm_freq_hz,
+		      dev_priv->vbt.backlight.active_low_pwm ? "low" : "high",
+		      entry->min_brightness,
+		      backlight_data->level[panel_type]);
+}
+
 /* Try to find sdvo panel data */
 static void
 parse_sdvo_panel_data(struct drm_i915_private *dev_priv,
@@ -327,12 +355,12 @@ static int intel_bios_ssc_frequency(struct drm_device *dev,
 {
 	switch (INTEL_INFO(dev)->gen) {
 	case 2:
-		return alternate ? 66 : 48;
+		return alternate ? 66667 : 48000;
 	case 3:
 	case 4:
-		return alternate ? 100 : 96;
+		return alternate ? 100000 : 96000;
 	default:
-		return alternate ? 100 : 120;
+		return alternate ? 100000 : 120000;
 	}
 }
 
@@ -796,7 +824,7 @@ init_vbt_defaults(struct drm_i915_private *dev_priv)
 	 */
 	dev_priv->vbt.lvds_ssc_freq = intel_bios_ssc_frequency(dev,
 			!HAS_PCH_SPLIT(dev));
-	DRM_DEBUG_KMS("Set default to SSC at %dMHz\n", dev_priv->vbt.lvds_ssc_freq);
+	DRM_DEBUG_KMS("Set default to SSC at %d kHz\n", dev_priv->vbt.lvds_ssc_freq);
 
 	for (port = PORT_A; port < I915_MAX_PORTS; port++) {
 		struct ddi_vbt_port_info *info =
@@ -894,6 +922,7 @@ intel_parse_bios(struct drm_device *dev)
 	parse_general_features(dev_priv, bdb);
 	parse_general_definitions(dev_priv, bdb);
 	parse_lfp_panel_data(dev_priv, bdb);
+	parse_lfp_backlight(dev_priv, bdb);
 	parse_sdvo_panel_data(dev_priv, bdb);
 	parse_sdvo_device_mapping(dev_priv, bdb);
 	parse_device_mapping(dev_priv, bdb);
diff --git a/drivers/gpu/drm/i915/intel_bios.h b/drivers/gpu/drm/i915/intel_bios.h
index f580a2b0ddd3..282de5e9f39d 100644
--- a/drivers/gpu/drm/i915/intel_bios.h
+++ b/drivers/gpu/drm/i915/intel_bios.h
@@ -39,7 +39,7 @@ struct vbt_header {
 	u8 reserved0;
 	u32 bdb_offset;			/**< from beginning of VBT */
 	u32 aim_offset[4];		/**< from beginning of VBT */
-} __attribute__((packed));
+} __packed;
 
 struct bdb_header {
 	u8 signature[16];		/**< Always 'BIOS_DATA_BLOCK' */
@@ -65,7 +65,7 @@ struct vbios_data {
 	u8 rsvd4; /* popup memory size */
 	u8 resize_pci_bios;
 	u8 rsvd5; /* is crt already on ddc2 */
-} __attribute__((packed));
+} __packed;
 
 /*
  * There are several types of BIOS data blocks (BDBs), each block has
@@ -142,7 +142,7 @@ struct bdb_general_features {
 	u8 dp_ssc_enb:1;	/* PCH attached eDP supports SSC */
 	u8 dp_ssc_freq:1;	/* SSC freq for PCH attached eDP */
 	u8 rsvd11:3; /* finish byte */
-} __attribute__((packed));
+} __packed;
 
 /* pre-915 */
 #define GPIO_PIN_DVI_LVDS	0x03 /* "DVI/LVDS DDC GPIO pins" */
@@ -225,7 +225,7 @@ struct old_child_dev_config {
 	u8  dvo2_wiring;
 	u16 extended_type;
 	u8  dvo_function;
-} __attribute__((packed));
+} __packed;
 
 /* This one contains field offsets that are known to be common for all BDB
  * versions. Notice that the meaning of the contents contents may still change,
@@ -238,7 +238,7 @@ struct common_child_dev_config {
 	u8 not_common2[2];
 	u8 ddc_pin;
 	u16 edid_ptr;
-} __attribute__((packed));
+} __packed;
 
 /* This field changes depending on the BDB version, so the most reliable way to
  * read it is by checking the BDB version and reading the raw pointer. */
@@ -279,7 +279,7 @@ struct bdb_general_definitions {
 	 *	     sizeof(child_device_config);
 	 */
 	union child_device_config devices[0];
-} __attribute__((packed));
+} __packed;
 
 struct bdb_lvds_options {
 	u8 panel_type;
@@ -293,7 +293,7 @@ struct bdb_lvds_options {
 	u8 lvds_edid:1;
 	u8 rsvd2:1;
 	u8 rsvd4;
-} __attribute__((packed));
+} __packed;
 
 /* LFP pointer table contains entries to the struct below */
 struct bdb_lvds_lfp_data_ptr {
@@ -303,12 +303,12 @@ struct bdb_lvds_lfp_data_ptr {
 	u8 dvo_table_size;
 	u16 panel_pnp_id_offset;
 	u8 pnp_table_size;
-} __attribute__((packed));
+} __packed;
 
 struct bdb_lvds_lfp_data_ptrs {
 	u8 lvds_entries; /* followed by one or more lvds_data_ptr structs */
 	struct bdb_lvds_lfp_data_ptr ptr[16];
-} __attribute__((packed));
+} __packed;
 
 /* LFP data has 3 blocks per entry */
 struct lvds_fp_timing {
@@ -325,7 +325,7 @@ struct lvds_fp_timing {
 	u32 pfit_reg;
 	u32 pfit_reg_val;
 	u16 terminator;
-} __attribute__((packed));
+} __packed;
 
 struct lvds_dvo_timing {
 	u16 clock;		/**< In 10khz */
@@ -353,7 +353,7 @@ struct lvds_dvo_timing {
 	u8 vsync_positive:1;
 	u8 hsync_positive:1;
 	u8 rsvd2:1;
-} __attribute__((packed));
+} __packed;
 
 struct lvds_pnp_id {
 	u16 mfg_name;
@@ -361,17 +361,33 @@ struct lvds_pnp_id {
 	u32 serial;
 	u8 mfg_week;
 	u8 mfg_year;
-} __attribute__((packed));
+} __packed;
 
 struct bdb_lvds_lfp_data_entry {
 	struct lvds_fp_timing fp_timing;
 	struct lvds_dvo_timing dvo_timing;
 	struct lvds_pnp_id pnp_id;
-} __attribute__((packed));
+} __packed;
 
 struct bdb_lvds_lfp_data {
 	struct bdb_lvds_lfp_data_entry data[16];
-} __attribute__((packed));
+} __packed;
+
+struct bdb_lfp_backlight_data_entry {
+	u8 type:2;
+	u8 active_low_pwm:1;
+	u8 obsolete1:5;
+	u16 pwm_freq_hz;
+	u8 min_brightness;
+	u8 obsolete2;
+	u8 obsolete3;
+} __packed;
+
+struct bdb_lfp_backlight_data {
+	u8 entry_size;
+	struct bdb_lfp_backlight_data_entry data[16];
+	u8 level[16];
+} __packed;
 
 struct aimdb_header {
 	char signature[16];
@@ -379,12 +395,12 @@ struct aimdb_header {
 	u16 aimdb_version;
 	u16 aimdb_header_size;
 	u16 aimdb_size;
-} __attribute__((packed));
+} __packed;
 
 struct aimdb_block {
 	u8 aimdb_id;
 	u16 aimdb_size;
-} __attribute__((packed));
+} __packed;
 
 struct vch_panel_data {
 	u16 fp_timing_offset;
@@ -395,12 +411,12 @@ struct vch_panel_data {
 	u8 text_fitting_size;
 	u16 graphics_fitting_offset;
 	u8 graphics_fitting_size;
-} __attribute__((packed));
+} __packed;
 
 struct vch_bdb_22 {
 	struct aimdb_block aimdb_block;
 	struct vch_panel_data panels[16];
-} __attribute__((packed));
+} __packed;
 
 struct bdb_sdvo_lvds_options {
 	u8 panel_backlight;
@@ -416,7 +432,7 @@ struct bdb_sdvo_lvds_options {
 	u8 panel_misc_bits_2;
 	u8 panel_misc_bits_3;
 	u8 panel_misc_bits_4;
-} __attribute__((packed));
+} __packed;
 
 
 #define BDB_DRIVER_FEATURE_NO_LVDS		0
@@ -462,7 +478,7 @@ struct bdb_driver_features {
 
 	u8 hdmi_termination;
 	u8 custom_vbt_version;
-} __attribute__((packed));
+} __packed;
 
 #define EDP_18BPP	0
 #define EDP_24BPP	1
@@ -487,14 +503,14 @@ struct edp_power_seq {
 	u16 t9;
 	u16 t10;
 	u16 t11_t12;
-} __attribute__ ((packed));
+} __packed;
 
 struct edp_link_params {
 	u8 rate:4;
 	u8 lanes:4;
 	u8 preemphasis:4;
 	u8 vswing:4;
-} __attribute__ ((packed));
+} __packed;
 
 struct bdb_edp {
 	struct edp_power_seq power_seqs[16];
@@ -505,7 +521,7 @@ struct bdb_edp {
 	/* ith bit indicates enabled/disabled for (i+1)th panel */
 	u16 edp_s3d_feature;
 	u16 edp_t3_optimization;
-} __attribute__ ((packed));
+} __packed;
 
 void intel_setup_bios(struct drm_device *dev);
 int intel_parse_bios(struct drm_device *dev);
@@ -733,6 +749,6 @@ struct bdb_mipi {
 	u32 hl_switch_cnt;
 	u32 lp_byte_clk;
 	u32 clk_lane_switch_cnt;
-} __attribute__((packed));
+} __packed;
 
 #endif /* _I830_BIOS_H_ */
diff --git a/drivers/gpu/drm/i915/intel_crt.c b/drivers/gpu/drm/i915/intel_crt.c
index b5b1b9b23adf..e2e39e65f109 100644
--- a/drivers/gpu/drm/i915/intel_crt.c
+++ b/drivers/gpu/drm/i915/intel_crt.c
@@ -222,8 +222,9 @@ static void intel_crt_dpms(struct drm_connector *connector, int mode)
 	intel_modeset_check_state(connector->dev);
 }
 
-static int intel_crt_mode_valid(struct drm_connector *connector,
-				struct drm_display_mode *mode)
+static enum drm_mode_status
+intel_crt_mode_valid(struct drm_connector *connector,
+		     struct drm_display_mode *mode)
 {
 	struct drm_device *dev = connector->dev;
 
diff --git a/drivers/gpu/drm/i915/intel_ddi.c b/drivers/gpu/drm/i915/intel_ddi.c
index b69dc3e66c16..e06b9e017d6b 100644
--- a/drivers/gpu/drm/i915/intel_ddi.c
+++ b/drivers/gpu/drm/i915/intel_ddi.c
@@ -73,7 +73,7 @@ static const u32 hsw_ddi_translations_hdmi[] = {
 };
 
 static const u32 bdw_ddi_translations_edp[] = {
-	0x00FFFFFF, 0x00000012,		/* DP parameters */
+	0x00FFFFFF, 0x00000012,		/* eDP parameters */
 	0x00EBAFFF, 0x00020011,
 	0x00C71FFF, 0x0006000F,
 	0x00FFFFFF, 0x00020011,
@@ -696,25 +696,25 @@ intel_ddi_calculate_wrpll(int clock /* in Hz */,
 	*n2_out = best.n2;
 	*p_out = best.p;
 	*r2_out = best.r2;
-
-	DRM_DEBUG_KMS("WRPLL: %dHz refresh rate with p=%d, n2=%d r2=%d\n",
-		      clock, *p_out, *n2_out, *r2_out);
 }
 
-bool intel_ddi_pll_mode_set(struct drm_crtc *crtc)
+/*
+ * 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)
 {
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	struct drm_crtc *crtc = &intel_crtc->base;
 	struct intel_encoder *intel_encoder = intel_ddi_get_crtc_encoder(crtc);
 	struct drm_encoder *encoder = &intel_encoder->base;
 	struct drm_i915_private *dev_priv = crtc->dev->dev_private;
 	struct intel_ddi_plls *plls = &dev_priv->ddi_plls;
 	int type = intel_encoder->type;
 	enum pipe pipe = intel_crtc->pipe;
-	uint32_t reg, val;
 	int clock = intel_crtc->config.port_clock;
 
-	/* TODO: reuse PLLs when possible (compare values) */
-
 	intel_ddi_put_crtc_pll(crtc);
 
 	if (type == INTEL_OUTPUT_DISPLAYPORT || type == INTEL_OUTPUT_EDP) {
@@ -736,66 +736,145 @@ bool intel_ddi_pll_mode_set(struct drm_crtc *crtc)
 			return false;
 		}
 
-		/* We don't need to turn any PLL on because we'll use LCPLL. */
-		return true;
-
 	} else if (type == INTEL_OUTPUT_HDMI) {
+		uint32_t reg, val;
 		unsigned p, n2, r2;
 
-		if (plls->wrpll1_refcount == 0) {
+		intel_ddi_calculate_wrpll(clock * 1000, &r2, &n2, &p);
+
+		val = WRPLL_PLL_ENABLE | WRPLL_PLL_SELECT_LCPLL_2700 |
+		      WRPLL_DIVIDER_REFERENCE(r2) | WRPLL_DIVIDER_FEEDBACK(n2) |
+		      WRPLL_DIVIDER_POST(p);
+
+		if (val == I915_READ(WRPLL_CTL1)) {
+			DRM_DEBUG_KMS("Reusing WRPLL 1 on pipe %c\n",
+				      pipe_name(pipe));
+			reg = WRPLL_CTL1;
+		} else if (val == I915_READ(WRPLL_CTL2)) {
+			DRM_DEBUG_KMS("Reusing WRPLL 2 on pipe %c\n",
+				      pipe_name(pipe));
+			reg = WRPLL_CTL2;
+		} else if (plls->wrpll1_refcount == 0) {
 			DRM_DEBUG_KMS("Using WRPLL 1 on pipe %c\n",
 				      pipe_name(pipe));
-			plls->wrpll1_refcount++;
 			reg = WRPLL_CTL1;
-			intel_crtc->ddi_pll_sel = PORT_CLK_SEL_WRPLL1;
 		} else if (plls->wrpll2_refcount == 0) {
 			DRM_DEBUG_KMS("Using WRPLL 2 on pipe %c\n",
 				      pipe_name(pipe));
-			plls->wrpll2_refcount++;
 			reg = WRPLL_CTL2;
-			intel_crtc->ddi_pll_sel = PORT_CLK_SEL_WRPLL2;
 		} else {
 			DRM_ERROR("No WRPLLs available!\n");
 			return false;
 		}
 
-		WARN(I915_READ(reg) & WRPLL_PLL_ENABLE,
-		     "WRPLL already enabled\n");
+		DRM_DEBUG_KMS("WRPLL: %dKHz refresh rate with p=%d, n2=%d r2=%d\n",
+			      clock, p, n2, r2);
 
-		intel_ddi_calculate_wrpll(clock * 1000, &r2, &n2, &p);
-
-		val = WRPLL_PLL_ENABLE | WRPLL_PLL_SELECT_LCPLL_2700 |
-		      WRPLL_DIVIDER_REFERENCE(r2) | WRPLL_DIVIDER_FEEDBACK(n2) |
-		      WRPLL_DIVIDER_POST(p);
+		if (reg == WRPLL_CTL1) {
+			plls->wrpll1_refcount++;
+			intel_crtc->ddi_pll_sel = PORT_CLK_SEL_WRPLL1;
+		} else {
+			plls->wrpll2_refcount++;
+			intel_crtc->ddi_pll_sel = PORT_CLK_SEL_WRPLL2;
+		}
 
 	} else if (type == INTEL_OUTPUT_ANALOG) {
 		if (plls->spll_refcount == 0) {
 			DRM_DEBUG_KMS("Using SPLL on pipe %c\n",
 				      pipe_name(pipe));
 			plls->spll_refcount++;
-			reg = SPLL_CTL;
 			intel_crtc->ddi_pll_sel = PORT_CLK_SEL_SPLL;
 		} else {
 			DRM_ERROR("SPLL already in use\n");
 			return false;
 		}
 
-		WARN(I915_READ(reg) & SPLL_PLL_ENABLE,
-		     "SPLL already enabled\n");
-
-		val = SPLL_PLL_ENABLE | SPLL_PLL_FREQ_1350MHz | SPLL_PLL_SSC;
-
 	} else {
 		WARN(1, "Invalid DDI encoder type %d\n", type);
 		return false;
 	}
 
-	I915_WRITE(reg, val);
-	udelay(20);
-
 	return true;
 }
 
+/*
+ * To be called after intel_ddi_pll_select(). That one selects the PLL to be
+ * used, this one actually enables the PLL.
+ */
+void intel_ddi_pll_enable(struct intel_crtc *crtc)
+{
+	struct drm_device *dev = crtc->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_ddi_plls *plls = &dev_priv->ddi_plls;
+	int clock = crtc->config.port_clock;
+	uint32_t reg, cur_val, new_val;
+	int refcount;
+	const char *pll_name;
+	uint32_t enable_bit = (1 << 31);
+	unsigned int p, n2, r2;
+
+	BUILD_BUG_ON(enable_bit != SPLL_PLL_ENABLE);
+	BUILD_BUG_ON(enable_bit != WRPLL_PLL_ENABLE);
+
+	switch (crtc->ddi_pll_sel) {
+	case PORT_CLK_SEL_LCPLL_2700:
+	case PORT_CLK_SEL_LCPLL_1350:
+	case PORT_CLK_SEL_LCPLL_810:
+		/*
+		 * LCPLL should always be enabled at this point of the mode set
+		 * sequence, so nothing to do.
+		 */
+		return;
+
+	case PORT_CLK_SEL_SPLL:
+		pll_name = "SPLL";
+		reg = SPLL_CTL;
+		refcount = plls->spll_refcount;
+		new_val = SPLL_PLL_ENABLE | SPLL_PLL_FREQ_1350MHz |
+			  SPLL_PLL_SSC;
+		break;
+
+	case PORT_CLK_SEL_WRPLL1:
+	case PORT_CLK_SEL_WRPLL2:
+		if (crtc->ddi_pll_sel == PORT_CLK_SEL_WRPLL1) {
+			pll_name = "WRPLL1";
+			reg = WRPLL_CTL1;
+			refcount = plls->wrpll1_refcount;
+		} else {
+			pll_name = "WRPLL2";
+			reg = WRPLL_CTL2;
+			refcount = plls->wrpll2_refcount;
+		}
+
+		intel_ddi_calculate_wrpll(clock * 1000, &r2, &n2, &p);
+
+		new_val = WRPLL_PLL_ENABLE | WRPLL_PLL_SELECT_LCPLL_2700 |
+			  WRPLL_DIVIDER_REFERENCE(r2) |
+			  WRPLL_DIVIDER_FEEDBACK(n2) | WRPLL_DIVIDER_POST(p);
+
+		break;
+
+	case PORT_CLK_SEL_NONE:
+		WARN(1, "Bad selected pll: PORT_CLK_SEL_NONE\n");
+		return;
+	default:
+		WARN(1, "Bad selected pll: 0x%08x\n", crtc->ddi_pll_sel);
+		return;
+	}
+
+	cur_val = I915_READ(reg);
+
+	WARN(refcount < 1, "Bad %s refcount: %d\n", pll_name, refcount);
+	if (refcount == 1) {
+		WARN(cur_val & enable_bit, "%s already enabled\n", pll_name);
+		I915_WRITE(reg, new_val);
+		POSTING_READ(reg);
+		udelay(20);
+	} else {
+		WARN((cur_val & enable_bit) == 0, "%s disabled\n", pll_name);
+	}
+}
+
 void intel_ddi_set_pipe_settings(struct drm_crtc *crtc)
 {
 	struct drm_i915_private *dev_priv = crtc->dev->dev_private;
@@ -1121,9 +1200,7 @@ static void intel_ddi_pre_enable(struct intel_encoder *intel_encoder)
 
 	if (type == INTEL_OUTPUT_EDP) {
 		struct intel_dp *intel_dp = enc_to_intel_dp(encoder);
-		ironlake_edp_panel_vdd_on(intel_dp);
 		ironlake_edp_panel_on(intel_dp);
-		ironlake_edp_panel_vdd_off(intel_dp, true);
 	}
 
 	WARN_ON(intel_crtc->ddi_pll_sel == PORT_CLK_SEL_NONE);
@@ -1166,7 +1243,6 @@ static void intel_ddi_post_disable(struct intel_encoder *intel_encoder)
 
 	if (type == INTEL_OUTPUT_DISPLAYPORT || type == INTEL_OUTPUT_EDP) {
 		struct intel_dp *intel_dp = enc_to_intel_dp(encoder);
-		ironlake_edp_panel_vdd_on(intel_dp);
 		intel_dp_sink_dpms(intel_dp, DRM_MODE_DPMS_OFF);
 		ironlake_edp_panel_off(intel_dp);
 	}
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 2bde35d34eb9..9fa24347963a 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -90,8 +90,8 @@ intel_fdi_link_freq(struct drm_device *dev)
 
 static const intel_limit_t intel_limits_i8xx_dac = {
 	.dot = { .min = 25000, .max = 350000 },
-	.vco = { .min = 930000, .max = 1400000 },
-	.n = { .min = 3, .max = 16 },
+	.vco = { .min = 908000, .max = 1512000 },
+	.n = { .min = 2, .max = 16 },
 	.m = { .min = 96, .max = 140 },
 	.m1 = { .min = 18, .max = 26 },
 	.m2 = { .min = 6, .max = 16 },
@@ -103,8 +103,8 @@ static const intel_limit_t intel_limits_i8xx_dac = {
 
 static const intel_limit_t intel_limits_i8xx_dvo = {
 	.dot = { .min = 25000, .max = 350000 },
-	.vco = { .min = 930000, .max = 1400000 },
-	.n = { .min = 3, .max = 16 },
+	.vco = { .min = 908000, .max = 1512000 },
+	.n = { .min = 2, .max = 16 },
 	.m = { .min = 96, .max = 140 },
 	.m1 = { .min = 18, .max = 26 },
 	.m2 = { .min = 6, .max = 16 },
@@ -116,8 +116,8 @@ static const intel_limit_t intel_limits_i8xx_dvo = {
 
 static const intel_limit_t intel_limits_i8xx_lvds = {
 	.dot = { .min = 25000, .max = 350000 },
-	.vco = { .min = 930000, .max = 1400000 },
-	.n = { .min = 3, .max = 16 },
+	.vco = { .min = 908000, .max = 1512000 },
+	.n = { .min = 2, .max = 16 },
 	.m = { .min = 96, .max = 140 },
 	.m1 = { .min = 18, .max = 26 },
 	.m2 = { .min = 6, .max = 16 },
@@ -329,6 +329,8 @@ static void vlv_clock(int refclk, intel_clock_t *clock)
 {
 	clock->m = clock->m1 * clock->m2;
 	clock->p = clock->p1 * clock->p2;
+	if (WARN_ON(clock->n == 0 || clock->p == 0))
+		return;
 	clock->vco = DIV_ROUND_CLOSEST(refclk * clock->m, clock->n);
 	clock->dot = DIV_ROUND_CLOSEST(clock->vco, clock->p);
 }
@@ -430,6 +432,8 @@ static void pineview_clock(int refclk, intel_clock_t *clock)
 {
 	clock->m = clock->m2 + 2;
 	clock->p = clock->p1 * clock->p2;
+	if (WARN_ON(clock->n == 0 || clock->p == 0))
+		return;
 	clock->vco = DIV_ROUND_CLOSEST(refclk * clock->m, clock->n);
 	clock->dot = DIV_ROUND_CLOSEST(clock->vco, clock->p);
 }
@@ -443,6 +447,8 @@ static void i9xx_clock(int refclk, intel_clock_t *clock)
 {
 	clock->m = i9xx_dpll_compute_m(clock);
 	clock->p = clock->p1 * clock->p2;
+	if (WARN_ON(clock->n + 2 == 0 || clock->p == 0))
+		return;
 	clock->vco = DIV_ROUND_CLOSEST(refclk * clock->m, clock->n + 2);
 	clock->dot = DIV_ROUND_CLOSEST(clock->vco, clock->p);
 }
@@ -748,10 +754,10 @@ enum transcoder intel_pipe_to_cpu_transcoder(struct drm_i915_private *dev_priv,
 	return intel_crtc->config.cpu_transcoder;
 }
 
-static void ironlake_wait_for_vblank(struct drm_device *dev, int pipe)
+static void g4x_wait_for_vblank(struct drm_device *dev, int pipe)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	u32 frame, frame_reg = PIPEFRAME(pipe);
+	u32 frame, frame_reg = PIPE_FRMCOUNT_GM45(pipe);
 
 	frame = I915_READ(frame_reg);
 
@@ -772,8 +778,8 @@ void intel_wait_for_vblank(struct drm_device *dev, int pipe)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int pipestat_reg = PIPESTAT(pipe);
 
-	if (INTEL_INFO(dev)->gen >= 5) {
-		ironlake_wait_for_vblank(dev, pipe);
+	if (IS_G4X(dev) || INTEL_INFO(dev)->gen >= 5) {
+		g4x_wait_for_vblank(dev, pipe);
 		return;
 	}
 
@@ -1205,15 +1211,12 @@ static void assert_sprites_disabled(struct drm_i915_private *dev_priv,
 	}
 }
 
-static void assert_pch_refclk_enabled(struct drm_i915_private *dev_priv)
+static void ibx_assert_pch_refclk_enabled(struct drm_i915_private *dev_priv)
 {
 	u32 val;
 	bool enabled;
 
-	if (HAS_PCH_LPT(dev_priv->dev)) {
-		DRM_DEBUG_DRIVER("LPT does not has PCH refclk, skipping check\n");
-		return;
-	}
+	WARN_ON(!(HAS_PCH_IBX(dev_priv->dev) || HAS_PCH_CPT(dev_priv->dev)));
 
 	val = I915_READ(PCH_DREF_CONTROL);
 	enabled = !!(val & (DREF_SSC_SOURCE_MASK | DREF_NONSPREAD_SOURCE_MASK |
@@ -1361,6 +1364,24 @@ static void intel_init_dpio(struct drm_device *dev)
 	if (!IS_VALLEYVIEW(dev))
 		return;
 
+	DPIO_PHY_IOSF_PORT(DPIO_PHY0) = IOSF_PORT_DPIO;
+}
+
+static void intel_reset_dpio(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	if (!IS_VALLEYVIEW(dev))
+		return;
+
+	/*
+	 * Enable the CRI clock source so we can get at the display and the
+	 * reference clock for VGA hotplug / manual detection.
+	 */
+	I915_WRITE(DPLL(PIPE_B), I915_READ(DPLL(PIPE_B)) |
+		   DPLL_REFA_CLK_ENABLE_VLV |
+		   DPLL_INTEGRATED_CRI_CLK_VLV);
+
 	/*
 	 * From VLV2A0_DP_eDP_DPIO_driver_vbios_notes_10.docx -
 	 *  6.	De-assert cmn_reset/side_reset. Same as VLV X0.
@@ -1487,25 +1508,35 @@ static void vlv_disable_pll(struct drm_i915_private *dev_priv, enum pipe pipe)
 	/* Make sure the pipe isn't still relying on us */
 	assert_pipe_disabled(dev_priv, pipe);
 
-	/* Leave integrated clock source enabled */
+	/*
+	 * Leave integrated clock source and reference clock enabled for pipe B.
+	 * The latter is needed for VGA hotplug / manual detection.
+	 */
 	if (pipe == PIPE_B)
-		val = DPLL_INTEGRATED_CRI_CLK_VLV;
+		val = DPLL_INTEGRATED_CRI_CLK_VLV | DPLL_REFA_CLK_ENABLE_VLV;
 	I915_WRITE(DPLL(pipe), val);
 	POSTING_READ(DPLL(pipe));
 }
 
-void vlv_wait_port_ready(struct drm_i915_private *dev_priv, int port)
+void vlv_wait_port_ready(struct drm_i915_private *dev_priv,
+		struct intel_digital_port *dport)
 {
 	u32 port_mask;
 
-	if (!port)
+	switch (dport->port) {
+	case PORT_B:
 		port_mask = DPLL_PORTB_READY_MASK;
-	else
+		break;
+	case PORT_C:
 		port_mask = DPLL_PORTC_READY_MASK;
+		break;
+	default:
+		BUG();
+	}
 
 	if (wait_for((I915_READ(DPLL(0)) & port_mask) == 0, 1000))
 		WARN(1, "timed out waiting for port %c ready: 0x%08x\n",
-		     'B' + port, I915_READ(DPLL(0)));
+		     port_name(dport->port), I915_READ(DPLL(0)));
 }
 
 /**
@@ -2083,8 +2114,8 @@ static int i9xx_update_plane(struct drm_crtc *crtc, struct drm_framebuffer *fb,
 		      fb->pitches[0]);
 	I915_WRITE(DSPSTRIDE(plane), fb->pitches[0]);
 	if (INTEL_INFO(dev)->gen >= 4) {
-		I915_MODIFY_DISPBASE(DSPSURF(plane),
-				     i915_gem_obj_ggtt_offset(obj) + intel_crtc->dspaddr_offset);
+		I915_WRITE(DSPSURF(plane),
+			   i915_gem_obj_ggtt_offset(obj) + intel_crtc->dspaddr_offset);
 		I915_WRITE(DSPTILEOFF(plane), (y << 16) | x);
 		I915_WRITE(DSPLINOFF(plane), linear_offset);
 	} else
@@ -2174,8 +2205,8 @@ static int ironlake_update_plane(struct drm_crtc *crtc,
 		      i915_gem_obj_ggtt_offset(obj), linear_offset, x, y,
 		      fb->pitches[0]);
 	I915_WRITE(DSPSTRIDE(plane), fb->pitches[0]);
-	I915_MODIFY_DISPBASE(DSPSURF(plane),
-			     i915_gem_obj_ggtt_offset(obj) + intel_crtc->dspaddr_offset);
+	I915_WRITE(DSPSURF(plane),
+		   i915_gem_obj_ggtt_offset(obj) + intel_crtc->dspaddr_offset);
 	if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
 		I915_WRITE(DSPOFFSET(plane), (y << 16) | x);
 	} else {
@@ -2233,7 +2264,12 @@ void intel_display_handle_reset(struct drm_device *dev)
 		struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 
 		mutex_lock(&crtc->mutex);
-		if (intel_crtc->active)
+		/*
+		 * FIXME: Once we have proper support for primary planes (and
+		 * disabling them without disabling the entire crtc) allow again
+		 * a NULL crtc->fb.
+		 */
+		if (intel_crtc->active && crtc->fb)
 			dev_priv->display.update_plane(crtc, crtc->fb,
 						       crtc->x, crtc->y);
 		mutex_unlock(&crtc->mutex);
@@ -2350,6 +2386,8 @@ intel_pipe_set_base(struct drm_crtc *crtc, int x, int y,
 			I915_WRITE(PF_WIN_POS(intel_crtc->pipe), 0);
 			I915_WRITE(PF_WIN_SZ(intel_crtc->pipe), 0);
 		}
+		intel_crtc->config.pipe_src_w = adjusted_mode->crtc_hdisplay;
+		intel_crtc->config.pipe_src_h = adjusted_mode->crtc_vdisplay;
 	}
 
 	ret = dev_priv->display.update_plane(crtc, fb, x, y);
@@ -2944,6 +2982,30 @@ static bool intel_crtc_has_pending_flip(struct drm_crtc *crtc)
 	return pending;
 }
 
+bool intel_has_pending_fb_unpin(struct drm_device *dev)
+{
+	struct intel_crtc *crtc;
+
+	/* Note that we don't need to be called with mode_config.lock here
+	 * as our list of CRTC objects is static for the lifetime of the
+	 * device and so cannot disappear as we iterate. Similarly, we can
+	 * happily treat the predicates as racy, atomic checks as userspace
+	 * cannot claim and pin a new fb without at least acquring the
+	 * struct_mutex and so serialising with us.
+	 */
+	list_for_each_entry(crtc, &dev->mode_config.crtc_list, base.head) {
+		if (atomic_read(&crtc->unpin_work_count) == 0)
+			continue;
+
+		if (crtc->unpin_work)
+			intel_wait_for_vblank(dev, crtc->pipe);
+
+		return true;
+	}
+
+	return false;
+}
+
 static void intel_crtc_wait_for_pending_flips(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
@@ -3399,9 +3461,8 @@ void hsw_enable_ips(struct intel_crtc *crtc)
 		mutex_unlock(&dev_priv->rps.hw_lock);
 		/* Quoting Art Runyan: "its not safe to expect any particular
 		 * value in IPS_CTL bit 31 after enabling IPS through the
-		 * mailbox." Therefore we need to defer waiting on the state
-		 * change.
-		 * TODO: need to fix this for state checker
+		 * mailbox." Moreover, the mailbox may return a bogus state,
+		 * so we need to just enable it and continue on.
 		 */
 	} else {
 		I915_WRITE(IPS_CTL, IPS_ENABLE);
@@ -3428,9 +3489,10 @@ void hsw_disable_ips(struct intel_crtc *crtc)
 		mutex_lock(&dev_priv->rps.hw_lock);
 		WARN_ON(sandybridge_pcode_write(dev_priv, DISPLAY_IPS_CONTROL, 0));
 		mutex_unlock(&dev_priv->rps.hw_lock);
-	} else
+	} else {
 		I915_WRITE(IPS_CTL, 0);
-	POSTING_READ(IPS_CTL);
+		POSTING_READ(IPS_CTL);
+	}
 
 	/* We need to wait for a vblank before we can disable the plane. */
 	intel_wait_for_vblank(dev, crtc->pipe);
@@ -3465,7 +3527,7 @@ static void intel_crtc_load_lut(struct drm_crtc *crtc)
 	/* Workaround : Do not read or write the pipe palette/gamma data while
 	 * GAMMA_MODE is configured for split gamma and IPS_CTL has IPS enabled.
 	 */
-	if (intel_crtc->config.ips_enabled &&
+	if (IS_HASWELL(dev) && intel_crtc->config.ips_enabled &&
 	    ((I915_READ(GAMMA_MODE(pipe)) & GAMMA_MODE_MODE_MASK) ==
 	     GAMMA_MODE_MODE_SPLIT)) {
 		hsw_disable_ips(intel_crtc);
@@ -3910,6 +3972,174 @@ static void i9xx_pfit_enable(struct intel_crtc *crtc)
 	I915_WRITE(BCLRPAT(crtc->pipe), 0);
 }
 
+int valleyview_get_vco(struct drm_i915_private *dev_priv)
+{
+	int hpll_freq, vco_freq[] = { 800, 1600, 2000, 2400 };
+
+	/* Obtain SKU information */
+	mutex_lock(&dev_priv->dpio_lock);
+	hpll_freq = vlv_cck_read(dev_priv, CCK_FUSE_REG) &
+		CCK_FUSE_HPLL_FREQ_MASK;
+	mutex_unlock(&dev_priv->dpio_lock);
+
+	return vco_freq[hpll_freq];
+}
+
+/* Adjust CDclk dividers to allow high res or save power if possible */
+static void valleyview_set_cdclk(struct drm_device *dev, int cdclk)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 val, cmd;
+
+	if (cdclk >= 320) /* jump to highest voltage for 400MHz too */
+		cmd = 2;
+	else if (cdclk == 266)
+		cmd = 1;
+	else
+		cmd = 0;
+
+	mutex_lock(&dev_priv->rps.hw_lock);
+	val = vlv_punit_read(dev_priv, PUNIT_REG_DSPFREQ);
+	val &= ~DSPFREQGUAR_MASK;
+	val |= (cmd << DSPFREQGUAR_SHIFT);
+	vlv_punit_write(dev_priv, PUNIT_REG_DSPFREQ, val);
+	if (wait_for((vlv_punit_read(dev_priv, PUNIT_REG_DSPFREQ) &
+		      DSPFREQSTAT_MASK) == (cmd << DSPFREQSTAT_SHIFT),
+		     50)) {
+		DRM_ERROR("timed out waiting for CDclk change\n");
+	}
+	mutex_unlock(&dev_priv->rps.hw_lock);
+
+	if (cdclk == 400) {
+		u32 divider, vco;
+
+		vco = valleyview_get_vco(dev_priv);
+		divider = ((vco << 1) / cdclk) - 1;
+
+		mutex_lock(&dev_priv->dpio_lock);
+		/* adjust cdclk divider */
+		val = vlv_cck_read(dev_priv, CCK_DISPLAY_CLOCK_CONTROL);
+		val &= ~0xf;
+		val |= divider;
+		vlv_cck_write(dev_priv, CCK_DISPLAY_CLOCK_CONTROL, val);
+		mutex_unlock(&dev_priv->dpio_lock);
+	}
+
+	mutex_lock(&dev_priv->dpio_lock);
+	/* adjust self-refresh exit latency value */
+	val = vlv_bunit_read(dev_priv, BUNIT_REG_BISOC);
+	val &= ~0x7f;
+
+	/*
+	 * For high bandwidth configs, we set a higher latency in the bunit
+	 * so that the core display fetch happens in time to avoid underruns.
+	 */
+	if (cdclk == 400)
+		val |= 4500 / 250; /* 4.5 usec */
+	else
+		val |= 3000 / 250; /* 3.0 usec */
+	vlv_bunit_write(dev_priv, BUNIT_REG_BISOC, val);
+	mutex_unlock(&dev_priv->dpio_lock);
+
+	/* Since we changed the CDclk, we need to update the GMBUSFREQ too */
+	intel_i2c_reset(dev);
+}
+
+static int valleyview_cur_cdclk(struct drm_i915_private *dev_priv)
+{
+	int cur_cdclk, vco;
+	int divider;
+
+	vco = valleyview_get_vco(dev_priv);
+
+	mutex_lock(&dev_priv->dpio_lock);
+	divider = vlv_cck_read(dev_priv, CCK_DISPLAY_CLOCK_CONTROL);
+	mutex_unlock(&dev_priv->dpio_lock);
+
+	divider &= 0xf;
+
+	cur_cdclk = (vco << 1) / (divider + 1);
+
+	return cur_cdclk;
+}
+
+static int valleyview_calc_cdclk(struct drm_i915_private *dev_priv,
+				 int max_pixclk)
+{
+	int cur_cdclk;
+
+	cur_cdclk = valleyview_cur_cdclk(dev_priv);
+
+	/*
+	 * Really only a few cases to deal with, as only 4 CDclks are supported:
+	 *   200MHz
+	 *   267MHz
+	 *   320MHz
+	 *   400MHz
+	 * So we check to see whether we're above 90% of the lower bin and
+	 * adjust if needed.
+	 */
+	if (max_pixclk > 288000) {
+		return 400;
+	} else if (max_pixclk > 240000) {
+		return 320;
+	} else
+		return 266;
+	/* Looks like the 200MHz CDclk freq doesn't work on some configs */
+}
+
+static int intel_mode_max_pixclk(struct drm_i915_private *dev_priv,
+				 unsigned modeset_pipes,
+				 struct intel_crtc_config *pipe_config)
+{
+	struct drm_device *dev = dev_priv->dev;
+	struct intel_crtc *intel_crtc;
+	int max_pixclk = 0;
+
+	list_for_each_entry(intel_crtc, &dev->mode_config.crtc_list,
+			    base.head) {
+		if (modeset_pipes & (1 << intel_crtc->pipe))
+			max_pixclk = max(max_pixclk,
+					 pipe_config->adjusted_mode.crtc_clock);
+		else if (intel_crtc->base.enabled)
+			max_pixclk = max(max_pixclk,
+					 intel_crtc->config.adjusted_mode.crtc_clock);
+	}
+
+	return max_pixclk;
+}
+
+static void valleyview_modeset_global_pipes(struct drm_device *dev,
+					    unsigned *prepare_pipes,
+					    unsigned modeset_pipes,
+					    struct intel_crtc_config *pipe_config)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_crtc *intel_crtc;
+	int max_pixclk = intel_mode_max_pixclk(dev_priv, modeset_pipes,
+					       pipe_config);
+	int cur_cdclk = valleyview_cur_cdclk(dev_priv);
+
+	if (valleyview_calc_cdclk(dev_priv, max_pixclk) == cur_cdclk)
+		return;
+
+	list_for_each_entry(intel_crtc, &dev->mode_config.crtc_list,
+			    base.head)
+		if (intel_crtc->base.enabled)
+			*prepare_pipes |= (1 << intel_crtc->pipe);
+}
+
+static void valleyview_modeset_global_resources(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int max_pixclk = intel_mode_max_pixclk(dev_priv, 0, NULL);
+	int cur_cdclk = valleyview_cur_cdclk(dev_priv);
+	int req_cdclk = valleyview_calc_cdclk(dev_priv, max_pixclk);
+
+	if (req_cdclk != cur_cdclk)
+		valleyview_set_cdclk(dev, req_cdclk);
+}
+
 static void valleyview_crtc_enable(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
@@ -4570,9 +4800,8 @@ static int i9xx_get_refclk(struct drm_crtc *crtc, int num_connectors)
 		refclk = 100000;
 	} else if (intel_pipe_has_type(crtc, INTEL_OUTPUT_LVDS) &&
 	    intel_panel_use_ssc(dev_priv) && num_connectors < 2) {
-		refclk = dev_priv->vbt.lvds_ssc_freq * 1000;
-		DRM_DEBUG_KMS("using SSC reference clock of %d MHz\n",
-			      refclk / 1000);
+		refclk = dev_priv->vbt.lvds_ssc_freq;
+		DRM_DEBUG_KMS("using SSC reference clock of %d kHz\n", refclk);
 	} else if (!IS_GEN2(dev)) {
 		refclk = 96000;
 	} else {
@@ -4634,24 +4863,24 @@ static void vlv_pllb_recal_opamp(struct drm_i915_private *dev_priv, enum pipe
 	 * PLLB opamp always calibrates to max value of 0x3f, force enable it
 	 * and set it to a reasonable value instead.
 	 */
-	reg_val = vlv_dpio_read(dev_priv, pipe, DPIO_IREF(1));
+	reg_val = vlv_dpio_read(dev_priv, pipe, VLV_PLL_DW9(1));
 	reg_val &= 0xffffff00;
 	reg_val |= 0x00000030;
-	vlv_dpio_write(dev_priv, pipe, DPIO_IREF(1), reg_val);
+	vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW9(1), reg_val);
 
-	reg_val = vlv_dpio_read(dev_priv, pipe, DPIO_CALIBRATION);
+	reg_val = vlv_dpio_read(dev_priv, pipe, VLV_REF_DW13);
 	reg_val &= 0x8cffffff;
 	reg_val = 0x8c000000;
-	vlv_dpio_write(dev_priv, pipe, DPIO_CALIBRATION, reg_val);
+	vlv_dpio_write(dev_priv, pipe, VLV_REF_DW13, reg_val);
 
-	reg_val = vlv_dpio_read(dev_priv, pipe, DPIO_IREF(1));
+	reg_val = vlv_dpio_read(dev_priv, pipe, VLV_PLL_DW9(1));
 	reg_val &= 0xffffff00;
-	vlv_dpio_write(dev_priv, pipe, DPIO_IREF(1), reg_val);
+	vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW9(1), reg_val);
 
-	reg_val = vlv_dpio_read(dev_priv, pipe, DPIO_CALIBRATION);
+	reg_val = vlv_dpio_read(dev_priv, pipe, VLV_REF_DW13);
 	reg_val &= 0x00ffffff;
 	reg_val |= 0xb0000000;
-	vlv_dpio_write(dev_priv, pipe, DPIO_CALIBRATION, reg_val);
+	vlv_dpio_write(dev_priv, pipe, VLV_REF_DW13, reg_val);
 }
 
 static void intel_pch_transcoder_set_m_n(struct intel_crtc *crtc,
@@ -4720,15 +4949,15 @@ static void vlv_update_pll(struct intel_crtc *crtc)
 		vlv_pllb_recal_opamp(dev_priv, pipe);
 
 	/* Set up Tx target for periodic Rcomp update */
-	vlv_dpio_write(dev_priv, pipe, DPIO_IREF_BCAST, 0x0100000f);
+	vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW9_BCAST, 0x0100000f);
 
 	/* Disable target IRef on PLL */
-	reg_val = vlv_dpio_read(dev_priv, pipe, DPIO_IREF_CTL(pipe));
+	reg_val = vlv_dpio_read(dev_priv, pipe, VLV_PLL_DW8(pipe));
 	reg_val &= 0x00ffffff;
-	vlv_dpio_write(dev_priv, pipe, DPIO_IREF_CTL(pipe), reg_val);
+	vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW8(pipe), reg_val);
 
 	/* Disable fast lock */
-	vlv_dpio_write(dev_priv, pipe, DPIO_FASTCLK_DISABLE, 0x610);
+	vlv_dpio_write(dev_priv, pipe, VLV_CMN_DW0, 0x610);
 
 	/* Set idtafcrecal before PLL is enabled */
 	mdiv = ((bestm1 << DPIO_M1DIV_SHIFT) | (bestm2 & DPIO_M2DIV_MASK));
@@ -4742,50 +4971,54 @@ static void vlv_update_pll(struct intel_crtc *crtc)
 	 * Note: don't use the DAC post divider as it seems unstable.
 	 */
 	mdiv |= (DPIO_POST_DIV_HDMIDP << DPIO_POST_DIV_SHIFT);
-	vlv_dpio_write(dev_priv, pipe, DPIO_DIV(pipe), mdiv);
+	vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW3(pipe), mdiv);
 
 	mdiv |= DPIO_ENABLE_CALIBRATION;
-	vlv_dpio_write(dev_priv, pipe, DPIO_DIV(pipe), mdiv);
+	vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW3(pipe), mdiv);
 
 	/* Set HBR and RBR LPF coefficients */
 	if (crtc->config.port_clock == 162000 ||
 	    intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_ANALOG) ||
 	    intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_HDMI))
-		vlv_dpio_write(dev_priv, pipe, DPIO_LPF_COEFF(pipe),
+		vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW10(pipe),
 				 0x009f0003);
 	else
-		vlv_dpio_write(dev_priv, pipe, DPIO_LPF_COEFF(pipe),
+		vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW10(pipe),
 				 0x00d0000f);
 
 	if (intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_EDP) ||
 	    intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_DISPLAYPORT)) {
 		/* Use SSC source */
 		if (!pipe)
-			vlv_dpio_write(dev_priv, pipe, DPIO_REFSFR(pipe),
+			vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW5(pipe),
 					 0x0df40000);
 		else
-			vlv_dpio_write(dev_priv, pipe, DPIO_REFSFR(pipe),
+			vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW5(pipe),
 					 0x0df70000);
 	} else { /* HDMI or VGA */
 		/* Use bend source */
 		if (!pipe)
-			vlv_dpio_write(dev_priv, pipe, DPIO_REFSFR(pipe),
+			vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW5(pipe),
 					 0x0df70000);
 		else
-			vlv_dpio_write(dev_priv, pipe, DPIO_REFSFR(pipe),
+			vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW5(pipe),
 					 0x0df40000);
 	}
 
-	coreclk = vlv_dpio_read(dev_priv, pipe, DPIO_CORE_CLK(pipe));
+	coreclk = vlv_dpio_read(dev_priv, pipe, VLV_PLL_DW7(pipe));
 	coreclk = (coreclk & 0x0000ff00) | 0x01c00000;
 	if (intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_DISPLAYPORT) ||
 	    intel_pipe_has_type(&crtc->base, INTEL_OUTPUT_EDP))
 		coreclk |= 0x01000000;
-	vlv_dpio_write(dev_priv, pipe, DPIO_CORE_CLK(pipe), coreclk);
+	vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW7(pipe), coreclk);
 
-	vlv_dpio_write(dev_priv, pipe, DPIO_PLL_CML(pipe), 0x87871000);
+	vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW11(pipe), 0x87871000);
 
-	/* Enable DPIO clock input */
+	/*
+	 * Enable DPIO clock input. We should never disable the reference
+	 * clock for pipe B, since VGA hotplug / manual detection depends
+	 * on it.
+	 */
 	dpll = DPLL_EXT_BUFFER_ENABLE_VLV | DPLL_REFA_CLK_ENABLE_VLV |
 		DPLL_VGA_MODE_DIS | DPLL_INTEGRATED_CLOCK_VLV;
 	/* We should never disable this, set it here for state tracking */
@@ -5230,6 +5463,9 @@ static void i9xx_get_pfit_config(struct intel_crtc *crtc,
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	uint32_t tmp;
 
+	if (INTEL_INFO(dev)->gen <= 3 && (IS_I830(dev) || !IS_MOBILE(dev)))
+		return;
+
 	tmp = I915_READ(PFIT_CONTROL);
 	if (!(tmp & PFIT_ENABLE))
 		return;
@@ -5261,7 +5497,7 @@ static void vlv_crtc_clock_get(struct intel_crtc *crtc,
 	int refclk = 100000;
 
 	mutex_lock(&dev_priv->dpio_lock);
-	mdiv = vlv_dpio_read(dev_priv, pipe, DPIO_DIV(pipe));
+	mdiv = vlv_dpio_read(dev_priv, pipe, VLV_PLL_DW3(pipe));
 	mutex_unlock(&dev_priv->dpio_lock);
 
 	clock.m1 = (mdiv >> DPIO_M1DIV_SHIFT) & 7;
@@ -5718,9 +5954,9 @@ static int ironlake_get_refclk(struct drm_crtc *crtc)
 	}
 
 	if (is_lvds && intel_panel_use_ssc(dev_priv) && num_connectors < 2) {
-		DRM_DEBUG_KMS("using SSC reference clock of %d MHz\n",
+		DRM_DEBUG_KMS("using SSC reference clock of %d kHz\n",
 			      dev_priv->vbt.lvds_ssc_freq);
-		return dev_priv->vbt.lvds_ssc_freq * 1000;
+		return dev_priv->vbt.lvds_ssc_freq;
 	}
 
 	return 120000;
@@ -5982,7 +6218,7 @@ static uint32_t ironlake_compute_dpll(struct intel_crtc *intel_crtc,
 	factor = 21;
 	if (is_lvds) {
 		if ((intel_panel_use_ssc(dev_priv) &&
-		     dev_priv->vbt.lvds_ssc_freq == 100) ||
+		     dev_priv->vbt.lvds_ssc_freq == 100000) ||
 		    (HAS_PCH_IBX(dev) && intel_is_dual_link_lvds(dev)))
 			factor = 25;
 	} else if (intel_crtc->config.sdvo_tv_clock)
@@ -6323,7 +6559,7 @@ static void assert_can_disable_lcpll(struct drm_i915_private *dev_priv)
 
 	spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
 	val = I915_READ(DEIMR);
-	WARN((val & ~DE_PCH_EVENT_IVB) != val,
+	WARN((val | DE_PCH_EVENT_IVB) != 0xffffffff,
 	     "Unexpected DEIMR bits enabled: 0x%x\n", val);
 	val = I915_READ(SDEIMR);
 	WARN((val | SDE_HOTPLUG_MASK_CPT) != 0xffffffff,
@@ -6402,7 +6638,7 @@ static void hsw_restore_lcpll(struct drm_i915_private *dev_priv)
 
 	/* Make sure we're not on PC8 state before disabling PC8, otherwise
 	 * we'll hang the machine! */
-	gen6_gt_force_wake_get(dev_priv);
+	gen6_gt_force_wake_get(dev_priv, FORCEWAKE_ALL);
 
 	if (val & LCPLL_POWER_DOWN_ALLOW) {
 		val &= ~LCPLL_POWER_DOWN_ALLOW;
@@ -6436,7 +6672,7 @@ static void hsw_restore_lcpll(struct drm_i915_private *dev_priv)
 			DRM_ERROR("Switching back to LCPLL failed\n");
 	}
 
-	gen6_gt_force_wake_put(dev_priv);
+	gen6_gt_force_wake_put(dev_priv, FORCEWAKE_ALL);
 }
 
 void hsw_enable_pc8_work(struct work_struct *__work)
@@ -6447,6 +6683,8 @@ void hsw_enable_pc8_work(struct work_struct *__work)
 	struct drm_device *dev = dev_priv->dev;
 	uint32_t val;
 
+	WARN_ON(!HAS_PC8(dev));
+
 	if (dev_priv->pc8.enabled)
 		return;
 
@@ -6463,6 +6701,8 @@ void hsw_enable_pc8_work(struct work_struct *__work)
 	lpt_disable_clkout_dp(dev);
 	hsw_pc8_disable_interrupts(dev);
 	hsw_disable_lcpll(dev_priv, true, true);
+
+	intel_runtime_pm_put(dev_priv);
 }
 
 static void __hsw_enable_package_c8(struct drm_i915_private *dev_priv)
@@ -6492,12 +6732,16 @@ static void __hsw_disable_package_c8(struct drm_i915_private *dev_priv)
 	if (dev_priv->pc8.disable_count != 1)
 		return;
 
+	WARN_ON(!HAS_PC8(dev));
+
 	cancel_delayed_work_sync(&dev_priv->pc8.enable_work);
 	if (!dev_priv->pc8.enabled)
 		return;
 
 	DRM_DEBUG_KMS("Disabling package C8+\n");
 
+	intel_runtime_pm_get(dev_priv);
+
 	hsw_restore_lcpll(dev_priv);
 	hsw_pc8_restore_interrupts(dev);
 	lpt_init_pch_refclk(dev);
@@ -6704,8 +6948,9 @@ static int haswell_crtc_mode_set(struct drm_crtc *crtc,
 	int plane = intel_crtc->plane;
 	int ret;
 
-	if (!intel_ddi_pll_mode_set(crtc))
+	if (!intel_ddi_pll_select(intel_crtc))
 		return -EINVAL;
+	intel_ddi_pll_enable(intel_crtc);
 
 	if (intel_crtc->config.has_dp_encoder)
 		intel_dp_set_m_n(intel_crtc);
@@ -6796,8 +7041,9 @@ static bool haswell_get_pipe_config(struct intel_crtc *crtc,
 	if (intel_display_power_enabled(dev, pfit_domain))
 		ironlake_get_pfit_config(crtc, pipe_config);
 
-	pipe_config->ips_enabled = hsw_crtc_supports_ips(crtc) &&
-				   (I915_READ(IPS_CTL) & IPS_ENABLE);
+	if (IS_HASWELL(dev))
+		pipe_config->ips_enabled = hsw_crtc_supports_ips(crtc) &&
+			(I915_READ(IPS_CTL) & IPS_ENABLE);
 
 	pipe_config->pixel_multiplier = 1;
 
@@ -7689,7 +7935,7 @@ static int i9xx_pll_refclk(struct drm_device *dev,
 	u32 dpll = pipe_config->dpll_hw_state.dpll;
 
 	if ((dpll & PLL_REF_INPUT_MASK) == PLLB_REF_INPUT_SPREADSPECTRUMIN)
-		return dev_priv->vbt.lvds_ssc_freq * 1000;
+		return dev_priv->vbt.lvds_ssc_freq;
 	else if (HAS_PCH_SPLIT(dev))
 		return 120000;
 	else if (!IS_GEN2(dev))
@@ -7752,12 +7998,17 @@ static void i9xx_crtc_clock_get(struct intel_crtc *crtc,
 		else
 			i9xx_clock(refclk, &clock);
 	} else {
-		bool is_lvds = (pipe == 1) && (I915_READ(LVDS) & LVDS_PORT_EN);
+		u32 lvds = IS_I830(dev) ? 0 : I915_READ(LVDS);
+		bool is_lvds = (pipe == 1) && (lvds & LVDS_PORT_EN);
 
 		if (is_lvds) {
 			clock.p1 = ffs((dpll & DPLL_FPA01_P1_POST_DIV_MASK_I830_LVDS) >>
 				       DPLL_FPA01_P1_POST_DIV_SHIFT);
-			clock.p2 = 14;
+
+			if (lvds & LVDS_CLKB_POWER_UP)
+				clock.p2 = 7;
+			else
+				clock.p2 = 14;
 		} else {
 			if (dpll & PLL_P1_DIVIDE_BY_TWO)
 				clock.p1 = 2;
@@ -8493,28 +8744,6 @@ static struct drm_crtc_helper_funcs intel_helper_funcs = {
 	.load_lut = intel_crtc_load_lut,
 };
 
-static bool intel_encoder_crtc_ok(struct drm_encoder *encoder,
-				  struct drm_crtc *crtc)
-{
-	struct drm_device *dev;
-	struct drm_crtc *tmp;
-	int crtc_mask = 1;
-
-	WARN(!crtc, "checking null crtc?\n");
-
-	dev = crtc->dev;
-
-	list_for_each_entry(tmp, &dev->mode_config.crtc_list, head) {
-		if (tmp == crtc)
-			break;
-		crtc_mask <<= 1;
-	}
-
-	if (encoder->possible_crtcs & crtc_mask)
-		return true;
-	return false;
-}
-
 /**
  * intel_modeset_update_staged_output_state
  *
@@ -9122,7 +9351,9 @@ intel_pipe_config_compare(struct drm_device *dev,
 		PIPE_CONF_CHECK_I(pch_pfit.size);
 	}
 
-	PIPE_CONF_CHECK_I(ips_enabled);
+	/* BDW+ don't expose a synchronous way to read the state */
+	if (IS_HASWELL(dev))
+		PIPE_CONF_CHECK_I(ips_enabled);
 
 	PIPE_CONF_CHECK_I(double_wide);
 
@@ -9368,21 +9599,19 @@ static int __intel_set_mode(struct drm_crtc *crtc,
 {
 	struct drm_device *dev = crtc->dev;
 	drm_i915_private_t *dev_priv = dev->dev_private;
-	struct drm_display_mode *saved_mode, *saved_hwmode;
+	struct drm_display_mode *saved_mode;
 	struct intel_crtc_config *pipe_config = NULL;
 	struct intel_crtc *intel_crtc;
 	unsigned disable_pipes, prepare_pipes, modeset_pipes;
 	int ret = 0;
 
-	saved_mode = kcalloc(2, sizeof(*saved_mode), GFP_KERNEL);
+	saved_mode = kmalloc(sizeof(*saved_mode), GFP_KERNEL);
 	if (!saved_mode)
 		return -ENOMEM;
-	saved_hwmode = saved_mode + 1;
 
 	intel_modeset_affected_pipes(crtc, &modeset_pipes,
 				     &prepare_pipes, &disable_pipes);
 
-	*saved_hwmode = crtc->hwmode;
 	*saved_mode = crtc->mode;
 
 	/* Hack: Because we don't (yet) support global modeset on multiple
@@ -9402,6 +9631,21 @@ static int __intel_set_mode(struct drm_crtc *crtc,
 				       "[modeset]");
 	}
 
+	/*
+	 * See if the config requires any additional preparation, e.g.
+	 * to adjust global state with pipes off.  We need to do this
+	 * here so we can get the modeset_pipe updated config for the new
+	 * mode set on this crtc.  For other crtcs we need to use the
+	 * adjusted_mode bits in the crtc directly.
+	 */
+	if (IS_VALLEYVIEW(dev)) {
+		valleyview_modeset_global_pipes(dev, &prepare_pipes,
+						modeset_pipes, pipe_config);
+
+		/* may have added more to prepare_pipes than we should */
+		prepare_pipes &= ~disable_pipes;
+	}
+
 	for_each_intel_crtc_masked(dev, disable_pipes, intel_crtc)
 		intel_crtc_disable(&intel_crtc->base);
 
@@ -9418,6 +9662,14 @@ static int __intel_set_mode(struct drm_crtc *crtc,
 		/* mode_set/enable/disable functions rely on a correct pipe
 		 * config. */
 		to_intel_crtc(crtc)->config = *pipe_config;
+
+		/*
+		 * Calculate and store various constants which
+		 * are later needed by vblank and swap-completion
+		 * timestamping. They are derived from true hwmode.
+		 */
+		drm_calc_timestamping_constants(crtc,
+						&pipe_config->adjusted_mode);
 	}
 
 	/* Only after disabling all output pipelines that will be changed can we
@@ -9441,23 +9693,10 @@ static int __intel_set_mode(struct drm_crtc *crtc,
 	for_each_intel_crtc_masked(dev, prepare_pipes, intel_crtc)
 		dev_priv->display.crtc_enable(&intel_crtc->base);
 
-	if (modeset_pipes) {
-		/* Store real post-adjustment hardware mode. */
-		crtc->hwmode = pipe_config->adjusted_mode;
-
-		/* Calculate and store various constants which
-		 * are later needed by vblank and swap-completion
-		 * timestamping. They are derived from true hwmode.
-		 */
-		drm_calc_timestamping_constants(crtc);
-	}
-
 	/* FIXME: add subpixel order */
 done:
-	if (ret && crtc->enabled) {
-		crtc->hwmode = *saved_hwmode;
+	if (ret && crtc->enabled)
 		crtc->mode = *saved_mode;
-	}
 
 out:
 	kfree(pipe_config);
@@ -9679,8 +9918,8 @@ intel_modeset_stage_output_state(struct drm_device *dev,
 		}
 
 		/* Make sure the new CRTC will work with the encoder */
-		if (!intel_encoder_crtc_ok(&connector->new_encoder->base,
-					   new_crtc)) {
+		if (!drm_encoder_crtc_ok(&connector->new_encoder->base,
+					 new_crtc)) {
 			return -EINVAL;
 		}
 		connector->encoder->new_crtc = to_intel_crtc(new_crtc);
@@ -9694,17 +9933,21 @@ 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) {
+		int num_connectors = 0;
 		list_for_each_entry(connector,
 				    &dev->mode_config.connector_list,
 				    base.head) {
 			if (connector->new_encoder == encoder) {
 				WARN_ON(!connector->new_encoder->new_crtc);
-
-				goto next_encoder;
+				num_connectors++;
 			}
 		}
-		encoder->new_crtc = NULL;
-next_encoder:
+
+		if (num_connectors == 0)
+			encoder->new_crtc = NULL;
+		else if (num_connectors > 1)
+			return -EINVAL;
+
 		/* Only now check for crtc changes so we don't miss encoders
 		 * that will be disabled. */
 		if (&encoder->new_crtc->base != encoder->base.crtc) {
@@ -9775,6 +10018,16 @@ static int intel_crtc_set_config(struct drm_mode_set *set)
 
 		ret = intel_pipe_set_base(set->crtc,
 					  set->x, set->y, set->fb);
+		/*
+		 * In the fastboot case this may be our only check of the
+		 * state after boot.  It would be better to only do it on
+		 * the first update, but we don't have a nice way of doing that
+		 * (and really, set_config isn't used much for high freq page
+		 * flipping, so increasing its cost here shouldn't be a big
+		 * deal).
+		 */
+		if (i915_fastboot && ret == 0)
+			intel_modeset_check_state(set->crtc->dev);
 	}
 
 	if (ret) {
@@ -9835,7 +10088,7 @@ static void ibx_pch_dpll_enable(struct drm_i915_private *dev_priv,
 				struct intel_shared_dpll *pll)
 {
 	/* PCH refclock must be enabled first */
-	assert_pch_refclk_enabled(dev_priv);
+	ibx_assert_pch_refclk_enabled(dev_priv);
 
 	I915_WRITE(PCH_DPLL(pll->id), pll->hw_state.dpll);
 
@@ -9903,8 +10156,6 @@ static void intel_shared_dpll_init(struct drm_device *dev)
 		dev_priv->num_shared_dpll = 0;
 
 	BUG_ON(dev_priv->num_shared_dpll > I915_NUM_PLLS);
-	DRM_DEBUG_KMS("%i shared PLLs initialized\n",
-		      dev_priv->num_shared_dpll);
 }
 
 static void intel_crtc_init(struct drm_device *dev, int pipe)
@@ -9926,10 +10177,13 @@ static void intel_crtc_init(struct drm_device *dev, int pipe)
 		intel_crtc->lut_b[i] = i;
 	}
 
-	/* Swap pipes & planes for FBC on pre-965 */
+	/*
+	 * On gen2/3 only plane A can do fbc, but the panel fitter and lvds port
+	 * is hooked to plane B. Hence we want plane A feeding pipe B.
+	 */
 	intel_crtc->pipe = pipe;
 	intel_crtc->plane = pipe;
-	if (IS_MOBILE(dev) && IS_GEN3(dev)) {
+	if (HAS_FBC(dev) && INTEL_INFO(dev)->gen < 4) {
 		DRM_DEBUG_KMS("swapping pipes & planes for FBC\n");
 		intel_crtc->plane = !pipe;
 	}
@@ -10018,6 +10272,28 @@ static bool has_edp_a(struct drm_device *dev)
 	return true;
 }
 
+const char *intel_output_name(int output)
+{
+	static const char *names[] = {
+		[INTEL_OUTPUT_UNUSED] = "Unused",
+		[INTEL_OUTPUT_ANALOG] = "Analog",
+		[INTEL_OUTPUT_DVO] = "DVO",
+		[INTEL_OUTPUT_SDVO] = "SDVO",
+		[INTEL_OUTPUT_LVDS] = "LVDS",
+		[INTEL_OUTPUT_TVOUT] = "TV",
+		[INTEL_OUTPUT_HDMI] = "HDMI",
+		[INTEL_OUTPUT_DISPLAYPORT] = "DisplayPort",
+		[INTEL_OUTPUT_EDP] = "eDP",
+		[INTEL_OUTPUT_DSI] = "DSI",
+		[INTEL_OUTPUT_UNKNOWN] = "Unknown",
+	};
+
+	if (output < 0 || output >= ARRAY_SIZE(names) || !names[output])
+		return "Invalid";
+
+	return names[output];
+}
+
 static void intel_setup_outputs(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -10412,8 +10688,11 @@ static void intel_init_display(struct drm_device *dev)
 		}
 	} else if (IS_G4X(dev)) {
 		dev_priv->display.write_eld = g4x_write_eld;
-	} else if (IS_VALLEYVIEW(dev))
+	} else if (IS_VALLEYVIEW(dev)) {
+		dev_priv->display.modeset_global_resources =
+			valleyview_modeset_global_resources;
 		dev_priv->display.write_eld = ironlake_write_eld;
+	}
 
 	/* Default just returns -ENODEV to indicate unsupported */
 	dev_priv->display.queue_flip = intel_default_queue_flip;
@@ -10440,6 +10719,8 @@ static void intel_init_display(struct drm_device *dev)
 		dev_priv->display.queue_flip = intel_gen7_queue_flip;
 		break;
 	}
+
+	intel_panel_init_backlight_funcs(dev);
 }
 
 /*
@@ -10476,17 +10757,6 @@ static void quirk_invert_brightness(struct drm_device *dev)
 	DRM_INFO("applying inverted panel brightness quirk\n");
 }
 
-/*
- * Some machines (Dell XPS13) suffer broken backlight controls if
- * BLM_PCH_PWM_ENABLE is set.
- */
-static void quirk_no_pcm_pwm_enable(struct drm_device *dev)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	dev_priv->quirks |= QUIRK_NO_PCH_PWM_ENABLE;
-	DRM_INFO("applying no-PCH_PWM_ENABLE quirk\n");
-}
-
 struct intel_quirk {
 	int device;
 	int subsystem_vendor;
@@ -10555,11 +10825,6 @@ static struct intel_quirk intel_quirks[] = {
 
 	/* Acer Aspire 4736Z */
 	{ 0x2a42, 0x1025, 0x0260, quirk_invert_brightness },
-
-	/* Dell XPS13 HD Sandy Bridge */
-	{ 0x0116, 0x1028, 0x052e, quirk_no_pcm_pwm_enable },
-	/* Dell XPS13 HD and XPS13 FHD Ivy Bridge */
-	{ 0x0166, 0x1028, 0x058b, quirk_no_pcm_pwm_enable },
 };
 
 static void intel_init_quirks(struct drm_device *dev)
@@ -10603,18 +10868,11 @@ static void i915_disable_vga(struct drm_device *dev)
 
 void intel_modeset_init_hw(struct drm_device *dev)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
-
 	intel_prepare_ddi(dev);
 
 	intel_init_clock_gating(dev);
 
-	/* Enable the CRI clock source so we can get at the display */
-	if (IS_VALLEYVIEW(dev))
-		I915_WRITE(DPLL(PIPE_B), I915_READ(DPLL(PIPE_B)) |
-			   DPLL_INTEGRATED_CRI_CLK_VLV);
-
-	intel_init_dpio(dev);
+	intel_reset_dpio(dev);
 
 	mutex_lock(&dev->struct_mutex);
 	intel_enable_gt_powersave(dev);
@@ -10676,6 +10934,9 @@ void intel_modeset_init(struct drm_device *dev)
 		}
 	}
 
+	intel_init_dpio(dev);
+	intel_reset_dpio(dev);
+
 	intel_cpu_pll_init(dev);
 	intel_shared_dpll_init(dev);
 
@@ -10879,7 +11140,7 @@ void i915_redisable_vga(struct drm_device *dev)
 	 * level, just check if the power well is enabled instead of trying to
 	 * follow the "don't touch the power well if we don't need it" policy
 	 * the rest of the driver uses. */
-	if (HAS_POWER_WELL(dev) &&
+	if ((IS_HASWELL(dev) || IS_BROADWELL(dev)) &&
 	    (I915_READ(HSW_PWR_WELL_DRIVER) & HSW_PWR_WELL_STATE_ENABLED) == 0)
 		return;
 
@@ -11023,7 +11284,7 @@ void intel_modeset_setup_hw_state(struct drm_device *dev,
 		pll->on = false;
 	}
 
-	if (IS_HASWELL(dev))
+	if (HAS_PCH_SPLIT(dev))
 		ilk_wm_get_hw_state(dev);
 
 	if (force_restore) {
@@ -11101,12 +11362,11 @@ void intel_modeset_cleanup(struct drm_device *dev)
 	/* flush any delayed tasks or pending work */
 	flush_scheduled_work();
 
-	/* destroy backlight, if any, before the connectors */
-	intel_panel_destroy_backlight(dev);
-
-	/* destroy the sysfs files before encoders/connectors */
-	list_for_each_entry(connector, &dev->mode_config.connector_list, head)
+	/* destroy the backlight and sysfs files before encoders/connectors */
+	list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+		intel_panel_destroy_backlight(connector);
 		drm_sysfs_connector_remove(connector);
+	}
 
 	drm_mode_config_cleanup(dev);
 
@@ -11161,6 +11421,7 @@ struct intel_display_error_state {
 	} cursor[I915_MAX_PIPES];
 
 	struct intel_pipe_error_state {
+		bool power_domain_on;
 		u32 source;
 	} pipe[I915_MAX_PIPES];
 
@@ -11175,6 +11436,7 @@ struct intel_display_error_state {
 	} plane[I915_MAX_PIPES];
 
 	struct intel_transcoder_error_state {
+		bool power_domain_on;
 		enum transcoder cpu_transcoder;
 
 		u32 conf;
@@ -11208,11 +11470,13 @@ intel_display_capture_error_state(struct drm_device *dev)
 	if (error == NULL)
 		return NULL;
 
-	if (HAS_POWER_WELL(dev))
+	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
 		error->power_well_driver = I915_READ(HSW_PWR_WELL_DRIVER);
 
 	for_each_pipe(i) {
-		if (!intel_display_power_enabled(dev, POWER_DOMAIN_PIPE(i)))
+		error->pipe[i].power_domain_on =
+			intel_display_power_enabled_sw(dev, POWER_DOMAIN_PIPE(i));
+		if (!error->pipe[i].power_domain_on)
 			continue;
 
 		if (INTEL_INFO(dev)->gen <= 6 || IS_VALLEYVIEW(dev)) {
@@ -11248,8 +11512,10 @@ intel_display_capture_error_state(struct drm_device *dev)
 	for (i = 0; i < error->num_transcoders; i++) {
 		enum transcoder cpu_transcoder = transcoders[i];
 
-		if (!intel_display_power_enabled(dev,
-				POWER_DOMAIN_TRANSCODER(cpu_transcoder)))
+		error->transcoder[i].power_domain_on =
+			intel_display_power_enabled_sw(dev,
+				POWER_DOMAIN_TRANSCODER(cpu_transcoder));
+		if (!error->transcoder[i].power_domain_on)
 			continue;
 
 		error->transcoder[i].cpu_transcoder = cpu_transcoder;
@@ -11279,11 +11545,13 @@ intel_display_print_error_state(struct drm_i915_error_state_buf *m,
 		return;
 
 	err_printf(m, "Num Pipes: %d\n", INTEL_INFO(dev)->num_pipes);
-	if (HAS_POWER_WELL(dev))
+	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
 		err_printf(m, "PWR_WELL_CTL2: %08x\n",
 			   error->power_well_driver);
 	for_each_pipe(i) {
 		err_printf(m, "Pipe [%d]:\n", i);
+		err_printf(m, "  Power: %s\n",
+			   error->pipe[i].power_domain_on ? "on" : "off");
 		err_printf(m, "  SRC: %08x\n", error->pipe[i].source);
 
 		err_printf(m, "Plane [%d]:\n", i);
@@ -11309,6 +11577,8 @@ intel_display_print_error_state(struct drm_i915_error_state_buf *m,
 	for (i = 0; i < error->num_transcoders; i++) {
 		err_printf(m, "CPU transcoder: %c\n",
 			   transcoder_name(error->transcoder[i].cpu_transcoder));
+		err_printf(m, "  Power: %s\n",
+			   error->transcoder[i].power_domain_on ? "on" : "off");
 		err_printf(m, "  CONF: %08x\n", error->transcoder[i].conf);
 		err_printf(m, "  HTOTAL: %08x\n", error->transcoder[i].htotal);
 		err_printf(m, "  HBLANK: %08x\n", error->transcoder[i].hblank);
diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
index 30c627c7b7ba..5ede4e8e290d 100644
--- a/drivers/gpu/drm/i915/intel_dp.c
+++ b/drivers/gpu/drm/i915/intel_dp.c
@@ -142,7 +142,7 @@ intel_dp_max_data_rate(int max_link_clock, int max_lanes)
 	return (max_link_clock * max_lanes * 8) / 10;
 }
 
-static int
+static enum drm_mode_status
 intel_dp_mode_valid(struct drm_connector *connector,
 		    struct drm_display_mode *mode)
 {
@@ -404,7 +404,7 @@ intel_dp_aux_ch(struct intel_dp *intel_dp,
 	int i, ret, recv_bytes;
 	uint32_t status;
 	int try, precharge, clock = 0;
-	bool has_aux_irq = INTEL_INFO(dev)->gen >= 5 && !IS_VALLEYVIEW(dev);
+	bool has_aux_irq = true;
 	uint32_t timeout;
 
 	/* dp aux is extremely sensitive to irq latency, hence request the
@@ -542,7 +542,7 @@ intel_dp_aux_native_write(struct intel_dp *intel_dp,
 		return -E2BIG;
 
 	intel_dp_check_edp(intel_dp);
-	msg[0] = AUX_NATIVE_WRITE << 4;
+	msg[0] = DP_AUX_NATIVE_WRITE << 4;
 	msg[1] = address >> 8;
 	msg[2] = address & 0xff;
 	msg[3] = send_bytes - 1;
@@ -552,9 +552,10 @@ intel_dp_aux_native_write(struct intel_dp *intel_dp,
 		ret = intel_dp_aux_ch(intel_dp, msg, msg_bytes, &ack, 1);
 		if (ret < 0)
 			return ret;
-		if ((ack & AUX_NATIVE_REPLY_MASK) == AUX_NATIVE_REPLY_ACK)
+		ack >>= 4;
+		if ((ack & DP_AUX_NATIVE_REPLY_MASK) == DP_AUX_NATIVE_REPLY_ACK)
 			break;
-		else if ((ack & AUX_NATIVE_REPLY_MASK) == AUX_NATIVE_REPLY_DEFER)
+		else if ((ack & DP_AUX_NATIVE_REPLY_MASK) == DP_AUX_NATIVE_REPLY_DEFER)
 			udelay(100);
 		else
 			return -EIO;
@@ -586,7 +587,7 @@ intel_dp_aux_native_read(struct intel_dp *intel_dp,
 		return -E2BIG;
 
 	intel_dp_check_edp(intel_dp);
-	msg[0] = AUX_NATIVE_READ << 4;
+	msg[0] = DP_AUX_NATIVE_READ << 4;
 	msg[1] = address >> 8;
 	msg[2] = address & 0xff;
 	msg[3] = recv_bytes - 1;
@@ -601,12 +602,12 @@ intel_dp_aux_native_read(struct intel_dp *intel_dp,
 			return -EPROTO;
 		if (ret < 0)
 			return ret;
-		ack = reply[0];
-		if ((ack & AUX_NATIVE_REPLY_MASK) == AUX_NATIVE_REPLY_ACK) {
+		ack = reply[0] >> 4;
+		if ((ack & DP_AUX_NATIVE_REPLY_MASK) == DP_AUX_NATIVE_REPLY_ACK) {
 			memcpy(recv, reply + 1, ret - 1);
 			return ret - 1;
 		}
-		else if ((ack & AUX_NATIVE_REPLY_MASK) == AUX_NATIVE_REPLY_DEFER)
+		else if ((ack & DP_AUX_NATIVE_REPLY_MASK) == DP_AUX_NATIVE_REPLY_DEFER)
 			udelay(100);
 		else
 			return -EIO;
@@ -633,12 +634,12 @@ intel_dp_i2c_aux_ch(struct i2c_adapter *adapter, int mode,
 	intel_dp_check_edp(intel_dp);
 	/* Set up the command byte */
 	if (mode & MODE_I2C_READ)
-		msg[0] = AUX_I2C_READ << 4;
+		msg[0] = DP_AUX_I2C_READ << 4;
 	else
-		msg[0] = AUX_I2C_WRITE << 4;
+		msg[0] = DP_AUX_I2C_WRITE << 4;
 
 	if (!(mode & MODE_I2C_STOP))
-		msg[0] |= AUX_I2C_MOT << 4;
+		msg[0] |= DP_AUX_I2C_MOT << 4;
 
 	msg[1] = address >> 8;
 	msg[2] = address;
@@ -675,17 +676,17 @@ intel_dp_i2c_aux_ch(struct i2c_adapter *adapter, int mode,
 			goto out;
 		}
 
-		switch (reply[0] & AUX_NATIVE_REPLY_MASK) {
-		case AUX_NATIVE_REPLY_ACK:
+		switch ((reply[0] >> 4) & DP_AUX_NATIVE_REPLY_MASK) {
+		case DP_AUX_NATIVE_REPLY_ACK:
 			/* I2C-over-AUX Reply field is only valid
 			 * when paired with AUX ACK.
 			 */
 			break;
-		case AUX_NATIVE_REPLY_NACK:
+		case DP_AUX_NATIVE_REPLY_NACK:
 			DRM_DEBUG_KMS("aux_ch native nack\n");
 			ret = -EREMOTEIO;
 			goto out;
-		case AUX_NATIVE_REPLY_DEFER:
+		case DP_AUX_NATIVE_REPLY_DEFER:
 			/*
 			 * For now, just give more slack to branch devices. We
 			 * could check the DPCD for I2C bit rate capabilities,
@@ -706,18 +707,18 @@ intel_dp_i2c_aux_ch(struct i2c_adapter *adapter, int mode,
 			goto out;
 		}
 
-		switch (reply[0] & AUX_I2C_REPLY_MASK) {
-		case AUX_I2C_REPLY_ACK:
+		switch ((reply[0] >> 4) & DP_AUX_I2C_REPLY_MASK) {
+		case DP_AUX_I2C_REPLY_ACK:
 			if (mode == MODE_I2C_READ) {
 				*read_byte = reply[1];
 			}
 			ret = reply_bytes - 1;
 			goto out;
-		case AUX_I2C_REPLY_NACK:
+		case DP_AUX_I2C_REPLY_NACK:
 			DRM_DEBUG_KMS("aux_i2c nack\n");
 			ret = -EREMOTEIO;
 			goto out;
-		case AUX_I2C_REPLY_DEFER:
+		case DP_AUX_I2C_REPLY_DEFER:
 			DRM_DEBUG_KMS("aux_i2c defer\n");
 			udelay(100);
 			break;
@@ -1037,6 +1038,8 @@ static void ironlake_wait_panel_status(struct intel_dp *intel_dp,
 				I915_READ(pp_stat_reg),
 				I915_READ(pp_ctrl_reg));
 	}
+
+	DRM_DEBUG_KMS("Wait complete\n");
 }
 
 static void ironlake_wait_panel_on(struct intel_dp *intel_dp)
@@ -1092,6 +1095,8 @@ void ironlake_edp_panel_vdd_on(struct intel_dp *intel_dp)
 	if (ironlake_edp_have_panel_vdd(intel_dp))
 		return;
 
+	intel_runtime_pm_get(dev_priv);
+
 	DRM_DEBUG_KMS("Turning eDP VDD on\n");
 
 	if (!ironlake_edp_have_panel_power(intel_dp))
@@ -1140,7 +1145,11 @@ static void ironlake_panel_vdd_off_sync(struct intel_dp *intel_dp)
 		/* 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));
-		msleep(intel_dp->panel_power_down_delay);
+
+		if ((pp & POWER_TARGET_ON) == 0)
+			msleep(intel_dp->panel_power_cycle_delay);
+
+		intel_runtime_pm_put(dev_priv);
 	}
 }
 
@@ -1233,20 +1242,16 @@ void ironlake_edp_panel_off(struct intel_dp *intel_dp)
 
 	DRM_DEBUG_KMS("Turn eDP power off\n");
 
-	WARN(!intel_dp->want_panel_vdd, "Need VDD to turn off panel\n");
-
 	pp = ironlake_get_pp_control(intel_dp);
 	/* We need to switch off panel power _and_ force vdd, for otherwise some
 	 * panels get very unhappy and cease to work. */
-	pp &= ~(POWER_TARGET_ON | EDP_FORCE_VDD | PANEL_POWER_RESET | EDP_BLC_ENABLE);
+	pp &= ~(POWER_TARGET_ON | PANEL_POWER_RESET | EDP_BLC_ENABLE);
 
 	pp_ctrl_reg = _pp_ctrl_reg(intel_dp);
 
 	I915_WRITE(pp_ctrl_reg, pp);
 	POSTING_READ(pp_ctrl_reg);
 
-	intel_dp->want_panel_vdd = false;
-
 	ironlake_wait_panel_off(intel_dp);
 }
 
@@ -1772,7 +1777,6 @@ static void intel_disable_dp(struct intel_encoder *encoder)
 
 	/* Make sure the panel is off before trying to change the mode. But also
 	 * ensure that we have vdd while we switch off the panel. */
-	ironlake_edp_panel_vdd_on(intel_dp);
 	ironlake_edp_backlight_off(intel_dp);
 	intel_dp_sink_dpms(intel_dp, DRM_MODE_DPMS_OFF);
 	ironlake_edp_panel_off(intel_dp);
@@ -1845,23 +1849,23 @@ static void vlv_pre_enable_dp(struct intel_encoder *encoder)
 	struct drm_device *dev = encoder->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->base.crtc);
-	int port = vlv_dport_to_channel(dport);
+	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);
 
-	val = vlv_dpio_read(dev_priv, pipe, DPIO_DATA_LANE_A(port));
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS01_DW8(port));
 	val = 0;
 	if (pipe)
 		val |= (1<<21);
 	else
 		val &= ~(1<<21);
 	val |= 0x001000c4;
-	vlv_dpio_write(dev_priv, pipe, DPIO_DATA_CHANNEL(port), val);
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_CLOCKBUF0(port), 0x00760018);
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_CLOCKBUF8(port), 0x00400888);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW8(port), val);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW14(port), 0x00760018);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW23(port), 0x00400888);
 
 	mutex_unlock(&dev_priv->dpio_lock);
 
@@ -1872,7 +1876,7 @@ static void vlv_pre_enable_dp(struct intel_encoder *encoder)
 
 	intel_enable_dp(encoder);
 
-	vlv_wait_port_ready(dev_priv, port);
+	vlv_wait_port_ready(dev_priv, dport);
 }
 
 static void vlv_dp_pre_pll_enable(struct intel_encoder *encoder)
@@ -1882,24 +1886,24 @@ static void vlv_dp_pre_pll_enable(struct intel_encoder *encoder)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc =
 		to_intel_crtc(encoder->base.crtc);
-	int port = vlv_dport_to_channel(dport);
+	enum dpio_channel port = vlv_dport_to_channel(dport);
 	int pipe = intel_crtc->pipe;
 
 	/* Program Tx lane resets to default */
 	mutex_lock(&dev_priv->dpio_lock);
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_TX(port),
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW0(port),
 			 DPIO_PCS_TX_LANE2_RESET |
 			 DPIO_PCS_TX_LANE1_RESET);
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_CLK(port),
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW1(port),
 			 DPIO_PCS_CLK_CRI_RXEB_EIOS_EN |
 			 DPIO_PCS_CLK_CRI_RXDIGFILTSG_EN |
 			 (1<<DPIO_PCS_CLK_DATAWIDTH_SHIFT) |
 				 DPIO_PCS_CLK_SOFT_RESET);
 
 	/* Fix up inter-pair skew failure */
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_STAGGER1(port), 0x00750f00);
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX_CTL(port), 0x00001500);
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX_LANE(port), 0x40400000);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW12(port), 0x00750f00);
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW11(port), 0x00001500);
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW14(port), 0x40400000);
 	mutex_unlock(&dev_priv->dpio_lock);
 }
 
@@ -1941,18 +1945,6 @@ intel_dp_get_link_status(struct intel_dp *intel_dp, uint8_t link_status[DP_LINK_
 					      DP_LINK_STATUS_SIZE);
 }
 
-#if 0
-static char	*voltage_names[] = {
-	"0.4V", "0.6V", "0.8V", "1.2V"
-};
-static char	*pre_emph_names[] = {
-	"0dB", "3.5dB", "6dB", "9.5dB"
-};
-static char	*link_train_names[] = {
-	"pattern 1", "pattern 2", "idle", "off"
-};
-#endif
-
 /*
  * These are source-specific values; current Intel hardware supports
  * a maximum voltage of 800mV and a maximum pre-emphasis of 6dB
@@ -2050,7 +2042,7 @@ static uint32_t intel_vlv_signal_levels(struct intel_dp *intel_dp)
 	unsigned long demph_reg_value, preemph_reg_value,
 		uniqtranscale_reg_value;
 	uint8_t train_set = intel_dp->train_set[0];
-	int port = vlv_dport_to_channel(dport);
+	enum dpio_channel port = vlv_dport_to_channel(dport);
 	int pipe = intel_crtc->pipe;
 
 	switch (train_set & DP_TRAIN_PRE_EMPHASIS_MASK) {
@@ -2127,14 +2119,14 @@ static uint32_t intel_vlv_signal_levels(struct intel_dp *intel_dp)
 	}
 
 	mutex_lock(&dev_priv->dpio_lock);
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX_OCALINIT(port), 0x00000000);
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX_SWING_CTL4(port), demph_reg_value);
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX_SWING_CTL2(port),
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW5(port), 0x00000000);
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW4(port), demph_reg_value);
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW2(port),
 			 uniqtranscale_reg_value);
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX_SWING_CTL3(port), 0x0C782040);
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_STAGGER0(port), 0x00030000);
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_CTL_OVER1(port), preemph_reg_value);
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX_OCALINIT(port), 0x80000000);
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW3(port), 0x0C782040);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW11(port), 0x00030000);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW9(port), preemph_reg_value);
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW5(port), 0x80000000);
 	mutex_unlock(&dev_priv->dpio_lock);
 
 	return 0;
@@ -2646,7 +2638,6 @@ intel_dp_complete_link_train(struct intel_dp *intel_dp)
 
 		if (cr_tries > 5) {
 			DRM_ERROR("failed to train DP, aborting\n");
-			intel_dp_link_down(intel_dp);
 			break;
 		}
 
@@ -2899,13 +2890,11 @@ intel_dp_check_link_status(struct intel_dp *intel_dp)
 
 	/* Try to read receiver status if the link appears to be up */
 	if (!intel_dp_get_link_status(intel_dp, link_status)) {
-		intel_dp_link_down(intel_dp);
 		return;
 	}
 
 	/* Now read the DPCD to see if it's actually running */
 	if (!intel_dp_get_dpcd(intel_dp)) {
-		intel_dp_link_down(intel_dp);
 		return;
 	}
 
@@ -3020,18 +3009,34 @@ g4x_dp_detect(struct intel_dp *intel_dp)
 		return status;
 	}
 
-	switch (intel_dig_port->port) {
-	case PORT_B:
-		bit = PORTB_HOTPLUG_LIVE_STATUS;
-		break;
-	case PORT_C:
-		bit = PORTC_HOTPLUG_LIVE_STATUS;
-		break;
-	case PORT_D:
-		bit = PORTD_HOTPLUG_LIVE_STATUS;
-		break;
-	default:
-		return connector_status_unknown;
+	if (IS_VALLEYVIEW(dev)) {
+		switch (intel_dig_port->port) {
+		case PORT_B:
+			bit = PORTB_HOTPLUG_LIVE_STATUS_VLV;
+			break;
+		case PORT_C:
+			bit = PORTC_HOTPLUG_LIVE_STATUS_VLV;
+			break;
+		case PORT_D:
+			bit = PORTD_HOTPLUG_LIVE_STATUS_VLV;
+			break;
+		default:
+			return connector_status_unknown;
+		}
+	} else {
+		switch (intel_dig_port->port) {
+		case PORT_B:
+			bit = PORTB_HOTPLUG_LIVE_STATUS_G4X;
+			break;
+		case PORT_C:
+			bit = PORTC_HOTPLUG_LIVE_STATUS_G4X;
+			break;
+		case PORT_D:
+			bit = PORTD_HOTPLUG_LIVE_STATUS_G4X;
+			break;
+		default:
+			return connector_status_unknown;
+		}
 	}
 
 	if ((I915_READ(PORT_HOTPLUG_STAT) & bit) == 0)
@@ -3082,9 +3087,12 @@ 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;
 	struct edid *edid = NULL;
 
+	intel_runtime_pm_get(dev_priv);
+
 	DRM_DEBUG_KMS("[CONNECTOR:%d:%s]\n",
 		      connector->base.id, drm_get_connector_name(connector));
 
@@ -3096,7 +3104,7 @@ intel_dp_detect(struct drm_connector *connector, bool force)
 		status = g4x_dp_detect(intel_dp);
 
 	if (status != connector_status_connected)
-		return status;
+		goto out;
 
 	intel_dp_probe_oui(intel_dp);
 
@@ -3112,7 +3120,11 @@ intel_dp_detect(struct drm_connector *connector, bool force)
 
 	if (intel_encoder->type != INTEL_OUTPUT_EDP)
 		intel_encoder->type = INTEL_OUTPUT_DISPLAYPORT;
-	return connector_status_connected;
+	status = connector_status_connected;
+
+out:
+	intel_runtime_pm_put(dev_priv);
+	return status;
 }
 
 static int intel_dp_get_modes(struct drm_connector *connector)
diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h
index 79f91f26e288..fbfaaba5cc3b 100644
--- a/drivers/gpu/drm/i915/intel_drv.h
+++ b/drivers/gpu/drm/i915/intel_drv.h
@@ -65,8 +65,8 @@
 #define wait_for_atomic_us(COND, US) _wait_for((COND), \
 					       DIV_ROUND_UP((US), 1000), 0)
 
-#define KHz(x) (1000*x)
-#define MHz(x) KHz(1000*x)
+#define KHz(x) (1000 * (x))
+#define MHz(x) KHz(1000 * (x))
 
 /*
  * Display related stuff
@@ -155,7 +155,19 @@ struct intel_encoder {
 
 struct intel_panel {
 	struct drm_display_mode *fixed_mode;
+	struct drm_display_mode *downclock_mode;
 	int fitting_mode;
+
+	/* backlight */
+	struct {
+		bool present;
+		u32 level;
+		u32 max;
+		bool enabled;
+		bool combination_mode;	/* gen 2/4 only */
+		bool active_low_pwm;
+		struct backlight_device *device;
+	} backlight;
 };
 
 struct intel_connector {
@@ -443,7 +455,7 @@ struct intel_hdmi {
 	bool rgb_quant_range_selectable;
 	void (*write_infoframe)(struct drm_encoder *encoder,
 				enum hdmi_infoframe_type type,
-				const uint8_t *frame, ssize_t len);
+				const void *frame, ssize_t len);
 	void (*set_infoframes)(struct drm_encoder *encoder,
 			       struct drm_display_mode *adjusted_mode);
 };
@@ -490,9 +502,9 @@ vlv_dport_to_channel(struct intel_digital_port *dport)
 {
 	switch (dport->port) {
 	case PORT_B:
-		return 0;
+		return DPIO_CH0;
 	case PORT_C:
-		return 1;
+		return DPIO_CH1;
 	default:
 		BUG();
 	}
@@ -601,7 +613,8 @@ void intel_ddi_disable_transcoder_func(struct drm_i915_private *dev_priv,
 void intel_ddi_enable_pipe_clock(struct intel_crtc *intel_crtc);
 void intel_ddi_disable_pipe_clock(struct intel_crtc *intel_crtc);
 void intel_ddi_setup_hw_pll_state(struct drm_device *dev);
-bool intel_ddi_pll_mode_set(struct drm_crtc *crtc);
+bool intel_ddi_pll_select(struct intel_crtc *crtc);
+void intel_ddi_pll_enable(struct intel_crtc *crtc);
 void intel_ddi_put_crtc_pll(struct drm_crtc *crtc);
 void intel_ddi_set_pipe_settings(struct drm_crtc *crtc);
 void intel_ddi_prepare_link_retrain(struct drm_encoder *encoder);
@@ -612,6 +625,8 @@ void intel_ddi_get_config(struct intel_encoder *encoder,
 
 
 /* intel_display.c */
+const char *intel_output_name(int output);
+bool intel_has_pending_fb_unpin(struct drm_device *dev);
 int intel_pch_rawclk(struct drm_device *dev);
 void intel_mark_busy(struct drm_device *dev);
 void intel_mark_fb_busy(struct drm_i915_gem_object *obj,
@@ -638,7 +653,8 @@ enum transcoder intel_pipe_to_cpu_transcoder(struct drm_i915_private *dev_priv,
 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, int port);
+void vlv_wait_port_ready(struct drm_i915_private *dev_priv,
+			 struct intel_digital_port *dport);
 bool intel_get_load_detect_pipe(struct drm_connector *connector,
 				struct drm_display_mode *mode,
 				struct intel_load_detect_pipe *old);
@@ -690,11 +706,10 @@ void
 ironlake_check_encoder_dotclock(const struct intel_crtc_config *pipe_config,
 				int dotclock);
 bool intel_crtc_active(struct drm_crtc *crtc);
-void i915_disable_vga_mem(struct drm_device *dev);
 void hsw_enable_ips(struct intel_crtc *crtc);
 void hsw_disable_ips(struct intel_crtc *crtc);
 void intel_display_set_init_power(struct drm_device *dev, bool enable);
-
+int valleyview_get_vco(struct drm_i915_private *dev_priv);
 
 /* intel_dp.c */
 void intel_dp_init(struct drm_device *dev, int output_reg, enum port port);
@@ -808,9 +823,13 @@ void intel_panel_set_backlight(struct intel_connector *connector, u32 level,
 int intel_panel_setup_backlight(struct drm_connector *connector);
 void intel_panel_enable_backlight(struct intel_connector *connector);
 void intel_panel_disable_backlight(struct intel_connector *connector);
-void intel_panel_destroy_backlight(struct drm_device *dev);
+void intel_panel_destroy_backlight(struct drm_connector *connector);
+void intel_panel_init_backlight_funcs(struct drm_device *dev);
 enum drm_connector_status intel_panel_detect(struct drm_device *dev);
-
+extern struct drm_display_mode *intel_find_panel_downclock(
+				struct drm_device *dev,
+				struct drm_display_mode *fixed_mode,
+				struct drm_connector *connector);
 
 /* intel_pm.c */
 void intel_init_clock_gating(struct drm_device *dev);
@@ -830,6 +849,8 @@ int intel_power_domains_init(struct drm_device *dev);
 void intel_power_domains_remove(struct drm_device *dev);
 bool intel_display_power_enabled(struct drm_device *dev,
 				 enum intel_display_power_domain domain);
+bool intel_display_power_enabled_sw(struct drm_device *dev,
+				    enum intel_display_power_domain domain);
 void intel_display_power_get(struct drm_device *dev,
 			     enum intel_display_power_domain domain);
 void intel_display_power_put(struct drm_device *dev,
@@ -844,6 +865,10 @@ void gen6_rps_idle(struct drm_i915_private *dev_priv);
 void gen6_rps_boost(struct drm_i915_private *dev_priv);
 void intel_aux_display_runtime_get(struct drm_i915_private *dev_priv);
 void intel_aux_display_runtime_put(struct drm_i915_private *dev_priv);
+void intel_runtime_pm_get(struct drm_i915_private *dev_priv);
+void intel_runtime_pm_put(struct drm_i915_private *dev_priv);
+void intel_init_runtime_pm(struct drm_i915_private *dev_priv);
+void intel_fini_runtime_pm(struct drm_i915_private *dev_priv);
 void ilk_wm_get_hw_state(struct drm_device *dev);
 
 
diff --git a/drivers/gpu/drm/i915/intel_dsi.c b/drivers/gpu/drm/i915/intel_dsi.c
index d257b093ca68..fabbf0d895cf 100644
--- a/drivers/gpu/drm/i915/intel_dsi.c
+++ b/drivers/gpu/drm/i915/intel_dsi.c
@@ -37,49 +37,18 @@
 static const struct intel_dsi_device intel_dsi_devices[] = {
 };
 
-
-static void vlv_cck_modify(struct drm_i915_private *dev_priv, u32 reg, u32 val,
-			   u32 mask)
-{
-	u32 tmp = vlv_cck_read(dev_priv, reg);
-	tmp &= ~mask;
-	tmp |= val;
-	vlv_cck_write(dev_priv, reg, tmp);
-}
-
-static void band_gap_wa(struct drm_i915_private *dev_priv)
+static void band_gap_reset(struct drm_i915_private *dev_priv)
 {
 	mutex_lock(&dev_priv->dpio_lock);
 
-	/* Enable bandgap fix in GOP driver */
-	vlv_cck_modify(dev_priv, 0x6D, 0x00010000, 0x00030000);
-	msleep(20);
-	vlv_cck_modify(dev_priv, 0x6E, 0x00010000, 0x00030000);
-	msleep(20);
-	vlv_cck_modify(dev_priv, 0x6F, 0x00010000, 0x00030000);
-	msleep(20);
-	vlv_cck_modify(dev_priv, 0x00, 0x00008000, 0x00008000);
-	msleep(20);
-	vlv_cck_modify(dev_priv, 0x00, 0x00000000, 0x00008000);
-	msleep(20);
-
-	/* Turn Display Trunk on */
-	vlv_cck_modify(dev_priv, 0x6B, 0x00020000, 0x00030000);
-	msleep(20);
-
-	vlv_cck_modify(dev_priv, 0x6C, 0x00020000, 0x00030000);
-	msleep(20);
-
-	vlv_cck_modify(dev_priv, 0x6D, 0x00020000, 0x00030000);
-	msleep(20);
-	vlv_cck_modify(dev_priv, 0x6E, 0x00020000, 0x00030000);
-	msleep(20);
-	vlv_cck_modify(dev_priv, 0x6F, 0x00020000, 0x00030000);
+	vlv_flisdsi_write(dev_priv, 0x08, 0x0001);
+	vlv_flisdsi_write(dev_priv, 0x0F, 0x0005);
+	vlv_flisdsi_write(dev_priv, 0x0F, 0x0025);
+	udelay(150);
+	vlv_flisdsi_write(dev_priv, 0x0F, 0x0000);
+	vlv_flisdsi_write(dev_priv, 0x08, 0x0000);
 
 	mutex_unlock(&dev_priv->dpio_lock);
-
-	/* Need huge delay, otherwise clock is not stable */
-	msleep(100);
 }
 
 static struct intel_dsi *intel_attached_dsi(struct drm_connector *connector)
@@ -132,14 +101,47 @@ static void intel_dsi_pre_pll_enable(struct intel_encoder *encoder)
 	vlv_enable_dsi_pll(encoder);
 }
 
+static void intel_dsi_device_ready(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);
+	int pipe = intel_crtc->pipe;
+	u32 val;
+
+	DRM_DEBUG_KMS("\n");
+
+	val = I915_READ(MIPI_PORT_CTRL(pipe));
+	I915_WRITE(MIPI_PORT_CTRL(pipe), val | LP_OUTPUT_HOLD);
+	usleep_range(1000, 1500);
+	I915_WRITE(MIPI_DEVICE_READY(pipe), DEVICE_READY | ULPS_STATE_EXIT);
+	usleep_range(2000, 2500);
+	I915_WRITE(MIPI_DEVICE_READY(pipe), DEVICE_READY);
+	usleep_range(2000, 2500);
+	I915_WRITE(MIPI_DEVICE_READY(pipe), 0x00);
+	usleep_range(2000, 2500);
+	I915_WRITE(MIPI_DEVICE_READY(pipe), DEVICE_READY);
+	usleep_range(2000, 2500);
+}
 static void intel_dsi_pre_enable(struct intel_encoder *encoder)
 {
+	struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base);
+
 	DRM_DEBUG_KMS("\n");
+
+	if (intel_dsi->dev.dev_ops->panel_reset)
+		intel_dsi->dev.dev_ops->panel_reset(&intel_dsi->dev);
+
+	/* put device in ready state */
+	intel_dsi_device_ready(encoder);
+
+	if (intel_dsi->dev.dev_ops->send_otp_cmds)
+		intel_dsi->dev.dev_ops->send_otp_cmds(&intel_dsi->dev);
 }
 
 static void intel_dsi_enable(struct intel_encoder *encoder)
 {
-	struct drm_i915_private *dev_priv = encoder->base.dev->dev_private;
+	struct drm_device *dev = encoder->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->base.crtc);
 	struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base);
 	int pipe = intel_crtc->pipe;
@@ -147,41 +149,28 @@ static void intel_dsi_enable(struct intel_encoder *encoder)
 
 	DRM_DEBUG_KMS("\n");
 
-	temp = I915_READ(MIPI_DEVICE_READY(pipe));
-	if ((temp & DEVICE_READY) == 0) {
-		temp &= ~ULPS_STATE_MASK;
-		I915_WRITE(MIPI_DEVICE_READY(pipe), temp | DEVICE_READY);
-	} else if (temp & ULPS_STATE_MASK) {
-		temp &= ~ULPS_STATE_MASK;
-		I915_WRITE(MIPI_DEVICE_READY(pipe), temp | ULPS_STATE_EXIT);
-		/*
-		 * We need to ensure that there is a minimum of 1 ms time
-		 * available before clearing the UPLS exit state.
-		 */
-		msleep(2);
-		I915_WRITE(MIPI_DEVICE_READY(pipe), temp);
-	}
-
 	if (is_cmd_mode(intel_dsi))
 		I915_WRITE(MIPI_MAX_RETURN_PKT_SIZE(pipe), 8 * 4);
-
-	if (is_vid_mode(intel_dsi)) {
+	else {
 		msleep(20); /* XXX */
 		dpi_send_cmd(intel_dsi, TURN_ON);
 		msleep(100);
 
 		/* assert ip_tg_enable signal */
-		temp = I915_READ(MIPI_PORT_CTRL(pipe));
+		temp = I915_READ(MIPI_PORT_CTRL(pipe)) & ~LANE_CONFIGURATION_MASK;
+		temp = temp | intel_dsi->port_bits;
 		I915_WRITE(MIPI_PORT_CTRL(pipe), temp | DPI_ENABLE);
 		POSTING_READ(MIPI_PORT_CTRL(pipe));
 	}
 
-	intel_dsi->dev.dev_ops->enable(&intel_dsi->dev);
+	if (intel_dsi->dev.dev_ops->enable)
+		intel_dsi->dev.dev_ops->enable(&intel_dsi->dev);
 }
 
 static void intel_dsi_disable(struct intel_encoder *encoder)
 {
-	struct drm_i915_private *dev_priv = encoder->base.dev->dev_private;
+	struct drm_device *dev = encoder->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->base.crtc);
 	struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base);
 	int pipe = intel_crtc->pipe;
@@ -189,8 +178,6 @@ static void intel_dsi_disable(struct intel_encoder *encoder)
 
 	DRM_DEBUG_KMS("\n");
 
-	intel_dsi->dev.dev_ops->disable(&intel_dsi->dev);
-
 	if (is_vid_mode(intel_dsi)) {
 		dpi_send_cmd(intel_dsi, SHUTDOWN);
 		msleep(10);
@@ -203,20 +190,54 @@ static void intel_dsi_disable(struct intel_encoder *encoder)
 		msleep(2);
 	}
 
-	temp = I915_READ(MIPI_DEVICE_READY(pipe));
-	if (temp & DEVICE_READY) {
-		temp &= ~DEVICE_READY;
-		temp &= ~ULPS_STATE_MASK;
-		I915_WRITE(MIPI_DEVICE_READY(pipe), temp);
-	}
+	/* if disable packets are sent before sending shutdown packet then in
+	 * some next enable sequence send turn on packet error is observed */
+	if (intel_dsi->dev.dev_ops->disable)
+		intel_dsi->dev.dev_ops->disable(&intel_dsi->dev);
 }
 
-static void intel_dsi_post_disable(struct intel_encoder *encoder)
+static void intel_dsi_clear_device_ready(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);
+	int pipe = intel_crtc->pipe;
+	u32 val;
+
 	DRM_DEBUG_KMS("\n");
 
+	I915_WRITE(MIPI_DEVICE_READY(pipe), ULPS_STATE_ENTER);
+	usleep_range(2000, 2500);
+
+	I915_WRITE(MIPI_DEVICE_READY(pipe), ULPS_STATE_EXIT);
+	usleep_range(2000, 2500);
+
+	I915_WRITE(MIPI_DEVICE_READY(pipe), ULPS_STATE_ENTER);
+	usleep_range(2000, 2500);
+
+	val = I915_READ(MIPI_PORT_CTRL(pipe));
+	I915_WRITE(MIPI_PORT_CTRL(pipe), val & ~LP_OUTPUT_HOLD);
+	usleep_range(1000, 1500);
+
+	if (wait_for(((I915_READ(MIPI_PORT_CTRL(pipe)) & AFE_LATCHOUT)
+					== 0x00000), 30))
+		DRM_ERROR("DSI LP not going Low\n");
+
+	I915_WRITE(MIPI_DEVICE_READY(pipe), 0x00);
+	usleep_range(2000, 2500);
+
 	vlv_disable_dsi_pll(encoder);
 }
+static void intel_dsi_post_disable(struct intel_encoder *encoder)
+{
+	struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base);
+
+	DRM_DEBUG_KMS("\n");
+
+	intel_dsi_clear_device_ready(encoder);
+
+	if (intel_dsi->dev.dev_ops->disable_panel_power)
+		intel_dsi->dev.dev_ops->disable_panel_power(&intel_dsi->dev);
+}
 
 static bool intel_dsi_get_hw_state(struct intel_encoder *encoder,
 				   enum pipe *pipe)
@@ -251,8 +272,9 @@ static void intel_dsi_get_config(struct intel_encoder *encoder,
 	/* XXX: read flags, set to adjusted_mode */
 }
 
-static int intel_dsi_mode_valid(struct drm_connector *connector,
-				struct drm_display_mode *mode)
+static enum drm_mode_status
+intel_dsi_mode_valid(struct drm_connector *connector,
+		     struct drm_display_mode *mode)
 {
 	struct intel_connector *intel_connector = to_intel_connector(connector);
 	struct drm_display_mode *fixed_mode = intel_connector->panel.fixed_mode;
@@ -352,11 +374,8 @@ static void intel_dsi_mode_set(struct intel_encoder *intel_encoder)
 
 	DRM_DEBUG_KMS("pipe %c\n", pipe_name(pipe));
 
-	/* Update the DSI PLL */
-	vlv_enable_dsi_pll(intel_encoder);
-
 	/* XXX: Location of the call */
-	band_gap_wa(dev_priv);
+	band_gap_reset(dev_priv);
 
 	/* escape clock divider, 20MHz, shared for A and C. device ready must be
 	 * off when doing this! txclkesc? */
@@ -373,11 +392,7 @@ static void intel_dsi_mode_set(struct intel_encoder *intel_encoder)
 	I915_WRITE(MIPI_INTR_STAT(pipe), 0xffffffff);
 	I915_WRITE(MIPI_INTR_EN(pipe), 0xffffffff);
 
-	I915_WRITE(MIPI_DPHY_PARAM(pipe),
-		   0x3c << EXIT_ZERO_COUNT_SHIFT |
-		   0x1f << TRAIL_COUNT_SHIFT |
-		   0xc5 << CLK_ZERO_COUNT_SHIFT |
-		   0x1f << PREPARE_COUNT_SHIFT);
+	I915_WRITE(MIPI_DPHY_PARAM(pipe), intel_dsi->dphy_reg);
 
 	I915_WRITE(MIPI_DPI_RESOLUTION(pipe),
 		   adjusted_mode->vdisplay << VERTICAL_ADDRESS_SHIFT |
@@ -425,9 +440,9 @@ static void intel_dsi_mode_set(struct intel_encoder *intel_encoder)
 				       adjusted_mode->htotal,
 				       bpp, intel_dsi->lane_count) + 1);
 	}
-	I915_WRITE(MIPI_LP_RX_TIMEOUT(pipe), 8309); /* max */
-	I915_WRITE(MIPI_TURN_AROUND_TIMEOUT(pipe), 0x14); /* max */
-	I915_WRITE(MIPI_DEVICE_RESET_TIMER(pipe), 0xffff); /* max */
+	I915_WRITE(MIPI_LP_RX_TIMEOUT(pipe), intel_dsi->lp_rx_timeout);
+	I915_WRITE(MIPI_TURN_AROUND_TIMEOUT(pipe), intel_dsi->turn_arnd_val);
+	I915_WRITE(MIPI_DEVICE_RESET_TIMER(pipe), intel_dsi->rst_timer_val);
 
 	/* dphy stuff */
 
@@ -442,29 +457,31 @@ static void intel_dsi_mode_set(struct intel_encoder *intel_encoder)
 	 *
 	 * XXX: write MIPI_STOP_STATE_STALL?
 	 */
-	I915_WRITE(MIPI_HIGH_LOW_SWITCH_COUNT(pipe), 0x46);
+	I915_WRITE(MIPI_HIGH_LOW_SWITCH_COUNT(pipe),
+						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
 	 * and txclkesc. txclkesc time / txbyteclk time * (105 +
 	 * MIPI_STOP_STATE_STALL) / 105.???
 	 */
-	I915_WRITE(MIPI_LP_BYTECLK(pipe), 4);
+	I915_WRITE(MIPI_LP_BYTECLK(pipe), intel_dsi->lp_byte_clk);
 
 	/* the bw essential for transmitting 16 long packets containing 252
 	 * bytes meant for dcs write memory command is programmed in this
 	 * register in terms of byte clocks. based on dsi transfer rate and the
 	 * number of lanes configured the time taken to transmit 16 long packets
 	 * in a dsi stream varies. */
-	I915_WRITE(MIPI_DBI_BW_CTRL(pipe), 0x820);
+	I915_WRITE(MIPI_DBI_BW_CTRL(pipe), intel_dsi->bw_timer);
 
 	I915_WRITE(MIPI_CLK_LANE_SWITCH_TIME_CNT(pipe),
-		   0xa << LP_HS_SSW_CNT_SHIFT |
-		   0x14 << HS_LP_PWR_SW_CNT_SHIFT);
+		   intel_dsi->clk_lp_to_hs_count << LP_HS_SSW_CNT_SHIFT |
+		   intel_dsi->clk_hs_to_lp_count << HS_LP_PWR_SW_CNT_SHIFT);
 
 	if (is_vid_mode(intel_dsi))
 		I915_WRITE(MIPI_VIDEO_MODE_FORMAT(pipe),
-			   intel_dsi->video_mode_format);
+				intel_dsi->video_frmt_cfg_bits |
+				intel_dsi->video_mode_format);
 }
 
 static enum drm_connector_status
diff --git a/drivers/gpu/drm/i915/intel_dsi.h b/drivers/gpu/drm/i915/intel_dsi.h
index c7765f33d524..b4a27cec882f 100644
--- a/drivers/gpu/drm/i915/intel_dsi.h
+++ b/drivers/gpu/drm/i915/intel_dsi.h
@@ -39,6 +39,13 @@ struct intel_dsi_device {
 struct intel_dsi_dev_ops {
 	bool (*init)(struct intel_dsi_device *dsi);
 
+	void (*panel_reset)(struct intel_dsi_device *dsi);
+
+	void (*disable_panel_power)(struct intel_dsi_device *dsi);
+
+	/* one time programmable commands if needed */
+	void (*send_otp_cmds)(struct intel_dsi_device *dsi);
+
 	/* This callback must be able to assume DSI commands can be sent */
 	void (*enable)(struct intel_dsi_device *dsi);
 
@@ -89,6 +96,20 @@ struct intel_dsi {
 
 	/* eot for MIPI_EOT_DISABLE register */
 	u32 eot_disable;
+
+	u32 port_bits;
+	u32 bw_timer;
+	u32 dphy_reg;
+	u32 video_frmt_cfg_bits;
+	u16 lp_byte_clk;
+
+	/* timeouts in byte clocks */
+	u16 lp_rx_timeout;
+	u16 turn_arnd_val;
+	u16 rst_timer_val;
+	u16 hs_to_lp_count;
+	u16 clk_lp_to_hs_count;
+	u16 clk_hs_to_lp_count;
 };
 
 static inline struct intel_dsi *enc_to_intel_dsi(struct drm_encoder *encoder)
diff --git a/drivers/gpu/drm/i915/intel_dsi_pll.c b/drivers/gpu/drm/i915/intel_dsi_pll.c
index 44279b2ade88..ba79ec19da3b 100644
--- a/drivers/gpu/drm/i915/intel_dsi_pll.c
+++ b/drivers/gpu/drm/i915/intel_dsi_pll.c
@@ -50,6 +50,8 @@ static const u32 lfsr_converts[] = {
 	71, 35							/* 91 - 92 */
 };
 
+#ifdef DSI_CLK_FROM_RR
+
 static u32 dsi_rr_formula(const struct drm_display_mode *mode,
 			  int pixel_format, int video_mode_format,
 			  int lane_count, bool eotp)
@@ -121,7 +123,7 @@ static u32 dsi_rr_formula(const struct drm_display_mode *mode,
 
 	/* the dsi clock is divided by 2 in the hardware to get dsi ddr clock */
 	dsi_bit_clock_hz = bytes_per_x_frames_x_lanes * 8;
-	dsi_clk = dsi_bit_clock_hz / (1000 * 1000);
+	dsi_clk = dsi_bit_clock_hz / 1000;
 
 	if (eotp && video_mode_format == VIDEO_MODE_BURST)
 		dsi_clk *= 2;
@@ -129,64 +131,37 @@ static u32 dsi_rr_formula(const struct drm_display_mode *mode,
 	return dsi_clk;
 }
 
-#ifdef MNP_FROM_TABLE
-
-struct dsi_clock_table {
-	u32 freq;
-	u8 m;
-	u8 p;
-};
-
-static const struct dsi_clock_table dsi_clk_tbl[] = {
-	{300, 72, 6}, {313, 75, 6}, {323, 78, 6}, {333, 80, 6},
-	{343, 82, 6}, {353, 85, 6}, {363, 87, 6}, {373, 90, 6},
-	{383, 92, 6}, {390, 78, 5}, {393, 79, 5}, {400, 80, 5},
-	{401, 80, 5}, {402, 80, 5}, {403, 81, 5}, {404, 81, 5},
-	{405, 81, 5}, {406, 81, 5}, {407, 81, 5}, {408, 82, 5},
-	{409, 82, 5}, {410, 82, 5}, {411, 82, 5}, {412, 82, 5},
-	{413, 83, 5}, {414, 83, 5}, {415, 83, 5}, {416, 83, 5},
-	{417, 83, 5}, {418, 84, 5}, {419, 84, 5}, {420, 84, 5},
-	{430, 86, 5}, {440, 88, 5}, {450, 90, 5}, {460, 92, 5},
-	{470, 75, 4}, {480, 77, 4}, {490, 78, 4}, {500, 80, 4},
-	{510, 82, 4}, {520, 83, 4}, {530, 85, 4}, {540, 86, 4},
-	{550, 88, 4}, {560, 90, 4}, {570, 91, 4}, {580, 70, 3},
-	{590, 71, 3}, {600, 72, 3}, {610, 73, 3}, {620, 74, 3},
-	{630, 76, 3}, {640, 77, 3}, {650, 78, 3}, {660, 79, 3},
-	{670, 80, 3}, {680, 82, 3}, {690, 83, 3}, {700, 84, 3},
-	{710, 85, 3}, {720, 86, 3}, {730, 88, 3}, {740, 89, 3},
-	{750, 90, 3}, {760, 91, 3}, {770, 92, 3}, {780, 62, 2},
-	{790, 63, 2}, {800, 64, 2}, {880, 70, 2}, {900, 72, 2},
-	{1000, 80, 2},		/* dsi clock frequency in Mhz*/
-};
+#else
 
-static int dsi_calc_mnp(u32 dsi_clk, struct dsi_mnp *dsi_mnp)
+/* Get DSI clock from pixel clock */
+static u32 dsi_clk_from_pclk(const struct drm_display_mode *mode,
+			  int pixel_format, int lane_count)
 {
-	unsigned int i;
-	u8 m;
-	u8 n;
-	u8 p;
-	u32 m_seed;
-
-	if (dsi_clk < 300 || dsi_clk > 1000)
-		return -ECHRNG;
+	u32 dsi_clk_khz;
+	u32 bpp;
 
-	for (i = 0; i <= ARRAY_SIZE(dsi_clk_tbl); i++) {
-		if (dsi_clk_tbl[i].freq > dsi_clk)
-			break;
+	switch (pixel_format) {
+	default:
+	case VID_MODE_FORMAT_RGB888:
+	case VID_MODE_FORMAT_RGB666_LOOSE:
+		bpp = 24;
+		break;
+	case VID_MODE_FORMAT_RGB666:
+		bpp = 18;
+		break;
+	case VID_MODE_FORMAT_RGB565:
+		bpp = 16;
+		break;
 	}
 
-	m = dsi_clk_tbl[i].m;
-	p = dsi_clk_tbl[i].p;
-	m_seed = lfsr_converts[m - 62];
-	n = 1;
-	dsi_mnp->dsi_pll_ctrl = 1 << (DSI_PLL_P1_POST_DIV_SHIFT + p - 2);
-	dsi_mnp->dsi_pll_div = (n - 1) << DSI_PLL_N1_DIV_SHIFT |
-		m_seed << DSI_PLL_M1_DIV_SHIFT;
+	/* 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);
 
-	return 0;
+	return dsi_clk_khz;
 }
 
-#else
+#endif
 
 static int dsi_calc_mnp(u32 dsi_clk, struct dsi_mnp *dsi_mnp)
 {
@@ -194,36 +169,47 @@ static int dsi_calc_mnp(u32 dsi_clk, struct dsi_mnp *dsi_mnp)
 	u32 ref_clk;
 	u32 error;
 	u32 tmp_error;
-	u32 target_dsi_clk;
-	u32 calc_dsi_clk;
+	int target_dsi_clk;
+	int calc_dsi_clk;
 	u32 calc_m;
 	u32 calc_p;
 	u32 m_seed;
 
-	if (dsi_clk < 300 || dsi_clk > 1150) {
+	/* dsi_clk is expected in KHZ */
+	if (dsi_clk < 300000 || dsi_clk > 1150000) {
 		DRM_ERROR("DSI CLK Out of Range\n");
 		return -ECHRNG;
 	}
 
 	ref_clk = 25000;
-	target_dsi_clk = dsi_clk * 1000;
+	target_dsi_clk = dsi_clk;
 	error = 0xFFFFFFFF;
+	tmp_error = 0xFFFFFFFF;
 	calc_m = 0;
 	calc_p = 0;
 
 	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 */
 			calc_dsi_clk = (m * ref_clk) / p;
-			if (calc_dsi_clk >= target_dsi_clk) {
-				tmp_error = calc_dsi_clk - target_dsi_clk;
-				if (tmp_error < error) {
-					error = tmp_error;
-					calc_m = m;
-					calc_p = p;
-				}
+			if (calc_dsi_clk == target_dsi_clk) {
+				calc_m = m;
+				calc_p = p;
+				error = 0;
+				break;
+			} else
+				tmp_error = abs(target_dsi_clk - calc_dsi_clk);
+
+			if (tmp_error < error) {
+				error = tmp_error;
+				calc_m = m;
+				calc_p = p;
 			}
 		}
+
+		if (error == 0)
+			break;
 	}
 
 	m_seed = lfsr_converts[calc_m - 62];
@@ -235,8 +221,6 @@ static int dsi_calc_mnp(u32 dsi_clk, struct dsi_mnp *dsi_mnp)
 	return 0;
 }
 
-#endif
-
 /*
  * XXX: The muxing and gating is hard coded for now. Need to add support for
  * sharing PLLs with two DSI outputs.
@@ -251,9 +235,8 @@ static void vlv_configure_dsi_pll(struct intel_encoder *encoder)
 	struct dsi_mnp dsi_mnp;
 	u32 dsi_clk;
 
-	dsi_clk = dsi_rr_formula(mode, intel_dsi->pixel_format,
-				 intel_dsi->video_mode_format,
-				 intel_dsi->lane_count, !intel_dsi->eot_disable);
+	dsi_clk = dsi_clk_from_pclk(mode, intel_dsi->pixel_format,
+						intel_dsi->lane_count);
 
 	ret = dsi_calc_mnp(dsi_clk, &dsi_mnp);
 	if (ret) {
diff --git a/drivers/gpu/drm/i915/intel_dvo.c b/drivers/gpu/drm/i915/intel_dvo.c
index 3c7736546856..eeff998e52ef 100644
--- a/drivers/gpu/drm/i915/intel_dvo.c
+++ b/drivers/gpu/drm/i915/intel_dvo.c
@@ -234,8 +234,9 @@ static void intel_dvo_dpms(struct drm_connector *connector, int mode)
 	intel_modeset_check_state(connector->dev);
 }
 
-static int intel_dvo_mode_valid(struct drm_connector *connector,
-				struct drm_display_mode *mode)
+static enum drm_mode_status
+intel_dvo_mode_valid(struct drm_connector *connector,
+		     struct drm_display_mode *mode)
 {
 	struct intel_dvo *intel_dvo = intel_attached_dvo(connector);
 
diff --git a/drivers/gpu/drm/i915/intel_fbdev.c b/drivers/gpu/drm/i915/intel_fbdev.c
index 895fcb4fbd94..39eac9937a4a 100644
--- a/drivers/gpu/drm/i915/intel_fbdev.c
+++ b/drivers/gpu/drm/i915/intel_fbdev.c
@@ -57,18 +57,14 @@ static struct fb_ops intelfb_ops = {
 	.fb_debug_leave = drm_fb_helper_debug_leave,
 };
 
-static int intelfb_create(struct drm_fb_helper *helper,
-			  struct drm_fb_helper_surface_size *sizes)
+static int intelfb_alloc(struct drm_fb_helper *helper,
+			 struct drm_fb_helper_surface_size *sizes)
 {
 	struct intel_fbdev *ifbdev =
 		container_of(helper, struct intel_fbdev, helper);
 	struct drm_device *dev = helper->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct fb_info *info;
-	struct drm_framebuffer *fb;
 	struct drm_mode_fb_cmd2 mode_cmd = {};
 	struct drm_i915_gem_object *obj;
-	struct device *device = &dev->pdev->dev;
 	int size, ret;
 
 	/* we don't do packed 24bpp */
@@ -94,8 +90,6 @@ static int intelfb_create(struct drm_fb_helper *helper,
 		goto out;
 	}
 
-	mutex_lock(&dev->struct_mutex);
-
 	/* Flush everything out, we'll be doing GTT only from now on */
 	ret = intel_pin_and_fence_fb_obj(dev, obj, NULL);
 	if (ret) {
@@ -103,7 +97,50 @@ static int intelfb_create(struct drm_fb_helper *helper,
 		goto out_unref;
 	}
 
-	info = framebuffer_alloc(0, device);
+	ret = intel_framebuffer_init(dev, &ifbdev->ifb, &mode_cmd, obj);
+	if (ret)
+		goto out_unpin;
+
+	return 0;
+
+out_unpin:
+	i915_gem_object_unpin(obj);
+out_unref:
+	drm_gem_object_unreference(&obj->base);
+out:
+	return ret;
+}
+
+static int intelfb_create(struct drm_fb_helper *helper,
+			  struct drm_fb_helper_surface_size *sizes)
+{
+	struct intel_fbdev *ifbdev =
+		container_of(helper, struct intel_fbdev, helper);
+	struct intel_framebuffer *intel_fb = &ifbdev->ifb;
+	struct drm_device *dev = helper->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct fb_info *info;
+	struct drm_framebuffer *fb;
+	struct drm_i915_gem_object *obj;
+	int size, ret;
+
+	mutex_lock(&dev->struct_mutex);
+
+	if (!intel_fb->obj) {
+		DRM_DEBUG_KMS("no BIOS fb, allocating a new one\n");
+		ret = intelfb_alloc(helper, sizes);
+		if (ret)
+			goto out_unlock;
+	} else {
+		DRM_DEBUG_KMS("re-using BIOS fb\n");
+		sizes->fb_width = intel_fb->base.width;
+		sizes->fb_height = intel_fb->base.height;
+	}
+
+	obj = intel_fb->obj;
+	size = obj->base.size;
+
+	info = framebuffer_alloc(0, &dev->pdev->dev);
 	if (!info) {
 		ret = -ENOMEM;
 		goto out_unpin;
@@ -111,10 +148,6 @@ static int intelfb_create(struct drm_fb_helper *helper,
 
 	info->par = helper;
 
-	ret = intel_framebuffer_init(dev, &ifbdev->ifb, &mode_cmd, obj);
-	if (ret)
-		goto out_unpin;
-
 	fb = &ifbdev->ifb.base;
 
 	ifbdev->helper.fb = fb;
@@ -170,17 +203,15 @@ static int intelfb_create(struct drm_fb_helper *helper,
 		      fb->width, fb->height,
 		      i915_gem_obj_ggtt_offset(obj), obj);
 
-
 	mutex_unlock(&dev->struct_mutex);
 	vga_switcheroo_client_fb_set(dev->pdev, info);
 	return 0;
 
 out_unpin:
 	i915_gem_object_unpin(obj);
-out_unref:
 	drm_gem_object_unreference(&obj->base);
+out_unlock:
 	mutex_unlock(&dev->struct_mutex);
-out:
 	return ret;
 }
 
@@ -297,8 +328,6 @@ void intel_fbdev_set_suspend(struct drm_device *dev, int state)
 	fb_set_suspend(info, state);
 }
 
-MODULE_LICENSE("GPL and additional rights");
-
 void intel_fbdev_output_poll_changed(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
diff --git a/drivers/gpu/drm/i915/intel_hdmi.c b/drivers/gpu/drm/i915/intel_hdmi.c
index 03f9ca70530c..6db0d9d17f47 100644
--- a/drivers/gpu/drm/i915/intel_hdmi.c
+++ b/drivers/gpu/drm/i915/intel_hdmi.c
@@ -130,9 +130,9 @@ static u32 hsw_infoframe_data_reg(enum hdmi_infoframe_type type,
 
 static void g4x_write_infoframe(struct drm_encoder *encoder,
 				enum hdmi_infoframe_type type,
-				const uint8_t *frame, ssize_t len)
+				const void *frame, ssize_t len)
 {
-	uint32_t *data = (uint32_t *)frame;
+	const uint32_t *data = frame;
 	struct drm_device *dev = encoder->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 val = I915_READ(VIDEO_DIP_CTL);
@@ -167,9 +167,9 @@ static void g4x_write_infoframe(struct drm_encoder *encoder,
 
 static void ibx_write_infoframe(struct drm_encoder *encoder,
 				enum hdmi_infoframe_type type,
-				const uint8_t *frame, ssize_t len)
+				const void *frame, ssize_t len)
 {
-	uint32_t *data = (uint32_t *)frame;
+	const uint32_t *data = frame;
 	struct drm_device *dev = encoder->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->crtc);
@@ -205,9 +205,9 @@ static void ibx_write_infoframe(struct drm_encoder *encoder,
 
 static void cpt_write_infoframe(struct drm_encoder *encoder,
 				enum hdmi_infoframe_type type,
-				const uint8_t *frame, ssize_t len)
+				const void *frame, ssize_t len)
 {
-	uint32_t *data = (uint32_t *)frame;
+	const uint32_t *data = frame;
 	struct drm_device *dev = encoder->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->crtc);
@@ -246,9 +246,9 @@ static void cpt_write_infoframe(struct drm_encoder *encoder,
 
 static void vlv_write_infoframe(struct drm_encoder *encoder,
 				enum hdmi_infoframe_type type,
-				const uint8_t *frame, ssize_t len)
+				const void *frame, ssize_t len)
 {
-	uint32_t *data = (uint32_t *)frame;
+	const uint32_t *data = frame;
 	struct drm_device *dev = encoder->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->crtc);
@@ -284,9 +284,9 @@ static void vlv_write_infoframe(struct drm_encoder *encoder,
 
 static void hsw_write_infoframe(struct drm_encoder *encoder,
 				enum hdmi_infoframe_type type,
-				const uint8_t *frame, ssize_t len)
+				const void *frame, ssize_t len)
 {
-	uint32_t *data = (uint32_t *)frame;
+	const uint32_t *data = frame;
 	struct drm_device *dev = encoder->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->crtc);
@@ -853,8 +853,9 @@ static int hdmi_portclock_limit(struct intel_hdmi *hdmi)
 		return 225000;
 }
 
-static int intel_hdmi_mode_valid(struct drm_connector *connector,
-				 struct drm_display_mode *mode)
+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)))
 		return MODE_CLOCK_HIGH;
@@ -1081,7 +1082,7 @@ static void vlv_hdmi_pre_enable(struct intel_encoder *encoder)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc =
 		to_intel_crtc(encoder->base.crtc);
-	int port = vlv_dport_to_channel(dport);
+	enum dpio_channel port = vlv_dport_to_channel(dport);
 	int pipe = intel_crtc->pipe;
 	u32 val;
 
@@ -1090,41 +1091,33 @@ static void vlv_hdmi_pre_enable(struct intel_encoder *encoder)
 
 	/* Enable clock channels for this port */
 	mutex_lock(&dev_priv->dpio_lock);
-	val = vlv_dpio_read(dev_priv, pipe, DPIO_DATA_LANE_A(port));
+	val = vlv_dpio_read(dev_priv, pipe, VLV_PCS01_DW8(port));
 	val = 0;
 	if (pipe)
 		val |= (1<<21);
 	else
 		val &= ~(1<<21);
 	val |= 0x001000c4;
-	vlv_dpio_write(dev_priv, pipe, DPIO_DATA_CHANNEL(port), val);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW8(port), val);
 
 	/* HDMI 1.0V-2dB */
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX_OCALINIT(port), 0);
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX_SWING_CTL4(port),
-			 0x2b245f5f);
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX_SWING_CTL2(port),
-			 0x5578b83a);
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX_SWING_CTL3(port),
-			 0x0c782040);
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX3_SWING_CTL4(port),
-			 0x2b247878);
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_STAGGER0(port), 0x00030000);
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_CTL_OVER1(port),
-			 0x00002000);
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX_OCALINIT(port),
-			 DPIO_TX_OCALINIT_EN);
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW5(port), 0);
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW4(port), 0x2b245f5f);
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW2(port), 0x5578b83a);
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW3(port), 0x0c782040);
+	vlv_dpio_write(dev_priv, pipe, VLV_TX3_DW4(port), 0x2b247878);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW11(port), 0x00030000);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW9(port), 0x00002000);
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW5(port), DPIO_TX_OCALINIT_EN);
 
 	/* Program lane clock */
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_CLOCKBUF0(port),
-			 0x00760018);
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_CLOCKBUF8(port),
-			 0x00400888);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW14(port), 0x00760018);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW23(port), 0x00400888);
 	mutex_unlock(&dev_priv->dpio_lock);
 
 	intel_enable_hdmi(encoder);
 
-	vlv_wait_port_ready(dev_priv, port);
+	vlv_wait_port_ready(dev_priv, dport);
 }
 
 static void vlv_hdmi_pre_pll_enable(struct intel_encoder *encoder)
@@ -1134,7 +1127,7 @@ static void vlv_hdmi_pre_pll_enable(struct intel_encoder *encoder)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_crtc *intel_crtc =
 		to_intel_crtc(encoder->base.crtc);
-	int port = vlv_dport_to_channel(dport);
+	enum dpio_channel port = vlv_dport_to_channel(dport);
 	int pipe = intel_crtc->pipe;
 
 	if (!IS_VALLEYVIEW(dev))
@@ -1142,24 +1135,22 @@ static void vlv_hdmi_pre_pll_enable(struct intel_encoder *encoder)
 
 	/* Program Tx lane resets to default */
 	mutex_lock(&dev_priv->dpio_lock);
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_TX(port),
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW0(port),
 			 DPIO_PCS_TX_LANE2_RESET |
 			 DPIO_PCS_TX_LANE1_RESET);
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_CLK(port),
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW1(port),
 			 DPIO_PCS_CLK_CRI_RXEB_EIOS_EN |
 			 DPIO_PCS_CLK_CRI_RXDIGFILTSG_EN |
 			 (1<<DPIO_PCS_CLK_DATAWIDTH_SHIFT) |
 			 DPIO_PCS_CLK_SOFT_RESET);
 
 	/* Fix up inter-pair skew failure */
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_STAGGER1(port), 0x00750f00);
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX_CTL(port), 0x00001500);
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX_LANE(port), 0x40400000);
-
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_CTL_OVER1(port),
-			 0x00002000);
-	vlv_dpio_write(dev_priv, pipe, DPIO_TX_OCALINIT(port),
-			 DPIO_TX_OCALINIT_EN);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW12(port), 0x00750f00);
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW11(port), 0x00001500);
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW14(port), 0x40400000);
+
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW9(port), 0x00002000);
+	vlv_dpio_write(dev_priv, pipe, VLV_TX_DW5(port), DPIO_TX_OCALINIT_EN);
 	mutex_unlock(&dev_priv->dpio_lock);
 }
 
@@ -1169,13 +1160,13 @@ static void vlv_hdmi_post_disable(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);
-	int port = vlv_dport_to_channel(dport);
+	enum dpio_channel port = vlv_dport_to_channel(dport);
 	int pipe = intel_crtc->pipe;
 
 	/* Reset lanes to avoid HDMI flicker (VLV w/a) */
 	mutex_lock(&dev_priv->dpio_lock);
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_TX(port), 0x00000000);
-	vlv_dpio_write(dev_priv, pipe, DPIO_PCS_CLK(port), 0x00e00060);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW0(port), 0x00000000);
+	vlv_dpio_write(dev_priv, pipe, VLV_PCS_DW1(port), 0x00e00060);
 	mutex_unlock(&dev_priv->dpio_lock);
 }
 
diff --git a/drivers/gpu/drm/i915/intel_i2c.c b/drivers/gpu/drm/i915/intel_i2c.c
index 2ca17b14b6c1..b1dc33f47899 100644
--- a/drivers/gpu/drm/i915/intel_i2c.c
+++ b/drivers/gpu/drm/i915/intel_i2c.c
@@ -82,20 +82,11 @@ static int get_disp_clk_div(struct drm_i915_private *dev_priv,
 
 static void gmbus_set_freq(struct drm_i915_private *dev_priv)
 {
-	int vco_freq[] = { 800, 1600, 2000, 2400 };
-	int gmbus_freq = 0, cdclk_div, hpll_freq;
+	int vco, gmbus_freq = 0, cdclk_div;
 
 	BUG_ON(!IS_VALLEYVIEW(dev_priv->dev));
 
-	/* Skip setting the gmbus freq if BIOS has already programmed it */
-	if (I915_READ(GMBUSFREQ_VLV) != 0xA0)
-		return;
-
-	/* Obtain SKU information */
-	mutex_lock(&dev_priv->dpio_lock);
-	hpll_freq =
-		vlv_cck_read(dev_priv, CCK_FUSE_REG) & CCK_FUSE_HPLL_FREQ_MASK;
-	mutex_unlock(&dev_priv->dpio_lock);
+	vco = valleyview_get_vco(dev_priv);
 
 	/* Get the CDCLK divide ratio */
 	cdclk_div = get_disp_clk_div(dev_priv, CDCLK);
@@ -106,7 +97,7 @@ static void gmbus_set_freq(struct drm_i915_private *dev_priv)
 	 * in fact 1MHz is the correct frequency.
 	 */
 	if (cdclk_div)
-		gmbus_freq = (vco_freq[hpll_freq] << 1) / cdclk_div;
+		gmbus_freq = (vco << 1) / cdclk_div;
 
 	if (WARN_ON(gmbus_freq == 0))
 		return;
diff --git a/drivers/gpu/drm/i915/intel_lvds.c b/drivers/gpu/drm/i915/intel_lvds.c
index c3b4da7895ed..8bcb93a2a9f6 100644
--- a/drivers/gpu/drm/i915/intel_lvds.c
+++ b/drivers/gpu/drm/i915/intel_lvds.c
@@ -256,8 +256,9 @@ static void intel_disable_lvds(struct intel_encoder *encoder)
 	POSTING_READ(lvds_encoder->reg);
 }
 
-static int intel_lvds_mode_valid(struct drm_connector *connector,
-				 struct drm_display_mode *mode)
+static enum drm_mode_status
+intel_lvds_mode_valid(struct drm_connector *connector,
+		      struct drm_display_mode *mode)
 {
 	struct intel_connector *intel_connector = to_intel_connector(connector);
 	struct drm_display_mode *fixed_mode = intel_connector->panel.fixed_mode;
@@ -446,9 +447,19 @@ static int intel_lid_notify(struct notifier_block *nb, unsigned long val,
 	if (dev_priv->modeset_restore == MODESET_DONE)
 		goto exit;
 
-	drm_modeset_lock_all(dev);
-	intel_modeset_setup_hw_state(dev, true);
-	drm_modeset_unlock_all(dev);
+	/*
+	 * Some old platform's BIOS love to wreak havoc while the lid is closed.
+	 * We try to detect this here and undo any damage. The split for PCH
+	 * platforms is rather conservative and a bit arbitrary expect that on
+	 * those platforms VGA disabling requires actual legacy VGA I/O access,
+	 * and as part of the cleanup in the hw state restore we also redisable
+	 * the vga plane.
+	 */
+	if (!HAS_PCH_SPLIT(dev)) {
+		drm_modeset_lock_all(dev);
+		intel_modeset_setup_hw_state(dev, true);
+		drm_modeset_unlock_all(dev);
+	}
 
 	dev_priv->modeset_restore = MODESET_DONE;
 
@@ -744,57 +755,6 @@ static const struct dmi_system_id intel_no_lvds[] = {
 	{ }	/* terminating entry */
 };
 
-/**
- * intel_find_lvds_downclock - find the reduced downclock for LVDS in EDID
- * @dev: drm device
- * @connector: LVDS connector
- *
- * Find the reduced downclock for LVDS in EDID.
- */
-static void intel_find_lvds_downclock(struct drm_device *dev,
-				      struct drm_display_mode *fixed_mode,
-				      struct drm_connector *connector)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_display_mode *scan;
-	int temp_downclock;
-
-	temp_downclock = fixed_mode->clock;
-	list_for_each_entry(scan, &connector->probed_modes, head) {
-		/*
-		 * If one mode has the same resolution with the fixed_panel
-		 * mode while they have the different refresh rate, it means
-		 * that the reduced downclock is found for the LVDS. In such
-		 * case we can set the different FPx0/1 to dynamically select
-		 * between low and high frequency.
-		 */
-		if (scan->hdisplay == fixed_mode->hdisplay &&
-		    scan->hsync_start == fixed_mode->hsync_start &&
-		    scan->hsync_end == fixed_mode->hsync_end &&
-		    scan->htotal == fixed_mode->htotal &&
-		    scan->vdisplay == fixed_mode->vdisplay &&
-		    scan->vsync_start == fixed_mode->vsync_start &&
-		    scan->vsync_end == fixed_mode->vsync_end &&
-		    scan->vtotal == fixed_mode->vtotal) {
-			if (scan->clock < temp_downclock) {
-				/*
-				 * The downclock is already found. But we
-				 * expect to find the lower downclock.
-				 */
-				temp_downclock = scan->clock;
-			}
-		}
-	}
-	if (temp_downclock < fixed_mode->clock && i915_lvds_downclock) {
-		/* We found the downclock for LVDS. */
-		dev_priv->lvds_downclock_avail = 1;
-		dev_priv->lvds_downclock = temp_downclock;
-		DRM_DEBUG_KMS("LVDS downclock is found in EDID. "
-			      "Normal clock %dKhz, downclock %dKhz\n",
-			      fixed_mode->clock, temp_downclock);
-	}
-}
-
 /*
  * Enumerate the child dev array parsed from VBT to check whether
  * the LVDS is present.
@@ -1072,8 +1032,22 @@ void intel_lvds_init(struct drm_device *dev)
 
 			fixed_mode = drm_mode_duplicate(dev, scan);
 			if (fixed_mode) {
-				intel_find_lvds_downclock(dev, fixed_mode,
-							  connector);
+				intel_connector->panel.downclock_mode =
+					intel_find_panel_downclock(dev,
+					fixed_mode, connector);
+				if (intel_connector->panel.downclock_mode !=
+					NULL &&	i915_lvds_downclock) {
+					/* We found the downclock for LVDS. */
+					dev_priv->lvds_downclock_avail = true;
+					dev_priv->lvds_downclock =
+						intel_connector->panel.
+						downclock_mode->clock;
+					DRM_DEBUG_KMS("LVDS downclock is found"
+					" in EDID. Normal clock %dKhz, "
+					"downclock %dKhz\n",
+					fixed_mode->clock,
+					dev_priv->lvds_downclock);
+				}
 				goto out;
 			}
 		}
diff --git a/drivers/gpu/drm/i915/intel_opregion.c b/drivers/gpu/drm/i915/intel_opregion.c
index 9a8804bee5cd..4e960ec7419f 100644
--- a/drivers/gpu/drm/i915/intel_opregion.c
+++ b/drivers/gpu/drm/i915/intel_opregion.c
@@ -63,7 +63,7 @@ struct opregion_header {
 	u8 driver_ver[16];
 	u32 mboxes;
 	u8 reserved[164];
-} __attribute__((packed));
+} __packed;
 
 /* OpRegion mailbox #1: public ACPI methods */
 struct opregion_acpi {
@@ -85,7 +85,7 @@ struct opregion_acpi {
 	u32 cnot;       /* current OS notification */
 	u32 nrdy;       /* driver status */
 	u8 rsvd2[60];
-} __attribute__((packed));
+} __packed;
 
 /* OpRegion mailbox #2: SWSCI */
 struct opregion_swsci {
@@ -93,7 +93,7 @@ struct opregion_swsci {
 	u32 parm;       /* command parameters */
 	u32 dslp;       /* driver sleep time-out */
 	u8 rsvd[244];
-} __attribute__((packed));
+} __packed;
 
 /* OpRegion mailbox #3: ASLE */
 struct opregion_asle {
@@ -114,7 +114,7 @@ struct opregion_asle {
 	u32 srot;       /* supported rotation angles */
 	u32 iuer;       /* IUER events */
 	u8 rsvd[86];
-} __attribute__((packed));
+} __packed;
 
 /* Driver readiness indicator */
 #define ASLE_ARDY_READY		(1 << 0)
@@ -395,13 +395,8 @@ int intel_opregion_notify_adapter(struct drm_device *dev, pci_power_t state)
 static u32 asle_set_backlight(struct drm_device *dev, u32 bclp)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_encoder *encoder;
-	struct drm_connector *connector;
-	struct intel_connector *intel_connector = NULL;
-	struct drm_crtc *crtc = dev_priv->pipe_to_crtc_mapping[0];
+	struct intel_connector *intel_connector;
 	struct opregion_asle __iomem *asle = dev_priv->opregion.asle;
-	u32 ret = 0;
-	bool found = false;
 
 	DRM_DEBUG_DRIVER("bclp = 0x%08x\n", bclp);
 
@@ -413,38 +408,20 @@ static u32 asle_set_backlight(struct drm_device *dev, u32 bclp)
 		return ASLC_BACKLIGHT_FAILED;
 
 	mutex_lock(&dev->mode_config.mutex);
+
 	/*
-	 * Could match the OpRegion connector here instead, but we'd also need
-	 * to verify the connector could handle a backlight call.
+	 * Update backlight on all connectors that support backlight (usually
+	 * only one).
 	 */
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list, head)
-		if (encoder->crtc == crtc) {
-			found = true;
-			break;
-		}
-
-	if (!found) {
-		ret = ASLC_BACKLIGHT_FAILED;
-		goto out;
-	}
-
-	list_for_each_entry(connector, &dev->mode_config.connector_list, head)
-		if (connector->encoder == encoder)
-			intel_connector = to_intel_connector(connector);
-
-	if (!intel_connector) {
-		ret = ASLC_BACKLIGHT_FAILED;
-		goto out;
-	}
-
 	DRM_DEBUG_KMS("updating opregion backlight %d/255\n", bclp);
-	intel_panel_set_backlight(intel_connector, bclp, 255);
+	list_for_each_entry(intel_connector, &dev->mode_config.connector_list, base.head)
+		intel_panel_set_backlight(intel_connector, bclp, 255);
 	iowrite32(DIV_ROUND_UP(bclp * 100, 255) | ASLE_CBLV_VALID, &asle->cblv);
 
-out:
 	mutex_unlock(&dev->mode_config.mutex);
 
-	return ret;
+
+	return 0;
 }
 
 static u32 asle_set_als_illum(struct drm_device *dev, u32 alsi)
diff --git a/drivers/gpu/drm/i915/intel_overlay.c b/drivers/gpu/drm/i915/intel_overlay.c
index a98a990fbab3..a759ecdb7a6e 100644
--- a/drivers/gpu/drm/i915/intel_overlay.c
+++ b/drivers/gpu/drm/i915/intel_overlay.c
@@ -1005,7 +1005,7 @@ static int intel_panel_fitter_pipe(struct drm_device *dev)
 	u32  pfit_control;
 
 	/* i830 doesn't have a panel fitter */
-	if (IS_I830(dev))
+	if (INTEL_INFO(dev)->gen <= 3 && (IS_I830(dev) || !IS_MOBILE(dev)))
 		return -1;
 
 	pfit_control = I915_READ(PFIT_CONTROL);
diff --git a/drivers/gpu/drm/i915/intel_panel.c b/drivers/gpu/drm/i915/intel_panel.c
index e6f782d1c669..350de359123a 100644
--- a/drivers/gpu/drm/i915/intel_panel.c
+++ b/drivers/gpu/drm/i915/intel_panel.c
@@ -325,214 +325,170 @@ out:
 	pipe_config->gmch_pfit.lvds_border_bits = border;
 }
 
-static int is_backlight_combination_mode(struct drm_device *dev)
+static int i915_panel_invert_brightness;
+MODULE_PARM_DESC(invert_brightness, "Invert backlight brightness "
+	"(-1 force normal, 0 machine defaults, 1 force inversion), please "
+	"report PCI device ID, subsystem vendor and subsystem device ID "
+	"to dri-devel@lists.freedesktop.org, if your machine needs it. "
+	"It will then be included in an upcoming module version.");
+module_param_named(invert_brightness, i915_panel_invert_brightness, int, 0600);
+static u32 intel_panel_compute_brightness(struct intel_connector *connector,
+					  u32 val)
 {
+	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_panel *panel = &connector->panel;
 
-	if (IS_GEN4(dev))
-		return I915_READ(BLC_PWM_CTL2) & BLM_COMBINATION_MODE;
+	WARN_ON(panel->backlight.max == 0);
 
-	if (IS_GEN2(dev))
-		return I915_READ(BLC_PWM_CTL) & BLM_LEGACY_MODE;
+	if (i915_panel_invert_brightness < 0)
+		return val;
 
-	return 0;
+	if (i915_panel_invert_brightness > 0 ||
+	    dev_priv->quirks & QUIRK_INVERT_BRIGHTNESS) {
+		return panel->backlight.max - val;
+	}
+
+	return val;
 }
 
-/* XXX: query mode clock or hardware clock and program max PWM appropriately
- * when it's 0.
- */
-static u32 i915_read_blc_pwm_ctl(struct drm_device *dev, enum pipe pipe)
+static u32 bdw_get_backlight(struct intel_connector *connector)
 {
+	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	u32 val;
 
-	WARN_ON_SMP(!spin_is_locked(&dev_priv->backlight.lock));
-
-	/* Restore the CTL value if it lost, e.g. GPU reset */
-
-	if (HAS_PCH_SPLIT(dev_priv->dev)) {
-		val = I915_READ(BLC_PWM_PCH_CTL2);
-		if (dev_priv->regfile.saveBLC_PWM_CTL2 == 0) {
-			dev_priv->regfile.saveBLC_PWM_CTL2 = val;
-		} else if (val == 0) {
-			val = dev_priv->regfile.saveBLC_PWM_CTL2;
-			I915_WRITE(BLC_PWM_PCH_CTL2, val);
-		}
-	} else if (IS_VALLEYVIEW(dev)) {
-		val = I915_READ(VLV_BLC_PWM_CTL(pipe));
-		if (dev_priv->regfile.saveBLC_PWM_CTL == 0) {
-			dev_priv->regfile.saveBLC_PWM_CTL = val;
-			dev_priv->regfile.saveBLC_PWM_CTL2 =
-				I915_READ(VLV_BLC_PWM_CTL2(pipe));
-		} else if (val == 0) {
-			val = dev_priv->regfile.saveBLC_PWM_CTL;
-			I915_WRITE(VLV_BLC_PWM_CTL(pipe), val);
-			I915_WRITE(VLV_BLC_PWM_CTL2(pipe),
-				   dev_priv->regfile.saveBLC_PWM_CTL2);
-		}
+	return I915_READ(BLC_PWM_PCH_CTL2) & BACKLIGHT_DUTY_CYCLE_MASK;
+}
 
-		if (!val)
-			val = 0x0f42ffff;
-	} else {
-		val = I915_READ(BLC_PWM_CTL);
-		if (dev_priv->regfile.saveBLC_PWM_CTL == 0) {
-			dev_priv->regfile.saveBLC_PWM_CTL = val;
-			if (INTEL_INFO(dev)->gen >= 4)
-				dev_priv->regfile.saveBLC_PWM_CTL2 =
-					I915_READ(BLC_PWM_CTL2);
-		} else if (val == 0) {
-			val = dev_priv->regfile.saveBLC_PWM_CTL;
-			I915_WRITE(BLC_PWM_CTL, val);
-			if (INTEL_INFO(dev)->gen >= 4)
-				I915_WRITE(BLC_PWM_CTL2,
-					   dev_priv->regfile.saveBLC_PWM_CTL2);
-		}
-	}
+static u32 pch_get_backlight(struct intel_connector *connector)
+{
+	struct drm_device *dev = connector->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	return val;
+	return I915_READ(BLC_PWM_CPU_CTL) & BACKLIGHT_DUTY_CYCLE_MASK;
 }
 
-static u32 intel_panel_get_max_backlight(struct drm_device *dev,
-					 enum pipe pipe)
+static u32 i9xx_get_backlight(struct intel_connector *connector)
 {
-	u32 max;
+	struct drm_device *dev = connector->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_panel *panel = &connector->panel;
+	u32 val;
 
-	max = i915_read_blc_pwm_ctl(dev, pipe);
+	val = I915_READ(BLC_PWM_CTL) & BACKLIGHT_DUTY_CYCLE_MASK;
+	if (INTEL_INFO(dev)->gen < 4)
+		val >>= 1;
 
-	if (HAS_PCH_SPLIT(dev)) {
-		max >>= 16;
-	} else {
-		if (INTEL_INFO(dev)->gen < 4)
-			max >>= 17;
-		else
-			max >>= 16;
+	if (panel->backlight.combination_mode) {
+		u8 lbpc;
 
-		if (is_backlight_combination_mode(dev))
-			max *= 0xff;
+		pci_read_config_byte(dev->pdev, PCI_LBPC, &lbpc);
+		val *= lbpc;
 	}
 
-	DRM_DEBUG_DRIVER("max backlight PWM = %d\n", max);
-
-	return max;
+	return val;
 }
 
-static int i915_panel_invert_brightness;
-MODULE_PARM_DESC(invert_brightness, "Invert backlight brightness "
-	"(-1 force normal, 0 machine defaults, 1 force inversion), please "
-	"report PCI device ID, subsystem vendor and subsystem device ID "
-	"to dri-devel@lists.freedesktop.org, if your machine needs it. "
-	"It will then be included in an upcoming module version.");
-module_param_named(invert_brightness, i915_panel_invert_brightness, int, 0600);
-static u32 intel_panel_compute_brightness(struct drm_device *dev,
-					  enum pipe pipe, u32 val)
+static u32 _vlv_get_backlight(struct drm_device *dev, enum pipe pipe)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	if (i915_panel_invert_brightness < 0)
-		return val;
+	return I915_READ(VLV_BLC_PWM_CTL(pipe)) & BACKLIGHT_DUTY_CYCLE_MASK;
+}
 
-	if (i915_panel_invert_brightness > 0 ||
-	    dev_priv->quirks & QUIRK_INVERT_BRIGHTNESS) {
-		u32 max = intel_panel_get_max_backlight(dev, pipe);
-		if (max)
-			return max - val;
-	}
+static u32 vlv_get_backlight(struct intel_connector *connector)
+{
+	struct drm_device *dev = connector->base.dev;
+	enum pipe pipe = intel_get_pipe_from_connector(connector);
 
-	return val;
+	return _vlv_get_backlight(dev, pipe);
 }
 
-static u32 intel_panel_get_backlight(struct drm_device *dev,
-				     enum pipe pipe)
+static u32 intel_panel_get_backlight(struct intel_connector *connector)
 {
+	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 val;
 	unsigned long flags;
-	int reg;
-
-	spin_lock_irqsave(&dev_priv->backlight.lock, flags);
-
-	if (IS_BROADWELL(dev)) {
-		val = I915_READ(BLC_PWM_PCH_CTL2) & BACKLIGHT_DUTY_CYCLE_MASK;
-	} else if (HAS_PCH_SPLIT(dev)) {
-		val = I915_READ(BLC_PWM_CPU_CTL) & BACKLIGHT_DUTY_CYCLE_MASK;
-	} else {
-		if (IS_VALLEYVIEW(dev))
-			reg = VLV_BLC_PWM_CTL(pipe);
-		else
-			reg = BLC_PWM_CTL;
-
-		val = I915_READ(reg) & BACKLIGHT_DUTY_CYCLE_MASK;
-		if (INTEL_INFO(dev)->gen < 4)
-			val >>= 1;
-
-		if (is_backlight_combination_mode(dev)) {
-			u8 lbpc;
 
-			pci_read_config_byte(dev->pdev, PCI_LBPC, &lbpc);
-			val *= lbpc;
-		}
-	}
+	spin_lock_irqsave(&dev_priv->backlight_lock, flags);
 
-	val = intel_panel_compute_brightness(dev, pipe, val);
+	val = dev_priv->display.get_backlight(connector);
+	val = intel_panel_compute_brightness(connector, val);
 
-	spin_unlock_irqrestore(&dev_priv->backlight.lock, flags);
+	spin_unlock_irqrestore(&dev_priv->backlight_lock, flags);
 
 	DRM_DEBUG_DRIVER("get backlight PWM = %d\n", val);
 	return val;
 }
 
-static void intel_bdw_panel_set_backlight(struct drm_device *dev, u32 level)
+static void bdw_set_backlight(struct intel_connector *connector, u32 level)
 {
+	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 val = I915_READ(BLC_PWM_PCH_CTL2) & ~BACKLIGHT_DUTY_CYCLE_MASK;
 	I915_WRITE(BLC_PWM_PCH_CTL2, val | level);
 }
 
-static void intel_pch_panel_set_backlight(struct drm_device *dev, u32 level)
+static void pch_set_backlight(struct intel_connector *connector, u32 level)
 {
+	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	u32 val = I915_READ(BLC_PWM_CPU_CTL) & ~BACKLIGHT_DUTY_CYCLE_MASK;
-	I915_WRITE(BLC_PWM_CPU_CTL, val | level);
+	u32 tmp;
+
+	tmp = I915_READ(BLC_PWM_CPU_CTL) & ~BACKLIGHT_DUTY_CYCLE_MASK;
+	I915_WRITE(BLC_PWM_CPU_CTL, tmp | level);
 }
 
-static void intel_panel_actually_set_backlight(struct drm_device *dev,
-					       enum pipe pipe, u32 level)
+static void i9xx_set_backlight(struct intel_connector *connector, u32 level)
 {
+	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	u32 tmp;
-	int reg;
-
-	DRM_DEBUG_DRIVER("set backlight PWM = %d\n", level);
-	level = intel_panel_compute_brightness(dev, pipe, level);
+	struct intel_panel *panel = &connector->panel;
+	u32 tmp, mask;
 
-	if (IS_BROADWELL(dev))
-		return intel_bdw_panel_set_backlight(dev, level);
-	else if (HAS_PCH_SPLIT(dev))
-		return intel_pch_panel_set_backlight(dev, level);
+	WARN_ON(panel->backlight.max == 0);
 
-	if (is_backlight_combination_mode(dev)) {
-		u32 max = intel_panel_get_max_backlight(dev, pipe);
+	if (panel->backlight.combination_mode) {
 		u8 lbpc;
 
-		/* we're screwed, but keep behaviour backwards compatible */
-		if (!max)
-			max = 1;
-
-		lbpc = level * 0xfe / max + 1;
+		lbpc = level * 0xfe / panel->backlight.max + 1;
 		level /= lbpc;
 		pci_write_config_byte(dev->pdev, PCI_LBPC, lbpc);
 	}
 
-	if (IS_VALLEYVIEW(dev))
-		reg = VLV_BLC_PWM_CTL(pipe);
-	else
-		reg = BLC_PWM_CTL;
-
-	tmp = I915_READ(reg);
-	if (INTEL_INFO(dev)->gen < 4)
+	if (IS_GEN4(dev)) {
+		mask = BACKLIGHT_DUTY_CYCLE_MASK;
+	} else {
 		level <<= 1;
-	tmp &= ~BACKLIGHT_DUTY_CYCLE_MASK;
-	I915_WRITE(reg, tmp | level);
+		mask = BACKLIGHT_DUTY_CYCLE_MASK_PNV;
+	}
+
+	tmp = I915_READ(BLC_PWM_CTL) & ~mask;
+	I915_WRITE(BLC_PWM_CTL, tmp | level);
+}
+
+static void vlv_set_backlight(struct intel_connector *connector, u32 level)
+{
+	struct drm_device *dev = connector->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	enum pipe pipe = intel_get_pipe_from_connector(connector);
+	u32 tmp;
+
+	tmp = I915_READ(VLV_BLC_PWM_CTL(pipe)) & ~BACKLIGHT_DUTY_CYCLE_MASK;
+	I915_WRITE(VLV_BLC_PWM_CTL(pipe), tmp | level);
+}
+
+static void
+intel_panel_actually_set_backlight(struct intel_connector *connector, u32 level)
+{
+	struct drm_device *dev = connector->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	DRM_DEBUG_DRIVER("set backlight PWM = %d\n", level);
+
+	level = intel_panel_compute_brightness(connector, level);
+	dev_priv->display.set_backlight(connector, level);
 }
 
 /* set backlight brightness to level in range [0..max] */
@@ -541,45 +497,89 @@ void intel_panel_set_backlight(struct intel_connector *connector, u32 level,
 {
 	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_panel *panel = &connector->panel;
 	enum pipe pipe = intel_get_pipe_from_connector(connector);
 	u32 freq;
 	unsigned long flags;
 
-	if (pipe == INVALID_PIPE)
+	if (!panel->backlight.present || pipe == INVALID_PIPE)
 		return;
 
-	spin_lock_irqsave(&dev_priv->backlight.lock, flags);
+	spin_lock_irqsave(&dev_priv->backlight_lock, flags);
 
-	freq = intel_panel_get_max_backlight(dev, pipe);
-	if (!freq) {
-		/* we are screwed, bail out */
-		goto out;
-	}
+	WARN_ON(panel->backlight.max == 0);
 
-	/* scale to hardware, but be careful to not overflow */
+	/* scale to hardware max, but be careful to not overflow */
+	freq = panel->backlight.max;
 	if (freq < max)
 		level = level * freq / max;
 	else
 		level = freq / max * level;
 
-	dev_priv->backlight.level = level;
-	if (dev_priv->backlight.device)
-		dev_priv->backlight.device->props.brightness = level;
+	panel->backlight.level = level;
+	if (panel->backlight.device)
+		panel->backlight.device->props.brightness = level;
 
-	if (dev_priv->backlight.enabled)
-		intel_panel_actually_set_backlight(dev, pipe, level);
-out:
-	spin_unlock_irqrestore(&dev_priv->backlight.lock, flags);
+	if (panel->backlight.enabled)
+		intel_panel_actually_set_backlight(connector, level);
+
+	spin_unlock_irqrestore(&dev_priv->backlight_lock, flags);
+}
+
+static void pch_disable_backlight(struct intel_connector *connector)
+{
+	struct drm_device *dev = connector->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 tmp;
+
+	intel_panel_actually_set_backlight(connector, 0);
+
+	tmp = I915_READ(BLC_PWM_CPU_CTL2);
+	I915_WRITE(BLC_PWM_CPU_CTL2, tmp & ~BLM_PWM_ENABLE);
+
+	tmp = I915_READ(BLC_PWM_PCH_CTL1);
+	I915_WRITE(BLC_PWM_PCH_CTL1, tmp & ~BLM_PCH_PWM_ENABLE);
+}
+
+static void i9xx_disable_backlight(struct intel_connector *connector)
+{
+	intel_panel_actually_set_backlight(connector, 0);
+}
+
+static void i965_disable_backlight(struct intel_connector *connector)
+{
+	struct drm_device *dev = connector->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	u32 tmp;
+
+	intel_panel_actually_set_backlight(connector, 0);
+
+	tmp = I915_READ(BLC_PWM_CTL2);
+	I915_WRITE(BLC_PWM_CTL2, tmp & ~BLM_PWM_ENABLE);
+}
+
+static void vlv_disable_backlight(struct intel_connector *connector)
+{
+	struct drm_device *dev = connector->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	enum pipe pipe = intel_get_pipe_from_connector(connector);
+	u32 tmp;
+
+	intel_panel_actually_set_backlight(connector, 0);
+
+	tmp = I915_READ(VLV_BLC_PWM_CTL2(pipe));
+	I915_WRITE(VLV_BLC_PWM_CTL2(pipe), tmp & ~BLM_PWM_ENABLE);
 }
 
 void intel_panel_disable_backlight(struct intel_connector *connector)
 {
 	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_panel *panel = &connector->panel;
 	enum pipe pipe = intel_get_pipe_from_connector(connector);
 	unsigned long flags;
 
-	if (pipe == INVALID_PIPE)
+	if (!panel->backlight.present || pipe == INVALID_PIPE)
 		return;
 
 	/*
@@ -593,150 +593,215 @@ void intel_panel_disable_backlight(struct intel_connector *connector)
 		return;
 	}
 
-	spin_lock_irqsave(&dev_priv->backlight.lock, flags);
+	spin_lock_irqsave(&dev_priv->backlight_lock, flags);
 
-	dev_priv->backlight.enabled = false;
-	intel_panel_actually_set_backlight(dev, pipe, 0);
+	panel->backlight.enabled = false;
+	dev_priv->display.disable_backlight(connector);
 
-	if (INTEL_INFO(dev)->gen >= 4) {
-		uint32_t reg, tmp;
+	spin_unlock_irqrestore(&dev_priv->backlight_lock, flags);
+}
 
-		if (HAS_PCH_SPLIT(dev))
-			reg = BLC_PWM_CPU_CTL2;
-		else if (IS_VALLEYVIEW(dev))
-			reg = VLV_BLC_PWM_CTL2(pipe);
-		else
-			reg = BLC_PWM_CTL2;
+static void bdw_enable_backlight(struct intel_connector *connector)
+{
+	struct drm_device *dev = connector->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_panel *panel = &connector->panel;
+	u32 pch_ctl1, pch_ctl2;
+
+	pch_ctl1 = I915_READ(BLC_PWM_PCH_CTL1);
+	if (pch_ctl1 & BLM_PCH_PWM_ENABLE) {
+		DRM_DEBUG_KMS("pch backlight already enabled\n");
+		pch_ctl1 &= ~BLM_PCH_PWM_ENABLE;
+		I915_WRITE(BLC_PWM_PCH_CTL1, pch_ctl1);
+	}
 
-		I915_WRITE(reg, I915_READ(reg) & ~BLM_PWM_ENABLE);
+	pch_ctl2 = panel->backlight.max << 16;
+	I915_WRITE(BLC_PWM_PCH_CTL2, pch_ctl2);
 
-		if (HAS_PCH_SPLIT(dev)) {
-			tmp = I915_READ(BLC_PWM_PCH_CTL1);
-			tmp &= ~BLM_PCH_PWM_ENABLE;
-			I915_WRITE(BLC_PWM_PCH_CTL1, tmp);
-		}
-	}
+	pch_ctl1 = 0;
+	if (panel->backlight.active_low_pwm)
+		pch_ctl1 |= BLM_PCH_POLARITY;
 
-	spin_unlock_irqrestore(&dev_priv->backlight.lock, flags);
+	/* BDW always uses the pch pwm controls. */
+	pch_ctl1 |= BLM_PCH_OVERRIDE_ENABLE;
+
+	I915_WRITE(BLC_PWM_PCH_CTL1, pch_ctl1);
+	POSTING_READ(BLC_PWM_PCH_CTL1);
+	I915_WRITE(BLC_PWM_PCH_CTL1, pch_ctl1 | BLM_PCH_PWM_ENABLE);
+
+	/* This won't stick until the above enable. */
+	intel_panel_actually_set_backlight(connector, panel->backlight.level);
 }
 
-void intel_panel_enable_backlight(struct intel_connector *connector)
+static void pch_enable_backlight(struct intel_connector *connector)
 {
 	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_panel *panel = &connector->panel;
 	enum pipe pipe = intel_get_pipe_from_connector(connector);
 	enum transcoder cpu_transcoder =
 		intel_pipe_to_cpu_transcoder(dev_priv, pipe);
-	unsigned long flags;
+	u32 cpu_ctl2, pch_ctl1, pch_ctl2;
 
-	if (pipe == INVALID_PIPE)
-		return;
+	cpu_ctl2 = I915_READ(BLC_PWM_CPU_CTL2);
+	if (cpu_ctl2 & BLM_PWM_ENABLE) {
+		WARN(1, "cpu backlight already enabled\n");
+		cpu_ctl2 &= ~BLM_PWM_ENABLE;
+		I915_WRITE(BLC_PWM_CPU_CTL2, cpu_ctl2);
+	}
 
-	DRM_DEBUG_KMS("pipe %c\n", pipe_name(pipe));
+	pch_ctl1 = I915_READ(BLC_PWM_PCH_CTL1);
+	if (pch_ctl1 & BLM_PCH_PWM_ENABLE) {
+		DRM_DEBUG_KMS("pch backlight already enabled\n");
+		pch_ctl1 &= ~BLM_PCH_PWM_ENABLE;
+		I915_WRITE(BLC_PWM_PCH_CTL1, pch_ctl1);
+	}
+
+	if (cpu_transcoder == TRANSCODER_EDP)
+		cpu_ctl2 = BLM_TRANSCODER_EDP;
+	else
+		cpu_ctl2 = BLM_PIPE(cpu_transcoder);
+	I915_WRITE(BLC_PWM_CPU_CTL2, cpu_ctl2);
+	POSTING_READ(BLC_PWM_CPU_CTL2);
+	I915_WRITE(BLC_PWM_CPU_CTL2, cpu_ctl2 | BLM_PWM_ENABLE);
+
+	/* This won't stick until the above enable. */
+	intel_panel_actually_set_backlight(connector, panel->backlight.level);
+
+	pch_ctl2 = panel->backlight.max << 16;
+	I915_WRITE(BLC_PWM_PCH_CTL2, pch_ctl2);
+
+	pch_ctl1 = 0;
+	if (panel->backlight.active_low_pwm)
+		pch_ctl1 |= BLM_PCH_POLARITY;
+
+	I915_WRITE(BLC_PWM_PCH_CTL1, pch_ctl1);
+	POSTING_READ(BLC_PWM_PCH_CTL1);
+	I915_WRITE(BLC_PWM_PCH_CTL1, pch_ctl1 | BLM_PCH_PWM_ENABLE);
+}
 
-	spin_lock_irqsave(&dev_priv->backlight.lock, flags);
+static void i9xx_enable_backlight(struct intel_connector *connector)
+{
+	struct drm_device *dev = connector->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_panel *panel = &connector->panel;
+	u32 ctl, freq;
 
-	if (dev_priv->backlight.level == 0) {
-		dev_priv->backlight.level = intel_panel_get_max_backlight(dev,
-									  pipe);
-		if (dev_priv->backlight.device)
-			dev_priv->backlight.device->props.brightness =
-				dev_priv->backlight.level;
+	ctl = I915_READ(BLC_PWM_CTL);
+	if (ctl & BACKLIGHT_DUTY_CYCLE_MASK_PNV) {
+		WARN(1, "backlight already enabled\n");
+		I915_WRITE(BLC_PWM_CTL, 0);
 	}
 
-	if (INTEL_INFO(dev)->gen >= 4) {
-		uint32_t reg, tmp;
+	freq = panel->backlight.max;
+	if (panel->backlight.combination_mode)
+		freq /= 0xff;
 
-		if (HAS_PCH_SPLIT(dev))
-			reg = BLC_PWM_CPU_CTL2;
-		else if (IS_VALLEYVIEW(dev))
-			reg = VLV_BLC_PWM_CTL2(pipe);
-		else
-			reg = BLC_PWM_CTL2;
+	ctl = freq << 17;
+	if (IS_GEN2(dev) && panel->backlight.combination_mode)
+		ctl |= BLM_LEGACY_MODE;
+	if (IS_PINEVIEW(dev) && panel->backlight.active_low_pwm)
+		ctl |= BLM_POLARITY_PNV;
 
-		tmp = I915_READ(reg);
+	I915_WRITE(BLC_PWM_CTL, ctl);
+	POSTING_READ(BLC_PWM_CTL);
 
-		/* Note that this can also get called through dpms changes. And
-		 * we don't track the backlight dpms state, hence check whether
-		 * we have to do anything first. */
-		if (tmp & BLM_PWM_ENABLE)
-			goto set_level;
+	/* XXX: combine this into above write? */
+	intel_panel_actually_set_backlight(connector, panel->backlight.level);
+}
 
-		if (INTEL_INFO(dev)->num_pipes == 3)
-			tmp &= ~BLM_PIPE_SELECT_IVB;
-		else
-			tmp &= ~BLM_PIPE_SELECT;
+static void i965_enable_backlight(struct intel_connector *connector)
+{
+	struct drm_device *dev = connector->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_panel *panel = &connector->panel;
+	enum pipe pipe = intel_get_pipe_from_connector(connector);
+	u32 ctl, ctl2, freq;
 
-		if (cpu_transcoder == TRANSCODER_EDP)
-			tmp |= BLM_TRANSCODER_EDP;
-		else
-			tmp |= BLM_PIPE(cpu_transcoder);
-		tmp &= ~BLM_PWM_ENABLE;
-
-		I915_WRITE(reg, tmp);
-		POSTING_READ(reg);
-		I915_WRITE(reg, tmp | BLM_PWM_ENABLE);
-
-		if (IS_BROADWELL(dev)) {
-			/*
-			 * Broadwell requires PCH override to drive the PCH
-			 * backlight pin. The above will configure the CPU
-			 * backlight pin, which we don't plan to use.
-			 */
-			tmp = I915_READ(BLC_PWM_PCH_CTL1);
-			tmp |= BLM_PCH_OVERRIDE_ENABLE | BLM_PCH_PWM_ENABLE;
-			I915_WRITE(BLC_PWM_PCH_CTL1, tmp);
-		} else if (HAS_PCH_SPLIT(dev) &&
-		    !(dev_priv->quirks & QUIRK_NO_PCH_PWM_ENABLE)) {
-			tmp = I915_READ(BLC_PWM_PCH_CTL1);
-			tmp |= BLM_PCH_PWM_ENABLE;
-			tmp &= ~BLM_PCH_OVERRIDE_ENABLE;
-			I915_WRITE(BLC_PWM_PCH_CTL1, tmp);
-		}
+	ctl2 = I915_READ(BLC_PWM_CTL2);
+	if (ctl2 & BLM_PWM_ENABLE) {
+		WARN(1, "backlight already enabled\n");
+		ctl2 &= ~BLM_PWM_ENABLE;
+		I915_WRITE(BLC_PWM_CTL2, ctl2);
 	}
 
-set_level:
-	/* Call below after setting BLC_PWM_CPU_CTL2 and BLC_PWM_PCH_CTL1.
-	 * BLC_PWM_CPU_CTL may be cleared to zero automatically when these
-	 * registers are set.
-	 */
-	dev_priv->backlight.enabled = true;
-	intel_panel_actually_set_backlight(dev, pipe,
-					   dev_priv->backlight.level);
+	freq = panel->backlight.max;
+	if (panel->backlight.combination_mode)
+		freq /= 0xff;
+
+	ctl = freq << 16;
+	I915_WRITE(BLC_PWM_CTL, ctl);
+
+	/* XXX: combine this into above write? */
+	intel_panel_actually_set_backlight(connector, panel->backlight.level);
 
-	spin_unlock_irqrestore(&dev_priv->backlight.lock, flags);
+	ctl2 = BLM_PIPE(pipe);
+	if (panel->backlight.combination_mode)
+		ctl2 |= BLM_COMBINATION_MODE;
+	if (panel->backlight.active_low_pwm)
+		ctl2 |= BLM_POLARITY_I965;
+	I915_WRITE(BLC_PWM_CTL2, ctl2);
+	POSTING_READ(BLC_PWM_CTL2);
+	I915_WRITE(BLC_PWM_CTL2, ctl2 | BLM_PWM_ENABLE);
 }
 
-/* FIXME: use VBT vals to init PWM_CTL and PWM_CTL2 correctly */
-static void intel_panel_init_backlight_regs(struct drm_device *dev)
+static void vlv_enable_backlight(struct intel_connector *connector)
 {
+	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_panel *panel = &connector->panel;
+	enum pipe pipe = intel_get_pipe_from_connector(connector);
+	u32 ctl, ctl2;
 
-	if (IS_VALLEYVIEW(dev)) {
-		enum pipe pipe;
+	ctl2 = I915_READ(VLV_BLC_PWM_CTL2(pipe));
+	if (ctl2 & BLM_PWM_ENABLE) {
+		WARN(1, "backlight already enabled\n");
+		ctl2 &= ~BLM_PWM_ENABLE;
+		I915_WRITE(VLV_BLC_PWM_CTL2(pipe), ctl2);
+	}
 
-		for_each_pipe(pipe) {
-			u32 cur_val = I915_READ(VLV_BLC_PWM_CTL(pipe));
+	ctl = panel->backlight.max << 16;
+	I915_WRITE(VLV_BLC_PWM_CTL(pipe), ctl);
 
-			/* Skip if the modulation freq is already set */
-			if (cur_val & ~BACKLIGHT_DUTY_CYCLE_MASK)
-				continue;
+	/* XXX: combine this into above write? */
+	intel_panel_actually_set_backlight(connector, panel->backlight.level);
 
-			cur_val &= BACKLIGHT_DUTY_CYCLE_MASK;
-			I915_WRITE(VLV_BLC_PWM_CTL(pipe), (0xf42 << 16) |
-				   cur_val);
-		}
-	}
+	ctl2 = 0;
+	if (panel->backlight.active_low_pwm)
+		ctl2 |= BLM_POLARITY_I965;
+	I915_WRITE(VLV_BLC_PWM_CTL2(pipe), ctl2);
+	POSTING_READ(VLV_BLC_PWM_CTL2(pipe));
+	I915_WRITE(VLV_BLC_PWM_CTL2(pipe), ctl2 | BLM_PWM_ENABLE);
 }
 
-static void intel_panel_init_backlight(struct drm_device *dev)
+void intel_panel_enable_backlight(struct intel_connector *connector)
 {
+	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_panel *panel = &connector->panel;
+	enum pipe pipe = intel_get_pipe_from_connector(connector);
+	unsigned long flags;
+
+	if (!panel->backlight.present || pipe == INVALID_PIPE)
+		return;
+
+	DRM_DEBUG_KMS("pipe %c\n", pipe_name(pipe));
+
+	spin_lock_irqsave(&dev_priv->backlight_lock, flags);
+
+	WARN_ON(panel->backlight.max == 0);
 
-	intel_panel_init_backlight_regs(dev);
+	if (panel->backlight.level == 0) {
+		panel->backlight.level = panel->backlight.max;
+		if (panel->backlight.device)
+			panel->backlight.device->props.brightness =
+				panel->backlight.level;
+	}
+
+	dev_priv->display.enable_backlight(connector);
+	panel->backlight.enabled = true;
 
-	dev_priv->backlight.level = intel_panel_get_backlight(dev, 0);
-	dev_priv->backlight.enabled = dev_priv->backlight.level != 0;
+	spin_unlock_irqrestore(&dev_priv->backlight_lock, flags);
 }
 
 enum drm_connector_status
@@ -762,7 +827,7 @@ intel_panel_detect(struct drm_device *dev)
 }
 
 #if IS_ENABLED(CONFIG_BACKLIGHT_CLASS_DEVICE)
-static int intel_panel_update_status(struct backlight_device *bd)
+static int intel_backlight_device_update_status(struct backlight_device *bd)
 {
 	struct intel_connector *connector = bl_get_data(bd);
 	struct drm_device *dev = connector->base.dev;
@@ -776,85 +841,362 @@ static int intel_panel_update_status(struct backlight_device *bd)
 	return 0;
 }
 
-static int intel_panel_get_brightness(struct backlight_device *bd)
+static int intel_backlight_device_get_brightness(struct backlight_device *bd)
 {
 	struct intel_connector *connector = bl_get_data(bd);
 	struct drm_device *dev = connector->base.dev;
-	enum pipe pipe;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int ret;
 
+	intel_runtime_pm_get(dev_priv);
 	mutex_lock(&dev->mode_config.mutex);
-	pipe = intel_get_pipe_from_connector(connector);
+	ret = intel_panel_get_backlight(connector);
 	mutex_unlock(&dev->mode_config.mutex);
-	if (pipe == INVALID_PIPE)
-		return 0;
+	intel_runtime_pm_put(dev_priv);
 
-	return intel_panel_get_backlight(connector->base.dev, pipe);
+	return ret;
 }
 
-static const struct backlight_ops intel_panel_bl_ops = {
-	.update_status = intel_panel_update_status,
-	.get_brightness = intel_panel_get_brightness,
+static const struct backlight_ops intel_backlight_device_ops = {
+	.update_status = intel_backlight_device_update_status,
+	.get_brightness = intel_backlight_device_get_brightness,
 };
 
-int intel_panel_setup_backlight(struct drm_connector *connector)
+static int intel_backlight_device_register(struct intel_connector *connector)
 {
-	struct drm_device *dev = connector->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_panel *panel = &connector->panel;
 	struct backlight_properties props;
-	unsigned long flags;
 
-	intel_panel_init_backlight(dev);
-
-	if (WARN_ON(dev_priv->backlight.device))
+	if (WARN_ON(panel->backlight.device))
 		return -ENODEV;
 
+	BUG_ON(panel->backlight.max == 0);
+
 	memset(&props, 0, sizeof(props));
 	props.type = BACKLIGHT_RAW;
-	props.brightness = dev_priv->backlight.level;
+	props.brightness = panel->backlight.level;
+	props.max_brightness = panel->backlight.max;
 
-	spin_lock_irqsave(&dev_priv->backlight.lock, flags);
-	props.max_brightness = intel_panel_get_max_backlight(dev, 0);
-	spin_unlock_irqrestore(&dev_priv->backlight.lock, flags);
-
-	if (props.max_brightness == 0) {
-		DRM_DEBUG_DRIVER("Failed to get maximum backlight value\n");
-		return -ENODEV;
-	}
-	dev_priv->backlight.device =
+	/*
+	 * Note: using the same name independent of the connector prevents
+	 * registration of multiple backlight devices in the driver.
+	 */
+	panel->backlight.device =
 		backlight_device_register("intel_backlight",
-					  connector->kdev,
-					  to_intel_connector(connector),
-					  &intel_panel_bl_ops, &props);
+					  connector->base.kdev,
+					  connector,
+					  &intel_backlight_device_ops, &props);
 
-	if (IS_ERR(dev_priv->backlight.device)) {
+	if (IS_ERR(panel->backlight.device)) {
 		DRM_ERROR("Failed to register backlight: %ld\n",
-			  PTR_ERR(dev_priv->backlight.device));
-		dev_priv->backlight.device = NULL;
+			  PTR_ERR(panel->backlight.device));
+		panel->backlight.device = NULL;
 		return -ENODEV;
 	}
 	return 0;
 }
 
-void intel_panel_destroy_backlight(struct drm_device *dev)
+static void intel_backlight_device_unregister(struct intel_connector *connector)
+{
+	struct intel_panel *panel = &connector->panel;
+
+	if (panel->backlight.device) {
+		backlight_device_unregister(panel->backlight.device);
+		panel->backlight.device = NULL;
+	}
+}
+#else /* CONFIG_BACKLIGHT_CLASS_DEVICE */
+static int intel_backlight_device_register(struct intel_connector *connector)
+{
+	return 0;
+}
+static void intel_backlight_device_unregister(struct intel_connector *connector)
+{
+}
+#endif /* CONFIG_BACKLIGHT_CLASS_DEVICE */
+
+/*
+ * Note: The setup hooks can't assume pipe is set!
+ *
+ * XXX: Query mode clock or hardware clock and program PWM modulation frequency
+ * appropriately when it's 0. Use VBT and/or sane defaults.
+ */
+static int bdw_setup_backlight(struct intel_connector *connector)
+{
+	struct drm_device *dev = connector->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_panel *panel = &connector->panel;
+	u32 pch_ctl1, pch_ctl2, val;
+
+	pch_ctl1 = I915_READ(BLC_PWM_PCH_CTL1);
+	panel->backlight.active_low_pwm = pch_ctl1 & BLM_PCH_POLARITY;
+
+	pch_ctl2 = I915_READ(BLC_PWM_PCH_CTL2);
+	panel->backlight.max = pch_ctl2 >> 16;
+	if (!panel->backlight.max)
+		return -ENODEV;
+
+	val = bdw_get_backlight(connector);
+	panel->backlight.level = intel_panel_compute_brightness(connector, val);
+
+	panel->backlight.enabled = (pch_ctl1 & BLM_PCH_PWM_ENABLE) &&
+		panel->backlight.level != 0;
+
+	return 0;
+}
+
+static int pch_setup_backlight(struct intel_connector *connector)
+{
+	struct drm_device *dev = connector->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_panel *panel = &connector->panel;
+	u32 cpu_ctl2, pch_ctl1, pch_ctl2, val;
+
+	pch_ctl1 = I915_READ(BLC_PWM_PCH_CTL1);
+	panel->backlight.active_low_pwm = pch_ctl1 & BLM_PCH_POLARITY;
+
+	pch_ctl2 = I915_READ(BLC_PWM_PCH_CTL2);
+	panel->backlight.max = pch_ctl2 >> 16;
+	if (!panel->backlight.max)
+		return -ENODEV;
+
+	val = pch_get_backlight(connector);
+	panel->backlight.level = intel_panel_compute_brightness(connector, val);
+
+	cpu_ctl2 = I915_READ(BLC_PWM_CPU_CTL2);
+	panel->backlight.enabled = (cpu_ctl2 & BLM_PWM_ENABLE) &&
+		(pch_ctl1 & BLM_PCH_PWM_ENABLE) && panel->backlight.level != 0;
+
+	return 0;
+}
+
+static int i9xx_setup_backlight(struct intel_connector *connector)
+{
+	struct drm_device *dev = connector->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_panel *panel = &connector->panel;
+	u32 ctl, val;
+
+	ctl = I915_READ(BLC_PWM_CTL);
+
+	if (IS_GEN2(dev))
+		panel->backlight.combination_mode = ctl & BLM_LEGACY_MODE;
+
+	if (IS_PINEVIEW(dev))
+		panel->backlight.active_low_pwm = ctl & BLM_POLARITY_PNV;
+
+	panel->backlight.max = ctl >> 17;
+	if (panel->backlight.combination_mode)
+		panel->backlight.max *= 0xff;
+
+	if (!panel->backlight.max)
+		return -ENODEV;
+
+	val = i9xx_get_backlight(connector);
+	panel->backlight.level = intel_panel_compute_brightness(connector, val);
+
+	panel->backlight.enabled = panel->backlight.level != 0;
+
+	return 0;
+}
+
+static int i965_setup_backlight(struct intel_connector *connector)
+{
+	struct drm_device *dev = connector->base.dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_panel *panel = &connector->panel;
+	u32 ctl, ctl2, val;
+
+	ctl2 = I915_READ(BLC_PWM_CTL2);
+	panel->backlight.combination_mode = ctl2 & BLM_COMBINATION_MODE;
+	panel->backlight.active_low_pwm = ctl2 & BLM_POLARITY_I965;
+
+	ctl = I915_READ(BLC_PWM_CTL);
+	panel->backlight.max = ctl >> 16;
+	if (panel->backlight.combination_mode)
+		panel->backlight.max *= 0xff;
+
+	if (!panel->backlight.max)
+		return -ENODEV;
+
+	val = i9xx_get_backlight(connector);
+	panel->backlight.level = intel_panel_compute_brightness(connector, val);
+
+	panel->backlight.enabled = (ctl2 & BLM_PWM_ENABLE) &&
+		panel->backlight.level != 0;
+
+	return 0;
+}
+
+static int vlv_setup_backlight(struct intel_connector *connector)
 {
+	struct drm_device *dev = connector->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	if (dev_priv->backlight.device) {
-		backlight_device_unregister(dev_priv->backlight.device);
-		dev_priv->backlight.device = NULL;
+	struct intel_panel *panel = &connector->panel;
+	enum pipe pipe;
+	u32 ctl, ctl2, val;
+
+	for_each_pipe(pipe) {
+		u32 cur_val = I915_READ(VLV_BLC_PWM_CTL(pipe));
+
+		/* Skip if the modulation freq is already set */
+		if (cur_val & ~BACKLIGHT_DUTY_CYCLE_MASK)
+			continue;
+
+		cur_val &= BACKLIGHT_DUTY_CYCLE_MASK;
+		I915_WRITE(VLV_BLC_PWM_CTL(pipe), (0xf42 << 16) |
+			   cur_val);
 	}
+
+	ctl2 = I915_READ(VLV_BLC_PWM_CTL2(PIPE_A));
+	panel->backlight.active_low_pwm = ctl2 & BLM_POLARITY_I965;
+
+	ctl = I915_READ(VLV_BLC_PWM_CTL(PIPE_A));
+	panel->backlight.max = ctl >> 16;
+	if (!panel->backlight.max)
+		return -ENODEV;
+
+	val = _vlv_get_backlight(dev, PIPE_A);
+	panel->backlight.level = intel_panel_compute_brightness(connector, val);
+
+	panel->backlight.enabled = (ctl2 & BLM_PWM_ENABLE) &&
+		panel->backlight.level != 0;
+
+	return 0;
 }
-#else
+
 int intel_panel_setup_backlight(struct drm_connector *connector)
 {
-	intel_panel_init_backlight(connector->dev);
+	struct drm_device *dev = connector->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_connector *intel_connector = to_intel_connector(connector);
+	struct intel_panel *panel = &intel_connector->panel;
+	unsigned long flags;
+	int ret;
+
+	/* set level and max in panel struct */
+	spin_lock_irqsave(&dev_priv->backlight_lock, flags);
+	ret = dev_priv->display.setup_backlight(intel_connector);
+	spin_unlock_irqrestore(&dev_priv->backlight_lock, flags);
+
+	if (ret) {
+		DRM_DEBUG_KMS("failed to setup backlight for connector %s\n",
+			      drm_get_connector_name(connector));
+		return ret;
+	}
+
+	intel_backlight_device_register(intel_connector);
+
+	panel->backlight.present = true;
+
+	DRM_DEBUG_KMS("backlight initialized, %s, brightness %u/%u, "
+		      "sysfs interface %sregistered\n",
+		      panel->backlight.enabled ? "enabled" : "disabled",
+		      panel->backlight.level, panel->backlight.max,
+		      panel->backlight.device ? "" : "not ");
+
 	return 0;
 }
 
-void intel_panel_destroy_backlight(struct drm_device *dev)
+void intel_panel_destroy_backlight(struct drm_connector *connector)
+{
+	struct intel_connector *intel_connector = to_intel_connector(connector);
+	struct intel_panel *panel = &intel_connector->panel;
+
+	panel->backlight.present = false;
+	intel_backlight_device_unregister(intel_connector);
+}
+
+/**
+ * intel_find_panel_downclock - find the reduced downclock for LVDS in EDID
+ * @dev: drm device
+ * @fixed_mode : panel native mode
+ * @connector: LVDS/eDP connector
+ *
+ * Return downclock_avail
+ * Find the reduced downclock for LVDS/eDP in EDID.
+ */
+struct drm_display_mode *
+intel_find_panel_downclock(struct drm_device *dev,
+			struct drm_display_mode *fixed_mode,
+			struct drm_connector *connector)
+{
+	struct drm_display_mode *scan, *tmp_mode;
+	int temp_downclock;
+
+	temp_downclock = fixed_mode->clock;
+	tmp_mode = NULL;
+
+	list_for_each_entry(scan, &connector->probed_modes, head) {
+		/*
+		 * If one mode has the same resolution with the fixed_panel
+		 * mode while they have the different refresh rate, it means
+		 * that the reduced downclock is found. In such
+		 * case we can set the different FPx0/1 to dynamically select
+		 * between low and high frequency.
+		 */
+		if (scan->hdisplay == fixed_mode->hdisplay &&
+		    scan->hsync_start == fixed_mode->hsync_start &&
+		    scan->hsync_end == fixed_mode->hsync_end &&
+		    scan->htotal == fixed_mode->htotal &&
+		    scan->vdisplay == fixed_mode->vdisplay &&
+		    scan->vsync_start == fixed_mode->vsync_start &&
+		    scan->vsync_end == fixed_mode->vsync_end &&
+		    scan->vtotal == fixed_mode->vtotal) {
+			if (scan->clock < temp_downclock) {
+				/*
+				 * The downclock is already found. But we
+				 * expect to find the lower downclock.
+				 */
+				temp_downclock = scan->clock;
+				tmp_mode = scan;
+			}
+		}
+	}
+
+	if (temp_downclock < fixed_mode->clock)
+		return drm_mode_duplicate(dev, tmp_mode);
+	else
+		return NULL;
+}
+
+/* Set up chip specific backlight functions */
+void intel_panel_init_backlight_funcs(struct drm_device *dev)
 {
-	return;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	if (IS_BROADWELL(dev)) {
+		dev_priv->display.setup_backlight = bdw_setup_backlight;
+		dev_priv->display.enable_backlight = bdw_enable_backlight;
+		dev_priv->display.disable_backlight = pch_disable_backlight;
+		dev_priv->display.set_backlight = bdw_set_backlight;
+		dev_priv->display.get_backlight = bdw_get_backlight;
+	} else if (HAS_PCH_SPLIT(dev)) {
+		dev_priv->display.setup_backlight = pch_setup_backlight;
+		dev_priv->display.enable_backlight = pch_enable_backlight;
+		dev_priv->display.disable_backlight = pch_disable_backlight;
+		dev_priv->display.set_backlight = pch_set_backlight;
+		dev_priv->display.get_backlight = pch_get_backlight;
+	} else if (IS_VALLEYVIEW(dev)) {
+		dev_priv->display.setup_backlight = vlv_setup_backlight;
+		dev_priv->display.enable_backlight = vlv_enable_backlight;
+		dev_priv->display.disable_backlight = vlv_disable_backlight;
+		dev_priv->display.set_backlight = vlv_set_backlight;
+		dev_priv->display.get_backlight = vlv_get_backlight;
+	} else if (IS_GEN4(dev)) {
+		dev_priv->display.setup_backlight = i965_setup_backlight;
+		dev_priv->display.enable_backlight = i965_enable_backlight;
+		dev_priv->display.disable_backlight = i965_disable_backlight;
+		dev_priv->display.set_backlight = i9xx_set_backlight;
+		dev_priv->display.get_backlight = i9xx_get_backlight;
+	} else {
+		dev_priv->display.setup_backlight = i9xx_setup_backlight;
+		dev_priv->display.enable_backlight = i9xx_enable_backlight;
+		dev_priv->display.disable_backlight = i9xx_disable_backlight;
+		dev_priv->display.set_backlight = i9xx_set_backlight;
+		dev_priv->display.get_backlight = i9xx_get_backlight;
+	}
 }
-#endif
 
 int intel_panel_init(struct intel_panel *panel,
 		     struct drm_display_mode *fixed_mode)
@@ -871,4 +1213,8 @@ void intel_panel_fini(struct intel_panel *panel)
 
 	if (panel->fixed_mode)
 		drm_mode_destroy(intel_connector->base.dev, panel->fixed_mode);
+
+	if (panel->downclock_mode)
+		drm_mode_destroy(intel_connector->base.dev,
+				panel->downclock_mode);
 }
diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
index 26c29c173221..d77cc81900f9 100644
--- a/drivers/gpu/drm/i915/intel_pm.c
+++ b/drivers/gpu/drm/i915/intel_pm.c
@@ -30,7 +30,9 @@
 #include "intel_drv.h"
 #include "../../../platform/x86/intel_ips.h"
 #include <linux/module.h>
+#include <linux/vgaarb.h>
 #include <drm/i915_powerwell.h>
+#include <linux/pm_runtime.h>
 
 /**
  * RC6 is a special power stage which allows the GPU to enter an very
@@ -86,7 +88,7 @@ static void i8xx_disable_fbc(struct drm_device *dev)
 	DRM_DEBUG_KMS("disabled FBC\n");
 }
 
-static void i8xx_enable_fbc(struct drm_crtc *crtc, unsigned long interval)
+static void i8xx_enable_fbc(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -96,32 +98,40 @@ static void i8xx_enable_fbc(struct drm_crtc *crtc, unsigned long interval)
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	int cfb_pitch;
 	int plane, i;
-	u32 fbc_ctl, fbc_ctl2;
+	u32 fbc_ctl;
 
 	cfb_pitch = dev_priv->fbc.size / FBC_LL_SIZE;
 	if (fb->pitches[0] < cfb_pitch)
 		cfb_pitch = fb->pitches[0];
 
-	/* FBC_CTL wants 64B units */
-	cfb_pitch = (cfb_pitch / 64) - 1;
+	/* FBC_CTL wants 32B or 64B units */
+	if (IS_GEN2(dev))
+		cfb_pitch = (cfb_pitch / 32) - 1;
+	else
+		cfb_pitch = (cfb_pitch / 64) - 1;
 	plane = intel_crtc->plane == 0 ? FBC_CTL_PLANEA : FBC_CTL_PLANEB;
 
 	/* Clear old tags */
 	for (i = 0; i < (FBC_LL_SIZE / 32) + 1; i++)
 		I915_WRITE(FBC_TAG + (i * 4), 0);
 
-	/* Set it up... */
-	fbc_ctl2 = FBC_CTL_FENCE_DBL | FBC_CTL_IDLE_IMM | FBC_CTL_CPU_FENCE;
-	fbc_ctl2 |= plane;
-	I915_WRITE(FBC_CONTROL2, fbc_ctl2);
-	I915_WRITE(FBC_FENCE_OFF, crtc->y);
+	if (IS_GEN4(dev)) {
+		u32 fbc_ctl2;
+
+		/* Set it up... */
+		fbc_ctl2 = FBC_CTL_FENCE_DBL | FBC_CTL_IDLE_IMM | FBC_CTL_CPU_FENCE;
+		fbc_ctl2 |= plane;
+		I915_WRITE(FBC_CONTROL2, fbc_ctl2);
+		I915_WRITE(FBC_FENCE_OFF, crtc->y);
+	}
 
 	/* enable it... */
-	fbc_ctl = FBC_CTL_EN | FBC_CTL_PERIODIC;
+	fbc_ctl = I915_READ(FBC_CONTROL);
+	fbc_ctl &= 0x3fff << FBC_CTL_INTERVAL_SHIFT;
+	fbc_ctl |= FBC_CTL_EN | FBC_CTL_PERIODIC;
 	if (IS_I945GM(dev))
 		fbc_ctl |= FBC_CTL_C3_IDLE; /* 945 needs special SR handling */
 	fbc_ctl |= (cfb_pitch & 0xff) << FBC_CTL_STRIDE_SHIFT;
-	fbc_ctl |= (interval & 0x2fff) << FBC_CTL_INTERVAL_SHIFT;
 	fbc_ctl |= obj->fence_reg;
 	I915_WRITE(FBC_CONTROL, fbc_ctl);
 
@@ -136,7 +146,7 @@ static bool i8xx_fbc_enabled(struct drm_device *dev)
 	return I915_READ(FBC_CONTROL) & FBC_CTL_EN;
 }
 
-static void g4x_enable_fbc(struct drm_crtc *crtc, unsigned long interval)
+static void g4x_enable_fbc(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -145,16 +155,12 @@ static void g4x_enable_fbc(struct drm_crtc *crtc, unsigned long interval)
 	struct drm_i915_gem_object *obj = intel_fb->obj;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	int plane = intel_crtc->plane == 0 ? DPFC_CTL_PLANEA : DPFC_CTL_PLANEB;
-	unsigned long stall_watermark = 200;
 	u32 dpfc_ctl;
 
 	dpfc_ctl = plane | DPFC_SR_EN | DPFC_CTL_LIMIT_1X;
 	dpfc_ctl |= DPFC_CTL_FENCE_EN | obj->fence_reg;
 	I915_WRITE(DPFC_CHICKEN, DPFC_HT_MODIFY);
 
-	I915_WRITE(DPFC_RECOMP_CTL, DPFC_RECOMP_STALL_EN |
-		   (stall_watermark << DPFC_RECOMP_STALL_WM_SHIFT) |
-		   (interval << DPFC_RECOMP_TIMER_COUNT_SHIFT));
 	I915_WRITE(DPFC_FENCE_YOFF, crtc->y);
 
 	/* enable it... */
@@ -191,7 +197,11 @@ static void sandybridge_blit_fbc_update(struct drm_device *dev)
 	u32 blt_ecoskpd;
 
 	/* Make sure blitter notifies FBC of writes */
-	gen6_gt_force_wake_get(dev_priv);
+
+	/* Blitter is part of Media powerwell on VLV. No impact of
+	 * his param in other platforms for now */
+	gen6_gt_force_wake_get(dev_priv, FORCEWAKE_MEDIA);
+
 	blt_ecoskpd = I915_READ(GEN6_BLITTER_ECOSKPD);
 	blt_ecoskpd |= GEN6_BLITTER_FBC_NOTIFY <<
 		GEN6_BLITTER_LOCK_SHIFT;
@@ -202,10 +212,11 @@ static void sandybridge_blit_fbc_update(struct drm_device *dev)
 			 GEN6_BLITTER_LOCK_SHIFT);
 	I915_WRITE(GEN6_BLITTER_ECOSKPD, blt_ecoskpd);
 	POSTING_READ(GEN6_BLITTER_ECOSKPD);
-	gen6_gt_force_wake_put(dev_priv);
+
+	gen6_gt_force_wake_put(dev_priv, FORCEWAKE_MEDIA);
 }
 
-static void ironlake_enable_fbc(struct drm_crtc *crtc, unsigned long interval)
+static void ironlake_enable_fbc(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -214,7 +225,6 @@ static void ironlake_enable_fbc(struct drm_crtc *crtc, unsigned long interval)
 	struct drm_i915_gem_object *obj = intel_fb->obj;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	int plane = intel_crtc->plane == 0 ? DPFC_CTL_PLANEA : DPFC_CTL_PLANEB;
-	unsigned long stall_watermark = 200;
 	u32 dpfc_ctl;
 
 	dpfc_ctl = I915_READ(ILK_DPFC_CONTROL);
@@ -222,12 +232,11 @@ static void ironlake_enable_fbc(struct drm_crtc *crtc, unsigned long interval)
 	dpfc_ctl |= (plane | DPFC_CTL_LIMIT_1X);
 	/* Set persistent mode for front-buffer rendering, ala X. */
 	dpfc_ctl |= DPFC_CTL_PERSISTENT_MODE;
-	dpfc_ctl |= (DPFC_CTL_FENCE_EN | obj->fence_reg);
+	dpfc_ctl |= DPFC_CTL_FENCE_EN;
+	if (IS_GEN5(dev))
+		dpfc_ctl |= obj->fence_reg;
 	I915_WRITE(ILK_DPFC_CHICKEN, DPFC_HT_MODIFY);
 
-	I915_WRITE(ILK_DPFC_RECOMP_CTL, DPFC_RECOMP_STALL_EN |
-		   (stall_watermark << DPFC_RECOMP_STALL_WM_SHIFT) |
-		   (interval << DPFC_RECOMP_TIMER_COUNT_SHIFT));
 	I915_WRITE(ILK_DPFC_FENCE_YOFF, crtc->y);
 	I915_WRITE(ILK_FBC_RT_BASE, i915_gem_obj_ggtt_offset(obj) | ILK_FBC_RT_VALID);
 	/* enable it... */
@@ -265,7 +274,7 @@ static bool ironlake_fbc_enabled(struct drm_device *dev)
 	return I915_READ(ILK_DPFC_CONTROL) & DPFC_CTL_EN;
 }
 
-static void gen7_enable_fbc(struct drm_crtc *crtc, unsigned long interval)
+static void gen7_enable_fbc(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -295,7 +304,7 @@ static void gen7_enable_fbc(struct drm_crtc *crtc, unsigned long interval)
 
 	sandybridge_blit_fbc_update(dev);
 
-	DRM_DEBUG_KMS("enabled fbc on plane %d\n", intel_crtc->plane);
+	DRM_DEBUG_KMS("enabled fbc on plane %c\n", plane_name(intel_crtc->plane));
 }
 
 bool intel_fbc_enabled(struct drm_device *dev)
@@ -322,8 +331,7 @@ static void intel_fbc_work_fn(struct work_struct *__work)
 		 * the prior work.
 		 */
 		if (work->crtc->fb == work->fb) {
-			dev_priv->display.enable_fbc(work->crtc,
-						     work->interval);
+			dev_priv->display.enable_fbc(work->crtc);
 
 			dev_priv->fbc.plane = to_intel_crtc(work->crtc)->plane;
 			dev_priv->fbc.fb_id = work->crtc->fb->base.id;
@@ -360,7 +368,7 @@ static void intel_cancel_fbc_work(struct drm_i915_private *dev_priv)
 	dev_priv->fbc.fbc_work = NULL;
 }
 
-static void intel_enable_fbc(struct drm_crtc *crtc, unsigned long interval)
+static void intel_enable_fbc(struct drm_crtc *crtc)
 {
 	struct intel_fbc_work *work;
 	struct drm_device *dev = crtc->dev;
@@ -374,13 +382,12 @@ static void intel_enable_fbc(struct drm_crtc *crtc, unsigned long interval)
 	work = kzalloc(sizeof(*work), GFP_KERNEL);
 	if (work == NULL) {
 		DRM_ERROR("Failed to allocate FBC work structure\n");
-		dev_priv->display.enable_fbc(crtc, interval);
+		dev_priv->display.enable_fbc(crtc);
 		return;
 	}
 
 	work->crtc = crtc;
 	work->fb = crtc->fb;
-	work->interval = interval;
 	INIT_DELAYED_WORK(&work->work, intel_fbc_work_fn);
 
 	dev_priv->fbc.fbc_work = work;
@@ -454,7 +461,7 @@ void intel_update_fbc(struct drm_device *dev)
 	const struct drm_display_mode *adjusted_mode;
 	unsigned int max_width, max_height;
 
-	if (!I915_HAS_FBC(dev)) {
+	if (!HAS_FBC(dev)) {
 		set_no_fbc_reason(dev_priv, FBC_UNSUPPORTED);
 		return;
 	}
@@ -530,10 +537,10 @@ void intel_update_fbc(struct drm_device *dev)
 			DRM_DEBUG_KMS("mode too large for compression, disabling\n");
 		goto out_disable;
 	}
-	if ((IS_I915GM(dev) || IS_I945GM(dev) || IS_HASWELL(dev)) &&
-	    intel_crtc->plane != 0) {
+	if ((INTEL_INFO(dev)->gen < 4 || IS_HASWELL(dev)) &&
+	    intel_crtc->plane != PLANE_A) {
 		if (set_no_fbc_reason(dev_priv, FBC_BAD_PLANE))
-			DRM_DEBUG_KMS("plane not 0, disabling compression\n");
+			DRM_DEBUG_KMS("plane not A, disabling compression\n");
 		goto out_disable;
 	}
 
@@ -595,7 +602,7 @@ void intel_update_fbc(struct drm_device *dev)
 		intel_disable_fbc(dev);
 	}
 
-	intel_enable_fbc(crtc, 500);
+	intel_enable_fbc(crtc);
 	dev_priv->fbc.no_fbc_reason = FBC_OK;
 	return;
 
@@ -817,7 +824,7 @@ static int i9xx_get_fifo_size(struct drm_device *dev, int plane)
 	return size;
 }
 
-static int i85x_get_fifo_size(struct drm_device *dev, int plane)
+static int i830_get_fifo_size(struct drm_device *dev, int plane)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	uint32_t dsparb = I915_READ(DSPARB);
@@ -850,21 +857,6 @@ static int i845_get_fifo_size(struct drm_device *dev, int plane)
 	return size;
 }
 
-static int i830_get_fifo_size(struct drm_device *dev, int plane)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	uint32_t dsparb = I915_READ(DSPARB);
-	int size;
-
-	size = dsparb & 0x7f;
-	size >>= 1; /* Convert to cachelines */
-
-	DRM_DEBUG_KMS("FIFO size - (0x%08x) %s: %d\n", dsparb,
-		      plane ? "B" : "A", size);
-
-	return size;
-}
-
 /* Pineview has different values for various configs */
 static const struct intel_watermark_params pineview_display_wm = {
 	PINEVIEW_DISPLAY_FIFO,
@@ -943,14 +935,14 @@ static const struct intel_watermark_params i915_wm_info = {
 	2,
 	I915_FIFO_LINE_SIZE
 };
-static const struct intel_watermark_params i855_wm_info = {
+static const struct intel_watermark_params i830_wm_info = {
 	I855GM_FIFO_SIZE,
 	I915_MAX_WM,
 	1,
 	2,
 	I830_FIFO_LINE_SIZE
 };
-static const struct intel_watermark_params i830_wm_info = {
+static const struct intel_watermark_params i845_wm_info = {
 	I830_FIFO_SIZE,
 	I915_MAX_WM,
 	1,
@@ -958,65 +950,6 @@ static const struct intel_watermark_params i830_wm_info = {
 	I830_FIFO_LINE_SIZE
 };
 
-static const struct intel_watermark_params ironlake_display_wm_info = {
-	ILK_DISPLAY_FIFO,
-	ILK_DISPLAY_MAXWM,
-	ILK_DISPLAY_DFTWM,
-	2,
-	ILK_FIFO_LINE_SIZE
-};
-static const struct intel_watermark_params ironlake_cursor_wm_info = {
-	ILK_CURSOR_FIFO,
-	ILK_CURSOR_MAXWM,
-	ILK_CURSOR_DFTWM,
-	2,
-	ILK_FIFO_LINE_SIZE
-};
-static const struct intel_watermark_params ironlake_display_srwm_info = {
-	ILK_DISPLAY_SR_FIFO,
-	ILK_DISPLAY_MAX_SRWM,
-	ILK_DISPLAY_DFT_SRWM,
-	2,
-	ILK_FIFO_LINE_SIZE
-};
-static const struct intel_watermark_params ironlake_cursor_srwm_info = {
-	ILK_CURSOR_SR_FIFO,
-	ILK_CURSOR_MAX_SRWM,
-	ILK_CURSOR_DFT_SRWM,
-	2,
-	ILK_FIFO_LINE_SIZE
-};
-
-static const struct intel_watermark_params sandybridge_display_wm_info = {
-	SNB_DISPLAY_FIFO,
-	SNB_DISPLAY_MAXWM,
-	SNB_DISPLAY_DFTWM,
-	2,
-	SNB_FIFO_LINE_SIZE
-};
-static const struct intel_watermark_params sandybridge_cursor_wm_info = {
-	SNB_CURSOR_FIFO,
-	SNB_CURSOR_MAXWM,
-	SNB_CURSOR_DFTWM,
-	2,
-	SNB_FIFO_LINE_SIZE
-};
-static const struct intel_watermark_params sandybridge_display_srwm_info = {
-	SNB_DISPLAY_SR_FIFO,
-	SNB_DISPLAY_MAX_SRWM,
-	SNB_DISPLAY_DFT_SRWM,
-	2,
-	SNB_FIFO_LINE_SIZE
-};
-static const struct intel_watermark_params sandybridge_cursor_srwm_info = {
-	SNB_CURSOR_SR_FIFO,
-	SNB_CURSOR_MAX_SRWM,
-	SNB_CURSOR_DFT_SRWM,
-	2,
-	SNB_FIFO_LINE_SIZE
-};
-
-
 /**
  * intel_calculate_wm - calculate watermark level
  * @clock_in_khz: pixel clock
@@ -1567,7 +1500,7 @@ static void i9xx_update_wm(struct drm_crtc *unused_crtc)
 	else if (!IS_GEN2(dev))
 		wm_info = &i915_wm_info;
 	else
-		wm_info = &i855_wm_info;
+		wm_info = &i830_wm_info;
 
 	fifo_size = dev_priv->display.get_fifo_size(dev, 0);
 	crtc = intel_get_crtc_for_plane(dev, 0);
@@ -1615,7 +1548,7 @@ static void i9xx_update_wm(struct drm_crtc *unused_crtc)
 	if (IS_I945G(dev) || IS_I945GM(dev))
 		I915_WRITE(FW_BLC_SELF, FW_BLC_SELF_EN_MASK | 0);
 	else if (IS_I915GM(dev))
-		I915_WRITE(INSTPM, I915_READ(INSTPM) & ~INSTPM_SELF_EN);
+		I915_WRITE(INSTPM, _MASKED_BIT_DISABLE(INSTPM_SELF_EN));
 
 	/* Calc sr entries for one plane configs */
 	if (HAS_FW_BLC(dev) && enabled) {
@@ -1667,14 +1600,14 @@ static void i9xx_update_wm(struct drm_crtc *unused_crtc)
 				I915_WRITE(FW_BLC_SELF,
 					   FW_BLC_SELF_EN_MASK | FW_BLC_SELF_EN);
 			else if (IS_I915GM(dev))
-				I915_WRITE(INSTPM, I915_READ(INSTPM) | INSTPM_SELF_EN);
+				I915_WRITE(INSTPM, _MASKED_BIT_ENABLE(INSTPM_SELF_EN));
 			DRM_DEBUG_KMS("memory self refresh enabled\n");
 		} else
 			DRM_DEBUG_KMS("memory self refresh disabled\n");
 	}
 }
 
-static void i830_update_wm(struct drm_crtc *unused_crtc)
+static void i845_update_wm(struct drm_crtc *unused_crtc)
 {
 	struct drm_device *dev = unused_crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -1689,7 +1622,7 @@ static void i830_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,
-				       &i830_wm_info,
+				       &i845_wm_info,
 				       dev_priv->display.get_fifo_size(dev, 0),
 				       4, latency_ns);
 	fwater_lo = I915_READ(FW_BLC) & ~0xfff;
@@ -1700,423 +1633,6 @@ static void i830_update_wm(struct drm_crtc *unused_crtc)
 	I915_WRITE(FW_BLC, fwater_lo);
 }
 
-/*
- * Check the wm result.
- *
- * If any calculated watermark values is larger than the maximum value that
- * can be programmed into the associated watermark register, that watermark
- * must be disabled.
- */
-static bool ironlake_check_srwm(struct drm_device *dev, int level,
-				int fbc_wm, int display_wm, int cursor_wm,
-				const struct intel_watermark_params *display,
-				const struct intel_watermark_params *cursor)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-
-	DRM_DEBUG_KMS("watermark %d: display plane %d, fbc lines %d,"
-		      " cursor %d\n", level, display_wm, fbc_wm, cursor_wm);
-
-	if (fbc_wm > SNB_FBC_MAX_SRWM) {
-		DRM_DEBUG_KMS("fbc watermark(%d) is too large(%d), disabling wm%d+\n",
-			      fbc_wm, SNB_FBC_MAX_SRWM, level);
-
-		/* fbc has it's own way to disable FBC WM */
-		I915_WRITE(DISP_ARB_CTL,
-			   I915_READ(DISP_ARB_CTL) | DISP_FBC_WM_DIS);
-		return false;
-	} else if (INTEL_INFO(dev)->gen >= 6) {
-		/* enable FBC WM (except on ILK, where it must remain off) */
-		I915_WRITE(DISP_ARB_CTL,
-			   I915_READ(DISP_ARB_CTL) & ~DISP_FBC_WM_DIS);
-	}
-
-	if (display_wm > display->max_wm) {
-		DRM_DEBUG_KMS("display watermark(%d) is too large(%d), disabling wm%d+\n",
-			      display_wm, SNB_DISPLAY_MAX_SRWM, level);
-		return false;
-	}
-
-	if (cursor_wm > cursor->max_wm) {
-		DRM_DEBUG_KMS("cursor watermark(%d) is too large(%d), disabling wm%d+\n",
-			      cursor_wm, SNB_CURSOR_MAX_SRWM, level);
-		return false;
-	}
-
-	if (!(fbc_wm || display_wm || cursor_wm)) {
-		DRM_DEBUG_KMS("latency %d is 0, disabling wm%d+\n", level, level);
-		return false;
-	}
-
-	return true;
-}
-
-/*
- * Compute watermark values of WM[1-3],
- */
-static bool ironlake_compute_srwm(struct drm_device *dev, int level, int plane,
-				  int latency_ns,
-				  const struct intel_watermark_params *display,
-				  const struct intel_watermark_params *cursor,
-				  int *fbc_wm, int *display_wm, int *cursor_wm)
-{
-	struct drm_crtc *crtc;
-	const struct drm_display_mode *adjusted_mode;
-	unsigned long line_time_us;
-	int hdisplay, htotal, pixel_size, clock;
-	int line_count, line_size;
-	int small, large;
-	int entries;
-
-	if (!latency_ns) {
-		*fbc_wm = *display_wm = *cursor_wm = 0;
-		return false;
-	}
-
-	crtc = intel_get_crtc_for_plane(dev, plane);
-	adjusted_mode = &to_intel_crtc(crtc)->config.adjusted_mode;
-	clock = adjusted_mode->crtc_clock;
-	htotal = adjusted_mode->crtc_htotal;
-	hdisplay = to_intel_crtc(crtc)->config.pipe_src_w;
-	pixel_size = crtc->fb->bits_per_pixel / 8;
-
-	line_time_us = (htotal * 1000) / clock;
-	line_count = (latency_ns / line_time_us + 1000) / 1000;
-	line_size = hdisplay * pixel_size;
-
-	/* Use the minimum of the small and large buffer method for primary */
-	small = ((clock * pixel_size / 1000) * latency_ns) / 1000;
-	large = line_count * line_size;
-
-	entries = DIV_ROUND_UP(min(small, large), display->cacheline_size);
-	*display_wm = entries + display->guard_size;
-
-	/*
-	 * Spec says:
-	 * FBC WM = ((Final Primary WM * 64) / number of bytes per line) + 2
-	 */
-	*fbc_wm = DIV_ROUND_UP(*display_wm * 64, line_size) + 2;
-
-	/* calculate the self-refresh watermark for display cursor */
-	entries = line_count * pixel_size * 64;
-	entries = DIV_ROUND_UP(entries, cursor->cacheline_size);
-	*cursor_wm = entries + cursor->guard_size;
-
-	return ironlake_check_srwm(dev, level,
-				   *fbc_wm, *display_wm, *cursor_wm,
-				   display, cursor);
-}
-
-static void ironlake_update_wm(struct drm_crtc *crtc)
-{
-	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	int fbc_wm, plane_wm, cursor_wm;
-	unsigned int enabled;
-
-	enabled = 0;
-	if (g4x_compute_wm0(dev, PIPE_A,
-			    &ironlake_display_wm_info,
-			    dev_priv->wm.pri_latency[0] * 100,
-			    &ironlake_cursor_wm_info,
-			    dev_priv->wm.cur_latency[0] * 100,
-			    &plane_wm, &cursor_wm)) {
-		I915_WRITE(WM0_PIPEA_ILK,
-			   (plane_wm << WM0_PIPE_PLANE_SHIFT) | cursor_wm);
-		DRM_DEBUG_KMS("FIFO watermarks For pipe A -"
-			      " plane %d, " "cursor: %d\n",
-			      plane_wm, cursor_wm);
-		enabled |= 1 << PIPE_A;
-	}
-
-	if (g4x_compute_wm0(dev, PIPE_B,
-			    &ironlake_display_wm_info,
-			    dev_priv->wm.pri_latency[0] * 100,
-			    &ironlake_cursor_wm_info,
-			    dev_priv->wm.cur_latency[0] * 100,
-			    &plane_wm, &cursor_wm)) {
-		I915_WRITE(WM0_PIPEB_ILK,
-			   (plane_wm << WM0_PIPE_PLANE_SHIFT) | cursor_wm);
-		DRM_DEBUG_KMS("FIFO watermarks For pipe B -"
-			      " plane %d, cursor: %d\n",
-			      plane_wm, cursor_wm);
-		enabled |= 1 << PIPE_B;
-	}
-
-	/*
-	 * Calculate and update the self-refresh watermark only when one
-	 * display plane is used.
-	 */
-	I915_WRITE(WM3_LP_ILK, 0);
-	I915_WRITE(WM2_LP_ILK, 0);
-	I915_WRITE(WM1_LP_ILK, 0);
-
-	if (!single_plane_enabled(enabled))
-		return;
-	enabled = ffs(enabled) - 1;
-
-	/* WM1 */
-	if (!ironlake_compute_srwm(dev, 1, enabled,
-				   dev_priv->wm.pri_latency[1] * 500,
-				   &ironlake_display_srwm_info,
-				   &ironlake_cursor_srwm_info,
-				   &fbc_wm, &plane_wm, &cursor_wm))
-		return;
-
-	I915_WRITE(WM1_LP_ILK,
-		   WM1_LP_SR_EN |
-		   (dev_priv->wm.pri_latency[1] << WM1_LP_LATENCY_SHIFT) |
-		   (fbc_wm << WM1_LP_FBC_SHIFT) |
-		   (plane_wm << WM1_LP_SR_SHIFT) |
-		   cursor_wm);
-
-	/* WM2 */
-	if (!ironlake_compute_srwm(dev, 2, enabled,
-				   dev_priv->wm.pri_latency[2] * 500,
-				   &ironlake_display_srwm_info,
-				   &ironlake_cursor_srwm_info,
-				   &fbc_wm, &plane_wm, &cursor_wm))
-		return;
-
-	I915_WRITE(WM2_LP_ILK,
-		   WM2_LP_EN |
-		   (dev_priv->wm.pri_latency[2] << WM1_LP_LATENCY_SHIFT) |
-		   (fbc_wm << WM1_LP_FBC_SHIFT) |
-		   (plane_wm << WM1_LP_SR_SHIFT) |
-		   cursor_wm);
-
-	/*
-	 * WM3 is unsupported on ILK, probably because we don't have latency
-	 * data for that power state
-	 */
-}
-
-static void sandybridge_update_wm(struct drm_crtc *crtc)
-{
-	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	int latency = dev_priv->wm.pri_latency[0] * 100;	/* In unit 0.1us */
-	u32 val;
-	int fbc_wm, plane_wm, cursor_wm;
-	unsigned int enabled;
-
-	enabled = 0;
-	if (g4x_compute_wm0(dev, PIPE_A,
-			    &sandybridge_display_wm_info, latency,
-			    &sandybridge_cursor_wm_info, latency,
-			    &plane_wm, &cursor_wm)) {
-		val = I915_READ(WM0_PIPEA_ILK);
-		val &= ~(WM0_PIPE_PLANE_MASK | WM0_PIPE_CURSOR_MASK);
-		I915_WRITE(WM0_PIPEA_ILK, val |
-			   ((plane_wm << WM0_PIPE_PLANE_SHIFT) | cursor_wm));
-		DRM_DEBUG_KMS("FIFO watermarks For pipe A -"
-			      " plane %d, " "cursor: %d\n",
-			      plane_wm, cursor_wm);
-		enabled |= 1 << PIPE_A;
-	}
-
-	if (g4x_compute_wm0(dev, PIPE_B,
-			    &sandybridge_display_wm_info, latency,
-			    &sandybridge_cursor_wm_info, latency,
-			    &plane_wm, &cursor_wm)) {
-		val = I915_READ(WM0_PIPEB_ILK);
-		val &= ~(WM0_PIPE_PLANE_MASK | WM0_PIPE_CURSOR_MASK);
-		I915_WRITE(WM0_PIPEB_ILK, val |
-			   ((plane_wm << WM0_PIPE_PLANE_SHIFT) | cursor_wm));
-		DRM_DEBUG_KMS("FIFO watermarks For pipe B -"
-			      " plane %d, cursor: %d\n",
-			      plane_wm, cursor_wm);
-		enabled |= 1 << PIPE_B;
-	}
-
-	/*
-	 * Calculate and update the self-refresh watermark only when one
-	 * display plane is used.
-	 *
-	 * SNB support 3 levels of watermark.
-	 *
-	 * WM1/WM2/WM2 watermarks have to be enabled in the ascending order,
-	 * and disabled in the descending order
-	 *
-	 */
-	I915_WRITE(WM3_LP_ILK, 0);
-	I915_WRITE(WM2_LP_ILK, 0);
-	I915_WRITE(WM1_LP_ILK, 0);
-
-	if (!single_plane_enabled(enabled) ||
-	    dev_priv->sprite_scaling_enabled)
-		return;
-	enabled = ffs(enabled) - 1;
-
-	/* WM1 */
-	if (!ironlake_compute_srwm(dev, 1, enabled,
-				   dev_priv->wm.pri_latency[1] * 500,
-				   &sandybridge_display_srwm_info,
-				   &sandybridge_cursor_srwm_info,
-				   &fbc_wm, &plane_wm, &cursor_wm))
-		return;
-
-	I915_WRITE(WM1_LP_ILK,
-		   WM1_LP_SR_EN |
-		   (dev_priv->wm.pri_latency[1] << WM1_LP_LATENCY_SHIFT) |
-		   (fbc_wm << WM1_LP_FBC_SHIFT) |
-		   (plane_wm << WM1_LP_SR_SHIFT) |
-		   cursor_wm);
-
-	/* WM2 */
-	if (!ironlake_compute_srwm(dev, 2, enabled,
-				   dev_priv->wm.pri_latency[2] * 500,
-				   &sandybridge_display_srwm_info,
-				   &sandybridge_cursor_srwm_info,
-				   &fbc_wm, &plane_wm, &cursor_wm))
-		return;
-
-	I915_WRITE(WM2_LP_ILK,
-		   WM2_LP_EN |
-		   (dev_priv->wm.pri_latency[2] << WM1_LP_LATENCY_SHIFT) |
-		   (fbc_wm << WM1_LP_FBC_SHIFT) |
-		   (plane_wm << WM1_LP_SR_SHIFT) |
-		   cursor_wm);
-
-	/* WM3 */
-	if (!ironlake_compute_srwm(dev, 3, enabled,
-				   dev_priv->wm.pri_latency[3] * 500,
-				   &sandybridge_display_srwm_info,
-				   &sandybridge_cursor_srwm_info,
-				   &fbc_wm, &plane_wm, &cursor_wm))
-		return;
-
-	I915_WRITE(WM3_LP_ILK,
-		   WM3_LP_EN |
-		   (dev_priv->wm.pri_latency[3] << WM1_LP_LATENCY_SHIFT) |
-		   (fbc_wm << WM1_LP_FBC_SHIFT) |
-		   (plane_wm << WM1_LP_SR_SHIFT) |
-		   cursor_wm);
-}
-
-static void ivybridge_update_wm(struct drm_crtc *crtc)
-{
-	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	int latency = dev_priv->wm.pri_latency[0] * 100;	/* In unit 0.1us */
-	u32 val;
-	int fbc_wm, plane_wm, cursor_wm;
-	int ignore_fbc_wm, ignore_plane_wm, ignore_cursor_wm;
-	unsigned int enabled;
-
-	enabled = 0;
-	if (g4x_compute_wm0(dev, PIPE_A,
-			    &sandybridge_display_wm_info, latency,
-			    &sandybridge_cursor_wm_info, latency,
-			    &plane_wm, &cursor_wm)) {
-		val = I915_READ(WM0_PIPEA_ILK);
-		val &= ~(WM0_PIPE_PLANE_MASK | WM0_PIPE_CURSOR_MASK);
-		I915_WRITE(WM0_PIPEA_ILK, val |
-			   ((plane_wm << WM0_PIPE_PLANE_SHIFT) | cursor_wm));
-		DRM_DEBUG_KMS("FIFO watermarks For pipe A -"
-			      " plane %d, " "cursor: %d\n",
-			      plane_wm, cursor_wm);
-		enabled |= 1 << PIPE_A;
-	}
-
-	if (g4x_compute_wm0(dev, PIPE_B,
-			    &sandybridge_display_wm_info, latency,
-			    &sandybridge_cursor_wm_info, latency,
-			    &plane_wm, &cursor_wm)) {
-		val = I915_READ(WM0_PIPEB_ILK);
-		val &= ~(WM0_PIPE_PLANE_MASK | WM0_PIPE_CURSOR_MASK);
-		I915_WRITE(WM0_PIPEB_ILK, val |
-			   ((plane_wm << WM0_PIPE_PLANE_SHIFT) | cursor_wm));
-		DRM_DEBUG_KMS("FIFO watermarks For pipe B -"
-			      " plane %d, cursor: %d\n",
-			      plane_wm, cursor_wm);
-		enabled |= 1 << PIPE_B;
-	}
-
-	if (g4x_compute_wm0(dev, PIPE_C,
-			    &sandybridge_display_wm_info, latency,
-			    &sandybridge_cursor_wm_info, latency,
-			    &plane_wm, &cursor_wm)) {
-		val = I915_READ(WM0_PIPEC_IVB);
-		val &= ~(WM0_PIPE_PLANE_MASK | WM0_PIPE_CURSOR_MASK);
-		I915_WRITE(WM0_PIPEC_IVB, val |
-			   ((plane_wm << WM0_PIPE_PLANE_SHIFT) | cursor_wm));
-		DRM_DEBUG_KMS("FIFO watermarks For pipe C -"
-			      " plane %d, cursor: %d\n",
-			      plane_wm, cursor_wm);
-		enabled |= 1 << PIPE_C;
-	}
-
-	/*
-	 * Calculate and update the self-refresh watermark only when one
-	 * display plane is used.
-	 *
-	 * SNB support 3 levels of watermark.
-	 *
-	 * WM1/WM2/WM2 watermarks have to be enabled in the ascending order,
-	 * and disabled in the descending order
-	 *
-	 */
-	I915_WRITE(WM3_LP_ILK, 0);
-	I915_WRITE(WM2_LP_ILK, 0);
-	I915_WRITE(WM1_LP_ILK, 0);
-
-	if (!single_plane_enabled(enabled) ||
-	    dev_priv->sprite_scaling_enabled)
-		return;
-	enabled = ffs(enabled) - 1;
-
-	/* WM1 */
-	if (!ironlake_compute_srwm(dev, 1, enabled,
-				   dev_priv->wm.pri_latency[1] * 500,
-				   &sandybridge_display_srwm_info,
-				   &sandybridge_cursor_srwm_info,
-				   &fbc_wm, &plane_wm, &cursor_wm))
-		return;
-
-	I915_WRITE(WM1_LP_ILK,
-		   WM1_LP_SR_EN |
-		   (dev_priv->wm.pri_latency[1] << WM1_LP_LATENCY_SHIFT) |
-		   (fbc_wm << WM1_LP_FBC_SHIFT) |
-		   (plane_wm << WM1_LP_SR_SHIFT) |
-		   cursor_wm);
-
-	/* WM2 */
-	if (!ironlake_compute_srwm(dev, 2, enabled,
-				   dev_priv->wm.pri_latency[2] * 500,
-				   &sandybridge_display_srwm_info,
-				   &sandybridge_cursor_srwm_info,
-				   &fbc_wm, &plane_wm, &cursor_wm))
-		return;
-
-	I915_WRITE(WM2_LP_ILK,
-		   WM2_LP_EN |
-		   (dev_priv->wm.pri_latency[2] << WM1_LP_LATENCY_SHIFT) |
-		   (fbc_wm << WM1_LP_FBC_SHIFT) |
-		   (plane_wm << WM1_LP_SR_SHIFT) |
-		   cursor_wm);
-
-	/* WM3, note we have to correct the cursor latency */
-	if (!ironlake_compute_srwm(dev, 3, enabled,
-				   dev_priv->wm.pri_latency[3] * 500,
-				   &sandybridge_display_srwm_info,
-				   &sandybridge_cursor_srwm_info,
-				   &fbc_wm, &plane_wm, &ignore_cursor_wm) ||
-	    !ironlake_compute_srwm(dev, 3, enabled,
-				   dev_priv->wm.cur_latency[3] * 500,
-				   &sandybridge_display_srwm_info,
-				   &sandybridge_cursor_srwm_info,
-				   &ignore_fbc_wm, &ignore_plane_wm, &cursor_wm))
-		return;
-
-	I915_WRITE(WM3_LP_ILK,
-		   WM3_LP_EN |
-		   (dev_priv->wm.pri_latency[3] << WM1_LP_LATENCY_SHIFT) |
-		   (fbc_wm << WM1_LP_FBC_SHIFT) |
-		   (plane_wm << WM1_LP_SR_SHIFT) |
-		   cursor_wm);
-}
-
 static uint32_t ilk_pipe_pixel_rate(struct drm_device *dev,
 				    struct drm_crtc *crtc)
 {
@@ -2185,7 +1701,7 @@ static uint32_t ilk_wm_fbc(uint32_t pri_val, uint32_t horiz_pixels,
 	return DIV_ROUND_UP(pri_val * 64, horiz_pixels * bytes_per_pixel) + 2;
 }
 
-struct hsw_pipe_wm_parameters {
+struct ilk_pipe_wm_parameters {
 	bool active;
 	uint32_t pipe_htotal;
 	uint32_t pixel_rate;
@@ -2194,7 +1710,7 @@ struct hsw_pipe_wm_parameters {
 	struct intel_plane_wm_parameters cur;
 };
 
-struct hsw_wm_maximums {
+struct ilk_wm_maximums {
 	uint16_t pri;
 	uint16_t spr;
 	uint16_t cur;
@@ -2212,7 +1728,7 @@ struct intel_wm_config {
  * For both WM_PIPE and WM_LP.
  * mem_value must be in 0.1us units.
  */
-static uint32_t ilk_compute_pri_wm(const struct hsw_pipe_wm_parameters *params,
+static uint32_t ilk_compute_pri_wm(const struct ilk_pipe_wm_parameters *params,
 				   uint32_t mem_value,
 				   bool is_lp)
 {
@@ -2241,7 +1757,7 @@ static uint32_t ilk_compute_pri_wm(const struct hsw_pipe_wm_parameters *params,
  * For both WM_PIPE and WM_LP.
  * mem_value must be in 0.1us units.
  */
-static uint32_t ilk_compute_spr_wm(const struct hsw_pipe_wm_parameters *params,
+static uint32_t ilk_compute_spr_wm(const struct ilk_pipe_wm_parameters *params,
 				   uint32_t mem_value)
 {
 	uint32_t method1, method2;
@@ -2264,7 +1780,7 @@ static uint32_t ilk_compute_spr_wm(const struct hsw_pipe_wm_parameters *params,
  * For both WM_PIPE and WM_LP.
  * mem_value must be in 0.1us units.
  */
-static uint32_t ilk_compute_cur_wm(const struct hsw_pipe_wm_parameters *params,
+static uint32_t ilk_compute_cur_wm(const struct ilk_pipe_wm_parameters *params,
 				   uint32_t mem_value)
 {
 	if (!params->active || !params->cur.enabled)
@@ -2278,7 +1794,7 @@ static uint32_t ilk_compute_cur_wm(const struct hsw_pipe_wm_parameters *params,
 }
 
 /* Only for WM_LP. */
-static uint32_t ilk_compute_fbc_wm(const struct hsw_pipe_wm_parameters *params,
+static uint32_t ilk_compute_fbc_wm(const struct ilk_pipe_wm_parameters *params,
 				   uint32_t pri_val)
 {
 	if (!params->active || !params->pri.enabled)
@@ -2383,7 +1899,7 @@ static void ilk_compute_wm_maximums(struct drm_device *dev,
 				    int level,
 				    const struct intel_wm_config *config,
 				    enum intel_ddb_partitioning ddb_partitioning,
-				    struct hsw_wm_maximums *max)
+				    struct ilk_wm_maximums *max)
 {
 	max->pri = ilk_plane_wm_max(dev, level, config, ddb_partitioning, false);
 	max->spr = ilk_plane_wm_max(dev, level, config, ddb_partitioning, true);
@@ -2392,7 +1908,7 @@ static void ilk_compute_wm_maximums(struct drm_device *dev,
 }
 
 static bool ilk_validate_wm_level(int level,
-				  const struct hsw_wm_maximums *max,
+				  const struct ilk_wm_maximums *max,
 				  struct intel_wm_level *result)
 {
 	bool ret;
@@ -2434,7 +1950,7 @@ static bool ilk_validate_wm_level(int level,
 
 static void ilk_compute_wm_level(struct drm_i915_private *dev_priv,
 				 int level,
-				 const struct hsw_pipe_wm_parameters *p,
+				 const struct ilk_pipe_wm_parameters *p,
 				 struct intel_wm_level *result)
 {
 	uint16_t pri_latency = dev_priv->wm.pri_latency[level];
@@ -2482,7 +1998,7 @@ static void intel_read_wm_latency(struct drm_device *dev, uint16_t wm[5])
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	if (IS_HASWELL(dev)) {
+	if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
 		uint64_t sskpd = I915_READ64(MCH_SSKPD);
 
 		wm[0] = (sskpd >> 56) & 0xFF;
@@ -2530,7 +2046,7 @@ static void intel_fixup_cur_wm_latency(struct drm_device *dev, uint16_t wm[5])
 static int ilk_wm_max_level(const struct drm_device *dev)
 {
 	/* how many WM levels are we expecting */
-	if (IS_HASWELL(dev))
+	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
 		return 4;
 	else if (INTEL_INFO(dev)->gen >= 6)
 		return 3;
@@ -2582,8 +2098,8 @@ static void intel_setup_wm_latency(struct drm_device *dev)
 	intel_print_wm_latency(dev, "Cursor", dev_priv->wm.cur_latency);
 }
 
-static void hsw_compute_wm_parameters(struct drm_crtc *crtc,
-				      struct hsw_pipe_wm_parameters *p,
+static void ilk_compute_wm_parameters(struct drm_crtc *crtc,
+				      struct ilk_pipe_wm_parameters *p,
 				      struct intel_wm_config *config)
 {
 	struct drm_device *dev = crtc->dev;
@@ -2593,7 +2109,7 @@ static void hsw_compute_wm_parameters(struct drm_crtc *crtc,
 
 	p->active = intel_crtc_active(crtc);
 	if (p->active) {
-		p->pipe_htotal = intel_crtc->config.adjusted_mode.htotal;
+		p->pipe_htotal = intel_crtc->config.adjusted_mode.crtc_htotal;
 		p->pixel_rate = ilk_pipe_pixel_rate(dev, crtc);
 		p->pri.bytes_per_pixel = crtc->fb->bits_per_pixel / 8;
 		p->cur.bytes_per_pixel = 4;
@@ -2620,7 +2136,7 @@ static void hsw_compute_wm_parameters(struct drm_crtc *crtc,
 
 /* Compute new watermarks for the pipe */
 static bool intel_compute_pipe_wm(struct drm_crtc *crtc,
-				  const struct hsw_pipe_wm_parameters *params,
+				  const struct ilk_pipe_wm_parameters *params,
 				  struct intel_pipe_wm *pipe_wm)
 {
 	struct drm_device *dev = crtc->dev;
@@ -2632,16 +2148,25 @@ static bool intel_compute_pipe_wm(struct drm_crtc *crtc,
 		.sprites_enabled = params->spr.enabled,
 		.sprites_scaled = params->spr.scaled,
 	};
-	struct hsw_wm_maximums max;
+	struct ilk_wm_maximums max;
 
 	/* LP0 watermarks always use 1/2 DDB partitioning */
 	ilk_compute_wm_maximums(dev, 0, &config, INTEL_DDB_PART_1_2, &max);
 
+	/* ILK/SNB: LP2+ watermarks only w/o sprites */
+	if (INTEL_INFO(dev)->gen <= 6 && params->spr.enabled)
+		max_level = 1;
+
+	/* ILK/SNB/IVB: LP1+ watermarks only w/o scaling */
+	if (params->spr.scaled)
+		max_level = 0;
+
 	for (level = 0; level <= max_level; level++)
 		ilk_compute_wm_level(dev_priv, level, params,
 				     &pipe_wm->wm[level]);
 
-	pipe_wm->linetime = hsw_compute_linetime_wm(dev, crtc);
+	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
+		pipe_wm->linetime = hsw_compute_linetime_wm(dev, crtc);
 
 	/* At least LP0 must be valid */
 	return ilk_validate_wm_level(0, &max, &pipe_wm->wm[0]);
@@ -2676,12 +2201,19 @@ static void ilk_merge_wm_level(struct drm_device *dev,
  * Merge all low power watermarks for all active pipes.
  */
 static void ilk_wm_merge(struct drm_device *dev,
-			 const struct hsw_wm_maximums *max,
+			 const struct intel_wm_config *config,
+			 const struct ilk_wm_maximums *max,
 			 struct intel_pipe_wm *merged)
 {
 	int level, max_level = ilk_wm_max_level(dev);
 
-	merged->fbc_wm_enabled = true;
+	/* ILK/SNB/IVB: LP1+ watermarks only w/ single pipe */
+	if ((INTEL_INFO(dev)->gen <= 6 || IS_IVYBRIDGE(dev)) &&
+	    config->num_pipes_active > 1)
+		return;
+
+	/* ILK: FBC WM must be disabled always */
+	merged->fbc_wm_enabled = INTEL_INFO(dev)->gen >= 6;
 
 	/* merge each WM1+ level */
 	for (level = 1; level <= max_level; level++) {
@@ -2701,6 +2233,20 @@ static void ilk_wm_merge(struct drm_device *dev,
 			wm->fbc_val = 0;
 		}
 	}
+
+	/* ILK: LP2+ must be disabled when FBC WM is disabled but FBC enabled */
+	/*
+	 * FIXME this is racy. FBC might get enabled later.
+	 * What we should check here is whether FBC can be
+	 * enabled sometime later.
+	 */
+	if (IS_GEN5(dev) && !merged->fbc_wm_enabled && intel_fbc_enabled(dev)) {
+		for (level = 2; level <= max_level; level++) {
+			struct intel_wm_level *wm = &merged->wm[level];
+
+			wm->enable = false;
+		}
+	}
 }
 
 static int ilk_wm_lp_to_level(int wm_lp, const struct intel_pipe_wm *pipe_wm)
@@ -2709,10 +2255,21 @@ static int ilk_wm_lp_to_level(int wm_lp, const struct intel_pipe_wm *pipe_wm)
 	return wm_lp + (wm_lp >= 2 && pipe_wm->wm[4].enable);
 }
 
-static void hsw_compute_wm_results(struct drm_device *dev,
+/* The value we need to program into the WM_LPx latency field */
+static unsigned int ilk_wm_lp_latency(struct drm_device *dev, int level)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
+		return 2 * level;
+	else
+		return dev_priv->wm.pri_latency[level];
+}
+
+static void ilk_compute_wm_results(struct drm_device *dev,
 				   const struct intel_pipe_wm *merged,
 				   enum intel_ddb_partitioning partitioning,
-				   struct hsw_wm_values *results)
+				   struct ilk_wm_values *results)
 {
 	struct intel_crtc *intel_crtc;
 	int level, wm_lp;
@@ -2731,7 +2288,7 @@ static void hsw_compute_wm_results(struct drm_device *dev,
 			break;
 
 		results->wm_lp[wm_lp - 1] = WM3_LP_EN |
-			((level * 2) << WM1_LP_LATENCY_SHIFT) |
+			(ilk_wm_lp_latency(dev, level) << WM1_LP_LATENCY_SHIFT) |
 			(r->pri_val << WM1_LP_SR_SHIFT) |
 			r->cur_val;
 
@@ -2742,7 +2299,11 @@ static void hsw_compute_wm_results(struct drm_device *dev,
 			results->wm_lp[wm_lp - 1] |=
 				r->fbc_val << WM1_LP_FBC_SHIFT;
 
-		results->wm_lp_spr[wm_lp - 1] = r->spr_val;
+		if (INTEL_INFO(dev)->gen <= 6 && r->spr_val) {
+			WARN_ON(wm_lp != 1);
+			results->wm_lp_spr[wm_lp - 1] = WM1S_LP_EN | r->spr_val;
+		} else
+			results->wm_lp_spr[wm_lp - 1] = r->spr_val;
 	}
 
 	/* LP0 register values */
@@ -2765,7 +2326,7 @@ static void hsw_compute_wm_results(struct drm_device *dev,
 
 /* Find the result with the highest level enabled. Check for enable_fbc_wm in
  * case both are at the same level. Prefer r1 in case they're the same. */
-static struct intel_pipe_wm *hsw_find_best_result(struct drm_device *dev,
+static struct intel_pipe_wm *ilk_find_best_result(struct drm_device *dev,
 						  struct intel_pipe_wm *r1,
 						  struct intel_pipe_wm *r2)
 {
@@ -2800,8 +2361,8 @@ static struct intel_pipe_wm *hsw_find_best_result(struct drm_device *dev,
 #define WM_DIRTY_DDB (1 << 25)
 
 static unsigned int ilk_compute_wm_dirty(struct drm_device *dev,
-					 const struct hsw_wm_values *old,
-					 const struct hsw_wm_values *new)
+					 const struct ilk_wm_values *old,
+					 const struct ilk_wm_values *new)
 {
 	unsigned int dirty = 0;
 	enum pipe pipe;
@@ -2851,27 +2412,53 @@ static unsigned int ilk_compute_wm_dirty(struct drm_device *dev,
 	return dirty;
 }
 
+static bool _ilk_disable_lp_wm(struct drm_i915_private *dev_priv,
+			       unsigned int dirty)
+{
+	struct ilk_wm_values *previous = &dev_priv->wm.hw;
+	bool changed = false;
+
+	if (dirty & WM_DIRTY_LP(3) && previous->wm_lp[2] & WM1_LP_SR_EN) {
+		previous->wm_lp[2] &= ~WM1_LP_SR_EN;
+		I915_WRITE(WM3_LP_ILK, previous->wm_lp[2]);
+		changed = true;
+	}
+	if (dirty & WM_DIRTY_LP(2) && previous->wm_lp[1] & WM1_LP_SR_EN) {
+		previous->wm_lp[1] &= ~WM1_LP_SR_EN;
+		I915_WRITE(WM2_LP_ILK, previous->wm_lp[1]);
+		changed = true;
+	}
+	if (dirty & WM_DIRTY_LP(1) && previous->wm_lp[0] & WM1_LP_SR_EN) {
+		previous->wm_lp[0] &= ~WM1_LP_SR_EN;
+		I915_WRITE(WM1_LP_ILK, previous->wm_lp[0]);
+		changed = true;
+	}
+
+	/*
+	 * Don't touch WM1S_LP_EN here.
+	 * Doing so could cause underruns.
+	 */
+
+	return changed;
+}
+
 /*
  * The spec says we shouldn't write when we don't need, because every write
  * causes WMs to be re-evaluated, expending some power.
  */
-static void hsw_write_wm_values(struct drm_i915_private *dev_priv,
-				struct hsw_wm_values *results)
+static void ilk_write_wm_values(struct drm_i915_private *dev_priv,
+				struct ilk_wm_values *results)
 {
-	struct hsw_wm_values *previous = &dev_priv->wm.hw;
+	struct drm_device *dev = dev_priv->dev;
+	struct ilk_wm_values *previous = &dev_priv->wm.hw;
 	unsigned int dirty;
 	uint32_t val;
 
-	dirty = ilk_compute_wm_dirty(dev_priv->dev, previous, results);
+	dirty = ilk_compute_wm_dirty(dev, previous, results);
 	if (!dirty)
 		return;
 
-	if (dirty & WM_DIRTY_LP(3) && previous->wm_lp[2] != 0)
-		I915_WRITE(WM3_LP_ILK, 0);
-	if (dirty & WM_DIRTY_LP(2) && previous->wm_lp[1] != 0)
-		I915_WRITE(WM2_LP_ILK, 0);
-	if (dirty & WM_DIRTY_LP(1) && previous->wm_lp[0] != 0)
-		I915_WRITE(WM1_LP_ILK, 0);
+	_ilk_disable_lp_wm(dev_priv, dirty);
 
 	if (dirty & WM_DIRTY_PIPE(PIPE_A))
 		I915_WRITE(WM0_PIPEA_ILK, results->wm_pipe[0]);
@@ -2888,12 +2475,21 @@ static void hsw_write_wm_values(struct drm_i915_private *dev_priv,
 		I915_WRITE(PIPE_WM_LINETIME(PIPE_C), results->wm_linetime[2]);
 
 	if (dirty & WM_DIRTY_DDB) {
-		val = I915_READ(WM_MISC);
-		if (results->partitioning == INTEL_DDB_PART_1_2)
-			val &= ~WM_MISC_DATA_PARTITION_5_6;
-		else
-			val |= WM_MISC_DATA_PARTITION_5_6;
-		I915_WRITE(WM_MISC, val);
+		if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
+			val = I915_READ(WM_MISC);
+			if (results->partitioning == INTEL_DDB_PART_1_2)
+				val &= ~WM_MISC_DATA_PARTITION_5_6;
+			else
+				val |= WM_MISC_DATA_PARTITION_5_6;
+			I915_WRITE(WM_MISC, val);
+		} else {
+			val = I915_READ(DISP_ARB_CTL2);
+			if (results->partitioning == INTEL_DDB_PART_1_2)
+				val &= ~DISP_DATA_PARTITION_5_6;
+			else
+				val |= DISP_DATA_PARTITION_5_6;
+			I915_WRITE(DISP_ARB_CTL2, val);
+		}
 	}
 
 	if (dirty & WM_DIRTY_FBC) {
@@ -2905,37 +2501,48 @@ static void hsw_write_wm_values(struct drm_i915_private *dev_priv,
 		I915_WRITE(DISP_ARB_CTL, val);
 	}
 
-	if (dirty & WM_DIRTY_LP(1) && previous->wm_lp_spr[0] != results->wm_lp_spr[0])
+	if (dirty & WM_DIRTY_LP(1) &&
+	    previous->wm_lp_spr[0] != results->wm_lp_spr[0])
 		I915_WRITE(WM1S_LP_ILK, results->wm_lp_spr[0]);
-	if (dirty & WM_DIRTY_LP(2) && previous->wm_lp_spr[1] != results->wm_lp_spr[1])
-		I915_WRITE(WM2S_LP_IVB, results->wm_lp_spr[1]);
-	if (dirty & WM_DIRTY_LP(3) && previous->wm_lp_spr[2] != results->wm_lp_spr[2])
-		I915_WRITE(WM3S_LP_IVB, results->wm_lp_spr[2]);
 
-	if (dirty & WM_DIRTY_LP(1) && results->wm_lp[0] != 0)
+	if (INTEL_INFO(dev)->gen >= 7) {
+		if (dirty & WM_DIRTY_LP(2) && previous->wm_lp_spr[1] != results->wm_lp_spr[1])
+			I915_WRITE(WM2S_LP_IVB, results->wm_lp_spr[1]);
+		if (dirty & WM_DIRTY_LP(3) && previous->wm_lp_spr[2] != results->wm_lp_spr[2])
+			I915_WRITE(WM3S_LP_IVB, results->wm_lp_spr[2]);
+	}
+
+	if (dirty & WM_DIRTY_LP(1) && previous->wm_lp[0] != results->wm_lp[0])
 		I915_WRITE(WM1_LP_ILK, results->wm_lp[0]);
-	if (dirty & WM_DIRTY_LP(2) && results->wm_lp[1] != 0)
+	if (dirty & WM_DIRTY_LP(2) && previous->wm_lp[1] != results->wm_lp[1])
 		I915_WRITE(WM2_LP_ILK, results->wm_lp[1]);
-	if (dirty & WM_DIRTY_LP(3) && results->wm_lp[2] != 0)
+	if (dirty & WM_DIRTY_LP(3) && previous->wm_lp[2] != results->wm_lp[2])
 		I915_WRITE(WM3_LP_ILK, results->wm_lp[2]);
 
 	dev_priv->wm.hw = *results;
 }
 
-static void haswell_update_wm(struct drm_crtc *crtc)
+static bool ilk_disable_lp_wm(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	return _ilk_disable_lp_wm(dev_priv, WM_DIRTY_LP_ALL);
+}
+
+static void ilk_update_wm(struct drm_crtc *crtc)
 {
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct hsw_wm_maximums max;
-	struct hsw_pipe_wm_parameters params = {};
-	struct hsw_wm_values results = {};
+	struct ilk_wm_maximums max;
+	struct ilk_pipe_wm_parameters params = {};
+	struct ilk_wm_values results = {};
 	enum intel_ddb_partitioning partitioning;
 	struct intel_pipe_wm pipe_wm = {};
 	struct intel_pipe_wm lp_wm_1_2 = {}, lp_wm_5_6 = {}, *best_lp_wm;
 	struct intel_wm_config config = {};
 
-	hsw_compute_wm_parameters(crtc, &params, &config);
+	ilk_compute_wm_parameters(crtc, &params, &config);
 
 	intel_compute_pipe_wm(crtc, &params, &pipe_wm);
 
@@ -2945,15 +2552,15 @@ static void haswell_update_wm(struct drm_crtc *crtc)
 	intel_crtc->wm.active = pipe_wm;
 
 	ilk_compute_wm_maximums(dev, 1, &config, INTEL_DDB_PART_1_2, &max);
-	ilk_wm_merge(dev, &max, &lp_wm_1_2);
+	ilk_wm_merge(dev, &config, &max, &lp_wm_1_2);
 
 	/* 5/6 split only in single pipe config on IVB+ */
 	if (INTEL_INFO(dev)->gen >= 7 &&
 	    config.num_pipes_active == 1 && config.sprites_enabled) {
 		ilk_compute_wm_maximums(dev, 1, &config, INTEL_DDB_PART_5_6, &max);
-		ilk_wm_merge(dev, &max, &lp_wm_5_6);
+		ilk_wm_merge(dev, &config, &max, &lp_wm_5_6);
 
-		best_lp_wm = hsw_find_best_result(dev, &lp_wm_1_2, &lp_wm_5_6);
+		best_lp_wm = ilk_find_best_result(dev, &lp_wm_1_2, &lp_wm_5_6);
 	} else {
 		best_lp_wm = &lp_wm_1_2;
 	}
@@ -2961,16 +2568,17 @@ static void haswell_update_wm(struct drm_crtc *crtc)
 	partitioning = (best_lp_wm == &lp_wm_1_2) ?
 		       INTEL_DDB_PART_1_2 : INTEL_DDB_PART_5_6;
 
-	hsw_compute_wm_results(dev, best_lp_wm, partitioning, &results);
+	ilk_compute_wm_results(dev, best_lp_wm, partitioning, &results);
 
-	hsw_write_wm_values(dev_priv, &results);
+	ilk_write_wm_values(dev_priv, &results);
 }
 
-static void haswell_update_sprite_wm(struct drm_plane *plane,
+static void ilk_update_sprite_wm(struct drm_plane *plane,
 				     struct drm_crtc *crtc,
 				     uint32_t sprite_width, int pixel_size,
 				     bool enabled, bool scaled)
 {
+	struct drm_device *dev = plane->dev;
 	struct intel_plane *intel_plane = to_intel_plane(plane);
 
 	intel_plane->wm.enabled = enabled;
@@ -2978,176 +2586,24 @@ static void haswell_update_sprite_wm(struct drm_plane *plane,
 	intel_plane->wm.horiz_pixels = sprite_width;
 	intel_plane->wm.bytes_per_pixel = pixel_size;
 
-	haswell_update_wm(crtc);
-}
-
-static bool
-sandybridge_compute_sprite_wm(struct drm_device *dev, int plane,
-			      uint32_t sprite_width, int pixel_size,
-			      const struct intel_watermark_params *display,
-			      int display_latency_ns, int *sprite_wm)
-{
-	struct drm_crtc *crtc;
-	int clock;
-	int entries, tlb_miss;
-
-	crtc = intel_get_crtc_for_plane(dev, plane);
-	if (!intel_crtc_active(crtc)) {
-		*sprite_wm = display->guard_size;
-		return false;
-	}
-
-	clock = to_intel_crtc(crtc)->config.adjusted_mode.crtc_clock;
-
-	/* Use the small buffer method to calculate the sprite watermark */
-	entries = ((clock * pixel_size / 1000) * display_latency_ns) / 1000;
-	tlb_miss = display->fifo_size*display->cacheline_size -
-		sprite_width * 8;
-	if (tlb_miss > 0)
-		entries += tlb_miss;
-	entries = DIV_ROUND_UP(entries, display->cacheline_size);
-	*sprite_wm = entries + display->guard_size;
-	if (*sprite_wm > (int)display->max_wm)
-		*sprite_wm = display->max_wm;
-
-	return true;
-}
-
-static bool
-sandybridge_compute_sprite_srwm(struct drm_device *dev, int plane,
-				uint32_t sprite_width, int pixel_size,
-				const struct intel_watermark_params *display,
-				int latency_ns, int *sprite_wm)
-{
-	struct drm_crtc *crtc;
-	unsigned long line_time_us;
-	int clock;
-	int line_count, line_size;
-	int small, large;
-	int entries;
-
-	if (!latency_ns) {
-		*sprite_wm = 0;
-		return false;
-	}
-
-	crtc = intel_get_crtc_for_plane(dev, plane);
-	clock = to_intel_crtc(crtc)->config.adjusted_mode.crtc_clock;
-	if (!clock) {
-		*sprite_wm = 0;
-		return false;
-	}
-
-	line_time_us = (sprite_width * 1000) / clock;
-	if (!line_time_us) {
-		*sprite_wm = 0;
-		return false;
-	}
-
-	line_count = (latency_ns / line_time_us + 1000) / 1000;
-	line_size = sprite_width * pixel_size;
-
-	/* Use the minimum of the small and large buffer method for primary */
-	small = ((clock * pixel_size / 1000) * latency_ns) / 1000;
-	large = line_count * line_size;
-
-	entries = DIV_ROUND_UP(min(small, large), display->cacheline_size);
-	*sprite_wm = entries + display->guard_size;
-
-	return *sprite_wm > 0x3ff ? false : true;
-}
-
-static void sandybridge_update_sprite_wm(struct drm_plane *plane,
-					 struct drm_crtc *crtc,
-					 uint32_t sprite_width, int pixel_size,
-					 bool enabled, bool scaled)
-{
-	struct drm_device *dev = plane->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	int pipe = to_intel_plane(plane)->pipe;
-	int latency = dev_priv->wm.spr_latency[0] * 100;	/* In unit 0.1us */
-	u32 val;
-	int sprite_wm, reg;
-	int ret;
-
-	if (!enabled)
-		return;
-
-	switch (pipe) {
-	case 0:
-		reg = WM0_PIPEA_ILK;
-		break;
-	case 1:
-		reg = WM0_PIPEB_ILK;
-		break;
-	case 2:
-		reg = WM0_PIPEC_IVB;
-		break;
-	default:
-		return; /* bad pipe */
-	}
-
-	ret = sandybridge_compute_sprite_wm(dev, pipe, sprite_width, pixel_size,
-					    &sandybridge_display_wm_info,
-					    latency, &sprite_wm);
-	if (!ret) {
-		DRM_DEBUG_KMS("failed to compute sprite wm for pipe %c\n",
-			      pipe_name(pipe));
-		return;
-	}
-
-	val = I915_READ(reg);
-	val &= ~WM0_PIPE_SPRITE_MASK;
-	I915_WRITE(reg, val | (sprite_wm << WM0_PIPE_SPRITE_SHIFT));
-	DRM_DEBUG_KMS("sprite watermarks For pipe %c - %d\n", pipe_name(pipe), sprite_wm);
-
-
-	ret = sandybridge_compute_sprite_srwm(dev, pipe, sprite_width,
-					      pixel_size,
-					      &sandybridge_display_srwm_info,
-					      dev_priv->wm.spr_latency[1] * 500,
-					      &sprite_wm);
-	if (!ret) {
-		DRM_DEBUG_KMS("failed to compute sprite lp1 wm on pipe %c\n",
-			      pipe_name(pipe));
-		return;
-	}
-	I915_WRITE(WM1S_LP_ILK, sprite_wm);
-
-	/* Only IVB has two more LP watermarks for sprite */
-	if (!IS_IVYBRIDGE(dev))
-		return;
-
-	ret = sandybridge_compute_sprite_srwm(dev, pipe, sprite_width,
-					      pixel_size,
-					      &sandybridge_display_srwm_info,
-					      dev_priv->wm.spr_latency[2] * 500,
-					      &sprite_wm);
-	if (!ret) {
-		DRM_DEBUG_KMS("failed to compute sprite lp2 wm on pipe %c\n",
-			      pipe_name(pipe));
-		return;
-	}
-	I915_WRITE(WM2S_LP_IVB, sprite_wm);
+	/*
+	 * IVB workaround: must disable low power watermarks for at least
+	 * one frame before enabling scaling.  LP watermarks can be re-enabled
+	 * when scaling is disabled.
+	 *
+	 * WaCxSRDisabledForSpriteScaling:ivb
+	 */
+	if (IS_IVYBRIDGE(dev) && scaled && ilk_disable_lp_wm(dev))
+		intel_wait_for_vblank(dev, intel_plane->pipe);
 
-	ret = sandybridge_compute_sprite_srwm(dev, pipe, sprite_width,
-					      pixel_size,
-					      &sandybridge_display_srwm_info,
-					      dev_priv->wm.spr_latency[3] * 500,
-					      &sprite_wm);
-	if (!ret) {
-		DRM_DEBUG_KMS("failed to compute sprite lp3 wm on pipe %c\n",
-			      pipe_name(pipe));
-		return;
-	}
-	I915_WRITE(WM3S_LP_IVB, sprite_wm);
+	ilk_update_wm(crtc);
 }
 
 static void ilk_pipe_wm_get_hw_state(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct hsw_wm_values *hw = &dev_priv->wm.hw;
+	struct ilk_wm_values *hw = &dev_priv->wm.hw;
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	struct intel_pipe_wm *active = &intel_crtc->wm.active;
 	enum pipe pipe = intel_crtc->pipe;
@@ -3158,7 +2614,8 @@ static void ilk_pipe_wm_get_hw_state(struct drm_crtc *crtc)
 	};
 
 	hw->wm_pipe[pipe] = I915_READ(wm0_pipe_reg[pipe]);
-	hw->wm_linetime[pipe] = I915_READ(PIPE_WM_LINETIME(pipe));
+	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
+		hw->wm_linetime[pipe] = I915_READ(PIPE_WM_LINETIME(pipe));
 
 	if (intel_crtc_active(crtc)) {
 		u32 tmp = hw->wm_pipe[pipe];
@@ -3190,7 +2647,7 @@ static void ilk_pipe_wm_get_hw_state(struct drm_crtc *crtc)
 void ilk_wm_get_hw_state(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct hsw_wm_values *hw = &dev_priv->wm.hw;
+	struct ilk_wm_values *hw = &dev_priv->wm.hw;
 	struct drm_crtc *crtc;
 
 	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head)
@@ -3204,8 +2661,12 @@ void ilk_wm_get_hw_state(struct drm_device *dev)
 	hw->wm_lp_spr[1] = I915_READ(WM2S_LP_IVB);
 	hw->wm_lp_spr[2] = I915_READ(WM3S_LP_IVB);
 
-	hw->partitioning = (I915_READ(WM_MISC) & WM_MISC_DATA_PARTITION_5_6) ?
-		INTEL_DDB_PART_5_6 : INTEL_DDB_PART_1_2;
+	if (IS_HASWELL(dev) || IS_BROADWELL(dev))
+		hw->partitioning = (I915_READ(WM_MISC) & WM_MISC_DATA_PARTITION_5_6) ?
+			INTEL_DDB_PART_5_6 : INTEL_DDB_PART_1_2;
+	else if (IS_IVYBRIDGE(dev))
+		hw->partitioning = (I915_READ(DISP_ARB_CTL2) & DISP_DATA_PARTITION_5_6) ?
+			INTEL_DDB_PART_5_6 : INTEL_DDB_PART_1_2;
 
 	hw->enable_fbc_wm =
 		!(I915_READ(DISP_ARB_CTL) & DISP_FBC_WM_DIS);
@@ -3430,26 +2891,19 @@ static void ironlake_disable_drps(struct drm_device *dev)
  * ourselves, instead of doing a rmw cycle (which might result in us clearing
  * all limits and the gpu stuck at whatever frequency it is at atm).
  */
-static u32 gen6_rps_limits(struct drm_i915_private *dev_priv, u8 *val)
+static u32 gen6_rps_limits(struct drm_i915_private *dev_priv, u8 val)
 {
 	u32 limits;
 
-	limits = 0;
-
-	if (*val >= dev_priv->rps.max_delay)
-		*val = dev_priv->rps.max_delay;
-	limits |= dev_priv->rps.max_delay << 24;
-
 	/* Only set the down limit when we've reached the lowest level to avoid
 	 * getting more interrupts, otherwise leave this clear. This prevents a
 	 * race in the hw when coming out of rc6: There's a tiny window where
 	 * the hw runs at the minimal clock before selecting the desired
 	 * frequency, if the down threshold expires in that window we will not
 	 * receive a down interrupt. */
-	if (*val <= dev_priv->rps.min_delay) {
-		*val = dev_priv->rps.min_delay;
+	limits = dev_priv->rps.max_delay << 24;
+	if (val <= dev_priv->rps.min_delay)
 		limits |= dev_priv->rps.min_delay << 16;
-	}
 
 	return limits;
 }
@@ -3549,7 +3003,6 @@ static void gen6_set_rps_thresholds(struct drm_i915_private *dev_priv, u8 val)
 void gen6_set_rps(struct drm_device *dev, u8 val)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	u32 limits = gen6_rps_limits(dev_priv, &val);
 
 	WARN_ON(!mutex_is_locked(&dev_priv->rps.hw_lock));
 	WARN_ON(val > dev_priv->rps.max_delay);
@@ -3572,7 +3025,8 @@ void gen6_set_rps(struct drm_device *dev, u8 val)
 	/* Make sure we continue to get interrupts
 	 * until we hit the minimum or maximum frequencies.
 	 */
-	I915_WRITE(GEN6_RP_INTERRUPT_LIMITS, limits);
+	I915_WRITE(GEN6_RP_INTERRUPT_LIMITS,
+		   gen6_rps_limits(dev_priv, val));
 
 	POSTING_READ(GEN6_RPNSWREQ);
 
@@ -3583,9 +3037,11 @@ void gen6_set_rps(struct drm_device *dev, u8 val)
 
 void gen6_rps_idle(struct drm_i915_private *dev_priv)
 {
+	struct drm_device *dev = dev_priv->dev;
+
 	mutex_lock(&dev_priv->rps.hw_lock);
 	if (dev_priv->rps.enabled) {
-		if (dev_priv->info->is_valleyview)
+		if (IS_VALLEYVIEW(dev))
 			valleyview_set_rps(dev_priv->dev, dev_priv->rps.min_delay);
 		else
 			gen6_set_rps(dev_priv->dev, dev_priv->rps.min_delay);
@@ -3596,9 +3052,11 @@ void gen6_rps_idle(struct drm_i915_private *dev_priv)
 
 void gen6_rps_boost(struct drm_i915_private *dev_priv)
 {
+	struct drm_device *dev = dev_priv->dev;
+
 	mutex_lock(&dev_priv->rps.hw_lock);
 	if (dev_priv->rps.enabled) {
-		if (dev_priv->info->is_valleyview)
+		if (IS_VALLEYVIEW(dev))
 			valleyview_set_rps(dev_priv->dev, dev_priv->rps.max_delay);
 		else
 			gen6_set_rps(dev_priv->dev, dev_priv->rps.max_delay);
@@ -3607,48 +3065,18 @@ void gen6_rps_boost(struct drm_i915_private *dev_priv)
 	mutex_unlock(&dev_priv->rps.hw_lock);
 }
 
-/*
- * Wait until the previous freq change has completed,
- * or the timeout elapsed, and then update our notion
- * of the current GPU frequency.
- */
-static void vlv_update_rps_cur_delay(struct drm_i915_private *dev_priv)
-{
-	u32 pval;
-
-	WARN_ON(!mutex_is_locked(&dev_priv->rps.hw_lock));
-
-	if (wait_for(((pval = vlv_punit_read(dev_priv, PUNIT_REG_GPU_FREQ_STS)) & GENFREQSTATUS) == 0, 10))
-		DRM_DEBUG_DRIVER("timed out waiting for Punit\n");
-
-	pval >>= 8;
-
-	if (pval != dev_priv->rps.cur_delay)
-		DRM_DEBUG_DRIVER("Punit overrode GPU freq: %d MHz (%u) requested, but got %d Mhz (%u)\n",
-				 vlv_gpu_freq(dev_priv->mem_freq, dev_priv->rps.cur_delay),
-				 dev_priv->rps.cur_delay,
-				 vlv_gpu_freq(dev_priv->mem_freq, pval), pval);
-
-	dev_priv->rps.cur_delay = pval;
-}
-
 void valleyview_set_rps(struct drm_device *dev, u8 val)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	gen6_rps_limits(dev_priv, &val);
-
 	WARN_ON(!mutex_is_locked(&dev_priv->rps.hw_lock));
 	WARN_ON(val > dev_priv->rps.max_delay);
 	WARN_ON(val < dev_priv->rps.min_delay);
 
-	vlv_update_rps_cur_delay(dev_priv);
-
 	DRM_DEBUG_DRIVER("GPU freq request from %d MHz (%u) to %d MHz (%u)\n",
-			 vlv_gpu_freq(dev_priv->mem_freq,
-				      dev_priv->rps.cur_delay),
+			 vlv_gpu_freq(dev_priv, dev_priv->rps.cur_delay),
 			 dev_priv->rps.cur_delay,
-			 vlv_gpu_freq(dev_priv->mem_freq, val), val);
+			 vlv_gpu_freq(dev_priv, val), val);
 
 	if (val == dev_priv->rps.cur_delay)
 		return;
@@ -3657,7 +3085,7 @@ void valleyview_set_rps(struct drm_device *dev, u8 val)
 
 	dev_priv->rps.cur_delay = val;
 
-	trace_intel_gpu_freq_change(vlv_gpu_freq(dev_priv->mem_freq, val));
+	trace_intel_gpu_freq_change(vlv_gpu_freq(dev_priv, val));
 }
 
 static void gen6_disable_rps_interrupts(struct drm_device *dev)
@@ -3775,7 +3203,7 @@ static void gen8_enable_rps(struct drm_device *dev)
 
 	/* 1c & 1d: Get forcewake during program sequence. Although the driver
 	 * hasn't enabled a state yet where we need forcewake, BIOS may have.*/
-	gen6_gt_force_wake_get(dev_priv);
+	gen6_gt_force_wake_get(dev_priv, FORCEWAKE_ALL);
 
 	/* 2a: Disable RC states. */
 	I915_WRITE(GEN6_RC_CONTROL, 0);
@@ -3832,7 +3260,7 @@ static void gen8_enable_rps(struct drm_device *dev)
 
 	gen6_enable_rps_interrupts(dev);
 
-	gen6_gt_force_wake_put(dev_priv);
+	gen6_gt_force_wake_put(dev_priv, FORCEWAKE_ALL);
 }
 
 static void gen6_enable_rps(struct drm_device *dev)
@@ -3862,7 +3290,7 @@ static void gen6_enable_rps(struct drm_device *dev)
 		I915_WRITE(GTFIFODBG, gtfifodbg);
 	}
 
-	gen6_gt_force_wake_get(dev_priv);
+	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);
@@ -3954,7 +3382,7 @@ static void gen6_enable_rps(struct drm_device *dev)
 			DRM_ERROR("Couldn't fix incorrect rc6 voltage\n");
 	}
 
-	gen6_gt_force_wake_put(dev_priv);
+	gen6_gt_force_wake_put(dev_priv, FORCEWAKE_ALL);
 }
 
 void gen6_update_ring_freq(struct drm_device *dev)
@@ -4116,7 +3544,8 @@ static void valleyview_enable_rps(struct drm_device *dev)
 
 	valleyview_setup_pctx(dev);
 
-	gen6_gt_force_wake_get(dev_priv);
+	/* If VLV, Forcewake all wells, else re-direct to regular path */
+	gen6_gt_force_wake_get(dev_priv, FORCEWAKE_ALL);
 
 	I915_WRITE(GEN6_RP_UP_THRESHOLD, 59400);
 	I915_WRITE(GEN6_RP_DOWN_THRESHOLD, 245000);
@@ -4140,7 +3569,7 @@ static void valleyview_enable_rps(struct drm_device *dev)
 	for_each_ring(ring, dev_priv, i)
 		I915_WRITE(RING_MAX_IDLE(ring->mmio_base), 10);
 
-	I915_WRITE(GEN6_RC6_THRESHOLD, 0xc350);
+	I915_WRITE(GEN6_RC6_THRESHOLD, 0x557);
 
 	/* allows RC6 residency counter to work */
 	I915_WRITE(VLV_COUNTER_CONTROL,
@@ -4148,65 +3577,47 @@ static void valleyview_enable_rps(struct drm_device *dev)
 				      VLV_MEDIA_RC6_COUNT_EN |
 				      VLV_RENDER_RC6_COUNT_EN));
 	if (intel_enable_rc6(dev) & INTEL_RC6_ENABLE)
-		rc6_mode = GEN7_RC_CTL_TO_MODE;
+		rc6_mode = GEN7_RC_CTL_TO_MODE | VLV_RC_CTL_CTX_RST_PARALLEL;
 
 	intel_print_rc6_info(dev, rc6_mode);
 
 	I915_WRITE(GEN6_RC_CONTROL, rc6_mode);
 
 	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);
 
 	DRM_DEBUG_DRIVER("GPLL enabled? %s\n", val & 0x10 ? "yes" : "no");
 	DRM_DEBUG_DRIVER("GPU status: 0x%08x\n", val);
 
 	dev_priv->rps.cur_delay = (val >> 8) & 0xff;
 	DRM_DEBUG_DRIVER("current GPU freq: %d MHz (%u)\n",
-			 vlv_gpu_freq(dev_priv->mem_freq,
-				      dev_priv->rps.cur_delay),
+			 vlv_gpu_freq(dev_priv, dev_priv->rps.cur_delay),
 			 dev_priv->rps.cur_delay);
 
 	dev_priv->rps.max_delay = valleyview_rps_max_freq(dev_priv);
 	dev_priv->rps.hw_max = dev_priv->rps.max_delay;
 	DRM_DEBUG_DRIVER("max GPU freq: %d MHz (%u)\n",
-			 vlv_gpu_freq(dev_priv->mem_freq,
-				      dev_priv->rps.max_delay),
+			 vlv_gpu_freq(dev_priv, dev_priv->rps.max_delay),
 			 dev_priv->rps.max_delay);
 
 	dev_priv->rps.rpe_delay = valleyview_rps_rpe_freq(dev_priv);
 	DRM_DEBUG_DRIVER("RPe GPU freq: %d MHz (%u)\n",
-			 vlv_gpu_freq(dev_priv->mem_freq,
-				      dev_priv->rps.rpe_delay),
+			 vlv_gpu_freq(dev_priv, dev_priv->rps.rpe_delay),
 			 dev_priv->rps.rpe_delay);
 
 	dev_priv->rps.min_delay = valleyview_rps_min_freq(dev_priv);
 	DRM_DEBUG_DRIVER("min GPU freq: %d MHz (%u)\n",
-			 vlv_gpu_freq(dev_priv->mem_freq,
-				      dev_priv->rps.min_delay),
+			 vlv_gpu_freq(dev_priv, dev_priv->rps.min_delay),
 			 dev_priv->rps.min_delay);
 
 	DRM_DEBUG_DRIVER("setting GPU freq to %d MHz (%u)\n",
-			 vlv_gpu_freq(dev_priv->mem_freq,
-				      dev_priv->rps.rpe_delay),
+			 vlv_gpu_freq(dev_priv, dev_priv->rps.rpe_delay),
 			 dev_priv->rps.rpe_delay);
 
 	valleyview_set_rps(dev_priv->dev, dev_priv->rps.rpe_delay);
 
 	gen6_enable_rps_interrupts(dev);
 
-	gen6_gt_force_wake_put(dev_priv);
+	gen6_gt_force_wake_put(dev_priv, FORCEWAKE_ALL);
 }
 
 void ironlake_teardown_rc6(struct drm_device *dev)
@@ -5019,6 +4430,20 @@ static void g4x_disable_trickle_feed(struct drm_device *dev)
 	}
 }
 
+static void ilk_init_lp_watermarks(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	I915_WRITE(WM3_LP_ILK, I915_READ(WM3_LP_ILK) & ~WM1_LP_SR_EN);
+	I915_WRITE(WM2_LP_ILK, I915_READ(WM2_LP_ILK) & ~WM1_LP_SR_EN);
+	I915_WRITE(WM1_LP_ILK, I915_READ(WM1_LP_ILK) & ~WM1_LP_SR_EN);
+
+	/*
+	 * Don't touch WM1S_LP_EN here.
+	 * Doing so could cause underruns.
+	 */
+}
+
 static void ironlake_init_clock_gating(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -5052,9 +4477,8 @@ static void ironlake_init_clock_gating(struct drm_device *dev)
 	I915_WRITE(DISP_ARB_CTL,
 		   (I915_READ(DISP_ARB_CTL) |
 		    DISP_FBC_WM_DIS));
-	I915_WRITE(WM3_LP_ILK, 0);
-	I915_WRITE(WM2_LP_ILK, 0);
-	I915_WRITE(WM1_LP_ILK, 0);
+
+	ilk_init_lp_watermarks(dev);
 
 	/*
 	 * Based on the document from hardware guys the following bits
@@ -5161,9 +4585,7 @@ static void gen6_init_clock_gating(struct drm_device *dev)
 		I915_WRITE(GEN6_GT_MODE,
 			   _MASKED_BIT_ENABLE(GEN6_TD_FOUR_ROW_DISPATCH_DISABLE));
 
-	I915_WRITE(WM3_LP_ILK, 0);
-	I915_WRITE(WM2_LP_ILK, 0);
-	I915_WRITE(WM1_LP_ILK, 0);
+	ilk_init_lp_watermarks(dev);
 
 	I915_WRITE(CACHE_MODE_0,
 		   _MASKED_BIT_DISABLE(CM0_STC_EVICT_DISABLE_LRA_SNB));
@@ -5304,28 +4726,40 @@ static void gen8_init_clock_gating(struct drm_device *dev)
 	I915_WRITE(GEN7_HALF_SLICE_CHICKEN1,
 		   _MASKED_BIT_ENABLE(GEN7_SINGLE_SUBSCAN_DISPATCH_ENABLE));
 
-	/* WaSwitchSolVfFArbitrationPriority */
+	/* WaSwitchSolVfFArbitrationPriority:bdw */
 	I915_WRITE(GAM_ECOCHK, I915_READ(GAM_ECOCHK) | HSW_ECOCHK_ARB_PRIO_SOL);
 
-	/* WaPsrDPAMaskVBlankInSRD */
+	/* WaPsrDPAMaskVBlankInSRD:bdw */
 	I915_WRITE(CHICKEN_PAR1_1,
 		   I915_READ(CHICKEN_PAR1_1) | DPA_MASK_VBLANK_SRD);
 
-	/* WaPsrDPRSUnmaskVBlankInSRD */
+	/* WaPsrDPRSUnmaskVBlankInSRD:bdw */
 	for_each_pipe(i) {
 		I915_WRITE(CHICKEN_PIPESL_1(i),
 			   I915_READ(CHICKEN_PIPESL_1(i) |
 				     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));
 }
 
 static void haswell_init_clock_gating(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	I915_WRITE(WM3_LP_ILK, 0);
-	I915_WRITE(WM2_LP_ILK, 0);
-	I915_WRITE(WM1_LP_ILK, 0);
+	ilk_init_lp_watermarks(dev);
 
 	/* According to the spec, bit 13 (RCZUNIT) must be set on IVB.
 	 * This implements the WaDisableRCZUnitClockGating:hsw workaround.
@@ -5374,9 +4808,7 @@ static void ivybridge_init_clock_gating(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	uint32_t snpcr;
 
-	I915_WRITE(WM3_LP_ILK, 0);
-	I915_WRITE(WM2_LP_ILK, 0);
-	I915_WRITE(WM1_LP_ILK, 0);
+	ilk_init_lp_watermarks(dev);
 
 	I915_WRITE(ILK_DSPCLK_GATE_D, ILK_VRHUNIT_CLOCK_GATE_DISABLE);
 
@@ -5463,6 +4895,26 @@ 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:
+		dev_priv->mem_freq = 800;
+		break;
+	case 1:
+		dev_priv->mem_freq = 1066;
+		break;
+	case 2:
+		dev_priv->mem_freq = 1333;
+		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);
 
@@ -5642,50 +5094,133 @@ void intel_suspend_hw(struct drm_device *dev)
 		lpt_suspend_hw(dev);
 }
 
-static bool is_always_on_power_domain(struct drm_device *dev,
-				      enum intel_display_power_domain domain)
-{
-	unsigned long always_on_domains;
+#define for_each_power_well(i, power_well, domain_mask, power_domains)	\
+	for (i = 0;							\
+	     i < (power_domains)->power_well_count &&			\
+		 ((power_well) = &(power_domains)->power_wells[i]);	\
+	     i++)							\
+		if ((power_well)->domains & (domain_mask))
 
-	BUG_ON(BIT(domain) & ~POWER_DOMAIN_MASK);
-
-	if (IS_BROADWELL(dev)) {
-		always_on_domains = BDW_ALWAYS_ON_POWER_DOMAINS;
-	} else if (IS_HASWELL(dev)) {
-		always_on_domains = HSW_ALWAYS_ON_POWER_DOMAINS;
-	} else {
-		WARN_ON(1);
-		return true;
-	}
-
-	return BIT(domain) & always_on_domains;
-}
+#define for_each_power_well_rev(i, power_well, domain_mask, power_domains) \
+	for (i = (power_domains)->power_well_count - 1;			 \
+	     i >= 0 && ((power_well) = &(power_domains)->power_wells[i]);\
+	     i--)							 \
+		if ((power_well)->domains & (domain_mask))
 
 /**
  * We should only use the power well if we explicitly asked the hardware to
  * enable it, so check if it's enabled and also check if we've requested it to
  * be enabled.
  */
+static bool hsw_power_well_enabled(struct drm_device *dev,
+				   struct i915_power_well *power_well)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+
+	return I915_READ(HSW_PWR_WELL_DRIVER) ==
+		     (HSW_PWR_WELL_ENABLE_REQUEST | HSW_PWR_WELL_STATE_ENABLED);
+}
+
+bool intel_display_power_enabled_sw(struct drm_device *dev,
+				    enum intel_display_power_domain domain)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct i915_power_domains *power_domains;
+
+	power_domains = &dev_priv->power_domains;
+
+	return power_domains->domain_use_count[domain];
+}
+
 bool intel_display_power_enabled(struct drm_device *dev,
 				 enum intel_display_power_domain domain)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct i915_power_domains *power_domains;
+	struct i915_power_well *power_well;
+	bool is_enabled;
+	int i;
 
-	if (!HAS_POWER_WELL(dev))
-		return true;
+	power_domains = &dev_priv->power_domains;
 
-	if (is_always_on_power_domain(dev, domain))
-		return true;
+	is_enabled = true;
 
-	return I915_READ(HSW_PWR_WELL_DRIVER) ==
-		     (HSW_PWR_WELL_ENABLE_REQUEST | HSW_PWR_WELL_STATE_ENABLED);
+	mutex_lock(&power_domains->lock);
+	for_each_power_well_rev(i, power_well, BIT(domain), power_domains) {
+		if (power_well->always_on)
+			continue;
+
+		if (!power_well->is_enabled(dev, power_well)) {
+			is_enabled = false;
+			break;
+		}
+	}
+	mutex_unlock(&power_domains->lock);
+
+	return is_enabled;
+}
+
+static void hsw_power_well_post_enable(struct drm_i915_private *dev_priv)
+{
+	struct drm_device *dev = dev_priv->dev;
+	unsigned long irqflags;
+
+	/*
+	 * After we re-enable the power well, if we touch VGA register 0x3d5
+	 * we'll get unclaimed register interrupts. This stops after we write
+	 * anything to the VGA MSR register. The vgacon module uses this
+	 * register all the time, so if we unbind our driver and, as a
+	 * consequence, bind vgacon, we'll get stuck in an infinite loop at
+	 * console_unlock(). So make here we touch the VGA MSR register, making
+	 * sure vgacon can keep working normally without triggering interrupts
+	 * and error messages.
+	 */
+	vga_get_uninterruptible(dev->pdev, VGA_RSRC_LEGACY_IO);
+	outb(inb(VGA_MSR_READ), VGA_MSR_WRITE);
+	vga_put(dev->pdev, VGA_RSRC_LEGACY_IO);
+
+	if (IS_BROADWELL(dev)) {
+		spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
+		I915_WRITE(GEN8_DE_PIPE_IMR(PIPE_B),
+			   dev_priv->de_irq_mask[PIPE_B]);
+		I915_WRITE(GEN8_DE_PIPE_IER(PIPE_B),
+			   ~dev_priv->de_irq_mask[PIPE_B] |
+			   GEN8_PIPE_VBLANK);
+		I915_WRITE(GEN8_DE_PIPE_IMR(PIPE_C),
+			   dev_priv->de_irq_mask[PIPE_C]);
+		I915_WRITE(GEN8_DE_PIPE_IER(PIPE_C),
+			   ~dev_priv->de_irq_mask[PIPE_C] |
+			   GEN8_PIPE_VBLANK);
+		POSTING_READ(GEN8_DE_PIPE_IER(PIPE_C));
+		spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
+	}
 }
 
-static void __intel_set_power_well(struct drm_device *dev, bool enable)
+static void hsw_power_well_post_disable(struct drm_i915_private *dev_priv)
+{
+	struct drm_device *dev = dev_priv->dev;
+	enum pipe p;
+	unsigned long irqflags;
+
+	/*
+	 * After this, the registers on the pipes that are part of the power
+	 * well will become zero, so we have to adjust our counters according to
+	 * that.
+	 *
+	 * FIXME: Should we do this in general in drm_vblank_post_modeset?
+	 */
+	spin_lock_irqsave(&dev->vbl_lock, irqflags);
+	for_each_pipe(p)
+		if (p != PIPE_A)
+			dev->vblank[p].last = 0;
+	spin_unlock_irqrestore(&dev->vbl_lock, irqflags);
+}
+
+static void hsw_set_power_well(struct drm_device *dev,
+			       struct i915_power_well *power_well, bool enable)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	bool is_enabled, enable_requested;
-	unsigned long irqflags;
 	uint32_t tmp;
 
 	WARN_ON(dev_priv->pc8.enabled);
@@ -5706,42 +5241,14 @@ static void __intel_set_power_well(struct drm_device *dev, bool enable)
 				DRM_ERROR("Timeout enabling power well\n");
 		}
 
-		if (IS_BROADWELL(dev)) {
-			spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
-			I915_WRITE(GEN8_DE_PIPE_IMR(PIPE_B),
-				   dev_priv->de_irq_mask[PIPE_B]);
-			I915_WRITE(GEN8_DE_PIPE_IER(PIPE_B),
-				   ~dev_priv->de_irq_mask[PIPE_B] |
-				   GEN8_PIPE_VBLANK);
-			I915_WRITE(GEN8_DE_PIPE_IMR(PIPE_C),
-				   dev_priv->de_irq_mask[PIPE_C]);
-			I915_WRITE(GEN8_DE_PIPE_IER(PIPE_C),
-				   ~dev_priv->de_irq_mask[PIPE_C] |
-				   GEN8_PIPE_VBLANK);
-			POSTING_READ(GEN8_DE_PIPE_IER(PIPE_C));
-			spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
-		}
+		hsw_power_well_post_enable(dev_priv);
 	} else {
 		if (enable_requested) {
-			enum pipe p;
-
 			I915_WRITE(HSW_PWR_WELL_DRIVER, 0);
 			POSTING_READ(HSW_PWR_WELL_DRIVER);
 			DRM_DEBUG_KMS("Requesting to disable the power well\n");
 
-			/*
-			 * After this, the registers on the pipes that are part
-			 * of the power well will become zero, so we have to
-			 * adjust our counters according to that.
-			 *
-			 * FIXME: Should we do this in general in
-			 * drm_vblank_post_modeset?
-			 */
-			spin_lock_irqsave(&dev->vbl_lock, irqflags);
-			for_each_pipe(p)
-				if (p != PIPE_A)
-					dev->vblank[p].last = 0;
-			spin_unlock_irqrestore(&dev->vbl_lock, irqflags);
+			hsw_power_well_post_disable(dev_priv);
 		}
 	}
 }
@@ -5751,9 +5258,9 @@ static void __intel_power_well_get(struct drm_device *dev,
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	if (!power_well->count++) {
+	if (!power_well->count++ && power_well->set) {
 		hsw_disable_package_c8(dev_priv);
-		__intel_set_power_well(dev, true);
+		power_well->set(dev, power_well, true);
 	}
 }
 
@@ -5763,8 +5270,10 @@ static void __intel_power_well_put(struct drm_device *dev,
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
 	WARN_ON(!power_well->count);
-	if (!--power_well->count && i915_disable_power_well) {
-		__intel_set_power_well(dev, false);
+
+	if (!--power_well->count && power_well->set &&
+	    i915_disable_power_well) {
+		power_well->set(dev, power_well, false);
 		hsw_enable_package_c8(dev_priv);
 	}
 }
@@ -5774,17 +5283,18 @@ void intel_display_power_get(struct drm_device *dev,
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct i915_power_domains *power_domains;
-
-	if (!HAS_POWER_WELL(dev))
-		return;
-
-	if (is_always_on_power_domain(dev, domain))
-		return;
+	struct i915_power_well *power_well;
+	int i;
 
 	power_domains = &dev_priv->power_domains;
 
 	mutex_lock(&power_domains->lock);
-	__intel_power_well_get(dev, &power_domains->power_wells[0]);
+
+	for_each_power_well(i, power_well, BIT(domain), power_domains)
+		__intel_power_well_get(dev, power_well);
+
+	power_domains->domain_use_count[domain]++;
+
 	mutex_unlock(&power_domains->lock);
 }
 
@@ -5793,17 +5303,19 @@ void intel_display_power_put(struct drm_device *dev,
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct i915_power_domains *power_domains;
-
-	if (!HAS_POWER_WELL(dev))
-		return;
-
-	if (is_always_on_power_domain(dev, domain))
-		return;
+	struct i915_power_well *power_well;
+	int i;
 
 	power_domains = &dev_priv->power_domains;
 
 	mutex_lock(&power_domains->lock);
-	__intel_power_well_put(dev, &power_domains->power_wells[0]);
+
+	WARN_ON(!power_domains->domain_use_count[domain]);
+	power_domains->domain_use_count[domain]--;
+
+	for_each_power_well_rev(i, power_well, BIT(domain), power_domains)
+		__intel_power_well_put(dev, power_well);
+
 	mutex_unlock(&power_domains->lock);
 }
 
@@ -5819,10 +5331,7 @@ void i915_request_power_well(void)
 
 	dev_priv = container_of(hsw_pwr, struct drm_i915_private,
 				power_domains);
-
-	mutex_lock(&hsw_pwr->lock);
-	__intel_power_well_get(dev_priv->dev, &hsw_pwr->power_wells[0]);
-	mutex_unlock(&hsw_pwr->lock);
+	intel_display_power_get(dev_priv->dev, POWER_DOMAIN_AUDIO);
 }
 EXPORT_SYMBOL_GPL(i915_request_power_well);
 
@@ -5836,24 +5345,71 @@ void i915_release_power_well(void)
 
 	dev_priv = container_of(hsw_pwr, struct drm_i915_private,
 				power_domains);
-
-	mutex_lock(&hsw_pwr->lock);
-	__intel_power_well_put(dev_priv->dev, &hsw_pwr->power_wells[0]);
-	mutex_unlock(&hsw_pwr->lock);
+	intel_display_power_put(dev_priv->dev, POWER_DOMAIN_AUDIO);
 }
 EXPORT_SYMBOL_GPL(i915_release_power_well);
 
+static struct i915_power_well i9xx_always_on_power_well[] = {
+	{
+		.name = "always-on",
+		.always_on = 1,
+		.domains = POWER_DOMAIN_MASK,
+	},
+};
+
+static struct i915_power_well hsw_power_wells[] = {
+	{
+		.name = "always-on",
+		.always_on = 1,
+		.domains = HSW_ALWAYS_ON_POWER_DOMAINS,
+	},
+	{
+		.name = "display",
+		.domains = POWER_DOMAIN_MASK & ~HSW_ALWAYS_ON_POWER_DOMAINS,
+		.is_enabled = hsw_power_well_enabled,
+		.set = hsw_set_power_well,
+	},
+};
+
+static struct i915_power_well bdw_power_wells[] = {
+	{
+		.name = "always-on",
+		.always_on = 1,
+		.domains = BDW_ALWAYS_ON_POWER_DOMAINS,
+	},
+	{
+		.name = "display",
+		.domains = POWER_DOMAIN_MASK & ~BDW_ALWAYS_ON_POWER_DOMAINS,
+		.is_enabled = hsw_power_well_enabled,
+		.set = hsw_set_power_well,
+	},
+};
+
+#define set_power_wells(power_domains, __power_wells) ({		\
+	(power_domains)->power_wells = (__power_wells);			\
+	(power_domains)->power_well_count = ARRAY_SIZE(__power_wells);	\
+})
+
 int intel_power_domains_init(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct i915_power_domains *power_domains = &dev_priv->power_domains;
-	struct i915_power_well *power_well;
 
 	mutex_init(&power_domains->lock);
-	hsw_pwr = power_domains;
 
-	power_well = &power_domains->power_wells[0];
-	power_well->count = 0;
+	/*
+	 * The enabling order will be from lower to higher indexed wells,
+	 * the disabling order is reversed.
+	 */
+	if (IS_HASWELL(dev)) {
+		set_power_wells(power_domains, hsw_power_wells);
+		hsw_pwr = power_domains;
+	} else if (IS_BROADWELL(dev)) {
+		set_power_wells(power_domains, bdw_power_wells);
+		hsw_pwr = power_domains;
+	} else {
+		set_power_wells(power_domains, i9xx_always_on_power_well);
+	}
 
 	return 0;
 }
@@ -5868,15 +5424,13 @@ static void intel_power_domains_resume(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct i915_power_domains *power_domains = &dev_priv->power_domains;
 	struct i915_power_well *power_well;
-
-	if (!HAS_POWER_WELL(dev))
-		return;
+	int i;
 
 	mutex_lock(&power_domains->lock);
-
-	power_well = &power_domains->power_wells[0];
-	__intel_set_power_well(dev, power_well->count > 0);
-
+	for_each_power_well(i, power_well, POWER_DOMAIN_MASK, power_domains) {
+		if (power_well->set)
+			power_well->set(dev, power_well, power_well->count > 0);
+	}
 	mutex_unlock(&power_domains->lock);
 }
 
@@ -5890,13 +5444,13 @@ void intel_power_domains_init_hw(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	if (!HAS_POWER_WELL(dev))
-		return;
-
 	/* For now, we need the power well to be always enabled. */
 	intel_display_set_init_power(dev, true);
 	intel_power_domains_resume(dev);
 
+	if (!(IS_HASWELL(dev) || IS_BROADWELL(dev)))
+		return;
+
 	/* We're taking over the BIOS, so clear any requests made by it since
 	 * the driver is in charge now. */
 	if (I915_READ(HSW_PWR_WELL_BIOS) & HSW_PWR_WELL_ENABLE_REQUEST)
@@ -5914,31 +5468,86 @@ void intel_aux_display_runtime_put(struct drm_i915_private *dev_priv)
 	hsw_enable_package_c8(dev_priv);
 }
 
+void intel_runtime_pm_get(struct drm_i915_private *dev_priv)
+{
+	struct drm_device *dev = dev_priv->dev;
+	struct device *device = &dev->pdev->dev;
+
+	if (!HAS_RUNTIME_PM(dev))
+		return;
+
+	pm_runtime_get_sync(device);
+	WARN(dev_priv->pm.suspended, "Device still suspended.\n");
+}
+
+void intel_runtime_pm_put(struct drm_i915_private *dev_priv)
+{
+	struct drm_device *dev = dev_priv->dev;
+	struct device *device = &dev->pdev->dev;
+
+	if (!HAS_RUNTIME_PM(dev))
+		return;
+
+	pm_runtime_mark_last_busy(device);
+	pm_runtime_put_autosuspend(device);
+}
+
+void intel_init_runtime_pm(struct drm_i915_private *dev_priv)
+{
+	struct drm_device *dev = dev_priv->dev;
+	struct device *device = &dev->pdev->dev;
+
+	dev_priv->pm.suspended = false;
+
+	if (!HAS_RUNTIME_PM(dev))
+		return;
+
+	pm_runtime_set_active(device);
+
+	pm_runtime_set_autosuspend_delay(device, 10000); /* 10s */
+	pm_runtime_mark_last_busy(device);
+	pm_runtime_use_autosuspend(device);
+}
+
+void intel_fini_runtime_pm(struct drm_i915_private *dev_priv)
+{
+	struct drm_device *dev = dev_priv->dev;
+	struct device *device = &dev->pdev->dev;
+
+	if (!HAS_RUNTIME_PM(dev))
+		return;
+
+	/* Make sure we're not suspended first. */
+	pm_runtime_get_sync(device);
+	pm_runtime_disable(device);
+}
+
 /* Set up chip specific power management-related functions */
 void intel_init_pm(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
-	if (I915_HAS_FBC(dev)) {
-		if (HAS_PCH_SPLIT(dev)) {
+	if (HAS_FBC(dev)) {
+		if (INTEL_INFO(dev)->gen >= 7) {
 			dev_priv->display.fbc_enabled = ironlake_fbc_enabled;
-			if (IS_IVYBRIDGE(dev) || IS_HASWELL(dev))
-				dev_priv->display.enable_fbc =
-					gen7_enable_fbc;
-			else
-				dev_priv->display.enable_fbc =
-					ironlake_enable_fbc;
+			dev_priv->display.enable_fbc = gen7_enable_fbc;
+			dev_priv->display.disable_fbc = ironlake_disable_fbc;
+		} else if (INTEL_INFO(dev)->gen >= 5) {
+			dev_priv->display.fbc_enabled = ironlake_fbc_enabled;
+			dev_priv->display.enable_fbc = ironlake_enable_fbc;
 			dev_priv->display.disable_fbc = ironlake_disable_fbc;
 		} else if (IS_GM45(dev)) {
 			dev_priv->display.fbc_enabled = g4x_fbc_enabled;
 			dev_priv->display.enable_fbc = g4x_enable_fbc;
 			dev_priv->display.disable_fbc = g4x_disable_fbc;
-		} else if (IS_CRESTLINE(dev)) {
+		} else {
 			dev_priv->display.fbc_enabled = i8xx_fbc_enabled;
 			dev_priv->display.enable_fbc = i8xx_enable_fbc;
 			dev_priv->display.disable_fbc = i8xx_disable_fbc;
+
+			/* This value was pulled out of someone's hat */
+			I915_WRITE(FBC_CONTROL, 500 << FBC_CTL_INTERVAL_SHIFT);
 		}
-		/* 855GM needs testing */
 	}
 
 	/* For cxsr */
@@ -5951,58 +5560,27 @@ void intel_init_pm(struct drm_device *dev)
 	if (HAS_PCH_SPLIT(dev)) {
 		intel_setup_wm_latency(dev);
 
-		if (IS_GEN5(dev)) {
-			if (dev_priv->wm.pri_latency[1] &&
-			    dev_priv->wm.spr_latency[1] &&
-			    dev_priv->wm.cur_latency[1])
-				dev_priv->display.update_wm = ironlake_update_wm;
-			else {
-				DRM_DEBUG_KMS("Failed to get proper latency. "
-					      "Disable CxSR\n");
-				dev_priv->display.update_wm = NULL;
-			}
+		if ((IS_GEN5(dev) && dev_priv->wm.pri_latency[1] &&
+		     dev_priv->wm.spr_latency[1] && dev_priv->wm.cur_latency[1]) ||
+		    (!IS_GEN5(dev) && dev_priv->wm.pri_latency[0] &&
+		     dev_priv->wm.spr_latency[0] && dev_priv->wm.cur_latency[0])) {
+			dev_priv->display.update_wm = ilk_update_wm;
+			dev_priv->display.update_sprite_wm = ilk_update_sprite_wm;
+		} else {
+			DRM_DEBUG_KMS("Failed to read display plane latency. "
+				      "Disable CxSR\n");
+		}
+
+		if (IS_GEN5(dev))
 			dev_priv->display.init_clock_gating = ironlake_init_clock_gating;
-		} else if (IS_GEN6(dev)) {
-			if (dev_priv->wm.pri_latency[0] &&
-			    dev_priv->wm.spr_latency[0] &&
-			    dev_priv->wm.cur_latency[0]) {
-				dev_priv->display.update_wm = sandybridge_update_wm;
-				dev_priv->display.update_sprite_wm = sandybridge_update_sprite_wm;
-			} else {
-				DRM_DEBUG_KMS("Failed to read display plane latency. "
-					      "Disable CxSR\n");
-				dev_priv->display.update_wm = NULL;
-			}
+		else if (IS_GEN6(dev))
 			dev_priv->display.init_clock_gating = gen6_init_clock_gating;
-		} else if (IS_IVYBRIDGE(dev)) {
-			if (dev_priv->wm.pri_latency[0] &&
-			    dev_priv->wm.spr_latency[0] &&
-			    dev_priv->wm.cur_latency[0]) {
-				dev_priv->display.update_wm = ivybridge_update_wm;
-				dev_priv->display.update_sprite_wm = sandybridge_update_sprite_wm;
-			} else {
-				DRM_DEBUG_KMS("Failed to read display plane latency. "
-					      "Disable CxSR\n");
-				dev_priv->display.update_wm = NULL;
-			}
+		else if (IS_IVYBRIDGE(dev))
 			dev_priv->display.init_clock_gating = ivybridge_init_clock_gating;
-		} else if (IS_HASWELL(dev)) {
-			if (dev_priv->wm.pri_latency[0] &&
-			    dev_priv->wm.spr_latency[0] &&
-			    dev_priv->wm.cur_latency[0]) {
-				dev_priv->display.update_wm = haswell_update_wm;
-				dev_priv->display.update_sprite_wm =
-					haswell_update_sprite_wm;
-			} else {
-				DRM_DEBUG_KMS("Failed to read display plane latency. "
-					      "Disable CxSR\n");
-				dev_priv->display.update_wm = NULL;
-			}
+		else if (IS_HASWELL(dev))
 			dev_priv->display.init_clock_gating = haswell_init_clock_gating;
-		} else if (INTEL_INFO(dev)->gen == 8) {
+		else if (INTEL_INFO(dev)->gen == 8)
 			dev_priv->display.init_clock_gating = gen8_init_clock_gating;
-		} else
-			dev_priv->display.update_wm = NULL;
 	} else if (IS_VALLEYVIEW(dev)) {
 		dev_priv->display.update_wm = valleyview_update_wm;
 		dev_priv->display.init_clock_gating =
@@ -6036,21 +5614,21 @@ void intel_init_pm(struct drm_device *dev)
 		dev_priv->display.update_wm = i9xx_update_wm;
 		dev_priv->display.get_fifo_size = i9xx_get_fifo_size;
 		dev_priv->display.init_clock_gating = gen3_init_clock_gating;
-	} else if (IS_I865G(dev)) {
-		dev_priv->display.update_wm = i830_update_wm;
-		dev_priv->display.init_clock_gating = i85x_init_clock_gating;
-		dev_priv->display.get_fifo_size = i830_get_fifo_size;
-	} else if (IS_I85X(dev)) {
-		dev_priv->display.update_wm = i9xx_update_wm;
-		dev_priv->display.get_fifo_size = i85x_get_fifo_size;
-		dev_priv->display.init_clock_gating = i85x_init_clock_gating;
-	} else {
-		dev_priv->display.update_wm = i830_update_wm;
-		dev_priv->display.init_clock_gating = i830_init_clock_gating;
-		if (IS_845G(dev))
+	} else if (IS_GEN2(dev)) {
+		if (INTEL_INFO(dev)->num_pipes == 1) {
+			dev_priv->display.update_wm = i845_update_wm;
 			dev_priv->display.get_fifo_size = i845_get_fifo_size;
-		else
+		} else {
+			dev_priv->display.update_wm = i9xx_update_wm;
 			dev_priv->display.get_fifo_size = i830_get_fifo_size;
+		}
+
+		if (IS_I85X(dev) || IS_I865G(dev))
+			dev_priv->display.init_clock_gating = i85x_init_clock_gating;
+		else
+			dev_priv->display.init_clock_gating = i830_init_clock_gating;
+	} else {
+		DRM_ERROR("unexpected fall-through in intel_init_pm\n");
 	}
 }
 
@@ -6101,59 +5679,48 @@ int sandybridge_pcode_write(struct drm_i915_private *dev_priv, u8 mbox, u32 val)
 	return 0;
 }
 
-int vlv_gpu_freq(int ddr_freq, int val)
+int vlv_gpu_freq(struct drm_i915_private *dev_priv, int val)
 {
-	int mult, base;
+	int div;
 
-	switch (ddr_freq) {
+	/* 4 x czclk */
+	switch (dev_priv->mem_freq) {
 	case 800:
-		mult = 20;
-		base = 120;
+		div = 10;
 		break;
 	case 1066:
-		mult = 22;
-		base = 133;
+		div = 12;
 		break;
 	case 1333:
-		mult = 21;
-		base = 125;
+		div = 16;
 		break;
 	default:
 		return -1;
 	}
 
-	return ((val - 0xbd) * mult) + base;
+	return DIV_ROUND_CLOSEST(dev_priv->mem_freq * (val + 6 - 0xbd), 4 * div);
 }
 
-int vlv_freq_opcode(int ddr_freq, int val)
+int vlv_freq_opcode(struct drm_i915_private *dev_priv, int val)
 {
-	int mult, base;
+	int mul;
 
-	switch (ddr_freq) {
+	/* 4 x czclk */
+	switch (dev_priv->mem_freq) {
 	case 800:
-		mult = 20;
-		base = 120;
+		mul = 10;
 		break;
 	case 1066:
-		mult = 22;
-		base = 133;
+		mul = 12;
 		break;
 	case 1333:
-		mult = 21;
-		base = 125;
+		mul = 16;
 		break;
 	default:
 		return -1;
 	}
 
-	val /= mult;
-	val -= base / mult;
-	val += 0xbd;
-
-	if (val > 0xea)
-		val = 0xea;
-
-	return val;
+	return DIV_ROUND_CLOSEST(4 * mul * val, dev_priv->mem_freq) + 0xbd - 6;
 }
 
 void intel_pm_setup(struct drm_device *dev)
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index c2f09d456300..b7f1742caf87 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -285,14 +285,16 @@ static int gen7_ring_fbc_flush(struct intel_ring_buffer *ring, u32 value)
 	if (!ring->fbc_dirty)
 		return 0;
 
-	ret = intel_ring_begin(ring, 4);
+	ret = intel_ring_begin(ring, 6);
 	if (ret)
 		return ret;
-	intel_ring_emit(ring, MI_NOOP);
 	/* WaFbcNukeOn3DBlt:ivb/hsw */
 	intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(1));
 	intel_ring_emit(ring, MSG_FBC_REND_STATE);
 	intel_ring_emit(ring, value);
+	intel_ring_emit(ring, MI_STORE_REGISTER_MEM(1) | MI_SRM_LRM_GLOBAL_GTT);
+	intel_ring_emit(ring, MSG_FBC_REND_STATE);
+	intel_ring_emit(ring, ring->scratch.gtt_offset + 256);
 	intel_ring_advance(ring);
 
 	ring->fbc_dirty = false;
@@ -354,7 +356,7 @@ gen7_render_ring_flush(struct intel_ring_buffer *ring,
 	intel_ring_emit(ring, 0);
 	intel_ring_advance(ring);
 
-	if (flush_domains)
+	if (!invalidate_domains && flush_domains)
 		return gen7_ring_fbc_flush(ring, FBC_REND_NUKE);
 
 	return 0;
@@ -436,7 +438,7 @@ static int init_ring_common(struct intel_ring_buffer *ring)
 	int ret = 0;
 	u32 head;
 
-	gen6_gt_force_wake_get(dev_priv);
+	gen6_gt_force_wake_get(dev_priv, FORCEWAKE_ALL);
 
 	if (I915_NEED_GFX_HWS(dev))
 		intel_ring_setup_status_page(ring);
@@ -509,7 +511,7 @@ static int init_ring_common(struct intel_ring_buffer *ring)
 	memset(&ring->hangcheck, 0, sizeof(ring->hangcheck));
 
 out:
-	gen6_gt_force_wake_put(dev_priv);
+	gen6_gt_force_wake_put(dev_priv, FORCEWAKE_ALL);
 
 	return ret;
 }
@@ -661,19 +663,22 @@ gen6_add_request(struct intel_ring_buffer *ring)
 	struct drm_device *dev = ring->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_ring_buffer *useless;
-	int i, ret;
+	int i, ret, num_dwords = 4;
 
-	ret = intel_ring_begin(ring, ((I915_NUM_RINGS-1) *
-				      MBOX_UPDATE_DWORDS) +
-				      4);
+	if (i915_semaphore_is_enabled(dev))
+		num_dwords += ((I915_NUM_RINGS-1) * MBOX_UPDATE_DWORDS);
+#undef MBOX_UPDATE_DWORDS
+
+	ret = intel_ring_begin(ring, num_dwords);
 	if (ret)
 		return ret;
-#undef MBOX_UPDATE_DWORDS
 
-	for_each_ring(useless, dev_priv, i) {
-		u32 mbox_reg = ring->signal_mbox[i];
-		if (mbox_reg != GEN6_NOSYNC)
-			update_mboxes(ring, mbox_reg);
+	if (i915_semaphore_is_enabled(dev)) {
+		for_each_ring(useless, dev_priv, i) {
+			u32 mbox_reg = ring->signal_mbox[i];
+			if (mbox_reg != GEN6_NOSYNC)
+				update_mboxes(ring, mbox_reg);
+		}
 	}
 
 	intel_ring_emit(ring, MI_STORE_DWORD_INDEX);
@@ -1030,11 +1035,6 @@ gen6_ring_get_irq(struct intel_ring_buffer *ring)
 	if (!dev->irq_enabled)
 	       return false;
 
-	/* It looks like we need to prevent the gt from suspending while waiting
-	 * for an notifiy irq, otherwise irqs seem to get lost on at least the
-	 * blt/bsd rings on ivb. */
-	gen6_gt_force_wake_get(dev_priv);
-
 	spin_lock_irqsave(&dev_priv->irq_lock, flags);
 	if (ring->irq_refcount++ == 0) {
 		if (HAS_L3_DPF(dev) && ring->id == RCS)
@@ -1066,8 +1066,6 @@ gen6_ring_put_irq(struct intel_ring_buffer *ring)
 		ilk_disable_gt_irq(dev_priv, ring->irq_enable_mask);
 	}
 	spin_unlock_irqrestore(&dev_priv->irq_lock, flags);
-
-	gen6_gt_force_wake_put(dev_priv);
 }
 
 static bool
@@ -1611,8 +1609,8 @@ intel_ring_alloc_seqno(struct intel_ring_buffer *ring)
 	return i915_gem_get_seqno(ring->dev, &ring->outstanding_lazy_seqno);
 }
 
-static int __intel_ring_begin(struct intel_ring_buffer *ring,
-			      int bytes)
+static int __intel_ring_prepare(struct intel_ring_buffer *ring,
+				int bytes)
 {
 	int ret;
 
@@ -1628,7 +1626,6 @@ static int __intel_ring_begin(struct intel_ring_buffer *ring,
 			return ret;
 	}
 
-	ring->space -= bytes;
 	return 0;
 }
 
@@ -1643,12 +1640,17 @@ int intel_ring_begin(struct intel_ring_buffer *ring,
 	if (ret)
 		return ret;
 
+	ret = __intel_ring_prepare(ring, num_dwords * sizeof(uint32_t));
+	if (ret)
+		return ret;
+
 	/* Preallocate the olr before touching the ring */
 	ret = intel_ring_alloc_seqno(ring);
 	if (ret)
 		return ret;
 
-	return __intel_ring_begin(ring, num_dwords * sizeof(uint32_t));
+	ring->space -= num_dwords * sizeof(uint32_t);
+	return 0;
 }
 
 void intel_ring_init_seqno(struct intel_ring_buffer *ring, u32 seqno)
@@ -1838,7 +1840,7 @@ static int gen6_ring_flush(struct intel_ring_buffer *ring,
 	}
 	intel_ring_advance(ring);
 
-	if (IS_GEN7(dev) && flush)
+	if (IS_GEN7(dev) && !invalidate && flush)
 		return gen7_ring_fbc_flush(ring, FBC_REND_CACHE_CLEAN);
 
 	return 0;
diff --git a/drivers/gpu/drm/i915/intel_sdvo.c b/drivers/gpu/drm/i915/intel_sdvo.c
index a583e8f718a7..95bdfb3c431c 100644
--- a/drivers/gpu/drm/i915/intel_sdvo.c
+++ b/drivers/gpu/drm/i915/intel_sdvo.c
@@ -413,23 +413,34 @@ static const struct _sdvo_cmd_name {
 static void intel_sdvo_debug_write(struct intel_sdvo *intel_sdvo, u8 cmd,
 				   const void *args, int args_len)
 {
-	int i;
+	int i, pos = 0;
+#define BUF_LEN 256
+	char buffer[BUF_LEN];
+
+#define BUF_PRINT(args...) \
+	pos += snprintf(buffer + pos, max_t(int, BUF_LEN - pos, 0), args)
+
 
-	DRM_DEBUG_KMS("%s: W: %02X ",
-				SDVO_NAME(intel_sdvo), cmd);
-	for (i = 0; i < args_len; i++)
-		DRM_LOG_KMS("%02X ", ((u8 *)args)[i]);
-	for (; i < 8; i++)
-		DRM_LOG_KMS("   ");
+	for (i = 0; i < args_len; i++) {
+		BUF_PRINT("%02X ", ((u8 *)args)[i]);
+	}
+	for (; i < 8; i++) {
+		BUF_PRINT("   ");
+	}
 	for (i = 0; i < ARRAY_SIZE(sdvo_cmd_names); i++) {
 		if (cmd == sdvo_cmd_names[i].cmd) {
-			DRM_LOG_KMS("(%s)", sdvo_cmd_names[i].name);
+			BUF_PRINT("(%s)", sdvo_cmd_names[i].name);
 			break;
 		}
 	}
-	if (i == ARRAY_SIZE(sdvo_cmd_names))
-		DRM_LOG_KMS("(%02X)", cmd);
-	DRM_LOG_KMS("\n");
+	if (i == ARRAY_SIZE(sdvo_cmd_names)) {
+		BUF_PRINT("(%02X)", cmd);
+	}
+	BUG_ON(pos >= BUF_LEN - 1);
+#undef BUF_PRINT
+#undef BUF_LEN
+
+	DRM_DEBUG_KMS("%s: W: %02X %s\n", SDVO_NAME(intel_sdvo), cmd, buffer);
 }
 
 static const char *cmd_status_names[] = {
@@ -512,9 +523,10 @@ static bool intel_sdvo_read_response(struct intel_sdvo *intel_sdvo,
 {
 	u8 retry = 15; /* 5 quick checks, followed by 10 long checks */
 	u8 status;
-	int i;
+	int i, pos = 0;
+#define BUF_LEN 256
+	char buffer[BUF_LEN];
 
-	DRM_DEBUG_KMS("%s: R: ", SDVO_NAME(intel_sdvo));
 
 	/*
 	 * The documentation states that all commands will be
@@ -551,10 +563,13 @@ static bool intel_sdvo_read_response(struct intel_sdvo *intel_sdvo,
 			goto log_fail;
 	}
 
+#define BUF_PRINT(args...) \
+	pos += snprintf(buffer + pos, max_t(int, BUF_LEN - pos, 0), args)
+
 	if (status <= SDVO_CMD_STATUS_SCALING_NOT_SUPP)
-		DRM_LOG_KMS("(%s)", cmd_status_names[status]);
+		BUF_PRINT("(%s)", cmd_status_names[status]);
 	else
-		DRM_LOG_KMS("(??? %d)", status);
+		BUF_PRINT("(??? %d)", status);
 
 	if (status != SDVO_CMD_STATUS_SUCCESS)
 		goto log_fail;
@@ -565,13 +580,17 @@ static bool intel_sdvo_read_response(struct intel_sdvo *intel_sdvo,
 					  SDVO_I2C_RETURN_0 + i,
 					  &((u8 *)response)[i]))
 			goto log_fail;
-		DRM_LOG_KMS(" %02X", ((u8 *)response)[i]);
+		BUF_PRINT(" %02X", ((u8 *)response)[i]);
 	}
-	DRM_LOG_KMS("\n");
+	BUG_ON(pos >= BUF_LEN - 1);
+#undef BUF_PRINT
+#undef BUF_LEN
+
+	DRM_DEBUG_KMS("%s: R: %s\n", SDVO_NAME(intel_sdvo), buffer);
 	return true;
 
 log_fail:
-	DRM_LOG_KMS("... failed\n");
+	DRM_DEBUG_KMS("%s: R: ... failed\n", SDVO_NAME(intel_sdvo));
 	return false;
 }
 
@@ -933,7 +952,7 @@ static void intel_sdvo_dump_hdmi_buf(struct intel_sdvo *intel_sdvo)
 
 static bool intel_sdvo_write_infoframe(struct intel_sdvo *intel_sdvo,
 				       unsigned if_index, uint8_t tx_rate,
-				       uint8_t *data, unsigned length)
+				       const uint8_t *data, unsigned length)
 {
 	uint8_t set_buf_index[2] = { if_index, 0 };
 	uint8_t hbuf_size, tmp[8];
@@ -1517,8 +1536,9 @@ static void intel_sdvo_dpms(struct drm_connector *connector, int mode)
 	intel_modeset_check_state(connector->dev);
 }
 
-static int intel_sdvo_mode_valid(struct drm_connector *connector,
-				 struct drm_display_mode *mode)
+static enum drm_mode_status
+intel_sdvo_mode_valid(struct drm_connector *connector,
+		      struct drm_display_mode *mode)
 {
 	struct intel_sdvo *intel_sdvo = intel_attached_sdvo(connector);
 
diff --git a/drivers/gpu/drm/i915/intel_sdvo_regs.h b/drivers/gpu/drm/i915/intel_sdvo_regs.h
index 770bdd6ecd9f..2e2d4eb4a00d 100644
--- a/drivers/gpu/drm/i915/intel_sdvo_regs.h
+++ b/drivers/gpu/drm/i915/intel_sdvo_regs.h
@@ -59,7 +59,7 @@ struct intel_sdvo_caps {
 	unsigned int stall_support:1;
 	unsigned int pad:1;
 	u16 output_flags;
-} __attribute__((packed));
+} __packed;
 
 /* Note: SDVO detailed timing flags match EDID misc flags. */
 #define DTD_FLAG_HSYNC_POSITIVE (1 << 1)
@@ -94,12 +94,12 @@ struct intel_sdvo_dtd {
 		u8 v_sync_off_high;
 		u8 reserved;
 	} part2;
-} __attribute__((packed));
+} __packed;
 
 struct intel_sdvo_pixel_clock_range {
 	u16 min;	/**< pixel clock, in 10kHz units */
 	u16 max;	/**< pixel clock, in 10kHz units */
-} __attribute__((packed));
+} __packed;
 
 struct intel_sdvo_preferred_input_timing_args {
 	u16 clock;
@@ -108,7 +108,7 @@ struct intel_sdvo_preferred_input_timing_args {
 	u8	interlace:1;
 	u8	scaled:1;
 	u8	pad:6;
-} __attribute__((packed));
+} __packed;
 
 /* I2C registers for SDVO */
 #define SDVO_I2C_ARG_0				0x07
@@ -162,7 +162,7 @@ struct intel_sdvo_get_trained_inputs_response {
 	unsigned int input0_trained:1;
 	unsigned int input1_trained:1;
 	unsigned int pad:6;
-} __attribute__((packed));
+} __packed;
 
 /** Returns a struct intel_sdvo_output_flags of active outputs. */
 #define SDVO_CMD_GET_ACTIVE_OUTPUTS			0x04
@@ -219,7 +219,7 @@ struct intel_sdvo_get_interrupt_event_source_response {
 	unsigned int ambient_light_interrupt:1;
 	unsigned int hdmi_audio_encrypt_change:1;
 	unsigned int pad:6;
-} __attribute__((packed));
+} __packed;
 
 /**
  * Selects which input is affected by future input commands.
@@ -232,7 +232,7 @@ struct intel_sdvo_get_interrupt_event_source_response {
 struct intel_sdvo_set_target_input_args {
 	unsigned int target_1:1;
 	unsigned int pad:7;
-} __attribute__((packed));
+} __packed;
 
 /**
  * Takes a struct intel_sdvo_output_flags of which outputs are targeted by
@@ -370,7 +370,7 @@ struct intel_sdvo_tv_format {
 	unsigned int hdtv_std_eia_7702a_480i_60:1;
 	unsigned int hdtv_std_eia_7702a_480p_60:1;
 	unsigned int pad:3;
-} __attribute__((packed));
+} __packed;
 
 #define SDVO_CMD_GET_TV_FORMAT				0x28
 
@@ -401,7 +401,7 @@ struct intel_sdvo_sdtv_resolution_request {
 	unsigned int secam_l:1;
 	unsigned int secam_60:1;
 	unsigned int pad:5;
-} __attribute__((packed));
+} __packed;
 
 struct intel_sdvo_sdtv_resolution_reply {
 	unsigned int res_320x200:1;
@@ -426,7 +426,7 @@ struct intel_sdvo_sdtv_resolution_reply {
 	unsigned int res_1024x768:1;
 	unsigned int res_1280x1024:1;
 	unsigned int pad:5;
-} __attribute__((packed));
+} __packed;
 
 /* Get supported resolution with squire pixel aspect ratio that can be
    scaled for the requested HDTV format */
@@ -463,7 +463,7 @@ struct intel_sdvo_hdtv_resolution_request {
 	unsigned int hdtv_std_eia_7702a_480i_60:1;
 	unsigned int hdtv_std_eia_7702a_480p_60:1;
 	unsigned int pad:6;
-} __attribute__((packed));
+} __packed;
 
 struct intel_sdvo_hdtv_resolution_reply {
 	unsigned int res_640x480:1;
@@ -517,7 +517,7 @@ struct intel_sdvo_hdtv_resolution_reply {
 
 	unsigned int res_1280x768:1;
 	unsigned int pad5:7;
-} __attribute__((packed));
+} __packed;
 
 /* Get supported power state returns info for encoder and monitor, rely on
    last SetTargetInput and SetTargetOutput calls */
@@ -557,13 +557,13 @@ struct sdvo_panel_power_sequencing {
 
 	unsigned int t4_high:2;
 	unsigned int pad:6;
-} __attribute__((packed));
+} __packed;
 
 #define SDVO_CMD_GET_MAX_BACKLIGHT_LEVEL		0x30
 struct sdvo_max_backlight_reply {
 	u8 max_value;
 	u8 default_value;
-} __attribute__((packed));
+} __packed;
 
 #define SDVO_CMD_GET_BACKLIGHT_LEVEL			0x31
 #define SDVO_CMD_SET_BACKLIGHT_LEVEL			0x32
@@ -573,14 +573,14 @@ struct sdvo_get_ambient_light_reply {
 	u16 trip_low;
 	u16 trip_high;
 	u16 value;
-} __attribute__((packed));
+} __packed;
 #define SDVO_CMD_SET_AMBIENT_LIGHT			0x34
 struct sdvo_set_ambient_light_reply {
 	u16 trip_low;
 	u16 trip_high;
 	unsigned int enable:1;
 	unsigned int pad:7;
-} __attribute__((packed));
+} __packed;
 
 /* Set display power state */
 #define SDVO_CMD_SET_DISPLAY_POWER_STATE		0x7d
@@ -608,7 +608,7 @@ struct intel_sdvo_enhancements_reply {
 	unsigned int dither:1;
 	unsigned int tv_chroma_filter:1;
 	unsigned int tv_luma_filter:1;
-} __attribute__((packed));
+} __packed;
 
 /* Picture enhancement limits below are dependent on the current TV format,
  * and thus need to be queried and set after it.
@@ -630,7 +630,7 @@ struct intel_sdvo_enhancements_reply {
 struct intel_sdvo_enhancement_limits_reply {
 	u16 max_value;
 	u16 default_value;
-} __attribute__((packed));
+} __packed;
 
 #define SDVO_CMD_GET_LVDS_PANEL_INFORMATION		0x7f
 #define SDVO_CMD_SET_LVDS_PANEL_INFORMATION		0x80
@@ -671,7 +671,7 @@ struct intel_sdvo_enhancement_limits_reply {
 #define SDVO_CMD_SET_TV_LUMA_FILTER			0x79
 struct intel_sdvo_enhancements_arg {
 	u16 value;
-} __attribute__((packed));
+} __packed;
 
 #define SDVO_CMD_GET_DOT_CRAWL				0x70
 #define SDVO_CMD_SET_DOT_CRAWL				0x71
@@ -727,4 +727,4 @@ struct intel_sdvo_enhancements_arg {
 struct intel_sdvo_encode {
 	u8 dvi_rev;
 	u8 hdmi_rev;
-} __attribute__ ((packed));
+} __packed;
diff --git a/drivers/gpu/drm/i915/intel_sideband.c b/drivers/gpu/drm/i915/intel_sideband.c
index 9944d8135e87..0954f132726e 100644
--- a/drivers/gpu/drm/i915/intel_sideband.c
+++ b/drivers/gpu/drm/i915/intel_sideband.c
@@ -90,6 +90,22 @@ void vlv_punit_write(struct drm_i915_private *dev_priv, u8 addr, u32 val)
 	mutex_unlock(&dev_priv->dpio_lock);
 }
 
+u32 vlv_bunit_read(struct drm_i915_private *dev_priv, u32 reg)
+{
+	u32 val = 0;
+
+	vlv_sideband_rw(dev_priv, PCI_DEVFN(2, 0), IOSF_PORT_BUNIT,
+			PUNIT_OPCODE_REG_READ, reg, &val);
+
+	return val;
+}
+
+void vlv_bunit_write(struct drm_i915_private *dev_priv, u32 reg, u32 val)
+{
+	vlv_sideband_rw(dev_priv, PCI_DEVFN(2, 0), IOSF_PORT_BUNIT,
+			PUNIT_OPCODE_REG_WRITE, reg, &val);
+}
+
 u32 vlv_nc_read(struct drm_i915_private *dev_priv, u8 addr)
 {
 	u32 val = 0;
@@ -160,27 +176,18 @@ void vlv_gps_core_write(struct drm_i915_private *dev_priv, u32 reg, u32 val)
 			PUNIT_OPCODE_REG_WRITE, reg, &val);
 }
 
-static u32 vlv_get_phy_port(enum pipe pipe)
-{
-	u32 port = IOSF_PORT_DPIO;
-
-	WARN_ON ((pipe != PIPE_A) && (pipe != PIPE_B));
-
-	return port;
-}
-
 u32 vlv_dpio_read(struct drm_i915_private *dev_priv, enum pipe pipe, int reg)
 {
 	u32 val = 0;
 
-	vlv_sideband_rw(dev_priv, DPIO_DEVFN, vlv_get_phy_port(pipe),
+	vlv_sideband_rw(dev_priv, DPIO_DEVFN, DPIO_PHY_IOSF_PORT(DPIO_PHY(pipe)),
 			DPIO_OPCODE_REG_READ, reg, &val);
 	return val;
 }
 
 void vlv_dpio_write(struct drm_i915_private *dev_priv, enum pipe pipe, int reg, u32 val)
 {
-	vlv_sideband_rw(dev_priv, DPIO_DEVFN, vlv_get_phy_port(pipe),
+	vlv_sideband_rw(dev_priv, DPIO_DEVFN, DPIO_PHY_IOSF_PORT(DPIO_PHY(pipe)),
 			DPIO_OPCODE_REG_WRITE, reg, &val);
 }
 
@@ -242,3 +249,17 @@ void intel_sbi_write(struct drm_i915_private *dev_priv, u16 reg, u32 value,
 		return;
 	}
 }
+
+u32 vlv_flisdsi_read(struct drm_i915_private *dev_priv, u32 reg)
+{
+	u32 val = 0;
+	vlv_sideband_rw(dev_priv, DPIO_DEVFN, IOSF_PORT_FLISDSI,
+					DPIO_OPCODE_REG_READ, reg, &val);
+	return val;
+}
+
+void vlv_flisdsi_write(struct drm_i915_private *dev_priv, u32 reg, u32 val)
+{
+	vlv_sideband_rw(dev_priv, DPIO_DEVFN, IOSF_PORT_FLISDSI,
+					DPIO_OPCODE_REG_WRITE, reg, &val);
+}
diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c
index b9fabf826f7d..716a3c9c0751 100644
--- a/drivers/gpu/drm/i915/intel_sprite.c
+++ b/drivers/gpu/drm/i915/intel_sprite.c
@@ -104,6 +104,12 @@ vlv_update_plane(struct drm_plane *dplane, struct drm_crtc *crtc,
 		break;
 	}
 
+	/*
+	 * Enable gamma to match primary/cursor plane behaviour.
+	 * FIXME should be user controllable via propertiesa.
+	 */
+	sprctl |= SP_GAMMA_ENABLE;
+
 	if (obj->tiling_mode != I915_TILING_NONE)
 		sprctl |= SP_TILED;
 
@@ -135,8 +141,8 @@ vlv_update_plane(struct drm_plane *dplane, struct drm_crtc *crtc,
 
 	I915_WRITE(SPSIZE(pipe, plane), (crtc_h << 16) | crtc_w);
 	I915_WRITE(SPCNTR(pipe, plane), sprctl);
-	I915_MODIFY_DISPBASE(SPSURF(pipe, plane), i915_gem_obj_ggtt_offset(obj) +
-			     sprsurf_offset);
+	I915_WRITE(SPSURF(pipe, plane), i915_gem_obj_ggtt_offset(obj) +
+		   sprsurf_offset);
 	POSTING_READ(SPSURF(pipe, plane));
 }
 
@@ -152,7 +158,7 @@ vlv_disable_plane(struct drm_plane *dplane, struct drm_crtc *crtc)
 	I915_WRITE(SPCNTR(pipe, plane), I915_READ(SPCNTR(pipe, plane)) &
 		   ~SP_ENABLE);
 	/* Activate double buffered register update */
-	I915_MODIFY_DISPBASE(SPSURF(pipe, plane), 0);
+	I915_WRITE(SPSURF(pipe, plane), 0);
 	POSTING_READ(SPSURF(pipe, plane));
 
 	intel_update_sprite_watermarks(dplane, crtc, 0, 0, false, false);
@@ -224,7 +230,6 @@ ivb_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 	u32 sprctl, sprscale = 0;
 	unsigned long sprsurf_offset, linear_offset;
 	int pixel_size = drm_format_plane_cpp(fb->pixel_format, 0);
-	bool scaling_was_enabled = dev_priv->sprite_scaling_enabled;
 
 	sprctl = I915_READ(SPRCTL(pipe));
 
@@ -257,6 +262,12 @@ ivb_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 		BUG();
 	}
 
+	/*
+	 * Enable gamma to match primary/cursor plane behaviour.
+	 * FIXME should be user controllable via propertiesa.
+	 */
+	sprctl |= SPRITE_GAMMA_ENABLE;
+
 	if (obj->tiling_mode != I915_TILING_NONE)
 		sprctl |= SPRITE_TILED;
 
@@ -279,21 +290,8 @@ ivb_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 	crtc_w--;
 	crtc_h--;
 
-	/*
-	 * IVB workaround: must disable low power watermarks for at least
-	 * one frame before enabling scaling.  LP watermarks can be re-enabled
-	 * when scaling is disabled.
-	 */
-	if (crtc_w != src_w || crtc_h != src_h) {
-		dev_priv->sprite_scaling_enabled |= 1 << pipe;
-
-		if (!scaling_was_enabled) {
-			intel_update_watermarks(crtc);
-			intel_wait_for_vblank(dev, pipe);
-		}
+	if (crtc_w != src_w || crtc_h != src_h)
 		sprscale = SPRITE_SCALE_ENABLE | (src_w << 16) | src_h;
-	} else
-		dev_priv->sprite_scaling_enabled &= ~(1 << pipe);
 
 	I915_WRITE(SPRSTRIDE(pipe), fb->pitches[0]);
 	I915_WRITE(SPRPOS(pipe), (crtc_y << 16) | crtc_x);
@@ -317,13 +315,9 @@ ivb_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 	if (intel_plane->can_scale)
 		I915_WRITE(SPRSCALE(pipe), sprscale);
 	I915_WRITE(SPRCTL(pipe), sprctl);
-	I915_MODIFY_DISPBASE(SPRSURF(pipe),
-			     i915_gem_obj_ggtt_offset(obj) + sprsurf_offset);
+	I915_WRITE(SPRSURF(pipe),
+		   i915_gem_obj_ggtt_offset(obj) + sprsurf_offset);
 	POSTING_READ(SPRSURF(pipe));
-
-	/* potentially re-enable LP watermarks */
-	if (scaling_was_enabled && !dev_priv->sprite_scaling_enabled)
-		intel_update_watermarks(crtc);
 }
 
 static void
@@ -333,23 +327,22 @@ ivb_disable_plane(struct drm_plane *plane, struct drm_crtc *crtc)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct intel_plane *intel_plane = to_intel_plane(plane);
 	int pipe = intel_plane->pipe;
-	bool scaling_was_enabled = dev_priv->sprite_scaling_enabled;
 
 	I915_WRITE(SPRCTL(pipe), I915_READ(SPRCTL(pipe)) & ~SPRITE_ENABLE);
 	/* Can't leave the scaler enabled... */
 	if (intel_plane->can_scale)
 		I915_WRITE(SPRSCALE(pipe), 0);
 	/* Activate double buffered register update */
-	I915_MODIFY_DISPBASE(SPRSURF(pipe), 0);
+	I915_WRITE(SPRSURF(pipe), 0);
 	POSTING_READ(SPRSURF(pipe));
 
-	dev_priv->sprite_scaling_enabled &= ~(1 << pipe);
+	/*
+	 * Avoid underruns when disabling the sprite.
+	 * FIXME remove once watermark updates are done properly.
+	 */
+	intel_wait_for_vblank(dev, pipe);
 
 	intel_update_sprite_watermarks(plane, crtc, 0, 0, false, false);
-
-	/* potentially re-enable LP watermarks */
-	if (scaling_was_enabled && !dev_priv->sprite_scaling_enabled)
-		intel_update_watermarks(crtc);
 }
 
 static int
@@ -453,6 +446,12 @@ ilk_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 		BUG();
 	}
 
+	/*
+	 * Enable gamma to match primary/cursor plane behaviour.
+	 * FIXME should be user controllable via propertiesa.
+	 */
+	dvscntr |= DVS_GAMMA_ENABLE;
+
 	if (obj->tiling_mode != I915_TILING_NONE)
 		dvscntr |= DVS_TILED;
 
@@ -470,7 +469,7 @@ ilk_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 	crtc_h--;
 
 	dvsscale = 0;
-	if (IS_GEN5(dev) || crtc_w != src_w || crtc_h != src_h)
+	if (crtc_w != src_w || crtc_h != src_h)
 		dvsscale = DVS_SCALE_ENABLE | (src_w << 16) | src_h;
 
 	I915_WRITE(DVSSTRIDE(pipe), fb->pitches[0]);
@@ -490,8 +489,8 @@ ilk_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 	I915_WRITE(DVSSIZE(pipe), (crtc_h << 16) | crtc_w);
 	I915_WRITE(DVSSCALE(pipe), dvsscale);
 	I915_WRITE(DVSCNTR(pipe), dvscntr);
-	I915_MODIFY_DISPBASE(DVSSURF(pipe),
-			     i915_gem_obj_ggtt_offset(obj) + dvssurf_offset);
+	I915_WRITE(DVSSURF(pipe),
+		   i915_gem_obj_ggtt_offset(obj) + dvssurf_offset);
 	POSTING_READ(DVSSURF(pipe));
 }
 
@@ -507,9 +506,15 @@ ilk_disable_plane(struct drm_plane *plane, struct drm_crtc *crtc)
 	/* Disable the scaler */
 	I915_WRITE(DVSSCALE(pipe), 0);
 	/* Flush double buffered register updates */
-	I915_MODIFY_DISPBASE(DVSSURF(pipe), 0);
+	I915_WRITE(DVSSURF(pipe), 0);
 	POSTING_READ(DVSSURF(pipe));
 
+	/*
+	 * Avoid underruns when disabling the sprite.
+	 * FIXME remove once watermark updates are done properly.
+	 */
+	intel_wait_for_vblank(dev, pipe);
+
 	intel_update_sprite_watermarks(plane, crtc, 0, 0, false, false);
 }
 
@@ -643,6 +648,15 @@ format_is_yuv(uint32_t format)
 	}
 }
 
+static bool colorkey_enabled(struct intel_plane *intel_plane)
+{
+	struct drm_intel_sprite_colorkey key;
+
+	intel_plane->get_colorkey(&intel_plane->base, &key);
+
+	return key.flags != I915_SET_COLORKEY_NONE;
+}
+
 static int
 intel_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 		   struct drm_framebuffer *fb, int crtc_x, int crtc_y,
@@ -828,7 +842,7 @@ intel_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 	 * If the sprite is completely covering the primary plane,
 	 * we can disable the primary and save power.
 	 */
-	disable_primary = drm_rect_equals(&dst, &clip);
+	disable_primary = drm_rect_equals(&dst, &clip) && !colorkey_enabled(intel_plane);
 	WARN_ON(disable_primary && !visible && intel_crtc->active);
 
 	mutex_lock(&dev->struct_mutex);
diff --git a/drivers/gpu/drm/i915/intel_uncore.c b/drivers/gpu/drm/i915/intel_uncore.c
index 25cbe073c388..87df68f5f504 100644
--- a/drivers/gpu/drm/i915/intel_uncore.c
+++ b/drivers/gpu/drm/i915/intel_uncore.c
@@ -64,7 +64,8 @@ static void __gen6_gt_force_wake_reset(struct drm_i915_private *dev_priv)
 	__raw_posting_read(dev_priv, ECOBUS);
 }
 
-static void __gen6_gt_force_wake_get(struct drm_i915_private *dev_priv)
+static void __gen6_gt_force_wake_get(struct drm_i915_private *dev_priv,
+							int fw_engine)
 {
 	if (wait_for_atomic((__raw_i915_read32(dev_priv, FORCEWAKE_ACK) & 1) == 0,
 			    FORCEWAKE_ACK_TIMEOUT_MS))
@@ -89,7 +90,8 @@ static void __gen6_gt_force_wake_mt_reset(struct drm_i915_private *dev_priv)
 	__raw_posting_read(dev_priv, ECOBUS);
 }
 
-static void __gen6_gt_force_wake_mt_get(struct drm_i915_private *dev_priv)
+static void __gen6_gt_force_wake_mt_get(struct drm_i915_private *dev_priv,
+							int fw_engine)
 {
 	u32 forcewake_ack;
 
@@ -121,12 +123,12 @@ static void gen6_gt_check_fifodbg(struct drm_i915_private *dev_priv)
 	u32 gtfifodbg;
 
 	gtfifodbg = __raw_i915_read32(dev_priv, GTFIFODBG);
-	if (WARN(gtfifodbg & GT_FIFO_CPU_ERROR_MASK,
-	     "MMIO read or write has been dropped %x\n", gtfifodbg))
-		__raw_i915_write32(dev_priv, GTFIFODBG, GT_FIFO_CPU_ERROR_MASK);
+	if (WARN(gtfifodbg, "GT wake FIFO error 0x%x\n", gtfifodbg))
+		__raw_i915_write32(dev_priv, GTFIFODBG, gtfifodbg);
 }
 
-static void __gen6_gt_force_wake_put(struct drm_i915_private *dev_priv)
+static void __gen6_gt_force_wake_put(struct drm_i915_private *dev_priv,
+							int fw_engine)
 {
 	__raw_i915_write32(dev_priv, FORCEWAKE, 0);
 	/* something from same cacheline, but !FORCEWAKE */
@@ -134,7 +136,8 @@ static void __gen6_gt_force_wake_put(struct drm_i915_private *dev_priv)
 	gen6_gt_check_fifodbg(dev_priv);
 }
 
-static void __gen6_gt_force_wake_mt_put(struct drm_i915_private *dev_priv)
+static void __gen6_gt_force_wake_mt_put(struct drm_i915_private *dev_priv,
+							int fw_engine)
 {
 	__raw_i915_write32(dev_priv, FORCEWAKE_MT,
 			   _MASKED_BIT_DISABLE(FORCEWAKE_KERNEL));
@@ -147,12 +150,19 @@ static int __gen6_gt_wait_for_fifo(struct drm_i915_private *dev_priv)
 {
 	int ret = 0;
 
+	/* On VLV, FIFO will be shared by both SW and HW.
+	 * So, we need to read the FREE_ENTRIES everytime */
+	if (IS_VALLEYVIEW(dev_priv->dev))
+		dev_priv->uncore.fifo_count =
+			__raw_i915_read32(dev_priv, GTFIFOCTL) &
+						GT_FIFO_FREE_ENTRIES_MASK;
+
 	if (dev_priv->uncore.fifo_count < GT_FIFO_NUM_RESERVED_ENTRIES) {
 		int loop = 500;
-		u32 fifo = __raw_i915_read32(dev_priv, GT_FIFO_FREE_ENTRIES);
+		u32 fifo = __raw_i915_read32(dev_priv, GTFIFOCTL) & GT_FIFO_FREE_ENTRIES_MASK;
 		while (fifo <= GT_FIFO_NUM_RESERVED_ENTRIES && loop--) {
 			udelay(10);
-			fifo = __raw_i915_read32(dev_priv, GT_FIFO_FREE_ENTRIES);
+			fifo = __raw_i915_read32(dev_priv, GTFIFOCTL) & GT_FIFO_FREE_ENTRIES_MASK;
 		}
 		if (WARN_ON(loop < 0 && fifo <= GT_FIFO_NUM_RESERVED_ENTRIES))
 			++ret;
@@ -171,38 +181,112 @@ static void vlv_force_wake_reset(struct drm_i915_private *dev_priv)
 	__raw_posting_read(dev_priv, FORCEWAKE_ACK_VLV);
 }
 
-static void vlv_force_wake_get(struct drm_i915_private *dev_priv)
+static void __vlv_force_wake_get(struct drm_i915_private *dev_priv,
+						int fw_engine)
 {
-	if (wait_for_atomic((__raw_i915_read32(dev_priv, FORCEWAKE_ACK_VLV) & FORCEWAKE_KERNEL) == 0,
-			    FORCEWAKE_ACK_TIMEOUT_MS))
-		DRM_ERROR("Timed out waiting for forcewake old ack to clear.\n");
+	/* Check for Render Engine */
+	if (FORCEWAKE_RENDER & fw_engine) {
+		if (wait_for_atomic((__raw_i915_read32(dev_priv,
+						FORCEWAKE_ACK_VLV) &
+						FORCEWAKE_KERNEL) == 0,
+					FORCEWAKE_ACK_TIMEOUT_MS))
+			DRM_ERROR("Timed out: Render forcewake old ack to clear.\n");
 
-	__raw_i915_write32(dev_priv, FORCEWAKE_VLV,
-			   _MASKED_BIT_ENABLE(FORCEWAKE_KERNEL));
-	__raw_i915_write32(dev_priv, FORCEWAKE_MEDIA_VLV,
-			   _MASKED_BIT_ENABLE(FORCEWAKE_KERNEL));
+		__raw_i915_write32(dev_priv, FORCEWAKE_VLV,
+				   _MASKED_BIT_ENABLE(FORCEWAKE_KERNEL));
 
-	if (wait_for_atomic((__raw_i915_read32(dev_priv, FORCEWAKE_ACK_VLV) & FORCEWAKE_KERNEL),
-			    FORCEWAKE_ACK_TIMEOUT_MS))
-		DRM_ERROR("Timed out waiting for GT to ack forcewake request.\n");
+		if (wait_for_atomic((__raw_i915_read32(dev_priv,
+						FORCEWAKE_ACK_VLV) &
+						FORCEWAKE_KERNEL),
+					FORCEWAKE_ACK_TIMEOUT_MS))
+			DRM_ERROR("Timed out: waiting for Render to ack.\n");
+	}
 
-	if (wait_for_atomic((__raw_i915_read32(dev_priv, FORCEWAKE_ACK_MEDIA_VLV) &
-			     FORCEWAKE_KERNEL),
-			    FORCEWAKE_ACK_TIMEOUT_MS))
-		DRM_ERROR("Timed out waiting for media to ack forcewake request.\n");
+	/* Check for Media Engine */
+	if (FORCEWAKE_MEDIA & fw_engine) {
+		if (wait_for_atomic((__raw_i915_read32(dev_priv,
+						FORCEWAKE_ACK_MEDIA_VLV) &
+						FORCEWAKE_KERNEL) == 0,
+					FORCEWAKE_ACK_TIMEOUT_MS))
+			DRM_ERROR("Timed out: Media forcewake old ack to clear.\n");
+
+		__raw_i915_write32(dev_priv, FORCEWAKE_MEDIA_VLV,
+				   _MASKED_BIT_ENABLE(FORCEWAKE_KERNEL));
+
+		if (wait_for_atomic((__raw_i915_read32(dev_priv,
+						FORCEWAKE_ACK_MEDIA_VLV) &
+						FORCEWAKE_KERNEL),
+					FORCEWAKE_ACK_TIMEOUT_MS))
+			DRM_ERROR("Timed out: waiting for media to ack.\n");
+	}
 
 	/* WaRsForcewakeWaitTC0:vlv */
 	__gen6_gt_wait_for_thread_c0(dev_priv);
+
 }
 
-static void vlv_force_wake_put(struct drm_i915_private *dev_priv)
+static void __vlv_force_wake_put(struct drm_i915_private *dev_priv,
+					int fw_engine)
 {
-	__raw_i915_write32(dev_priv, FORCEWAKE_VLV,
-			   _MASKED_BIT_DISABLE(FORCEWAKE_KERNEL));
-	__raw_i915_write32(dev_priv, FORCEWAKE_MEDIA_VLV,
-			   _MASKED_BIT_DISABLE(FORCEWAKE_KERNEL));
+
+	/* Check for Render Engine */
+	if (FORCEWAKE_RENDER & fw_engine)
+		__raw_i915_write32(dev_priv, FORCEWAKE_VLV,
+					_MASKED_BIT_DISABLE(FORCEWAKE_KERNEL));
+
+
+	/* Check for Media Engine */
+	if (FORCEWAKE_MEDIA & fw_engine)
+		__raw_i915_write32(dev_priv, FORCEWAKE_MEDIA_VLV,
+				_MASKED_BIT_DISABLE(FORCEWAKE_KERNEL));
+
 	/* The below doubles as a POSTING_READ */
 	gen6_gt_check_fifodbg(dev_priv);
+
+}
+
+void vlv_force_wake_get(struct drm_i915_private *dev_priv,
+						int fw_engine)
+{
+	unsigned long irqflags;
+
+	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
+	if (FORCEWAKE_RENDER & fw_engine) {
+		if (dev_priv->uncore.fw_rendercount++ == 0)
+			dev_priv->uncore.funcs.force_wake_get(dev_priv,
+							FORCEWAKE_RENDER);
+	}
+	if (FORCEWAKE_MEDIA & fw_engine) {
+		if (dev_priv->uncore.fw_mediacount++ == 0)
+			dev_priv->uncore.funcs.force_wake_get(dev_priv,
+							FORCEWAKE_MEDIA);
+	}
+
+	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags);
+}
+
+void vlv_force_wake_put(struct drm_i915_private *dev_priv,
+						int fw_engine)
+{
+	unsigned long irqflags;
+
+	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
+
+	if (FORCEWAKE_RENDER & fw_engine) {
+		WARN_ON(dev_priv->uncore.fw_rendercount == 0);
+		if (--dev_priv->uncore.fw_rendercount == 0)
+			dev_priv->uncore.funcs.force_wake_put(dev_priv,
+							FORCEWAKE_RENDER);
+	}
+
+	if (FORCEWAKE_MEDIA & fw_engine) {
+		WARN_ON(dev_priv->uncore.fw_mediacount == 0);
+		if (--dev_priv->uncore.fw_mediacount == 0)
+			dev_priv->uncore.funcs.force_wake_put(dev_priv,
+							FORCEWAKE_MEDIA);
+	}
+
+	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags);
 }
 
 static void gen6_force_wake_work(struct work_struct *work)
@@ -213,7 +297,7 @@ static void gen6_force_wake_work(struct work_struct *work)
 
 	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
 	if (--dev_priv->uncore.forcewake_count == 0)
-		dev_priv->uncore.funcs.force_wake_put(dev_priv);
+		dev_priv->uncore.funcs.force_wake_put(dev_priv, FORCEWAKE_ALL);
 	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags);
 }
 
@@ -248,6 +332,11 @@ void intel_uncore_early_sanitize(struct drm_device *dev)
 		DRM_INFO("Found %zuMB of eLLC\n", dev_priv->ellc_size);
 	}
 
+	/* clear out old GT FIFO errors */
+	if (IS_GEN6(dev) || IS_GEN7(dev))
+		__raw_i915_write32(dev_priv, GTFIFODBG,
+				   __raw_i915_read32(dev_priv, GTFIFODBG));
+
 	intel_uncore_forcewake_reset(dev);
 }
 
@@ -256,8 +345,6 @@ void intel_uncore_sanitize(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 reg_val;
 
-	intel_uncore_forcewake_reset(dev);
-
 	/* BIOS often leaves RC6 enabled, but disable it for hw init */
 	intel_disable_gt_powersave(dev);
 
@@ -281,29 +368,40 @@ void intel_uncore_sanitize(struct drm_device *dev)
  * be called at the beginning of the sequence followed by a call to
  * gen6_gt_force_wake_put() at the end of the sequence.
  */
-void gen6_gt_force_wake_get(struct drm_i915_private *dev_priv)
+void gen6_gt_force_wake_get(struct drm_i915_private *dev_priv, int fw_engine)
 {
 	unsigned long irqflags;
 
 	if (!dev_priv->uncore.funcs.force_wake_get)
 		return;
 
+	intel_runtime_pm_get(dev_priv);
+
+	/* Redirect to VLV specific routine */
+	if (IS_VALLEYVIEW(dev_priv->dev))
+		return vlv_force_wake_get(dev_priv, fw_engine);
+
 	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
 	if (dev_priv->uncore.forcewake_count++ == 0)
-		dev_priv->uncore.funcs.force_wake_get(dev_priv);
+		dev_priv->uncore.funcs.force_wake_get(dev_priv, FORCEWAKE_ALL);
 	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags);
 }
 
 /*
  * see gen6_gt_force_wake_get()
  */
-void gen6_gt_force_wake_put(struct drm_i915_private *dev_priv)
+void gen6_gt_force_wake_put(struct drm_i915_private *dev_priv, int fw_engine)
 {
 	unsigned long irqflags;
 
 	if (!dev_priv->uncore.funcs.force_wake_put)
 		return;
 
+	/* Redirect to VLV specific routine */
+	if (IS_VALLEYVIEW(dev_priv->dev))
+		return vlv_force_wake_put(dev_priv, fw_engine);
+
+
 	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
 	if (--dev_priv->uncore.forcewake_count == 0) {
 		dev_priv->uncore.forcewake_count++;
@@ -312,6 +410,8 @@ void gen6_gt_force_wake_put(struct drm_i915_private *dev_priv)
 				 1);
 	}
 	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags);
+
+	intel_runtime_pm_put(dev_priv);
 }
 
 /* We give fast paths for the really cool registers */
@@ -346,6 +446,13 @@ hsw_unclaimed_reg_check(struct drm_i915_private *dev_priv, u32 reg)
 	}
 }
 
+static void
+assert_device_not_suspended(struct drm_i915_private *dev_priv)
+{
+	WARN(HAS_RUNTIME_PM(dev_priv->dev) && dev_priv->pm.suspended,
+	     "Device suspended\n");
+}
+
 #define REG_READ_HEADER(x) \
 	unsigned long irqflags; \
 	u##x val = 0; \
@@ -379,16 +486,51 @@ gen6_read##x(struct drm_i915_private *dev_priv, off_t reg, bool trace) { \
 	REG_READ_HEADER(x); \
 	if (NEEDS_FORCE_WAKE((dev_priv), (reg))) { \
 		if (dev_priv->uncore.forcewake_count == 0) \
-			dev_priv->uncore.funcs.force_wake_get(dev_priv); \
+			dev_priv->uncore.funcs.force_wake_get(dev_priv, \
+							FORCEWAKE_ALL); \
 		val = __raw_i915_read##x(dev_priv, reg); \
 		if (dev_priv->uncore.forcewake_count == 0) \
-			dev_priv->uncore.funcs.force_wake_put(dev_priv); \
+			dev_priv->uncore.funcs.force_wake_put(dev_priv, \
+							FORCEWAKE_ALL); \
 	} else { \
 		val = __raw_i915_read##x(dev_priv, reg); \
 	} \
 	REG_READ_FOOTER; \
 }
 
+#define __vlv_read(x) \
+static u##x \
+vlv_read##x(struct drm_i915_private *dev_priv, off_t reg, bool trace) { \
+	unsigned fwengine = 0; \
+	unsigned *fwcount; \
+	REG_READ_HEADER(x); \
+	if (FORCEWAKE_VLV_RENDER_RANGE_OFFSET(reg)) {   \
+		fwengine = FORCEWAKE_RENDER;            \
+		fwcount = &dev_priv->uncore.fw_rendercount;    \
+	}                                               \
+	else if (FORCEWAKE_VLV_MEDIA_RANGE_OFFSET(reg)) {       \
+		fwengine = FORCEWAKE_MEDIA;             \
+		fwcount = &dev_priv->uncore.fw_mediacount;     \
+	}  \
+	if (fwengine != 0) {		\
+		if ((*fwcount)++ == 0) \
+			(dev_priv)->uncore.funcs.force_wake_get(dev_priv, \
+								fwengine); \
+		val = __raw_i915_read##x(dev_priv, reg); \
+		if (--(*fwcount) == 0) \
+			(dev_priv)->uncore.funcs.force_wake_put(dev_priv, \
+							fwengine); \
+	} else { \
+		val = __raw_i915_read##x(dev_priv, reg); \
+	} \
+	REG_READ_FOOTER; \
+}
+
+
+__vlv_read(8)
+__vlv_read(16)
+__vlv_read(32)
+__vlv_read(64)
 __gen6_read(8)
 __gen6_read(16)
 __gen6_read(32)
@@ -402,6 +544,7 @@ __gen4_read(16)
 __gen4_read(32)
 __gen4_read(64)
 
+#undef __vlv_read
 #undef __gen6_read
 #undef __gen5_read
 #undef __gen4_read
@@ -413,12 +556,15 @@ __gen4_read(64)
 	trace_i915_reg_rw(true, reg, val, sizeof(val), trace); \
 	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags)
 
+#define REG_WRITE_FOOTER \
+	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags)
+
 #define __gen4_write(x) \
 static void \
 gen4_write##x(struct drm_i915_private *dev_priv, off_t reg, u##x val, bool trace) { \
 	REG_WRITE_HEADER; \
 	__raw_i915_write##x(dev_priv, reg, val); \
-	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags); \
+	REG_WRITE_FOOTER; \
 }
 
 #define __gen5_write(x) \
@@ -427,7 +573,7 @@ gen5_write##x(struct drm_i915_private *dev_priv, off_t reg, u##x val, bool trace
 	REG_WRITE_HEADER; \
 	ilk_dummy_write(dev_priv); \
 	__raw_i915_write##x(dev_priv, reg, val); \
-	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags); \
+	REG_WRITE_FOOTER; \
 }
 
 #define __gen6_write(x) \
@@ -438,11 +584,12 @@ gen6_write##x(struct drm_i915_private *dev_priv, off_t reg, u##x val, bool trace
 	if (NEEDS_FORCE_WAKE((dev_priv), (reg))) { \
 		__fifo_ret = __gen6_gt_wait_for_fifo(dev_priv); \
 	} \
+	assert_device_not_suspended(dev_priv); \
 	__raw_i915_write##x(dev_priv, reg, val); \
 	if (unlikely(__fifo_ret)) { \
 		gen6_gt_check_fifodbg(dev_priv); \
 	} \
-	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags); \
+	REG_WRITE_FOOTER; \
 }
 
 #define __hsw_write(x) \
@@ -453,13 +600,14 @@ hsw_write##x(struct drm_i915_private *dev_priv, off_t reg, u##x val, bool trace)
 	if (NEEDS_FORCE_WAKE((dev_priv), (reg))) { \
 		__fifo_ret = __gen6_gt_wait_for_fifo(dev_priv); \
 	} \
+	assert_device_not_suspended(dev_priv); \
 	hsw_unclaimed_reg_clear(dev_priv, reg); \
 	__raw_i915_write##x(dev_priv, reg, val); \
 	if (unlikely(__fifo_ret)) { \
 		gen6_gt_check_fifodbg(dev_priv); \
 	} \
 	hsw_unclaimed_reg_check(dev_priv, reg); \
-	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags); \
+	REG_WRITE_FOOTER; \
 }
 
 static const u32 gen8_shadowed_regs[] = {
@@ -486,16 +634,18 @@ static bool is_gen8_shadowed(struct drm_i915_private *dev_priv, u32 reg)
 #define __gen8_write(x) \
 static void \
 gen8_write##x(struct drm_i915_private *dev_priv, off_t reg, u##x val, bool trace) { \
-	bool __needs_put = !is_gen8_shadowed(dev_priv, reg); \
+	bool __needs_put = reg < 0x40000 && !is_gen8_shadowed(dev_priv, reg); \
 	REG_WRITE_HEADER; \
 	if (__needs_put) { \
-		dev_priv->uncore.funcs.force_wake_get(dev_priv); \
+		dev_priv->uncore.funcs.force_wake_get(dev_priv, \
+							FORCEWAKE_ALL); \
 	} \
 	__raw_i915_write##x(dev_priv, reg, val); \
 	if (__needs_put) { \
-		dev_priv->uncore.funcs.force_wake_put(dev_priv); \
+		dev_priv->uncore.funcs.force_wake_put(dev_priv, \
+							FORCEWAKE_ALL); \
 	} \
-	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags); \
+	REG_WRITE_FOOTER; \
 }
 
 __gen8_write(8)
@@ -524,6 +674,7 @@ __gen4_write(64)
 #undef __gen6_write
 #undef __gen5_write
 #undef __gen4_write
+#undef REG_WRITE_FOOTER
 #undef REG_WRITE_HEADER
 
 void intel_uncore_init(struct drm_device *dev)
@@ -534,8 +685,8 @@ void intel_uncore_init(struct drm_device *dev)
 			  gen6_force_wake_work);
 
 	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;
+		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)) {
 		dev_priv->uncore.funcs.force_wake_get = __gen6_gt_force_wake_mt_get;
 		dev_priv->uncore.funcs.force_wake_put = __gen6_gt_force_wake_mt_put;
@@ -552,9 +703,9 @@ void intel_uncore_init(struct drm_device *dev)
 		 * forcewake being disabled.
 		 */
 		mutex_lock(&dev->struct_mutex);
-		__gen6_gt_force_wake_mt_get(dev_priv);
+		__gen6_gt_force_wake_mt_get(dev_priv, FORCEWAKE_ALL);
 		ecobus = __raw_i915_read32(dev_priv, ECOBUS);
-		__gen6_gt_force_wake_mt_put(dev_priv);
+		__gen6_gt_force_wake_mt_put(dev_priv, FORCEWAKE_ALL);
 		mutex_unlock(&dev->struct_mutex);
 
 		if (ecobus & FORCEWAKE_MT_ENABLE) {
@@ -601,10 +752,18 @@ void intel_uncore_init(struct drm_device *dev)
 			dev_priv->uncore.funcs.mmio_writel  = gen6_write32;
 			dev_priv->uncore.funcs.mmio_writeq  = gen6_write64;
 		}
-		dev_priv->uncore.funcs.mmio_readb  = gen6_read8;
-		dev_priv->uncore.funcs.mmio_readw  = gen6_read16;
-		dev_priv->uncore.funcs.mmio_readl  = gen6_read32;
-		dev_priv->uncore.funcs.mmio_readq  = gen6_read64;
+
+		if (IS_VALLEYVIEW(dev)) {
+			dev_priv->uncore.funcs.mmio_readb  = vlv_read8;
+			dev_priv->uncore.funcs.mmio_readw  = vlv_read16;
+			dev_priv->uncore.funcs.mmio_readl  = vlv_read32;
+			dev_priv->uncore.funcs.mmio_readq  = vlv_read64;
+		} else {
+			dev_priv->uncore.funcs.mmio_readb  = gen6_read8;
+			dev_priv->uncore.funcs.mmio_readw  = gen6_read16;
+			dev_priv->uncore.funcs.mmio_readl  = gen6_read32;
+			dev_priv->uncore.funcs.mmio_readq  = gen6_read64;
+		}
 		break;
 	case 5:
 		dev_priv->uncore.funcs.mmio_writeb  = gen5_write8;
@@ -646,7 +805,7 @@ static const struct register_whitelist {
 	uint32_t size;
 	uint32_t gen_bitmask; /* support gens, 0x10 for 4, 0x30 for 4 and 5, etc. */
 } whitelist[] = {
-	{ RING_TIMESTAMP(RENDER_RING_BASE), 8, 0xF0 },
+	{ RING_TIMESTAMP(RENDER_RING_BASE), 8, 0x1F0 },
 };
 
 int i915_reg_read_ioctl(struct drm_device *dev,
@@ -687,6 +846,43 @@ int i915_reg_read_ioctl(struct drm_device *dev,
 	return 0;
 }
 
+int i915_get_reset_stats_ioctl(struct drm_device *dev,
+			       void *data, struct drm_file *file)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct drm_i915_reset_stats *args = data;
+	struct i915_ctx_hang_stats *hs;
+	int ret;
+
+	if (args->flags || args->pad)
+		return -EINVAL;
+
+	if (args->ctx_id == DEFAULT_CONTEXT_ID && !capable(CAP_SYS_ADMIN))
+		return -EPERM;
+
+	ret = mutex_lock_interruptible(&dev->struct_mutex);
+	if (ret)
+		return ret;
+
+	hs = i915_gem_context_get_hang_stats(dev, file, args->ctx_id);
+	if (IS_ERR(hs)) {
+		mutex_unlock(&dev->struct_mutex);
+		return PTR_ERR(hs);
+	}
+
+	if (capable(CAP_SYS_ADMIN))
+		args->reset_count = i915_reset_count(&dev_priv->gpu_error);
+	else
+		args->reset_count = 0;
+
+	args->batch_active = hs->batch_active;
+	args->batch_pending = hs->batch_pending;
+
+	mutex_unlock(&dev->struct_mutex);
+
+	return 0;
+}
+
 static int i965_reset_complete(struct drm_device *dev)
 {
 	u8 gdrst;
@@ -770,12 +966,12 @@ static int gen6_do_reset(struct drm_device *dev)
 
 	/* If reset with a user forcewake, try to restore, otherwise turn it off */
 	if (dev_priv->uncore.forcewake_count)
-		dev_priv->uncore.funcs.force_wake_get(dev_priv);
+		dev_priv->uncore.funcs.force_wake_get(dev_priv, FORCEWAKE_ALL);
 	else
-		dev_priv->uncore.funcs.force_wake_put(dev_priv);
+		dev_priv->uncore.funcs.force_wake_put(dev_priv, FORCEWAKE_ALL);
 
 	/* Restore fifo count */
-	dev_priv->uncore.fifo_count = __raw_i915_read32(dev_priv, GT_FIFO_FREE_ENTRIES);
+	dev_priv->uncore.fifo_count = __raw_i915_read32(dev_priv, GTFIFOCTL) & GT_FIFO_FREE_ENTRIES_MASK;
 
 	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags);
 	return ret;
@@ -793,15 +989,6 @@ int intel_gpu_reset(struct drm_device *dev)
 	}
 }
 
-void intel_uncore_clear_errors(struct drm_device *dev)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-
-	/* XXX needs spinlock around caller's grouping */
-	if (HAS_FPGA_DBG_UNCLAIMED(dev))
-		__raw_i915_write32(dev_priv, FPGA_DBG, FPGA_DBG_RM_NOCLAIM);
-}
-
 void intel_uncore_check_errors(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
diff --git a/drivers/gpu/drm/mga/mga_dma.c b/drivers/gpu/drm/mga/mga_dma.c
index 087db33f6cff..c3bf059ba720 100644
--- a/drivers/gpu/drm/mga/mga_dma.c
+++ b/drivers/gpu/drm/mga/mga_dma.c
@@ -1075,10 +1075,10 @@ static int mga_dma_get_buffers(struct drm_device *dev,
 
 		buf->file_priv = file_priv;
 
-		if (DRM_COPY_TO_USER(&d->request_indices[i],
+		if (copy_to_user(&d->request_indices[i],
 				     &buf->idx, sizeof(buf->idx)))
 			return -EFAULT;
-		if (DRM_COPY_TO_USER(&d->request_sizes[i],
+		if (copy_to_user(&d->request_sizes[i],
 				     &buf->total, sizeof(buf->total)))
 			return -EFAULT;
 
diff --git a/drivers/gpu/drm/mga/mga_drv.h b/drivers/gpu/drm/mga/mga_drv.h
index ca4bc54ea214..fe453213600a 100644
--- a/drivers/gpu/drm/mga/mga_drv.h
+++ b/drivers/gpu/drm/mga/mga_drv.h
@@ -186,14 +186,14 @@ extern void mga_disable_vblank(struct drm_device *dev, int crtc);
 extern u32 mga_get_vblank_counter(struct drm_device *dev, int crtc);
 extern int mga_driver_fence_wait(struct drm_device *dev, unsigned int *sequence);
 extern int mga_driver_vblank_wait(struct drm_device *dev, unsigned int *sequence);
-extern irqreturn_t mga_driver_irq_handler(DRM_IRQ_ARGS);
+extern irqreturn_t mga_driver_irq_handler(int irq, void *arg);
 extern void mga_driver_irq_preinstall(struct drm_device *dev);
 extern int mga_driver_irq_postinstall(struct drm_device *dev);
 extern void mga_driver_irq_uninstall(struct drm_device *dev);
 extern long mga_compat_ioctl(struct file *filp, unsigned int cmd,
 			     unsigned long arg);
 
-#define mga_flush_write_combine()	DRM_WRITEMEMORYBARRIER()
+#define mga_flush_write_combine()	wmb()
 
 #define MGA_READ8(reg)		DRM_READ8(dev_priv->mmio, (reg))
 #define MGA_READ(reg)		DRM_READ32(dev_priv->mmio, (reg))
diff --git a/drivers/gpu/drm/mga/mga_ioc32.c b/drivers/gpu/drm/mga/mga_ioc32.c
index 709e90db8c40..86b4bb804852 100644
--- a/drivers/gpu/drm/mga/mga_ioc32.c
+++ b/drivers/gpu/drm/mga/mga_ioc32.c
@@ -34,6 +34,7 @@
 
 #include <drm/drmP.h>
 #include <drm/mga_drm.h>
+#include "mga_drv.h"
 
 typedef struct drm32_mga_init {
 	int func;
diff --git a/drivers/gpu/drm/mga/mga_irq.c b/drivers/gpu/drm/mga/mga_irq.c
index 2b0ceb8dc11b..1b071b8ff9dc 100644
--- a/drivers/gpu/drm/mga/mga_irq.c
+++ b/drivers/gpu/drm/mga/mga_irq.c
@@ -47,7 +47,7 @@ u32 mga_get_vblank_counter(struct drm_device *dev, int crtc)
 }
 
 
-irqreturn_t mga_driver_irq_handler(DRM_IRQ_ARGS)
+irqreturn_t mga_driver_irq_handler(int irq, void *arg)
 {
 	struct drm_device *dev = (struct drm_device *) arg;
 	drm_mga_private_t *dev_priv = (drm_mga_private_t *) dev->dev_private;
@@ -79,7 +79,7 @@ irqreturn_t mga_driver_irq_handler(DRM_IRQ_ARGS)
 			MGA_WRITE(MGA_PRIMEND, prim_end);
 
 		atomic_inc(&dev_priv->last_fence_retired);
-		DRM_WAKEUP(&dev_priv->fence_queue);
+		wake_up(&dev_priv->fence_queue);
 		handled = 1;
 	}
 
@@ -128,7 +128,7 @@ int mga_driver_fence_wait(struct drm_device *dev, unsigned int *sequence)
 	 * by about a day rather than she wants to wait for years
 	 * using fences.
 	 */
-	DRM_WAIT_ON(ret, dev_priv->fence_queue, 3 * DRM_HZ,
+	DRM_WAIT_ON(ret, dev_priv->fence_queue, 3 * HZ,
 		    (((cur_fence = atomic_read(&dev_priv->last_fence_retired))
 		      - *sequence) <= (1 << 23)));
 
@@ -151,7 +151,7 @@ int mga_driver_irq_postinstall(struct drm_device *dev)
 {
 	drm_mga_private_t *dev_priv = (drm_mga_private_t *) dev->dev_private;
 
-	DRM_INIT_WAITQUEUE(&dev_priv->fence_queue);
+	init_waitqueue_head(&dev_priv->fence_queue);
 
 	/* Turn on soft trap interrupt.  Vertical blank interrupts are enabled
 	 * in mga_enable_vblank.
diff --git a/drivers/gpu/drm/mga/mga_state.c b/drivers/gpu/drm/mga/mga_state.c
index 37cc2fb4eadd..314685b7f41f 100644
--- a/drivers/gpu/drm/mga/mga_state.c
+++ b/drivers/gpu/drm/mga/mga_state.c
@@ -1029,7 +1029,7 @@ static int mga_getparam(struct drm_device *dev, void *data, struct drm_file *fil
 		return -EINVAL;
 	}
 
-	if (DRM_COPY_TO_USER(param->value, &value, sizeof(int))) {
+	if (copy_to_user(param->value, &value, sizeof(int))) {
 		DRM_ERROR("copy_to_user\n");
 		return -EFAULT;
 	}
diff --git a/drivers/gpu/drm/mgag200/mgag200_cursor.c b/drivers/gpu/drm/mgag200/mgag200_cursor.c
index 801731aeab61..9f9780b7ddf0 100644
--- a/drivers/gpu/drm/mgag200/mgag200_cursor.c
+++ b/drivers/gpu/drm/mgag200/mgag200_cursor.c
@@ -22,8 +22,10 @@ static void mga_hide_cursor(struct mga_device *mdev)
 {
 	WREG8(MGA_CURPOSXL, 0);
 	WREG8(MGA_CURPOSXH, 0);
-	mgag200_bo_unpin(mdev->cursor.pixels_1);
-	mgag200_bo_unpin(mdev->cursor.pixels_2);
+	if (mdev->cursor.pixels_1->pin_count)
+		mgag200_bo_unpin(mdev->cursor.pixels_1);
+	if (mdev->cursor.pixels_2->pin_count)
+		mgag200_bo_unpin(mdev->cursor.pixels_2);
 }
 
 int mga_crtc_cursor_set(struct drm_crtc *crtc,
@@ -32,7 +34,7 @@ int mga_crtc_cursor_set(struct drm_crtc *crtc,
 			uint32_t width,
 			uint32_t height)
 {
-	struct drm_device *dev = (struct drm_device *)file_priv->minor->dev;
+	struct drm_device *dev = crtc->dev;
 	struct mga_device *mdev = (struct mga_device *)dev->dev_private;
 	struct mgag200_bo *pixels_1 = mdev->cursor.pixels_1;
 	struct mgag200_bo *pixels_2 = mdev->cursor.pixels_2;
diff --git a/drivers/gpu/drm/mgag200/mgag200_fb.c b/drivers/gpu/drm/mgag200/mgag200_fb.c
index 964f58cee5ea..f9adc27ef32a 100644
--- a/drivers/gpu/drm/mgag200/mgag200_fb.c
+++ b/drivers/gpu/drm/mgag200/mgag200_fb.c
@@ -41,7 +41,7 @@ static void mga_dirty_update(struct mga_fbdev *mfbdev,
 	 * then the BO is being moved and we should
 	 * store up the damage until later.
 	 */
-	if (!in_interrupt())
+	if (!drm_can_sleep())
 		ret = mgag200_bo_reserve(bo, true);
 	if (ret) {
 		if (ret != -EBUSY)
@@ -282,6 +282,11 @@ int mgag200_fbdev_init(struct mga_device *mdev)
 {
 	struct mga_fbdev *mfbdev;
 	int ret;
+	int bpp_sel = 32;
+
+	/* prefer 16bpp on low end gpus with limited VRAM */
+	if (IS_G200_SE(mdev) && mdev->mc.vram_size < (2048*1024))
+		bpp_sel = 16;
 
 	mfbdev = devm_kzalloc(mdev->dev->dev, sizeof(struct mga_fbdev), GFP_KERNEL);
 	if (!mfbdev)
@@ -301,7 +306,7 @@ int mgag200_fbdev_init(struct mga_device *mdev)
 	/* disable all the possible outputs/crtcs before entering KMS mode */
 	drm_helper_disable_unused_functions(mdev->dev);
 
-	drm_fb_helper_initial_config(&mfbdev->helper, 32);
+	drm_fb_helper_initial_config(&mfbdev->helper, bpp_sel);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/mgag200/mgag200_main.c b/drivers/gpu/drm/mgag200/mgag200_main.c
index b1120cb1db6d..26868e5c55b0 100644
--- a/drivers/gpu/drm/mgag200/mgag200_main.c
+++ b/drivers/gpu/drm/mgag200/mgag200_main.c
@@ -217,7 +217,10 @@ int mgag200_driver_load(struct drm_device *dev, unsigned long flags)
 
 	drm_mode_config_init(dev);
 	dev->mode_config.funcs = (void *)&mga_mode_funcs;
-	dev->mode_config.preferred_depth = 24;
+	if (IS_G200_SE(mdev) && mdev->mc.vram_size < (2048*1024))
+		dev->mode_config.preferred_depth = 16;
+	else
+		dev->mode_config.preferred_depth = 24;
 	dev->mode_config.prefer_shadow = 1;
 
 	r = mgag200_modeset_init(mdev);
@@ -310,7 +313,7 @@ int mgag200_dumb_create(struct drm_file *file,
 	return 0;
 }
 
-void mgag200_bo_unref(struct mgag200_bo **bo)
+static void mgag200_bo_unref(struct mgag200_bo **bo)
 {
 	struct ttm_buffer_object *tbo;
 
diff --git a/drivers/gpu/drm/mgag200/mgag200_mode.c b/drivers/gpu/drm/mgag200/mgag200_mode.c
index ee6ed633b7b1..b8583f275e80 100644
--- a/drivers/gpu/drm/mgag200/mgag200_mode.c
+++ b/drivers/gpu/drm/mgag200/mgag200_mode.c
@@ -691,7 +691,7 @@ static void mga_g200wb_commit(struct drm_crtc *crtc)
    CRTCEXT0 has to be programmed last to trigger an update and make the
    new addr variable take effect.
  */
-void mga_set_start_address(struct drm_crtc *crtc, unsigned offset)
+static void mga_set_start_address(struct drm_crtc *crtc, unsigned offset)
 {
 	struct mga_device *mdev = crtc->dev->dev_private;
 	u32 addr;
@@ -1398,7 +1398,7 @@ static void mga_encoder_commit(struct drm_encoder *encoder)
 {
 }
 
-void mga_encoder_destroy(struct drm_encoder *encoder)
+static void mga_encoder_destroy(struct drm_encoder *encoder)
 {
 	struct mga_encoder *mga_encoder = to_mga_encoder(encoder);
 	drm_encoder_cleanup(encoder);
@@ -1558,7 +1558,7 @@ static int mga_vga_mode_valid(struct drm_connector *connector,
 	return MODE_OK;
 }
 
-struct drm_encoder *mga_connector_best_encoder(struct drm_connector
+static struct drm_encoder *mga_connector_best_encoder(struct drm_connector
 						  *connector)
 {
 	int enc_id = connector->encoder_ids[0];
diff --git a/drivers/gpu/drm/mgag200/mgag200_ttm.c b/drivers/gpu/drm/mgag200/mgag200_ttm.c
index 07b192fe15c6..adb5166a5dfd 100644
--- a/drivers/gpu/drm/mgag200/mgag200_ttm.c
+++ b/drivers/gpu/drm/mgag200/mgag200_ttm.c
@@ -80,7 +80,7 @@ static int mgag200_ttm_global_init(struct mga_device *ast)
 	return 0;
 }
 
-void
+static void
 mgag200_ttm_global_release(struct mga_device *ast)
 {
 	if (ast->ttm.mem_global_ref.release == NULL)
@@ -102,7 +102,7 @@ static void mgag200_bo_ttm_destroy(struct ttm_buffer_object *tbo)
 	kfree(bo);
 }
 
-bool mgag200_ttm_bo_is_mgag200_bo(struct ttm_buffer_object *bo)
+static bool mgag200_ttm_bo_is_mgag200_bo(struct ttm_buffer_object *bo)
 {
 	if (bo->destroy == &mgag200_bo_ttm_destroy)
 		return true;
@@ -208,7 +208,7 @@ static struct ttm_backend_func mgag200_tt_backend_func = {
 };
 
 
-struct ttm_tt *mgag200_ttm_tt_create(struct ttm_bo_device *bdev,
+static struct ttm_tt *mgag200_ttm_tt_create(struct ttm_bo_device *bdev,
 				 unsigned long size, uint32_t page_flags,
 				 struct page *dummy_read_page)
 {
diff --git a/drivers/gpu/drm/msm/Kconfig b/drivers/gpu/drm/msm/Kconfig
index d3de8e1ae915..c69d1e07a3a6 100644
--- a/drivers/gpu/drm/msm/Kconfig
+++ b/drivers/gpu/drm/msm/Kconfig
@@ -2,9 +2,8 @@
 config DRM_MSM
 	tristate "MSM DRM"
 	depends on DRM
-	depends on ARCH_MSM
-	depends on ARCH_MSM8960
 	depends on MSM_IOMMU
+	depends on (ARCH_MSM && ARCH_MSM8960) || (ARM && COMPILE_TEST)
 	select DRM_KMS_HELPER
 	select SHMEM
 	select TMPFS
diff --git a/drivers/gpu/drm/msm/Makefile b/drivers/gpu/drm/msm/Makefile
index e5fa12b0d21e..4f977a593bea 100644
--- a/drivers/gpu/drm/msm/Makefile
+++ b/drivers/gpu/drm/msm/Makefile
@@ -12,18 +12,27 @@ msm-y := \
 	hdmi/hdmi_i2c.o \
 	hdmi/hdmi_phy_8960.o \
 	hdmi/hdmi_phy_8x60.o \
-	mdp4/mdp4_crtc.o \
-	mdp4/mdp4_dtv_encoder.o \
-	mdp4/mdp4_format.o \
-	mdp4/mdp4_irq.o \
-	mdp4/mdp4_kms.o \
-	mdp4/mdp4_plane.o \
+	hdmi/hdmi_phy_8x74.o \
+	mdp/mdp_format.o \
+	mdp/mdp_kms.o \
+	mdp/mdp4/mdp4_crtc.o \
+	mdp/mdp4/mdp4_dtv_encoder.o \
+	mdp/mdp4/mdp4_irq.o \
+	mdp/mdp4/mdp4_kms.o \
+	mdp/mdp4/mdp4_plane.o \
+	mdp/mdp5/mdp5_crtc.o \
+	mdp/mdp5/mdp5_encoder.o \
+	mdp/mdp5/mdp5_irq.o \
+	mdp/mdp5/mdp5_kms.o \
+	mdp/mdp5/mdp5_plane.o \
+	mdp/mdp5/mdp5_smp.o \
 	msm_drv.o \
 	msm_fb.o \
 	msm_gem.o \
 	msm_gem_prime.o \
 	msm_gem_submit.o \
 	msm_gpu.o \
+	msm_iommu.o \
 	msm_ringbuffer.o
 
 msm-$(CONFIG_DRM_MSM_FBDEV) += msm_fbdev.o
diff --git a/drivers/gpu/drm/msm/NOTES b/drivers/gpu/drm/msm/NOTES
index e036f6c1db94..9c4255b98021 100644
--- a/drivers/gpu/drm/msm/NOTES
+++ b/drivers/gpu/drm/msm/NOTES
@@ -4,7 +4,7 @@ In the current snapdragon SoC's, we have (at least) 3 different
 display controller blocks at play:
  + MDP3 - ?? seems to be what is on geeksphone peak device
  + MDP4 - S3 (APQ8060, touchpad), S4-pro (APQ8064, nexus4 & ifc6410)
- + MDSS - snapdragon 800
+ + MDP5 - snapdragon 800
 
 (I don't have a completely clear picture on which display controller
 maps to which part #)
@@ -46,6 +46,24 @@ and treat the MDP4 block's irq as "the" irq.  Even though the connectors
 may have their own irqs which they install themselves.  For this reason
 the display controller is the "master" device.
 
+For MDP5, the mapping is:
+
+  plane   -> PIPE{RGBn,VIGn}             \
+  crtc    -> LM (layer mixer)            |-> MDP "device"
+  encoder -> INTF                        /
+  connector -> HDMI/DSI/eDP/etc          --> other device(s)
+
+Unlike MDP4, it appears we can get by with a single encoder, rather
+than needing a different implementation for DTV, DSI, etc.  (Ie. the
+register interface is same, just different bases.)
+
+Also unlike MDP4, with MDP5 all the IRQs for other blocks (HDMI, DSI,
+etc) are routed through MDP.
+
+And finally, MDP5 has this "Shared Memory Pool" (called "SMP"), from
+which blocks need to be allocated to the active pipes based on fetch
+stride.
+
 Each connector probably ends up being a separate device, just for the
 logistics of finding/mapping io region, irq, etc.  Idealy we would
 have a better way than just stashing the platform device in a global
diff --git a/drivers/gpu/drm/msm/adreno/a2xx.xml.h b/drivers/gpu/drm/msm/adreno/a2xx.xml.h
index 9588098741b5..85d615e7d62f 100644
--- a/drivers/gpu/drm/msm/adreno/a2xx.xml.h
+++ b/drivers/gpu/drm/msm/adreno/a2xx.xml.h
@@ -8,12 +8,13 @@ http://github.com/freedreno/envytools/
 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/adreno.xml              (    327 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml (   1453 bytes, from 2013-03-31 16:51:27)
-- /home/robclark/src/freedreno/envytools/rnndb/a2xx/a2xx.xml           (  31003 bytes, from 2013-09-19 18:50:16)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno_common.xml       (   8983 bytes, from 2013-07-24 01:38:36)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno_pm4.xml          (   9759 bytes, from 2013-09-10 00:52:33)
-- /home/robclark/src/freedreno/envytools/rnndb/a3xx/a3xx.xml           (  51983 bytes, from 2013-09-10 00:52:32)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno.xml               (    364 bytes, from 2013-11-30 14:47:15)
+- /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          (  32814 bytes, from 2013-11-30 15:07:33)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_common.xml (   8900 bytes, from 2013-10-22 23:57:49)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_pm4.xml    (  10574 bytes, from 2013-11-13 05:44:45)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a3xx.xml          (  53644 bytes, from 2013-11-30 15:07:33)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a4xx.xml          (   8344 bytes, from 2013-11-30 14:49:47)
 
 Copyright (C) 2013 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
@@ -202,6 +203,12 @@ enum a2xx_rb_copy_sample_select {
 	SAMPLE_0123 = 6,
 };
 
+enum adreno_mmu_clnt_beh {
+	BEH_NEVR = 0,
+	BEH_TRAN_RNG = 1,
+	BEH_TRAN_FLT = 2,
+};
+
 enum sq_tex_clamp {
 	SQ_TEX_WRAP = 0,
 	SQ_TEX_MIRROR = 1,
@@ -238,6 +245,92 @@ enum sq_tex_filter {
 
 #define REG_A2XX_CP_PFP_UCODE_DATA				0x000000c1
 
+#define REG_A2XX_MH_MMU_CONFIG					0x00000040
+#define A2XX_MH_MMU_CONFIG_MMU_ENABLE				0x00000001
+#define A2XX_MH_MMU_CONFIG_SPLIT_MODE_ENABLE			0x00000002
+#define A2XX_MH_MMU_CONFIG_RB_W_CLNT_BEHAVIOR__MASK		0x00000030
+#define A2XX_MH_MMU_CONFIG_RB_W_CLNT_BEHAVIOR__SHIFT		4
+static inline uint32_t A2XX_MH_MMU_CONFIG_RB_W_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
+{
+	return ((val) << A2XX_MH_MMU_CONFIG_RB_W_CLNT_BEHAVIOR__SHIFT) & A2XX_MH_MMU_CONFIG_RB_W_CLNT_BEHAVIOR__MASK;
+}
+#define A2XX_MH_MMU_CONFIG_CP_W_CLNT_BEHAVIOR__MASK		0x000000c0
+#define A2XX_MH_MMU_CONFIG_CP_W_CLNT_BEHAVIOR__SHIFT		6
+static inline uint32_t A2XX_MH_MMU_CONFIG_CP_W_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
+{
+	return ((val) << A2XX_MH_MMU_CONFIG_CP_W_CLNT_BEHAVIOR__SHIFT) & A2XX_MH_MMU_CONFIG_CP_W_CLNT_BEHAVIOR__MASK;
+}
+#define A2XX_MH_MMU_CONFIG_CP_R0_CLNT_BEHAVIOR__MASK		0x00000300
+#define A2XX_MH_MMU_CONFIG_CP_R0_CLNT_BEHAVIOR__SHIFT		8
+static inline uint32_t A2XX_MH_MMU_CONFIG_CP_R0_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
+{
+	return ((val) << A2XX_MH_MMU_CONFIG_CP_R0_CLNT_BEHAVIOR__SHIFT) & A2XX_MH_MMU_CONFIG_CP_R0_CLNT_BEHAVIOR__MASK;
+}
+#define A2XX_MH_MMU_CONFIG_CP_R1_CLNT_BEHAVIOR__MASK		0x00000c00
+#define A2XX_MH_MMU_CONFIG_CP_R1_CLNT_BEHAVIOR__SHIFT		10
+static inline uint32_t A2XX_MH_MMU_CONFIG_CP_R1_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
+{
+	return ((val) << A2XX_MH_MMU_CONFIG_CP_R1_CLNT_BEHAVIOR__SHIFT) & A2XX_MH_MMU_CONFIG_CP_R1_CLNT_BEHAVIOR__MASK;
+}
+#define A2XX_MH_MMU_CONFIG_CP_R2_CLNT_BEHAVIOR__MASK		0x00003000
+#define A2XX_MH_MMU_CONFIG_CP_R2_CLNT_BEHAVIOR__SHIFT		12
+static inline uint32_t A2XX_MH_MMU_CONFIG_CP_R2_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
+{
+	return ((val) << A2XX_MH_MMU_CONFIG_CP_R2_CLNT_BEHAVIOR__SHIFT) & A2XX_MH_MMU_CONFIG_CP_R2_CLNT_BEHAVIOR__MASK;
+}
+#define A2XX_MH_MMU_CONFIG_CP_R3_CLNT_BEHAVIOR__MASK		0x0000c000
+#define A2XX_MH_MMU_CONFIG_CP_R3_CLNT_BEHAVIOR__SHIFT		14
+static inline uint32_t A2XX_MH_MMU_CONFIG_CP_R3_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
+{
+	return ((val) << A2XX_MH_MMU_CONFIG_CP_R3_CLNT_BEHAVIOR__SHIFT) & A2XX_MH_MMU_CONFIG_CP_R3_CLNT_BEHAVIOR__MASK;
+}
+#define A2XX_MH_MMU_CONFIG_CP_R4_CLNT_BEHAVIOR__MASK		0x00030000
+#define A2XX_MH_MMU_CONFIG_CP_R4_CLNT_BEHAVIOR__SHIFT		16
+static inline uint32_t A2XX_MH_MMU_CONFIG_CP_R4_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
+{
+	return ((val) << A2XX_MH_MMU_CONFIG_CP_R4_CLNT_BEHAVIOR__SHIFT) & A2XX_MH_MMU_CONFIG_CP_R4_CLNT_BEHAVIOR__MASK;
+}
+#define A2XX_MH_MMU_CONFIG_VGT_R0_CLNT_BEHAVIOR__MASK		0x000c0000
+#define A2XX_MH_MMU_CONFIG_VGT_R0_CLNT_BEHAVIOR__SHIFT		18
+static inline uint32_t A2XX_MH_MMU_CONFIG_VGT_R0_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
+{
+	return ((val) << A2XX_MH_MMU_CONFIG_VGT_R0_CLNT_BEHAVIOR__SHIFT) & A2XX_MH_MMU_CONFIG_VGT_R0_CLNT_BEHAVIOR__MASK;
+}
+#define A2XX_MH_MMU_CONFIG_VGT_R1_CLNT_BEHAVIOR__MASK		0x00300000
+#define A2XX_MH_MMU_CONFIG_VGT_R1_CLNT_BEHAVIOR__SHIFT		20
+static inline uint32_t A2XX_MH_MMU_CONFIG_VGT_R1_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
+{
+	return ((val) << A2XX_MH_MMU_CONFIG_VGT_R1_CLNT_BEHAVIOR__SHIFT) & A2XX_MH_MMU_CONFIG_VGT_R1_CLNT_BEHAVIOR__MASK;
+}
+#define A2XX_MH_MMU_CONFIG_TC_R_CLNT_BEHAVIOR__MASK		0x00c00000
+#define A2XX_MH_MMU_CONFIG_TC_R_CLNT_BEHAVIOR__SHIFT		22
+static inline uint32_t A2XX_MH_MMU_CONFIG_TC_R_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
+{
+	return ((val) << A2XX_MH_MMU_CONFIG_TC_R_CLNT_BEHAVIOR__SHIFT) & A2XX_MH_MMU_CONFIG_TC_R_CLNT_BEHAVIOR__MASK;
+}
+#define A2XX_MH_MMU_CONFIG_PA_W_CLNT_BEHAVIOR__MASK		0x03000000
+#define A2XX_MH_MMU_CONFIG_PA_W_CLNT_BEHAVIOR__SHIFT		24
+static inline uint32_t A2XX_MH_MMU_CONFIG_PA_W_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
+{
+	return ((val) << A2XX_MH_MMU_CONFIG_PA_W_CLNT_BEHAVIOR__SHIFT) & A2XX_MH_MMU_CONFIG_PA_W_CLNT_BEHAVIOR__MASK;
+}
+
+#define REG_A2XX_MH_MMU_VA_RANGE				0x00000041
+
+#define REG_A2XX_MH_MMU_PT_BASE					0x00000042
+
+#define REG_A2XX_MH_MMU_PAGE_FAULT				0x00000043
+
+#define REG_A2XX_MH_MMU_TRAN_ERROR				0x00000044
+
+#define REG_A2XX_MH_MMU_INVALIDATE				0x00000045
+
+#define REG_A2XX_MH_MMU_MPU_BASE				0x00000046
+
+#define REG_A2XX_MH_MMU_MPU_END					0x00000047
+
+#define REG_A2XX_NQWAIT_UNTIL					0x00000394
+
 #define REG_A2XX_RBBM_PERFCOUNTER1_SELECT			0x00000395
 
 #define REG_A2XX_RBBM_PERFCOUNTER1_LO				0x00000397
@@ -276,20 +369,6 @@ enum sq_tex_filter {
 
 #define REG_A2XX_CP_PERFCOUNTER_HI				0x00000447
 
-#define REG_A2XX_CP_ST_BASE					0x0000044d
-
-#define REG_A2XX_CP_ST_BUFSZ					0x0000044e
-
-#define REG_A2XX_CP_IB1_BASE					0x00000458
-
-#define REG_A2XX_CP_IB1_BUFSZ					0x00000459
-
-#define REG_A2XX_CP_IB2_BASE					0x0000045a
-
-#define REG_A2XX_CP_IB2_BUFSZ					0x0000045b
-
-#define REG_A2XX_CP_STAT					0x0000047f
-
 #define REG_A2XX_RBBM_STATUS					0x000005d0
 #define A2XX_RBBM_STATUS_CMDFIFO_AVAIL__MASK			0x0000001f
 #define A2XX_RBBM_STATUS_CMDFIFO_AVAIL__SHIFT			0
@@ -808,6 +887,12 @@ static inline uint32_t A2XX_SQ_CONTEXT_MISC_PARAM_GEN_POS(uint32_t val)
 
 #define REG_A2XX_SQ_VS_PROGRAM					0x000021f7
 
+#define REG_A2XX_VGT_EVENT_INITIATOR				0x000021f9
+
+#define REG_A2XX_VGT_DRAW_INITIATOR				0x000021fc
+
+#define REG_A2XX_VGT_IMMED_DATA					0x000021fd
+
 #define REG_A2XX_RB_DEPTHCONTROL				0x00002200
 #define A2XX_RB_DEPTHCONTROL_STENCIL_ENABLE			0x00000001
 #define A2XX_RB_DEPTHCONTROL_Z_ENABLE				0x00000002
diff --git a/drivers/gpu/drm/msm/adreno/a3xx.xml.h b/drivers/gpu/drm/msm/adreno/a3xx.xml.h
index d4afdf657559..a7be56163d23 100644
--- a/drivers/gpu/drm/msm/adreno/a3xx.xml.h
+++ b/drivers/gpu/drm/msm/adreno/a3xx.xml.h
@@ -8,12 +8,13 @@ http://github.com/freedreno/envytools/
 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/adreno.xml              (    327 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml (   1453 bytes, from 2013-03-31 16:51:27)
-- /home/robclark/src/freedreno/envytools/rnndb/a2xx/a2xx.xml           (  31003 bytes, from 2013-09-19 18:50:16)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno_common.xml       (   8983 bytes, from 2013-07-24 01:38:36)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno_pm4.xml          (   9759 bytes, from 2013-09-10 00:52:33)
-- /home/robclark/src/freedreno/envytools/rnndb/a3xx/a3xx.xml           (  51983 bytes, from 2013-09-10 00:52:32)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno.xml               (    364 bytes, from 2013-11-30 14:47:15)
+- /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          (  32814 bytes, from 2013-11-30 15:07:33)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_common.xml (   8900 bytes, from 2013-10-22 23:57:49)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_pm4.xml    (  10574 bytes, from 2013-11-13 05:44:45)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a3xx.xml          (  53644 bytes, from 2013-11-30 15:07:33)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a4xx.xml          (   8344 bytes, from 2013-11-30 14:49:47)
 
 Copyright (C) 2013 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
@@ -292,6 +293,8 @@ enum a3xx_tex_type {
 #define A3XX_RBBM_STATUS_GPU_BUSY_NOHC				0x40000000
 #define A3XX_RBBM_STATUS_GPU_BUSY				0x80000000
 
+#define REG_A3XX_RBBM_NQWAIT_UNTIL				0x00000040
+
 #define REG_A3XX_RBBM_WAIT_IDLE_CLOCKS_CTL			0x00000033
 
 #define REG_A3XX_RBBM_INTERFACE_HANG_INT_CTL			0x00000050
@@ -304,6 +307,8 @@ enum a3xx_tex_type {
 
 #define REG_A3XX_RBBM_INTERFACE_HANG_MASK_CTL3			0x0000005a
 
+#define REG_A3XX_RBBM_INT_SET_CMD				0x00000060
+
 #define REG_A3XX_RBBM_INT_CLEAR_CMD				0x00000061
 
 #define REG_A3XX_RBBM_INT_0_MASK				0x00000063
@@ -937,13 +942,13 @@ static inline uint32_t A3XX_RB_BLEND_ALPHA_FLOAT(float val)
 	return ((util_float_to_half(val)) << A3XX_RB_BLEND_ALPHA_FLOAT__SHIFT) & A3XX_RB_BLEND_ALPHA_FLOAT__MASK;
 }
 
-#define REG_A3XX_UNKNOWN_20E8					0x000020e8
+#define REG_A3XX_RB_CLEAR_COLOR_DW0				0x000020e8
 
-#define REG_A3XX_UNKNOWN_20E9					0x000020e9
+#define REG_A3XX_RB_CLEAR_COLOR_DW1				0x000020e9
 
-#define REG_A3XX_UNKNOWN_20EA					0x000020ea
+#define REG_A3XX_RB_CLEAR_COLOR_DW2				0x000020ea
 
-#define REG_A3XX_UNKNOWN_20EB					0x000020eb
+#define REG_A3XX_RB_CLEAR_COLOR_DW3				0x000020eb
 
 #define REG_A3XX_RB_COPY_CONTROL				0x000020ec
 #define A3XX_RB_COPY_CONTROL_MSAA_RESOLVE__MASK			0x00000003
@@ -1026,7 +1031,7 @@ static inline uint32_t A3XX_RB_DEPTH_CONTROL_ZFUNC(enum adreno_compare_func val)
 #define A3XX_RB_DEPTH_CONTROL_BF_ENABLE				0x00000080
 #define A3XX_RB_DEPTH_CONTROL_Z_TEST_ENABLE			0x80000000
 
-#define REG_A3XX_UNKNOWN_2101					0x00002101
+#define REG_A3XX_RB_DEPTH_CLEAR					0x00002101
 
 #define REG_A3XX_RB_DEPTH_INFO					0x00002102
 #define A3XX_RB_DEPTH_INFO_DEPTH_FORMAT__MASK			0x00000001
@@ -1103,11 +1108,11 @@ static inline uint32_t A3XX_RB_STENCIL_CONTROL_ZFAIL_BF(enum adreno_stencil_op v
 	return ((val) << A3XX_RB_STENCIL_CONTROL_ZFAIL_BF__SHIFT) & A3XX_RB_STENCIL_CONTROL_ZFAIL_BF__MASK;
 }
 
-#define REG_A3XX_UNKNOWN_2105					0x00002105
+#define REG_A3XX_RB_STENCIL_CLEAR				0x00002105
 
-#define REG_A3XX_UNKNOWN_2106					0x00002106
+#define REG_A3XX_RB_STENCIL_BUF_INFO				0x00002106
 
-#define REG_A3XX_UNKNOWN_2107					0x00002107
+#define REG_A3XX_RB_STENCIL_BUF_PITCH				0x00002107
 
 #define REG_A3XX_RB_STENCILREFMASK				0x00002108
 #define A3XX_RB_STENCILREFMASK_STENCILREF__MASK			0x000000ff
@@ -1149,20 +1154,31 @@ static inline uint32_t A3XX_RB_STENCILREFMASK_BF_STENCILWRITEMASK(uint32_t val)
 	return ((val) << A3XX_RB_STENCILREFMASK_BF_STENCILWRITEMASK__SHIFT) & A3XX_RB_STENCILREFMASK_BF_STENCILWRITEMASK__MASK;
 }
 
-#define REG_A3XX_PA_SC_WINDOW_OFFSET				0x0000210e
-#define A3XX_PA_SC_WINDOW_OFFSET_X__MASK			0x0000ffff
-#define A3XX_PA_SC_WINDOW_OFFSET_X__SHIFT			0
-static inline uint32_t A3XX_PA_SC_WINDOW_OFFSET_X(uint32_t val)
+#define REG_A3XX_RB_LRZ_VSC_CONTROL				0x0000210c
+#define A3XX_RB_LRZ_VSC_CONTROL_BINNING_ENABLE			0x00000002
+
+#define REG_A3XX_RB_WINDOW_OFFSET				0x0000210e
+#define A3XX_RB_WINDOW_OFFSET_X__MASK				0x0000ffff
+#define A3XX_RB_WINDOW_OFFSET_X__SHIFT				0
+static inline uint32_t A3XX_RB_WINDOW_OFFSET_X(uint32_t val)
 {
-	return ((val) << A3XX_PA_SC_WINDOW_OFFSET_X__SHIFT) & A3XX_PA_SC_WINDOW_OFFSET_X__MASK;
+	return ((val) << A3XX_RB_WINDOW_OFFSET_X__SHIFT) & A3XX_RB_WINDOW_OFFSET_X__MASK;
 }
-#define A3XX_PA_SC_WINDOW_OFFSET_Y__MASK			0xffff0000
-#define A3XX_PA_SC_WINDOW_OFFSET_Y__SHIFT			16
-static inline uint32_t A3XX_PA_SC_WINDOW_OFFSET_Y(uint32_t val)
+#define A3XX_RB_WINDOW_OFFSET_Y__MASK				0xffff0000
+#define A3XX_RB_WINDOW_OFFSET_Y__SHIFT				16
+static inline uint32_t A3XX_RB_WINDOW_OFFSET_Y(uint32_t val)
 {
-	return ((val) << A3XX_PA_SC_WINDOW_OFFSET_Y__SHIFT) & A3XX_PA_SC_WINDOW_OFFSET_Y__MASK;
+	return ((val) << A3XX_RB_WINDOW_OFFSET_Y__SHIFT) & A3XX_RB_WINDOW_OFFSET_Y__MASK;
 }
 
+#define REG_A3XX_RB_SAMPLE_COUNT_CONTROL			0x00002110
+
+#define REG_A3XX_RB_SAMPLE_COUNT_ADDR				0x00002111
+
+#define REG_A3XX_RB_Z_CLAMP_MIN					0x00002114
+
+#define REG_A3XX_RB_Z_CLAMP_MAX					0x00002115
+
 #define REG_A3XX_PC_VSTREAM_CONTROL				0x000021e4
 
 #define REG_A3XX_PC_VERTEX_REUSE_BLOCK_CNTL			0x000021ea
@@ -1309,6 +1325,8 @@ static inline uint32_t A3XX_HLSQ_CONST_FSPRESV_RANGE_REG_ENDENTRY(uint32_t val)
 
 #define REG_A3XX_HLSQ_CL_KERNEL_GROUP_X_REG			0x00002215
 
+#define REG_A3XX_HLSQ_CL_KERNEL_GROUP_Y_REG			0x00002216
+
 #define REG_A3XX_HLSQ_CL_KERNEL_GROUP_Z_REG			0x00002217
 
 #define REG_A3XX_HLSQ_CL_WG_OFFSET_REG				0x0000221a
@@ -1491,12 +1509,13 @@ static inline uint32_t REG_A3XX_VPC_VARYING_PS_REPL_MODE(uint32_t i0) { return 0
 
 #define REG_A3XX_SP_SP_CTRL_REG					0x000022c0
 #define A3XX_SP_SP_CTRL_REG_RESOLVE				0x00010000
-#define A3XX_SP_SP_CTRL_REG_CONSTMODE__MASK			0x000c0000
+#define A3XX_SP_SP_CTRL_REG_CONSTMODE__MASK			0x00040000
 #define A3XX_SP_SP_CTRL_REG_CONSTMODE__SHIFT			18
 static inline uint32_t A3XX_SP_SP_CTRL_REG_CONSTMODE(uint32_t val)
 {
 	return ((val) << A3XX_SP_SP_CTRL_REG_CONSTMODE__SHIFT) & A3XX_SP_SP_CTRL_REG_CONSTMODE__MASK;
 }
+#define A3XX_SP_SP_CTRL_REG_BINNING				0x00080000
 #define A3XX_SP_SP_CTRL_REG_SLEEPMODE__MASK			0x00300000
 #define A3XX_SP_SP_CTRL_REG_SLEEPMODE__SHIFT			20
 static inline uint32_t A3XX_SP_SP_CTRL_REG_SLEEPMODE(uint32_t val)
@@ -1669,7 +1688,7 @@ static inline uint32_t A3XX_SP_VS_OBJ_OFFSET_REG_SHADEROBJOFFSET(uint32_t val)
 
 #define REG_A3XX_SP_VS_OBJ_START_REG				0x000022d5
 
-#define REG_A3XX_SP_VS_PVT_MEM_CTRL_REG				0x000022d6
+#define REG_A3XX_SP_VS_PVT_MEM_PARAM_REG			0x000022d6
 
 #define REG_A3XX_SP_VS_PVT_MEM_ADDR_REG				0x000022d7
 
@@ -1772,7 +1791,7 @@ static inline uint32_t A3XX_SP_FS_OBJ_OFFSET_REG_SHADEROBJOFFSET(uint32_t val)
 
 #define REG_A3XX_SP_FS_OBJ_START_REG				0x000022e3
 
-#define REG_A3XX_SP_FS_PVT_MEM_CTRL_REG				0x000022e4
+#define REG_A3XX_SP_FS_PVT_MEM_PARAM_REG			0x000022e4
 
 #define REG_A3XX_SP_FS_PVT_MEM_ADDR_REG				0x000022e5
 
@@ -1943,6 +1962,9 @@ static inline uint32_t REG_A3XX_VSC_PIPE_DATA_ADDRESS(uint32_t i0) { return 0x00
 
 static inline uint32_t REG_A3XX_VSC_PIPE_DATA_LENGTH(uint32_t i0) { return 0x00000c08 + 0x3*i0; }
 
+#define REG_A3XX_VSC_BIN_CONTROL				0x00000c3c
+#define A3XX_VSC_BIN_CONTROL_BINNING_ENABLE			0x00000001
+
 #define REG_A3XX_UNKNOWN_0C3D					0x00000c3d
 
 #define REG_A3XX_PC_PERFCOUNTER0_SELECT				0x00000c48
@@ -1953,7 +1975,7 @@ static inline uint32_t REG_A3XX_VSC_PIPE_DATA_LENGTH(uint32_t i0) { return 0x000
 
 #define REG_A3XX_PC_PERFCOUNTER3_SELECT				0x00000c4b
 
-#define REG_A3XX_UNKNOWN_0C81					0x00000c81
+#define REG_A3XX_GRAS_TSE_DEBUG_ECO				0x00000c81
 
 #define REG_A3XX_GRAS_PERFCOUNTER0_SELECT			0x00000c88
 
@@ -1975,22 +1997,24 @@ static inline uint32_t REG_A3XX_GRAS_CL_USER_PLANE_W(uint32_t i0) { return 0x000
 
 #define REG_A3XX_RB_GMEM_BASE_ADDR				0x00000cc0
 
+#define REG_A3XX_RB_DEBUG_ECO_CONTROLS_ADDR			0x00000cc1
+
 #define REG_A3XX_RB_PERFCOUNTER0_SELECT				0x00000cc6
 
 #define REG_A3XX_RB_PERFCOUNTER1_SELECT				0x00000cc7
 
-#define REG_A3XX_RB_WINDOW_SIZE					0x00000ce0
-#define A3XX_RB_WINDOW_SIZE_WIDTH__MASK				0x00003fff
-#define A3XX_RB_WINDOW_SIZE_WIDTH__SHIFT			0
-static inline uint32_t A3XX_RB_WINDOW_SIZE_WIDTH(uint32_t val)
+#define REG_A3XX_RB_FRAME_BUFFER_DIMENSION			0x00000ce0
+#define A3XX_RB_FRAME_BUFFER_DIMENSION_WIDTH__MASK		0x00003fff
+#define A3XX_RB_FRAME_BUFFER_DIMENSION_WIDTH__SHIFT		0
+static inline uint32_t A3XX_RB_FRAME_BUFFER_DIMENSION_WIDTH(uint32_t val)
 {
-	return ((val) << A3XX_RB_WINDOW_SIZE_WIDTH__SHIFT) & A3XX_RB_WINDOW_SIZE_WIDTH__MASK;
+	return ((val) << A3XX_RB_FRAME_BUFFER_DIMENSION_WIDTH__SHIFT) & A3XX_RB_FRAME_BUFFER_DIMENSION_WIDTH__MASK;
 }
-#define A3XX_RB_WINDOW_SIZE_HEIGHT__MASK			0x0fffc000
-#define A3XX_RB_WINDOW_SIZE_HEIGHT__SHIFT			14
-static inline uint32_t A3XX_RB_WINDOW_SIZE_HEIGHT(uint32_t val)
+#define A3XX_RB_FRAME_BUFFER_DIMENSION_HEIGHT__MASK		0x0fffc000
+#define A3XX_RB_FRAME_BUFFER_DIMENSION_HEIGHT__SHIFT		14
+static inline uint32_t A3XX_RB_FRAME_BUFFER_DIMENSION_HEIGHT(uint32_t val)
 {
-	return ((val) << A3XX_RB_WINDOW_SIZE_HEIGHT__SHIFT) & A3XX_RB_WINDOW_SIZE_HEIGHT__MASK;
+	return ((val) << A3XX_RB_FRAME_BUFFER_DIMENSION_HEIGHT__SHIFT) & A3XX_RB_FRAME_BUFFER_DIMENSION_HEIGHT__MASK;
 }
 
 #define REG_A3XX_HLSQ_PERFCOUNTER0_SELECT			0x00000e00
@@ -2088,6 +2112,14 @@ static inline uint32_t A3XX_UCHE_CACHE_INVALIDATE1_REG_OPCODE(enum a3xx_cache_op
 
 #define REG_A3XX_TP_PERFCOUNTER5_SELECT				0x00000f09
 
+#define REG_A3XX_VGT_CL_INITIATOR				0x000021f0
+
+#define REG_A3XX_VGT_EVENT_INITIATOR				0x000021f9
+
+#define REG_A3XX_VGT_DRAW_INITIATOR				0x000021fc
+
+#define REG_A3XX_VGT_IMMED_DATA					0x000021fd
+
 #define REG_A3XX_TEX_SAMP_0					0x00000000
 #define A3XX_TEX_SAMP_0_MIPFILTER_LINEAR			0x00000002
 #define A3XX_TEX_SAMP_0_XY_MAG__MASK				0x0000000c
@@ -2123,6 +2155,18 @@ static inline uint32_t A3XX_TEX_SAMP_0_WRAP_R(enum a3xx_tex_clamp val)
 #define A3XX_TEX_SAMP_0_UNNORM_COORDS				0x80000000
 
 #define REG_A3XX_TEX_SAMP_1					0x00000001
+#define A3XX_TEX_SAMP_1_MAX_LOD__MASK				0x003ff000
+#define A3XX_TEX_SAMP_1_MAX_LOD__SHIFT				12
+static inline uint32_t A3XX_TEX_SAMP_1_MAX_LOD(float val)
+{
+	return ((((uint32_t)(val * 12.0))) << A3XX_TEX_SAMP_1_MAX_LOD__SHIFT) & A3XX_TEX_SAMP_1_MAX_LOD__MASK;
+}
+#define A3XX_TEX_SAMP_1_MIN_LOD__MASK				0xffc00000
+#define A3XX_TEX_SAMP_1_MIN_LOD__SHIFT				22
+static inline uint32_t A3XX_TEX_SAMP_1_MIN_LOD(float val)
+{
+	return ((((uint32_t)(val * 12.0))) << A3XX_TEX_SAMP_1_MIN_LOD__SHIFT) & A3XX_TEX_SAMP_1_MIN_LOD__MASK;
+}
 
 #define REG_A3XX_TEX_CONST_0					0x00000000
 #define A3XX_TEX_CONST_0_TILED					0x00000001
diff --git a/drivers/gpu/drm/msm/adreno/a3xx_gpu.c b/drivers/gpu/drm/msm/adreno/a3xx_gpu.c
index 035bd13dc8bd..461df93e825e 100644
--- a/drivers/gpu/drm/msm/adreno/a3xx_gpu.c
+++ b/drivers/gpu/drm/msm/adreno/a3xx_gpu.c
@@ -15,6 +15,10 @@
  * this program.  If not, see <http://www.gnu.org/licenses/>.
  */
 
+#ifdef CONFIG_MSM_OCMEM
+#  include <mach/ocmem.h>
+#endif
+
 #include "a3xx_gpu.h"
 
 #define A3XX_INT0_MASK \
@@ -63,6 +67,7 @@ static void a3xx_me_init(struct msm_gpu *gpu)
 static int a3xx_hw_init(struct msm_gpu *gpu)
 {
 	struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
+	struct a3xx_gpu *a3xx_gpu = to_a3xx_gpu(adreno_gpu);
 	uint32_t *ptr, len;
 	int i, ret;
 
@@ -105,6 +110,21 @@ static int a3xx_hw_init(struct msm_gpu *gpu)
 		gpu_write(gpu, REG_A3XX_VBIF_ABIT_SORT, 0x000000ff);
 		gpu_write(gpu, REG_A3XX_VBIF_ABIT_SORT_CONF, 0x000000a4);
 
+	} else if (adreno_is_a330v2(adreno_gpu)) {
+		/*
+		 * Most of the VBIF registers on 8974v2 have the correct
+		 * values at power on, so we won't modify those if we don't
+		 * need to
+		 */
+		/* Enable 1k sort: */
+		gpu_write(gpu, REG_A3XX_VBIF_ABIT_SORT, 0x0001003f);
+		gpu_write(gpu, REG_A3XX_VBIF_ABIT_SORT_CONF, 0x000000a4);
+		/* Enable WR-REQ: */
+		gpu_write(gpu, REG_A3XX_VBIF_GATE_OFF_WRREQ_EN, 0x00003f);
+		gpu_write(gpu, REG_A3XX_VBIF_DDR_OUT_MAX_BURST, 0x0000303);
+		/* Set up VBIF_ROUND_ROBIN_QOS_ARB: */
+		gpu_write(gpu, REG_A3XX_VBIF_ROUND_ROBIN_QOS_ARB, 0x0003);
+
 	} else if (adreno_is_a330(adreno_gpu)) {
 		/* Set up 16 deep read/write request queues: */
 		gpu_write(gpu, REG_A3XX_VBIF_IN_RD_LIM_CONF0, 0x18181818);
@@ -121,10 +141,10 @@ static int a3xx_hw_init(struct msm_gpu *gpu)
 		/* Set up VBIF_ROUND_ROBIN_QOS_ARB: */
 		gpu_write(gpu, REG_A3XX_VBIF_ROUND_ROBIN_QOS_ARB, 0x0001);
 		/* Set up AOOO: */
-		gpu_write(gpu, REG_A3XX_VBIF_OUT_AXI_AOOO_EN, 0x0000ffff);
-		gpu_write(gpu, REG_A3XX_VBIF_OUT_AXI_AOOO, 0xffffffff);
+		gpu_write(gpu, REG_A3XX_VBIF_OUT_AXI_AOOO_EN, 0x0000003f);
+		gpu_write(gpu, REG_A3XX_VBIF_OUT_AXI_AOOO, 0x003f003f);
 		/* Enable 1K sort: */
-		gpu_write(gpu, REG_A3XX_VBIF_ABIT_SORT, 0x0001ffff);
+		gpu_write(gpu, REG_A3XX_VBIF_ABIT_SORT, 0x0001003f);
 		gpu_write(gpu, REG_A3XX_VBIF_ABIT_SORT_CONF, 0x000000a4);
 		/* Disable VBIF clock gating. This is to enable AXI running
 		 * higher frequency than GPU:
@@ -162,14 +182,23 @@ static int a3xx_hw_init(struct msm_gpu *gpu)
 	gpu_write(gpu, REG_A3XX_UCHE_CACHE_MODE_CONTROL_REG, 0x00000001);
 
 	/* Enable Clock gating: */
-	gpu_write(gpu, REG_A3XX_RBBM_CLOCK_CTL, 0xbfffffff);
-
-	/* Set the OCMEM base address for A330 */
-//TODO:
-//	if (adreno_is_a330(adreno_gpu)) {
-//		gpu_write(gpu, REG_A3XX_RB_GMEM_BASE_ADDR,
-//			(unsigned int)(a3xx_gpu->ocmem_base >> 14));
-//	}
+	if (adreno_is_a320(adreno_gpu))
+		gpu_write(gpu, REG_A3XX_RBBM_CLOCK_CTL, 0xbfffffff);
+	else if (adreno_is_a330v2(adreno_gpu))
+		gpu_write(gpu, REG_A3XX_RBBM_CLOCK_CTL, 0xaaaaaaaa);
+	else if (adreno_is_a330(adreno_gpu))
+		gpu_write(gpu, REG_A3XX_RBBM_CLOCK_CTL, 0xbffcffff);
+
+	if (adreno_is_a330v2(adreno_gpu))
+		gpu_write(gpu, REG_A3XX_RBBM_GPR0_CTL, 0x05515455);
+	else if (adreno_is_a330(adreno_gpu))
+		gpu_write(gpu, REG_A3XX_RBBM_GPR0_CTL, 0x00000000);
+
+	/* Set the OCMEM base address for A330, etc */
+	if (a3xx_gpu->ocmem_hdl) {
+		gpu_write(gpu, REG_A3XX_RB_GMEM_BASE_ADDR,
+			(unsigned int)(a3xx_gpu->ocmem_base >> 14));
+	}
 
 	/* Turn on performance counters: */
 	gpu_write(gpu, REG_A3XX_RBBM_PERFCTR_CTL, 0x01);
@@ -219,7 +248,7 @@ static int a3xx_hw_init(struct msm_gpu *gpu)
 	/* Load PM4: */
 	ptr = (uint32_t *)(adreno_gpu->pm4->data);
 	len = adreno_gpu->pm4->size / 4;
-	DBG("loading PM4 ucode version: %u", ptr[0]);
+	DBG("loading PM4 ucode version: %x", ptr[1]);
 
 	gpu_write(gpu, REG_AXXX_CP_DEBUG,
 			AXXX_CP_DEBUG_DYNAMIC_CLK_DISABLE |
@@ -231,19 +260,26 @@ static int a3xx_hw_init(struct msm_gpu *gpu)
 	/* Load PFP: */
 	ptr = (uint32_t *)(adreno_gpu->pfp->data);
 	len = adreno_gpu->pfp->size / 4;
-	DBG("loading PFP ucode version: %u", ptr[0]);
+	DBG("loading PFP ucode version: %x", ptr[5]);
 
 	gpu_write(gpu, REG_A3XX_CP_PFP_UCODE_ADDR, 0);
 	for (i = 1; i < len; i++)
 		gpu_write(gpu, REG_A3XX_CP_PFP_UCODE_DATA, ptr[i]);
 
 	/* CP ROQ queue sizes (bytes) - RB:16, ST:16, IB1:32, IB2:64 */
-	if (adreno_is_a305(adreno_gpu) || adreno_is_a320(adreno_gpu))
+	if (adreno_is_a305(adreno_gpu) || adreno_is_a320(adreno_gpu)) {
 		gpu_write(gpu, REG_AXXX_CP_QUEUE_THRESHOLDS,
 				AXXX_CP_QUEUE_THRESHOLDS_CSQ_IB1_START(2) |
 				AXXX_CP_QUEUE_THRESHOLDS_CSQ_IB2_START(6) |
 				AXXX_CP_QUEUE_THRESHOLDS_CSQ_ST_START(14));
-
+	} else if (adreno_is_a330(adreno_gpu)) {
+		/* NOTE: this (value take from downstream android driver)
+		 * includes some bits outside of the known bitfields.  But
+		 * A330 has this "MERCIU queue" thing too, which might
+		 * explain a new bitfield or reshuffling:
+		 */
+		gpu_write(gpu, REG_AXXX_CP_QUEUE_THRESHOLDS, 0x003e2008);
+	}
 
 	/* clear ME_HALT to start micro engine */
 	gpu_write(gpu, REG_AXXX_CP_ME_CNTL, 0);
@@ -253,6 +289,14 @@ static int a3xx_hw_init(struct msm_gpu *gpu)
 	return 0;
 }
 
+static void a3xx_recover(struct msm_gpu *gpu)
+{
+	gpu_write(gpu, REG_A3XX_RBBM_SW_RESET_CMD, 1);
+	gpu_read(gpu, REG_A3XX_RBBM_SW_RESET_CMD);
+	gpu_write(gpu, REG_A3XX_RBBM_SW_RESET_CMD, 0);
+	adreno_recover(gpu);
+}
+
 static void a3xx_destroy(struct msm_gpu *gpu)
 {
 	struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
@@ -261,6 +305,12 @@ static void a3xx_destroy(struct msm_gpu *gpu)
 	DBG("%s", gpu->name);
 
 	adreno_gpu_cleanup(adreno_gpu);
+
+#ifdef CONFIG_MSM_OCMEM
+	if (a3xx_gpu->ocmem_base)
+		ocmem_free(OCMEM_GRAPHICS, a3xx_gpu->ocmem_hdl);
+#endif
+
 	put_device(&a3xx_gpu->pdev->dev);
 	kfree(a3xx_gpu);
 }
@@ -371,7 +421,7 @@ static const struct adreno_gpu_funcs funcs = {
 		.hw_init = a3xx_hw_init,
 		.pm_suspend = msm_gpu_pm_suspend,
 		.pm_resume = msm_gpu_pm_resume,
-		.recover = adreno_recover,
+		.recover = a3xx_recover,
 		.last_fence = adreno_last_fence,
 		.submit = adreno_submit,
 		.flush = adreno_flush,
@@ -387,6 +437,7 @@ static const struct adreno_gpu_funcs funcs = {
 struct msm_gpu *a3xx_gpu_init(struct drm_device *dev)
 {
 	struct a3xx_gpu *a3xx_gpu = NULL;
+	struct adreno_gpu *adreno_gpu;
 	struct msm_gpu *gpu;
 	struct platform_device *pdev = a3xx_pdev;
 	struct adreno_platform_config *config;
@@ -406,7 +457,8 @@ struct msm_gpu *a3xx_gpu_init(struct drm_device *dev)
 		goto fail;
 	}
 
-	gpu = &a3xx_gpu->base.base;
+	adreno_gpu = &a3xx_gpu->base;
+	gpu = &adreno_gpu->base;
 
 	get_device(&pdev->dev);
 	a3xx_gpu->pdev = pdev;
@@ -414,16 +466,46 @@ struct msm_gpu *a3xx_gpu_init(struct drm_device *dev)
 	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);
 
-	ret = adreno_gpu_init(dev, pdev, &a3xx_gpu->base,
-			&funcs, config->rev);
+	ret = adreno_gpu_init(dev, pdev, adreno_gpu, &funcs, config->rev);
 	if (ret)
 		goto fail;
 
-	return &a3xx_gpu->base.base;
+	/* if needed, allocate gmem: */
+	if (adreno_is_a330(adreno_gpu)) {
+#ifdef CONFIG_MSM_OCMEM
+		/* TODO this is different/missing upstream: */
+		struct ocmem_buf *ocmem_hdl =
+				ocmem_allocate(OCMEM_GRAPHICS, adreno_gpu->gmem);
+
+		a3xx_gpu->ocmem_hdl = ocmem_hdl;
+		a3xx_gpu->ocmem_base = ocmem_hdl->addr;
+		adreno_gpu->gmem = ocmem_hdl->len;
+		DBG("using %dK of OCMEM at 0x%08x", adreno_gpu->gmem / 1024,
+				a3xx_gpu->ocmem_base);
+#endif
+	}
+
+	if (!gpu->mmu) {
+		/* TODO we think it is possible to configure the GPU to
+		 * restrict access to VRAM carveout.  But the required
+		 * registers are unknown.  For now just bail out and
+		 * limp along with just modesetting.  If it turns out
+		 * to not be possible to restrict access, then we must
+		 * implement a cmdstream validator.
+		 */
+		dev_err(dev->dev, "No memory protection without IOMMU\n");
+		ret = -ENXIO;
+		goto fail;
+	}
+
+	return gpu;
 
 fail:
 	if (a3xx_gpu)
@@ -436,19 +518,59 @@ fail:
  * The a3xx device:
  */
 
+#if defined(CONFIG_MSM_BUS_SCALING) && !defined(CONFIG_OF)
+#  include <mach/kgsl.h>
+#endif
+
 static int a3xx_probe(struct platform_device *pdev)
 {
 	static struct adreno_platform_config config = {};
 #ifdef CONFIG_OF
-	/* TODO */
+	struct device_node *child, *node = pdev->dev.of_node;
+	u32 val;
+	int ret;
+
+	ret = of_property_read_u32(node, "qcom,chipid", &val);
+	if (ret) {
+		dev_err(&pdev->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(&pdev->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(&pdev->dev, "could not find clk rates\n");
+		return -ENXIO;
+	}
+
 #else
+	struct kgsl_device_platform_data *pdata = pdev->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() || cpu_is_msm8960ab()) {
+	} else if (cpu_is_apq8064()) {
 		config.fast_rate = 400000000;
 		config.slow_rate = 27000000;
 		config.bus_freq  = 4;
@@ -461,6 +583,16 @@ static int a3xx_probe(struct platform_device *pdev)
 		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;
@@ -473,6 +605,9 @@ static int a3xx_probe(struct platform_device *pdev)
 			config.rev = ADRENO_REV(3, 0, 5, 0);
 
 	}
+#  ifdef CONFIG_MSM_BUS_SCALING
+	config.bus_scale_table = pdata->bus_scale_table;
+#  endif
 #endif
 	pdev->dev.platform_data = &config;
 	a3xx_pdev = pdev;
@@ -485,10 +620,19 @@ static int a3xx_remove(struct platform_device *pdev)
 	return 0;
 }
 
+static const struct of_device_id dt_match[] = {
+	{ .compatible = "qcom,kgsl-3d0" },
+	{}
+};
+MODULE_DEVICE_TABLE(of, dt_match);
+
 static struct platform_driver a3xx_driver = {
 	.probe = a3xx_probe,
 	.remove = a3xx_remove,
-	.driver.name = "kgsl-3d0",
+	.driver = {
+		.name = "kgsl-3d0",
+		.of_match_table = dt_match,
+	},
 };
 
 void __init a3xx_register(void)
diff --git a/drivers/gpu/drm/msm/adreno/a3xx_gpu.h b/drivers/gpu/drm/msm/adreno/a3xx_gpu.h
index 32c398c2d00a..bb9a8ca0507b 100644
--- a/drivers/gpu/drm/msm/adreno/a3xx_gpu.h
+++ b/drivers/gpu/drm/msm/adreno/a3xx_gpu.h
@@ -24,6 +24,10 @@
 struct a3xx_gpu {
 	struct adreno_gpu base;
 	struct platform_device *pdev;
+
+	/* if OCMEM is used for GMEM: */
+	uint32_t ocmem_base;
+	void *ocmem_hdl;
 };
 #define to_a3xx_gpu(x) container_of(x, struct a3xx_gpu, base)
 
diff --git a/drivers/gpu/drm/msm/adreno/adreno_common.xml.h b/drivers/gpu/drm/msm/adreno/adreno_common.xml.h
index 33dcc606c7c5..d6e6ce2d1abd 100644
--- a/drivers/gpu/drm/msm/adreno/adreno_common.xml.h
+++ b/drivers/gpu/drm/msm/adreno/adreno_common.xml.h
@@ -8,12 +8,13 @@ http://github.com/freedreno/envytools/
 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/adreno.xml              (    327 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml (   1453 bytes, from 2013-03-31 16:51:27)
-- /home/robclark/src/freedreno/envytools/rnndb/a2xx/a2xx.xml           (  31003 bytes, from 2013-09-19 18:50:16)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno_common.xml       (   8983 bytes, from 2013-07-24 01:38:36)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno_pm4.xml          (   9759 bytes, from 2013-09-10 00:52:33)
-- /home/robclark/src/freedreno/envytools/rnndb/a3xx/a3xx.xml           (  51983 bytes, from 2013-09-10 00:52:32)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno.xml               (    364 bytes, from 2013-11-30 14:47:15)
+- /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          (  32814 bytes, from 2013-11-30 15:07:33)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_common.xml (   8900 bytes, from 2013-10-22 23:57:49)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_pm4.xml    (  10574 bytes, from 2013-11-13 05:44:45)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a3xx.xml          (  53644 bytes, from 2013-11-30 15:07:33)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a4xx.xml          (   8344 bytes, from 2013-11-30 14:49:47)
 
 Copyright (C) 2013 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
@@ -115,96 +116,6 @@ enum adreno_rb_depth_format {
 	DEPTHX_24_8 = 1,
 };
 
-enum adreno_mmu_clnt_beh {
-	BEH_NEVR = 0,
-	BEH_TRAN_RNG = 1,
-	BEH_TRAN_FLT = 2,
-};
-
-#define REG_AXXX_MH_MMU_CONFIG					0x00000040
-#define AXXX_MH_MMU_CONFIG_MMU_ENABLE				0x00000001
-#define AXXX_MH_MMU_CONFIG_SPLIT_MODE_ENABLE			0x00000002
-#define AXXX_MH_MMU_CONFIG_RB_W_CLNT_BEHAVIOR__MASK		0x00000030
-#define AXXX_MH_MMU_CONFIG_RB_W_CLNT_BEHAVIOR__SHIFT		4
-static inline uint32_t AXXX_MH_MMU_CONFIG_RB_W_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
-{
-	return ((val) << AXXX_MH_MMU_CONFIG_RB_W_CLNT_BEHAVIOR__SHIFT) & AXXX_MH_MMU_CONFIG_RB_W_CLNT_BEHAVIOR__MASK;
-}
-#define AXXX_MH_MMU_CONFIG_CP_W_CLNT_BEHAVIOR__MASK		0x000000c0
-#define AXXX_MH_MMU_CONFIG_CP_W_CLNT_BEHAVIOR__SHIFT		6
-static inline uint32_t AXXX_MH_MMU_CONFIG_CP_W_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
-{
-	return ((val) << AXXX_MH_MMU_CONFIG_CP_W_CLNT_BEHAVIOR__SHIFT) & AXXX_MH_MMU_CONFIG_CP_W_CLNT_BEHAVIOR__MASK;
-}
-#define AXXX_MH_MMU_CONFIG_CP_R0_CLNT_BEHAVIOR__MASK		0x00000300
-#define AXXX_MH_MMU_CONFIG_CP_R0_CLNT_BEHAVIOR__SHIFT		8
-static inline uint32_t AXXX_MH_MMU_CONFIG_CP_R0_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
-{
-	return ((val) << AXXX_MH_MMU_CONFIG_CP_R0_CLNT_BEHAVIOR__SHIFT) & AXXX_MH_MMU_CONFIG_CP_R0_CLNT_BEHAVIOR__MASK;
-}
-#define AXXX_MH_MMU_CONFIG_CP_R1_CLNT_BEHAVIOR__MASK		0x00000c00
-#define AXXX_MH_MMU_CONFIG_CP_R1_CLNT_BEHAVIOR__SHIFT		10
-static inline uint32_t AXXX_MH_MMU_CONFIG_CP_R1_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
-{
-	return ((val) << AXXX_MH_MMU_CONFIG_CP_R1_CLNT_BEHAVIOR__SHIFT) & AXXX_MH_MMU_CONFIG_CP_R1_CLNT_BEHAVIOR__MASK;
-}
-#define AXXX_MH_MMU_CONFIG_CP_R2_CLNT_BEHAVIOR__MASK		0x00003000
-#define AXXX_MH_MMU_CONFIG_CP_R2_CLNT_BEHAVIOR__SHIFT		12
-static inline uint32_t AXXX_MH_MMU_CONFIG_CP_R2_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
-{
-	return ((val) << AXXX_MH_MMU_CONFIG_CP_R2_CLNT_BEHAVIOR__SHIFT) & AXXX_MH_MMU_CONFIG_CP_R2_CLNT_BEHAVIOR__MASK;
-}
-#define AXXX_MH_MMU_CONFIG_CP_R3_CLNT_BEHAVIOR__MASK		0x0000c000
-#define AXXX_MH_MMU_CONFIG_CP_R3_CLNT_BEHAVIOR__SHIFT		14
-static inline uint32_t AXXX_MH_MMU_CONFIG_CP_R3_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
-{
-	return ((val) << AXXX_MH_MMU_CONFIG_CP_R3_CLNT_BEHAVIOR__SHIFT) & AXXX_MH_MMU_CONFIG_CP_R3_CLNT_BEHAVIOR__MASK;
-}
-#define AXXX_MH_MMU_CONFIG_CP_R4_CLNT_BEHAVIOR__MASK		0x00030000
-#define AXXX_MH_MMU_CONFIG_CP_R4_CLNT_BEHAVIOR__SHIFT		16
-static inline uint32_t AXXX_MH_MMU_CONFIG_CP_R4_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
-{
-	return ((val) << AXXX_MH_MMU_CONFIG_CP_R4_CLNT_BEHAVIOR__SHIFT) & AXXX_MH_MMU_CONFIG_CP_R4_CLNT_BEHAVIOR__MASK;
-}
-#define AXXX_MH_MMU_CONFIG_VGT_R0_CLNT_BEHAVIOR__MASK		0x000c0000
-#define AXXX_MH_MMU_CONFIG_VGT_R0_CLNT_BEHAVIOR__SHIFT		18
-static inline uint32_t AXXX_MH_MMU_CONFIG_VGT_R0_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
-{
-	return ((val) << AXXX_MH_MMU_CONFIG_VGT_R0_CLNT_BEHAVIOR__SHIFT) & AXXX_MH_MMU_CONFIG_VGT_R0_CLNT_BEHAVIOR__MASK;
-}
-#define AXXX_MH_MMU_CONFIG_VGT_R1_CLNT_BEHAVIOR__MASK		0x00300000
-#define AXXX_MH_MMU_CONFIG_VGT_R1_CLNT_BEHAVIOR__SHIFT		20
-static inline uint32_t AXXX_MH_MMU_CONFIG_VGT_R1_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
-{
-	return ((val) << AXXX_MH_MMU_CONFIG_VGT_R1_CLNT_BEHAVIOR__SHIFT) & AXXX_MH_MMU_CONFIG_VGT_R1_CLNT_BEHAVIOR__MASK;
-}
-#define AXXX_MH_MMU_CONFIG_TC_R_CLNT_BEHAVIOR__MASK		0x00c00000
-#define AXXX_MH_MMU_CONFIG_TC_R_CLNT_BEHAVIOR__SHIFT		22
-static inline uint32_t AXXX_MH_MMU_CONFIG_TC_R_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
-{
-	return ((val) << AXXX_MH_MMU_CONFIG_TC_R_CLNT_BEHAVIOR__SHIFT) & AXXX_MH_MMU_CONFIG_TC_R_CLNT_BEHAVIOR__MASK;
-}
-#define AXXX_MH_MMU_CONFIG_PA_W_CLNT_BEHAVIOR__MASK		0x03000000
-#define AXXX_MH_MMU_CONFIG_PA_W_CLNT_BEHAVIOR__SHIFT		24
-static inline uint32_t AXXX_MH_MMU_CONFIG_PA_W_CLNT_BEHAVIOR(enum adreno_mmu_clnt_beh val)
-{
-	return ((val) << AXXX_MH_MMU_CONFIG_PA_W_CLNT_BEHAVIOR__SHIFT) & AXXX_MH_MMU_CONFIG_PA_W_CLNT_BEHAVIOR__MASK;
-}
-
-#define REG_AXXX_MH_MMU_VA_RANGE				0x00000041
-
-#define REG_AXXX_MH_MMU_PT_BASE					0x00000042
-
-#define REG_AXXX_MH_MMU_PAGE_FAULT				0x00000043
-
-#define REG_AXXX_MH_MMU_TRAN_ERROR				0x00000044
-
-#define REG_AXXX_MH_MMU_INVALIDATE				0x00000045
-
-#define REG_AXXX_MH_MMU_MPU_BASE				0x00000046
-
-#define REG_AXXX_MH_MMU_MPU_END					0x00000047
-
 #define REG_AXXX_CP_RB_BASE					0x000001c0
 
 #define REG_AXXX_CP_RB_CNTL					0x000001c1
@@ -275,6 +186,18 @@ static inline uint32_t AXXX_CP_QUEUE_THRESHOLDS_CSQ_ST_START(uint32_t val)
 }
 
 #define REG_AXXX_CP_MEQ_THRESHOLDS				0x000001d6
+#define AXXX_CP_MEQ_THRESHOLDS_MEQ_END__MASK			0x001f0000
+#define AXXX_CP_MEQ_THRESHOLDS_MEQ_END__SHIFT			16
+static inline uint32_t AXXX_CP_MEQ_THRESHOLDS_MEQ_END(uint32_t val)
+{
+	return ((val) << AXXX_CP_MEQ_THRESHOLDS_MEQ_END__SHIFT) & AXXX_CP_MEQ_THRESHOLDS_MEQ_END__MASK;
+}
+#define AXXX_CP_MEQ_THRESHOLDS_ROQ_END__MASK			0x1f000000
+#define AXXX_CP_MEQ_THRESHOLDS_ROQ_END__SHIFT			24
+static inline uint32_t AXXX_CP_MEQ_THRESHOLDS_ROQ_END(uint32_t val)
+{
+	return ((val) << AXXX_CP_MEQ_THRESHOLDS_ROQ_END__SHIFT) & AXXX_CP_MEQ_THRESHOLDS_ROQ_END__MASK;
+}
 
 #define REG_AXXX_CP_CSQ_AVAIL					0x000001d7
 #define AXXX_CP_CSQ_AVAIL_RING__MASK				0x0000007f
@@ -402,6 +325,36 @@ static inline uint32_t AXXX_CP_CSQ_IB2_STAT_WPTR(uint32_t val)
 	return ((val) << AXXX_CP_CSQ_IB2_STAT_WPTR__SHIFT) & AXXX_CP_CSQ_IB2_STAT_WPTR__MASK;
 }
 
+#define REG_AXXX_CP_NON_PREFETCH_CNTRS				0x00000440
+
+#define REG_AXXX_CP_STQ_ST_STAT					0x00000443
+
+#define REG_AXXX_CP_ST_BASE					0x0000044d
+
+#define REG_AXXX_CP_ST_BUFSZ					0x0000044e
+
+#define REG_AXXX_CP_MEQ_STAT					0x0000044f
+
+#define REG_AXXX_CP_MIU_TAG_STAT				0x00000452
+
+#define REG_AXXX_CP_BIN_MASK_LO					0x00000454
+
+#define REG_AXXX_CP_BIN_MASK_HI					0x00000455
+
+#define REG_AXXX_CP_BIN_SELECT_LO				0x00000456
+
+#define REG_AXXX_CP_BIN_SELECT_HI				0x00000457
+
+#define REG_AXXX_CP_IB1_BASE					0x00000458
+
+#define REG_AXXX_CP_IB1_BUFSZ					0x00000459
+
+#define REG_AXXX_CP_IB2_BASE					0x0000045a
+
+#define REG_AXXX_CP_IB2_BUFSZ					0x0000045b
+
+#define REG_AXXX_CP_STAT					0x0000047f
+
 #define REG_AXXX_CP_SCRATCH_REG0				0x00000578
 
 #define REG_AXXX_CP_SCRATCH_REG1				0x00000579
@@ -418,6 +371,26 @@ static inline uint32_t AXXX_CP_CSQ_IB2_STAT_WPTR(uint32_t val)
 
 #define REG_AXXX_CP_SCRATCH_REG7				0x0000057f
 
+#define REG_AXXX_CP_ME_VS_EVENT_SRC				0x00000600
+
+#define REG_AXXX_CP_ME_VS_EVENT_ADDR				0x00000601
+
+#define REG_AXXX_CP_ME_VS_EVENT_DATA				0x00000602
+
+#define REG_AXXX_CP_ME_VS_EVENT_ADDR_SWM			0x00000603
+
+#define REG_AXXX_CP_ME_VS_EVENT_DATA_SWM			0x00000604
+
+#define REG_AXXX_CP_ME_PS_EVENT_SRC				0x00000605
+
+#define REG_AXXX_CP_ME_PS_EVENT_ADDR				0x00000606
+
+#define REG_AXXX_CP_ME_PS_EVENT_DATA				0x00000607
+
+#define REG_AXXX_CP_ME_PS_EVENT_ADDR_SWM			0x00000608
+
+#define REG_AXXX_CP_ME_PS_EVENT_DATA_SWM			0x00000609
+
 #define REG_AXXX_CP_ME_CF_EVENT_SRC				0x0000060a
 
 #define REG_AXXX_CP_ME_CF_EVENT_ADDR				0x0000060b
@@ -428,5 +401,11 @@ static inline uint32_t AXXX_CP_CSQ_IB2_STAT_WPTR(uint32_t val)
 
 #define REG_AXXX_CP_ME_NRT_DATA					0x0000060e
 
+#define REG_AXXX_CP_ME_VS_FETCH_DONE_SRC			0x00000612
+
+#define REG_AXXX_CP_ME_VS_FETCH_DONE_ADDR			0x00000613
+
+#define REG_AXXX_CP_ME_VS_FETCH_DONE_DATA			0x00000614
+
 
 #endif /* ADRENO_COMMON_XML */
diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.c b/drivers/gpu/drm/msm/adreno/adreno_gpu.c
index a0b9d8a95b16..d321099abdd4 100644
--- a/drivers/gpu/drm/msm/adreno/adreno_gpu.c
+++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.c
@@ -17,6 +17,7 @@
 
 #include "adreno_gpu.h"
 #include "msm_gem.h"
+#include "msm_mmu.h"
 
 struct adreno_info {
 	struct adreno_rev rev;
@@ -44,7 +45,7 @@ static const struct adreno_info gpulist[] = {
 		.pfpfw = "a300_pfp.fw",
 		.gmem  = SZ_512K,
 	}, {
-		.rev   = ADRENO_REV(3, 3, 0, 0),
+		.rev   = ADRENO_REV(3, 3, 0, ANY_ID),
 		.revn  = 330,
 		.name  = "A330",
 		.pm4fw = "a330_pm4.fw",
@@ -53,6 +54,11 @@ static const struct adreno_info gpulist[] = {
 	},
 };
 
+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
 
@@ -65,7 +71,7 @@ int adreno_get_param(struct msm_gpu *gpu, uint32_t param, uint64_t *value)
 		*value = adreno_gpu->info->revn;
 		return 0;
 	case MSM_PARAM_GMEM_SIZE:
-		*value = adreno_gpu->info->gmem;
+		*value = adreno_gpu->gmem;
 		return 0;
 	default:
 		DBG("%s: invalid param: %u", gpu->name, param);
@@ -86,7 +92,7 @@ int adreno_hw_init(struct msm_gpu *gpu)
 	gpu_write(gpu, REG_AXXX_CP_RB_CNTL,
 			/* size is log2(quad-words): */
 			AXXX_CP_RB_CNTL_BUFSZ(ilog2(gpu->rb->size / 8)) |
-			AXXX_CP_RB_CNTL_BLKSZ(RB_BLKSIZE));
+			AXXX_CP_RB_CNTL_BLKSZ(ilog2(RB_BLKSIZE / 8)));
 
 	/* Setup ringbuffer address: */
 	gpu_write(gpu, REG_AXXX_CP_RB_BASE, gpu->rb_iova);
@@ -286,6 +292,7 @@ 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 msm_mmu *mmu;
 	int i, ret;
 
 	/* identify gpu: */
@@ -311,6 +318,7 @@ int adreno_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 			rev.core, rev.major, rev.minor, rev.patchid);
 
 	gpu->funcs = funcs;
+	gpu->gmem = gpu->info->gmem;
 	gpu->rev = rev;
 
 	ret = request_firmware(&gpu->pm4, gpu->info->pm4fw, drm->dev);
@@ -333,10 +341,13 @@ int adreno_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 	if (ret)
 		return ret;
 
-	ret = msm_iommu_attach(drm, gpu->base.iommu,
-			iommu_ports, ARRAY_SIZE(iommu_ports));
-	if (ret)
-		return ret;
+	mmu = gpu->base.mmu;
+	if (mmu) {
+		ret = mmu->funcs->attach(mmu, iommu_ports,
+				ARRAY_SIZE(iommu_ports));
+		if (ret)
+			return ret;
+	}
 
 	gpu->memptrs_bo = msm_gem_new(drm, sizeof(*gpu->memptrs),
 			MSM_BO_UNCACHED);
diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.h b/drivers/gpu/drm/msm/adreno/adreno_gpu.h
index f73abfba7c22..ca11ea4da165 100644
--- a/drivers/gpu/drm/msm/adreno/adreno_gpu.h
+++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.h
@@ -51,6 +51,7 @@ struct adreno_gpu {
 	struct msm_gpu base;
 	struct adreno_rev rev;
 	const struct adreno_info *info;
+	uint32_t gmem;  /* actual gmem size */
 	uint32_t revn;  /* numeric revision name */
 	const struct adreno_gpu_funcs *funcs;
 
@@ -70,6 +71,9 @@ struct adreno_gpu {
 struct adreno_platform_config {
 	struct adreno_rev rev;
 	uint32_t fast_rate, slow_rate, bus_freq;
+#ifdef CONFIG_MSM_BUS_SCALING
+	struct msm_bus_scale_pdata *bus_scale_table;
+#endif
 };
 
 #define ADRENO_IDLE_TIMEOUT (20 * 1000)
@@ -94,6 +98,11 @@ static inline bool adreno_is_a330(struct adreno_gpu *gpu)
 	return gpu->revn == 330;
 }
 
+static inline bool adreno_is_a330v2(struct adreno_gpu *gpu)
+{
+	return adreno_is_a330(gpu) && (gpu->rev.patchid > 0);
+}
+
 int adreno_get_param(struct msm_gpu *gpu, uint32_t param, uint64_t *value);
 int adreno_hw_init(struct msm_gpu *gpu);
 uint32_t adreno_last_fence(struct msm_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 259ad709b0cc..ae992c71703f 100644
--- a/drivers/gpu/drm/msm/adreno/adreno_pm4.xml.h
+++ b/drivers/gpu/drm/msm/adreno/adreno_pm4.xml.h
@@ -8,12 +8,13 @@ http://github.com/freedreno/envytools/
 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/adreno.xml              (    327 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/freedreno_copyright.xml (   1453 bytes, from 2013-03-31 16:51:27)
-- /home/robclark/src/freedreno/envytools/rnndb/a2xx/a2xx.xml           (  31003 bytes, from 2013-09-19 18:50:16)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno_common.xml       (   8983 bytes, from 2013-07-24 01:38:36)
-- /home/robclark/src/freedreno/envytools/rnndb/adreno_pm4.xml          (   9759 bytes, from 2013-09-10 00:52:33)
-- /home/robclark/src/freedreno/envytools/rnndb/a3xx/a3xx.xml           (  51983 bytes, from 2013-09-10 00:52:32)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno.xml               (    364 bytes, from 2013-11-30 14:47:15)
+- /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          (  32814 bytes, from 2013-11-30 15:07:33)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_common.xml (   8900 bytes, from 2013-10-22 23:57:49)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/adreno_pm4.xml    (  10574 bytes, from 2013-11-13 05:44:45)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a3xx.xml          (  53644 bytes, from 2013-11-30 15:07:33)
+- /home/robclark/src/freedreno/envytools/rnndb/adreno/a4xx.xml          (   8344 bytes, from 2013-11-30 14:49:47)
 
 Copyright (C) 2013 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
@@ -66,13 +67,15 @@ enum vgt_event_type {
 
 enum pc_di_primtype {
 	DI_PT_NONE = 0,
-	DI_PT_POINTLIST = 1,
+	DI_PT_POINTLIST_A2XX = 1,
 	DI_PT_LINELIST = 2,
 	DI_PT_LINESTRIP = 3,
 	DI_PT_TRILIST = 4,
 	DI_PT_TRIFAN = 5,
 	DI_PT_TRISTRIP = 6,
+	DI_PT_LINELOOP = 7,
 	DI_PT_RECTLIST = 8,
+	DI_PT_POINTLIST_A3XX = 9,
 	DI_PT_QUADLIST = 13,
 	DI_PT_QUADSTRIP = 14,
 	DI_PT_POLYGON = 15,
@@ -119,7 +122,7 @@ enum adreno_pm4_type3_packets {
 	CP_WAIT_FOR_IDLE = 38,
 	CP_WAIT_REG_MEM = 60,
 	CP_WAIT_REG_EQ = 82,
-	CP_WAT_REG_GTE = 83,
+	CP_WAIT_REG_GTE = 83,
 	CP_WAIT_UNTIL_READ = 92,
 	CP_WAIT_IB_PFD_COMPLETE = 93,
 	CP_REG_RMW = 33,
@@ -151,7 +154,6 @@ enum adreno_pm4_type3_packets {
 	CP_CONTEXT_UPDATE = 94,
 	CP_INTERRUPT = 64,
 	CP_IM_STORE = 44,
-	CP_SET_BIN_BASE_OFFSET = 75,
 	CP_SET_DRAW_INIT_FLAGS = 75,
 	CP_SET_PROTECTED_MODE = 95,
 	CP_LOAD_STATE = 48,
@@ -159,6 +161,16 @@ enum adreno_pm4_type3_packets {
 	CP_COND_INDIRECT_BUFFER_PFD = 50,
 	CP_INDIRECT_BUFFER_PFE = 63,
 	CP_SET_BIN = 76,
+	CP_TEST_TWO_MEMS = 113,
+	CP_WAIT_FOR_ME = 19,
+	IN_IB_PREFETCH_END = 23,
+	IN_SUBBLK_PREFETCH = 31,
+	IN_INSTR_PREFETCH = 32,
+	IN_INSTR_MATCH = 71,
+	IN_CONST_PREFETCH = 73,
+	IN_INCR_UPDT_STATE = 85,
+	IN_INCR_UPDT_CONST = 86,
+	IN_INCR_UPDT_INSTR = 87,
 };
 
 enum adreno_state_block {
diff --git a/drivers/gpu/drm/msm/dsi/dsi.xml.h b/drivers/gpu/drm/msm/dsi/dsi.xml.h
index 6d4c62bf70dc..87be647e3825 100644
--- a/drivers/gpu/drm/msm/dsi/dsi.xml.h
+++ b/drivers/gpu/drm/msm/dsi/dsi.xml.h
@@ -8,14 +8,16 @@ http://github.com/freedreno/envytools/
 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                 (    595 bytes, from 2013-07-05 19:21:12)
+- /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/mdp4/mdp4.xml           (  19332 bytes, from 2013-10-07 16:36:48)
+- /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 2013-12-03 20:59:13)
 - /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/hdmi/qfprom.xml         (    600 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  19288 bytes, from 2013-08-11 18:14:15)
+- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  20932 bytes, from 2013-12-01 15:13:04)
 
 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 d1df38bf5747..747a6ef4211f 100644
--- a/drivers/gpu/drm/msm/dsi/mmss_cc.xml.h
+++ b/drivers/gpu/drm/msm/dsi/mmss_cc.xml.h
@@ -8,14 +8,16 @@ http://github.com/freedreno/envytools/
 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                 (    595 bytes, from 2013-07-05 19:21:12)
+- /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/mdp4/mdp4.xml           (  19332 bytes, from 2013-10-07 16:36:48)
+- /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 2013-12-03 20:59:13)
 - /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/hdmi/qfprom.xml         (    600 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  19288 bytes, from 2013-08-11 18:14:15)
+- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  20932 bytes, from 2013-12-01 15:13:04)
 
 Copyright (C) 2013 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
diff --git a/drivers/gpu/drm/msm/dsi/sfpb.xml.h b/drivers/gpu/drm/msm/dsi/sfpb.xml.h
index 0030a111302d..48e03acf19bf 100644
--- a/drivers/gpu/drm/msm/dsi/sfpb.xml.h
+++ b/drivers/gpu/drm/msm/dsi/sfpb.xml.h
@@ -8,14 +8,16 @@ http://github.com/freedreno/envytools/
 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                 (    595 bytes, from 2013-07-05 19:21:12)
+- /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/mdp4/mdp4.xml           (  19332 bytes, from 2013-10-07 16:36:48)
+- /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 2013-12-03 20:59:13)
 - /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/hdmi/qfprom.xml         (    600 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  19288 bytes, from 2013-08-11 18:14:15)
+- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  20932 bytes, from 2013-12-01 15:13:04)
 
 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 50d11df35b21..6f1588aa9071 100644
--- a/drivers/gpu/drm/msm/hdmi/hdmi.c
+++ b/drivers/gpu/drm/msm/hdmi/hdmi.c
@@ -41,7 +41,7 @@ void hdmi_set_mode(struct hdmi *hdmi, bool power_on)
 			power_on ? "Enable" : "Disable", ctrl);
 }
 
-static irqreturn_t hdmi_irq(int irq, void *dev_id)
+irqreturn_t hdmi_irq(int irq, void *dev_id)
 {
 	struct hdmi *hdmi = dev_id;
 
@@ -71,13 +71,13 @@ void hdmi_destroy(struct kref *kref)
 }
 
 /* initialize connector */
-int hdmi_init(struct drm_device *dev, struct drm_encoder *encoder)
+struct hdmi *hdmi_init(struct drm_device *dev, struct drm_encoder *encoder)
 {
 	struct hdmi *hdmi = NULL;
 	struct msm_drm_private *priv = dev->dev_private;
 	struct platform_device *pdev = hdmi_pdev;
 	struct hdmi_platform_config *config;
-	int ret;
+	int i, ret;
 
 	if (!pdev) {
 		dev_err(dev->dev, "no hdmi device\n");
@@ -99,6 +99,7 @@ int hdmi_init(struct drm_device *dev, struct drm_encoder *encoder)
 
 	hdmi->dev = dev;
 	hdmi->pdev = pdev;
+	hdmi->config = config;
 	hdmi->encoder = encoder;
 
 	/* not sure about which phy maps to which msm.. probably I miss some */
@@ -114,44 +115,70 @@ int hdmi_init(struct drm_device *dev, struct drm_encoder *encoder)
 		goto fail;
 	}
 
-	hdmi->mmio = msm_ioremap(pdev, "hdmi_msm_hdmi_addr", "HDMI");
+	hdmi->mmio = msm_ioremap(pdev, config->mmio_name, "HDMI");
 	if (IS_ERR(hdmi->mmio)) {
 		ret = PTR_ERR(hdmi->mmio);
 		goto fail;
 	}
 
-	hdmi->mvs = devm_regulator_get(&pdev->dev, "8901_hdmi_mvs");
-	if (IS_ERR(hdmi->mvs))
-		hdmi->mvs = devm_regulator_get(&pdev->dev, "hdmi_mvs");
-	if (IS_ERR(hdmi->mvs)) {
-		ret = PTR_ERR(hdmi->mvs);
-		dev_err(dev->dev, "failed to get mvs regulator: %d\n", ret);
-		goto fail;
+	BUG_ON(config->hpd_reg_cnt > ARRAY_SIZE(hdmi->hpd_regs));
+	for (i = 0; i < config->hpd_reg_cnt; i++) {
+		struct regulator *reg;
+
+		reg = devm_regulator_get(&pdev->dev, config->hpd_reg_names[i]);
+		if (IS_ERR(reg)) {
+			ret = PTR_ERR(reg);
+			dev_err(dev->dev, "failed to get hpd regulator: %s (%d)\n",
+					config->hpd_reg_names[i], ret);
+			goto fail;
+		}
+
+		hdmi->hpd_regs[i] = reg;
 	}
 
-	hdmi->mpp0 = devm_regulator_get(&pdev->dev, "8901_mpp0");
-	if (IS_ERR(hdmi->mpp0))
-		hdmi->mpp0 = NULL;
+	BUG_ON(config->pwr_reg_cnt > ARRAY_SIZE(hdmi->pwr_regs));
+	for (i = 0; i < config->pwr_reg_cnt; i++) {
+		struct regulator *reg;
 
-	hdmi->clk = devm_clk_get(&pdev->dev, "core_clk");
-	if (IS_ERR(hdmi->clk)) {
-		ret = PTR_ERR(hdmi->clk);
-		dev_err(dev->dev, "failed to get 'clk': %d\n", ret);
-		goto fail;
+		reg = devm_regulator_get(&pdev->dev, config->pwr_reg_names[i]);
+		if (IS_ERR(reg)) {
+			ret = PTR_ERR(reg);
+			dev_err(dev->dev, "failed to get pwr regulator: %s (%d)\n",
+					config->pwr_reg_names[i], ret);
+			goto fail;
+		}
+
+		hdmi->pwr_regs[i] = reg;
 	}
 
-	hdmi->m_pclk = devm_clk_get(&pdev->dev, "master_iface_clk");
-	if (IS_ERR(hdmi->m_pclk)) {
-		ret = PTR_ERR(hdmi->m_pclk);
-		dev_err(dev->dev, "failed to get 'm_pclk': %d\n", ret);
-		goto fail;
+	BUG_ON(config->hpd_clk_cnt > ARRAY_SIZE(hdmi->hpd_clks));
+	for (i = 0; i < config->hpd_clk_cnt; i++) {
+		struct clk *clk;
+
+		clk = devm_clk_get(&pdev->dev, config->hpd_clk_names[i]);
+		if (IS_ERR(clk)) {
+			ret = PTR_ERR(clk);
+			dev_err(dev->dev, "failed to get hpd clk: %s (%d)\n",
+					config->hpd_clk_names[i], ret);
+			goto fail;
+		}
+
+		hdmi->hpd_clks[i] = clk;
 	}
 
-	hdmi->s_pclk = devm_clk_get(&pdev->dev, "slave_iface_clk");
-	if (IS_ERR(hdmi->s_pclk)) {
-		ret = PTR_ERR(hdmi->s_pclk);
-		dev_err(dev->dev, "failed to get 's_pclk': %d\n", ret);
-		goto fail;
+	BUG_ON(config->pwr_clk_cnt > ARRAY_SIZE(hdmi->pwr_clks));
+	for (i = 0; i < config->pwr_clk_cnt; i++) {
+		struct clk *clk;
+
+		clk = devm_clk_get(&pdev->dev, config->pwr_clk_names[i]);
+		if (IS_ERR(clk)) {
+			ret = PTR_ERR(clk);
+			dev_err(dev->dev, "failed to get pwr clk: %s (%d)\n",
+					config->pwr_clk_names[i], ret);
+			goto fail;
+		}
+
+		hdmi->pwr_clks[i] = clk;
 	}
 
 	hdmi->i2c = hdmi_i2c_init(hdmi);
@@ -178,20 +205,22 @@ int hdmi_init(struct drm_device *dev, struct drm_encoder *encoder)
 		goto fail;
 	}
 
-	hdmi->irq = platform_get_irq(pdev, 0);
-	if (hdmi->irq < 0) {
-		ret = hdmi->irq;
-		dev_err(dev->dev, "failed to get irq: %d\n", ret);
-		goto fail;
-	}
+	if (!config->shared_irq) {
+		hdmi->irq = platform_get_irq(pdev, 0);
+		if (hdmi->irq < 0) {
+			ret = hdmi->irq;
+			dev_err(dev->dev, "failed to get irq: %d\n", ret);
+			goto fail;
+		}
 
-	ret = devm_request_threaded_irq(&pdev->dev, hdmi->irq,
-			NULL, hdmi_irq, IRQF_TRIGGER_HIGH | IRQF_ONESHOT,
-			"hdmi_isr", hdmi);
-	if (ret < 0) {
-		dev_err(dev->dev, "failed to request IRQ%u: %d\n",
-				hdmi->irq, ret);
-		goto fail;
+		ret = devm_request_threaded_irq(&pdev->dev, hdmi->irq,
+				NULL, hdmi_irq, IRQF_TRIGGER_HIGH | IRQF_ONESHOT,
+				"hdmi_isr", hdmi);
+		if (ret < 0) {
+			dev_err(dev->dev, "failed to request IRQ%u: %d\n",
+					hdmi->irq, ret);
+			goto fail;
+		}
 	}
 
 	encoder->bridge = hdmi->bridge;
@@ -199,7 +228,7 @@ int hdmi_init(struct drm_device *dev, struct drm_encoder *encoder)
 	priv->bridges[priv->num_bridges++]       = hdmi->bridge;
 	priv->connectors[priv->num_connectors++] = hdmi->connector;
 
-	return 0;
+	return hdmi;
 
 fail:
 	if (hdmi) {
@@ -211,37 +240,100 @@ fail:
 		hdmi_destroy(&hdmi->refcount);
 	}
 
-	return ret;
+	return ERR_PTR(ret);
 }
 
 /*
  * The hdmi device:
  */
 
+#include <linux/of_gpio.h>
+
 static int hdmi_dev_probe(struct platform_device *pdev)
 {
 	static struct hdmi_platform_config config = {};
 #ifdef CONFIG_OF
-	/* TODO */
+	struct device_node *of_node = pdev->dev.of_node;
+
+	int get_gpio(const char *name)
+	{
+		int gpio = of_get_named_gpio(of_node, name, 0);
+		if (gpio < 0) {
+			dev_err(&pdev->dev, "failed to get gpio: %s (%d)\n",
+					name, gpio);
+			gpio = -1;
+		}
+		return gpio;
+	}
+
+	/* TODO actually use DT.. */
+	static const char *hpd_reg_names[] = {"hpd-gdsc", "hpd-5v"};
+	static const char *pwr_reg_names[] = {"core-vdda", "core-vcc"};
+	static const char *hpd_clk_names[] = {"iface_clk", "core_clk", "mdp_core_clk"};
+	static const char *pwr_clk_names[] = {"extp_clk", "alt_iface_clk"};
+
+	config.phy_init      = hdmi_phy_8x74_init;
+	config.mmio_name     = "core_physical";
+	config.hpd_reg_names = hpd_reg_names;
+	config.hpd_reg_cnt   = ARRAY_SIZE(hpd_reg_names);
+	config.pwr_reg_names = pwr_reg_names;
+	config.pwr_reg_cnt   = ARRAY_SIZE(pwr_reg_names);
+	config.hpd_clk_names = hpd_clk_names;
+	config.hpd_clk_cnt   = ARRAY_SIZE(hpd_clk_names);
+	config.pwr_clk_names = pwr_clk_names;
+	config.pwr_clk_cnt   = ARRAY_SIZE(pwr_clk_names);
+	config.ddc_clk_gpio  = get_gpio("qcom,hdmi-tx-ddc-clk");
+	config.ddc_data_gpio = get_gpio("qcom,hdmi-tx-ddc-data");
+	config.hpd_gpio      = get_gpio("qcom,hdmi-tx-hpd");
+	config.mux_en_gpio   = get_gpio("qcom,hdmi-tx-mux-en");
+	config.mux_sel_gpio  = get_gpio("qcom,hdmi-tx-mux-sel");
+	config.shared_irq    = true;
+
 #else
+	static const char *hpd_clk_names[] = {
+			"core_clk", "master_iface_clk", "slave_iface_clk",
+	};
 	if (cpu_is_apq8064()) {
+		static const char *hpd_reg_names[] = {"8921_hdmi_mvs"};
 		config.phy_init      = hdmi_phy_8960_init;
+		config.mmio_name     = "hdmi_msm_hdmi_addr";
+		config.hpd_reg_names = hpd_reg_names;
+		config.hpd_reg_cnt   = ARRAY_SIZE(hpd_reg_names);
+		config.hpd_clk_names = hpd_clk_names;
+		config.hpd_clk_cnt   = ARRAY_SIZE(hpd_clk_names);
 		config.ddc_clk_gpio  = 70;
 		config.ddc_data_gpio = 71;
 		config.hpd_gpio      = 72;
-		config.pmic_gpio     = 13 + NR_GPIO_IRQS;
-	} else if (cpu_is_msm8960()) {
+		config.mux_en_gpio   = -1;
+		config.mux_sel_gpio  = 13 + NR_GPIO_IRQS;
+	} else if (cpu_is_msm8960() || cpu_is_msm8960ab()) {
+		static const char *hpd_reg_names[] = {"8921_hdmi_mvs"};
 		config.phy_init      = hdmi_phy_8960_init;
+		config.mmio_name     = "hdmi_msm_hdmi_addr";
+		config.hpd_reg_names = hpd_reg_names;
+		config.hpd_reg_cnt   = ARRAY_SIZE(hpd_reg_names);
+		config.hpd_clk_names = hpd_clk_names;
+		config.hpd_clk_cnt   = ARRAY_SIZE(hpd_clk_names);
 		config.ddc_clk_gpio  = 100;
 		config.ddc_data_gpio = 101;
 		config.hpd_gpio      = 102;
-		config.pmic_gpio     = -1;
+		config.mux_en_gpio   = -1;
+		config.mux_sel_gpio  = -1;
 	} else if (cpu_is_msm8x60()) {
+		static const char *hpd_reg_names[] = {
+				"8901_hdmi_mvs", "8901_mpp0"
+		};
 		config.phy_init      = hdmi_phy_8x60_init;
+		config.mmio_name     = "hdmi_msm_hdmi_addr";
+		config.hpd_reg_names = hpd_reg_names;
+		config.hpd_reg_cnt   = ARRAY_SIZE(hpd_reg_names);
+		config.hpd_clk_names = hpd_clk_names;
+		config.hpd_clk_cnt   = ARRAY_SIZE(hpd_clk_names);
 		config.ddc_clk_gpio  = 170;
 		config.ddc_data_gpio = 171;
 		config.hpd_gpio      = 172;
-		config.pmic_gpio     = -1;
+		config.mux_en_gpio   = -1;
+		config.mux_sel_gpio  = -1;
 	}
 #endif
 	pdev->dev.platform_data = &config;
@@ -255,10 +347,19 @@ static int hdmi_dev_remove(struct platform_device *pdev)
 	return 0;
 }
 
+static const struct of_device_id dt_match[] = {
+	{ .compatible = "qcom,hdmi-tx" },
+	{}
+};
+MODULE_DEVICE_TABLE(of, dt_match);
+
 static struct platform_driver hdmi_driver = {
 	.probe = hdmi_dev_probe,
 	.remove = hdmi_dev_remove,
-	.driver.name = "hdmi_msm",
+	.driver = {
+		.name = "hdmi_msm",
+		.of_match_table = dt_match,
+	},
 };
 
 void __init hdmi_register(void)
diff --git a/drivers/gpu/drm/msm/hdmi/hdmi.h b/drivers/gpu/drm/msm/hdmi/hdmi.h
index 2c2ec566394c..41b29add70b1 100644
--- a/drivers/gpu/drm/msm/hdmi/hdmi.h
+++ b/drivers/gpu/drm/msm/hdmi/hdmi.h
@@ -28,6 +28,7 @@
 
 
 struct hdmi_phy;
+struct hdmi_platform_config;
 
 struct hdmi {
 	struct kref refcount;
@@ -35,14 +36,14 @@ struct hdmi {
 	struct drm_device *dev;
 	struct platform_device *pdev;
 
-	void __iomem *mmio;
+	const struct hdmi_platform_config *config;
 
-	struct regulator *mvs;        /* HDMI_5V */
-	struct regulator *mpp0;       /* External 5V */
+	void __iomem *mmio;
 
-	struct clk *clk;
-	struct clk *m_pclk;
-	struct clk *s_pclk;
+	struct regulator *hpd_regs[2];
+	struct regulator *pwr_regs[2];
+	struct clk *hpd_clks[3];
+	struct clk *pwr_clks[2];
 
 	struct hdmi_phy *phy;
 	struct i2c_adapter *i2c;
@@ -60,7 +61,29 @@ struct hdmi {
 /* platform config data (ie. from DT, or pdata) */
 struct hdmi_platform_config {
 	struct hdmi_phy *(*phy_init)(struct hdmi *hdmi);
-	int ddc_clk_gpio, ddc_data_gpio, hpd_gpio, pmic_gpio;
+	const char *mmio_name;
+
+	/* regulators that need to be on for hpd: */
+	const char **hpd_reg_names;
+	int hpd_reg_cnt;
+
+	/* regulators that need to be on for screen pwr: */
+	const char **pwr_reg_names;
+	int pwr_reg_cnt;
+
+	/* clks that need to be on for hpd: */
+	const char **hpd_clk_names;
+	int hpd_clk_cnt;
+
+	/* clks that need to be on for screen pwr (ie pixel clk): */
+	const char **pwr_clk_names;
+	int pwr_clk_cnt;
+
+	/* gpio's: */
+	int ddc_clk_gpio, ddc_data_gpio, hpd_gpio, mux_en_gpio, mux_sel_gpio;
+
+	/* older devices had their own irq, mdp5+ it is shared w/ mdp: */
+	bool shared_irq;
 };
 
 void hdmi_set_mode(struct hdmi *hdmi, bool power_on);
@@ -106,6 +129,7 @@ struct hdmi_phy {
 
 struct hdmi_phy *hdmi_phy_8960_init(struct hdmi *hdmi);
 struct hdmi_phy *hdmi_phy_8x60_init(struct hdmi *hdmi);
+struct hdmi_phy *hdmi_phy_8x74_init(struct hdmi *hdmi);
 
 /*
  * hdmi bridge:
diff --git a/drivers/gpu/drm/msm/hdmi/hdmi.xml.h b/drivers/gpu/drm/msm/hdmi/hdmi.xml.h
index 4e939f82918c..e2636582cfd7 100644
--- a/drivers/gpu/drm/msm/hdmi/hdmi.xml.h
+++ b/drivers/gpu/drm/msm/hdmi/hdmi.xml.h
@@ -8,14 +8,16 @@ http://github.com/freedreno/envytools/
 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                 (    595 bytes, from 2013-07-05 19:21:12)
+- /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/mdp4/mdp4.xml           (  19332 bytes, from 2013-10-07 16:36:48)
+- /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 2013-12-03 20:59:13)
 - /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/hdmi/qfprom.xml         (    600 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  19288 bytes, from 2013-08-11 18:14:15)
+- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  20932 bytes, from 2013-12-01 15:13:04)
 
 Copyright (C) 2013 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
@@ -212,6 +214,20 @@ static inline uint32_t HDMI_HDCP_LINK0_STATUS_KEY_STATE(enum hdmi_hdcp_key_state
 #define REG_HDMI_HDCP_RESET					0x00000130
 #define HDMI_HDCP_RESET_LINK0_DEAUTHENTICATE			0x00000001
 
+#define REG_HDMI_VENSPEC_INFO0					0x0000016c
+
+#define REG_HDMI_VENSPEC_INFO1					0x00000170
+
+#define REG_HDMI_VENSPEC_INFO2					0x00000174
+
+#define REG_HDMI_VENSPEC_INFO3					0x00000178
+
+#define REG_HDMI_VENSPEC_INFO4					0x0000017c
+
+#define REG_HDMI_VENSPEC_INFO5					0x00000180
+
+#define REG_HDMI_VENSPEC_INFO6					0x00000184
+
 #define REG_HDMI_AUDIO_CFG					0x000001d0
 #define HDMI_AUDIO_CFG_ENGINE_ENABLE				0x00000001
 #define HDMI_AUDIO_CFG_FIFO_WATERMARK__MASK			0x000000f0
@@ -235,6 +251,9 @@ static inline uint32_t HDMI_DDC_CTRL_TRANSACTION_CNT(uint32_t val)
 	return ((val) << HDMI_DDC_CTRL_TRANSACTION_CNT__SHIFT) & HDMI_DDC_CTRL_TRANSACTION_CNT__MASK;
 }
 
+#define REG_HDMI_DDC_ARBITRATION				0x00000210
+#define HDMI_DDC_ARBITRATION_HW_ARBITRATION			0x00000010
+
 #define REG_HDMI_DDC_INT_CTRL					0x00000214
 #define HDMI_DDC_INT_CTRL_SW_DONE_INT				0x00000001
 #define HDMI_DDC_INT_CTRL_SW_DONE_ACK				0x00000002
@@ -340,6 +359,20 @@ static inline uint32_t HDMI_DDC_REF_REFTIMER(uint32_t val)
 	return ((val) << HDMI_DDC_REF_REFTIMER__SHIFT) & HDMI_DDC_REF_REFTIMER__MASK;
 }
 
+#define REG_HDMI_CEC_STATUS					0x00000298
+
+#define REG_HDMI_CEC_INT					0x0000029c
+
+#define REG_HDMI_CEC_ADDR					0x000002a0
+
+#define REG_HDMI_CEC_TIME					0x000002a4
+
+#define REG_HDMI_CEC_REFTIMER					0x000002a8
+
+#define REG_HDMI_CEC_RD_DATA					0x000002ac
+
+#define REG_HDMI_CEC_RD_FILTER					0x000002b0
+
 #define REG_HDMI_ACTIVE_HSYNC					0x000002b4
 #define HDMI_ACTIVE_HSYNC_START__MASK				0x00000fff
 #define HDMI_ACTIVE_HSYNC_START__SHIFT				0
@@ -410,17 +443,33 @@ static inline uint32_t HDMI_VSYNC_TOTAL_F2_V_TOTAL(uint32_t val)
 #define HDMI_FRAME_CTRL_HSYNC_LOW				0x20000000
 #define HDMI_FRAME_CTRL_INTERLACED_EN				0x80000000
 
+#define REG_HDMI_AUD_INT					0x000002cc
+#define HDMI_AUD_INT_AUD_FIFO_URUN_INT				0x00000001
+#define HDMI_AUD_INT_AUD_FIFO_URAN_MASK				0x00000002
+#define HDMI_AUD_INT_AUD_SAM_DROP_INT				0x00000004
+#define HDMI_AUD_INT_AUD_SAM_DROP_MASK				0x00000008
+
 #define REG_HDMI_PHY_CTRL					0x000002d4
 #define HDMI_PHY_CTRL_SW_RESET_PLL				0x00000001
 #define HDMI_PHY_CTRL_SW_RESET_PLL_LOW				0x00000002
 #define HDMI_PHY_CTRL_SW_RESET					0x00000004
 #define HDMI_PHY_CTRL_SW_RESET_LOW				0x00000008
 
-#define REG_HDMI_AUD_INT					0x000002cc
-#define HDMI_AUD_INT_AUD_FIFO_URUN_INT				0x00000001
-#define HDMI_AUD_INT_AUD_FIFO_URAN_MASK				0x00000002
-#define HDMI_AUD_INT_AUD_SAM_DROP_INT				0x00000004
-#define HDMI_AUD_INT_AUD_SAM_DROP_MASK				0x00000008
+#define REG_HDMI_CEC_WR_RANGE					0x000002dc
+
+#define REG_HDMI_CEC_RD_RANGE					0x000002e0
+
+#define REG_HDMI_VERSION					0x000002e4
+
+#define REG_HDMI_CEC_COMPL_CTL					0x00000360
+
+#define REG_HDMI_CEC_RD_START_RANGE				0x00000364
+
+#define REG_HDMI_CEC_RD_TOTAL_RANGE				0x00000368
+
+#define REG_HDMI_CEC_RD_ERR_RESP_LO				0x0000036c
+
+#define REG_HDMI_CEC_WR_CHECK_CONFIG				0x00000370
 
 #define REG_HDMI_8x60_PHY_REG0					0x00000300
 #define HDMI_8x60_PHY_REG0_DESER_DEL_CTRL__MASK			0x0000001c
@@ -504,5 +553,23 @@ static inline uint32_t HDMI_8x60_PHY_REG1_OUTVOL_SWING_CTRL(uint32_t val)
 
 #define REG_HDMI_8960_PHY_REG12					0x00000430
 
+#define REG_HDMI_8x74_ANA_CFG0					0x00000000
+
+#define REG_HDMI_8x74_ANA_CFG1					0x00000004
+
+#define REG_HDMI_8x74_PD_CTRL0					0x00000010
+
+#define REG_HDMI_8x74_PD_CTRL1					0x00000014
+
+#define REG_HDMI_8x74_BIST_CFG0					0x00000034
+
+#define REG_HDMI_8x74_BIST_PATN0				0x0000003c
+
+#define REG_HDMI_8x74_BIST_PATN1				0x00000040
+
+#define REG_HDMI_8x74_BIST_PATN2				0x00000044
+
+#define REG_HDMI_8x74_BIST_PATN3				0x00000048
+
 
 #endif /* HDMI_XML */
diff --git a/drivers/gpu/drm/msm/hdmi/hdmi_bridge.c b/drivers/gpu/drm/msm/hdmi/hdmi_bridge.c
index 5a8ee3473cf5..7d10e55403c6 100644
--- a/drivers/gpu/drm/msm/hdmi/hdmi_bridge.c
+++ b/drivers/gpu/drm/msm/hdmi/hdmi_bridge.c
@@ -21,6 +21,7 @@ struct hdmi_bridge {
 	struct drm_bridge base;
 
 	struct hdmi *hdmi;
+	bool power_on;
 
 	unsigned long int pixclock;
 };
@@ -34,6 +35,65 @@ static void hdmi_bridge_destroy(struct drm_bridge *bridge)
 	kfree(hdmi_bridge);
 }
 
+static void power_on(struct drm_bridge *bridge)
+{
+	struct drm_device *dev = bridge->dev;
+	struct hdmi_bridge *hdmi_bridge = to_hdmi_bridge(bridge);
+	struct hdmi *hdmi = hdmi_bridge->hdmi;
+	const struct hdmi_platform_config *config = hdmi->config;
+	int i, ret;
+
+	for (i = 0; i < config->pwr_reg_cnt; i++) {
+		ret = regulator_enable(hdmi->pwr_regs[i]);
+		if (ret) {
+			dev_err(dev->dev, "failed to enable pwr regulator: %s (%d)\n",
+					config->pwr_reg_names[i], ret);
+		}
+	}
+
+	if (config->pwr_clk_cnt > 0) {
+		DBG("pixclock: %lu", hdmi_bridge->pixclock);
+		ret = clk_set_rate(hdmi->pwr_clks[0], hdmi_bridge->pixclock);
+		if (ret) {
+			dev_err(dev->dev, "failed to set pixel clk: %s (%d)\n",
+					config->pwr_clk_names[0], ret);
+		}
+	}
+
+	for (i = 0; i < config->pwr_clk_cnt; i++) {
+		ret = clk_prepare_enable(hdmi->pwr_clks[i]);
+		if (ret) {
+			dev_err(dev->dev, "failed to enable pwr clk: %s (%d)\n",
+					config->pwr_clk_names[i], ret);
+		}
+	}
+}
+
+static void power_off(struct drm_bridge *bridge)
+{
+	struct drm_device *dev = bridge->dev;
+	struct hdmi_bridge *hdmi_bridge = to_hdmi_bridge(bridge);
+	struct hdmi *hdmi = hdmi_bridge->hdmi;
+	const struct hdmi_platform_config *config = hdmi->config;
+	int i, ret;
+
+	/* TODO do we need to wait for final vblank somewhere before
+	 * cutting the clocks?
+	 */
+	mdelay(16 + 4);
+
+	for (i = 0; i < config->pwr_clk_cnt; i++)
+		clk_disable_unprepare(hdmi->pwr_clks[i]);
+
+	for (i = 0; i < config->pwr_reg_cnt; i++) {
+		ret = regulator_disable(hdmi->pwr_regs[i]);
+		if (ret) {
+			dev_err(dev->dev, "failed to disable pwr regulator: %s (%d)\n",
+					config->pwr_reg_names[i], ret);
+		}
+	}
+}
+
 static void hdmi_bridge_pre_enable(struct drm_bridge *bridge)
 {
 	struct hdmi_bridge *hdmi_bridge = to_hdmi_bridge(bridge);
@@ -41,6 +101,12 @@ static void hdmi_bridge_pre_enable(struct drm_bridge *bridge)
 	struct hdmi_phy *phy = hdmi->phy;
 
 	DBG("power up");
+
+	if (!hdmi_bridge->power_on) {
+		power_on(bridge);
+		hdmi_bridge->power_on = true;
+	}
+
 	phy->funcs->powerup(phy, hdmi_bridge->pixclock);
 	hdmi_set_mode(hdmi, true);
 }
@@ -62,6 +128,11 @@ static void hdmi_bridge_post_disable(struct drm_bridge *bridge)
 	DBG("power down");
 	hdmi_set_mode(hdmi, false);
 	phy->funcs->powerdown(phy);
+
+	if (hdmi_bridge->power_on) {
+		power_off(bridge);
+		hdmi_bridge->power_on = false;
+	}
 }
 
 static void hdmi_bridge_mode_set(struct drm_bridge *bridge,
diff --git a/drivers/gpu/drm/msm/hdmi/hdmi_connector.c b/drivers/gpu/drm/msm/hdmi/hdmi_connector.c
index 823eee521a31..7dedfdd12075 100644
--- a/drivers/gpu/drm/msm/hdmi/hdmi_connector.c
+++ b/drivers/gpu/drm/msm/hdmi/hdmi_connector.c
@@ -17,19 +17,20 @@
 
 #include <linux/gpio.h>
 
+#include "msm_kms.h"
 #include "hdmi.h"
 
 struct hdmi_connector {
 	struct drm_connector base;
 	struct hdmi *hdmi;
+	struct work_struct hpd_work;
 };
 #define to_hdmi_connector(x) container_of(x, struct hdmi_connector, base)
 
 static int gpio_config(struct hdmi *hdmi, bool on)
 {
 	struct drm_device *dev = hdmi->dev;
-	struct hdmi_platform_config *config =
-			hdmi->pdev->dev.platform_data;
+	const struct hdmi_platform_config *config = hdmi->config;
 	int ret;
 
 	if (on) {
@@ -39,26 +40,43 @@ static int gpio_config(struct hdmi *hdmi, bool on)
 				"HDMI_DDC_CLK", config->ddc_clk_gpio, ret);
 			goto error1;
 		}
+		gpio_set_value_cansleep(config->ddc_clk_gpio, 1);
+
 		ret = gpio_request(config->ddc_data_gpio, "HDMI_DDC_DATA");
 		if (ret) {
 			dev_err(dev->dev, "'%s'(%d) gpio_request failed: %d\n",
 				"HDMI_DDC_DATA", config->ddc_data_gpio, ret);
 			goto error2;
 		}
+		gpio_set_value_cansleep(config->ddc_data_gpio, 1);
+
 		ret = gpio_request(config->hpd_gpio, "HDMI_HPD");
 		if (ret) {
 			dev_err(dev->dev, "'%s'(%d) gpio_request failed: %d\n",
 				"HDMI_HPD", config->hpd_gpio, ret);
 			goto error3;
 		}
-		if (config->pmic_gpio != -1) {
-			ret = gpio_request(config->pmic_gpio, "PMIC_HDMI_MUX_SEL");
+		gpio_direction_input(config->hpd_gpio);
+		gpio_set_value_cansleep(config->hpd_gpio, 1);
+
+		if (config->mux_en_gpio != -1) {
+			ret = gpio_request(config->mux_en_gpio, "HDMI_MUX_EN");
 			if (ret) {
 				dev_err(dev->dev, "'%s'(%d) gpio_request failed: %d\n",
-					"PMIC_HDMI_MUX_SEL", config->pmic_gpio, ret);
+					"HDMI_MUX_SEL", config->mux_en_gpio, ret);
 				goto error4;
 			}
-			gpio_set_value_cansleep(config->pmic_gpio, 0);
+			gpio_set_value_cansleep(config->mux_en_gpio, 1);
+		}
+
+		if (config->mux_sel_gpio != -1) {
+			ret = gpio_request(config->mux_sel_gpio, "HDMI_MUX_SEL");
+			if (ret) {
+				dev_err(dev->dev, "'%s'(%d) gpio_request failed: %d\n",
+					"HDMI_MUX_SEL", config->mux_sel_gpio, ret);
+				goto error5;
+			}
+			gpio_set_value_cansleep(config->mux_sel_gpio, 0);
 		}
 		DBG("gpio on");
 	} else {
@@ -66,15 +84,23 @@ static int gpio_config(struct hdmi *hdmi, bool on)
 		gpio_free(config->ddc_data_gpio);
 		gpio_free(config->hpd_gpio);
 
-		if (config->pmic_gpio != -1) {
-			gpio_set_value_cansleep(config->pmic_gpio, 1);
-			gpio_free(config->pmic_gpio);
+		if (config->mux_en_gpio != -1) {
+			gpio_set_value_cansleep(config->mux_en_gpio, 0);
+			gpio_free(config->mux_en_gpio);
+		}
+
+		if (config->mux_sel_gpio != -1) {
+			gpio_set_value_cansleep(config->mux_sel_gpio, 1);
+			gpio_free(config->mux_sel_gpio);
 		}
 		DBG("gpio off");
 	}
 
 	return 0;
 
+error5:
+	if (config->mux_en_gpio != -1)
+		gpio_free(config->mux_en_gpio);
 error4:
 	gpio_free(config->hpd_gpio);
 error3:
@@ -88,10 +114,11 @@ error1:
 static int hpd_enable(struct hdmi_connector *hdmi_connector)
 {
 	struct hdmi *hdmi = hdmi_connector->hdmi;
+	const struct hdmi_platform_config *config = hdmi->config;
 	struct drm_device *dev = hdmi_connector->base.dev;
 	struct hdmi_phy *phy = hdmi->phy;
 	uint32_t hpd_ctrl;
-	int ret;
+	int i, ret;
 
 	ret = gpio_config(hdmi, true);
 	if (ret) {
@@ -99,31 +126,22 @@ static int hpd_enable(struct hdmi_connector *hdmi_connector)
 		goto fail;
 	}
 
-	ret = clk_prepare_enable(hdmi->clk);
-	if (ret) {
-		dev_err(dev->dev, "failed to enable 'clk': %d\n", ret);
-		goto fail;
-	}
-
-	ret = clk_prepare_enable(hdmi->m_pclk);
-	if (ret) {
-		dev_err(dev->dev, "failed to enable 'm_pclk': %d\n", ret);
-		goto fail;
-	}
-
-	ret = clk_prepare_enable(hdmi->s_pclk);
-	if (ret) {
-		dev_err(dev->dev, "failed to enable 's_pclk': %d\n", ret);
-		goto fail;
+	for (i = 0; i < config->hpd_clk_cnt; i++) {
+		ret = clk_prepare_enable(hdmi->hpd_clks[i]);
+		if (ret) {
+			dev_err(dev->dev, "failed to enable hpd clk: %s (%d)\n",
+					config->hpd_clk_names[i], ret);
+			goto fail;
+		}
 	}
 
-	if (hdmi->mpp0)
-		ret = regulator_enable(hdmi->mpp0);
-	if (!ret)
-		ret = regulator_enable(hdmi->mvs);
-	if (ret) {
-		dev_err(dev->dev, "failed to enable regulators: %d\n", ret);
-		goto fail;
+	for (i = 0; i < config->hpd_reg_cnt; i++) {
+		ret = regulator_enable(hdmi->hpd_regs[i]);
+		if (ret) {
+			dev_err(dev->dev, "failed to enable hpd regulator: %s (%d)\n",
+					config->hpd_reg_names[i], ret);
+			goto fail;
+		}
 	}
 
 	hdmi_set_mode(hdmi, false);
@@ -156,26 +174,26 @@ fail:
 static int hdp_disable(struct hdmi_connector *hdmi_connector)
 {
 	struct hdmi *hdmi = hdmi_connector->hdmi;
+	const struct hdmi_platform_config *config = hdmi->config;
 	struct drm_device *dev = hdmi_connector->base.dev;
-	int ret = 0;
+	int i, ret = 0;
 
 	/* Disable HPD interrupt */
 	hdmi_write(hdmi, REG_HDMI_HPD_INT_CTRL, 0);
 
 	hdmi_set_mode(hdmi, false);
 
-	if (hdmi->mpp0)
-		ret = regulator_disable(hdmi->mpp0);
-	if (!ret)
-		ret = regulator_disable(hdmi->mvs);
-	if (ret) {
-		dev_err(dev->dev, "failed to enable regulators: %d\n", ret);
-		goto fail;
+	for (i = 0; i < config->hpd_reg_cnt; i++) {
+		ret = regulator_disable(hdmi->hpd_regs[i]);
+		if (ret) {
+			dev_err(dev->dev, "failed to disable hpd regulator: %s (%d)\n",
+					config->hpd_reg_names[i], ret);
+			goto fail;
+		}
 	}
 
-	clk_disable_unprepare(hdmi->clk);
-	clk_disable_unprepare(hdmi->m_pclk);
-	clk_disable_unprepare(hdmi->s_pclk);
+	for (i = 0; i < config->hpd_clk_cnt; i++)
+		clk_disable_unprepare(hdmi->hpd_clks[i]);
 
 	ret = gpio_config(hdmi, false);
 	if (ret) {
@@ -189,9 +207,19 @@ fail:
 	return ret;
 }
 
+static void
+hotplug_work(struct work_struct *work)
+{
+	struct hdmi_connector *hdmi_connector =
+		container_of(work, struct hdmi_connector, hpd_work);
+	struct drm_connector *connector = &hdmi_connector->base;
+	drm_helper_hpd_irq_event(connector->dev);
+}
+
 void hdmi_connector_irq(struct drm_connector *connector)
 {
 	struct hdmi_connector *hdmi_connector = to_hdmi_connector(connector);
+	struct msm_drm_private *priv = connector->dev->dev_private;
 	struct hdmi *hdmi = hdmi_connector->hdmi;
 	uint32_t hpd_int_status, hpd_int_ctrl;
 
@@ -209,13 +237,13 @@ void hdmi_connector_irq(struct drm_connector *connector)
 		hdmi_write(hdmi, REG_HDMI_HPD_INT_CTRL,
 				hpd_int_ctrl | HDMI_HPD_INT_CTRL_INT_ACK);
 
-		drm_helper_hpd_irq_event(connector->dev);
-
 		/* detect disconnect if we are connected or visa versa: */
 		hpd_int_ctrl = HDMI_HPD_INT_CTRL_INT_EN;
 		if (!detected)
 			hpd_int_ctrl |= HDMI_HPD_INT_CTRL_INT_CONNECT;
 		hdmi_write(hdmi, REG_HDMI_HPD_INT_CTRL, hpd_int_ctrl);
+
+		queue_work(priv->wq, &hdmi_connector->hpd_work);
 	}
 }
 
@@ -224,6 +252,7 @@ static enum drm_connector_status hdmi_connector_detect(
 {
 	struct hdmi_connector *hdmi_connector = to_hdmi_connector(connector);
 	struct hdmi *hdmi = hdmi_connector->hdmi;
+	const struct hdmi_platform_config *config = hdmi->config;
 	uint32_t hpd_int_status;
 	int retry = 20;
 
@@ -233,6 +262,14 @@ static enum drm_connector_status hdmi_connector_detect(
 	 * let that trick us into thinking the monitor is gone:
 	 */
 	while (retry-- && !(hpd_int_status & HDMI_HPD_INT_STATUS_CABLE_DETECTED)) {
+		/* hdmi debounce logic seems to get stuck sometimes,
+		 * read directly the gpio to get a second opinion:
+		 */
+		if (gpio_get_value(config->hpd_gpio)) {
+			DBG("gpio tells us we are connected!");
+			hpd_int_status |= HDMI_HPD_INT_STATUS_CABLE_DETECTED;
+			break;
+		}
 		mdelay(10);
 		hpd_int_status = hdmi_read(hdmi, REG_HDMI_HPD_INT_STATUS);
 		DBG("status=%08x", hpd_int_status);
@@ -285,6 +322,8 @@ static int hdmi_connector_mode_valid(struct drm_connector *connector,
 				 struct drm_display_mode *mode)
 {
 	struct hdmi_connector *hdmi_connector = to_hdmi_connector(connector);
+	struct hdmi *hdmi = hdmi_connector->hdmi;
+	const struct hdmi_platform_config *config = hdmi->config;
 	struct msm_drm_private *priv = connector->dev->dev_private;
 	struct msm_kms *kms = priv->kms;
 	long actual, requested;
@@ -293,6 +332,13 @@ static int hdmi_connector_mode_valid(struct drm_connector *connector,
 	actual = kms->funcs->round_pixclk(kms,
 			requested, hdmi_connector->hdmi->encoder);
 
+	/* for mdp5/apq8074, we manage our own pixel clk (as opposed to
+	 * mdp4/dtv stuff where pixel clk is assigned to mdp/encoder
+	 * instead):
+	 */
+	if (config->pwr_clk_cnt > 0)
+		actual = clk_round_rate(hdmi->pwr_clks[0], actual);
+
 	DBG("requested=%ld, actual=%ld", requested, actual);
 
 	if (actual != requested)
@@ -335,6 +381,7 @@ struct drm_connector *hdmi_connector_init(struct hdmi *hdmi)
 	}
 
 	hdmi_connector->hdmi = hdmi_reference(hdmi);
+	INIT_WORK(&hdmi_connector->hpd_work, hotplug_work);
 
 	connector = &hdmi_connector->base;
 
diff --git a/drivers/gpu/drm/msm/hdmi/hdmi_phy_8x74.c b/drivers/gpu/drm/msm/hdmi/hdmi_phy_8x74.c
new file mode 100644
index 000000000000..59fa6cdacb2a
--- /dev/null
+++ b/drivers/gpu/drm/msm/hdmi/hdmi_phy_8x74.c
@@ -0,0 +1,157 @@
+/*
+ * Copyright (C) 2013 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 "hdmi.h"
+
+struct hdmi_phy_8x74 {
+	struct hdmi_phy base;
+	struct hdmi *hdmi;
+	void __iomem *mmio;
+};
+#define to_hdmi_phy_8x74(x) container_of(x, struct hdmi_phy_8x74, base)
+
+
+static void phy_write(struct hdmi_phy_8x74 *phy, u32 reg, u32 data)
+{
+	msm_writel(data, phy->mmio + reg);
+}
+
+//static u32 phy_read(struct hdmi_phy_8x74 *phy, u32 reg)
+//{
+//	return msm_readl(phy->mmio + reg);
+//}
+
+static void hdmi_phy_8x74_destroy(struct hdmi_phy *phy)
+{
+	struct hdmi_phy_8x74 *phy_8x74 = to_hdmi_phy_8x74(phy);
+	kfree(phy_8x74);
+}
+
+static void hdmi_phy_8x74_reset(struct hdmi_phy *phy)
+{
+	struct hdmi_phy_8x74 *phy_8x74 = to_hdmi_phy_8x74(phy);
+	struct hdmi *hdmi = phy_8x74->hdmi;
+	unsigned int val;
+
+	/* NOTE that HDMI_PHY_CTL is in core mmio, not phy mmio: */
+
+	val = hdmi_read(hdmi, REG_HDMI_PHY_CTRL);
+
+	if (val & HDMI_PHY_CTRL_SW_RESET_LOW) {
+		/* pull low */
+		hdmi_write(hdmi, REG_HDMI_PHY_CTRL,
+				val & ~HDMI_PHY_CTRL_SW_RESET);
+	} else {
+		/* pull high */
+		hdmi_write(hdmi, REG_HDMI_PHY_CTRL,
+				val | HDMI_PHY_CTRL_SW_RESET);
+	}
+
+	if (val & HDMI_PHY_CTRL_SW_RESET_PLL_LOW) {
+		/* pull low */
+		hdmi_write(hdmi, REG_HDMI_PHY_CTRL,
+				val & ~HDMI_PHY_CTRL_SW_RESET_PLL);
+	} else {
+		/* pull high */
+		hdmi_write(hdmi, REG_HDMI_PHY_CTRL,
+				val | HDMI_PHY_CTRL_SW_RESET_PLL);
+	}
+
+	msleep(100);
+
+	if (val & HDMI_PHY_CTRL_SW_RESET_LOW) {
+		/* pull high */
+		hdmi_write(hdmi, REG_HDMI_PHY_CTRL,
+				val | HDMI_PHY_CTRL_SW_RESET);
+	} else {
+		/* pull low */
+		hdmi_write(hdmi, REG_HDMI_PHY_CTRL,
+				val & ~HDMI_PHY_CTRL_SW_RESET);
+	}
+
+	if (val & HDMI_PHY_CTRL_SW_RESET_PLL_LOW) {
+		/* pull high */
+		hdmi_write(hdmi, REG_HDMI_PHY_CTRL,
+				val | HDMI_PHY_CTRL_SW_RESET_PLL);
+	} else {
+		/* pull low */
+		hdmi_write(hdmi, REG_HDMI_PHY_CTRL,
+				val & ~HDMI_PHY_CTRL_SW_RESET_PLL);
+	}
+}
+
+static void hdmi_phy_8x74_powerup(struct hdmi_phy *phy,
+		unsigned long int pixclock)
+{
+	struct hdmi_phy_8x74 *phy_8x74 = to_hdmi_phy_8x74(phy);
+
+	phy_write(phy_8x74, REG_HDMI_8x74_ANA_CFG0,   0x1b);
+	phy_write(phy_8x74, REG_HDMI_8x74_ANA_CFG1,   0xf2);
+	phy_write(phy_8x74, REG_HDMI_8x74_BIST_CFG0,  0x0);
+	phy_write(phy_8x74, REG_HDMI_8x74_BIST_PATN0, 0x0);
+	phy_write(phy_8x74, REG_HDMI_8x74_BIST_PATN1, 0x0);
+	phy_write(phy_8x74, REG_HDMI_8x74_BIST_PATN2, 0x0);
+	phy_write(phy_8x74, REG_HDMI_8x74_BIST_PATN3, 0x0);
+	phy_write(phy_8x74, REG_HDMI_8x74_PD_CTRL1,   0x20);
+}
+
+static void hdmi_phy_8x74_powerdown(struct hdmi_phy *phy)
+{
+	struct hdmi_phy_8x74 *phy_8x74 = to_hdmi_phy_8x74(phy);
+	phy_write(phy_8x74, REG_HDMI_8x74_PD_CTRL0, 0x7f);
+}
+
+static const struct hdmi_phy_funcs hdmi_phy_8x74_funcs = {
+		.destroy = hdmi_phy_8x74_destroy,
+		.reset = hdmi_phy_8x74_reset,
+		.powerup = hdmi_phy_8x74_powerup,
+		.powerdown = hdmi_phy_8x74_powerdown,
+};
+
+struct hdmi_phy *hdmi_phy_8x74_init(struct hdmi *hdmi)
+{
+	struct hdmi_phy_8x74 *phy_8x74;
+	struct hdmi_phy *phy = NULL;
+	int ret;
+
+	phy_8x74 = kzalloc(sizeof(*phy_8x74), GFP_KERNEL);
+	if (!phy_8x74) {
+		ret = -ENOMEM;
+		goto fail;
+	}
+
+	phy = &phy_8x74->base;
+
+	phy->funcs = &hdmi_phy_8x74_funcs;
+
+	phy_8x74->hdmi = hdmi;
+
+	/* for 8x74, the phy mmio is mapped separately: */
+	phy_8x74->mmio = msm_ioremap(hdmi->pdev,
+			"phy_physical", "HDMI_8x74");
+	if (IS_ERR(phy_8x74->mmio)) {
+		ret = PTR_ERR(phy_8x74->mmio);
+		goto fail;
+	}
+
+	return phy;
+
+fail:
+	if (phy)
+		hdmi_phy_8x74_destroy(phy);
+	return ERR_PTR(ret);
+}
diff --git a/drivers/gpu/drm/msm/hdmi/qfprom.xml.h b/drivers/gpu/drm/msm/hdmi/qfprom.xml.h
index dbde4f6339b9..d591567173c4 100644
--- a/drivers/gpu/drm/msm/hdmi/qfprom.xml.h
+++ b/drivers/gpu/drm/msm/hdmi/qfprom.xml.h
@@ -8,14 +8,16 @@ http://github.com/freedreno/envytools/
 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                 (    595 bytes, from 2013-07-05 19:21:12)
+- /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/mdp4/mdp4.xml           (  19332 bytes, from 2013-10-07 16:36:48)
+- /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 2013-12-03 20:59:13)
 - /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/hdmi/qfprom.xml         (    600 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  19288 bytes, from 2013-08-11 18:14:15)
+- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  20932 bytes, from 2013-12-01 15:13:04)
 
 Copyright (C) 2013 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
diff --git a/drivers/gpu/drm/msm/mdp4/mdp4.xml.h b/drivers/gpu/drm/msm/mdp/mdp4/mdp4.xml.h
index 9908ffe1c3ad..416a26e1e58d 100644
--- a/drivers/gpu/drm/msm/mdp4/mdp4.xml.h
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4.xml.h
@@ -8,14 +8,16 @@ http://github.com/freedreno/envytools/
 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                 (    595 bytes, from 2013-07-05 19:21:12)
+- /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/mdp4/mdp4.xml           (  19332 bytes, from 2013-10-07 16:36:48)
+- /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 2013-12-03 20:59:13)
 - /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/hdmi/qfprom.xml         (    600 bytes, from 2013-07-05 19:21:12)
-- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  19288 bytes, from 2013-08-11 18:14:15)
+- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  20932 bytes, from 2013-12-01 15:13:04)
 
 Copyright (C) 2013 by the following authors:
 - Rob Clark <robdclark@gmail.com> (robclark)
@@ -42,27 +44,6 @@ WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
 */
 
 
-enum mdp4_bpc {
-	BPC1 = 0,
-	BPC5 = 1,
-	BPC6 = 2,
-	BPC8 = 3,
-};
-
-enum mdp4_bpc_alpha {
-	BPC1A = 0,
-	BPC4A = 1,
-	BPC6A = 2,
-	BPC8A = 3,
-};
-
-enum mdp4_alpha_type {
-	FG_CONST = 0,
-	BG_CONST = 1,
-	FG_PIXEL = 2,
-	BG_PIXEL = 3,
-};
-
 enum mdp4_pipe {
 	VG1 = 0,
 	VG2 = 1,
@@ -79,15 +60,6 @@ enum mdp4_mixer {
 	MIXER2 = 2,
 };
 
-enum mdp4_mixer_stage_id {
-	STAGE_UNUSED = 0,
-	STAGE_BASE = 1,
-	STAGE0 = 2,
-	STAGE1 = 3,
-	STAGE2 = 4,
-	STAGE3 = 5,
-};
-
 enum mdp4_intf {
 	INTF_LCDC_DTV = 0,
 	INTF_DSI_VIDEO = 1,
@@ -194,56 +166,56 @@ static inline uint32_t MDP4_DISP_INTF_SEL_EXT(enum mdp4_intf val)
 #define REG_MDP4_LAYERMIXER2_IN_CFG				0x000100f0
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE0__MASK			0x00000007
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE0__SHIFT			0
-static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE0(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE0(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER2_IN_CFG_PIPE0__SHIFT) & MDP4_LAYERMIXER2_IN_CFG_PIPE0__MASK;
 }
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE0_MIXER1			0x00000008
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE1__MASK			0x00000070
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE1__SHIFT			4
-static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE1(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE1(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER2_IN_CFG_PIPE1__SHIFT) & MDP4_LAYERMIXER2_IN_CFG_PIPE1__MASK;
 }
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE1_MIXER1			0x00000080
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE2__MASK			0x00000700
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE2__SHIFT			8
-static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE2(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE2(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER2_IN_CFG_PIPE2__SHIFT) & MDP4_LAYERMIXER2_IN_CFG_PIPE2__MASK;
 }
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE2_MIXER1			0x00000800
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE3__MASK			0x00007000
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE3__SHIFT			12
-static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE3(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE3(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER2_IN_CFG_PIPE3__SHIFT) & MDP4_LAYERMIXER2_IN_CFG_PIPE3__MASK;
 }
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE3_MIXER1			0x00008000
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE4__MASK			0x00070000
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE4__SHIFT			16
-static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE4(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE4(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER2_IN_CFG_PIPE4__SHIFT) & MDP4_LAYERMIXER2_IN_CFG_PIPE4__MASK;
 }
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE4_MIXER1			0x00080000
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE5__MASK			0x00700000
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE5__SHIFT			20
-static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE5(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE5(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER2_IN_CFG_PIPE5__SHIFT) & MDP4_LAYERMIXER2_IN_CFG_PIPE5__MASK;
 }
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE5_MIXER1			0x00800000
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE6__MASK			0x07000000
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE6__SHIFT			24
-static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE6(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE6(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER2_IN_CFG_PIPE6__SHIFT) & MDP4_LAYERMIXER2_IN_CFG_PIPE6__MASK;
 }
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE6_MIXER1			0x08000000
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE7__MASK			0x70000000
 #define MDP4_LAYERMIXER2_IN_CFG_PIPE7__SHIFT			28
-static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE7(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE7(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER2_IN_CFG_PIPE7__SHIFT) & MDP4_LAYERMIXER2_IN_CFG_PIPE7__MASK;
 }
@@ -254,56 +226,56 @@ static inline uint32_t MDP4_LAYERMIXER2_IN_CFG_PIPE7(enum mdp4_mixer_stage_id va
 #define REG_MDP4_LAYERMIXER_IN_CFG				0x00010100
 #define MDP4_LAYERMIXER_IN_CFG_PIPE0__MASK			0x00000007
 #define MDP4_LAYERMIXER_IN_CFG_PIPE0__SHIFT			0
-static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE0(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE0(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER_IN_CFG_PIPE0__SHIFT) & MDP4_LAYERMIXER_IN_CFG_PIPE0__MASK;
 }
 #define MDP4_LAYERMIXER_IN_CFG_PIPE0_MIXER1			0x00000008
 #define MDP4_LAYERMIXER_IN_CFG_PIPE1__MASK			0x00000070
 #define MDP4_LAYERMIXER_IN_CFG_PIPE1__SHIFT			4
-static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE1(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE1(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER_IN_CFG_PIPE1__SHIFT) & MDP4_LAYERMIXER_IN_CFG_PIPE1__MASK;
 }
 #define MDP4_LAYERMIXER_IN_CFG_PIPE1_MIXER1			0x00000080
 #define MDP4_LAYERMIXER_IN_CFG_PIPE2__MASK			0x00000700
 #define MDP4_LAYERMIXER_IN_CFG_PIPE2__SHIFT			8
-static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE2(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE2(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER_IN_CFG_PIPE2__SHIFT) & MDP4_LAYERMIXER_IN_CFG_PIPE2__MASK;
 }
 #define MDP4_LAYERMIXER_IN_CFG_PIPE2_MIXER1			0x00000800
 #define MDP4_LAYERMIXER_IN_CFG_PIPE3__MASK			0x00007000
 #define MDP4_LAYERMIXER_IN_CFG_PIPE3__SHIFT			12
-static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE3(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE3(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER_IN_CFG_PIPE3__SHIFT) & MDP4_LAYERMIXER_IN_CFG_PIPE3__MASK;
 }
 #define MDP4_LAYERMIXER_IN_CFG_PIPE3_MIXER1			0x00008000
 #define MDP4_LAYERMIXER_IN_CFG_PIPE4__MASK			0x00070000
 #define MDP4_LAYERMIXER_IN_CFG_PIPE4__SHIFT			16
-static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE4(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE4(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER_IN_CFG_PIPE4__SHIFT) & MDP4_LAYERMIXER_IN_CFG_PIPE4__MASK;
 }
 #define MDP4_LAYERMIXER_IN_CFG_PIPE4_MIXER1			0x00080000
 #define MDP4_LAYERMIXER_IN_CFG_PIPE5__MASK			0x00700000
 #define MDP4_LAYERMIXER_IN_CFG_PIPE5__SHIFT			20
-static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE5(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE5(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER_IN_CFG_PIPE5__SHIFT) & MDP4_LAYERMIXER_IN_CFG_PIPE5__MASK;
 }
 #define MDP4_LAYERMIXER_IN_CFG_PIPE5_MIXER1			0x00800000
 #define MDP4_LAYERMIXER_IN_CFG_PIPE6__MASK			0x07000000
 #define MDP4_LAYERMIXER_IN_CFG_PIPE6__SHIFT			24
-static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE6(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE6(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER_IN_CFG_PIPE6__SHIFT) & MDP4_LAYERMIXER_IN_CFG_PIPE6__MASK;
 }
 #define MDP4_LAYERMIXER_IN_CFG_PIPE6_MIXER1			0x08000000
 #define MDP4_LAYERMIXER_IN_CFG_PIPE7__MASK			0x70000000
 #define MDP4_LAYERMIXER_IN_CFG_PIPE7__SHIFT			28
-static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE7(enum mdp4_mixer_stage_id val)
+static inline uint32_t MDP4_LAYERMIXER_IN_CFG_PIPE7(enum mdp_mixer_stage_id val)
 {
 	return ((val) << MDP4_LAYERMIXER_IN_CFG_PIPE7__SHIFT) & MDP4_LAYERMIXER_IN_CFG_PIPE7__MASK;
 }
@@ -369,7 +341,7 @@ static inline uint32_t REG_MDP4_OVLP_STAGE(uint32_t i0, uint32_t i1) { return 0x
 static inline uint32_t REG_MDP4_OVLP_STAGE_OP(uint32_t i0, uint32_t i1) { return 0x00000000 + __offset_OVLP(i0) + __offset_STAGE(i1); }
 #define MDP4_OVLP_STAGE_OP_FG_ALPHA__MASK			0x00000003
 #define MDP4_OVLP_STAGE_OP_FG_ALPHA__SHIFT			0
-static inline uint32_t MDP4_OVLP_STAGE_OP_FG_ALPHA(enum mdp4_alpha_type val)
+static inline uint32_t MDP4_OVLP_STAGE_OP_FG_ALPHA(enum mdp_alpha_type val)
 {
 	return ((val) << MDP4_OVLP_STAGE_OP_FG_ALPHA__SHIFT) & MDP4_OVLP_STAGE_OP_FG_ALPHA__MASK;
 }
@@ -377,7 +349,7 @@ static inline uint32_t MDP4_OVLP_STAGE_OP_FG_ALPHA(enum mdp4_alpha_type val)
 #define MDP4_OVLP_STAGE_OP_FG_MOD_ALPHA				0x00000008
 #define MDP4_OVLP_STAGE_OP_BG_ALPHA__MASK			0x00000030
 #define MDP4_OVLP_STAGE_OP_BG_ALPHA__SHIFT			4
-static inline uint32_t MDP4_OVLP_STAGE_OP_BG_ALPHA(enum mdp4_alpha_type val)
+static inline uint32_t MDP4_OVLP_STAGE_OP_BG_ALPHA(enum mdp_alpha_type val)
 {
 	return ((val) << MDP4_OVLP_STAGE_OP_BG_ALPHA__SHIFT) & MDP4_OVLP_STAGE_OP_BG_ALPHA__MASK;
 }
@@ -472,19 +444,19 @@ static inline uint32_t REG_MDP4_DMA(enum mdp4_dma i0) { return 0x00000000 + __of
 static inline uint32_t REG_MDP4_DMA_CONFIG(enum mdp4_dma i0) { return 0x00000000 + __offset_DMA(i0); }
 #define MDP4_DMA_CONFIG_G_BPC__MASK				0x00000003
 #define MDP4_DMA_CONFIG_G_BPC__SHIFT				0
-static inline uint32_t MDP4_DMA_CONFIG_G_BPC(enum mdp4_bpc val)
+static inline uint32_t MDP4_DMA_CONFIG_G_BPC(enum mdp_bpc val)
 {
 	return ((val) << MDP4_DMA_CONFIG_G_BPC__SHIFT) & MDP4_DMA_CONFIG_G_BPC__MASK;
 }
 #define MDP4_DMA_CONFIG_B_BPC__MASK				0x0000000c
 #define MDP4_DMA_CONFIG_B_BPC__SHIFT				2
-static inline uint32_t MDP4_DMA_CONFIG_B_BPC(enum mdp4_bpc val)
+static inline uint32_t MDP4_DMA_CONFIG_B_BPC(enum mdp_bpc val)
 {
 	return ((val) << MDP4_DMA_CONFIG_B_BPC__SHIFT) & MDP4_DMA_CONFIG_B_BPC__MASK;
 }
 #define MDP4_DMA_CONFIG_R_BPC__MASK				0x00000030
 #define MDP4_DMA_CONFIG_R_BPC__SHIFT				4
-static inline uint32_t MDP4_DMA_CONFIG_R_BPC(enum mdp4_bpc val)
+static inline uint32_t MDP4_DMA_CONFIG_R_BPC(enum mdp_bpc val)
 {
 	return ((val) << MDP4_DMA_CONFIG_R_BPC__SHIFT) & MDP4_DMA_CONFIG_R_BPC__MASK;
 }
@@ -710,25 +682,25 @@ static inline uint32_t MDP4_PIPE_FRAME_SIZE_WIDTH(uint32_t val)
 static inline uint32_t REG_MDP4_PIPE_SRC_FORMAT(enum mdp4_pipe i0) { return 0x00020050 + 0x10000*i0; }
 #define MDP4_PIPE_SRC_FORMAT_G_BPC__MASK			0x00000003
 #define MDP4_PIPE_SRC_FORMAT_G_BPC__SHIFT			0
-static inline uint32_t MDP4_PIPE_SRC_FORMAT_G_BPC(enum mdp4_bpc val)
+static inline uint32_t MDP4_PIPE_SRC_FORMAT_G_BPC(enum mdp_bpc val)
 {
 	return ((val) << MDP4_PIPE_SRC_FORMAT_G_BPC__SHIFT) & MDP4_PIPE_SRC_FORMAT_G_BPC__MASK;
 }
 #define MDP4_PIPE_SRC_FORMAT_B_BPC__MASK			0x0000000c
 #define MDP4_PIPE_SRC_FORMAT_B_BPC__SHIFT			2
-static inline uint32_t MDP4_PIPE_SRC_FORMAT_B_BPC(enum mdp4_bpc val)
+static inline uint32_t MDP4_PIPE_SRC_FORMAT_B_BPC(enum mdp_bpc val)
 {
 	return ((val) << MDP4_PIPE_SRC_FORMAT_B_BPC__SHIFT) & MDP4_PIPE_SRC_FORMAT_B_BPC__MASK;
 }
 #define MDP4_PIPE_SRC_FORMAT_R_BPC__MASK			0x00000030
 #define MDP4_PIPE_SRC_FORMAT_R_BPC__SHIFT			4
-static inline uint32_t MDP4_PIPE_SRC_FORMAT_R_BPC(enum mdp4_bpc val)
+static inline uint32_t MDP4_PIPE_SRC_FORMAT_R_BPC(enum mdp_bpc val)
 {
 	return ((val) << MDP4_PIPE_SRC_FORMAT_R_BPC__SHIFT) & MDP4_PIPE_SRC_FORMAT_R_BPC__MASK;
 }
 #define MDP4_PIPE_SRC_FORMAT_A_BPC__MASK			0x000000c0
 #define MDP4_PIPE_SRC_FORMAT_A_BPC__SHIFT			6
-static inline uint32_t MDP4_PIPE_SRC_FORMAT_A_BPC(enum mdp4_bpc_alpha val)
+static inline uint32_t MDP4_PIPE_SRC_FORMAT_A_BPC(enum mdp_bpc_alpha val)
 {
 	return ((val) << MDP4_PIPE_SRC_FORMAT_A_BPC__SHIFT) & MDP4_PIPE_SRC_FORMAT_A_BPC__MASK;
 }
diff --git a/drivers/gpu/drm/msm/mdp4/mdp4_crtc.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c
index 019d530187ff..1964f4f0d452 100644
--- a/drivers/gpu/drm/msm/mdp4/mdp4_crtc.c
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c
@@ -66,15 +66,15 @@ struct mdp4_crtc {
 	/* for unref'ing cursor bo's after scanout completes: */
 	struct drm_flip_work unref_cursor_work;
 
-	struct mdp4_irq vblank;
-	struct mdp4_irq err;
+	struct mdp_irq vblank;
+	struct mdp_irq err;
 };
 #define to_mdp4_crtc(x) container_of(x, struct mdp4_crtc, base)
 
 static struct mdp4_kms *get_kms(struct drm_crtc *crtc)
 {
 	struct msm_drm_private *priv = crtc->dev->dev_private;
-	return to_mdp4_kms(priv->kms);
+	return to_mdp4_kms(to_mdp_kms(priv->kms));
 }
 
 static void update_fb(struct drm_crtc *crtc, bool async,
@@ -93,7 +93,7 @@ static void update_fb(struct drm_crtc *crtc, bool async,
 
 	if (!async) {
 		/* enable vblank to pick up the old_fb */
-		mdp4_irq_register(get_kms(crtc), &mdp4_crtc->vblank);
+		mdp_irq_register(&get_kms(crtc)->base, &mdp4_crtc->vblank);
 	}
 }
 
@@ -145,7 +145,7 @@ static void request_pending(struct drm_crtc *crtc, uint32_t pending)
 	struct mdp4_crtc *mdp4_crtc = to_mdp4_crtc(crtc);
 
 	atomic_or(pending, &mdp4_crtc->pending);
-	mdp4_irq_register(get_kms(crtc), &mdp4_crtc->vblank);
+	mdp_irq_register(&get_kms(crtc)->base, &mdp4_crtc->vblank);
 }
 
 static void pageflip_cb(struct msm_fence_cb *cb)
@@ -210,9 +210,9 @@ static void mdp4_crtc_dpms(struct drm_crtc *crtc, int mode)
 	if (enabled != mdp4_crtc->enabled) {
 		if (enabled) {
 			mdp4_enable(mdp4_kms);
-			mdp4_irq_register(mdp4_kms, &mdp4_crtc->err);
+			mdp_irq_register(&mdp4_kms->base, &mdp4_crtc->err);
 		} else {
-			mdp4_irq_unregister(mdp4_kms, &mdp4_crtc->err);
+			mdp_irq_unregister(&mdp4_kms->base, &mdp4_crtc->err);
 			mdp4_disable(mdp4_kms);
 		}
 		mdp4_crtc->enabled = enabled;
@@ -232,7 +232,7 @@ static void blend_setup(struct drm_crtc *crtc)
 	struct mdp4_kms *mdp4_kms = get_kms(crtc);
 	int i, ovlp = mdp4_crtc->ovlp;
 	uint32_t mixer_cfg = 0;
-	static const enum mdp4_mixer_stage_id stages[] = {
+	static const enum mdp_mixer_stage_id stages[] = {
 			STAGE_BASE, STAGE0, STAGE1, STAGE2, STAGE3,
 	};
 	/* statically (for now) map planes to mixer stage (z-order): */
@@ -262,8 +262,8 @@ static void blend_setup(struct drm_crtc *crtc)
 			enum mdp4_pipe pipe_id = mdp4_plane_pipe(plane);
 			int idx = idxs[pipe_id];
 			if (idx > 0) {
-				const struct mdp4_format *format =
-					to_mdp4_format(msm_framebuffer_format(plane->fb));
+				const struct mdp_format *format =
+					to_mdp_format(msm_framebuffer_format(plane->fb));
 				alpha[idx-1] = format->alpha_enable;
 			}
 			mixer_cfg |= mixercfg(mdp4_crtc->mixer, pipe_id, stages[idx]);
@@ -571,14 +571,14 @@ static const struct drm_crtc_helper_funcs mdp4_crtc_helper_funcs = {
 	.load_lut = mdp4_crtc_load_lut,
 };
 
-static void mdp4_crtc_vblank_irq(struct mdp4_irq *irq, uint32_t irqstatus)
+static void mdp4_crtc_vblank_irq(struct mdp_irq *irq, uint32_t irqstatus)
 {
 	struct mdp4_crtc *mdp4_crtc = container_of(irq, struct mdp4_crtc, vblank);
 	struct drm_crtc *crtc = &mdp4_crtc->base;
 	struct msm_drm_private *priv = crtc->dev->dev_private;
 	unsigned pending;
 
-	mdp4_irq_unregister(get_kms(crtc), &mdp4_crtc->vblank);
+	mdp_irq_unregister(&get_kms(crtc)->base, &mdp4_crtc->vblank);
 
 	pending = atomic_xchg(&mdp4_crtc->pending, 0);
 
@@ -593,7 +593,7 @@ static void mdp4_crtc_vblank_irq(struct mdp4_irq *irq, uint32_t irqstatus)
 	}
 }
 
-static void mdp4_crtc_err_irq(struct mdp4_irq *irq, uint32_t irqstatus)
+static void mdp4_crtc_err_irq(struct mdp_irq *irq, uint32_t irqstatus)
 {
 	struct mdp4_crtc *mdp4_crtc = container_of(irq, struct mdp4_crtc, err);
 	struct drm_crtc *crtc = &mdp4_crtc->base;
diff --git a/drivers/gpu/drm/msm/mdp4/mdp4_dtv_encoder.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_dtv_encoder.c
index 5e0dcae70ab5..067ed03b35fe 100644
--- a/drivers/gpu/drm/msm/mdp4/mdp4_dtv_encoder.c
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_dtv_encoder.c
@@ -15,8 +15,6 @@
  * this program.  If not, see <http://www.gnu.org/licenses/>.
  */
 
-#include <mach/clk.h>
-
 #include "mdp4_kms.h"
 
 #include "drm_crtc.h"
@@ -37,7 +35,7 @@ struct mdp4_dtv_encoder {
 static struct mdp4_kms *get_kms(struct drm_encoder *encoder)
 {
 	struct msm_drm_private *priv = encoder->dev->dev_private;
-	return to_mdp4_kms(priv->kms);
+	return to_mdp4_kms(to_mdp_kms(priv->kms));
 }
 
 #ifdef CONFIG_MSM_BUS_SCALING
@@ -139,7 +137,7 @@ static void mdp4_dtv_encoder_dpms(struct drm_encoder *encoder, int mode)
 		 * the settings changes for the new modeset (like new
 		 * scanout buffer) don't latch properly..
 		 */
-		mdp4_irq_wait(mdp4_kms, MDP4_IRQ_EXTERNAL_VSYNC);
+		mdp_irq_wait(&mdp4_kms->base, MDP4_IRQ_EXTERNAL_VSYNC);
 
 		clk_disable_unprepare(mdp4_dtv_encoder->src_clk);
 		clk_disable_unprepare(mdp4_dtv_encoder->hdmi_clk);
diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_irq.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_irq.c
new file mode 100644
index 000000000000..c740ccd1cc67
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_irq.c
@@ -0,0 +1,93 @@
+/*
+ * Copyright (C) 2013 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 "msm_drv.h"
+#include "mdp4_kms.h"
+
+void mdp4_set_irqmask(struct mdp_kms *mdp_kms, uint32_t irqmask)
+{
+	mdp4_write(to_mdp4_kms(mdp_kms), REG_MDP4_INTR_ENABLE, irqmask);
+}
+
+static void mdp4_irq_error_handler(struct mdp_irq *irq, uint32_t irqstatus)
+{
+	DRM_ERROR("errors: %08x\n", irqstatus);
+}
+
+void mdp4_irq_preinstall(struct msm_kms *kms)
+{
+	struct mdp4_kms *mdp4_kms = to_mdp4_kms(to_mdp_kms(kms));
+	mdp4_write(mdp4_kms, REG_MDP4_INTR_CLEAR, 0xffffffff);
+}
+
+int mdp4_irq_postinstall(struct msm_kms *kms)
+{
+	struct mdp_kms *mdp_kms = to_mdp_kms(kms);
+	struct mdp4_kms *mdp4_kms = to_mdp4_kms(mdp_kms);
+	struct mdp_irq *error_handler = &mdp4_kms->error_handler;
+
+	error_handler->irq = mdp4_irq_error_handler;
+	error_handler->irqmask = MDP4_IRQ_PRIMARY_INTF_UDERRUN |
+			MDP4_IRQ_EXTERNAL_INTF_UDERRUN;
+
+	mdp_irq_register(mdp_kms, error_handler);
+
+	return 0;
+}
+
+void mdp4_irq_uninstall(struct msm_kms *kms)
+{
+	struct mdp4_kms *mdp4_kms = to_mdp4_kms(to_mdp_kms(kms));
+	mdp4_write(mdp4_kms, REG_MDP4_INTR_ENABLE, 0x00000000);
+}
+
+irqreturn_t mdp4_irq(struct msm_kms *kms)
+{
+	struct mdp_kms *mdp_kms = to_mdp_kms(kms);
+	struct mdp4_kms *mdp4_kms = to_mdp4_kms(mdp_kms);
+	struct drm_device *dev = mdp4_kms->dev;
+	struct msm_drm_private *priv = dev->dev_private;
+	unsigned int id;
+	uint32_t status;
+
+	status = mdp4_read(mdp4_kms, REG_MDP4_INTR_STATUS);
+	mdp4_write(mdp4_kms, REG_MDP4_INTR_CLEAR, status);
+
+	VERB("status=%08x", status);
+
+	for (id = 0; id < priv->num_crtcs; id++)
+		if (status & mdp4_crtc_vblank(priv->crtcs[id]))
+			drm_handle_vblank(dev, id);
+
+	mdp_dispatch_irqs(mdp_kms, status);
+
+	return IRQ_HANDLED;
+}
+
+int mdp4_enable_vblank(struct msm_kms *kms, struct drm_crtc *crtc)
+{
+	mdp_update_vblank_mask(to_mdp_kms(kms),
+			mdp4_crtc_vblank(crtc), true);
+	return 0;
+}
+
+void mdp4_disable_vblank(struct msm_kms *kms, struct drm_crtc *crtc)
+{
+	mdp_update_vblank_mask(to_mdp_kms(kms),
+			mdp4_crtc_vblank(crtc), false);
+}
diff --git a/drivers/gpu/drm/msm/mdp4/mdp4_kms.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.c
index 8972ac35a43d..272e707c9487 100644
--- a/drivers/gpu/drm/msm/mdp4/mdp4_kms.c
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.c
@@ -17,13 +17,14 @@
 
 
 #include "msm_drv.h"
+#include "msm_mmu.h"
 #include "mdp4_kms.h"
 
 static struct mdp4_platform_config *mdp4_get_config(struct platform_device *dev);
 
 static int mdp4_hw_init(struct msm_kms *kms)
 {
-	struct mdp4_kms *mdp4_kms = to_mdp4_kms(kms);
+	struct mdp4_kms *mdp4_kms = to_mdp4_kms(to_mdp_kms(kms));
 	struct drm_device *dev = mdp4_kms->dev;
 	uint32_t version, major, minor, dmap_cfg, vg_cfg;
 	unsigned long clk;
@@ -31,12 +32,14 @@ static int mdp4_hw_init(struct msm_kms *kms)
 
 	pm_runtime_get_sync(dev->dev);
 
+	mdp4_enable(mdp4_kms);
 	version = mdp4_read(mdp4_kms, REG_MDP4_VERSION);
+	mdp4_disable(mdp4_kms);
 
 	major = FIELD(version, MDP4_VERSION_MAJOR);
 	minor = FIELD(version, MDP4_VERSION_MINOR);
 
-	DBG("found MDP version v%d.%d", major, minor);
+	DBG("found MDP4 version v%d.%d", major, minor);
 
 	if (major != 4) {
 		dev_err(dev->dev, "unexpected MDP version: v%d.%d\n",
@@ -130,7 +133,7 @@ static long mdp4_round_pixclk(struct msm_kms *kms, unsigned long rate,
 
 static void mdp4_preclose(struct msm_kms *kms, struct drm_file *file)
 {
-	struct mdp4_kms *mdp4_kms = to_mdp4_kms(kms);
+	struct mdp4_kms *mdp4_kms = to_mdp4_kms(to_mdp_kms(kms));
 	struct msm_drm_private *priv = mdp4_kms->dev->dev_private;
 	unsigned i;
 
@@ -140,11 +143,12 @@ static void mdp4_preclose(struct msm_kms *kms, struct drm_file *file)
 
 static void mdp4_destroy(struct msm_kms *kms)
 {
-	struct mdp4_kms *mdp4_kms = to_mdp4_kms(kms);
+	struct mdp4_kms *mdp4_kms = to_mdp4_kms(to_mdp_kms(kms));
 	kfree(mdp4_kms);
 }
 
-static const struct msm_kms_funcs kms_funcs = {
+static const struct mdp_kms_funcs kms_funcs = {
+	.base = {
 		.hw_init         = mdp4_hw_init,
 		.irq_preinstall  = mdp4_irq_preinstall,
 		.irq_postinstall = mdp4_irq_postinstall,
@@ -152,10 +156,12 @@ static const struct msm_kms_funcs kms_funcs = {
 		.irq             = mdp4_irq,
 		.enable_vblank   = mdp4_enable_vblank,
 		.disable_vblank  = mdp4_disable_vblank,
-		.get_format      = mdp4_get_format,
+		.get_format      = mdp_get_format,
 		.round_pixclk    = mdp4_round_pixclk,
 		.preclose        = mdp4_preclose,
 		.destroy         = mdp4_destroy,
+	},
+	.set_irqmask         = mdp4_set_irqmask,
 };
 
 int mdp4_disable(struct mdp4_kms *mdp4_kms)
@@ -189,6 +195,7 @@ static int modeset_init(struct mdp4_kms *mdp4_kms)
 	struct drm_plane *plane;
 	struct drm_crtc *crtc;
 	struct drm_encoder *encoder;
+	struct hdmi *hdmi;
 	int ret;
 
 	/*
@@ -238,9 +245,10 @@ static int modeset_init(struct mdp4_kms *mdp4_kms)
 	encoder->possible_crtcs = 0x1;     /* DTV can be hooked to DMA_E */
 	priv->encoders[priv->num_encoders++] = encoder;
 
-	ret = hdmi_init(dev, encoder);
-	if (ret) {
-		dev_err(dev->dev, "failed to initialize HDMI\n");
+	hdmi = hdmi_init(dev, encoder);
+	if (IS_ERR(hdmi)) {
+		ret = PTR_ERR(hdmi);
+		dev_err(dev->dev, "failed to initialize HDMI: %d\n", ret);
 		goto fail;
 	}
 
@@ -260,6 +268,7 @@ struct msm_kms *mdp4_kms_init(struct drm_device *dev)
 	struct mdp4_platform_config *config = mdp4_get_config(pdev);
 	struct mdp4_kms *mdp4_kms;
 	struct msm_kms *kms = NULL;
+	struct msm_mmu *mmu;
 	int ret;
 
 	mdp4_kms = kzalloc(sizeof(*mdp4_kms), GFP_KERNEL);
@@ -269,8 +278,9 @@ struct msm_kms *mdp4_kms_init(struct drm_device *dev)
 		goto fail;
 	}
 
-	kms = &mdp4_kms->base;
-	kms->funcs = &kms_funcs;
+	mdp_kms_init(&mdp4_kms->base, &kms_funcs);
+
+	kms = &mdp4_kms->base.base;
 
 	mdp4_kms->dev = dev;
 
@@ -322,27 +332,34 @@ struct msm_kms *mdp4_kms_init(struct drm_device *dev)
 	clk_set_rate(mdp4_kms->clk, config->max_clk);
 	clk_set_rate(mdp4_kms->lut_clk, config->max_clk);
 
-	if (!config->iommu) {
-		dev_err(dev->dev, "no iommu\n");
-		ret = -ENXIO;
-		goto fail;
-	}
-
 	/* make sure things are off before attaching iommu (bootloader could
 	 * have left things on, in which case we'll start getting faults if
 	 * we don't disable):
 	 */
+	mdp4_enable(mdp4_kms);
 	mdp4_write(mdp4_kms, REG_MDP4_DTV_ENABLE, 0);
 	mdp4_write(mdp4_kms, REG_MDP4_LCDC_ENABLE, 0);
 	mdp4_write(mdp4_kms, REG_MDP4_DSI_ENABLE, 0);
+	mdp4_disable(mdp4_kms);
 	mdelay(16);
 
-	ret = msm_iommu_attach(dev, config->iommu,
-			iommu_ports, ARRAY_SIZE(iommu_ports));
-	if (ret)
-		goto fail;
+	if (config->iommu) {
+		mmu = msm_iommu_new(dev, config->iommu);
+		if (IS_ERR(mmu)) {
+			ret = PTR_ERR(mmu);
+			goto fail;
+		}
+		ret = mmu->funcs->attach(mmu, iommu_ports,
+				ARRAY_SIZE(iommu_ports));
+		if (ret)
+			goto fail;
+	} else {
+		dev_info(dev->dev, "no iommu, fallback to phys "
+				"contig buffers for scanout\n");
+		mmu = NULL;
+	}
 
-	mdp4_kms->id = msm_register_iommu(dev, config->iommu);
+	mdp4_kms->id = msm_register_mmu(dev, mmu);
 	if (mdp4_kms->id < 0) {
 		ret = mdp4_kms->id;
 		dev_err(dev->dev, "failed to register mdp4 iommu: %d\n", ret);
diff --git a/drivers/gpu/drm/msm/mdp4/mdp4_kms.h b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.h
index eb015c834087..66a4d31aec80 100644
--- a/drivers/gpu/drm/msm/mdp4/mdp4_kms.h
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.h
@@ -18,29 +18,13 @@
 #ifndef __MDP4_KMS_H__
 #define __MDP4_KMS_H__
 
-#include <linux/clk.h>
-#include <linux/platform_device.h>
-#include <linux/regulator/consumer.h>
-
 #include "msm_drv.h"
+#include "msm_kms.h"
+#include "mdp/mdp_kms.h"
 #include "mdp4.xml.h"
 
-
-/* For transiently registering for different MDP4 irqs that various parts
- * of the KMS code need during setup/configuration.  We these are not
- * necessarily the same as what drm_vblank_get/put() are requesting, and
- * the hysteresis in drm_vblank_put() is not necessarily desirable for
- * internal housekeeping related irq usage.
- */
-struct mdp4_irq {
-	struct list_head node;
-	uint32_t irqmask;
-	bool registered;
-	void (*irq)(struct mdp4_irq *irq, uint32_t irqstatus);
-};
-
 struct mdp4_kms {
-	struct msm_kms base;
+	struct mdp_kms base;
 
 	struct drm_device *dev;
 
@@ -59,11 +43,7 @@ struct mdp4_kms {
 	struct clk *pclk;
 	struct clk *lut_clk;
 
-	/* irq handling: */
-	bool in_irq;
-	struct list_head irq_list;    /* list of mdp4_irq */
-	uint32_t vblank_mask;         /* irq bits set for userspace vblank */
-	struct mdp4_irq error_handler;
+	struct mdp_irq error_handler;
 };
 #define to_mdp4_kms(x) container_of(x, struct mdp4_kms, base)
 
@@ -73,16 +53,6 @@ struct mdp4_platform_config {
 	uint32_t max_clk;
 };
 
-struct mdp4_format {
-	struct msm_format base;
-	enum mdp4_bpc bpc_r, bpc_g, bpc_b;
-	enum mdp4_bpc_alpha bpc_a;
-	uint8_t unpack[4];
-	bool alpha_enable, unpack_tight;
-	uint8_t cpp, unpack_count;
-};
-#define to_mdp4_format(x) container_of(x, struct mdp4_format, base)
-
 static inline void mdp4_write(struct mdp4_kms *mdp4_kms, u32 reg, u32 data)
 {
 	msm_writel(data, mdp4_kms->mmio + reg);
@@ -134,7 +104,7 @@ static inline uint32_t dma2err(enum mdp4_dma dma)
 }
 
 static inline uint32_t mixercfg(int mixer, enum mdp4_pipe pipe,
-		enum mdp4_mixer_stage_id stage)
+		enum mdp_mixer_stage_id stage)
 {
 	uint32_t mixer_cfg = 0;
 
@@ -178,19 +148,23 @@ static inline uint32_t mixercfg(int mixer, enum mdp4_pipe pipe,
 int mdp4_disable(struct mdp4_kms *mdp4_kms);
 int mdp4_enable(struct mdp4_kms *mdp4_kms);
 
+void mdp4_set_irqmask(struct mdp_kms *mdp_kms, uint32_t irqmask);
 void mdp4_irq_preinstall(struct msm_kms *kms);
 int mdp4_irq_postinstall(struct msm_kms *kms);
 void mdp4_irq_uninstall(struct msm_kms *kms);
 irqreturn_t mdp4_irq(struct msm_kms *kms);
-void mdp4_irq_wait(struct mdp4_kms *mdp4_kms, uint32_t irqmask);
-void mdp4_irq_register(struct mdp4_kms *mdp4_kms, struct mdp4_irq *irq);
-void mdp4_irq_unregister(struct mdp4_kms *mdp4_kms, struct mdp4_irq *irq);
 int mdp4_enable_vblank(struct msm_kms *kms, struct drm_crtc *crtc);
 void mdp4_disable_vblank(struct msm_kms *kms, struct drm_crtc *crtc);
 
-uint32_t mdp4_get_formats(enum mdp4_pipe pipe_id, uint32_t *formats,
-		uint32_t max_formats);
-const struct msm_format *mdp4_get_format(struct msm_kms *kms, uint32_t format);
+static inline
+uint32_t mdp4_get_formats(enum mdp4_pipe pipe_id, uint32_t *pixel_formats,
+		uint32_t max_formats)
+{
+	/* TODO when we have YUV, we need to filter supported formats
+	 * based on pipe_id..
+	 */
+	return mdp_get_formats(pixel_formats, max_formats);
+}
 
 void mdp4_plane_install_properties(struct drm_plane *plane,
 		struct drm_mode_object *obj);
diff --git a/drivers/gpu/drm/msm/mdp4/mdp4_plane.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_plane.c
index 0f0af243f6fc..2406027200ec 100644
--- a/drivers/gpu/drm/msm/mdp4/mdp4_plane.c
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_plane.c
@@ -34,7 +34,7 @@ struct mdp4_plane {
 static struct mdp4_kms *get_kms(struct drm_plane *plane)
 {
 	struct msm_drm_private *priv = plane->dev->dev_private;
-	return to_mdp4_kms(priv->kms);
+	return to_mdp4_kms(to_mdp_kms(priv->kms));
 }
 
 static int mdp4_plane_update(struct drm_plane *plane,
@@ -132,7 +132,7 @@ int mdp4_plane_mode_set(struct drm_plane *plane,
 	struct mdp4_plane *mdp4_plane = to_mdp4_plane(plane);
 	struct mdp4_kms *mdp4_kms = get_kms(plane);
 	enum mdp4_pipe pipe = mdp4_plane->pipe;
-	const struct mdp4_format *format;
+	const struct mdp_format *format;
 	uint32_t op_mode = 0;
 	uint32_t phasex_step = MDP4_VG_PHASE_STEP_DEFAULT;
 	uint32_t phasey_step = MDP4_VG_PHASE_STEP_DEFAULT;
@@ -175,7 +175,7 @@ int mdp4_plane_mode_set(struct drm_plane *plane,
 
 	mdp4_plane_set_scanout(plane, fb);
 
-	format = to_mdp4_format(msm_framebuffer_format(fb));
+	format = to_mdp_format(msm_framebuffer_format(fb));
 
 	mdp4_write(mdp4_kms, REG_MDP4_PIPE_SRC_FORMAT(pipe),
 			MDP4_PIPE_SRC_FORMAT_A_BPC(format->bpc_a) |
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5.xml.h b/drivers/gpu/drm/msm/mdp/mdp5/mdp5.xml.h
new file mode 100644
index 000000000000..0aa51517f826
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5.xml.h
@@ -0,0 +1,1036 @@
+#ifndef MDP5_XML
+#define MDP5_XML
+
+/* Autogenerated file, DO NOT EDIT manually!
+
+This file was generated by the rules-ng-ng headergen tool in this git repository:
+http://github.com/freedreno/envytools/
+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 2013-12-03 20:59:13)
+- /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/hdmi/qfprom.xml         (    600 bytes, from 2013-07-05 19:21:12)
+- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  20932 bytes, from 2013-12-01 15:13:04)
+
+Copyright (C) 2013 by the following authors:
+- Rob Clark <robdclark@gmail.com> (robclark)
+
+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 COPYRIGHT OWNER(S) 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.
+*/
+
+
+enum mdp5_intf {
+	INTF_DSI = 1,
+	INTF_HDMI = 3,
+	INTF_LCDC = 5,
+	INTF_eDP = 9,
+};
+
+enum mdp5_intfnum {
+	NO_INTF = 0,
+	INTF0 = 1,
+	INTF1 = 2,
+	INTF2 = 3,
+	INTF3 = 4,
+};
+
+enum mdp5_pipe {
+	SSPP_VIG0 = 0,
+	SSPP_VIG1 = 1,
+	SSPP_VIG2 = 2,
+	SSPP_RGB0 = 3,
+	SSPP_RGB1 = 4,
+	SSPP_RGB2 = 5,
+	SSPP_DMA0 = 6,
+	SSPP_DMA1 = 7,
+};
+
+enum mdp5_ctl_mode {
+	MODE_NONE = 0,
+	MODE_ROT0 = 1,
+	MODE_ROT1 = 2,
+	MODE_WB0 = 3,
+	MODE_WB1 = 4,
+	MODE_WFD = 5,
+};
+
+enum mdp5_pack_3d {
+	PACK_3D_FRAME_INT = 0,
+	PACK_3D_H_ROW_INT = 1,
+	PACK_3D_V_ROW_INT = 2,
+	PACK_3D_COL_INT = 3,
+};
+
+enum mdp5_chroma_samp_type {
+	CHROMA_RGB = 0,
+	CHROMA_H2V1 = 1,
+	CHROMA_H1V2 = 2,
+	CHROMA_420 = 3,
+};
+
+enum mdp5_scale_filter {
+	SCALE_FILTER_NEAREST = 0,
+	SCALE_FILTER_BIL = 1,
+	SCALE_FILTER_PCMN = 2,
+	SCALE_FILTER_CA = 3,
+};
+
+enum mdp5_pipe_bwc {
+	BWC_LOSSLESS = 0,
+	BWC_Q_HIGH = 1,
+	BWC_Q_MED = 2,
+};
+
+enum mdp5_client_id {
+	CID_UNUSED = 0,
+	CID_VIG0_Y = 1,
+	CID_VIG0_CR = 2,
+	CID_VIG0_CB = 3,
+	CID_VIG1_Y = 4,
+	CID_VIG1_CR = 5,
+	CID_VIG1_CB = 6,
+	CID_VIG2_Y = 7,
+	CID_VIG2_CR = 8,
+	CID_VIG2_CB = 9,
+	CID_DMA0_Y = 10,
+	CID_DMA0_CR = 11,
+	CID_DMA0_CB = 12,
+	CID_DMA1_Y = 13,
+	CID_DMA1_CR = 14,
+	CID_DMA1_CB = 15,
+	CID_RGB0 = 16,
+	CID_RGB1 = 17,
+	CID_RGB2 = 18,
+	CID_MAX = 19,
+};
+
+enum mdp5_igc_type {
+	IGC_VIG = 0,
+	IGC_RGB = 1,
+	IGC_DMA = 2,
+	IGC_DSPP = 3,
+};
+
+#define MDP5_IRQ_INTF0_WB_ROT_COMP				0x00000001
+#define MDP5_IRQ_INTF1_WB_ROT_COMP				0x00000002
+#define MDP5_IRQ_INTF2_WB_ROT_COMP				0x00000004
+#define MDP5_IRQ_INTF3_WB_ROT_COMP				0x00000008
+#define MDP5_IRQ_INTF0_WB_WFD					0x00000010
+#define MDP5_IRQ_INTF1_WB_WFD					0x00000020
+#define MDP5_IRQ_INTF2_WB_WFD					0x00000040
+#define MDP5_IRQ_INTF3_WB_WFD					0x00000080
+#define MDP5_IRQ_INTF0_PING_PONG_COMP				0x00000100
+#define MDP5_IRQ_INTF1_PING_PONG_COMP				0x00000200
+#define MDP5_IRQ_INTF2_PING_PONG_COMP				0x00000400
+#define MDP5_IRQ_INTF3_PING_PONG_COMP				0x00000800
+#define MDP5_IRQ_INTF0_PING_PONG_RD_PTR				0x00001000
+#define MDP5_IRQ_INTF1_PING_PONG_RD_PTR				0x00002000
+#define MDP5_IRQ_INTF2_PING_PONG_RD_PTR				0x00004000
+#define MDP5_IRQ_INTF3_PING_PONG_RD_PTR				0x00008000
+#define MDP5_IRQ_INTF0_PING_PONG_WR_PTR				0x00010000
+#define MDP5_IRQ_INTF1_PING_PONG_WR_PTR				0x00020000
+#define MDP5_IRQ_INTF2_PING_PONG_WR_PTR				0x00040000
+#define MDP5_IRQ_INTF3_PING_PONG_WR_PTR				0x00080000
+#define MDP5_IRQ_INTF0_PING_PONG_AUTO_REF			0x00100000
+#define MDP5_IRQ_INTF1_PING_PONG_AUTO_REF			0x00200000
+#define MDP5_IRQ_INTF2_PING_PONG_AUTO_REF			0x00400000
+#define MDP5_IRQ_INTF3_PING_PONG_AUTO_REF			0x00800000
+#define MDP5_IRQ_INTF0_UNDER_RUN				0x01000000
+#define MDP5_IRQ_INTF0_VSYNC					0x02000000
+#define MDP5_IRQ_INTF1_UNDER_RUN				0x04000000
+#define MDP5_IRQ_INTF1_VSYNC					0x08000000
+#define MDP5_IRQ_INTF2_UNDER_RUN				0x10000000
+#define MDP5_IRQ_INTF2_VSYNC					0x20000000
+#define MDP5_IRQ_INTF3_UNDER_RUN				0x40000000
+#define MDP5_IRQ_INTF3_VSYNC					0x80000000
+#define REG_MDP5_HW_VERSION					0x00000000
+
+#define REG_MDP5_HW_INTR_STATUS					0x00000010
+#define MDP5_HW_INTR_STATUS_INTR_MDP				0x00000001
+#define MDP5_HW_INTR_STATUS_INTR_DSI0				0x00000010
+#define MDP5_HW_INTR_STATUS_INTR_DSI1				0x00000020
+#define MDP5_HW_INTR_STATUS_INTR_HDMI				0x00000100
+#define MDP5_HW_INTR_STATUS_INTR_EDP				0x00001000
+
+#define REG_MDP5_MDP_VERSION					0x00000100
+#define MDP5_MDP_VERSION_MINOR__MASK				0x00ff0000
+#define MDP5_MDP_VERSION_MINOR__SHIFT				16
+static inline uint32_t MDP5_MDP_VERSION_MINOR(uint32_t val)
+{
+	return ((val) << MDP5_MDP_VERSION_MINOR__SHIFT) & MDP5_MDP_VERSION_MINOR__MASK;
+}
+#define MDP5_MDP_VERSION_MAJOR__MASK				0xf0000000
+#define MDP5_MDP_VERSION_MAJOR__SHIFT				28
+static inline uint32_t MDP5_MDP_VERSION_MAJOR(uint32_t val)
+{
+	return ((val) << MDP5_MDP_VERSION_MAJOR__SHIFT) & MDP5_MDP_VERSION_MAJOR__MASK;
+}
+
+#define REG_MDP5_DISP_INTF_SEL					0x00000104
+#define MDP5_DISP_INTF_SEL_INTF0__MASK				0x000000ff
+#define MDP5_DISP_INTF_SEL_INTF0__SHIFT				0
+static inline uint32_t MDP5_DISP_INTF_SEL_INTF0(enum mdp5_intf val)
+{
+	return ((val) << MDP5_DISP_INTF_SEL_INTF0__SHIFT) & MDP5_DISP_INTF_SEL_INTF0__MASK;
+}
+#define MDP5_DISP_INTF_SEL_INTF1__MASK				0x0000ff00
+#define MDP5_DISP_INTF_SEL_INTF1__SHIFT				8
+static inline uint32_t MDP5_DISP_INTF_SEL_INTF1(enum mdp5_intf val)
+{
+	return ((val) << MDP5_DISP_INTF_SEL_INTF1__SHIFT) & MDP5_DISP_INTF_SEL_INTF1__MASK;
+}
+#define MDP5_DISP_INTF_SEL_INTF2__MASK				0x00ff0000
+#define MDP5_DISP_INTF_SEL_INTF2__SHIFT				16
+static inline uint32_t MDP5_DISP_INTF_SEL_INTF2(enum mdp5_intf val)
+{
+	return ((val) << MDP5_DISP_INTF_SEL_INTF2__SHIFT) & MDP5_DISP_INTF_SEL_INTF2__MASK;
+}
+#define MDP5_DISP_INTF_SEL_INTF3__MASK				0xff000000
+#define MDP5_DISP_INTF_SEL_INTF3__SHIFT				24
+static inline uint32_t MDP5_DISP_INTF_SEL_INTF3(enum mdp5_intf val)
+{
+	return ((val) << MDP5_DISP_INTF_SEL_INTF3__SHIFT) & MDP5_DISP_INTF_SEL_INTF3__MASK;
+}
+
+#define REG_MDP5_INTR_EN					0x00000110
+
+#define REG_MDP5_INTR_STATUS					0x00000114
+
+#define REG_MDP5_INTR_CLEAR					0x00000118
+
+#define REG_MDP5_HIST_INTR_EN					0x0000011c
+
+#define REG_MDP5_HIST_INTR_STATUS				0x00000120
+
+#define REG_MDP5_HIST_INTR_CLEAR				0x00000124
+
+static inline uint32_t REG_MDP5_SMP_ALLOC_W(uint32_t i0) { return 0x00000180 + 0x4*i0; }
+
+static inline uint32_t REG_MDP5_SMP_ALLOC_W_REG(uint32_t i0) { return 0x00000180 + 0x4*i0; }
+#define MDP5_SMP_ALLOC_W_REG_CLIENT0__MASK			0x000000ff
+#define MDP5_SMP_ALLOC_W_REG_CLIENT0__SHIFT			0
+static inline uint32_t MDP5_SMP_ALLOC_W_REG_CLIENT0(enum mdp5_client_id val)
+{
+	return ((val) << MDP5_SMP_ALLOC_W_REG_CLIENT0__SHIFT) & MDP5_SMP_ALLOC_W_REG_CLIENT0__MASK;
+}
+#define MDP5_SMP_ALLOC_W_REG_CLIENT1__MASK			0x0000ff00
+#define MDP5_SMP_ALLOC_W_REG_CLIENT1__SHIFT			8
+static inline uint32_t MDP5_SMP_ALLOC_W_REG_CLIENT1(enum mdp5_client_id val)
+{
+	return ((val) << MDP5_SMP_ALLOC_W_REG_CLIENT1__SHIFT) & MDP5_SMP_ALLOC_W_REG_CLIENT1__MASK;
+}
+#define MDP5_SMP_ALLOC_W_REG_CLIENT2__MASK			0x00ff0000
+#define MDP5_SMP_ALLOC_W_REG_CLIENT2__SHIFT			16
+static inline uint32_t MDP5_SMP_ALLOC_W_REG_CLIENT2(enum mdp5_client_id val)
+{
+	return ((val) << MDP5_SMP_ALLOC_W_REG_CLIENT2__SHIFT) & MDP5_SMP_ALLOC_W_REG_CLIENT2__MASK;
+}
+
+static inline uint32_t REG_MDP5_SMP_ALLOC_R(uint32_t i0) { return 0x00000230 + 0x4*i0; }
+
+static inline uint32_t REG_MDP5_SMP_ALLOC_R_REG(uint32_t i0) { return 0x00000230 + 0x4*i0; }
+#define MDP5_SMP_ALLOC_R_REG_CLIENT0__MASK			0x000000ff
+#define MDP5_SMP_ALLOC_R_REG_CLIENT0__SHIFT			0
+static inline uint32_t MDP5_SMP_ALLOC_R_REG_CLIENT0(enum mdp5_client_id val)
+{
+	return ((val) << MDP5_SMP_ALLOC_R_REG_CLIENT0__SHIFT) & MDP5_SMP_ALLOC_R_REG_CLIENT0__MASK;
+}
+#define MDP5_SMP_ALLOC_R_REG_CLIENT1__MASK			0x0000ff00
+#define MDP5_SMP_ALLOC_R_REG_CLIENT1__SHIFT			8
+static inline uint32_t MDP5_SMP_ALLOC_R_REG_CLIENT1(enum mdp5_client_id val)
+{
+	return ((val) << MDP5_SMP_ALLOC_R_REG_CLIENT1__SHIFT) & MDP5_SMP_ALLOC_R_REG_CLIENT1__MASK;
+}
+#define MDP5_SMP_ALLOC_R_REG_CLIENT2__MASK			0x00ff0000
+#define MDP5_SMP_ALLOC_R_REG_CLIENT2__SHIFT			16
+static inline uint32_t MDP5_SMP_ALLOC_R_REG_CLIENT2(enum mdp5_client_id val)
+{
+	return ((val) << MDP5_SMP_ALLOC_R_REG_CLIENT2__SHIFT) & MDP5_SMP_ALLOC_R_REG_CLIENT2__MASK;
+}
+
+static inline uint32_t __offset_IGC(enum mdp5_igc_type idx)
+{
+	switch (idx) {
+		case IGC_VIG: return 0x00000300;
+		case IGC_RGB: return 0x00000310;
+		case IGC_DMA: return 0x00000320;
+		case IGC_DSPP: return 0x00000400;
+		default: return INVALID_IDX(idx);
+	}
+}
+static inline uint32_t REG_MDP5_IGC(enum mdp5_igc_type i0) { return 0x00000000 + __offset_IGC(i0); }
+
+static inline uint32_t REG_MDP5_IGC_LUT(enum mdp5_igc_type i0, uint32_t i1) { return 0x00000000 + __offset_IGC(i0) + 0x4*i1; }
+
+static inline uint32_t REG_MDP5_IGC_LUT_REG(enum mdp5_igc_type i0, uint32_t i1) { return 0x00000000 + __offset_IGC(i0) + 0x4*i1; }
+#define MDP5_IGC_LUT_REG_VAL__MASK				0x00000fff
+#define MDP5_IGC_LUT_REG_VAL__SHIFT				0
+static inline uint32_t MDP5_IGC_LUT_REG_VAL(uint32_t val)
+{
+	return ((val) << MDP5_IGC_LUT_REG_VAL__SHIFT) & MDP5_IGC_LUT_REG_VAL__MASK;
+}
+#define MDP5_IGC_LUT_REG_INDEX_UPDATE				0x02000000
+#define MDP5_IGC_LUT_REG_DISABLE_PIPE_0				0x10000000
+#define MDP5_IGC_LUT_REG_DISABLE_PIPE_1				0x20000000
+#define MDP5_IGC_LUT_REG_DISABLE_PIPE_2				0x40000000
+
+static inline uint32_t REG_MDP5_CTL(uint32_t i0) { return 0x00000600 + 0x100*i0; }
+
+static inline uint32_t REG_MDP5_CTL_LAYER(uint32_t i0, uint32_t i1) { return 0x00000600 + 0x100*i0 + 0x4*i1; }
+
+static inline uint32_t REG_MDP5_CTL_LAYER_REG(uint32_t i0, uint32_t i1) { return 0x00000600 + 0x100*i0 + 0x4*i1; }
+#define MDP5_CTL_LAYER_REG_VIG0__MASK				0x00000007
+#define MDP5_CTL_LAYER_REG_VIG0__SHIFT				0
+static inline uint32_t MDP5_CTL_LAYER_REG_VIG0(enum mdp_mixer_stage_id val)
+{
+	return ((val) << MDP5_CTL_LAYER_REG_VIG0__SHIFT) & MDP5_CTL_LAYER_REG_VIG0__MASK;
+}
+#define MDP5_CTL_LAYER_REG_VIG1__MASK				0x00000038
+#define MDP5_CTL_LAYER_REG_VIG1__SHIFT				3
+static inline uint32_t MDP5_CTL_LAYER_REG_VIG1(enum mdp_mixer_stage_id val)
+{
+	return ((val) << MDP5_CTL_LAYER_REG_VIG1__SHIFT) & MDP5_CTL_LAYER_REG_VIG1__MASK;
+}
+#define MDP5_CTL_LAYER_REG_VIG2__MASK				0x000001c0
+#define MDP5_CTL_LAYER_REG_VIG2__SHIFT				6
+static inline uint32_t MDP5_CTL_LAYER_REG_VIG2(enum mdp_mixer_stage_id val)
+{
+	return ((val) << MDP5_CTL_LAYER_REG_VIG2__SHIFT) & MDP5_CTL_LAYER_REG_VIG2__MASK;
+}
+#define MDP5_CTL_LAYER_REG_RGB0__MASK				0x00000e00
+#define MDP5_CTL_LAYER_REG_RGB0__SHIFT				9
+static inline uint32_t MDP5_CTL_LAYER_REG_RGB0(enum mdp_mixer_stage_id val)
+{
+	return ((val) << MDP5_CTL_LAYER_REG_RGB0__SHIFT) & MDP5_CTL_LAYER_REG_RGB0__MASK;
+}
+#define MDP5_CTL_LAYER_REG_RGB1__MASK				0x00007000
+#define MDP5_CTL_LAYER_REG_RGB1__SHIFT				12
+static inline uint32_t MDP5_CTL_LAYER_REG_RGB1(enum mdp_mixer_stage_id val)
+{
+	return ((val) << MDP5_CTL_LAYER_REG_RGB1__SHIFT) & MDP5_CTL_LAYER_REG_RGB1__MASK;
+}
+#define MDP5_CTL_LAYER_REG_RGB2__MASK				0x00038000
+#define MDP5_CTL_LAYER_REG_RGB2__SHIFT				15
+static inline uint32_t MDP5_CTL_LAYER_REG_RGB2(enum mdp_mixer_stage_id val)
+{
+	return ((val) << MDP5_CTL_LAYER_REG_RGB2__SHIFT) & MDP5_CTL_LAYER_REG_RGB2__MASK;
+}
+#define MDP5_CTL_LAYER_REG_DMA0__MASK				0x001c0000
+#define MDP5_CTL_LAYER_REG_DMA0__SHIFT				18
+static inline uint32_t MDP5_CTL_LAYER_REG_DMA0(enum mdp_mixer_stage_id val)
+{
+	return ((val) << MDP5_CTL_LAYER_REG_DMA0__SHIFT) & MDP5_CTL_LAYER_REG_DMA0__MASK;
+}
+#define MDP5_CTL_LAYER_REG_DMA1__MASK				0x00e00000
+#define MDP5_CTL_LAYER_REG_DMA1__SHIFT				21
+static inline uint32_t MDP5_CTL_LAYER_REG_DMA1(enum mdp_mixer_stage_id val)
+{
+	return ((val) << MDP5_CTL_LAYER_REG_DMA1__SHIFT) & MDP5_CTL_LAYER_REG_DMA1__MASK;
+}
+#define MDP5_CTL_LAYER_REG_BORDER_COLOR				0x01000000
+#define MDP5_CTL_LAYER_REG_CURSOR_OUT				0x02000000
+
+static inline uint32_t REG_MDP5_CTL_OP(uint32_t i0) { return 0x00000614 + 0x100*i0; }
+#define MDP5_CTL_OP_MODE__MASK					0x0000000f
+#define MDP5_CTL_OP_MODE__SHIFT					0
+static inline uint32_t MDP5_CTL_OP_MODE(enum mdp5_ctl_mode val)
+{
+	return ((val) << MDP5_CTL_OP_MODE__SHIFT) & MDP5_CTL_OP_MODE__MASK;
+}
+#define MDP5_CTL_OP_INTF_NUM__MASK				0x00000070
+#define MDP5_CTL_OP_INTF_NUM__SHIFT				4
+static inline uint32_t MDP5_CTL_OP_INTF_NUM(enum mdp5_intfnum val)
+{
+	return ((val) << MDP5_CTL_OP_INTF_NUM__SHIFT) & MDP5_CTL_OP_INTF_NUM__MASK;
+}
+#define MDP5_CTL_OP_CMD_MODE					0x00020000
+#define MDP5_CTL_OP_PACK_3D_ENABLE				0x00080000
+#define MDP5_CTL_OP_PACK_3D__MASK				0x00300000
+#define MDP5_CTL_OP_PACK_3D__SHIFT				20
+static inline uint32_t MDP5_CTL_OP_PACK_3D(enum mdp5_pack_3d val)
+{
+	return ((val) << MDP5_CTL_OP_PACK_3D__SHIFT) & MDP5_CTL_OP_PACK_3D__MASK;
+}
+
+static inline uint32_t REG_MDP5_CTL_FLUSH(uint32_t i0) { return 0x00000618 + 0x100*i0; }
+#define MDP5_CTL_FLUSH_VIG0					0x00000001
+#define MDP5_CTL_FLUSH_VIG1					0x00000002
+#define MDP5_CTL_FLUSH_VIG2					0x00000004
+#define MDP5_CTL_FLUSH_RGB0					0x00000008
+#define MDP5_CTL_FLUSH_RGB1					0x00000010
+#define MDP5_CTL_FLUSH_RGB2					0x00000020
+#define MDP5_CTL_FLUSH_LM0					0x00000040
+#define MDP5_CTL_FLUSH_LM1					0x00000080
+#define MDP5_CTL_FLUSH_LM2					0x00000100
+#define MDP5_CTL_FLUSH_DMA0					0x00000800
+#define MDP5_CTL_FLUSH_DMA1					0x00001000
+#define MDP5_CTL_FLUSH_DSPP0					0x00002000
+#define MDP5_CTL_FLUSH_DSPP1					0x00004000
+#define MDP5_CTL_FLUSH_DSPP2					0x00008000
+#define MDP5_CTL_FLUSH_CTL					0x00020000
+
+static inline uint32_t REG_MDP5_CTL_START(uint32_t i0) { return 0x0000061c + 0x100*i0; }
+
+static inline uint32_t REG_MDP5_CTL_PACK_3D(uint32_t i0) { return 0x00000620 + 0x100*i0; }
+
+static inline uint32_t REG_MDP5_PIPE(enum mdp5_pipe i0) { return 0x00001200 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_HIST_CTL_BASE(enum mdp5_pipe i0) { return 0x000014c4 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_HIST_LUT_BASE(enum mdp5_pipe i0) { return 0x000014f0 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_HIST_LUT_SWAP(enum mdp5_pipe i0) { return 0x00001500 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_SRC_SIZE(enum mdp5_pipe i0) { return 0x00001200 + 0x400*i0; }
+#define MDP5_PIPE_SRC_SIZE_HEIGHT__MASK				0xffff0000
+#define MDP5_PIPE_SRC_SIZE_HEIGHT__SHIFT			16
+static inline uint32_t MDP5_PIPE_SRC_SIZE_HEIGHT(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_SIZE_HEIGHT__SHIFT) & MDP5_PIPE_SRC_SIZE_HEIGHT__MASK;
+}
+#define MDP5_PIPE_SRC_SIZE_WIDTH__MASK				0x0000ffff
+#define MDP5_PIPE_SRC_SIZE_WIDTH__SHIFT				0
+static inline uint32_t MDP5_PIPE_SRC_SIZE_WIDTH(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_SIZE_WIDTH__SHIFT) & MDP5_PIPE_SRC_SIZE_WIDTH__MASK;
+}
+
+static inline uint32_t REG_MDP5_PIPE_SRC_IMG_SIZE(enum mdp5_pipe i0) { return 0x00001204 + 0x400*i0; }
+#define MDP5_PIPE_SRC_IMG_SIZE_HEIGHT__MASK			0xffff0000
+#define MDP5_PIPE_SRC_IMG_SIZE_HEIGHT__SHIFT			16
+static inline uint32_t MDP5_PIPE_SRC_IMG_SIZE_HEIGHT(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_IMG_SIZE_HEIGHT__SHIFT) & MDP5_PIPE_SRC_IMG_SIZE_HEIGHT__MASK;
+}
+#define MDP5_PIPE_SRC_IMG_SIZE_WIDTH__MASK			0x0000ffff
+#define MDP5_PIPE_SRC_IMG_SIZE_WIDTH__SHIFT			0
+static inline uint32_t MDP5_PIPE_SRC_IMG_SIZE_WIDTH(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_IMG_SIZE_WIDTH__SHIFT) & MDP5_PIPE_SRC_IMG_SIZE_WIDTH__MASK;
+}
+
+static inline uint32_t REG_MDP5_PIPE_SRC_XY(enum mdp5_pipe i0) { return 0x00001208 + 0x400*i0; }
+#define MDP5_PIPE_SRC_XY_Y__MASK				0xffff0000
+#define MDP5_PIPE_SRC_XY_Y__SHIFT				16
+static inline uint32_t MDP5_PIPE_SRC_XY_Y(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_XY_Y__SHIFT) & MDP5_PIPE_SRC_XY_Y__MASK;
+}
+#define MDP5_PIPE_SRC_XY_X__MASK				0x0000ffff
+#define MDP5_PIPE_SRC_XY_X__SHIFT				0
+static inline uint32_t MDP5_PIPE_SRC_XY_X(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_XY_X__SHIFT) & MDP5_PIPE_SRC_XY_X__MASK;
+}
+
+static inline uint32_t REG_MDP5_PIPE_OUT_SIZE(enum mdp5_pipe i0) { return 0x0000120c + 0x400*i0; }
+#define MDP5_PIPE_OUT_SIZE_HEIGHT__MASK				0xffff0000
+#define MDP5_PIPE_OUT_SIZE_HEIGHT__SHIFT			16
+static inline uint32_t MDP5_PIPE_OUT_SIZE_HEIGHT(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_OUT_SIZE_HEIGHT__SHIFT) & MDP5_PIPE_OUT_SIZE_HEIGHT__MASK;
+}
+#define MDP5_PIPE_OUT_SIZE_WIDTH__MASK				0x0000ffff
+#define MDP5_PIPE_OUT_SIZE_WIDTH__SHIFT				0
+static inline uint32_t MDP5_PIPE_OUT_SIZE_WIDTH(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_OUT_SIZE_WIDTH__SHIFT) & MDP5_PIPE_OUT_SIZE_WIDTH__MASK;
+}
+
+static inline uint32_t REG_MDP5_PIPE_OUT_XY(enum mdp5_pipe i0) { return 0x00001210 + 0x400*i0; }
+#define MDP5_PIPE_OUT_XY_Y__MASK				0xffff0000
+#define MDP5_PIPE_OUT_XY_Y__SHIFT				16
+static inline uint32_t MDP5_PIPE_OUT_XY_Y(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_OUT_XY_Y__SHIFT) & MDP5_PIPE_OUT_XY_Y__MASK;
+}
+#define MDP5_PIPE_OUT_XY_X__MASK				0x0000ffff
+#define MDP5_PIPE_OUT_XY_X__SHIFT				0
+static inline uint32_t MDP5_PIPE_OUT_XY_X(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_OUT_XY_X__SHIFT) & MDP5_PIPE_OUT_XY_X__MASK;
+}
+
+static inline uint32_t REG_MDP5_PIPE_SRC0_ADDR(enum mdp5_pipe i0) { return 0x00001214 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_SRC1_ADDR(enum mdp5_pipe i0) { return 0x00001218 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_SRC2_ADDR(enum mdp5_pipe i0) { return 0x0000121c + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_SRC3_ADDR(enum mdp5_pipe i0) { return 0x00001220 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_SRC_STRIDE_A(enum mdp5_pipe i0) { return 0x00001224 + 0x400*i0; }
+#define MDP5_PIPE_SRC_STRIDE_A_P0__MASK				0x0000ffff
+#define MDP5_PIPE_SRC_STRIDE_A_P0__SHIFT			0
+static inline uint32_t MDP5_PIPE_SRC_STRIDE_A_P0(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_STRIDE_A_P0__SHIFT) & MDP5_PIPE_SRC_STRIDE_A_P0__MASK;
+}
+#define MDP5_PIPE_SRC_STRIDE_A_P1__MASK				0xffff0000
+#define MDP5_PIPE_SRC_STRIDE_A_P1__SHIFT			16
+static inline uint32_t MDP5_PIPE_SRC_STRIDE_A_P1(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_STRIDE_A_P1__SHIFT) & MDP5_PIPE_SRC_STRIDE_A_P1__MASK;
+}
+
+static inline uint32_t REG_MDP5_PIPE_SRC_STRIDE_B(enum mdp5_pipe i0) { return 0x00001228 + 0x400*i0; }
+#define MDP5_PIPE_SRC_STRIDE_B_P2__MASK				0x0000ffff
+#define MDP5_PIPE_SRC_STRIDE_B_P2__SHIFT			0
+static inline uint32_t MDP5_PIPE_SRC_STRIDE_B_P2(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_STRIDE_B_P2__SHIFT) & MDP5_PIPE_SRC_STRIDE_B_P2__MASK;
+}
+#define MDP5_PIPE_SRC_STRIDE_B_P3__MASK				0xffff0000
+#define MDP5_PIPE_SRC_STRIDE_B_P3__SHIFT			16
+static inline uint32_t MDP5_PIPE_SRC_STRIDE_B_P3(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_STRIDE_B_P3__SHIFT) & MDP5_PIPE_SRC_STRIDE_B_P3__MASK;
+}
+
+static inline uint32_t REG_MDP5_PIPE_STILE_FRAME_SIZE(enum mdp5_pipe i0) { return 0x0000122c + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_SRC_FORMAT(enum mdp5_pipe i0) { return 0x00001230 + 0x400*i0; }
+#define MDP5_PIPE_SRC_FORMAT_G_BPC__MASK			0x00000003
+#define MDP5_PIPE_SRC_FORMAT_G_BPC__SHIFT			0
+static inline uint32_t MDP5_PIPE_SRC_FORMAT_G_BPC(enum mdp_bpc val)
+{
+	return ((val) << MDP5_PIPE_SRC_FORMAT_G_BPC__SHIFT) & MDP5_PIPE_SRC_FORMAT_G_BPC__MASK;
+}
+#define MDP5_PIPE_SRC_FORMAT_B_BPC__MASK			0x0000000c
+#define MDP5_PIPE_SRC_FORMAT_B_BPC__SHIFT			2
+static inline uint32_t MDP5_PIPE_SRC_FORMAT_B_BPC(enum mdp_bpc val)
+{
+	return ((val) << MDP5_PIPE_SRC_FORMAT_B_BPC__SHIFT) & MDP5_PIPE_SRC_FORMAT_B_BPC__MASK;
+}
+#define MDP5_PIPE_SRC_FORMAT_R_BPC__MASK			0x00000030
+#define MDP5_PIPE_SRC_FORMAT_R_BPC__SHIFT			4
+static inline uint32_t MDP5_PIPE_SRC_FORMAT_R_BPC(enum mdp_bpc val)
+{
+	return ((val) << MDP5_PIPE_SRC_FORMAT_R_BPC__SHIFT) & MDP5_PIPE_SRC_FORMAT_R_BPC__MASK;
+}
+#define MDP5_PIPE_SRC_FORMAT_A_BPC__MASK			0x000000c0
+#define MDP5_PIPE_SRC_FORMAT_A_BPC__SHIFT			6
+static inline uint32_t MDP5_PIPE_SRC_FORMAT_A_BPC(enum mdp_bpc_alpha val)
+{
+	return ((val) << MDP5_PIPE_SRC_FORMAT_A_BPC__SHIFT) & MDP5_PIPE_SRC_FORMAT_A_BPC__MASK;
+}
+#define MDP5_PIPE_SRC_FORMAT_ALPHA_ENABLE			0x00000100
+#define MDP5_PIPE_SRC_FORMAT_CPP__MASK				0x00000600
+#define MDP5_PIPE_SRC_FORMAT_CPP__SHIFT				9
+static inline uint32_t MDP5_PIPE_SRC_FORMAT_CPP(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_FORMAT_CPP__SHIFT) & MDP5_PIPE_SRC_FORMAT_CPP__MASK;
+}
+#define MDP5_PIPE_SRC_FORMAT_ROT90				0x00000800
+#define MDP5_PIPE_SRC_FORMAT_UNPACK_COUNT__MASK			0x00003000
+#define MDP5_PIPE_SRC_FORMAT_UNPACK_COUNT__SHIFT		12
+static inline uint32_t MDP5_PIPE_SRC_FORMAT_UNPACK_COUNT(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_FORMAT_UNPACK_COUNT__SHIFT) & MDP5_PIPE_SRC_FORMAT_UNPACK_COUNT__MASK;
+}
+#define MDP5_PIPE_SRC_FORMAT_UNPACK_TIGHT			0x00020000
+#define MDP5_PIPE_SRC_FORMAT_UNPACK_ALIGN_MSB			0x00040000
+#define MDP5_PIPE_SRC_FORMAT_NUM_PLANES__MASK			0x00780000
+#define MDP5_PIPE_SRC_FORMAT_NUM_PLANES__SHIFT			19
+static inline uint32_t MDP5_PIPE_SRC_FORMAT_NUM_PLANES(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_FORMAT_NUM_PLANES__SHIFT) & MDP5_PIPE_SRC_FORMAT_NUM_PLANES__MASK;
+}
+#define MDP5_PIPE_SRC_FORMAT_CHROMA_SAMP__MASK			0x01800000
+#define MDP5_PIPE_SRC_FORMAT_CHROMA_SAMP__SHIFT			23
+static inline uint32_t MDP5_PIPE_SRC_FORMAT_CHROMA_SAMP(enum mdp5_chroma_samp_type val)
+{
+	return ((val) << MDP5_PIPE_SRC_FORMAT_CHROMA_SAMP__SHIFT) & MDP5_PIPE_SRC_FORMAT_CHROMA_SAMP__MASK;
+}
+
+static inline uint32_t REG_MDP5_PIPE_SRC_UNPACK(enum mdp5_pipe i0) { return 0x00001234 + 0x400*i0; }
+#define MDP5_PIPE_SRC_UNPACK_ELEM0__MASK			0x000000ff
+#define MDP5_PIPE_SRC_UNPACK_ELEM0__SHIFT			0
+static inline uint32_t MDP5_PIPE_SRC_UNPACK_ELEM0(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_UNPACK_ELEM0__SHIFT) & MDP5_PIPE_SRC_UNPACK_ELEM0__MASK;
+}
+#define MDP5_PIPE_SRC_UNPACK_ELEM1__MASK			0x0000ff00
+#define MDP5_PIPE_SRC_UNPACK_ELEM1__SHIFT			8
+static inline uint32_t MDP5_PIPE_SRC_UNPACK_ELEM1(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_UNPACK_ELEM1__SHIFT) & MDP5_PIPE_SRC_UNPACK_ELEM1__MASK;
+}
+#define MDP5_PIPE_SRC_UNPACK_ELEM2__MASK			0x00ff0000
+#define MDP5_PIPE_SRC_UNPACK_ELEM2__SHIFT			16
+static inline uint32_t MDP5_PIPE_SRC_UNPACK_ELEM2(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_UNPACK_ELEM2__SHIFT) & MDP5_PIPE_SRC_UNPACK_ELEM2__MASK;
+}
+#define MDP5_PIPE_SRC_UNPACK_ELEM3__MASK			0xff000000
+#define MDP5_PIPE_SRC_UNPACK_ELEM3__SHIFT			24
+static inline uint32_t MDP5_PIPE_SRC_UNPACK_ELEM3(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_SRC_UNPACK_ELEM3__SHIFT) & MDP5_PIPE_SRC_UNPACK_ELEM3__MASK;
+}
+
+static inline uint32_t REG_MDP5_PIPE_SRC_OP_MODE(enum mdp5_pipe i0) { return 0x00001238 + 0x400*i0; }
+#define MDP5_PIPE_SRC_OP_MODE_BWC_EN				0x00000001
+#define MDP5_PIPE_SRC_OP_MODE_BWC__MASK				0x00000006
+#define MDP5_PIPE_SRC_OP_MODE_BWC__SHIFT			1
+static inline uint32_t MDP5_PIPE_SRC_OP_MODE_BWC(enum mdp5_pipe_bwc val)
+{
+	return ((val) << MDP5_PIPE_SRC_OP_MODE_BWC__SHIFT) & MDP5_PIPE_SRC_OP_MODE_BWC__MASK;
+}
+#define MDP5_PIPE_SRC_OP_MODE_FLIP_LR				0x00002000
+#define MDP5_PIPE_SRC_OP_MODE_FLIP_UD				0x00004000
+#define MDP5_PIPE_SRC_OP_MODE_IGC_EN				0x00010000
+#define MDP5_PIPE_SRC_OP_MODE_IGC_ROM_0				0x00020000
+#define MDP5_PIPE_SRC_OP_MODE_IGC_ROM_1				0x00040000
+#define MDP5_PIPE_SRC_OP_MODE_DEINTERLACE			0x00400000
+#define MDP5_PIPE_SRC_OP_MODE_DEINTERLACE_ODD			0x00800000
+
+static inline uint32_t REG_MDP5_PIPE_SRC_CONSTANT_COLOR(enum mdp5_pipe i0) { return 0x0000123c + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_FETCH_CONFIG(enum mdp5_pipe i0) { return 0x00001248 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_VC1_RANGE(enum mdp5_pipe i0) { return 0x0000124c + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_REQPRIO_FIFO_WM_0(enum mdp5_pipe i0) { return 0x00001250 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_REQPRIO_FIFO_WM_1(enum mdp5_pipe i0) { return 0x00001254 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_REQPRIO_FIFO_WM_2(enum mdp5_pipe i0) { return 0x00001258 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_SRC_ADDR_SW_STATUS(enum mdp5_pipe i0) { return 0x00001270 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_CURRENT_SRC0_ADDR(enum mdp5_pipe i0) { return 0x000012a4 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_CURRENT_SRC1_ADDR(enum mdp5_pipe i0) { return 0x000012a8 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_CURRENT_SRC2_ADDR(enum mdp5_pipe i0) { return 0x000012ac + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_CURRENT_SRC3_ADDR(enum mdp5_pipe i0) { return 0x000012b0 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_DECIMATION(enum mdp5_pipe i0) { return 0x000012b4 + 0x400*i0; }
+#define MDP5_PIPE_DECIMATION_VERT__MASK				0x000000ff
+#define MDP5_PIPE_DECIMATION_VERT__SHIFT			0
+static inline uint32_t MDP5_PIPE_DECIMATION_VERT(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_DECIMATION_VERT__SHIFT) & MDP5_PIPE_DECIMATION_VERT__MASK;
+}
+#define MDP5_PIPE_DECIMATION_HORZ__MASK				0x0000ff00
+#define MDP5_PIPE_DECIMATION_HORZ__SHIFT			8
+static inline uint32_t MDP5_PIPE_DECIMATION_HORZ(uint32_t val)
+{
+	return ((val) << MDP5_PIPE_DECIMATION_HORZ__SHIFT) & MDP5_PIPE_DECIMATION_HORZ__MASK;
+}
+
+static inline uint32_t REG_MDP5_PIPE_SCALE_CONFIG(enum mdp5_pipe i0) { return 0x00001404 + 0x400*i0; }
+#define MDP5_PIPE_SCALE_CONFIG_SCALEX_EN			0x00000001
+#define MDP5_PIPE_SCALE_CONFIG_SCALEY_EN			0x00000002
+#define MDP5_PIPE_SCALE_CONFIG_SCALEX_MIN_FILTER__MASK		0x00000300
+#define MDP5_PIPE_SCALE_CONFIG_SCALEX_MIN_FILTER__SHIFT		8
+static inline uint32_t MDP5_PIPE_SCALE_CONFIG_SCALEX_MIN_FILTER(enum mdp5_scale_filter val)
+{
+	return ((val) << MDP5_PIPE_SCALE_CONFIG_SCALEX_MIN_FILTER__SHIFT) & MDP5_PIPE_SCALE_CONFIG_SCALEX_MIN_FILTER__MASK;
+}
+#define MDP5_PIPE_SCALE_CONFIG_SCALEY_MIN_FILTER__MASK		0x00000c00
+#define MDP5_PIPE_SCALE_CONFIG_SCALEY_MIN_FILTER__SHIFT		10
+static inline uint32_t MDP5_PIPE_SCALE_CONFIG_SCALEY_MIN_FILTER(enum mdp5_scale_filter val)
+{
+	return ((val) << MDP5_PIPE_SCALE_CONFIG_SCALEY_MIN_FILTER__SHIFT) & MDP5_PIPE_SCALE_CONFIG_SCALEY_MIN_FILTER__MASK;
+}
+#define MDP5_PIPE_SCALE_CONFIG_SCALEX_CR_FILTER__MASK		0x00003000
+#define MDP5_PIPE_SCALE_CONFIG_SCALEX_CR_FILTER__SHIFT		12
+static inline uint32_t MDP5_PIPE_SCALE_CONFIG_SCALEX_CR_FILTER(enum mdp5_scale_filter val)
+{
+	return ((val) << MDP5_PIPE_SCALE_CONFIG_SCALEX_CR_FILTER__SHIFT) & MDP5_PIPE_SCALE_CONFIG_SCALEX_CR_FILTER__MASK;
+}
+#define MDP5_PIPE_SCALE_CONFIG_SCALEY_CR_FILTER__MASK		0x0000c000
+#define MDP5_PIPE_SCALE_CONFIG_SCALEY_CR_FILTER__SHIFT		14
+static inline uint32_t MDP5_PIPE_SCALE_CONFIG_SCALEY_CR_FILTER(enum mdp5_scale_filter val)
+{
+	return ((val) << MDP5_PIPE_SCALE_CONFIG_SCALEY_CR_FILTER__SHIFT) & MDP5_PIPE_SCALE_CONFIG_SCALEY_CR_FILTER__MASK;
+}
+#define MDP5_PIPE_SCALE_CONFIG_SCALEX_MAX_FILTER__MASK		0x00030000
+#define MDP5_PIPE_SCALE_CONFIG_SCALEX_MAX_FILTER__SHIFT		16
+static inline uint32_t MDP5_PIPE_SCALE_CONFIG_SCALEX_MAX_FILTER(enum mdp5_scale_filter val)
+{
+	return ((val) << MDP5_PIPE_SCALE_CONFIG_SCALEX_MAX_FILTER__SHIFT) & MDP5_PIPE_SCALE_CONFIG_SCALEX_MAX_FILTER__MASK;
+}
+#define MDP5_PIPE_SCALE_CONFIG_SCALEY_MAX_FILTER__MASK		0x000c0000
+#define MDP5_PIPE_SCALE_CONFIG_SCALEY_MAX_FILTER__SHIFT		18
+static inline uint32_t MDP5_PIPE_SCALE_CONFIG_SCALEY_MAX_FILTER(enum mdp5_scale_filter val)
+{
+	return ((val) << MDP5_PIPE_SCALE_CONFIG_SCALEY_MAX_FILTER__SHIFT) & MDP5_PIPE_SCALE_CONFIG_SCALEY_MAX_FILTER__MASK;
+}
+
+static inline uint32_t REG_MDP5_PIPE_SCALE_PHASE_STEP_X(enum mdp5_pipe i0) { return 0x00001410 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_SCALE_PHASE_STEP_Y(enum mdp5_pipe i0) { return 0x00001414 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_SCALE_INIT_PHASE_X(enum mdp5_pipe i0) { return 0x00001420 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_PIPE_SCALE_INIT_PHASE_Y(enum mdp5_pipe i0) { return 0x00001424 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM(uint32_t i0) { return 0x00003200 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_BLEND_COLOR_OUT(uint32_t i0) { return 0x00003200 + 0x400*i0; }
+#define MDP5_LM_BLEND_COLOR_OUT_STAGE0_FG_ALPHA			0x00000002
+#define MDP5_LM_BLEND_COLOR_OUT_STAGE1_FG_ALPHA			0x00000004
+#define MDP5_LM_BLEND_COLOR_OUT_STAGE2_FG_ALPHA			0x00000008
+#define MDP5_LM_BLEND_COLOR_OUT_STAGE3_FG_ALPHA			0x00000010
+
+static inline uint32_t REG_MDP5_LM_OUT_SIZE(uint32_t i0) { return 0x00003204 + 0x400*i0; }
+#define MDP5_LM_OUT_SIZE_HEIGHT__MASK				0xffff0000
+#define MDP5_LM_OUT_SIZE_HEIGHT__SHIFT				16
+static inline uint32_t MDP5_LM_OUT_SIZE_HEIGHT(uint32_t val)
+{
+	return ((val) << MDP5_LM_OUT_SIZE_HEIGHT__SHIFT) & MDP5_LM_OUT_SIZE_HEIGHT__MASK;
+}
+#define MDP5_LM_OUT_SIZE_WIDTH__MASK				0x0000ffff
+#define MDP5_LM_OUT_SIZE_WIDTH__SHIFT				0
+static inline uint32_t MDP5_LM_OUT_SIZE_WIDTH(uint32_t val)
+{
+	return ((val) << MDP5_LM_OUT_SIZE_WIDTH__SHIFT) & MDP5_LM_OUT_SIZE_WIDTH__MASK;
+}
+
+static inline uint32_t REG_MDP5_LM_BORDER_COLOR_0(uint32_t i0) { return 0x00003208 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_BORDER_COLOR_1(uint32_t i0) { return 0x00003210 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_BLEND(uint32_t i0, uint32_t i1) { return 0x00003220 + 0x400*i0 + 0x30*i1; }
+
+static inline uint32_t REG_MDP5_LM_BLEND_OP_MODE(uint32_t i0, uint32_t i1) { return 0x00003220 + 0x400*i0 + 0x30*i1; }
+#define MDP5_LM_BLEND_OP_MODE_FG_ALPHA__MASK			0x00000003
+#define MDP5_LM_BLEND_OP_MODE_FG_ALPHA__SHIFT			0
+static inline uint32_t MDP5_LM_BLEND_OP_MODE_FG_ALPHA(enum mdp_alpha_type val)
+{
+	return ((val) << MDP5_LM_BLEND_OP_MODE_FG_ALPHA__SHIFT) & MDP5_LM_BLEND_OP_MODE_FG_ALPHA__MASK;
+}
+#define MDP5_LM_BLEND_OP_MODE_FG_INV_ALPHA			0x00000004
+#define MDP5_LM_BLEND_OP_MODE_FG_MOD_ALPHA			0x00000008
+#define MDP5_LM_BLEND_OP_MODE_FG_INV_MOD_ALPHA			0x00000010
+#define MDP5_LM_BLEND_OP_MODE_FG_TRANSP_EN			0x00000020
+#define MDP5_LM_BLEND_OP_MODE_BG_ALPHA__MASK			0x00000300
+#define MDP5_LM_BLEND_OP_MODE_BG_ALPHA__SHIFT			8
+static inline uint32_t MDP5_LM_BLEND_OP_MODE_BG_ALPHA(enum mdp_alpha_type val)
+{
+	return ((val) << MDP5_LM_BLEND_OP_MODE_BG_ALPHA__SHIFT) & MDP5_LM_BLEND_OP_MODE_BG_ALPHA__MASK;
+}
+#define MDP5_LM_BLEND_OP_MODE_BG_INV_ALPHA			0x00000400
+#define MDP5_LM_BLEND_OP_MODE_BG_MOD_ALPHA			0x00000800
+#define MDP5_LM_BLEND_OP_MODE_BG_INV_MOD_ALPHA			0x00001000
+#define MDP5_LM_BLEND_OP_MODE_BG_TRANSP_EN			0x00002000
+
+static inline uint32_t REG_MDP5_LM_BLEND_FG_ALPHA(uint32_t i0, uint32_t i1) { return 0x00003224 + 0x400*i0 + 0x30*i1; }
+
+static inline uint32_t REG_MDP5_LM_BLEND_BG_ALPHA(uint32_t i0, uint32_t i1) { return 0x00003228 + 0x400*i0 + 0x30*i1; }
+
+static inline uint32_t REG_MDP5_LM_BLEND_FG_TRANSP_LOW0(uint32_t i0, uint32_t i1) { return 0x0000322c + 0x400*i0 + 0x30*i1; }
+
+static inline uint32_t REG_MDP5_LM_BLEND_FG_TRANSP_LOW1(uint32_t i0, uint32_t i1) { return 0x00003230 + 0x400*i0 + 0x30*i1; }
+
+static inline uint32_t REG_MDP5_LM_BLEND_FG_TRANSP_HIGH0(uint32_t i0, uint32_t i1) { return 0x00003234 + 0x400*i0 + 0x30*i1; }
+
+static inline uint32_t REG_MDP5_LM_BLEND_FG_TRANSP_HIGH1(uint32_t i0, uint32_t i1) { return 0x00003238 + 0x400*i0 + 0x30*i1; }
+
+static inline uint32_t REG_MDP5_LM_BLEND_BG_TRANSP_LOW0(uint32_t i0, uint32_t i1) { return 0x0000323c + 0x400*i0 + 0x30*i1; }
+
+static inline uint32_t REG_MDP5_LM_BLEND_BG_TRANSP_LOW1(uint32_t i0, uint32_t i1) { return 0x00003240 + 0x400*i0 + 0x30*i1; }
+
+static inline uint32_t REG_MDP5_LM_BLEND_BG_TRANSP_HIGH0(uint32_t i0, uint32_t i1) { return 0x00003244 + 0x400*i0 + 0x30*i1; }
+
+static inline uint32_t REG_MDP5_LM_BLEND_BG_TRANSP_HIGH1(uint32_t i0, uint32_t i1) { return 0x00003248 + 0x400*i0 + 0x30*i1; }
+
+static inline uint32_t REG_MDP5_LM_CURSOR_IMG_SIZE(uint32_t i0) { return 0x000032e0 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_CURSOR_SIZE(uint32_t i0) { return 0x000032e4 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_CURSOR_XY(uint32_t i0) { return 0x000032e8 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_CURSOR_STRIDE(uint32_t i0) { return 0x000032dc + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_CURSOR_FORMAT(uint32_t i0) { return 0x000032ec + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_CURSOR_BASE_ADDR(uint32_t i0) { return 0x000032f0 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_CURSOR_START_XY(uint32_t i0) { return 0x000032f4 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_CURSOR_BLEND_CONFIG(uint32_t i0) { return 0x000032f8 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_CURSOR_BLEND_PARAM(uint32_t i0) { return 0x000032fc + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_CURSOR_BLEND_TRANSP_LOW0(uint32_t i0) { return 0x00003300 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_CURSOR_BLEND_TRANSP_LOW1(uint32_t i0) { return 0x00003304 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_CURSOR_BLEND_TRANSP_HIGH0(uint32_t i0) { return 0x00003308 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_CURSOR_BLEND_TRANSP_HIGH1(uint32_t i0) { return 0x0000330c + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_LM_GC_LUT_BASE(uint32_t i0) { return 0x00003310 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_DSPP(uint32_t i0) { return 0x00004600 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_DSPP_OP_MODE(uint32_t i0) { return 0x00004600 + 0x400*i0; }
+#define MDP5_DSPP_OP_MODE_IGC_LUT_EN				0x00000001
+#define MDP5_DSPP_OP_MODE_IGC_TBL_IDX__MASK			0x0000000e
+#define MDP5_DSPP_OP_MODE_IGC_TBL_IDX__SHIFT			1
+static inline uint32_t MDP5_DSPP_OP_MODE_IGC_TBL_IDX(uint32_t val)
+{
+	return ((val) << MDP5_DSPP_OP_MODE_IGC_TBL_IDX__SHIFT) & MDP5_DSPP_OP_MODE_IGC_TBL_IDX__MASK;
+}
+#define MDP5_DSPP_OP_MODE_PCC_EN				0x00000010
+#define MDP5_DSPP_OP_MODE_DITHER_EN				0x00000100
+#define MDP5_DSPP_OP_MODE_HIST_EN				0x00010000
+#define MDP5_DSPP_OP_MODE_AUTO_CLEAR				0x00020000
+#define MDP5_DSPP_OP_MODE_HIST_LUT_EN				0x00080000
+#define MDP5_DSPP_OP_MODE_PA_EN					0x00100000
+#define MDP5_DSPP_OP_MODE_GAMUT_EN				0x00800000
+#define MDP5_DSPP_OP_MODE_GAMUT_ORDER				0x01000000
+
+static inline uint32_t REG_MDP5_DSPP_PCC_BASE(uint32_t i0) { return 0x00004630 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_DSPP_DITHER_DEPTH(uint32_t i0) { return 0x00004750 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_DSPP_HIST_CTL_BASE(uint32_t i0) { return 0x00004810 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_DSPP_HIST_LUT_BASE(uint32_t i0) { return 0x00004830 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_DSPP_HIST_LUT_SWAP(uint32_t i0) { return 0x00004834 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_DSPP_PA_BASE(uint32_t i0) { return 0x00004838 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_DSPP_GAMUT_BASE(uint32_t i0) { return 0x000048dc + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_DSPP_GC_BASE(uint32_t i0) { return 0x000048b0 + 0x400*i0; }
+
+static inline uint32_t REG_MDP5_INTF(uint32_t i0) { return 0x00012500 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_TIMING_ENGINE_EN(uint32_t i0) { return 0x00012500 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_CONFIG(uint32_t i0) { return 0x00012504 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_HSYNC_CTL(uint32_t i0) { return 0x00012508 + 0x200*i0; }
+#define MDP5_INTF_HSYNC_CTL_PULSEW__MASK			0x0000ffff
+#define MDP5_INTF_HSYNC_CTL_PULSEW__SHIFT			0
+static inline uint32_t MDP5_INTF_HSYNC_CTL_PULSEW(uint32_t val)
+{
+	return ((val) << MDP5_INTF_HSYNC_CTL_PULSEW__SHIFT) & MDP5_INTF_HSYNC_CTL_PULSEW__MASK;
+}
+#define MDP5_INTF_HSYNC_CTL_PERIOD__MASK			0xffff0000
+#define MDP5_INTF_HSYNC_CTL_PERIOD__SHIFT			16
+static inline uint32_t MDP5_INTF_HSYNC_CTL_PERIOD(uint32_t val)
+{
+	return ((val) << MDP5_INTF_HSYNC_CTL_PERIOD__SHIFT) & MDP5_INTF_HSYNC_CTL_PERIOD__MASK;
+}
+
+static inline uint32_t REG_MDP5_INTF_VSYNC_PERIOD_F0(uint32_t i0) { return 0x0001250c + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_VSYNC_PERIOD_F1(uint32_t i0) { return 0x00012510 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_VSYNC_LEN_F0(uint32_t i0) { return 0x00012514 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_VSYNC_LEN_F1(uint32_t i0) { return 0x00012518 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_DISPLAY_VSTART_F0(uint32_t i0) { return 0x0001251c + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_DISPLAY_VSTART_F1(uint32_t i0) { return 0x00012520 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_DISPLAY_VEND_F0(uint32_t i0) { return 0x00012524 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_DISPLAY_VEND_F1(uint32_t i0) { return 0x00012528 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_ACTIVE_VSTART_F0(uint32_t i0) { return 0x0001252c + 0x200*i0; }
+#define MDP5_INTF_ACTIVE_VSTART_F0_VAL__MASK			0x7fffffff
+#define MDP5_INTF_ACTIVE_VSTART_F0_VAL__SHIFT			0
+static inline uint32_t MDP5_INTF_ACTIVE_VSTART_F0_VAL(uint32_t val)
+{
+	return ((val) << MDP5_INTF_ACTIVE_VSTART_F0_VAL__SHIFT) & MDP5_INTF_ACTIVE_VSTART_F0_VAL__MASK;
+}
+#define MDP5_INTF_ACTIVE_VSTART_F0_ACTIVE_V_ENABLE		0x80000000
+
+static inline uint32_t REG_MDP5_INTF_ACTIVE_VSTART_F1(uint32_t i0) { return 0x00012530 + 0x200*i0; }
+#define MDP5_INTF_ACTIVE_VSTART_F1_VAL__MASK			0x7fffffff
+#define MDP5_INTF_ACTIVE_VSTART_F1_VAL__SHIFT			0
+static inline uint32_t MDP5_INTF_ACTIVE_VSTART_F1_VAL(uint32_t val)
+{
+	return ((val) << MDP5_INTF_ACTIVE_VSTART_F1_VAL__SHIFT) & MDP5_INTF_ACTIVE_VSTART_F1_VAL__MASK;
+}
+
+static inline uint32_t REG_MDP5_INTF_ACTIVE_VEND_F0(uint32_t i0) { return 0x00012534 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_ACTIVE_VEND_F1(uint32_t i0) { return 0x00012538 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_DISPLAY_HCTL(uint32_t i0) { return 0x0001253c + 0x200*i0; }
+#define MDP5_INTF_DISPLAY_HCTL_START__MASK			0x0000ffff
+#define MDP5_INTF_DISPLAY_HCTL_START__SHIFT			0
+static inline uint32_t MDP5_INTF_DISPLAY_HCTL_START(uint32_t val)
+{
+	return ((val) << MDP5_INTF_DISPLAY_HCTL_START__SHIFT) & MDP5_INTF_DISPLAY_HCTL_START__MASK;
+}
+#define MDP5_INTF_DISPLAY_HCTL_END__MASK			0xffff0000
+#define MDP5_INTF_DISPLAY_HCTL_END__SHIFT			16
+static inline uint32_t MDP5_INTF_DISPLAY_HCTL_END(uint32_t val)
+{
+	return ((val) << MDP5_INTF_DISPLAY_HCTL_END__SHIFT) & MDP5_INTF_DISPLAY_HCTL_END__MASK;
+}
+
+static inline uint32_t REG_MDP5_INTF_ACTIVE_HCTL(uint32_t i0) { return 0x00012540 + 0x200*i0; }
+#define MDP5_INTF_ACTIVE_HCTL_START__MASK			0x00007fff
+#define MDP5_INTF_ACTIVE_HCTL_START__SHIFT			0
+static inline uint32_t MDP5_INTF_ACTIVE_HCTL_START(uint32_t val)
+{
+	return ((val) << MDP5_INTF_ACTIVE_HCTL_START__SHIFT) & MDP5_INTF_ACTIVE_HCTL_START__MASK;
+}
+#define MDP5_INTF_ACTIVE_HCTL_END__MASK				0x7fff0000
+#define MDP5_INTF_ACTIVE_HCTL_END__SHIFT			16
+static inline uint32_t MDP5_INTF_ACTIVE_HCTL_END(uint32_t val)
+{
+	return ((val) << MDP5_INTF_ACTIVE_HCTL_END__SHIFT) & MDP5_INTF_ACTIVE_HCTL_END__MASK;
+}
+#define MDP5_INTF_ACTIVE_HCTL_ACTIVE_H_ENABLE			0x80000000
+
+static inline uint32_t REG_MDP5_INTF_BORDER_COLOR(uint32_t i0) { return 0x00012544 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_UNDERFLOW_COLOR(uint32_t i0) { return 0x00012548 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_HSYNC_SKEW(uint32_t i0) { return 0x0001254c + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_POLARITY_CTL(uint32_t i0) { return 0x00012550 + 0x200*i0; }
+#define MDP5_INTF_POLARITY_CTL_HSYNC_LOW			0x00000001
+#define MDP5_INTF_POLARITY_CTL_VSYNC_LOW			0x00000002
+#define MDP5_INTF_POLARITY_CTL_DATA_EN_LOW			0x00000004
+
+static inline uint32_t REG_MDP5_INTF_TEST_CTL(uint32_t i0) { return 0x00012554 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_TP_COLOR0(uint32_t i0) { return 0x00012558 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_TP_COLOR1(uint32_t i0) { return 0x0001255c + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_DSI_CMD_MODE_TRIGGER_EN(uint32_t i0) { return 0x00012584 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_PANEL_FORMAT(uint32_t i0) { return 0x00012590 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_FRAME_LINE_COUNT_EN(uint32_t i0) { return 0x000125a8 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_FRAME_COUNT(uint32_t i0) { return 0x000125ac + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_LINE_COUNT(uint32_t i0) { return 0x000125b0 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_DEFLICKER_CONFIG(uint32_t i0) { return 0x000125f0 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_DEFLICKER_STRNG_COEFF(uint32_t i0) { return 0x000125f4 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_DEFLICKER_WEAK_COEFF(uint32_t i0) { return 0x000125f8 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_TPG_ENABLE(uint32_t i0) { return 0x00012600 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_TPG_MAIN_CONTROL(uint32_t i0) { return 0x00012604 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_TPG_VIDEO_CONFIG(uint32_t i0) { return 0x00012608 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_TPG_COMPONENT_LIMITS(uint32_t i0) { return 0x0001260c + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_TPG_RECTANGLE(uint32_t i0) { return 0x00012610 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_TPG_INITIAL_VALUE(uint32_t i0) { return 0x00012614 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_TPG_BLK_WHITE_PATTERN_FRAME(uint32_t i0) { return 0x00012618 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_INTF_TPG_RGB_MAPPING(uint32_t i0) { return 0x0001261c + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD(uint32_t i0) { return 0x00013100 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_BYPASS(uint32_t i0) { return 0x00013100 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_CTRL_0(uint32_t i0) { return 0x00013104 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_CTRL_1(uint32_t i0) { return 0x00013108 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_FRAME_SIZE(uint32_t i0) { return 0x0001310c + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_CON_CTRL_0(uint32_t i0) { return 0x00013110 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_CON_CTRL_1(uint32_t i0) { return 0x00013114 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_STR_MAN(uint32_t i0) { return 0x00013118 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_VAR(uint32_t i0) { return 0x0001311c + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_DITH(uint32_t i0) { return 0x00013120 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_DITH_CTRL(uint32_t i0) { return 0x00013124 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_AMP_LIM(uint32_t i0) { return 0x00013128 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_SLOPE(uint32_t i0) { return 0x0001312c + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_BW_LVL(uint32_t i0) { return 0x00013130 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_LOGO_POS(uint32_t i0) { return 0x00013134 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_LUT_FI(uint32_t i0) { return 0x00013138 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_LUT_CC(uint32_t i0) { return 0x0001317c + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_STR_LIM(uint32_t i0) { return 0x000131c8 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_CALIB_AB(uint32_t i0) { return 0x000131cc + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_CALIB_CD(uint32_t i0) { return 0x000131d0 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_MODE_SEL(uint32_t i0) { return 0x000131d4 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_TFILT_CTRL(uint32_t i0) { return 0x000131d8 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_BL_MINMAX(uint32_t i0) { return 0x000131dc + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_BL(uint32_t i0) { return 0x000131e0 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_BL_MAX(uint32_t i0) { return 0x000131e8 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_AL(uint32_t i0) { return 0x000131ec + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_AL_MIN(uint32_t i0) { return 0x000131f0 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_AL_FILT(uint32_t i0) { return 0x000131f4 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_CFG_BUF(uint32_t i0) { return 0x000131f8 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_LUT_AL(uint32_t i0) { return 0x00013200 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_TARG_STR(uint32_t i0) { return 0x00013244 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_START_CALC(uint32_t i0) { return 0x00013248 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_STR_OUT(uint32_t i0) { return 0x0001324c + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_BL_OUT(uint32_t i0) { return 0x00013254 + 0x200*i0; }
+
+static inline uint32_t REG_MDP5_AD_CALC_DONE(uint32_t i0) { return 0x00013258 + 0x200*i0; }
+
+
+#endif /* MDP5_XML */
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c
new file mode 100644
index 000000000000..71a3b2345eb3
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c
@@ -0,0 +1,569 @@
+/*
+ * Copyright (C) 2013 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 "mdp5_kms.h"
+
+#include <drm/drm_mode.h>
+#include "drm_crtc.h"
+#include "drm_crtc_helper.h"
+#include "drm_flip_work.h"
+
+struct mdp5_crtc {
+	struct drm_crtc base;
+	char name[8];
+	struct drm_plane *plane;
+	struct drm_plane *planes[8];
+	int id;
+	bool enabled;
+
+	/* which mixer/encoder we route output to: */
+	int mixer;
+
+	/* if there is a pending flip, these will be non-null: */
+	struct drm_pending_vblank_event *event;
+	struct msm_fence_cb pageflip_cb;
+
+#define PENDING_CURSOR 0x1
+#define PENDING_FLIP   0x2
+	atomic_t pending;
+
+	/* the fb that we logically (from PoV of KMS API) hold a ref
+	 * to.  Which we may not yet be scanning out (we may still
+	 * be scanning out previous in case of page_flip while waiting
+	 * for gpu rendering to complete:
+	 */
+	struct drm_framebuffer *fb;
+
+	/* the fb that we currently hold a scanout ref to: */
+	struct drm_framebuffer *scanout_fb;
+
+	/* for unref'ing framebuffers after scanout completes: */
+	struct drm_flip_work unref_fb_work;
+
+	struct mdp_irq vblank;
+	struct mdp_irq err;
+};
+#define to_mdp5_crtc(x) container_of(x, struct mdp5_crtc, base)
+
+static struct mdp5_kms *get_kms(struct drm_crtc *crtc)
+{
+	struct msm_drm_private *priv = crtc->dev->dev_private;
+	return to_mdp5_kms(to_mdp_kms(priv->kms));
+}
+
+static void request_pending(struct drm_crtc *crtc, uint32_t pending)
+{
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+
+	atomic_or(pending, &mdp5_crtc->pending);
+	mdp_irq_register(&get_kms(crtc)->base, &mdp5_crtc->vblank);
+}
+
+static void crtc_flush(struct drm_crtc *crtc)
+{
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+	struct mdp5_kms *mdp5_kms = get_kms(crtc);
+	int id = mdp5_crtc->id;
+	uint32_t i, flush = 0;
+
+	for (i = 0; i < ARRAY_SIZE(mdp5_crtc->planes); i++) {
+		struct drm_plane *plane = mdp5_crtc->planes[i];
+		if (plane) {
+			enum mdp5_pipe pipe = mdp5_plane_pipe(plane);
+			flush |= pipe2flush(pipe);
+		}
+	}
+	flush |= mixer2flush(mdp5_crtc->id);
+	flush |= MDP5_CTL_FLUSH_CTL;
+
+	DBG("%s: flush=%08x", mdp5_crtc->name, flush);
+
+	mdp5_write(mdp5_kms, REG_MDP5_CTL_FLUSH(id), flush);
+}
+
+static void update_fb(struct drm_crtc *crtc, struct drm_framebuffer *new_fb)
+{
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+	struct drm_framebuffer *old_fb = mdp5_crtc->fb;
+
+	/* grab reference to incoming scanout fb: */
+	drm_framebuffer_reference(new_fb);
+	mdp5_crtc->base.fb = new_fb;
+	mdp5_crtc->fb = new_fb;
+
+	if (old_fb)
+		drm_flip_work_queue(&mdp5_crtc->unref_fb_work, old_fb);
+}
+
+/* unlike update_fb(), take a ref to the new scanout fb *before* updating
+ * plane, then call this.  Needed to ensure we don't unref the buffer that
+ * is actually still being scanned out.
+ *
+ * Note that this whole thing goes away with atomic.. since we can defer
+ * calling into driver until rendering is done.
+ */
+static void update_scanout(struct drm_crtc *crtc, struct drm_framebuffer *fb)
+{
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+
+	/* flush updates, to make sure hw is updated to new scanout fb,
+	 * so that we can safely queue unref to current fb (ie. next
+	 * vblank we know hw is done w/ previous scanout_fb).
+	 */
+	crtc_flush(crtc);
+
+	if (mdp5_crtc->scanout_fb)
+		drm_flip_work_queue(&mdp5_crtc->unref_fb_work,
+				mdp5_crtc->scanout_fb);
+
+	mdp5_crtc->scanout_fb = fb;
+
+	/* enable vblank to complete flip: */
+	request_pending(crtc, PENDING_FLIP);
+}
+
+/* if file!=NULL, this is preclose potential cancel-flip path */
+static void complete_flip(struct drm_crtc *crtc, struct drm_file *file)
+{
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+	struct drm_device *dev = crtc->dev;
+	struct drm_pending_vblank_event *event;
+	unsigned long flags, i;
+
+	spin_lock_irqsave(&dev->event_lock, flags);
+	event = mdp5_crtc->event;
+	if (event) {
+		/* if regular vblank case (!file) or if cancel-flip from
+		 * preclose on file that requested flip, then send the
+		 * event:
+		 */
+		if (!file || (event->base.file_priv == file)) {
+			mdp5_crtc->event = NULL;
+			drm_send_vblank_event(dev, mdp5_crtc->id, event);
+		}
+	}
+	spin_unlock_irqrestore(&dev->event_lock, flags);
+
+	for (i = 0; i < ARRAY_SIZE(mdp5_crtc->planes); i++) {
+		struct drm_plane *plane = mdp5_crtc->planes[i];
+		if (plane)
+			mdp5_plane_complete_flip(plane);
+	}
+}
+
+static void pageflip_cb(struct msm_fence_cb *cb)
+{
+	struct mdp5_crtc *mdp5_crtc =
+		container_of(cb, struct mdp5_crtc, pageflip_cb);
+	struct drm_crtc *crtc = &mdp5_crtc->base;
+	struct drm_framebuffer *fb = mdp5_crtc->fb;
+
+	if (!fb)
+		return;
+
+	drm_framebuffer_reference(fb);
+	mdp5_plane_set_scanout(mdp5_crtc->plane, fb);
+	update_scanout(crtc, fb);
+}
+
+static void unref_fb_worker(struct drm_flip_work *work, void *val)
+{
+	struct mdp5_crtc *mdp5_crtc =
+		container_of(work, struct mdp5_crtc, unref_fb_work);
+	struct drm_device *dev = mdp5_crtc->base.dev;
+
+	mutex_lock(&dev->mode_config.mutex);
+	drm_framebuffer_unreference(val);
+	mutex_unlock(&dev->mode_config.mutex);
+}
+
+static void mdp5_crtc_destroy(struct drm_crtc *crtc)
+{
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+
+	mdp5_crtc->plane->funcs->destroy(mdp5_crtc->plane);
+
+	drm_crtc_cleanup(crtc);
+	drm_flip_work_cleanup(&mdp5_crtc->unref_fb_work);
+
+	kfree(mdp5_crtc);
+}
+
+static void mdp5_crtc_dpms(struct drm_crtc *crtc, int mode)
+{
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+	struct mdp5_kms *mdp5_kms = get_kms(crtc);
+	bool enabled = (mode == DRM_MODE_DPMS_ON);
+
+	DBG("%s: mode=%d", mdp5_crtc->name, mode);
+
+	if (enabled != mdp5_crtc->enabled) {
+		if (enabled) {
+			mdp5_enable(mdp5_kms);
+			mdp_irq_register(&mdp5_kms->base, &mdp5_crtc->err);
+		} else {
+			mdp_irq_unregister(&mdp5_kms->base, &mdp5_crtc->err);
+			mdp5_disable(mdp5_kms);
+		}
+		mdp5_crtc->enabled = enabled;
+	}
+}
+
+static bool mdp5_crtc_mode_fixup(struct drm_crtc *crtc,
+		const struct drm_display_mode *mode,
+		struct drm_display_mode *adjusted_mode)
+{
+	return true;
+}
+
+static void blend_setup(struct drm_crtc *crtc)
+{
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+	struct mdp5_kms *mdp5_kms = get_kms(crtc);
+	int id = mdp5_crtc->id;
+
+	/*
+	 * Hard-coded setup for now until I figure out how the
+	 * layer-mixer works
+	 */
+
+	/* LM[id]: */
+	mdp5_write(mdp5_kms, REG_MDP5_LM_BLEND_COLOR_OUT(id),
+			MDP5_LM_BLEND_COLOR_OUT_STAGE0_FG_ALPHA);
+	mdp5_write(mdp5_kms, REG_MDP5_LM_BLEND_OP_MODE(id, 0),
+			MDP5_LM_BLEND_OP_MODE_FG_ALPHA(FG_CONST) |
+			MDP5_LM_BLEND_OP_MODE_BG_ALPHA(FG_PIXEL) |
+			MDP5_LM_BLEND_OP_MODE_BG_INV_ALPHA);
+	mdp5_write(mdp5_kms, REG_MDP5_LM_BLEND_FG_ALPHA(id, 0), 0xff);
+	mdp5_write(mdp5_kms, REG_MDP5_LM_BLEND_BG_ALPHA(id, 0), 0x00);
+
+	/* NOTE: seems that LM[n] and CTL[m], we do not need n==m.. but
+	 * we want to be setting CTL[m].LAYER[n].  Not sure what the
+	 * point of having CTL[m].LAYER[o] (for o!=n).. maybe that is
+	 * used when chaining up mixers for high resolution displays?
+	 */
+
+	/* CTL[id]: */
+	mdp5_write(mdp5_kms, REG_MDP5_CTL_LAYER_REG(id, 0),
+			MDP5_CTL_LAYER_REG_RGB0(STAGE0) |
+			MDP5_CTL_LAYER_REG_BORDER_COLOR);
+	mdp5_write(mdp5_kms, REG_MDP5_CTL_LAYER_REG(id, 1), 0);
+	mdp5_write(mdp5_kms, REG_MDP5_CTL_LAYER_REG(id, 2), 0);
+	mdp5_write(mdp5_kms, REG_MDP5_CTL_LAYER_REG(id, 3), 0);
+	mdp5_write(mdp5_kms, REG_MDP5_CTL_LAYER_REG(id, 4), 0);
+}
+
+static int mdp5_crtc_mode_set(struct drm_crtc *crtc,
+		struct drm_display_mode *mode,
+		struct drm_display_mode *adjusted_mode,
+		int x, int y,
+		struct drm_framebuffer *old_fb)
+{
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+	struct mdp5_kms *mdp5_kms = get_kms(crtc);
+	int ret;
+
+	mode = adjusted_mode;
+
+	DBG("%s: set mode: %d:\"%s\" %d %d %d %d %d %d %d %d %d %d 0x%x 0x%x",
+			mdp5_crtc->name, 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);
+
+	/* grab extra ref for update_scanout() */
+	drm_framebuffer_reference(crtc->fb);
+
+	ret = mdp5_plane_mode_set(mdp5_crtc->plane, crtc, crtc->fb,
+			0, 0, mode->hdisplay, mode->vdisplay,
+			x << 16, y << 16,
+			mode->hdisplay << 16, mode->vdisplay << 16);
+	if (ret) {
+		dev_err(crtc->dev->dev, "%s: failed to set mode on plane: %d\n",
+				mdp5_crtc->name, ret);
+		return ret;
+	}
+
+	mdp5_write(mdp5_kms, REG_MDP5_LM_OUT_SIZE(mdp5_crtc->id),
+			MDP5_LM_OUT_SIZE_WIDTH(mode->hdisplay) |
+			MDP5_LM_OUT_SIZE_HEIGHT(mode->vdisplay));
+
+	update_fb(crtc, crtc->fb);
+	update_scanout(crtc, crtc->fb);
+
+	return 0;
+}
+
+static void mdp5_crtc_prepare(struct drm_crtc *crtc)
+{
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+	DBG("%s", mdp5_crtc->name);
+	/* make sure we hold a ref to mdp clks while setting up mode: */
+	mdp5_enable(get_kms(crtc));
+	mdp5_crtc_dpms(crtc, DRM_MODE_DPMS_OFF);
+}
+
+static void mdp5_crtc_commit(struct drm_crtc *crtc)
+{
+	mdp5_crtc_dpms(crtc, DRM_MODE_DPMS_ON);
+	crtc_flush(crtc);
+	/* drop the ref to mdp clk's that we got in prepare: */
+	mdp5_disable(get_kms(crtc));
+}
+
+static int mdp5_crtc_mode_set_base(struct drm_crtc *crtc, int x, int y,
+		struct drm_framebuffer *old_fb)
+{
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+	struct drm_plane *plane = mdp5_crtc->plane;
+	struct drm_display_mode *mode = &crtc->mode;
+	int ret;
+
+	/* grab extra ref for update_scanout() */
+	drm_framebuffer_reference(crtc->fb);
+
+	ret = mdp5_plane_mode_set(plane, crtc, crtc->fb,
+			0, 0, mode->hdisplay, mode->vdisplay,
+			x << 16, y << 16,
+			mode->hdisplay << 16, mode->vdisplay << 16);
+
+	update_fb(crtc, crtc->fb);
+	update_scanout(crtc, crtc->fb);
+
+	return ret;
+}
+
+static void mdp5_crtc_load_lut(struct drm_crtc *crtc)
+{
+}
+
+static int mdp5_crtc_page_flip(struct drm_crtc *crtc,
+		struct drm_framebuffer *new_fb,
+		struct drm_pending_vblank_event *event,
+		uint32_t page_flip_flags)
+{
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+	struct drm_device *dev = crtc->dev;
+	struct drm_gem_object *obj;
+	unsigned long flags;
+
+	if (mdp5_crtc->event) {
+		dev_err(dev->dev, "already pending flip!\n");
+		return -EBUSY;
+	}
+
+	obj = msm_framebuffer_bo(new_fb, 0);
+
+	spin_lock_irqsave(&dev->event_lock, flags);
+	mdp5_crtc->event = event;
+	spin_unlock_irqrestore(&dev->event_lock, flags);
+
+	update_fb(crtc, new_fb);
+
+	return msm_gem_queue_inactive_cb(obj, &mdp5_crtc->pageflip_cb);
+}
+
+static int mdp5_crtc_set_property(struct drm_crtc *crtc,
+		struct drm_property *property, uint64_t val)
+{
+	// XXX
+	return -EINVAL;
+}
+
+static const struct drm_crtc_funcs mdp5_crtc_funcs = {
+	.set_config = drm_crtc_helper_set_config,
+	.destroy = mdp5_crtc_destroy,
+	.page_flip = mdp5_crtc_page_flip,
+	.set_property = mdp5_crtc_set_property,
+};
+
+static const struct drm_crtc_helper_funcs mdp5_crtc_helper_funcs = {
+	.dpms = mdp5_crtc_dpms,
+	.mode_fixup = mdp5_crtc_mode_fixup,
+	.mode_set = mdp5_crtc_mode_set,
+	.prepare = mdp5_crtc_prepare,
+	.commit = mdp5_crtc_commit,
+	.mode_set_base = mdp5_crtc_mode_set_base,
+	.load_lut = mdp5_crtc_load_lut,
+};
+
+static void mdp5_crtc_vblank_irq(struct mdp_irq *irq, uint32_t irqstatus)
+{
+	struct mdp5_crtc *mdp5_crtc = container_of(irq, struct mdp5_crtc, vblank);
+	struct drm_crtc *crtc = &mdp5_crtc->base;
+	struct msm_drm_private *priv = crtc->dev->dev_private;
+	unsigned pending;
+
+	mdp_irq_unregister(&get_kms(crtc)->base, &mdp5_crtc->vblank);
+
+	pending = atomic_xchg(&mdp5_crtc->pending, 0);
+
+	if (pending & PENDING_FLIP) {
+		complete_flip(crtc, NULL);
+		drm_flip_work_commit(&mdp5_crtc->unref_fb_work, priv->wq);
+	}
+}
+
+static void mdp5_crtc_err_irq(struct mdp_irq *irq, uint32_t irqstatus)
+{
+	struct mdp5_crtc *mdp5_crtc = container_of(irq, struct mdp5_crtc, err);
+	struct drm_crtc *crtc = &mdp5_crtc->base;
+	DBG("%s: error: %08x", mdp5_crtc->name, irqstatus);
+	crtc_flush(crtc);
+}
+
+uint32_t mdp5_crtc_vblank(struct drm_crtc *crtc)
+{
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+	return mdp5_crtc->vblank.irqmask;
+}
+
+void mdp5_crtc_cancel_pending_flip(struct drm_crtc *crtc, struct drm_file *file)
+{
+	DBG("cancel: %p", file);
+	complete_flip(crtc, file);
+}
+
+/* set interface for routing crtc->encoder: */
+void mdp5_crtc_set_intf(struct drm_crtc *crtc, int intf,
+		enum mdp5_intf intf_id)
+{
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+	struct mdp5_kms *mdp5_kms = get_kms(crtc);
+	static const enum mdp5_intfnum intfnum[] = {
+			INTF0, INTF1, INTF2, INTF3,
+	};
+	uint32_t intf_sel;
+
+	/* now that we know what irq's we want: */
+	mdp5_crtc->err.irqmask = intf2err(intf);
+	mdp5_crtc->vblank.irqmask = intf2vblank(intf);
+
+	/* when called from modeset_init(), skip the rest until later: */
+	if (!mdp5_kms)
+		return;
+
+	intf_sel = mdp5_read(mdp5_kms, REG_MDP5_DISP_INTF_SEL);
+
+	switch (intf) {
+	case 0:
+		intf_sel &= ~MDP5_DISP_INTF_SEL_INTF0__MASK;
+		intf_sel |= MDP5_DISP_INTF_SEL_INTF0(intf_id);
+		break;
+	case 1:
+		intf_sel &= ~MDP5_DISP_INTF_SEL_INTF1__MASK;
+		intf_sel |= MDP5_DISP_INTF_SEL_INTF1(intf_id);
+		break;
+	case 2:
+		intf_sel &= ~MDP5_DISP_INTF_SEL_INTF2__MASK;
+		intf_sel |= MDP5_DISP_INTF_SEL_INTF2(intf_id);
+		break;
+	case 3:
+		intf_sel &= ~MDP5_DISP_INTF_SEL_INTF3__MASK;
+		intf_sel |= MDP5_DISP_INTF_SEL_INTF3(intf_id);
+		break;
+	default:
+		BUG();
+		break;
+	}
+
+	blend_setup(crtc);
+
+	DBG("%s: intf_sel=%08x", mdp5_crtc->name, intf_sel);
+
+	mdp5_write(mdp5_kms, REG_MDP5_DISP_INTF_SEL, intf_sel);
+	mdp5_write(mdp5_kms, REG_MDP5_CTL_OP(mdp5_crtc->id),
+			MDP5_CTL_OP_MODE(MODE_NONE) |
+			MDP5_CTL_OP_INTF_NUM(intfnum[intf]));
+
+	crtc_flush(crtc);
+}
+
+static void set_attach(struct drm_crtc *crtc, enum mdp5_pipe pipe_id,
+		struct drm_plane *plane)
+{
+	struct mdp5_crtc *mdp5_crtc = to_mdp5_crtc(crtc);
+
+	BUG_ON(pipe_id >= ARRAY_SIZE(mdp5_crtc->planes));
+
+	if (mdp5_crtc->planes[pipe_id] == plane)
+		return;
+
+	mdp5_crtc->planes[pipe_id] = plane;
+	blend_setup(crtc);
+	if (mdp5_crtc->enabled && (plane != mdp5_crtc->plane))
+		crtc_flush(crtc);
+}
+
+void mdp5_crtc_attach(struct drm_crtc *crtc, struct drm_plane *plane)
+{
+	set_attach(crtc, mdp5_plane_pipe(plane), plane);
+}
+
+void mdp5_crtc_detach(struct drm_crtc *crtc, struct drm_plane *plane)
+{
+	set_attach(crtc, mdp5_plane_pipe(plane), NULL);
+}
+
+/* initialize crtc */
+struct drm_crtc *mdp5_crtc_init(struct drm_device *dev,
+		struct drm_plane *plane, int id)
+{
+	struct drm_crtc *crtc = NULL;
+	struct mdp5_crtc *mdp5_crtc;
+	int ret;
+
+	mdp5_crtc = kzalloc(sizeof(*mdp5_crtc), GFP_KERNEL);
+	if (!mdp5_crtc) {
+		ret = -ENOMEM;
+		goto fail;
+	}
+
+	crtc = &mdp5_crtc->base;
+
+	mdp5_crtc->plane = plane;
+	mdp5_crtc->id = id;
+
+	mdp5_crtc->vblank.irq = mdp5_crtc_vblank_irq;
+	mdp5_crtc->err.irq = mdp5_crtc_err_irq;
+
+	snprintf(mdp5_crtc->name, sizeof(mdp5_crtc->name), "%s:%d",
+			pipe2name(mdp5_plane_pipe(plane)), id);
+
+	ret = drm_flip_work_init(&mdp5_crtc->unref_fb_work, 16,
+			"unref fb", unref_fb_worker);
+	if (ret)
+		goto fail;
+
+	INIT_FENCE_CB(&mdp5_crtc->pageflip_cb, pageflip_cb);
+
+	drm_crtc_init(dev, crtc, &mdp5_crtc_funcs);
+	drm_crtc_helper_add(crtc, &mdp5_crtc_helper_funcs);
+
+	mdp5_plane_install_properties(mdp5_crtc->plane, &crtc->base);
+
+	return crtc;
+
+fail:
+	if (crtc)
+		mdp5_crtc_destroy(crtc);
+
+	return ERR_PTR(ret);
+}
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_encoder.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_encoder.c
new file mode 100644
index 000000000000..edec7bfaa952
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_encoder.c
@@ -0,0 +1,258 @@
+/*
+ * Copyright (C) 2013 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 "mdp5_kms.h"
+
+#include "drm_crtc.h"
+#include "drm_crtc_helper.h"
+
+struct mdp5_encoder {
+	struct drm_encoder base;
+	int intf;
+	enum mdp5_intf intf_id;
+	bool enabled;
+	uint32_t bsc;
+};
+#define to_mdp5_encoder(x) container_of(x, struct mdp5_encoder, base)
+
+static struct mdp5_kms *get_kms(struct drm_encoder *encoder)
+{
+	struct msm_drm_private *priv = encoder->dev->dev_private;
+	return to_mdp5_kms(to_mdp_kms(priv->kms));
+}
+
+#ifdef CONFIG_MSM_BUS_SCALING
+#include <mach/board.h>
+#include <mach/msm_bus.h>
+#include <mach/msm_bus_board.h>
+#define MDP_BUS_VECTOR_ENTRY(ab_val, ib_val)		\
+	{						\
+		.src = MSM_BUS_MASTER_MDP_PORT0,	\
+		.dst = MSM_BUS_SLAVE_EBI_CH0,		\
+		.ab = (ab_val),				\
+		.ib = (ib_val),				\
+	}
+
+static struct msm_bus_vectors mdp_bus_vectors[] = {
+	MDP_BUS_VECTOR_ENTRY(0, 0),
+	MDP_BUS_VECTOR_ENTRY(2000000000, 2000000000),
+};
+static struct msm_bus_paths mdp_bus_usecases[] = { {
+		.num_paths = 1,
+		.vectors = &mdp_bus_vectors[0],
+}, {
+		.num_paths = 1,
+		.vectors = &mdp_bus_vectors[1],
+} };
+static struct msm_bus_scale_pdata mdp_bus_scale_table = {
+	.usecase = mdp_bus_usecases,
+	.num_usecases = ARRAY_SIZE(mdp_bus_usecases),
+	.name = "mdss_mdp",
+};
+
+static void bs_init(struct mdp5_encoder *mdp5_encoder)
+{
+	mdp5_encoder->bsc = msm_bus_scale_register_client(
+			&mdp_bus_scale_table);
+	DBG("bus scale client: %08x", mdp5_encoder->bsc);
+}
+
+static void bs_fini(struct mdp5_encoder *mdp5_encoder)
+{
+	if (mdp5_encoder->bsc) {
+		msm_bus_scale_unregister_client(mdp5_encoder->bsc);
+		mdp5_encoder->bsc = 0;
+	}
+}
+
+static void bs_set(struct mdp5_encoder *mdp5_encoder, int idx)
+{
+	if (mdp5_encoder->bsc) {
+		DBG("set bus scaling: %d", idx);
+		/* HACK: scaling down, and then immediately back up
+		 * seems to leave things broken (underflow).. so
+		 * never disable:
+		 */
+		idx = 1;
+		msm_bus_scale_client_update_request(mdp5_encoder->bsc, idx);
+	}
+}
+#else
+static void bs_init(struct mdp5_encoder *mdp5_encoder) {}
+static void bs_fini(struct mdp5_encoder *mdp5_encoder) {}
+static void bs_set(struct mdp5_encoder *mdp5_encoder, int idx) {}
+#endif
+
+static void mdp5_encoder_destroy(struct drm_encoder *encoder)
+{
+	struct mdp5_encoder *mdp5_encoder = to_mdp5_encoder(encoder);
+	bs_fini(mdp5_encoder);
+	drm_encoder_cleanup(encoder);
+	kfree(mdp5_encoder);
+}
+
+static const struct drm_encoder_funcs mdp5_encoder_funcs = {
+	.destroy = mdp5_encoder_destroy,
+};
+
+static void mdp5_encoder_dpms(struct drm_encoder *encoder, int mode)
+{
+	struct mdp5_encoder *mdp5_encoder = to_mdp5_encoder(encoder);
+	struct mdp5_kms *mdp5_kms = get_kms(encoder);
+	int intf = mdp5_encoder->intf;
+	bool enabled = (mode == DRM_MODE_DPMS_ON);
+
+	DBG("mode=%d", mode);
+
+	if (enabled == mdp5_encoder->enabled)
+		return;
+
+	if (enabled) {
+		bs_set(mdp5_encoder, 1);
+		mdp5_write(mdp5_kms, REG_MDP5_INTF_TIMING_ENGINE_EN(intf), 1);
+	} else {
+		mdp5_write(mdp5_kms, REG_MDP5_INTF_TIMING_ENGINE_EN(intf), 0);
+		bs_set(mdp5_encoder, 0);
+	}
+
+	mdp5_encoder->enabled = enabled;
+}
+
+static bool mdp5_encoder_mode_fixup(struct drm_encoder *encoder,
+		const struct drm_display_mode *mode,
+		struct drm_display_mode *adjusted_mode)
+{
+	return true;
+}
+
+static void mdp5_encoder_mode_set(struct drm_encoder *encoder,
+		struct drm_display_mode *mode,
+		struct drm_display_mode *adjusted_mode)
+{
+	struct mdp5_encoder *mdp5_encoder = to_mdp5_encoder(encoder);
+	struct mdp5_kms *mdp5_kms = get_kms(encoder);
+	int intf = mdp5_encoder->intf;
+	uint32_t dtv_hsync_skew, vsync_period, vsync_len, ctrl_pol;
+	uint32_t display_v_start, display_v_end;
+	uint32_t hsync_start_x, hsync_end_x;
+	uint32_t format;
+
+	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);
+
+	ctrl_pol = 0;
+	if (mode->flags & DRM_MODE_FLAG_NHSYNC)
+		ctrl_pol |= MDP5_INTF_POLARITY_CTL_HSYNC_LOW;
+	if (mode->flags & DRM_MODE_FLAG_NVSYNC)
+		ctrl_pol |= MDP5_INTF_POLARITY_CTL_VSYNC_LOW;
+	/* probably need to get DATA_EN polarity from panel.. */
+
+	dtv_hsync_skew = 0;  /* get this from panel? */
+	format = 0x213f;     /* 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 + dtv_hsync_skew;
+	display_v_end = vsync_period - ((mode->vsync_start - mode->vdisplay) * mode->htotal) + dtv_hsync_skew - 1;
+
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_HSYNC_CTL(intf),
+			MDP5_INTF_HSYNC_CTL_PULSEW(mode->hsync_end - mode->hsync_start) |
+			MDP5_INTF_HSYNC_CTL_PERIOD(mode->htotal));
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_VSYNC_PERIOD_F0(intf), vsync_period);
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_VSYNC_LEN_F0(intf), vsync_len);
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_DISPLAY_HCTL(intf),
+			MDP5_INTF_DISPLAY_HCTL_START(hsync_start_x) |
+			MDP5_INTF_DISPLAY_HCTL_END(hsync_end_x));
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_DISPLAY_VSTART_F0(intf), display_v_start);
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_DISPLAY_VEND_F0(intf), display_v_end);
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_BORDER_COLOR(intf), 0);
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_UNDERFLOW_COLOR(intf), 0xff);
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_HSYNC_SKEW(intf), dtv_hsync_skew);
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_POLARITY_CTL(intf), ctrl_pol);
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_ACTIVE_HCTL(intf),
+			MDP5_INTF_ACTIVE_HCTL_START(0) |
+			MDP5_INTF_ACTIVE_HCTL_END(0));
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_ACTIVE_VSTART_F0(intf), 0);
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_ACTIVE_VEND_F0(intf), 0);
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_PANEL_FORMAT(intf), format);
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_FRAME_LINE_COUNT_EN(intf), 0x3);  /* frame+line? */
+}
+
+static void mdp5_encoder_prepare(struct drm_encoder *encoder)
+{
+	mdp5_encoder_dpms(encoder, DRM_MODE_DPMS_OFF);
+}
+
+static void mdp5_encoder_commit(struct drm_encoder *encoder)
+{
+	struct mdp5_encoder *mdp5_encoder = to_mdp5_encoder(encoder);
+	mdp5_crtc_set_intf(encoder->crtc, mdp5_encoder->intf,
+			mdp5_encoder->intf_id);
+	mdp5_encoder_dpms(encoder, DRM_MODE_DPMS_ON);
+}
+
+static const struct drm_encoder_helper_funcs mdp5_encoder_helper_funcs = {
+	.dpms = mdp5_encoder_dpms,
+	.mode_fixup = mdp5_encoder_mode_fixup,
+	.mode_set = mdp5_encoder_mode_set,
+	.prepare = mdp5_encoder_prepare,
+	.commit = mdp5_encoder_commit,
+};
+
+/* initialize encoder */
+struct drm_encoder *mdp5_encoder_init(struct drm_device *dev, int intf,
+		enum mdp5_intf intf_id)
+{
+	struct drm_encoder *encoder = NULL;
+	struct mdp5_encoder *mdp5_encoder;
+	int ret;
+
+	mdp5_encoder = kzalloc(sizeof(*mdp5_encoder), GFP_KERNEL);
+	if (!mdp5_encoder) {
+		ret = -ENOMEM;
+		goto fail;
+	}
+
+	mdp5_encoder->intf = intf;
+	mdp5_encoder->intf_id = intf_id;
+	encoder = &mdp5_encoder->base;
+
+	drm_encoder_init(dev, encoder, &mdp5_encoder_funcs,
+			 DRM_MODE_ENCODER_TMDS);
+	drm_encoder_helper_add(encoder, &mdp5_encoder_helper_funcs);
+
+	bs_init(mdp5_encoder);
+
+	return encoder;
+
+fail:
+	if (encoder)
+		mdp5_encoder_destroy(encoder);
+
+	return ERR_PTR(ret);
+}
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_irq.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_irq.c
new file mode 100644
index 000000000000..353d494a497f
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_irq.c
@@ -0,0 +1,111 @@
+/*
+ * Copyright (C) 2013 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 "msm_drv.h"
+#include "mdp5_kms.h"
+
+void mdp5_set_irqmask(struct mdp_kms *mdp_kms, uint32_t irqmask)
+{
+	mdp5_write(to_mdp5_kms(mdp_kms), REG_MDP5_INTR_EN, irqmask);
+}
+
+static void mdp5_irq_error_handler(struct mdp_irq *irq, uint32_t irqstatus)
+{
+	DRM_ERROR("errors: %08x\n", irqstatus);
+}
+
+void mdp5_irq_preinstall(struct msm_kms *kms)
+{
+	struct mdp5_kms *mdp5_kms = to_mdp5_kms(to_mdp_kms(kms));
+	mdp5_write(mdp5_kms, REG_MDP5_INTR_CLEAR, 0xffffffff);
+}
+
+int mdp5_irq_postinstall(struct msm_kms *kms)
+{
+	struct mdp_kms *mdp_kms = to_mdp_kms(kms);
+	struct mdp5_kms *mdp5_kms = to_mdp5_kms(mdp_kms);
+	struct mdp_irq *error_handler = &mdp5_kms->error_handler;
+
+	error_handler->irq = mdp5_irq_error_handler;
+	error_handler->irqmask = MDP5_IRQ_INTF0_UNDER_RUN |
+			MDP5_IRQ_INTF1_UNDER_RUN |
+			MDP5_IRQ_INTF2_UNDER_RUN |
+			MDP5_IRQ_INTF3_UNDER_RUN;
+
+	mdp_irq_register(mdp_kms, error_handler);
+
+	return 0;
+}
+
+void mdp5_irq_uninstall(struct msm_kms *kms)
+{
+	struct mdp5_kms *mdp5_kms = to_mdp5_kms(to_mdp_kms(kms));
+	mdp5_write(mdp5_kms, REG_MDP5_INTR_EN, 0x00000000);
+}
+
+static void mdp5_irq_mdp(struct mdp_kms *mdp_kms)
+{
+	struct mdp5_kms *mdp5_kms = to_mdp5_kms(mdp_kms);
+	struct drm_device *dev = mdp5_kms->dev;
+	struct msm_drm_private *priv = dev->dev_private;
+	unsigned int id;
+	uint32_t status;
+
+	status = mdp5_read(mdp5_kms, REG_MDP5_INTR_STATUS);
+	mdp5_write(mdp5_kms, REG_MDP5_INTR_CLEAR, status);
+
+	VERB("status=%08x", status);
+
+	for (id = 0; id < priv->num_crtcs; id++)
+		if (status & mdp5_crtc_vblank(priv->crtcs[id]))
+			drm_handle_vblank(dev, id);
+
+	mdp_dispatch_irqs(mdp_kms, status);
+}
+
+irqreturn_t mdp5_irq(struct msm_kms *kms)
+{
+	struct mdp_kms *mdp_kms = to_mdp_kms(kms);
+	struct mdp5_kms *mdp5_kms = to_mdp5_kms(mdp_kms);
+	uint32_t intr;
+
+	intr = mdp5_read(mdp5_kms, REG_MDP5_HW_INTR_STATUS);
+
+	VERB("intr=%08x", intr);
+
+	if (intr & MDP5_HW_INTR_STATUS_INTR_MDP)
+		mdp5_irq_mdp(mdp_kms);
+
+	if (intr & MDP5_HW_INTR_STATUS_INTR_HDMI)
+		hdmi_irq(0, mdp5_kms->hdmi);
+
+	return IRQ_HANDLED;
+}
+
+int mdp5_enable_vblank(struct msm_kms *kms, struct drm_crtc *crtc)
+{
+	mdp_update_vblank_mask(to_mdp_kms(kms),
+			mdp5_crtc_vblank(crtc), true);
+	return 0;
+}
+
+void mdp5_disable_vblank(struct msm_kms *kms, struct drm_crtc *crtc)
+{
+	mdp_update_vblank_mask(to_mdp_kms(kms),
+			mdp5_crtc_vblank(crtc), false);
+}
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c
new file mode 100644
index 000000000000..ee8446c1b5f6
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c
@@ -0,0 +1,350 @@
+/*
+ * Copyright (C) 2013 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 "msm_drv.h"
+#include "msm_mmu.h"
+#include "mdp5_kms.h"
+
+static struct mdp5_platform_config *mdp5_get_config(struct platform_device *dev);
+
+static int mdp5_hw_init(struct msm_kms *kms)
+{
+	struct mdp5_kms *mdp5_kms = to_mdp5_kms(to_mdp_kms(kms));
+	struct drm_device *dev = mdp5_kms->dev;
+	uint32_t version, major, minor;
+	int ret = 0;
+
+	pm_runtime_get_sync(dev->dev);
+
+	mdp5_enable(mdp5_kms);
+	version = mdp5_read(mdp5_kms, REG_MDP5_MDP_VERSION);
+	mdp5_disable(mdp5_kms);
+
+	major = FIELD(version, MDP5_MDP_VERSION_MAJOR);
+	minor = FIELD(version, MDP5_MDP_VERSION_MINOR);
+
+	DBG("found MDP5 version v%d.%d", major, minor);
+
+	if ((major != 1) || ((minor != 0) && (minor != 2))) {
+		dev_err(dev->dev, "unexpected MDP version: v%d.%d\n",
+				major, minor);
+		ret = -ENXIO;
+		goto out;
+	}
+
+	mdp5_kms->rev = minor;
+
+	/* Magic unknown register writes:
+	 *
+	 *    W VBIF:0x004 00000001      (mdss_mdp.c:839)
+	 *    W MDP5:0x2e0 0xe9          (mdss_mdp.c:839)
+	 *    W MDP5:0x2e4 0x55          (mdss_mdp.c:839)
+	 *    W MDP5:0x3ac 0xc0000ccc    (mdss_mdp.c:839)
+	 *    W MDP5:0x3b4 0xc0000ccc    (mdss_mdp.c:839)
+	 *    W MDP5:0x3bc 0xcccccc      (mdss_mdp.c:839)
+	 *    W MDP5:0x4a8 0xcccc0c0     (mdss_mdp.c:839)
+	 *    W MDP5:0x4b0 0xccccc0c0    (mdss_mdp.c:839)
+	 *    W MDP5:0x4b8 0xccccc000    (mdss_mdp.c:839)
+	 *
+	 * Downstream fbdev driver gets these register offsets/values
+	 * from DT.. not really sure what these registers are or if
+	 * different values for different boards/SoC's, etc.  I guess
+	 * they are the golden registers.
+	 *
+	 * Not setting these does not seem to cause any problem.  But
+	 * we may be getting lucky with the bootloader initializing
+	 * them for us.  OTOH, if we can always count on the bootloader
+	 * setting the golden registers, then perhaps we don't need to
+	 * care.
+	 */
+
+	mdp5_write(mdp5_kms, REG_MDP5_DISP_INTF_SEL, 0);
+	mdp5_write(mdp5_kms, REG_MDP5_CTL_OP(0), 0);
+	mdp5_write(mdp5_kms, REG_MDP5_CTL_OP(1), 0);
+	mdp5_write(mdp5_kms, REG_MDP5_CTL_OP(2), 0);
+	mdp5_write(mdp5_kms, REG_MDP5_CTL_OP(3), 0);
+
+out:
+	pm_runtime_put_sync(dev->dev);
+
+	return ret;
+}
+
+static long mdp5_round_pixclk(struct msm_kms *kms, unsigned long rate,
+		struct drm_encoder *encoder)
+{
+	return rate;
+}
+
+static void mdp5_preclose(struct msm_kms *kms, struct drm_file *file)
+{
+	struct mdp5_kms *mdp5_kms = to_mdp5_kms(to_mdp_kms(kms));
+	struct msm_drm_private *priv = mdp5_kms->dev->dev_private;
+	unsigned i;
+
+	for (i = 0; i < priv->num_crtcs; i++)
+		mdp5_crtc_cancel_pending_flip(priv->crtcs[i], file);
+}
+
+static void mdp5_destroy(struct msm_kms *kms)
+{
+	struct mdp5_kms *mdp5_kms = to_mdp5_kms(to_mdp_kms(kms));
+	kfree(mdp5_kms);
+}
+
+static const struct mdp_kms_funcs kms_funcs = {
+	.base = {
+		.hw_init         = mdp5_hw_init,
+		.irq_preinstall  = mdp5_irq_preinstall,
+		.irq_postinstall = mdp5_irq_postinstall,
+		.irq_uninstall   = mdp5_irq_uninstall,
+		.irq             = mdp5_irq,
+		.enable_vblank   = mdp5_enable_vblank,
+		.disable_vblank  = mdp5_disable_vblank,
+		.get_format      = mdp_get_format,
+		.round_pixclk    = mdp5_round_pixclk,
+		.preclose        = mdp5_preclose,
+		.destroy         = mdp5_destroy,
+	},
+	.set_irqmask         = mdp5_set_irqmask,
+};
+
+int mdp5_disable(struct mdp5_kms *mdp5_kms)
+{
+	DBG("");
+
+	clk_disable_unprepare(mdp5_kms->ahb_clk);
+	clk_disable_unprepare(mdp5_kms->axi_clk);
+	clk_disable_unprepare(mdp5_kms->core_clk);
+	clk_disable_unprepare(mdp5_kms->lut_clk);
+
+	return 0;
+}
+
+int mdp5_enable(struct mdp5_kms *mdp5_kms)
+{
+	DBG("");
+
+	clk_prepare_enable(mdp5_kms->ahb_clk);
+	clk_prepare_enable(mdp5_kms->axi_clk);
+	clk_prepare_enable(mdp5_kms->core_clk);
+	clk_prepare_enable(mdp5_kms->lut_clk);
+
+	return 0;
+}
+
+static int modeset_init(struct mdp5_kms *mdp5_kms)
+{
+	static const enum mdp5_pipe crtcs[] = {
+			SSPP_RGB0, SSPP_RGB1, SSPP_RGB2,
+	};
+	struct drm_device *dev = mdp5_kms->dev;
+	struct msm_drm_private *priv = dev->dev_private;
+	struct drm_encoder *encoder;
+	int i, ret;
+
+	/* construct CRTCs: */
+	for (i = 0; i < ARRAY_SIZE(crtcs); i++) {
+		struct drm_plane *plane;
+		struct drm_crtc *crtc;
+
+		plane = mdp5_plane_init(dev, crtcs[i], true);
+		if (IS_ERR(plane)) {
+			ret = PTR_ERR(plane);
+			dev_err(dev->dev, "failed to construct plane for %s (%d)\n",
+					pipe2name(crtcs[i]), ret);
+			goto fail;
+		}
+
+		crtc  = mdp5_crtc_init(dev, plane, i);
+		if (IS_ERR(crtc)) {
+			ret = PTR_ERR(crtc);
+			dev_err(dev->dev, "failed to construct crtc for %s (%d)\n",
+					pipe2name(crtcs[i]), ret);
+			goto fail;
+		}
+		priv->crtcs[priv->num_crtcs++] = crtc;
+	}
+
+	/* Construct encoder for HDMI: */
+	encoder = mdp5_encoder_init(dev, 3, INTF_HDMI);
+	if (IS_ERR(encoder)) {
+		dev_err(dev->dev, "failed to construct encoder\n");
+		ret = PTR_ERR(encoder);
+		goto fail;
+	}
+
+	/* NOTE: the vsync and error irq's are actually associated with
+	 * the INTF/encoder.. the easiest way to deal with this (ie. what
+	 * we do now) is assume a fixed relationship between crtc's and
+	 * encoders.  I'm not sure if there is ever a need to more freely
+	 * assign crtcs to encoders, but if there is then we need to take
+	 * care of error and vblank irq's that the crtc has registered,
+	 * and also update user-requested vblank_mask.
+	 */
+	encoder->possible_crtcs = BIT(0);
+	mdp5_crtc_set_intf(priv->crtcs[0], 3, INTF_HDMI);
+
+	priv->encoders[priv->num_encoders++] = encoder;
+
+	/* Construct bridge/connector for HDMI: */
+	mdp5_kms->hdmi = hdmi_init(dev, encoder);
+	if (IS_ERR(mdp5_kms->hdmi)) {
+		ret = PTR_ERR(mdp5_kms->hdmi);
+		dev_err(dev->dev, "failed to initialize HDMI: %d\n", ret);
+		goto fail;
+	}
+
+	return 0;
+
+fail:
+	return ret;
+}
+
+static const char *iommu_ports[] = {
+		"mdp_0",
+};
+
+static int get_clk(struct platform_device *pdev, struct clk **clkp,
+		const char *name)
+{
+	struct device *dev = &pdev->dev;
+	struct clk *clk = devm_clk_get(dev, name);
+	if (IS_ERR(clk)) {
+		dev_err(dev, "failed to get %s (%ld)\n", name, PTR_ERR(clk));
+		return PTR_ERR(clk);
+	}
+	*clkp = clk;
+	return 0;
+}
+
+struct msm_kms *mdp5_kms_init(struct drm_device *dev)
+{
+	struct platform_device *pdev = dev->platformdev;
+	struct mdp5_platform_config *config = mdp5_get_config(pdev);
+	struct mdp5_kms *mdp5_kms;
+	struct msm_kms *kms = NULL;
+	struct msm_mmu *mmu;
+	int ret;
+
+	mdp5_kms = kzalloc(sizeof(*mdp5_kms), GFP_KERNEL);
+	if (!mdp5_kms) {
+		dev_err(dev->dev, "failed to allocate kms\n");
+		ret = -ENOMEM;
+		goto fail;
+	}
+
+	mdp_kms_init(&mdp5_kms->base, &kms_funcs);
+
+	kms = &mdp5_kms->base.base;
+
+	mdp5_kms->dev = dev;
+	mdp5_kms->smp_blk_cnt = config->smp_blk_cnt;
+
+	mdp5_kms->mmio = msm_ioremap(pdev, "mdp_phys", "MDP5");
+	if (IS_ERR(mdp5_kms->mmio)) {
+		ret = PTR_ERR(mdp5_kms->mmio);
+		goto fail;
+	}
+
+	mdp5_kms->vbif = msm_ioremap(pdev, "vbif_phys", "VBIF");
+	if (IS_ERR(mdp5_kms->vbif)) {
+		ret = PTR_ERR(mdp5_kms->vbif);
+		goto fail;
+	}
+
+	mdp5_kms->vdd = devm_regulator_get(&pdev->dev, "vdd");
+	if (IS_ERR(mdp5_kms->vdd)) {
+		ret = PTR_ERR(mdp5_kms->vdd);
+		goto fail;
+	}
+
+	ret = regulator_enable(mdp5_kms->vdd);
+	if (ret) {
+		dev_err(dev->dev, "failed to enable regulator vdd: %d\n", ret);
+		goto fail;
+	}
+
+	ret = get_clk(pdev, &mdp5_kms->axi_clk, "bus_clk") ||
+			get_clk(pdev, &mdp5_kms->ahb_clk, "iface_clk") ||
+			get_clk(pdev, &mdp5_kms->src_clk, "core_clk_src") ||
+			get_clk(pdev, &mdp5_kms->core_clk, "core_clk") ||
+			get_clk(pdev, &mdp5_kms->lut_clk, "lut_clk") ||
+			get_clk(pdev, &mdp5_kms->vsync_clk, "vsync_clk");
+	if (ret)
+		goto fail;
+
+	ret = clk_set_rate(mdp5_kms->src_clk, config->max_clk);
+
+	/* make sure things are off before attaching iommu (bootloader could
+	 * have left things on, in which case we'll start getting faults if
+	 * we don't disable):
+	 */
+	mdp5_enable(mdp5_kms);
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_TIMING_ENGINE_EN(0), 0);
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_TIMING_ENGINE_EN(1), 0);
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_TIMING_ENGINE_EN(2), 0);
+	mdp5_write(mdp5_kms, REG_MDP5_INTF_TIMING_ENGINE_EN(3), 0);
+	mdp5_disable(mdp5_kms);
+	mdelay(16);
+
+	if (config->iommu) {
+		mmu = msm_iommu_new(dev, config->iommu);
+		if (IS_ERR(mmu)) {
+			ret = PTR_ERR(mmu);
+			goto fail;
+		}
+		ret = mmu->funcs->attach(mmu, iommu_ports,
+				ARRAY_SIZE(iommu_ports));
+		if (ret)
+			goto fail;
+	} else {
+		dev_info(dev->dev, "no iommu, fallback to phys "
+				"contig buffers for scanout\n");
+		mmu = NULL;
+	}
+
+	mdp5_kms->id = msm_register_mmu(dev, mmu);
+	if (mdp5_kms->id < 0) {
+		ret = mdp5_kms->id;
+		dev_err(dev->dev, "failed to register mdp5 iommu: %d\n", ret);
+		goto fail;
+	}
+
+	ret = modeset_init(mdp5_kms);
+	if (ret) {
+		dev_err(dev->dev, "modeset_init failed: %d\n", ret);
+		goto fail;
+	}
+
+	return kms;
+
+fail:
+	if (kms)
+		mdp5_destroy(kms);
+	return ERR_PTR(ret);
+}
+
+static struct mdp5_platform_config *mdp5_get_config(struct platform_device *dev)
+{
+	static struct mdp5_platform_config config = {};
+#ifdef CONFIG_OF
+	/* TODO */
+#endif
+	return &config;
+}
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.h b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.h
new file mode 100644
index 000000000000..c8b1a2522c25
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.h
@@ -0,0 +1,213 @@
+/*
+ * Copyright (C) 2013 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/>.
+ */
+
+#ifndef __MDP5_KMS_H__
+#define __MDP5_KMS_H__
+
+#include "msm_drv.h"
+#include "msm_kms.h"
+#include "mdp/mdp_kms.h"
+#include "mdp5.xml.h"
+#include "mdp5_smp.h"
+
+struct mdp5_kms {
+	struct mdp_kms base;
+
+	struct drm_device *dev;
+
+	int rev;
+
+	/* mapper-id used to request GEM buffer mapped for scanout: */
+	int id;
+
+	/* for tracking smp allocation amongst pipes: */
+	mdp5_smp_state_t smp_state;
+	struct mdp5_client_smp_state smp_client_state[CID_MAX];
+	int smp_blk_cnt;
+
+	/* io/register spaces: */
+	void __iomem *mmio, *vbif;
+
+	struct regulator *vdd;
+
+	struct clk *axi_clk;
+	struct clk *ahb_clk;
+	struct clk *src_clk;
+	struct clk *core_clk;
+	struct clk *lut_clk;
+	struct clk *vsync_clk;
+
+	struct hdmi *hdmi;
+
+	struct mdp_irq error_handler;
+};
+#define to_mdp5_kms(x) container_of(x, struct mdp5_kms, base)
+
+/* platform config data (ie. from DT, or pdata) */
+struct mdp5_platform_config {
+	struct iommu_domain *iommu;
+	uint32_t max_clk;
+	int smp_blk_cnt;
+};
+
+static inline void mdp5_write(struct mdp5_kms *mdp5_kms, u32 reg, u32 data)
+{
+	msm_writel(data, mdp5_kms->mmio + reg);
+}
+
+static inline u32 mdp5_read(struct mdp5_kms *mdp5_kms, u32 reg)
+{
+	return msm_readl(mdp5_kms->mmio + reg);
+}
+
+static inline const char *pipe2name(enum mdp5_pipe pipe)
+{
+	static const char *names[] = {
+#define NAME(n) [SSPP_ ## n] = #n
+		NAME(VIG0), NAME(VIG1), NAME(VIG2),
+		NAME(RGB0), NAME(RGB1), NAME(RGB2),
+		NAME(DMA0), NAME(DMA1),
+#undef NAME
+	};
+	return names[pipe];
+}
+
+static inline uint32_t pipe2flush(enum mdp5_pipe pipe)
+{
+	switch (pipe) {
+	case SSPP_VIG0: return MDP5_CTL_FLUSH_VIG0;
+	case SSPP_VIG1: return MDP5_CTL_FLUSH_VIG1;
+	case SSPP_VIG2: return MDP5_CTL_FLUSH_VIG2;
+	case SSPP_RGB0: return MDP5_CTL_FLUSH_RGB0;
+	case SSPP_RGB1: return MDP5_CTL_FLUSH_RGB1;
+	case SSPP_RGB2: return MDP5_CTL_FLUSH_RGB2;
+	case SSPP_DMA0: return MDP5_CTL_FLUSH_DMA0;
+	case SSPP_DMA1: return MDP5_CTL_FLUSH_DMA1;
+	default:        return 0;
+	}
+}
+
+static inline int pipe2nclients(enum mdp5_pipe pipe)
+{
+	switch (pipe) {
+	case SSPP_RGB0:
+	case SSPP_RGB1:
+	case SSPP_RGB2:
+		return 1;
+	default:
+		return 3;
+	}
+}
+
+static inline enum mdp5_client_id pipe2client(enum mdp5_pipe pipe, int plane)
+{
+	WARN_ON(plane >= pipe2nclients(pipe));
+	switch (pipe) {
+	case SSPP_VIG0: return CID_VIG0_Y + plane;
+	case SSPP_VIG1: return CID_VIG1_Y + plane;
+	case SSPP_VIG2: return CID_VIG2_Y + plane;
+	case SSPP_RGB0: return CID_RGB0;
+	case SSPP_RGB1: return CID_RGB1;
+	case SSPP_RGB2: return CID_RGB2;
+	case SSPP_DMA0: return CID_DMA0_Y + plane;
+	case SSPP_DMA1: return CID_DMA1_Y + plane;
+	default:        return CID_UNUSED;
+	}
+}
+
+static inline uint32_t mixer2flush(int lm)
+{
+	switch (lm) {
+	case 0:  return MDP5_CTL_FLUSH_LM0;
+	case 1:  return MDP5_CTL_FLUSH_LM1;
+	case 2:  return MDP5_CTL_FLUSH_LM2;
+	default: return 0;
+	}
+}
+
+static inline uint32_t intf2err(int intf)
+{
+	switch (intf) {
+	case 0:  return MDP5_IRQ_INTF0_UNDER_RUN;
+	case 1:  return MDP5_IRQ_INTF1_UNDER_RUN;
+	case 2:  return MDP5_IRQ_INTF2_UNDER_RUN;
+	case 3:  return MDP5_IRQ_INTF3_UNDER_RUN;
+	default: return 0;
+	}
+}
+
+static inline uint32_t intf2vblank(int intf)
+{
+	switch (intf) {
+	case 0:  return MDP5_IRQ_INTF0_VSYNC;
+	case 1:  return MDP5_IRQ_INTF1_VSYNC;
+	case 2:  return MDP5_IRQ_INTF2_VSYNC;
+	case 3:  return MDP5_IRQ_INTF3_VSYNC;
+	default: return 0;
+	}
+}
+
+int mdp5_disable(struct mdp5_kms *mdp5_kms);
+int mdp5_enable(struct mdp5_kms *mdp5_kms);
+
+void mdp5_set_irqmask(struct mdp_kms *mdp_kms, uint32_t irqmask);
+void mdp5_irq_preinstall(struct msm_kms *kms);
+int mdp5_irq_postinstall(struct msm_kms *kms);
+void mdp5_irq_uninstall(struct msm_kms *kms);
+irqreturn_t mdp5_irq(struct msm_kms *kms);
+int mdp5_enable_vblank(struct msm_kms *kms, struct drm_crtc *crtc);
+void mdp5_disable_vblank(struct msm_kms *kms, struct drm_crtc *crtc);
+
+static inline
+uint32_t mdp5_get_formats(enum mdp5_pipe pipe, uint32_t *pixel_formats,
+		uint32_t max_formats)
+{
+	/* TODO when we have YUV, we need to filter supported formats
+	 * based on pipe id..
+	 */
+	return mdp_get_formats(pixel_formats, max_formats);
+}
+
+void mdp5_plane_install_properties(struct drm_plane *plane,
+		struct drm_mode_object *obj);
+void mdp5_plane_set_scanout(struct drm_plane *plane,
+		struct drm_framebuffer *fb);
+int mdp5_plane_mode_set(struct drm_plane *plane,
+		struct drm_crtc *crtc, struct drm_framebuffer *fb,
+		int crtc_x, int crtc_y,
+		unsigned int crtc_w, unsigned int crtc_h,
+		uint32_t src_x, uint32_t src_y,
+		uint32_t src_w, uint32_t src_h);
+void mdp5_plane_complete_flip(struct drm_plane *plane);
+enum mdp5_pipe mdp5_plane_pipe(struct drm_plane *plane);
+struct drm_plane *mdp5_plane_init(struct drm_device *dev,
+		enum mdp5_pipe pipe, bool private_plane);
+
+uint32_t mdp5_crtc_vblank(struct drm_crtc *crtc);
+
+void mdp5_crtc_cancel_pending_flip(struct drm_crtc *crtc, struct drm_file *file);
+void mdp5_crtc_set_intf(struct drm_crtc *crtc, int intf,
+		enum mdp5_intf intf_id);
+void mdp5_crtc_attach(struct drm_crtc *crtc, struct drm_plane *plane);
+void mdp5_crtc_detach(struct drm_crtc *crtc, struct drm_plane *plane);
+struct drm_crtc *mdp5_crtc_init(struct drm_device *dev,
+		struct drm_plane *plane, int id);
+
+struct drm_encoder *mdp5_encoder_init(struct drm_device *dev, int intf,
+		enum mdp5_intf intf_id);
+
+#endif /* __MDP5_KMS_H__ */
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_plane.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_plane.c
new file mode 100644
index 000000000000..0ac8bb5e7e85
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_plane.c
@@ -0,0 +1,389 @@
+/*
+ * Copyright (C) 2013 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 "mdp5_kms.h"
+
+
+struct mdp5_plane {
+	struct drm_plane base;
+	const char *name;
+
+	enum mdp5_pipe pipe;
+
+	uint32_t nformats;
+	uint32_t formats[32];
+
+	bool enabled;
+};
+#define to_mdp5_plane(x) container_of(x, struct mdp5_plane, base)
+
+static struct mdp5_kms *get_kms(struct drm_plane *plane)
+{
+	struct msm_drm_private *priv = plane->dev->dev_private;
+	return to_mdp5_kms(to_mdp_kms(priv->kms));
+}
+
+static int mdp5_plane_update(struct drm_plane *plane,
+		struct drm_crtc *crtc, struct drm_framebuffer *fb,
+		int crtc_x, int crtc_y,
+		unsigned int crtc_w, unsigned int crtc_h,
+		uint32_t src_x, uint32_t src_y,
+		uint32_t src_w, uint32_t src_h)
+{
+	struct mdp5_plane *mdp5_plane = to_mdp5_plane(plane);
+
+	mdp5_plane->enabled = true;
+
+	if (plane->fb)
+		drm_framebuffer_unreference(plane->fb);
+
+	drm_framebuffer_reference(fb);
+
+	return mdp5_plane_mode_set(plane, crtc, fb,
+			crtc_x, crtc_y, crtc_w, crtc_h,
+			src_x, src_y, src_w, src_h);
+}
+
+static int mdp5_plane_disable(struct drm_plane *plane)
+{
+	struct mdp5_plane *mdp5_plane = to_mdp5_plane(plane);
+	struct mdp5_kms *mdp5_kms = get_kms(plane);
+	enum mdp5_pipe pipe = mdp5_plane->pipe;
+	int i;
+
+	DBG("%s: disable", mdp5_plane->name);
+
+	/* update our SMP request to zero (release all our blks): */
+	for (i = 0; i < pipe2nclients(pipe); i++)
+		mdp5_smp_request(mdp5_kms, pipe2client(pipe, i), 0);
+
+	/* TODO detaching now will cause us not to get the last
+	 * vblank and mdp5_smp_commit().. so other planes will
+	 * still see smp blocks previously allocated to us as
+	 * in-use..
+	 */
+	if (plane->crtc)
+		mdp5_crtc_detach(plane->crtc, plane);
+
+	return 0;
+}
+
+static void mdp5_plane_destroy(struct drm_plane *plane)
+{
+	struct mdp5_plane *mdp5_plane = to_mdp5_plane(plane);
+
+	mdp5_plane_disable(plane);
+	drm_plane_cleanup(plane);
+
+	kfree(mdp5_plane);
+}
+
+/* helper to install properties which are common to planes and crtcs */
+void mdp5_plane_install_properties(struct drm_plane *plane,
+		struct drm_mode_object *obj)
+{
+	// XXX
+}
+
+int mdp5_plane_set_property(struct drm_plane *plane,
+		struct drm_property *property, uint64_t val)
+{
+	// XXX
+	return -EINVAL;
+}
+
+static const struct drm_plane_funcs mdp5_plane_funcs = {
+		.update_plane = mdp5_plane_update,
+		.disable_plane = mdp5_plane_disable,
+		.destroy = mdp5_plane_destroy,
+		.set_property = mdp5_plane_set_property,
+};
+
+void mdp5_plane_set_scanout(struct drm_plane *plane,
+		struct drm_framebuffer *fb)
+{
+	struct mdp5_plane *mdp5_plane = to_mdp5_plane(plane);
+	struct mdp5_kms *mdp5_kms = get_kms(plane);
+	enum mdp5_pipe pipe = mdp5_plane->pipe;
+	uint32_t nplanes = drm_format_num_planes(fb->pixel_format);
+	uint32_t iova[4];
+	int i;
+
+	for (i = 0; i < nplanes; i++) {
+		struct drm_gem_object *bo = msm_framebuffer_bo(fb, i);
+		msm_gem_get_iova(bo, mdp5_kms->id, &iova[i]);
+	}
+	for (; i < 4; i++)
+		iova[i] = 0;
+
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC_STRIDE_A(pipe),
+			MDP5_PIPE_SRC_STRIDE_A_P0(fb->pitches[0]) |
+			MDP5_PIPE_SRC_STRIDE_A_P1(fb->pitches[1]));
+
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC_STRIDE_B(pipe),
+			MDP5_PIPE_SRC_STRIDE_B_P2(fb->pitches[2]) |
+			MDP5_PIPE_SRC_STRIDE_B_P3(fb->pitches[3]));
+
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC0_ADDR(pipe), iova[0]);
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC1_ADDR(pipe), iova[1]);
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC2_ADDR(pipe), iova[2]);
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC3_ADDR(pipe), iova[3]);
+
+	plane->fb = fb;
+}
+
+/* NOTE: looks like if horizontal decimation is used (if we supported that)
+ * then the width used to calculate SMP block requirements is the post-
+ * decimated width.  Ie. SMP buffering sits downstream of decimation (which
+ * presumably happens during the dma from scanout buffer).
+ */
+static int request_smp_blocks(struct drm_plane *plane, uint32_t format,
+		uint32_t nplanes, uint32_t width)
+{
+	struct drm_device *dev = plane->dev;
+	struct mdp5_plane *mdp5_plane = to_mdp5_plane(plane);
+	struct mdp5_kms *mdp5_kms = get_kms(plane);
+	enum mdp5_pipe pipe = mdp5_plane->pipe;
+	int i, hsub, nlines, nblks, ret;
+
+	hsub = drm_format_horz_chroma_subsampling(format);
+
+	/* different if BWC (compressed framebuffer?) enabled: */
+	nlines = 2;
+
+	for (i = 0, nblks = 0; i < nplanes; i++) {
+		int n, fetch_stride, cpp;
+
+		cpp = drm_format_plane_cpp(format, i);
+		fetch_stride = width * cpp / (i ? hsub : 1);
+
+		n = DIV_ROUND_UP(fetch_stride * nlines, SMP_BLK_SIZE);
+
+		/* for hw rev v1.00 */
+		if (mdp5_kms->rev == 0)
+			n = roundup_pow_of_two(n);
+
+		DBG("%s[%d]: request %d SMP blocks", mdp5_plane->name, i, n);
+		ret = mdp5_smp_request(mdp5_kms, pipe2client(pipe, i), n);
+		if (ret) {
+			dev_err(dev->dev, "Could not allocate %d SMP blocks: %d\n",
+					n, ret);
+			return ret;
+		}
+
+		nblks += n;
+	}
+
+	/* in success case, return total # of blocks allocated: */
+	return nblks;
+}
+
+static void set_fifo_thresholds(struct drm_plane *plane, int nblks)
+{
+	struct mdp5_plane *mdp5_plane = to_mdp5_plane(plane);
+	struct mdp5_kms *mdp5_kms = get_kms(plane);
+	enum mdp5_pipe pipe = mdp5_plane->pipe;
+	uint32_t val;
+
+	/* 1/4 of SMP pool that is being fetched */
+	val = (nblks * SMP_ENTRIES_PER_BLK) / 4;
+
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_REQPRIO_FIFO_WM_0(pipe), val * 1);
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_REQPRIO_FIFO_WM_1(pipe), val * 2);
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_REQPRIO_FIFO_WM_2(pipe), val * 3);
+
+}
+
+int mdp5_plane_mode_set(struct drm_plane *plane,
+		struct drm_crtc *crtc, struct drm_framebuffer *fb,
+		int crtc_x, int crtc_y,
+		unsigned int crtc_w, unsigned int crtc_h,
+		uint32_t src_x, uint32_t src_y,
+		uint32_t src_w, uint32_t src_h)
+{
+	struct mdp5_plane *mdp5_plane = to_mdp5_plane(plane);
+	struct mdp5_kms *mdp5_kms = get_kms(plane);
+	enum mdp5_pipe pipe = mdp5_plane->pipe;
+	const struct mdp_format *format;
+	uint32_t nplanes, config = 0;
+	uint32_t phasex_step = 0, phasey_step = 0;
+	uint32_t hdecm = 0, vdecm = 0;
+	int i, nblks;
+
+	nplanes = drm_format_num_planes(fb->pixel_format);
+
+	/* bad formats should already be rejected: */
+	if (WARN_ON(nplanes > pipe2nclients(pipe)))
+		return -EINVAL;
+
+	/* src values are in Q16 fixed point, convert to integer: */
+	src_x = src_x >> 16;
+	src_y = src_y >> 16;
+	src_w = src_w >> 16;
+	src_h = src_h >> 16;
+
+	DBG("%s: FB[%u] %u,%u,%u,%u -> CRTC[%u] %d,%d,%u,%u", mdp5_plane->name,
+			fb->base.id, src_x, src_y, src_w, src_h,
+			crtc->base.id, crtc_x, crtc_y, crtc_w, crtc_h);
+
+	/*
+	 * Calculate and request required # of smp blocks:
+	 */
+	nblks = request_smp_blocks(plane, fb->pixel_format, nplanes, src_w);
+	if (nblks < 0)
+		return nblks;
+
+	/*
+	 * Currently we update the hw for allocations/requests immediately,
+	 * but once atomic modeset/pageflip is in place, the allocation
+	 * would move into atomic->check_plane_state(), while updating the
+	 * hw would remain here:
+	 */
+	for (i = 0; i < pipe2nclients(pipe); i++)
+		mdp5_smp_configure(mdp5_kms, pipe2client(pipe, i));
+
+	if (src_w != crtc_w) {
+		config |= MDP5_PIPE_SCALE_CONFIG_SCALEX_EN;
+		/* TODO calc phasex_step, hdecm */
+	}
+
+	if (src_h != crtc_h) {
+		config |= MDP5_PIPE_SCALE_CONFIG_SCALEY_EN;
+		/* TODO calc phasey_step, vdecm */
+	}
+
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC_IMG_SIZE(pipe),
+			MDP5_PIPE_SRC_IMG_SIZE_WIDTH(src_w) |
+			MDP5_PIPE_SRC_IMG_SIZE_HEIGHT(src_h));
+
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC_SIZE(pipe),
+			MDP5_PIPE_SRC_SIZE_WIDTH(src_w) |
+			MDP5_PIPE_SRC_SIZE_HEIGHT(src_h));
+
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC_XY(pipe),
+			MDP5_PIPE_SRC_XY_X(src_x) |
+			MDP5_PIPE_SRC_XY_Y(src_y));
+
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_OUT_SIZE(pipe),
+			MDP5_PIPE_OUT_SIZE_WIDTH(crtc_w) |
+			MDP5_PIPE_OUT_SIZE_HEIGHT(crtc_h));
+
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_OUT_XY(pipe),
+			MDP5_PIPE_OUT_XY_X(crtc_x) |
+			MDP5_PIPE_OUT_XY_Y(crtc_y));
+
+	mdp5_plane_set_scanout(plane, fb);
+
+	format = to_mdp_format(msm_framebuffer_format(fb));
+
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC_FORMAT(pipe),
+			MDP5_PIPE_SRC_FORMAT_A_BPC(format->bpc_a) |
+			MDP5_PIPE_SRC_FORMAT_R_BPC(format->bpc_r) |
+			MDP5_PIPE_SRC_FORMAT_G_BPC(format->bpc_g) |
+			MDP5_PIPE_SRC_FORMAT_B_BPC(format->bpc_b) |
+			COND(format->alpha_enable, MDP5_PIPE_SRC_FORMAT_ALPHA_ENABLE) |
+			MDP5_PIPE_SRC_FORMAT_CPP(format->cpp - 1) |
+			MDP5_PIPE_SRC_FORMAT_UNPACK_COUNT(format->unpack_count - 1) |
+			COND(format->unpack_tight, MDP5_PIPE_SRC_FORMAT_UNPACK_TIGHT) |
+			MDP5_PIPE_SRC_FORMAT_NUM_PLANES(nplanes - 1) |
+			MDP5_PIPE_SRC_FORMAT_CHROMA_SAMP(CHROMA_RGB));
+
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC_UNPACK(pipe),
+			MDP5_PIPE_SRC_UNPACK_ELEM0(format->unpack[0]) |
+			MDP5_PIPE_SRC_UNPACK_ELEM1(format->unpack[1]) |
+			MDP5_PIPE_SRC_UNPACK_ELEM2(format->unpack[2]) |
+			MDP5_PIPE_SRC_UNPACK_ELEM3(format->unpack[3]));
+
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC_OP_MODE(pipe),
+			MDP5_PIPE_SRC_OP_MODE_BWC(BWC_LOSSLESS));
+
+	/* not using secure mode: */
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SRC_ADDR_SW_STATUS(pipe), 0);
+
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SCALE_PHASE_STEP_X(pipe), phasex_step);
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SCALE_PHASE_STEP_Y(pipe), phasey_step);
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_DECIMATION(pipe),
+			MDP5_PIPE_DECIMATION_VERT(vdecm) |
+			MDP5_PIPE_DECIMATION_HORZ(hdecm));
+	mdp5_write(mdp5_kms, REG_MDP5_PIPE_SCALE_CONFIG(pipe),
+			MDP5_PIPE_SCALE_CONFIG_SCALEX_MIN_FILTER(SCALE_FILTER_NEAREST) |
+			MDP5_PIPE_SCALE_CONFIG_SCALEY_MIN_FILTER(SCALE_FILTER_NEAREST) |
+			MDP5_PIPE_SCALE_CONFIG_SCALEX_CR_FILTER(SCALE_FILTER_NEAREST) |
+			MDP5_PIPE_SCALE_CONFIG_SCALEY_CR_FILTER(SCALE_FILTER_NEAREST) |
+			MDP5_PIPE_SCALE_CONFIG_SCALEX_MAX_FILTER(SCALE_FILTER_NEAREST) |
+			MDP5_PIPE_SCALE_CONFIG_SCALEY_MAX_FILTER(SCALE_FILTER_NEAREST));
+
+	set_fifo_thresholds(plane, nblks);
+
+	/* TODO detach from old crtc (if we had more than one) */
+	mdp5_crtc_attach(crtc, plane);
+
+	return 0;
+}
+
+void mdp5_plane_complete_flip(struct drm_plane *plane)
+{
+	struct mdp5_kms *mdp5_kms = get_kms(plane);
+	enum mdp5_pipe pipe = to_mdp5_plane(plane)->pipe;
+	int i;
+
+	for (i = 0; i < pipe2nclients(pipe); i++)
+		mdp5_smp_commit(mdp5_kms, pipe2client(pipe, i));
+}
+
+enum mdp5_pipe mdp5_plane_pipe(struct drm_plane *plane)
+{
+	struct mdp5_plane *mdp5_plane = to_mdp5_plane(plane);
+	return mdp5_plane->pipe;
+}
+
+/* initialize plane */
+struct drm_plane *mdp5_plane_init(struct drm_device *dev,
+		enum mdp5_pipe pipe, bool private_plane)
+{
+	struct drm_plane *plane = NULL;
+	struct mdp5_plane *mdp5_plane;
+	int ret;
+
+	mdp5_plane = kzalloc(sizeof(*mdp5_plane), GFP_KERNEL);
+	if (!mdp5_plane) {
+		ret = -ENOMEM;
+		goto fail;
+	}
+
+	plane = &mdp5_plane->base;
+
+	mdp5_plane->pipe = pipe;
+	mdp5_plane->name = pipe2name(pipe);
+
+	mdp5_plane->nformats = mdp5_get_formats(pipe, mdp5_plane->formats,
+			ARRAY_SIZE(mdp5_plane->formats));
+
+	drm_plane_init(dev, plane, 0xff, &mdp5_plane_funcs,
+			mdp5_plane->formats, mdp5_plane->nformats,
+			private_plane);
+
+	mdp5_plane_install_properties(plane, &plane->base);
+
+	return plane;
+
+fail:
+	if (plane)
+		mdp5_plane_destroy(plane);
+
+	return ERR_PTR(ret);
+}
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_smp.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_smp.c
new file mode 100644
index 000000000000..2d0236b963a6
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_smp.c
@@ -0,0 +1,173 @@
+/*
+ * Copyright (C) 2013 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 "mdp5_kms.h"
+#include "mdp5_smp.h"
+
+
+/* SMP - Shared Memory Pool
+ *
+ * These are shared between all the clients, where each plane in a
+ * scanout buffer is a SMP client.  Ie. scanout of 3 plane I420 on
+ * pipe VIG0 => 3 clients: VIG0_Y, VIG0_CB, VIG0_CR.
+ *
+ * Based on the size of the attached scanout buffer, a certain # of
+ * blocks must be allocated to that client out of the shared pool.
+ *
+ * For each block, it can be either free, or pending/in-use by a
+ * client.  The updates happen in three steps:
+ *
+ *  1) mdp5_smp_request():
+ *     When plane scanout is setup, calculate required number of
+ *     blocks needed per client, and request.  Blocks not inuse or
+ *     pending by any other client are added to client's pending
+ *     set.
+ *
+ *  2) mdp5_smp_configure():
+ *     As hw is programmed, before FLUSH, MDP5_SMP_ALLOC registers
+ *     are configured for the union(pending, inuse)
+ *
+ *  3) mdp5_smp_commit():
+ *     After next vblank, copy pending -> inuse.  Optionally update
+ *     MDP5_SMP_ALLOC registers if there are newly unused blocks
+ *
+ * On the next vblank after changes have been committed to hw, the
+ * client's pending blocks become it's in-use blocks (and no-longer
+ * in-use blocks become available to other clients).
+ *
+ * btw, hurray for confusing overloaded acronyms!  :-/
+ *
+ * NOTE: for atomic modeset/pageflip NONBLOCK operations, step #1
+ * should happen at (or before)? atomic->check().  And we'd need
+ * an API to discard previous requests if update is aborted or
+ * (test-only).
+ *
+ * TODO would perhaps be nice to have debugfs to dump out kernel
+ * inuse and pending state of all clients..
+ */
+
+static DEFINE_SPINLOCK(smp_lock);
+
+
+/* step #1: update # of blocks pending for the client: */
+int mdp5_smp_request(struct mdp5_kms *mdp5_kms,
+		enum mdp5_client_id cid, int nblks)
+{
+	struct mdp5_client_smp_state *ps = &mdp5_kms->smp_client_state[cid];
+	int i, ret, avail, cur_nblks, cnt = mdp5_kms->smp_blk_cnt;
+	unsigned long flags;
+
+	spin_lock_irqsave(&smp_lock, flags);
+
+	avail = cnt - bitmap_weight(mdp5_kms->smp_state, cnt);
+	if (nblks > avail) {
+		ret = -ENOSPC;
+		goto fail;
+	}
+
+	cur_nblks = bitmap_weight(ps->pending, cnt);
+	if (nblks > cur_nblks) {
+		/* grow the existing pending reservation: */
+		for (i = cur_nblks; i < nblks; i++) {
+			int blk = find_first_zero_bit(mdp5_kms->smp_state, cnt);
+			set_bit(blk, ps->pending);
+			set_bit(blk, mdp5_kms->smp_state);
+		}
+	} else {
+		/* shrink the existing pending reservation: */
+		for (i = cur_nblks; i > nblks; i--) {
+			int blk = find_first_bit(ps->pending, cnt);
+			clear_bit(blk, ps->pending);
+			/* don't clear in global smp_state until _commit() */
+		}
+	}
+
+fail:
+	spin_unlock_irqrestore(&smp_lock, flags);
+	return 0;
+}
+
+static void update_smp_state(struct mdp5_kms *mdp5_kms,
+		enum mdp5_client_id cid, mdp5_smp_state_t *assigned)
+{
+	int cnt = mdp5_kms->smp_blk_cnt;
+	uint32_t blk, val;
+
+	for_each_set_bit(blk, *assigned, cnt) {
+		int idx = blk / 3;
+		int fld = blk % 3;
+
+		val = mdp5_read(mdp5_kms, REG_MDP5_SMP_ALLOC_W_REG(idx));
+
+		switch (fld) {
+		case 0:
+			val &= ~MDP5_SMP_ALLOC_W_REG_CLIENT0__MASK;
+			val |= MDP5_SMP_ALLOC_W_REG_CLIENT0(cid);
+			break;
+		case 1:
+			val &= ~MDP5_SMP_ALLOC_W_REG_CLIENT1__MASK;
+			val |= MDP5_SMP_ALLOC_W_REG_CLIENT1(cid);
+			break;
+		case 2:
+			val &= ~MDP5_SMP_ALLOC_W_REG_CLIENT2__MASK;
+			val |= MDP5_SMP_ALLOC_W_REG_CLIENT2(cid);
+			break;
+		}
+
+		mdp5_write(mdp5_kms, REG_MDP5_SMP_ALLOC_W_REG(idx), val);
+		mdp5_write(mdp5_kms, REG_MDP5_SMP_ALLOC_R_REG(idx), val);
+	}
+}
+
+/* step #2: configure hw for union(pending, inuse): */
+void mdp5_smp_configure(struct mdp5_kms *mdp5_kms, enum mdp5_client_id cid)
+{
+	struct mdp5_client_smp_state *ps = &mdp5_kms->smp_client_state[cid];
+	int cnt = mdp5_kms->smp_blk_cnt;
+	mdp5_smp_state_t assigned;
+
+	bitmap_or(assigned, ps->inuse, ps->pending, cnt);
+	update_smp_state(mdp5_kms, cid, &assigned);
+}
+
+/* step #3: after vblank, copy pending -> inuse: */
+void mdp5_smp_commit(struct mdp5_kms *mdp5_kms, enum mdp5_client_id cid)
+{
+	struct mdp5_client_smp_state *ps = &mdp5_kms->smp_client_state[cid];
+	int cnt = mdp5_kms->smp_blk_cnt;
+	mdp5_smp_state_t released;
+
+	/*
+	 * Figure out if there are any blocks we where previously
+	 * using, which can be released and made available to other
+	 * clients:
+	 */
+	if (bitmap_andnot(released, ps->inuse, ps->pending, cnt)) {
+		unsigned long flags;
+
+		spin_lock_irqsave(&smp_lock, flags);
+		/* clear released blocks: */
+		bitmap_andnot(mdp5_kms->smp_state, mdp5_kms->smp_state,
+				released, cnt);
+		spin_unlock_irqrestore(&smp_lock, flags);
+
+		update_smp_state(mdp5_kms, CID_UNUSED, &released);
+	}
+
+	bitmap_copy(ps->inuse, ps->pending, cnt);
+}
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_smp.h b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_smp.h
new file mode 100644
index 000000000000..0ab739e1a1dd
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_smp.h
@@ -0,0 +1,41 @@
+/*
+ * Copyright (C) 2013 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/>.
+ */
+
+#ifndef __MDP5_SMP_H__
+#define __MDP5_SMP_H__
+
+#include "msm_drv.h"
+
+#define MAX_SMP_BLOCKS  22
+#define SMP_BLK_SIZE    4096
+#define SMP_ENTRIES_PER_BLK (SMP_BLK_SIZE / 16)
+
+typedef DECLARE_BITMAP(mdp5_smp_state_t, MAX_SMP_BLOCKS);
+
+struct mdp5_client_smp_state {
+	mdp5_smp_state_t inuse;
+	mdp5_smp_state_t pending;
+};
+
+struct mdp5_kms;
+
+int mdp5_smp_request(struct mdp5_kms *mdp5_kms, enum mdp5_client_id cid, int nblks);
+void mdp5_smp_configure(struct mdp5_kms *mdp5_kms, enum mdp5_client_id cid);
+void mdp5_smp_commit(struct mdp5_kms *mdp5_kms, enum mdp5_client_id cid);
+
+
+#endif /* __MDP5_SMP_H__ */
diff --git a/drivers/gpu/drm/msm/mdp/mdp_common.xml.h b/drivers/gpu/drm/msm/mdp/mdp_common.xml.h
new file mode 100644
index 000000000000..a9629b85b983
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp_common.xml.h
@@ -0,0 +1,78 @@
+#ifndef MDP_COMMON_XML
+#define MDP_COMMON_XML
+
+/* Autogenerated file, DO NOT EDIT manually!
+
+This file was generated by the rules-ng-ng headergen tool in this git repository:
+http://github.com/freedreno/envytools/
+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 2013-12-03 20:59:13)
+- /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/hdmi/qfprom.xml         (    600 bytes, from 2013-07-05 19:21:12)
+- /home/robclark/src/freedreno/envytools/rnndb/hdmi/hdmi.xml           (  20932 bytes, from 2013-12-01 15:13:04)
+
+Copyright (C) 2013 by the following authors:
+- Rob Clark <robdclark@gmail.com> (robclark)
+
+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 COPYRIGHT OWNER(S) 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.
+*/
+
+
+enum mdp_mixer_stage_id {
+	STAGE_UNUSED = 0,
+	STAGE_BASE = 1,
+	STAGE0 = 2,
+	STAGE1 = 3,
+	STAGE2 = 4,
+	STAGE3 = 5,
+};
+
+enum mdp_alpha_type {
+	FG_CONST = 0,
+	BG_CONST = 1,
+	FG_PIXEL = 2,
+	BG_PIXEL = 3,
+};
+
+enum mdp_bpc {
+	BPC1 = 0,
+	BPC5 = 1,
+	BPC6 = 2,
+	BPC8 = 3,
+};
+
+enum mdp_bpc_alpha {
+	BPC1A = 0,
+	BPC4A = 1,
+	BPC6A = 2,
+	BPC8A = 3,
+};
+
+
+#endif /* MDP_COMMON_XML */
diff --git a/drivers/gpu/drm/msm/mdp4/mdp4_format.c b/drivers/gpu/drm/msm/mdp/mdp_format.c
index 17330b0927b2..e0a6ffbe6ab4 100644
--- a/drivers/gpu/drm/msm/mdp4/mdp4_format.c
+++ b/drivers/gpu/drm/msm/mdp/mdp_format.c
@@ -17,7 +17,7 @@
 
 
 #include "msm_drv.h"
-#include "mdp4_kms.h"
+#include "mdp_kms.h"
 
 #define FMT(name, a, r, g, b, e0, e1, e2, e3, alpha, tight, c, cnt) { \
 		.base = { .pixel_format = DRM_FORMAT_ ## name }, \
@@ -34,7 +34,7 @@
 
 #define BPC0A 0
 
-static const struct mdp4_format formats[] = {
+static const struct mdp_format formats[] = {
 	/*  name      a  r  g  b   e0 e1 e2 e3  alpha   tight  cpp cnt */
 	FMT(ARGB8888, 8, 8, 8, 8,  1, 0, 2, 3,  true,   true,  4,  4),
 	FMT(XRGB8888, 8, 8, 8, 8,  1, 0, 2, 3,  false,  true,  4,  4),
@@ -44,12 +44,11 @@ static const struct mdp4_format formats[] = {
 	FMT(BGR565,   0, 5, 6, 5,  2, 0, 1, 0,  false,  true,  2,  3),
 };
 
-uint32_t mdp4_get_formats(enum mdp4_pipe pipe_id, uint32_t *pixel_formats,
-		uint32_t max_formats)
+uint32_t mdp_get_formats(uint32_t *pixel_formats, uint32_t max_formats)
 {
 	uint32_t i;
 	for (i = 0; i < ARRAY_SIZE(formats); i++) {
-		const struct mdp4_format *f = &formats[i];
+		const struct mdp_format *f = &formats[i];
 
 		if (i == max_formats)
 			break;
@@ -60,11 +59,11 @@ uint32_t mdp4_get_formats(enum mdp4_pipe pipe_id, uint32_t *pixel_formats,
 	return i;
 }
 
-const struct msm_format *mdp4_get_format(struct msm_kms *kms, uint32_t format)
+const struct msm_format *mdp_get_format(struct msm_kms *kms, uint32_t format)
 {
 	int i;
 	for (i = 0; i < ARRAY_SIZE(formats); i++) {
-		const struct mdp4_format *f = &formats[i];
+		const struct mdp_format *f = &formats[i];
 		if (f->base.pixel_format == format)
 			return &f->base;
 	}
diff --git a/drivers/gpu/drm/msm/mdp/mdp_kms.c b/drivers/gpu/drm/msm/mdp/mdp_kms.c
new file mode 100644
index 000000000000..3be48f7c36be
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp_kms.c
@@ -0,0 +1,144 @@
+/*
+ * Copyright (C) 2013 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 "msm_drv.h"
+#include "mdp_kms.h"
+
+
+struct mdp_irq_wait {
+	struct mdp_irq irq;
+	int count;
+};
+
+static DECLARE_WAIT_QUEUE_HEAD(wait_event);
+
+static DEFINE_SPINLOCK(list_lock);
+
+static void update_irq(struct mdp_kms *mdp_kms)
+{
+	struct mdp_irq *irq;
+	uint32_t irqmask = mdp_kms->vblank_mask;
+
+	BUG_ON(!spin_is_locked(&list_lock));
+
+	list_for_each_entry(irq, &mdp_kms->irq_list, node)
+		irqmask |= irq->irqmask;
+
+	mdp_kms->funcs->set_irqmask(mdp_kms, irqmask);
+}
+
+static void update_irq_unlocked(struct mdp_kms *mdp_kms)
+{
+	unsigned long flags;
+	spin_lock_irqsave(&list_lock, flags);
+	update_irq(mdp_kms);
+	spin_unlock_irqrestore(&list_lock, flags);
+}
+
+void mdp_dispatch_irqs(struct mdp_kms *mdp_kms, uint32_t status)
+{
+	struct mdp_irq *handler, *n;
+	unsigned long flags;
+
+	spin_lock_irqsave(&list_lock, flags);
+	mdp_kms->in_irq = true;
+	list_for_each_entry_safe(handler, n, &mdp_kms->irq_list, node) {
+		if (handler->irqmask & status) {
+			spin_unlock_irqrestore(&list_lock, flags);
+			handler->irq(handler, handler->irqmask & status);
+			spin_lock_irqsave(&list_lock, flags);
+		}
+	}
+	mdp_kms->in_irq = false;
+	update_irq(mdp_kms);
+	spin_unlock_irqrestore(&list_lock, flags);
+
+}
+
+void mdp_update_vblank_mask(struct mdp_kms *mdp_kms, uint32_t mask, bool enable)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&list_lock, flags);
+	if (enable)
+		mdp_kms->vblank_mask |= mask;
+	else
+		mdp_kms->vblank_mask &= ~mask;
+	update_irq(mdp_kms);
+	spin_unlock_irqrestore(&list_lock, flags);
+}
+
+static void wait_irq(struct mdp_irq *irq, uint32_t irqstatus)
+{
+	struct mdp_irq_wait *wait =
+			container_of(irq, struct mdp_irq_wait, irq);
+	wait->count--;
+	wake_up_all(&wait_event);
+}
+
+void mdp_irq_wait(struct mdp_kms *mdp_kms, uint32_t irqmask)
+{
+	struct mdp_irq_wait wait = {
+		.irq = {
+			.irq = wait_irq,
+			.irqmask = irqmask,
+		},
+		.count = 1,
+	};
+	mdp_irq_register(mdp_kms, &wait.irq);
+	wait_event(wait_event, (wait.count <= 0));
+	mdp_irq_unregister(mdp_kms, &wait.irq);
+}
+
+void mdp_irq_register(struct mdp_kms *mdp_kms, struct mdp_irq *irq)
+{
+	unsigned long flags;
+	bool needs_update = false;
+
+	spin_lock_irqsave(&list_lock, flags);
+
+	if (!irq->registered) {
+		irq->registered = true;
+		list_add(&irq->node, &mdp_kms->irq_list);
+		needs_update = !mdp_kms->in_irq;
+	}
+
+	spin_unlock_irqrestore(&list_lock, flags);
+
+	if (needs_update)
+		update_irq_unlocked(mdp_kms);
+}
+
+void mdp_irq_unregister(struct mdp_kms *mdp_kms, struct mdp_irq *irq)
+{
+	unsigned long flags;
+	bool needs_update = false;
+
+	spin_lock_irqsave(&list_lock, flags);
+
+	if (irq->registered) {
+		irq->registered = false;
+		list_del(&irq->node);
+		needs_update = !mdp_kms->in_irq;
+	}
+
+	spin_unlock_irqrestore(&list_lock, flags);
+
+	if (needs_update)
+		update_irq_unlocked(mdp_kms);
+}
diff --git a/drivers/gpu/drm/msm/mdp/mdp_kms.h b/drivers/gpu/drm/msm/mdp/mdp_kms.h
new file mode 100644
index 000000000000..99557b5ad4fd
--- /dev/null
+++ b/drivers/gpu/drm/msm/mdp/mdp_kms.h
@@ -0,0 +1,97 @@
+/*
+ * Copyright (C) 2013 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/>.
+ */
+
+#ifndef __MDP_KMS_H__
+#define __MDP_KMS_H__
+
+#include <linux/clk.h>
+#include <linux/platform_device.h>
+#include <linux/regulator/consumer.h>
+
+#include "msm_drv.h"
+#include "msm_kms.h"
+#include "mdp_common.xml.h"
+
+struct mdp_kms;
+
+struct mdp_kms_funcs {
+	struct msm_kms_funcs base;
+	void (*set_irqmask)(struct mdp_kms *mdp_kms, uint32_t irqmask);
+};
+
+struct mdp_kms {
+	struct msm_kms base;
+
+	const struct mdp_kms_funcs *funcs;
+
+	/* irq handling: */
+	bool in_irq;
+	struct list_head irq_list;    /* list of mdp4_irq */
+	uint32_t vblank_mask;         /* irq bits set for userspace vblank */
+};
+#define to_mdp_kms(x) container_of(x, struct mdp_kms, base)
+
+static inline void mdp_kms_init(struct mdp_kms *mdp_kms,
+		const struct mdp_kms_funcs *funcs)
+{
+	mdp_kms->funcs = funcs;
+	INIT_LIST_HEAD(&mdp_kms->irq_list);
+	msm_kms_init(&mdp_kms->base, &funcs->base);
+}
+
+/*
+ * irq helpers:
+ */
+
+/* For transiently registering for different MDP irqs that various parts
+ * of the KMS code need during setup/configuration.  These are not
+ * necessarily the same as what drm_vblank_get/put() are requesting, and
+ * the hysteresis in drm_vblank_put() is not necessarily desirable for
+ * internal housekeeping related irq usage.
+ */
+struct mdp_irq {
+	struct list_head node;
+	uint32_t irqmask;
+	bool registered;
+	void (*irq)(struct mdp_irq *irq, uint32_t irqstatus);
+};
+
+void mdp_dispatch_irqs(struct mdp_kms *mdp_kms, uint32_t status);
+void mdp_update_vblank_mask(struct mdp_kms *mdp_kms, uint32_t mask, bool enable);
+void mdp_irq_wait(struct mdp_kms *mdp_kms, uint32_t irqmask);
+void mdp_irq_register(struct mdp_kms *mdp_kms, struct mdp_irq *irq);
+void mdp_irq_unregister(struct mdp_kms *mdp_kms, struct mdp_irq *irq);
+
+
+/*
+ * pixel format helpers:
+ */
+
+struct mdp_format {
+	struct msm_format base;
+	enum mdp_bpc bpc_r, bpc_g, bpc_b;
+	enum mdp_bpc_alpha bpc_a;
+	uint8_t unpack[4];
+	bool alpha_enable, unpack_tight;
+	uint8_t cpp, unpack_count;
+};
+#define to_mdp_format(x) container_of(x, struct mdp_format, base)
+
+uint32_t mdp_get_formats(uint32_t *formats, uint32_t max_formats);
+const struct msm_format *mdp_get_format(struct msm_kms *kms, uint32_t format);
+
+#endif /* __MDP_KMS_H__ */
diff --git a/drivers/gpu/drm/msm/mdp4/mdp4_irq.c b/drivers/gpu/drm/msm/mdp4/mdp4_irq.c
deleted file mode 100644
index 5c6b7fca4edd..000000000000
--- a/drivers/gpu/drm/msm/mdp4/mdp4_irq.c
+++ /dev/null
@@ -1,203 +0,0 @@
-/*
- * Copyright (C) 2013 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 "msm_drv.h"
-#include "mdp4_kms.h"
-
-
-struct mdp4_irq_wait {
-	struct mdp4_irq irq;
-	int count;
-};
-
-static DECLARE_WAIT_QUEUE_HEAD(wait_event);
-
-static DEFINE_SPINLOCK(list_lock);
-
-static void update_irq(struct mdp4_kms *mdp4_kms)
-{
-	struct mdp4_irq *irq;
-	uint32_t irqmask = mdp4_kms->vblank_mask;
-
-	BUG_ON(!spin_is_locked(&list_lock));
-
-	list_for_each_entry(irq, &mdp4_kms->irq_list, node)
-		irqmask |= irq->irqmask;
-
-	mdp4_write(mdp4_kms, REG_MDP4_INTR_ENABLE, irqmask);
-}
-
-static void update_irq_unlocked(struct mdp4_kms *mdp4_kms)
-{
-	unsigned long flags;
-	spin_lock_irqsave(&list_lock, flags);
-	update_irq(mdp4_kms);
-	spin_unlock_irqrestore(&list_lock, flags);
-}
-
-static void mdp4_irq_error_handler(struct mdp4_irq *irq, uint32_t irqstatus)
-{
-	DRM_ERROR("errors: %08x\n", irqstatus);
-}
-
-void mdp4_irq_preinstall(struct msm_kms *kms)
-{
-	struct mdp4_kms *mdp4_kms = to_mdp4_kms(kms);
-	mdp4_write(mdp4_kms, REG_MDP4_INTR_CLEAR, 0xffffffff);
-}
-
-int mdp4_irq_postinstall(struct msm_kms *kms)
-{
-	struct mdp4_kms *mdp4_kms = to_mdp4_kms(kms);
-	struct mdp4_irq *error_handler = &mdp4_kms->error_handler;
-
-	INIT_LIST_HEAD(&mdp4_kms->irq_list);
-
-	error_handler->irq = mdp4_irq_error_handler;
-	error_handler->irqmask = MDP4_IRQ_PRIMARY_INTF_UDERRUN |
-			MDP4_IRQ_EXTERNAL_INTF_UDERRUN;
-
-	mdp4_irq_register(mdp4_kms, error_handler);
-
-	return 0;
-}
-
-void mdp4_irq_uninstall(struct msm_kms *kms)
-{
-	struct mdp4_kms *mdp4_kms = to_mdp4_kms(kms);
-	mdp4_write(mdp4_kms, REG_MDP4_INTR_ENABLE, 0x00000000);
-}
-
-irqreturn_t mdp4_irq(struct msm_kms *kms)
-{
-	struct mdp4_kms *mdp4_kms = to_mdp4_kms(kms);
-	struct drm_device *dev = mdp4_kms->dev;
-	struct msm_drm_private *priv = dev->dev_private;
-	struct mdp4_irq *handler, *n;
-	unsigned long flags;
-	unsigned int id;
-	uint32_t status;
-
-	status = mdp4_read(mdp4_kms, REG_MDP4_INTR_STATUS);
-	mdp4_write(mdp4_kms, REG_MDP4_INTR_CLEAR, status);
-
-	VERB("status=%08x", status);
-
-	for (id = 0; id < priv->num_crtcs; id++)
-		if (status & mdp4_crtc_vblank(priv->crtcs[id]))
-			drm_handle_vblank(dev, id);
-
-	spin_lock_irqsave(&list_lock, flags);
-	mdp4_kms->in_irq = true;
-	list_for_each_entry_safe(handler, n, &mdp4_kms->irq_list, node) {
-		if (handler->irqmask & status) {
-			spin_unlock_irqrestore(&list_lock, flags);
-			handler->irq(handler, handler->irqmask & status);
-			spin_lock_irqsave(&list_lock, flags);
-		}
-	}
-	mdp4_kms->in_irq = false;
-	update_irq(mdp4_kms);
-	spin_unlock_irqrestore(&list_lock, flags);
-
-	return IRQ_HANDLED;
-}
-
-int mdp4_enable_vblank(struct msm_kms *kms, struct drm_crtc *crtc)
-{
-	struct mdp4_kms *mdp4_kms = to_mdp4_kms(kms);
-	unsigned long flags;
-
-	spin_lock_irqsave(&list_lock, flags);
-	mdp4_kms->vblank_mask |= mdp4_crtc_vblank(crtc);
-	update_irq(mdp4_kms);
-	spin_unlock_irqrestore(&list_lock, flags);
-
-	return 0;
-}
-
-void mdp4_disable_vblank(struct msm_kms *kms, struct drm_crtc *crtc)
-{
-	struct mdp4_kms *mdp4_kms = to_mdp4_kms(kms);
-	unsigned long flags;
-
-	spin_lock_irqsave(&list_lock, flags);
-	mdp4_kms->vblank_mask &= ~mdp4_crtc_vblank(crtc);
-	update_irq(mdp4_kms);
-	spin_unlock_irqrestore(&list_lock, flags);
-}
-
-static void wait_irq(struct mdp4_irq *irq, uint32_t irqstatus)
-{
-	struct mdp4_irq_wait *wait =
-			container_of(irq, struct mdp4_irq_wait, irq);
-	wait->count--;
-	wake_up_all(&wait_event);
-}
-
-void mdp4_irq_wait(struct mdp4_kms *mdp4_kms, uint32_t irqmask)
-{
-	struct mdp4_irq_wait wait = {
-		.irq = {
-			.irq = wait_irq,
-			.irqmask = irqmask,
-		},
-		.count = 1,
-	};
-	mdp4_irq_register(mdp4_kms, &wait.irq);
-	wait_event(wait_event, (wait.count <= 0));
-	mdp4_irq_unregister(mdp4_kms, &wait.irq);
-}
-
-void mdp4_irq_register(struct mdp4_kms *mdp4_kms, struct mdp4_irq *irq)
-{
-	unsigned long flags;
-	bool needs_update = false;
-
-	spin_lock_irqsave(&list_lock, flags);
-
-	if (!irq->registered) {
-		irq->registered = true;
-		list_add(&irq->node, &mdp4_kms->irq_list);
-		needs_update = !mdp4_kms->in_irq;
-	}
-
-	spin_unlock_irqrestore(&list_lock, flags);
-
-	if (needs_update)
-		update_irq_unlocked(mdp4_kms);
-}
-
-void mdp4_irq_unregister(struct mdp4_kms *mdp4_kms, struct mdp4_irq *irq)
-{
-	unsigned long flags;
-	bool needs_update = false;
-
-	spin_lock_irqsave(&list_lock, flags);
-
-	if (irq->registered) {
-		irq->registered = false;
-		list_del(&irq->node);
-		needs_update = !mdp4_kms->in_irq;
-	}
-
-	spin_unlock_irqrestore(&list_lock, flags);
-
-	if (needs_update)
-		update_irq_unlocked(mdp4_kms);
-}
diff --git a/drivers/gpu/drm/msm/msm_drv.c b/drivers/gpu/drm/msm/msm_drv.c
index 86537692e45c..e6adafc7eff3 100644
--- a/drivers/gpu/drm/msm/msm_drv.c
+++ b/drivers/gpu/drm/msm/msm_drv.c
@@ -17,6 +17,7 @@
 
 #include "msm_drv.h"
 #include "msm_gpu.h"
+#include "msm_kms.h"
 
 static void msm_fb_output_poll_changed(struct drm_device *dev)
 {
@@ -30,50 +31,19 @@ static const struct drm_mode_config_funcs mode_config_funcs = {
 	.output_poll_changed = msm_fb_output_poll_changed,
 };
 
-static int msm_fault_handler(struct iommu_domain *iommu, struct device *dev,
-		unsigned long iova, int flags, void *arg)
-{
-	DBG("*** fault: iova=%08lx, flags=%d", iova, flags);
-	return 0;
-}
-
-int msm_register_iommu(struct drm_device *dev, struct iommu_domain *iommu)
+int msm_register_mmu(struct drm_device *dev, struct msm_mmu *mmu)
 {
 	struct msm_drm_private *priv = dev->dev_private;
-	int idx = priv->num_iommus++;
+	int idx = priv->num_mmus++;
 
-	if (WARN_ON(idx >= ARRAY_SIZE(priv->iommus)))
+	if (WARN_ON(idx >= ARRAY_SIZE(priv->mmus)))
 		return -EINVAL;
 
-	priv->iommus[idx] = iommu;
-
-	iommu_set_fault_handler(iommu, msm_fault_handler, dev);
-
-	/* need to iommu_attach_device() somewhere??  on resume?? */
+	priv->mmus[idx] = mmu;
 
 	return idx;
 }
 
-int msm_iommu_attach(struct drm_device *dev, struct iommu_domain *iommu,
-		const char **names, int cnt)
-{
-	int i, ret;
-
-	for (i = 0; i < cnt; i++) {
-		/* TODO maybe some day msm iommu won't require this hack: */
-		struct device *msm_iommu_get_ctx(const char *ctx_name);
-		struct device *ctx = msm_iommu_get_ctx(names[i]);
-		if (!ctx)
-			continue;
-		ret = iommu_attach_device(iommu, ctx);
-		if (ret) {
-			dev_warn(dev->dev, "could not attach iommu to %s", names[i]);
-			return ret;
-		}
-	}
-	return 0;
-}
-
 #ifdef CONFIG_DRM_MSM_REGISTER_LOGGING
 static bool reglog = false;
 MODULE_PARM_DESC(reglog, "Enable register read/write logging");
@@ -82,6 +52,10 @@ module_param(reglog, bool, 0600);
 #define reglog 0
 #endif
 
+static char *vram;
+MODULE_PARM_DESC(vram, "Configure VRAM size (for devices without IOMMU/GPUMMU");
+module_param(vram, charp, 0);
+
 void __iomem *msm_ioremap(struct platform_device *pdev, const char *name,
 		const char *dbgname)
 {
@@ -161,6 +135,14 @@ static int msm_unload(struct drm_device *dev)
 		mutex_unlock(&dev->struct_mutex);
 	}
 
+	if (priv->vram.paddr) {
+		DEFINE_DMA_ATTRS(attrs);
+		dma_set_attr(DMA_ATTR_NO_KERNEL_MAPPING, &attrs);
+		drm_mm_takedown(&priv->vram.mm);
+		dma_free_attrs(dev->dev, priv->vram.size, NULL,
+				priv->vram.paddr, &attrs);
+	}
+
 	dev->dev_private = NULL;
 
 	kfree(priv);
@@ -168,6 +150,24 @@ static int msm_unload(struct drm_device *dev)
 	return 0;
 }
 
+static int get_mdp_ver(struct platform_device *pdev)
+{
+#ifdef CONFIG_OF
+	const static struct of_device_id match_types[] = { {
+		.compatible = "qcom,mdss_mdp",
+		.data	= (void	*)5,
+	}, {
+		/* end node */
+	} };
+	struct device *dev = &pdev->dev;
+	const struct of_device_id *match;
+	match = of_match_node(match_types, dev->of_node);
+	if (match)
+		return (int)match->data;
+#endif
+	return 4;
+}
+
 static int msm_load(struct drm_device *dev, unsigned long flags)
 {
 	struct platform_device *pdev = dev->platformdev;
@@ -191,7 +191,53 @@ static int msm_load(struct drm_device *dev, unsigned long flags)
 
 	drm_mode_config_init(dev);
 
-	kms = mdp4_kms_init(dev);
+	/* if we have no IOMMU, then we need to use carveout allocator.
+	 * Grab the entire CMA chunk carved out in early startup in
+	 * mach-msm:
+	 */
+	if (!iommu_present(&platform_bus_type)) {
+		DEFINE_DMA_ATTRS(attrs);
+		unsigned long size;
+		void *p;
+
+		DBG("using %s VRAM carveout", vram);
+		size = memparse(vram, NULL);
+		priv->vram.size = size;
+
+		drm_mm_init(&priv->vram.mm, 0, (size >> PAGE_SHIFT) - 1);
+
+		dma_set_attr(DMA_ATTR_NO_KERNEL_MAPPING, &attrs);
+		dma_set_attr(DMA_ATTR_WRITE_COMBINE, &attrs);
+
+		/* note that for no-kernel-mapping, the vaddr returned
+		 * is bogus, but non-null if allocation succeeded:
+		 */
+		p = dma_alloc_attrs(dev->dev, size,
+				&priv->vram.paddr, 0, &attrs);
+		if (!p) {
+			dev_err(dev->dev, "failed to allocate VRAM\n");
+			priv->vram.paddr = 0;
+			ret = -ENOMEM;
+			goto fail;
+		}
+
+		dev_info(dev->dev, "VRAM: %08x->%08x\n",
+				(uint32_t)priv->vram.paddr,
+				(uint32_t)(priv->vram.paddr + size));
+	}
+
+	switch (get_mdp_ver(pdev)) {
+	case 4:
+		kms = mdp4_kms_init(dev);
+		break;
+	case 5:
+		kms = mdp5_kms_init(dev);
+		break;
+	default:
+		kms = ERR_PTR(-ENODEV);
+		break;
+	}
+
 	if (IS_ERR(kms)) {
 		/*
 		 * NOTE: once we have GPU support, having no kms should not
@@ -326,7 +372,7 @@ static void msm_lastclose(struct drm_device *dev)
 	}
 }
 
-static irqreturn_t msm_irq(DRM_IRQ_ARGS)
+static irqreturn_t msm_irq(int irq, void *arg)
 {
 	struct drm_device *dev = arg;
 	struct msm_drm_private *priv = dev->dev_private;
@@ -415,7 +461,7 @@ static int msm_gem_show(struct drm_device *dev, struct seq_file *m)
 
 static int msm_mm_show(struct drm_device *dev, struct seq_file *m)
 {
-	return drm_mm_dump_table(m, dev->mm_private);
+	return drm_mm_dump_table(m, &dev->vma_offset_manager->vm_addr_space_mm);
 }
 
 static int msm_fb_show(struct drm_device *dev, struct seq_file *m)
@@ -778,12 +824,13 @@ static const struct dev_pm_ops msm_pm_ops = {
 
 static int msm_pdev_probe(struct platform_device *pdev)
 {
+	pdev->dev.coherent_dma_mask = DMA_BIT_MASK(32);
 	return drm_platform_init(&msm_driver, pdev);
 }
 
 static int msm_pdev_remove(struct platform_device *pdev)
 {
-	drm_platform_exit(&msm_driver, pdev);
+	drm_put_dev(platform_get_drvdata(pdev));
 
 	return 0;
 }
@@ -793,12 +840,19 @@ static const struct platform_device_id msm_id[] = {
 	{ }
 };
 
+static const struct of_device_id dt_match[] = {
+	{ .compatible = "qcom,mdss_mdp" },
+	{}
+};
+MODULE_DEVICE_TABLE(of, dt_match);
+
 static struct platform_driver msm_platform_driver = {
 	.probe      = msm_pdev_probe,
 	.remove     = msm_pdev_remove,
 	.driver     = {
 		.owner  = THIS_MODULE,
 		.name   = "msm",
+		.of_match_table = dt_match,
 		.pm     = &msm_pm_ops,
 	},
 	.id_table   = msm_id,
diff --git a/drivers/gpu/drm/msm/msm_drv.h b/drivers/gpu/drm/msm/msm_drv.h
index d39f0862b19e..3d63269c5b29 100644
--- a/drivers/gpu/drm/msm/msm_drv.h
+++ b/drivers/gpu/drm/msm/msm_drv.h
@@ -31,6 +31,15 @@
 #include <linux/types.h>
 #include <asm/sizes.h>
 
+
+#if defined(CONFIG_COMPILE_TEST) && !defined(CONFIG_ARCH_MSM)
+/* stubs we need for compile-test: */
+static inline struct device *msm_iommu_get_ctx(const char *ctx_name)
+{
+	return NULL;
+}
+#endif
+
 #ifndef CONFIG_OF
 #include <mach/board.h>
 #include <mach/socinfo.h>
@@ -44,6 +53,7 @@
 
 struct msm_kms;
 struct msm_gpu;
+struct msm_mmu;
 
 #define NUM_DOMAINS 2    /* one for KMS, then one per gpu core (?) */
 
@@ -76,9 +86,9 @@ struct msm_drm_private {
 	/* callbacks deferred until bo is inactive: */
 	struct list_head fence_cbs;
 
-	/* registered IOMMU domains: */
-	unsigned int num_iommus;
-	struct iommu_domain *iommus[NUM_DOMAINS];
+	/* registered MMUs: */
+	unsigned int num_mmus;
+	struct msm_mmu *mmus[NUM_DOMAINS];
 
 	unsigned int num_planes;
 	struct drm_plane *planes[8];
@@ -94,6 +104,16 @@ struct msm_drm_private {
 
 	unsigned int num_connectors;
 	struct drm_connector *connectors[8];
+
+	/* VRAM carveout, used when no IOMMU: */
+	struct {
+		unsigned long size;
+		dma_addr_t paddr;
+		/* NOTE: mm managed at the page level, size is in # of pages
+		 * and position mm_node->start is in # of pages:
+		 */
+		struct drm_mm mm;
+	} vram;
 };
 
 struct msm_format {
@@ -114,39 +134,7 @@ void __msm_fence_worker(struct work_struct *work);
 		(_cb)->func = _func;                         \
 	} while (0)
 
-/* As there are different display controller blocks depending on the
- * snapdragon version, the kms support is split out and the appropriate
- * implementation is loaded at runtime.  The kms module is responsible
- * for constructing the appropriate planes/crtcs/encoders/connectors.
- */
-struct msm_kms_funcs {
-	/* hw initialization: */
-	int (*hw_init)(struct msm_kms *kms);
-	/* irq handling: */
-	void (*irq_preinstall)(struct msm_kms *kms);
-	int (*irq_postinstall)(struct msm_kms *kms);
-	void (*irq_uninstall)(struct msm_kms *kms);
-	irqreturn_t (*irq)(struct msm_kms *kms);
-	int (*enable_vblank)(struct msm_kms *kms, struct drm_crtc *crtc);
-	void (*disable_vblank)(struct msm_kms *kms, struct drm_crtc *crtc);
-	/* misc: */
-	const struct msm_format *(*get_format)(struct msm_kms *kms, uint32_t format);
-	long (*round_pixclk)(struct msm_kms *kms, unsigned long rate,
-			struct drm_encoder *encoder);
-	/* cleanup: */
-	void (*preclose)(struct msm_kms *kms, struct drm_file *file);
-	void (*destroy)(struct msm_kms *kms);
-};
-
-struct msm_kms {
-	const struct msm_kms_funcs *funcs;
-};
-
-struct msm_kms *mdp4_kms_init(struct drm_device *dev);
-
-int msm_register_iommu(struct drm_device *dev, struct iommu_domain *iommu);
-int msm_iommu_attach(struct drm_device *dev, struct iommu_domain *iommu,
-		const char **names, int cnt);
+int msm_register_mmu(struct drm_device *dev, struct msm_mmu *mmu);
 
 int msm_wait_fence_interruptable(struct drm_device *dev, uint32_t fence,
 		struct timespec *timeout);
@@ -202,7 +190,9 @@ struct drm_framebuffer *msm_framebuffer_create(struct drm_device *dev,
 
 struct drm_fb_helper *msm_fbdev_init(struct drm_device *dev);
 
-int hdmi_init(struct drm_device *dev, struct drm_encoder *encoder);
+struct hdmi;
+struct hdmi *hdmi_init(struct drm_device *dev, struct drm_encoder *encoder);
+irqreturn_t hdmi_irq(int irq, void *dev_id);
 void __init hdmi_register(void);
 void __exit hdmi_unregister(void);
 
diff --git a/drivers/gpu/drm/msm/msm_fb.c b/drivers/gpu/drm/msm/msm_fb.c
index 0286c0eeb10c..81bafdf19ab3 100644
--- a/drivers/gpu/drm/msm/msm_fb.c
+++ b/drivers/gpu/drm/msm/msm_fb.c
@@ -16,6 +16,7 @@
  */
 
 #include "msm_drv.h"
+#include "msm_kms.h"
 
 #include "drm_crtc.h"
 #include "drm_crtc_helper.h"
diff --git a/drivers/gpu/drm/msm/msm_gem.c b/drivers/gpu/drm/msm/msm_gem.c
index e587d251c590..d8d60c969ac7 100644
--- a/drivers/gpu/drm/msm/msm_gem.c
+++ b/drivers/gpu/drm/msm/msm_gem.c
@@ -22,7 +22,45 @@
 #include "msm_drv.h"
 #include "msm_gem.h"
 #include "msm_gpu.h"
+#include "msm_mmu.h"
 
+static dma_addr_t physaddr(struct drm_gem_object *obj)
+{
+	struct msm_gem_object *msm_obj = to_msm_bo(obj);
+	struct msm_drm_private *priv = obj->dev->dev_private;
+	return (((dma_addr_t)msm_obj->vram_node->start) << PAGE_SHIFT) +
+			priv->vram.paddr;
+}
+
+/* allocate pages from VRAM carveout, used when no IOMMU: */
+static struct page **get_pages_vram(struct drm_gem_object *obj,
+		int npages)
+{
+	struct msm_gem_object *msm_obj = to_msm_bo(obj);
+	struct msm_drm_private *priv = obj->dev->dev_private;
+	dma_addr_t paddr;
+	struct page **p;
+	int ret, i;
+
+	p = drm_malloc_ab(npages, sizeof(struct page *));
+	if (!p)
+		return ERR_PTR(-ENOMEM);
+
+	ret = drm_mm_insert_node(&priv->vram.mm, msm_obj->vram_node,
+			npages, 0, DRM_MM_SEARCH_DEFAULT);
+	if (ret) {
+		drm_free_large(p);
+		return ERR_PTR(ret);
+	}
+
+	paddr = physaddr(obj);
+	for (i = 0; i < npages; i++) {
+		p[i] = phys_to_page(paddr);
+		paddr += PAGE_SIZE;
+	}
+
+	return p;
+}
 
 /* called with dev->struct_mutex held */
 static struct page **get_pages(struct drm_gem_object *obj)
@@ -31,9 +69,14 @@ static struct page **get_pages(struct drm_gem_object *obj)
 
 	if (!msm_obj->pages) {
 		struct drm_device *dev = obj->dev;
-		struct page **p = drm_gem_get_pages(obj, 0);
+		struct page **p;
 		int npages = obj->size >> PAGE_SHIFT;
 
+		if (iommu_present(&platform_bus_type))
+			p = drm_gem_get_pages(obj, 0);
+		else
+			p = get_pages_vram(obj, npages);
+
 		if (IS_ERR(p)) {
 			dev_err(dev->dev, "could not get pages: %ld\n",
 					PTR_ERR(p));
@@ -73,7 +116,11 @@ static void put_pages(struct drm_gem_object *obj)
 		sg_free_table(msm_obj->sgt);
 		kfree(msm_obj->sgt);
 
-		drm_gem_put_pages(obj, msm_obj->pages, true, false);
+		if (iommu_present(&platform_bus_type))
+			drm_gem_put_pages(obj, msm_obj->pages, true, false);
+		else
+			drm_mm_remove_node(msm_obj->vram_node);
+
 		msm_obj->pages = NULL;
 	}
 }
@@ -138,7 +185,6 @@ int msm_gem_mmap(struct file *filp, struct vm_area_struct *vma)
 int msm_gem_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 {
 	struct drm_gem_object *obj = vma->vm_private_data;
-	struct msm_gem_object *msm_obj = to_msm_bo(obj);
 	struct drm_device *dev = obj->dev;
 	struct page **pages;
 	unsigned long pfn;
@@ -163,7 +209,7 @@ int msm_gem_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 	pgoff = ((unsigned long)vmf->virtual_address -
 			vma->vm_start) >> PAGE_SHIFT;
 
-	pfn = page_to_pfn(msm_obj->pages[pgoff]);
+	pfn = page_to_pfn(pages[pgoff]);
 
 	VERB("Inserting %p pfn %lx, pa %lx", vmf->virtual_address,
 			pfn, pfn << PAGE_SHIFT);
@@ -219,67 +265,6 @@ uint64_t msm_gem_mmap_offset(struct drm_gem_object *obj)
 	return offset;
 }
 
-/* helpers for dealing w/ iommu: */
-static int map_range(struct iommu_domain *domain, unsigned int iova,
-		struct sg_table *sgt, unsigned int len, int prot)
-{
-	struct scatterlist *sg;
-	unsigned int da = iova;
-	unsigned int i, j;
-	int ret;
-
-	if (!domain || !sgt)
-		return -EINVAL;
-
-	for_each_sg(sgt->sgl, sg, sgt->nents, i) {
-		u32 pa = sg_phys(sg) - sg->offset;
-		size_t bytes = sg->length + sg->offset;
-
-		VERB("map[%d]: %08x %08x(%x)", i, iova, pa, bytes);
-
-		ret = iommu_map(domain, da, pa, bytes, prot);
-		if (ret)
-			goto fail;
-
-		da += bytes;
-	}
-
-	return 0;
-
-fail:
-	da = iova;
-
-	for_each_sg(sgt->sgl, sg, i, j) {
-		size_t bytes = sg->length + sg->offset;
-		iommu_unmap(domain, da, bytes);
-		da += bytes;
-	}
-	return ret;
-}
-
-static void unmap_range(struct iommu_domain *domain, unsigned int iova,
-		struct sg_table *sgt, unsigned int len)
-{
-	struct scatterlist *sg;
-	unsigned int da = iova;
-	int i;
-
-	for_each_sg(sgt->sgl, sg, sgt->nents, i) {
-		size_t bytes = sg->length + sg->offset;
-		size_t unmapped;
-
-		unmapped = iommu_unmap(domain, da, bytes);
-		if (unmapped < bytes)
-			break;
-
-		VERB("unmap[%d]: %08x(%x)", i, iova, bytes);
-
-		BUG_ON(!IS_ALIGNED(bytes, PAGE_SIZE));
-
-		da += bytes;
-	}
-}
-
 /* should be called under struct_mutex.. although it can be called
  * from atomic context without struct_mutex to acquire an extra
  * iova ref if you know one is already held.
@@ -295,15 +280,20 @@ int msm_gem_get_iova_locked(struct drm_gem_object *obj, int id,
 
 	if (!msm_obj->domain[id].iova) {
 		struct msm_drm_private *priv = obj->dev->dev_private;
-		uint32_t offset = (uint32_t)mmap_offset(obj);
-		struct page **pages;
-		pages = get_pages(obj);
+		struct msm_mmu *mmu = priv->mmus[id];
+		struct page **pages = get_pages(obj);
+
 		if (IS_ERR(pages))
 			return PTR_ERR(pages);
-		// XXX ideally we would not map buffers writable when not needed...
-		ret = map_range(priv->iommus[id], offset, msm_obj->sgt,
-				obj->size, IOMMU_READ | IOMMU_WRITE);
-		msm_obj->domain[id].iova = offset;
+
+		if (iommu_present(&platform_bus_type)) {
+			uint32_t offset = (uint32_t)mmap_offset(obj);
+			ret = mmu->funcs->map(mmu, offset, msm_obj->sgt,
+					obj->size, IOMMU_READ | IOMMU_WRITE);
+			msm_obj->domain[id].iova = offset;
+		} else {
+			msm_obj->domain[id].iova = physaddr(obj);
+		}
 	}
 
 	if (!ret)
@@ -514,6 +504,7 @@ void msm_gem_describe_objects(struct list_head *list, struct seq_file *m)
 void msm_gem_free_object(struct drm_gem_object *obj)
 {
 	struct drm_device *dev = obj->dev;
+	struct msm_drm_private *priv = obj->dev->dev_private;
 	struct msm_gem_object *msm_obj = to_msm_bo(obj);
 	int id;
 
@@ -525,11 +516,10 @@ void msm_gem_free_object(struct drm_gem_object *obj)
 	list_del(&msm_obj->mm_list);
 
 	for (id = 0; id < ARRAY_SIZE(msm_obj->domain); id++) {
-		if (msm_obj->domain[id].iova) {
-			struct msm_drm_private *priv = obj->dev->dev_private;
+		struct msm_mmu *mmu = priv->mmus[id];
+		if (mmu && msm_obj->domain[id].iova) {
 			uint32_t offset = (uint32_t)mmap_offset(obj);
-			unmap_range(priv->iommus[id], offset,
-					msm_obj->sgt, obj->size);
+			mmu->funcs->unmap(mmu, offset, msm_obj->sgt, obj->size);
 		}
 	}
 
@@ -591,6 +581,7 @@ static int msm_gem_new_impl(struct drm_device *dev,
 {
 	struct msm_drm_private *priv = dev->dev_private;
 	struct msm_gem_object *msm_obj;
+	unsigned sz;
 
 	switch (flags & MSM_BO_CACHE_MASK) {
 	case MSM_BO_UNCACHED:
@@ -603,10 +594,17 @@ static int msm_gem_new_impl(struct drm_device *dev,
 		return -EINVAL;
 	}
 
-	msm_obj = kzalloc(sizeof(*msm_obj), GFP_KERNEL);
+	sz = sizeof(*msm_obj);
+	if (!iommu_present(&platform_bus_type))
+		sz += sizeof(struct drm_mm_node);
+
+	msm_obj = kzalloc(sz, GFP_KERNEL);
 	if (!msm_obj)
 		return -ENOMEM;
 
+	if (!iommu_present(&platform_bus_type))
+		msm_obj->vram_node = (void *)&msm_obj[1];
+
 	msm_obj->flags = flags;
 
 	msm_obj->resv = &msm_obj->_resv;
@@ -623,7 +621,7 @@ static int msm_gem_new_impl(struct drm_device *dev,
 struct drm_gem_object *msm_gem_new(struct drm_device *dev,
 		uint32_t size, uint32_t flags)
 {
-	struct drm_gem_object *obj;
+	struct drm_gem_object *obj = NULL;
 	int ret;
 
 	WARN_ON(!mutex_is_locked(&dev->struct_mutex));
@@ -634,9 +632,13 @@ struct drm_gem_object *msm_gem_new(struct drm_device *dev,
 	if (ret)
 		goto fail;
 
-	ret = drm_gem_object_init(dev, obj, size);
-	if (ret)
-		goto fail;
+	if (iommu_present(&platform_bus_type)) {
+		ret = drm_gem_object_init(dev, obj, size);
+		if (ret)
+			goto fail;
+	} else {
+		drm_gem_private_object_init(dev, obj, size);
+	}
 
 	return obj;
 
@@ -654,6 +656,12 @@ struct drm_gem_object *msm_gem_import(struct drm_device *dev,
 	struct drm_gem_object *obj;
 	int ret, npages;
 
+	/* if we don't have IOMMU, don't bother pretending we can import: */
+	if (!iommu_present(&platform_bus_type)) {
+		dev_err(dev->dev, "cannot import without IOMMU\n");
+		return ERR_PTR(-EINVAL);
+	}
+
 	size = PAGE_ALIGN(size);
 
 	ret = msm_gem_new_impl(dev, size, MSM_BO_WC, &obj);
diff --git a/drivers/gpu/drm/msm/msm_gem.h b/drivers/gpu/drm/msm/msm_gem.h
index f4f23a578d9d..3246bb46c4f2 100644
--- a/drivers/gpu/drm/msm/msm_gem.h
+++ b/drivers/gpu/drm/msm/msm_gem.h
@@ -57,6 +57,11 @@ struct msm_gem_object {
 	/* normally (resv == &_resv) except for imported bo's */
 	struct reservation_object *resv;
 	struct reservation_object _resv;
+
+	/* For physically contiguous buffers.  Used when we don't have
+	 * an IOMMU.
+	 */
+	struct drm_mm_node *vram_node;
 };
 #define to_msm_bo(x) container_of(x, struct msm_gem_object, base)
 
diff --git a/drivers/gpu/drm/msm/msm_gpu.c b/drivers/gpu/drm/msm/msm_gpu.c
index 4583d61556f5..4ebce8be489d 100644
--- a/drivers/gpu/drm/msm/msm_gpu.c
+++ b/drivers/gpu/drm/msm/msm_gpu.c
@@ -17,6 +17,7 @@
 
 #include "msm_gpu.h"
 #include "msm_gem.h"
+#include "msm_mmu.h"
 
 
 /*
@@ -25,20 +26,10 @@
 
 #ifdef CONFIG_MSM_BUS_SCALING
 #include <mach/board.h>
-#include <mach/kgsl.h>
-static void bs_init(struct msm_gpu *gpu, struct platform_device *pdev)
+static void bs_init(struct msm_gpu *gpu)
 {
-	struct drm_device *dev = gpu->dev;
-	struct kgsl_device_platform_data *pdata;
-
-	if (!pdev) {
-		dev_err(dev->dev, "could not find dtv pdata\n");
-		return;
-	}
-
-	pdata = pdev->dev.platform_data;
-	if (pdata->bus_scale_table) {
-		gpu->bsc = msm_bus_scale_register_client(pdata->bus_scale_table);
+	if (gpu->bus_scale_table) {
+		gpu->bsc = msm_bus_scale_register_client(gpu->bus_scale_table);
 		DBG("bus scale client: %08x", gpu->bsc);
 	}
 }
@@ -59,7 +50,7 @@ static void bs_set(struct msm_gpu *gpu, int idx)
 	}
 }
 #else
-static void bs_init(struct msm_gpu *gpu, struct platform_device *pdev) {}
+static void bs_init(struct msm_gpu *gpu) {}
 static void bs_fini(struct msm_gpu *gpu) {}
 static void bs_set(struct msm_gpu *gpu, int idx) {}
 #endif
@@ -363,6 +354,7 @@ int msm_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 		struct msm_gpu *gpu, const struct msm_gpu_funcs *funcs,
 		const char *name, const char *ioname, const char *irqname, int ringsz)
 {
+	struct iommu_domain *iommu;
 	int i, ret;
 
 	gpu->dev = drm;
@@ -428,13 +420,14 @@ int msm_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 	 * and have separate page tables per context.  For now, to keep things
 	 * simple and to get something working, just use a single address space:
 	 */
-	gpu->iommu = iommu_domain_alloc(&platform_bus_type);
-	if (!gpu->iommu) {
-		dev_err(drm->dev, "failed to allocate IOMMU\n");
-		ret = -ENOMEM;
-		goto fail;
+	iommu = iommu_domain_alloc(&platform_bus_type);
+	if (iommu) {
+		dev_info(drm->dev, "%s: using IOMMU\n", name);
+		gpu->mmu = msm_iommu_new(drm, iommu);
+	} else {
+		dev_info(drm->dev, "%s: no IOMMU, fallback to VRAM carveout!\n", name);
 	}
-	gpu->id = msm_register_iommu(drm, gpu->iommu);
+	gpu->id = msm_register_mmu(drm, gpu->mmu);
 
 	/* Create ringbuffer: */
 	gpu->rb = msm_ringbuffer_new(gpu, ringsz);
@@ -452,7 +445,7 @@ int msm_gpu_init(struct drm_device *drm, struct platform_device *pdev,
 		goto fail;
 	}
 
-	bs_init(gpu, pdev);
+	bs_init(gpu);
 
 	return 0;
 
@@ -474,6 +467,6 @@ void msm_gpu_cleanup(struct msm_gpu *gpu)
 		msm_ringbuffer_destroy(gpu->rb);
 	}
 
-	if (gpu->iommu)
-		iommu_domain_free(gpu->iommu);
+	if (gpu->mmu)
+		gpu->mmu->funcs->destroy(gpu->mmu);
 }
diff --git a/drivers/gpu/drm/msm/msm_gpu.h b/drivers/gpu/drm/msm/msm_gpu.h
index 8cd829e520bb..458db8c64c28 100644
--- a/drivers/gpu/drm/msm/msm_gpu.h
+++ b/drivers/gpu/drm/msm/msm_gpu.h
@@ -78,14 +78,18 @@ struct msm_gpu {
 	void __iomem *mmio;
 	int irq;
 
-	struct iommu_domain *iommu;
+	struct msm_mmu *mmu;
 	int id;
 
 	/* Power Control: */
 	struct regulator *gpu_reg, *gpu_cx;
 	struct clk *ebi1_clk, *grp_clks[5];
 	uint32_t fast_rate, slow_rate, bus_freq;
+
+#ifdef CONFIG_MSM_BUS_SCALING
+	struct msm_bus_scale_pdata *bus_scale_table;
 	uint32_t bsc;
+#endif
 
 	/* Hang Detction: */
 #define DRM_MSM_HANGCHECK_PERIOD 500 /* in ms */
diff --git a/drivers/gpu/drm/msm/msm_iommu.c b/drivers/gpu/drm/msm/msm_iommu.c
new file mode 100644
index 000000000000..92b745986231
--- /dev/null
+++ b/drivers/gpu/drm/msm/msm_iommu.c
@@ -0,0 +1,148 @@
+/*
+ * Copyright (C) 2013 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 "msm_drv.h"
+#include "msm_mmu.h"
+
+struct msm_iommu {
+	struct msm_mmu base;
+	struct iommu_domain *domain;
+};
+#define to_msm_iommu(x) container_of(x, struct msm_iommu, base)
+
+static int msm_fault_handler(struct iommu_domain *iommu, struct device *dev,
+		unsigned long iova, int flags, void *arg)
+{
+	DBG("*** fault: iova=%08lx, flags=%d", iova, flags);
+	return 0;
+}
+
+static int msm_iommu_attach(struct msm_mmu *mmu, const char **names, int cnt)
+{
+	struct drm_device *dev = mmu->dev;
+	struct msm_iommu *iommu = to_msm_iommu(mmu);
+	int i, ret;
+
+	for (i = 0; i < cnt; i++) {
+		struct device *msm_iommu_get_ctx(const char *ctx_name);
+		struct device *ctx = msm_iommu_get_ctx(names[i]);
+		if (IS_ERR_OR_NULL(ctx))
+			continue;
+		ret = iommu_attach_device(iommu->domain, ctx);
+		if (ret) {
+			dev_warn(dev->dev, "could not attach iommu to %s", names[i]);
+			return ret;
+		}
+	}
+
+	return 0;
+}
+
+static int msm_iommu_map(struct msm_mmu *mmu, uint32_t iova,
+		struct sg_table *sgt, unsigned len, int prot)
+{
+	struct msm_iommu *iommu = to_msm_iommu(mmu);
+	struct iommu_domain *domain = iommu->domain;
+	struct scatterlist *sg;
+	unsigned int da = iova;
+	unsigned int i, j;
+	int ret;
+
+	if (!domain || !sgt)
+		return -EINVAL;
+
+	for_each_sg(sgt->sgl, sg, sgt->nents, i) {
+		u32 pa = sg_phys(sg) - sg->offset;
+		size_t bytes = sg->length + sg->offset;
+
+		VERB("map[%d]: %08x %08x(%x)", i, iova, pa, bytes);
+
+		ret = iommu_map(domain, da, pa, bytes, prot);
+		if (ret)
+			goto fail;
+
+		da += bytes;
+	}
+
+	return 0;
+
+fail:
+	da = iova;
+
+	for_each_sg(sgt->sgl, sg, i, j) {
+		size_t bytes = sg->length + sg->offset;
+		iommu_unmap(domain, da, bytes);
+		da += bytes;
+	}
+	return ret;
+}
+
+static int msm_iommu_unmap(struct msm_mmu *mmu, uint32_t iova,
+		struct sg_table *sgt, unsigned len)
+{
+	struct msm_iommu *iommu = to_msm_iommu(mmu);
+	struct iommu_domain *domain = iommu->domain;
+	struct scatterlist *sg;
+	unsigned int da = iova;
+	int i;
+
+	for_each_sg(sgt->sgl, sg, sgt->nents, i) {
+		size_t bytes = sg->length + sg->offset;
+		size_t unmapped;
+
+		unmapped = iommu_unmap(domain, da, bytes);
+		if (unmapped < bytes)
+			return unmapped;
+
+		VERB("unmap[%d]: %08x(%x)", i, iova, bytes);
+
+		BUG_ON(!IS_ALIGNED(bytes, PAGE_SIZE));
+
+		da += bytes;
+	}
+
+	return 0;
+}
+
+static void msm_iommu_destroy(struct msm_mmu *mmu)
+{
+	struct msm_iommu *iommu = to_msm_iommu(mmu);
+	iommu_domain_free(iommu->domain);
+	kfree(iommu);
+}
+
+static const struct msm_mmu_funcs funcs = {
+		.attach = msm_iommu_attach,
+		.map = msm_iommu_map,
+		.unmap = msm_iommu_unmap,
+		.destroy = msm_iommu_destroy,
+};
+
+struct msm_mmu *msm_iommu_new(struct drm_device *dev, struct iommu_domain *domain)
+{
+	struct msm_iommu *iommu;
+
+	iommu = kzalloc(sizeof(*iommu), GFP_KERNEL);
+	if (!iommu)
+		return ERR_PTR(-ENOMEM);
+
+	iommu->domain = domain;
+	msm_mmu_init(&iommu->base, dev, &funcs);
+	iommu_set_fault_handler(domain, msm_fault_handler, dev);
+
+	return &iommu->base;
+}
diff --git a/drivers/gpu/drm/msm/msm_kms.h b/drivers/gpu/drm/msm/msm_kms.h
new file mode 100644
index 000000000000..06437745bc2c
--- /dev/null
+++ b/drivers/gpu/drm/msm/msm_kms.h
@@ -0,0 +1,68 @@
+/*
+ * Copyright (C) 2013 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/>.
+ */
+
+#ifndef __MSM_KMS_H__
+#define __MSM_KMS_H__
+
+#include <linux/clk.h>
+#include <linux/regulator/consumer.h>
+
+#include "msm_drv.h"
+
+/* As there are different display controller blocks depending on the
+ * snapdragon version, the kms support is split out and the appropriate
+ * implementation is loaded at runtime.  The kms module is responsible
+ * for constructing the appropriate planes/crtcs/encoders/connectors.
+ */
+struct msm_kms_funcs {
+	/* hw initialization: */
+	int (*hw_init)(struct msm_kms *kms);
+	/* irq handling: */
+	void (*irq_preinstall)(struct msm_kms *kms);
+	int (*irq_postinstall)(struct msm_kms *kms);
+	void (*irq_uninstall)(struct msm_kms *kms);
+	irqreturn_t (*irq)(struct msm_kms *kms);
+	int (*enable_vblank)(struct msm_kms *kms, struct drm_crtc *crtc);
+	void (*disable_vblank)(struct msm_kms *kms, struct drm_crtc *crtc);
+	/* misc: */
+	const struct msm_format *(*get_format)(struct msm_kms *kms, uint32_t format);
+	long (*round_pixclk)(struct msm_kms *kms, unsigned long rate,
+			struct drm_encoder *encoder);
+	/* cleanup: */
+	void (*preclose)(struct msm_kms *kms, struct drm_file *file);
+	void (*destroy)(struct msm_kms *kms);
+};
+
+struct msm_kms {
+	const struct msm_kms_funcs *funcs;
+
+	/* irq handling: */
+	bool in_irq;
+	struct list_head irq_list;    /* list of mdp4_irq */
+	uint32_t vblank_mask;         /* irq bits set for userspace vblank */
+};
+
+static inline void msm_kms_init(struct msm_kms *kms,
+		const struct msm_kms_funcs *funcs)
+{
+	kms->funcs = funcs;
+}
+
+struct msm_kms *mdp4_kms_init(struct drm_device *dev);
+struct msm_kms *mdp5_kms_init(struct drm_device *dev);
+
+#endif /* __MSM_KMS_H__ */
diff --git a/drivers/gpu/drm/msm/msm_mmu.h b/drivers/gpu/drm/msm/msm_mmu.h
new file mode 100644
index 000000000000..030324482b4a
--- /dev/null
+++ b/drivers/gpu/drm/msm/msm_mmu.h
@@ -0,0 +1,47 @@
+/*
+ * Copyright (C) 2013 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/>.
+ */
+
+#ifndef __MSM_MMU_H__
+#define __MSM_MMU_H__
+
+#include <linux/iommu.h>
+
+struct msm_mmu_funcs {
+	int (*attach)(struct msm_mmu *mmu, const char **names, int cnt);
+	int (*map)(struct msm_mmu *mmu, uint32_t iova, struct sg_table *sgt,
+			unsigned len, int prot);
+	int (*unmap)(struct msm_mmu *mmu, uint32_t iova, struct sg_table *sgt,
+			unsigned len);
+	void (*destroy)(struct msm_mmu *mmu);
+};
+
+struct msm_mmu {
+	const struct msm_mmu_funcs *funcs;
+	struct drm_device *dev;
+};
+
+static inline void msm_mmu_init(struct msm_mmu *mmu, struct drm_device *dev,
+		const struct msm_mmu_funcs *funcs)
+{
+	mmu->dev = dev;
+	mmu->funcs = funcs;
+}
+
+struct msm_mmu *msm_iommu_new(struct drm_device *dev, struct iommu_domain *domain);
+struct msm_mmu *msm_gpummu_new(struct drm_device *dev, struct msm_gpu *gpu);
+
+#endif /* __MSM_MMU_H__ */
diff --git a/drivers/gpu/drm/nouveau/Makefile b/drivers/gpu/drm/nouveau/Makefile
index b3fa1ba191b7..e88145ba1bf5 100644
--- a/drivers/gpu/drm/nouveau/Makefile
+++ b/drivers/gpu/drm/nouveau/Makefile
@@ -41,6 +41,7 @@ nouveau-y += core/subdev/bios/init.o
 nouveau-y += core/subdev/bios/mxm.o
 nouveau-y += core/subdev/bios/perf.o
 nouveau-y += core/subdev/bios/pll.o
+nouveau-y += core/subdev/bios/ramcfg.o
 nouveau-y += core/subdev/bios/rammap.o
 nouveau-y += core/subdev/bios/timing.o
 nouveau-y += core/subdev/bios/therm.o
@@ -71,7 +72,10 @@ nouveau-y += core/subdev/devinit/nv10.o
 nouveau-y += core/subdev/devinit/nv1a.o
 nouveau-y += core/subdev/devinit/nv20.o
 nouveau-y += core/subdev/devinit/nv50.o
+nouveau-y += core/subdev/devinit/nv84.o
+nouveau-y += core/subdev/devinit/nv98.o
 nouveau-y += core/subdev/devinit/nva3.o
+nouveau-y += core/subdev/devinit/nvaf.o
 nouveau-y += core/subdev/devinit/nvc0.o
 nouveau-y += core/subdev/fb/base.o
 nouveau-y += core/subdev/fb/nv04.o
@@ -232,6 +236,7 @@ nouveau-y += core/engine/fifo/nv50.o
 nouveau-y += core/engine/fifo/nv84.o
 nouveau-y += core/engine/fifo/nvc0.o
 nouveau-y += core/engine/fifo/nve0.o
+nouveau-y += core/engine/fifo/nv108.o
 nouveau-y += core/engine/graph/ctxnv40.o
 nouveau-y += core/engine/graph/ctxnv50.o
 nouveau-y += core/engine/graph/ctxnvc0.o
@@ -242,6 +247,7 @@ nouveau-y += core/engine/graph/ctxnvd7.o
 nouveau-y += core/engine/graph/ctxnvd9.o
 nouveau-y += core/engine/graph/ctxnve4.o
 nouveau-y += core/engine/graph/ctxnvf0.o
+nouveau-y += core/engine/graph/ctxnv108.o
 nouveau-y += core/engine/graph/nv04.o
 nouveau-y += core/engine/graph/nv10.o
 nouveau-y += core/engine/graph/nv20.o
@@ -260,6 +266,7 @@ nouveau-y += core/engine/graph/nvd7.o
 nouveau-y += core/engine/graph/nvd9.o
 nouveau-y += core/engine/graph/nve4.o
 nouveau-y += core/engine/graph/nvf0.o
+nouveau-y += core/engine/graph/nv108.o
 nouveau-y += core/engine/mpeg/nv31.o
 nouveau-y += core/engine/mpeg/nv40.o
 nouveau-y += core/engine/mpeg/nv44.o
diff --git a/drivers/gpu/drm/nouveau/core/core/engine.c b/drivers/gpu/drm/nouveau/core/core/engine.c
index c8bed4a26833..1f6954ae9dd3 100644
--- a/drivers/gpu/drm/nouveau/core/core/engine.c
+++ b/drivers/gpu/drm/nouveau/core/core/engine.c
@@ -42,11 +42,24 @@ nouveau_engine_create_(struct nouveau_object *parent,
 	if (ret)
 		return ret;
 
-	if ( parent &&
-	    !nouveau_boolopt(nv_device(parent)->cfgopt, iname, enable)) {
-		if (!enable)
-			nv_warn(engine, "disabled, %s=1 to enable\n", iname);
-		return -ENODEV;
+	if (parent) {
+		struct nouveau_device *device = nv_device(parent);
+		int engidx = nv_engidx(nv_object(engine));
+
+		if (device->disable_mask & (1ULL << engidx)) {
+			if (!nouveau_boolopt(device->cfgopt, iname, false)) {
+				nv_debug(engine, "engine disabled by hw/fw\n");
+				return -ENODEV;
+			}
+
+			nv_warn(engine, "ignoring hw/fw engine disable\n");
+		}
+
+		if (!nouveau_boolopt(device->cfgopt, iname, enable)) {
+			if (!enable)
+				nv_warn(engine, "disabled, %s=1 to enable\n", iname);
+			return -ENODEV;
+		}
 	}
 
 	INIT_LIST_HEAD(&engine->contexts);
diff --git a/drivers/gpu/drm/nouveau/core/engine/copy/nvc0.c b/drivers/gpu/drm/nouveau/core/engine/copy/nvc0.c
index 993df09ad643..ac3291f781f6 100644
--- a/drivers/gpu/drm/nouveau/core/engine/copy/nvc0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/copy/nvc0.c
@@ -105,9 +105,6 @@ nvc0_copy0_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	struct nvc0_copy_priv *priv;
 	int ret;
 
-	if (nv_rd32(parent, 0x022500) & 0x00000100)
-		return -ENODEV;
-
 	ret = nouveau_falcon_create(parent, engine, oclass, 0x104000, true,
 				    "PCE0", "copy0", &priv);
 	*pobject = nv_object(priv);
@@ -133,9 +130,6 @@ nvc0_copy1_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	struct nvc0_copy_priv *priv;
 	int ret;
 
-	if (nv_rd32(parent, 0x022500) & 0x00000200)
-		return -ENODEV;
-
 	ret = nouveau_falcon_create(parent, engine, oclass, 0x105000, true,
 				    "PCE1", "copy1", &priv);
 	*pobject = nv_object(priv);
diff --git a/drivers/gpu/drm/nouveau/core/engine/copy/nve0.c b/drivers/gpu/drm/nouveau/core/engine/copy/nve0.c
index 30f1ef1edcc5..748a61eb3c6f 100644
--- a/drivers/gpu/drm/nouveau/core/engine/copy/nve0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/copy/nve0.c
@@ -88,9 +88,6 @@ nve0_copy0_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	struct nve0_copy_priv *priv;
 	int ret;
 
-	if (nv_rd32(parent, 0x022500) & 0x00000100)
-		return -ENODEV;
-
 	ret = nouveau_engine_create(parent, engine, oclass, true,
 				    "PCE0", "copy0", &priv);
 	*pobject = nv_object(priv);
@@ -112,9 +109,6 @@ nve0_copy1_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	struct nve0_copy_priv *priv;
 	int ret;
 
-	if (nv_rd32(parent, 0x022500) & 0x00000200)
-		return -ENODEV;
-
 	ret = nouveau_engine_create(parent, engine, oclass, true,
 				    "PCE1", "copy1", &priv);
 	*pobject = nv_object(priv);
diff --git a/drivers/gpu/drm/nouveau/core/engine/device/nv04.c b/drivers/gpu/drm/nouveau/core/engine/device/nv04.c
index dbd2dde7b7e7..32113b08c4d5 100644
--- a/drivers/gpu/drm/nouveau/core/engine/device/nv04.c
+++ b/drivers/gpu/drm/nouveau/core/engine/device/nv04.c
@@ -49,12 +49,12 @@ nv04_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv04_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv04_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv04_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv04_fifo_oclass;
@@ -67,12 +67,12 @@ nv04_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_VBIOS  ] = &nouveau_bios_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv05_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv05_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv04_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv04_fifo_oclass;
diff --git a/drivers/gpu/drm/nouveau/core/engine/device/nv10.c b/drivers/gpu/drm/nouveau/core/engine/device/nv10.c
index 6e03dd6abeea..744f15d7e131 100644
--- a/drivers/gpu/drm/nouveau/core/engine/device/nv10.c
+++ b/drivers/gpu/drm/nouveau/core/engine/device/nv10.c
@@ -51,12 +51,12 @@ nv10_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv10_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv10_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv10_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_GR     ] = &nv10_graph_oclass;
@@ -68,12 +68,12 @@ nv10_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv10_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv10_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv10_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv10_fifo_oclass;
@@ -87,12 +87,12 @@ nv10_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv10_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv10_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv10_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv10_fifo_oclass;
@@ -106,12 +106,12 @@ nv10_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv1a_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv10_fifo_oclass;
@@ -125,12 +125,12 @@ nv10_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv10_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv10_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv10_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv10_fifo_oclass;
@@ -144,12 +144,12 @@ nv10_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv10_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv10_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv10_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv17_fifo_oclass;
@@ -163,12 +163,12 @@ nv10_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv1a_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv17_fifo_oclass;
@@ -182,12 +182,12 @@ nv10_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv10_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv10_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv10_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv17_fifo_oclass;
diff --git a/drivers/gpu/drm/nouveau/core/engine/device/nv20.c b/drivers/gpu/drm/nouveau/core/engine/device/nv20.c
index dcde53b9f07f..27ba61fb2710 100644
--- a/drivers/gpu/drm/nouveau/core/engine/device/nv20.c
+++ b/drivers/gpu/drm/nouveau/core/engine/device/nv20.c
@@ -52,12 +52,12 @@ nv20_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv20_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv20_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv20_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv17_fifo_oclass;
@@ -71,12 +71,12 @@ nv20_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv20_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv20_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv25_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv17_fifo_oclass;
@@ -90,12 +90,12 @@ nv20_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv20_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv20_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv25_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv17_fifo_oclass;
@@ -109,12 +109,12 @@ nv20_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv20_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv20_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv25_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv17_fifo_oclass;
diff --git a/drivers/gpu/drm/nouveau/core/engine/device/nv30.c b/drivers/gpu/drm/nouveau/core/engine/device/nv30.c
index 7b8662ef4f59..fd47ace67543 100644
--- a/drivers/gpu/drm/nouveau/core/engine/device/nv30.c
+++ b/drivers/gpu/drm/nouveau/core/engine/device/nv30.c
@@ -52,12 +52,12 @@ nv30_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv20_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv20_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv30_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv17_fifo_oclass;
@@ -71,12 +71,12 @@ nv30_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv20_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv20_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv04_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv35_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv17_fifo_oclass;
@@ -90,12 +90,12 @@ nv30_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv20_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv20_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv30_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv17_fifo_oclass;
@@ -110,12 +110,12 @@ nv30_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv20_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv20_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv36_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv17_fifo_oclass;
@@ -130,12 +130,12 @@ nv30_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_GPIO   ] = &nv10_gpio_oclass;
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv04_clock_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv10_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv10_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv04_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv10_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv04_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv04_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
 		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv17_fifo_oclass;
diff --git a/drivers/gpu/drm/nouveau/core/engine/device/nv40.c b/drivers/gpu/drm/nouveau/core/engine/device/nv40.c
index c8c41e93695e..1b653dd74a70 100644
--- a/drivers/gpu/drm/nouveau/core/engine/device/nv40.c
+++ b/drivers/gpu/drm/nouveau/core/engine/device/nv40.c
@@ -57,12 +57,12 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv40_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv40_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
@@ -80,12 +80,12 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv40_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv41_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv41_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
@@ -103,12 +103,12 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv40_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv41_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv41_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
@@ -126,12 +126,12 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv40_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv41_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv41_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
@@ -149,12 +149,12 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv40_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv40_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv04_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
@@ -172,12 +172,12 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv40_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv47_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv41_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
@@ -195,12 +195,12 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv40_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv49_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv41_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
@@ -218,12 +218,12 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv40_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv49_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv41_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
@@ -241,12 +241,12 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv44_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv44_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv44_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
@@ -264,12 +264,12 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv44_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv46_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv44_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
@@ -287,12 +287,12 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv44_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv44_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv44_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
@@ -310,12 +310,12 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv44_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv46_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv44_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
@@ -333,12 +333,12 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv4e_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv44_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv4e_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv44_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
@@ -356,12 +356,12 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv44_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv46_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv44_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
@@ -379,12 +379,12 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv44_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv46_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv44_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_dmaeng_oclass;
@@ -402,12 +402,12 @@ nv40_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_I2C    ] = &nv04_i2c_oclass;
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nv40_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv40_therm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv1a_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv1a_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv44_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv31_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv46_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv40_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv40_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv44_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nv04_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 db3fc7be856a..81d5c26643d5 100644
--- a/drivers/gpu/drm/nouveau/core/engine/device/nv50.c
+++ b/drivers/gpu/drm/nouveau/core/engine/device/nv50.c
@@ -65,12 +65,12 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv50_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv50_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv50_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv50_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv50_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv50_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv50_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv50_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nv50_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
@@ -90,12 +90,12 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv84_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv50_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv84_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv50_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv50_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv84_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv50_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nv50_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
@@ -118,12 +118,12 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv84_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv50_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv84_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv50_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv50_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv84_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv50_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nv50_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
@@ -146,12 +146,12 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv84_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv50_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv84_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv50_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv50_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv84_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv50_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nv50_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
@@ -174,12 +174,12 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv84_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv50_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv84_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv94_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv94_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv84_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv50_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nv50_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
@@ -202,12 +202,12 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv84_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv50_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv84_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv94_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv94_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv84_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv50_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nv50_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
@@ -230,12 +230,12 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv84_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv50_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv98_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv98_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv94_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv84_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv50_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nv50_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
@@ -258,12 +258,12 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nv84_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv50_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv84_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv98_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv94_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nv84_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv50_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nv50_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
@@ -286,12 +286,12 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nvaa_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv50_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv98_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv98_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv94_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nvaa_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv50_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nv50_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
@@ -314,12 +314,12 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] =  nvaa_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nv84_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nv50_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nv98_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv98_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv94_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nvaa_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv50_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nv50_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
@@ -342,12 +342,12 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nva3_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nva3_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nva3_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv98_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv94_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nva3_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv50_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nv50_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nva3_pwr_oclass;
@@ -372,12 +372,12 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nva3_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nva3_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nva3_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv98_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv94_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nva3_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv50_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nv50_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nva3_pwr_oclass;
@@ -401,12 +401,12 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nva3_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nva3_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nva3_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv98_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv94_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nva3_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv50_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nv50_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nva3_pwr_oclass;
@@ -430,12 +430,12 @@ nv50_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nva3_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nva3_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvaf_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nv98_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nv94_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nvaf_fb_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nv50_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nv50_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nva3_pwr_oclass;
diff --git a/drivers/gpu/drm/nouveau/core/engine/device/nvc0.c b/drivers/gpu/drm/nouveau/core/engine/device/nvc0.c
index dbc5e33de94f..b7d66b59f43d 100644
--- a/drivers/gpu/drm/nouveau/core/engine/device/nvc0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/device/nvc0.c
@@ -65,14 +65,14 @@ nvc0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nvc0_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvc0_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nvc0_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nvc0_fb_oclass;
 		device->oclass[NVDEV_SUBDEV_LTCG   ] = &nvc0_ltcg_oclass;
 		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nvc0_ibus_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nvc0_pwr_oclass;
@@ -97,14 +97,14 @@ nvc0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nvc0_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvc0_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nvc0_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nvc0_fb_oclass;
 		device->oclass[NVDEV_SUBDEV_LTCG   ] = &nvc0_ltcg_oclass;
 		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nvc0_ibus_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nvc0_pwr_oclass;
@@ -129,14 +129,14 @@ nvc0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nvc0_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvc0_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nvc3_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nvc0_fb_oclass;
 		device->oclass[NVDEV_SUBDEV_LTCG   ] = &nvc0_ltcg_oclass;
 		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nvc0_ibus_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nvc0_pwr_oclass;
@@ -160,14 +160,14 @@ nvc0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nvc0_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvc0_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nvc0_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nvc0_fb_oclass;
 		device->oclass[NVDEV_SUBDEV_LTCG   ] = &nvc0_ltcg_oclass;
 		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nvc0_ibus_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nvc0_pwr_oclass;
@@ -192,14 +192,14 @@ nvc0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nvc0_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvc0_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nvc3_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nvc0_fb_oclass;
 		device->oclass[NVDEV_SUBDEV_LTCG   ] = &nvc0_ltcg_oclass;
 		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nvc0_ibus_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nvc0_pwr_oclass;
@@ -224,14 +224,14 @@ nvc0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nvc0_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvc0_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nvc3_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nvc0_fb_oclass;
 		device->oclass[NVDEV_SUBDEV_LTCG   ] = &nvc0_ltcg_oclass;
 		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nvc0_ibus_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nvc0_pwr_oclass;
@@ -255,14 +255,14 @@ nvc0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nva3_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nvc0_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvc0_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nvc0_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nvc0_fb_oclass;
 		device->oclass[NVDEV_SUBDEV_LTCG   ] = &nvc0_ltcg_oclass;
 		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nvc0_ibus_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nvc0_pwr_oclass;
@@ -287,14 +287,14 @@ nvc0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nvc0_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvc0_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nvc3_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nvc0_fb_oclass;
 		device->oclass[NVDEV_SUBDEV_LTCG   ] = &nvc0_ltcg_oclass;
 		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nvc0_ibus_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nvd0_pwr_oclass;
@@ -318,14 +318,14 @@ nvc0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nvc0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nvc0_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvc0_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nvc3_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nvc0_fb_oclass;
 		device->oclass[NVDEV_SUBDEV_LTCG   ] = &nvc0_ltcg_oclass;
 		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nvc0_ibus_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nvd0_dmaeng_oclass;
diff --git a/drivers/gpu/drm/nouveau/core/engine/device/nve0.c b/drivers/gpu/drm/nouveau/core/engine/device/nve0.c
index 3900104976fc..987edbc30a09 100644
--- a/drivers/gpu/drm/nouveau/core/engine/device/nve0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/device/nve0.c
@@ -65,14 +65,14 @@ nve0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nve0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nvc0_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvc0_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nvc3_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nve0_fb_oclass;
 		device->oclass[NVDEV_SUBDEV_LTCG   ] = &nvc0_ltcg_oclass;
 		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nve0_ibus_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nvd0_pwr_oclass;
@@ -98,14 +98,14 @@ nve0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nve0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nvc0_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvc0_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nvc3_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nve0_fb_oclass;
 		device->oclass[NVDEV_SUBDEV_LTCG   ] = &nvc0_ltcg_oclass;
 		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nve0_ibus_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nvd0_pwr_oclass;
@@ -131,14 +131,14 @@ nve0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nve0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nvc0_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvc0_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nvc3_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nve0_fb_oclass;
 		device->oclass[NVDEV_SUBDEV_LTCG   ] = &nvc0_ltcg_oclass;
 		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nve0_ibus_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nvd0_pwr_oclass;
@@ -164,14 +164,14 @@ nve0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nve0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nvc0_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvc0_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nvc3_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nve0_fb_oclass;
 		device->oclass[NVDEV_SUBDEV_LTCG   ] = &nvc0_ltcg_oclass;
 		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nve0_ibus_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nvd0_pwr_oclass;
@@ -199,29 +199,27 @@ nve0_identify(struct nouveau_device *device)
 		device->oclass[NVDEV_SUBDEV_CLOCK  ] = &nve0_clock_oclass;
 		device->oclass[NVDEV_SUBDEV_THERM  ] = &nvd0_therm_oclass;
 		device->oclass[NVDEV_SUBDEV_MXM    ] = &nv50_mxm_oclass;
-		device->oclass[NVDEV_SUBDEV_DEVINIT] = &nvc0_devinit_oclass;
+		device->oclass[NVDEV_SUBDEV_DEVINIT] =  nvc0_devinit_oclass;
 		device->oclass[NVDEV_SUBDEV_MC     ] =  nvc3_mc_oclass;
 		device->oclass[NVDEV_SUBDEV_BUS    ] =  nvc0_bus_oclass;
 		device->oclass[NVDEV_SUBDEV_TIMER  ] = &nv04_timer_oclass;
 		device->oclass[NVDEV_SUBDEV_FB     ] =  nve0_fb_oclass;
 		device->oclass[NVDEV_SUBDEV_LTCG   ] = &nvc0_ltcg_oclass;
 		device->oclass[NVDEV_SUBDEV_IBUS   ] = &nve0_ibus_oclass;
-		device->oclass[NVDEV_SUBDEV_INSTMEM] = &nv50_instmem_oclass;
+		device->oclass[NVDEV_SUBDEV_INSTMEM] =  nv50_instmem_oclass;
 		device->oclass[NVDEV_SUBDEV_VM     ] = &nvc0_vmmgr_oclass;
 		device->oclass[NVDEV_SUBDEV_BAR    ] = &nvc0_bar_oclass;
 		device->oclass[NVDEV_SUBDEV_PWR    ] = &nv108_pwr_oclass;
 		device->oclass[NVDEV_SUBDEV_VOLT   ] = &nv40_volt_oclass;
 		device->oclass[NVDEV_ENGINE_DMAOBJ ] = &nvd0_dmaeng_oclass;
-#if 0
-		device->oclass[NVDEV_ENGINE_FIFO   ] =  nve0_fifo_oclass;
+		device->oclass[NVDEV_ENGINE_FIFO   ] =  nv108_fifo_oclass;
 		device->oclass[NVDEV_ENGINE_SW     ] =  nvc0_software_oclass;
-		device->oclass[NVDEV_ENGINE_GR     ] =  nvf0_graph_oclass;
-#endif
+		device->oclass[NVDEV_ENGINE_GR     ] =  nv108_graph_oclass;
 		device->oclass[NVDEV_ENGINE_DISP   ] = &nvf0_disp_oclass;
-#if 0
 		device->oclass[NVDEV_ENGINE_COPY0  ] = &nve0_copy0_oclass;
 		device->oclass[NVDEV_ENGINE_COPY1  ] = &nve0_copy1_oclass;
 		device->oclass[NVDEV_ENGINE_COPY2  ] = &nve0_copy2_oclass;
+#if 0
 		device->oclass[NVDEV_ENGINE_BSP    ] = &nve0_bsp_oclass;
 		device->oclass[NVDEV_ENGINE_VP     ] = &nve0_vp_oclass;
 		device->oclass[NVDEV_ENGINE_PPP    ] = &nvc0_ppp_oclass;
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nv04.c b/drivers/gpu/drm/nouveau/core/engine/disp/nv04.c
index a0bc8a89b699..7cf8b1348632 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nv04.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nv04.c
@@ -31,9 +31,45 @@ struct nv04_disp_priv {
 	struct nouveau_disp base;
 };
 
+static int
+nv04_disp_scanoutpos(struct nouveau_object *object, u32 mthd,
+		     void *data, u32 size)
+{
+	struct nv04_disp_priv *priv = (void *)object->engine;
+	struct nv04_display_scanoutpos *args = data;
+	const int head = (mthd & NV04_DISP_MTHD_HEAD);
+	u32 line;
+
+	if (size < sizeof(*args))
+		return -EINVAL;
+
+	args->vblanks = nv_rd32(priv, 0x680800 + (head * 0x2000)) & 0xffff;
+	args->vtotal  = nv_rd32(priv, 0x680804 + (head * 0x2000)) & 0xffff;
+	args->vblanke = args->vtotal - 1;
+
+	args->hblanks = nv_rd32(priv, 0x680820 + (head * 0x2000)) & 0xffff;
+	args->htotal  = nv_rd32(priv, 0x680824 + (head * 0x2000)) & 0xffff;
+	args->hblanke = args->htotal - 1;
+
+	args->time[0] = ktime_to_ns(ktime_get());
+	line = nv_rd32(priv, 0x600868 + (head * 0x2000));
+	args->time[1] = ktime_to_ns(ktime_get());
+	args->hline = (line & 0xffff0000) >> 16;
+	args->vline = (line & 0x0000ffff);
+	return 0;
+}
+
+#define HEAD_MTHD(n) (n), (n) + 0x01
+
+static struct nouveau_omthds
+nv04_disp_omthds[] = {
+	{ HEAD_MTHD(NV04_DISP_SCANOUTPOS), nv04_disp_scanoutpos },
+	{}
+};
+
 static struct nouveau_oclass
 nv04_disp_sclass[] = {
-	{ NV04_DISP_CLASS, &nouveau_object_ofuncs },
+	{ NV04_DISP_CLASS, &nouveau_object_ofuncs, nv04_disp_omthds },
 	{},
 };
 
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nv50.c b/drivers/gpu/drm/nouveau/core/engine/disp/nv50.c
index c168ae3eaa97..940eaa5d8b9a 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nv50.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nv50.c
@@ -541,6 +541,35 @@ nv50_disp_curs_ofuncs = {
  * Base display object
  ******************************************************************************/
 
+int
+nv50_disp_base_scanoutpos(struct nouveau_object *object, u32 mthd,
+			  void *data, u32 size)
+{
+	struct nv50_disp_priv *priv = (void *)object->engine;
+	struct nv04_display_scanoutpos *args = data;
+	const int head = (mthd & NV50_DISP_MTHD_HEAD);
+	u32 blanke, blanks, total;
+
+	if (size < sizeof(*args) || head >= priv->head.nr)
+		return -EINVAL;
+	blanke = nv_rd32(priv, 0x610aec + (head * 0x540));
+	blanks = nv_rd32(priv, 0x610af4 + (head * 0x540));
+	total  = nv_rd32(priv, 0x610afc + (head * 0x540));
+
+	args->vblanke = (blanke & 0xffff0000) >> 16;
+	args->hblanke = (blanke & 0x0000ffff);
+	args->vblanks = (blanks & 0xffff0000) >> 16;
+	args->hblanks = (blanks & 0x0000ffff);
+	args->vtotal  = ( total & 0xffff0000) >> 16;
+	args->htotal  = ( total & 0x0000ffff);
+
+	args->time[0] = ktime_to_ns(ktime_get());
+	args->vline   = nv_rd32(priv, 0x616340 + (head * 0x800)) & 0xffff;
+	args->time[1] = ktime_to_ns(ktime_get()); /* vline read locks hline */
+	args->hline   = nv_rd32(priv, 0x616344 + (head * 0x800)) & 0xffff;
+	return 0;
+}
+
 static void
 nv50_disp_base_vblank_enable(struct nouveau_event *event, int head)
 {
@@ -675,6 +704,7 @@ nv50_disp_base_ofuncs = {
 
 static struct nouveau_omthds
 nv50_disp_base_omthds[] = {
+	{ HEAD_MTHD(NV50_DISP_SCANOUTPOS)     , nv50_disp_base_scanoutpos },
 	{ SOR_MTHD(NV50_DISP_SOR_PWR)         , nv50_sor_mthd },
 	{ SOR_MTHD(NV50_DISP_SOR_LVDS_SCRIPT) , nv50_sor_mthd },
 	{ DAC_MTHD(NV50_DISP_DAC_PWR)         , nv50_dac_mthd },
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nv50.h b/drivers/gpu/drm/nouveau/core/engine/disp/nv50.h
index 1ae6ceb56704..d31d426ea1f6 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nv50.h
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nv50.h
@@ -43,6 +43,10 @@ struct nv50_disp_priv {
 	} pior;
 };
 
+#define HEAD_MTHD(n) (n), (n) + 0x03
+
+int nv50_disp_base_scanoutpos(struct nouveau_object *, u32, void *, u32);
+
 #define DAC_MTHD(n) (n), (n) + 0x03
 
 int nv50_dac_mthd(struct nouveau_object *, u32, void *, u32);
@@ -132,13 +136,12 @@ void nv50_disp_intr(struct nouveau_subdev *);
 
 extern struct nouveau_omthds nv84_disp_base_omthds[];
 
-extern struct nouveau_omthds nva3_disp_base_omthds[];
-
 extern struct nouveau_ofuncs nvd0_disp_mast_ofuncs;
 extern struct nouveau_ofuncs nvd0_disp_sync_ofuncs;
 extern struct nouveau_ofuncs nvd0_disp_ovly_ofuncs;
 extern struct nouveau_ofuncs nvd0_disp_oimm_ofuncs;
 extern struct nouveau_ofuncs nvd0_disp_curs_ofuncs;
+extern struct nouveau_omthds nvd0_disp_base_omthds[];
 extern struct nouveau_ofuncs nvd0_disp_base_ofuncs;
 extern struct nouveau_oclass nvd0_disp_cclass;
 void nvd0_disp_intr_supervisor(struct work_struct *);
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nv84.c b/drivers/gpu/drm/nouveau/core/engine/disp/nv84.c
index d8c74c0883a1..ef9ce300a496 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nv84.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nv84.c
@@ -41,6 +41,7 @@ nv84_disp_sclass[] = {
 
 struct nouveau_omthds
 nv84_disp_base_omthds[] = {
+	{ HEAD_MTHD(NV50_DISP_SCANOUTPOS)     , nv50_disp_base_scanoutpos },
 	{ SOR_MTHD(NV50_DISP_SOR_PWR)         , nv50_sor_mthd },
 	{ SOR_MTHD(NV84_DISP_SOR_HDMI_PWR)    , nv50_sor_mthd },
 	{ SOR_MTHD(NV50_DISP_SOR_LVDS_SCRIPT) , nv50_sor_mthd },
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nv94.c b/drivers/gpu/drm/nouveau/core/engine/disp/nv94.c
index a66f949c1f84..a518543c00ab 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nv94.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nv94.c
@@ -41,6 +41,7 @@ nv94_disp_sclass[] = {
 
 static struct nouveau_omthds
 nv94_disp_base_omthds[] = {
+	{ HEAD_MTHD(NV50_DISP_SCANOUTPOS)     , nv50_disp_base_scanoutpos },
 	{ SOR_MTHD(NV50_DISP_SOR_PWR)         , nv50_sor_mthd },
 	{ SOR_MTHD(NV84_DISP_SOR_HDMI_PWR)    , nv50_sor_mthd },
 	{ SOR_MTHD(NV50_DISP_SOR_LVDS_SCRIPT) , nv50_sor_mthd },
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nva3.c b/drivers/gpu/drm/nouveau/core/engine/disp/nva3.c
index b75413169eae..6ad6dcece43b 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nva3.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nva3.c
@@ -39,8 +39,9 @@ nva3_disp_sclass[] = {
 	{}
 };
 
-struct nouveau_omthds
+static struct nouveau_omthds
 nva3_disp_base_omthds[] = {
+	{ HEAD_MTHD(NV50_DISP_SCANOUTPOS)     , nv50_disp_base_scanoutpos },
 	{ SOR_MTHD(NV50_DISP_SOR_PWR)         , nv50_sor_mthd },
 	{ SOR_MTHD(NVA3_DISP_SOR_HDA_ELD)     , nv50_sor_mthd },
 	{ SOR_MTHD(NV84_DISP_SOR_HDMI_PWR)    , nv50_sor_mthd },
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nvd0.c b/drivers/gpu/drm/nouveau/core/engine/disp/nvd0.c
index 378a015091d2..1c5e4e8b2c82 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nvd0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nvd0.c
@@ -440,6 +440,36 @@ nvd0_disp_curs_ofuncs = {
  * Base display object
  ******************************************************************************/
 
+static int
+nvd0_disp_base_scanoutpos(struct nouveau_object *object, u32 mthd,
+			  void *data, u32 size)
+{
+	struct nv50_disp_priv *priv = (void *)object->engine;
+	struct nv04_display_scanoutpos *args = data;
+	const int head = (mthd & NV50_DISP_MTHD_HEAD);
+	u32 blanke, blanks, total;
+
+	if (size < sizeof(*args) || head >= priv->head.nr)
+		return -EINVAL;
+
+	total  = nv_rd32(priv, 0x640414 + (head * 0x300));
+	blanke = nv_rd32(priv, 0x64041c + (head * 0x300));
+	blanks = nv_rd32(priv, 0x640420 + (head * 0x300));
+
+	args->vblanke = (blanke & 0xffff0000) >> 16;
+	args->hblanke = (blanke & 0x0000ffff);
+	args->vblanks = (blanks & 0xffff0000) >> 16;
+	args->hblanks = (blanks & 0x0000ffff);
+	args->vtotal  = ( total & 0xffff0000) >> 16;
+	args->htotal  = ( total & 0x0000ffff);
+
+	args->time[0] = ktime_to_ns(ktime_get());
+	args->vline   = nv_rd32(priv, 0x616340 + (head * 0x800)) & 0xffff;
+	args->time[1] = ktime_to_ns(ktime_get()); /* vline read locks hline */
+	args->hline   = nv_rd32(priv, 0x616344 + (head * 0x800)) & 0xffff;
+	return 0;
+}
+
 static void
 nvd0_disp_base_vblank_enable(struct nouveau_event *event, int head)
 {
@@ -573,9 +603,24 @@ nvd0_disp_base_ofuncs = {
 	.fini = nvd0_disp_base_fini,
 };
 
+struct nouveau_omthds
+nvd0_disp_base_omthds[] = {
+	{ HEAD_MTHD(NV50_DISP_SCANOUTPOS)     , nvd0_disp_base_scanoutpos },
+	{ SOR_MTHD(NV50_DISP_SOR_PWR)         , nv50_sor_mthd },
+	{ SOR_MTHD(NVA3_DISP_SOR_HDA_ELD)     , nv50_sor_mthd },
+	{ SOR_MTHD(NV84_DISP_SOR_HDMI_PWR)    , nv50_sor_mthd },
+	{ SOR_MTHD(NV50_DISP_SOR_LVDS_SCRIPT) , nv50_sor_mthd },
+	{ DAC_MTHD(NV50_DISP_DAC_PWR)         , nv50_dac_mthd },
+	{ DAC_MTHD(NV50_DISP_DAC_LOAD)        , nv50_dac_mthd },
+	{ PIOR_MTHD(NV50_DISP_PIOR_PWR)       , nv50_pior_mthd },
+	{ PIOR_MTHD(NV50_DISP_PIOR_TMDS_PWR)  , nv50_pior_mthd },
+	{ PIOR_MTHD(NV50_DISP_PIOR_DP_PWR)    , nv50_pior_mthd },
+	{},
+};
+
 static struct nouveau_oclass
 nvd0_disp_base_oclass[] = {
-	{ NVD0_DISP_CLASS, &nvd0_disp_base_ofuncs, nva3_disp_base_omthds },
+	{ NVD0_DISP_CLASS, &nvd0_disp_base_ofuncs, nvd0_disp_base_omthds },
 	{}
 };
 
@@ -967,9 +1012,6 @@ nvd0_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	int heads = nv_rd32(parent, 0x022448);
 	int ret;
 
-	if (nv_rd32(parent, 0x022500) & 0x00000001)
-		return -ENODEV;
-
 	ret = nouveau_disp_create(parent, engine, oclass, heads,
 				  "PDISP", "display", &priv);
 	*pobject = nv_object(priv);
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nve0.c b/drivers/gpu/drm/nouveau/core/engine/disp/nve0.c
index fb1fe6ae5e74..ab63f32c00b2 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nve0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nve0.c
@@ -41,7 +41,7 @@ nve0_disp_sclass[] = {
 
 static struct nouveau_oclass
 nve0_disp_base_oclass[] = {
-	{ NVE0_DISP_CLASS, &nvd0_disp_base_ofuncs, nva3_disp_base_omthds },
+	{ NVE0_DISP_CLASS, &nvd0_disp_base_ofuncs, nvd0_disp_base_omthds },
 	{}
 };
 
@@ -54,9 +54,6 @@ nve0_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	int heads = nv_rd32(parent, 0x022448);
 	int ret;
 
-	if (nv_rd32(parent, 0x022500) & 0x00000001)
-		return -ENODEV;
-
 	ret = nouveau_disp_create(parent, engine, oclass, heads,
 				  "PDISP", "display", &priv);
 	*pobject = nv_object(priv);
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nvf0.c b/drivers/gpu/drm/nouveau/core/engine/disp/nvf0.c
index 42aa6b97dbea..05fee10e0c97 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nvf0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nvf0.c
@@ -41,7 +41,7 @@ nvf0_disp_sclass[] = {
 
 static struct nouveau_oclass
 nvf0_disp_base_oclass[] = {
-	{ NVF0_DISP_CLASS, &nvd0_disp_base_ofuncs, nva3_disp_base_omthds },
+	{ NVF0_DISP_CLASS, &nvd0_disp_base_ofuncs, nvd0_disp_base_omthds },
 	{}
 };
 
@@ -54,9 +54,6 @@ nvf0_disp_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	int heads = nv_rd32(parent, 0x022448);
 	int ret;
 
-	if (nv_rd32(parent, 0x022500) & 0x00000001)
-		return -ENODEV;
-
 	ret = nouveau_disp_create(parent, engine, oclass, heads,
 				  "PDISP", "display", &priv);
 	*pobject = nv_object(priv);
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/vga.c b/drivers/gpu/drm/nouveau/core/engine/disp/vga.c
index 5a1c68474597..8836c3cb99c3 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/vga.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/vga.c
@@ -138,10 +138,15 @@ nv_wrvgai(void *obj, int head, u16 port, u8 index, u8 value)
 bool
 nv_lockvgac(void *obj, bool lock)
 {
+	struct nouveau_device *dev = nv_device(obj);
+
 	bool locked = !nv_rdvgac(obj, 0, 0x1f);
 	u8 data = lock ? 0x99 : 0x57;
-	nv_wrvgac(obj, 0, 0x1f, data);
-	if (nv_device(obj)->chipset == 0x11) {
+	if (dev->card_type < NV_50)
+		nv_wrvgac(obj, 0, 0x1f, data);
+	else
+		nv_wrvgac(obj, 0, 0x3f, data);
+	if (dev->chipset == 0x11) {
 		if (!(nv_rd32(obj, 0x001084) & 0x10000000))
 			nv_wrvgac(obj, 1, 0x1f, data);
 	}
diff --git a/drivers/gpu/drm/nouveau/core/engine/falcon.c b/drivers/gpu/drm/nouveau/core/engine/falcon.c
index e03fc8e4dc1d..5e077e4ed7f6 100644
--- a/drivers/gpu/drm/nouveau/core/engine/falcon.c
+++ b/drivers/gpu/drm/nouveau/core/engine/falcon.c
@@ -56,6 +56,16 @@ _nouveau_falcon_wr32(struct nouveau_object *object, u64 addr, u32 data)
 	nv_wr32(falcon, falcon->addr + addr, data);
 }
 
+static void *
+vmemdup(const void *src, size_t len)
+{
+	void *p = vmalloc(len);
+
+	if (p)
+		memcpy(p, src, len);
+	return p;
+}
+
 int
 _nouveau_falcon_init(struct nouveau_object *object)
 {
@@ -111,7 +121,7 @@ _nouveau_falcon_init(struct nouveau_object *object)
 
 		ret = request_firmware(&fw, name, &device->pdev->dev);
 		if (ret == 0) {
-			falcon->code.data = kmemdup(fw->data, fw->size, GFP_KERNEL);
+			falcon->code.data = vmemdup(fw->data, fw->size);
 			falcon->code.size = fw->size;
 			falcon->data.data = NULL;
 			falcon->data.size = 0;
@@ -134,7 +144,7 @@ _nouveau_falcon_init(struct nouveau_object *object)
 			return ret;
 		}
 
-		falcon->data.data = kmemdup(fw->data, fw->size, GFP_KERNEL);
+		falcon->data.data = vmemdup(fw->data, fw->size);
 		falcon->data.size = fw->size;
 		release_firmware(fw);
 		if (!falcon->data.data)
@@ -149,7 +159,7 @@ _nouveau_falcon_init(struct nouveau_object *object)
 			return ret;
 		}
 
-		falcon->code.data = kmemdup(fw->data, fw->size, GFP_KERNEL);
+		falcon->code.data = vmemdup(fw->data, fw->size);
 		falcon->code.size = fw->size;
 		release_firmware(fw);
 		if (!falcon->code.data)
@@ -235,8 +245,8 @@ _nouveau_falcon_fini(struct nouveau_object *object, bool suspend)
 	if (!suspend) {
 		nouveau_gpuobj_ref(NULL, &falcon->core);
 		if (falcon->external) {
-			kfree(falcon->data.data);
-			kfree(falcon->code.data);
+			vfree(falcon->data.data);
+			vfree(falcon->code.data);
 			falcon->code.data = NULL;
 		}
 	}
diff --git a/drivers/gpu/drm/nouveau/core/engine/fifo/nv108.c b/drivers/gpu/drm/nouveau/core/engine/fifo/nv108.c
new file mode 100644
index 000000000000..09362a51ba57
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/fifo/nv108.c
@@ -0,0 +1,37 @@
+/*
+ * 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 "nve0.h"
+
+struct nouveau_oclass *
+nv108_fifo_oclass = &(struct nve0_fifo_impl) {
+	.base.handle = NV_ENGINE(FIFO, 0x08),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nve0_fifo_ctor,
+		.dtor = nve0_fifo_dtor,
+		.init = nve0_fifo_init,
+		.fini = _nouveau_fifo_fini,
+	},
+	.channels = 1024,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/fifo/nvc0.c b/drivers/gpu/drm/nouveau/core/engine/fifo/nvc0.c
index 9ac94d4e5646..b22a33f0702d 100644
--- a/drivers/gpu/drm/nouveau/core/engine/fifo/nvc0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/fifo/nvc0.c
@@ -33,6 +33,7 @@
 
 #include <subdev/timer.h>
 #include <subdev/bar.h>
+#include <subdev/fb.h>
 #include <subdev/vm.h>
 
 #include <engine/dmaobj.h>
diff --git a/drivers/gpu/drm/nouveau/core/engine/fifo/nve0.c b/drivers/gpu/drm/nouveau/core/engine/fifo/nve0.c
index 04f412922d2d..9a850fe19515 100644
--- a/drivers/gpu/drm/nouveau/core/engine/fifo/nve0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/fifo/nve0.c
@@ -33,10 +33,12 @@
 
 #include <subdev/timer.h>
 #include <subdev/bar.h>
+#include <subdev/fb.h>
 #include <subdev/vm.h>
 
 #include <engine/dmaobj.h>
-#include <engine/fifo.h>
+
+#include "nve0.h"
 
 #define _(a,b) { (a), ((1ULL << (a)) | (b)) }
 static const struct {
@@ -56,8 +58,8 @@ static const struct {
 #define FIFO_ENGINE_NR ARRAY_SIZE(fifo_engine)
 
 struct nve0_fifo_engn {
-	struct nouveau_gpuobj *playlist[2];
-	int cur_playlist;
+	struct nouveau_gpuobj *runlist[2];
+	int cur_runlist;
 };
 
 struct nve0_fifo_priv {
@@ -86,7 +88,7 @@ struct nve0_fifo_chan {
  ******************************************************************************/
 
 static void
-nve0_fifo_playlist_update(struct nve0_fifo_priv *priv, u32 engine)
+nve0_fifo_runlist_update(struct nve0_fifo_priv *priv, u32 engine)
 {
 	struct nouveau_bar *bar = nouveau_bar(priv);
 	struct nve0_fifo_engn *engn = &priv->engine[engine];
@@ -95,8 +97,8 @@ nve0_fifo_playlist_update(struct nve0_fifo_priv *priv, u32 engine)
 	int i, p;
 
 	mutex_lock(&nv_subdev(priv)->mutex);
-	cur = engn->playlist[engn->cur_playlist];
-	engn->cur_playlist = !engn->cur_playlist;
+	cur = engn->runlist[engn->cur_runlist];
+	engn->cur_runlist = !engn->cur_runlist;
 
 	for (i = 0, p = 0; i < priv->base.max; i++) {
 		u32 ctrl = nv_rd32(priv, 0x800004 + (i * 8)) & 0x001f0001;
@@ -111,7 +113,7 @@ nve0_fifo_playlist_update(struct nve0_fifo_priv *priv, u32 engine)
 	nv_wr32(priv, 0x002270, cur->addr >> 12);
 	nv_wr32(priv, 0x002274, (engine << 20) | (p >> 3));
 	if (!nv_wait(priv, 0x002284 + (engine * 4), 0x00100000, 0x00000000))
-		nv_error(priv, "playlist %d update timeout\n", engine);
+		nv_error(priv, "runlist %d update timeout\n", engine);
 	mutex_unlock(&nv_subdev(priv)->mutex);
 }
 
@@ -278,7 +280,7 @@ nve0_fifo_chan_init(struct nouveau_object *object)
 	nv_mask(priv, 0x800004 + (chid * 8), 0x000f0000, chan->engine << 16);
 	nv_wr32(priv, 0x800000 + (chid * 8), 0x80000000 | base->addr >> 12);
 	nv_mask(priv, 0x800004 + (chid * 8), 0x00000400, 0x00000400);
-	nve0_fifo_playlist_update(priv, chan->engine);
+	nve0_fifo_runlist_update(priv, chan->engine);
 	nv_mask(priv, 0x800004 + (chid * 8), 0x00000400, 0x00000400);
 	return 0;
 }
@@ -291,7 +293,7 @@ nve0_fifo_chan_fini(struct nouveau_object *object, bool suspend)
 	u32 chid = chan->base.chid;
 
 	nv_mask(priv, 0x800004 + (chid * 8), 0x00000800, 0x00000800);
-	nve0_fifo_playlist_update(priv, chan->engine);
+	nve0_fifo_runlist_update(priv, chan->engine);
 	nv_wr32(priv, 0x800000 + (chid * 8), 0x00000000);
 
 	return nouveau_fifo_channel_fini(&chan->base, suspend);
@@ -375,54 +377,189 @@ nve0_fifo_cclass = {
  * PFIFO engine
  ******************************************************************************/
 
-static const struct nouveau_enum nve0_fifo_fault_unit[] = {
+static const struct nouveau_enum nve0_fifo_sched_reason[] = {
+	{ 0x0a, "CTXSW_TIMEOUT" },
+	{}
+};
+
+static const struct nouveau_enum nve0_fifo_fault_engine[] = {
+	{ 0x00, "GR", NULL, NVDEV_ENGINE_GR },
+	{ 0x03, "IFB" },
+	{ 0x04, "BAR1", NULL, NVDEV_SUBDEV_BAR },
+	{ 0x05, "BAR3", NULL, NVDEV_SUBDEV_INSTMEM },
+	{ 0x07, "PBDMA0", NULL, NVDEV_ENGINE_FIFO },
+	{ 0x08, "PBDMA1", NULL, NVDEV_ENGINE_FIFO },
+	{ 0x09, "PBDMA2", NULL, NVDEV_ENGINE_FIFO },
+	{ 0x10, "MSVLD", NULL, NVDEV_ENGINE_BSP },
+	{ 0x11, "MSPPP", NULL, NVDEV_ENGINE_PPP },
+	{ 0x13, "PERF" },
+	{ 0x14, "MSPDEC", NULL, NVDEV_ENGINE_VP },
+	{ 0x15, "CE0", NULL, NVDEV_ENGINE_COPY0 },
+	{ 0x16, "CE1", NULL, NVDEV_ENGINE_COPY1 },
+	{ 0x17, "PMU" },
+	{ 0x19, "MSENC", NULL, NVDEV_ENGINE_VENC },
+	{ 0x1b, "CE2", NULL, NVDEV_ENGINE_COPY2 },
 	{}
 };
 
 static const struct nouveau_enum nve0_fifo_fault_reason[] = {
-	{ 0x00, "PT_NOT_PRESENT" },
-	{ 0x01, "PT_TOO_SHORT" },
-	{ 0x02, "PAGE_NOT_PRESENT" },
-	{ 0x03, "VM_LIMIT_EXCEEDED" },
-	{ 0x04, "NO_CHANNEL" },
-	{ 0x05, "PAGE_SYSTEM_ONLY" },
-	{ 0x06, "PAGE_READ_ONLY" },
-	{ 0x0a, "COMPRESSED_SYSRAM" },
-	{ 0x0c, "INVALID_STORAGE_TYPE" },
+	{ 0x00, "PDE" },
+	{ 0x01, "PDE_SIZE" },
+	{ 0x02, "PTE" },
+	{ 0x03, "VA_LIMIT_VIOLATION" },
+	{ 0x04, "UNBOUND_INST_BLOCK" },
+	{ 0x05, "PRIV_VIOLATION" },
+	{ 0x06, "RO_VIOLATION" },
+	{ 0x07, "WO_VIOLATION" },
+	{ 0x08, "PITCH_MASK_VIOLATION" },
+	{ 0x09, "WORK_CREATION" },
+	{ 0x0a, "UNSUPPORTED_APERTURE" },
+	{ 0x0b, "COMPRESSION_FAILURE" },
+	{ 0x0c, "UNSUPPORTED_KIND" },
+	{ 0x0d, "REGION_VIOLATION" },
+	{ 0x0e, "BOTH_PTES_VALID" },
+	{ 0x0f, "INFO_TYPE_POISONED" },
 	{}
 };
 
 static const struct nouveau_enum nve0_fifo_fault_hubclient[] = {
+	{ 0x00, "VIP" },
+	{ 0x01, "CE0" },
+	{ 0x02, "CE1" },
+	{ 0x03, "DNISO" },
+	{ 0x04, "FE" },
+	{ 0x05, "FECS" },
+	{ 0x06, "HOST" },
+	{ 0x07, "HOST_CPU" },
+	{ 0x08, "HOST_CPU_NB" },
+	{ 0x09, "ISO" },
+	{ 0x0a, "MMU" },
+	{ 0x0b, "MSPDEC" },
+	{ 0x0c, "MSPPP" },
+	{ 0x0d, "MSVLD" },
+	{ 0x0e, "NISO" },
+	{ 0x0f, "P2P" },
+	{ 0x10, "PD" },
+	{ 0x11, "PERF" },
+	{ 0x12, "PMU" },
+	{ 0x13, "RASTERTWOD" },
+	{ 0x14, "SCC" },
+	{ 0x15, "SCC_NB" },
+	{ 0x16, "SEC" },
+	{ 0x17, "SSYNC" },
+	{ 0x18, "GR_COPY" },
+	{ 0x19, "CE2" },
+	{ 0x1a, "XV" },
+	{ 0x1b, "MMU_NB" },
+	{ 0x1c, "MSENC" },
+	{ 0x1d, "DFALCON" },
+	{ 0x1e, "SKED" },
+	{ 0x1f, "AFALCON" },
 	{}
 };
 
 static const struct nouveau_enum nve0_fifo_fault_gpcclient[] = {
+	{ 0x00, "L1_0" }, { 0x01, "T1_0" }, { 0x02, "PE_0" },
+	{ 0x03, "L1_1" }, { 0x04, "T1_1" }, { 0x05, "PE_1" },
+	{ 0x06, "L1_2" }, { 0x07, "T1_2" }, { 0x08, "PE_2" },
+	{ 0x09, "L1_3" }, { 0x0a, "T1_3" }, { 0x0b, "PE_3" },
+	{ 0x0c, "RAST" },
+	{ 0x0d, "GCC" },
+	{ 0x0e, "GPCCS" },
+	{ 0x0f, "PROP_0" },
+	{ 0x10, "PROP_1" },
+	{ 0x11, "PROP_2" },
+	{ 0x12, "PROP_3" },
+	{ 0x13, "L1_4" }, { 0x14, "T1_4" }, { 0x15, "PE_4" },
+	{ 0x16, "L1_5" }, { 0x17, "T1_5" }, { 0x18, "PE_5" },
+	{ 0x19, "L1_6" }, { 0x1a, "T1_6" }, { 0x1b, "PE_6" },
+	{ 0x1c, "L1_7" }, { 0x1d, "T1_7" }, { 0x1e, "PE_7" },
+	{ 0x1f, "GPM" },
+	{ 0x20, "LTP_UTLB_0" },
+	{ 0x21, "LTP_UTLB_1" },
+	{ 0x22, "LTP_UTLB_2" },
+	{ 0x23, "LTP_UTLB_3" },
+	{ 0x24, "GPC_RGG_UTLB" },
 	{}
 };
 
-static const struct nouveau_bitfield nve0_fifo_subfifo_intr[] = {
-	{ 0x00200000, "ILLEGAL_MTHD" },
-	{ 0x00800000, "EMPTY_SUBC" },
+static const struct nouveau_bitfield nve0_fifo_pbdma_intr[] = {
+	{ 0x00000001, "MEMREQ" },
+	{ 0x00000002, "MEMACK_TIMEOUT" },
+	{ 0x00000004, "MEMACK_EXTRA" },
+	{ 0x00000008, "MEMDAT_TIMEOUT" },
+	{ 0x00000010, "MEMDAT_EXTRA" },
+	{ 0x00000020, "MEMFLUSH" },
+	{ 0x00000040, "MEMOP" },
+	{ 0x00000080, "LBCONNECT" },
+	{ 0x00000100, "LBREQ" },
+	{ 0x00000200, "LBACK_TIMEOUT" },
+	{ 0x00000400, "LBACK_EXTRA" },
+	{ 0x00000800, "LBDAT_TIMEOUT" },
+	{ 0x00001000, "LBDAT_EXTRA" },
+	{ 0x00002000, "GPFIFO" },
+	{ 0x00004000, "GPPTR" },
+	{ 0x00008000, "GPENTRY" },
+	{ 0x00010000, "GPCRC" },
+	{ 0x00020000, "PBPTR" },
+	{ 0x00040000, "PBENTRY" },
+	{ 0x00080000, "PBCRC" },
+	{ 0x00100000, "XBARCONNECT" },
+	{ 0x00200000, "METHOD" },
+	{ 0x00400000, "METHODCRC" },
+	{ 0x00800000, "DEVICE" },
+	{ 0x02000000, "SEMAPHORE" },
+	{ 0x04000000, "ACQUIRE" },
+	{ 0x08000000, "PRI" },
+	{ 0x20000000, "NO_CTXSW_SEG" },
+	{ 0x40000000, "PBSEG" },
+	{ 0x80000000, "SIGNATURE" },
 	{}
 };
 
 static void
-nve0_fifo_isr_vm_fault(struct nve0_fifo_priv *priv, int unit)
+nve0_fifo_intr_sched(struct nve0_fifo_priv *priv)
+{
+	u32 intr = nv_rd32(priv, 0x00254c);
+	u32 code = intr & 0x000000ff;
+	nv_error(priv, "SCHED_ERROR [");
+	nouveau_enum_print(nve0_fifo_sched_reason, code);
+	pr_cont("]\n");
+}
+
+static void
+nve0_fifo_intr_chsw(struct nve0_fifo_priv *priv)
+{
+	u32 stat = nv_rd32(priv, 0x00256c);
+	nv_error(priv, "CHSW_ERROR 0x%08x\n", stat);
+	nv_wr32(priv, 0x00256c, stat);
+}
+
+static void
+nve0_fifo_intr_dropped_fault(struct nve0_fifo_priv *priv)
+{
+	u32 stat = nv_rd32(priv, 0x00259c);
+	nv_error(priv, "DROPPED_MMU_FAULT 0x%08x\n", stat);
+}
+
+static void
+nve0_fifo_intr_fault(struct nve0_fifo_priv *priv, int unit)
 {
 	u32 inst = nv_rd32(priv, 0x2800 + (unit * 0x10));
 	u32 valo = nv_rd32(priv, 0x2804 + (unit * 0x10));
 	u32 vahi = nv_rd32(priv, 0x2808 + (unit * 0x10));
 	u32 stat = nv_rd32(priv, 0x280c + (unit * 0x10));
 	u32 client = (stat & 0x00001f00) >> 8;
-	const struct nouveau_enum *en;
-	struct nouveau_engine *engine;
+	struct nouveau_engine *engine = NULL;
 	struct nouveau_object *engctx = NULL;
+	const struct nouveau_enum *en;
+	const char *name = "unknown";
 
 	nv_error(priv, "PFIFO: %s fault at 0x%010llx [", (stat & 0x00000080) ?
 		       "write" : "read", (u64)vahi << 32 | valo);
 	nouveau_enum_print(nve0_fifo_fault_reason, stat & 0x0000000f);
 	pr_cont("] from ");
-	en = nouveau_enum_print(nve0_fifo_fault_unit, unit);
+	en = nouveau_enum_print(nve0_fifo_fault_engine, unit);
 	if (stat & 0x00000040) {
 		pr_cont("/");
 		nouveau_enum_print(nve0_fifo_fault_hubclient, client);
@@ -432,14 +569,22 @@ nve0_fifo_isr_vm_fault(struct nve0_fifo_priv *priv, int unit)
 	}
 
 	if (en && en->data2) {
-		engine = nouveau_engine(priv, en->data2);
-		if (engine)
-			engctx = nouveau_engctx_get(engine, inst);
-
+		if (en->data2 == NVDEV_SUBDEV_BAR) {
+			nv_mask(priv, 0x001704, 0x00000000, 0x00000000);
+			name = "BAR1";
+		} else
+		if (en->data2 == NVDEV_SUBDEV_INSTMEM) {
+			nv_mask(priv, 0x001714, 0x00000000, 0x00000000);
+			name = "BAR3";
+		} else {
+			engine = nouveau_engine(priv, en->data2);
+			if (engine) {
+				engctx = nouveau_engctx_get(engine, inst);
+				name   = nouveau_client_name(engctx);
+			}
+		}
 	}
-
-	pr_cont(" on channel 0x%010llx [%s]\n", (u64)inst << 12,
-			nouveau_client_name(engctx));
+	pr_cont(" on channel 0x%010llx [%s]\n", (u64)inst << 12, name);
 
 	nouveau_engctx_put(engctx);
 }
@@ -471,7 +616,7 @@ out:
 }
 
 static void
-nve0_fifo_isr_subfifo_intr(struct nve0_fifo_priv *priv, int unit)
+nve0_fifo_intr_pbdma(struct nve0_fifo_priv *priv, int unit)
 {
 	u32 stat = nv_rd32(priv, 0x040108 + (unit * 0x2000));
 	u32 addr = nv_rd32(priv, 0x0400c0 + (unit * 0x2000));
@@ -487,11 +632,11 @@ nve0_fifo_isr_subfifo_intr(struct nve0_fifo_priv *priv, int unit)
 	}
 
 	if (show) {
-		nv_error(priv, "SUBFIFO%d:", unit);
-		nouveau_bitfield_print(nve0_fifo_subfifo_intr, show);
+		nv_error(priv, "PBDMA%d:", unit);
+		nouveau_bitfield_print(nve0_fifo_pbdma_intr, show);
 		pr_cont("\n");
 		nv_error(priv,
-			 "SUBFIFO%d: ch %d [%s] subc %d mthd 0x%04x data 0x%08x\n",
+			 "PBDMA%d: ch %d [%s] subc %d mthd 0x%04x data 0x%08x\n",
 			 unit, chid,
 			 nouveau_client_name_for_fifo_chid(&priv->base, chid),
 			 subc, mthd, data);
@@ -508,19 +653,56 @@ nve0_fifo_intr(struct nouveau_subdev *subdev)
 	u32 mask = nv_rd32(priv, 0x002140);
 	u32 stat = nv_rd32(priv, 0x002100) & mask;
 
+	if (stat & 0x00000001) {
+		u32 stat = nv_rd32(priv, 0x00252c);
+		nv_error(priv, "BIND_ERROR 0x%08x\n", stat);
+		nv_wr32(priv, 0x002100, 0x00000001);
+		stat &= ~0x00000001;
+	}
+
+	if (stat & 0x00000010) {
+		nv_error(priv, "PIO_ERROR\n");
+		nv_wr32(priv, 0x002100, 0x00000010);
+		stat &= ~0x00000010;
+	}
+
 	if (stat & 0x00000100) {
-		nv_warn(priv, "unknown status 0x00000100\n");
+		nve0_fifo_intr_sched(priv);
 		nv_wr32(priv, 0x002100, 0x00000100);
 		stat &= ~0x00000100;
 	}
 
+	if (stat & 0x00010000) {
+		nve0_fifo_intr_chsw(priv);
+		nv_wr32(priv, 0x002100, 0x00010000);
+		stat &= ~0x00010000;
+	}
+
+	if (stat & 0x00800000) {
+		nv_error(priv, "FB_FLUSH_TIMEOUT\n");
+		nv_wr32(priv, 0x002100, 0x00800000);
+		stat &= ~0x00800000;
+	}
+
+	if (stat & 0x01000000) {
+		nv_error(priv, "LB_ERROR\n");
+		nv_wr32(priv, 0x002100, 0x01000000);
+		stat &= ~0x01000000;
+	}
+
+	if (stat & 0x08000000) {
+		nve0_fifo_intr_dropped_fault(priv);
+		nv_wr32(priv, 0x002100, 0x08000000);
+		stat &= ~0x08000000;
+	}
+
 	if (stat & 0x10000000) {
 		u32 units = nv_rd32(priv, 0x00259c);
 		u32 u = units;
 
 		while (u) {
 			int i = ffs(u) - 1;
-			nve0_fifo_isr_vm_fault(priv, i);
+			nve0_fifo_intr_fault(priv, i);
 			u &= ~(1 << i);
 		}
 
@@ -529,22 +711,28 @@ nve0_fifo_intr(struct nouveau_subdev *subdev)
 	}
 
 	if (stat & 0x20000000) {
-		u32 units = nv_rd32(priv, 0x0025a0);
-		u32 u = units;
+		u32 mask = nv_rd32(priv, 0x0025a0);
+		u32 temp = mask;
 
-		while (u) {
-			int i = ffs(u) - 1;
-			nve0_fifo_isr_subfifo_intr(priv, i);
-			u &= ~(1 << i);
+		while (temp) {
+			u32 unit = ffs(temp) - 1;
+			nve0_fifo_intr_pbdma(priv, unit);
+			temp &= ~(1 << unit);
 		}
 
-		nv_wr32(priv, 0x0025a0, units);
+		nv_wr32(priv, 0x0025a0, mask);
 		stat &= ~0x20000000;
 	}
 
 	if (stat & 0x40000000) {
-		nv_warn(priv, "unknown status 0x40000000\n");
-		nv_mask(priv, 0x002a00, 0x00000000, 0x00000000);
+		u32 mask = nv_mask(priv, 0x002a00, 0x00000000, 0x00000000);
+
+		while (mask) {
+			u32 engn = ffs(mask) - 1;
+			/* runlist event, not currently used */
+			mask &= ~(1 << engn);
+		}
+
 		stat &= ~0x40000000;
 	}
 
@@ -575,53 +763,52 @@ nve0_fifo_uevent_disable(struct nouveau_event *event, int index)
 	nv_mask(priv, 0x002140, 0x80000000, 0x00000000);
 }
 
-static int
-nve0_fifo_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
-	       struct nouveau_oclass *oclass, void *data, u32 size,
-	       struct nouveau_object **pobject)
+int
+nve0_fifo_fini(struct nouveau_object *object, bool suspend)
 {
-	struct nve0_fifo_priv *priv;
-	int ret, i;
+	struct nve0_fifo_priv *priv = (void *)object;
+	int ret;
 
-	ret = nouveau_fifo_create(parent, engine, oclass, 0, 4095, &priv);
-	*pobject = nv_object(priv);
+	ret = nouveau_fifo_fini(&priv->base, suspend);
 	if (ret)
 		return ret;
 
-	for (i = 0; i < FIFO_ENGINE_NR; i++) {
-		ret = nouveau_gpuobj_new(nv_object(priv), NULL, 0x8000, 0x1000,
-					 0, &priv->engine[i].playlist[0]);
-		if (ret)
-			return ret;
+	/* allow mmu fault interrupts, even when we're not using fifo */
+	nv_mask(priv, 0x002140, 0x10000000, 0x10000000);
+	return 0;
+}
 
-		ret = nouveau_gpuobj_new(nv_object(priv), NULL, 0x8000, 0x1000,
-					 0, &priv->engine[i].playlist[1]);
-		if (ret)
-			return ret;
-	}
+int
+nve0_fifo_init(struct nouveau_object *object)
+{
+	struct nve0_fifo_priv *priv = (void *)object;
+	int ret, i;
 
-	ret = nouveau_gpuobj_new(nv_object(priv), NULL, 4096 * 0x200, 0x1000,
-				 NVOBJ_FLAG_ZERO_ALLOC, &priv->user.mem);
+	ret = nouveau_fifo_init(&priv->base);
 	if (ret)
 		return ret;
 
-	ret = nouveau_gpuobj_map(priv->user.mem, NV_MEM_ACCESS_RW,
-				&priv->user.bar);
-	if (ret)
-		return ret;
+	/* enable all available PBDMA units */
+	nv_wr32(priv, 0x000204, 0xffffffff);
+	priv->spoon_nr = hweight32(nv_rd32(priv, 0x000204));
+	nv_debug(priv, "%d PBDMA unit(s)\n", priv->spoon_nr);
 
-	priv->base.uevent->enable = nve0_fifo_uevent_enable;
-	priv->base.uevent->disable = nve0_fifo_uevent_disable;
-	priv->base.uevent->priv = priv;
+	/* PBDMA[n] */
+	for (i = 0; i < priv->spoon_nr; i++) {
+		nv_mask(priv, 0x04013c + (i * 0x2000), 0x10000100, 0x00000000);
+		nv_wr32(priv, 0x040108 + (i * 0x2000), 0xffffffff); /* INTR */
+		nv_wr32(priv, 0x04010c + (i * 0x2000), 0xfffffeff); /* INTREN */
+	}
 
-	nv_subdev(priv)->unit = 0x00000100;
-	nv_subdev(priv)->intr = nve0_fifo_intr;
-	nv_engine(priv)->cclass = &nve0_fifo_cclass;
-	nv_engine(priv)->sclass = nve0_fifo_sclass;
+	nv_wr32(priv, 0x002254, 0x10000000 | priv->user.bar.offset >> 12);
+
+	nv_wr32(priv, 0x002a00, 0xffffffff);
+	nv_wr32(priv, 0x002100, 0xffffffff);
+	nv_wr32(priv, 0x002140, 0x3fffffff);
 	return 0;
 }
 
-static void
+void
 nve0_fifo_dtor(struct nouveau_object *object)
 {
 	struct nve0_fifo_priv *priv = (void *)object;
@@ -631,50 +818,69 @@ nve0_fifo_dtor(struct nouveau_object *object)
 	nouveau_gpuobj_ref(NULL, &priv->user.mem);
 
 	for (i = 0; i < FIFO_ENGINE_NR; i++) {
-		nouveau_gpuobj_ref(NULL, &priv->engine[i].playlist[1]);
-		nouveau_gpuobj_ref(NULL, &priv->engine[i].playlist[0]);
+		nouveau_gpuobj_ref(NULL, &priv->engine[i].runlist[1]);
+		nouveau_gpuobj_ref(NULL, &priv->engine[i].runlist[0]);
 	}
 
 	nouveau_fifo_destroy(&priv->base);
 }
 
-static int
-nve0_fifo_init(struct nouveau_object *object)
+int
+nve0_fifo_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
+	       struct nouveau_oclass *oclass, void *data, u32 size,
+	       struct nouveau_object **pobject)
 {
-	struct nve0_fifo_priv *priv = (void *)object;
+	struct nve0_fifo_impl *impl = (void *)oclass;
+	struct nve0_fifo_priv *priv;
 	int ret, i;
 
-	ret = nouveau_fifo_init(&priv->base);
+	ret = nouveau_fifo_create(parent, engine, oclass, 0,
+				  impl->channels - 1, &priv);
+	*pobject = nv_object(priv);
 	if (ret)
 		return ret;
 
-	/* enable all available PSUBFIFOs */
-	nv_wr32(priv, 0x000204, 0xffffffff);
-	priv->spoon_nr = hweight32(nv_rd32(priv, 0x000204));
-	nv_debug(priv, "%d subfifo(s)\n", priv->spoon_nr);
+	for (i = 0; i < FIFO_ENGINE_NR; i++) {
+		ret = nouveau_gpuobj_new(nv_object(priv), NULL, 0x8000, 0x1000,
+					 0, &priv->engine[i].runlist[0]);
+		if (ret)
+			return ret;
 
-	/* PSUBFIFO[n] */
-	for (i = 0; i < priv->spoon_nr; i++) {
-		nv_mask(priv, 0x04013c + (i * 0x2000), 0x10000100, 0x00000000);
-		nv_wr32(priv, 0x040108 + (i * 0x2000), 0xffffffff); /* INTR */
-		nv_wr32(priv, 0x04010c + (i * 0x2000), 0xfffffeff); /* INTREN */
+		ret = nouveau_gpuobj_new(nv_object(priv), NULL, 0x8000, 0x1000,
+					 0, &priv->engine[i].runlist[1]);
+		if (ret)
+			return ret;
 	}
 
-	nv_wr32(priv, 0x002254, 0x10000000 | priv->user.bar.offset >> 12);
+	ret = nouveau_gpuobj_new(nv_object(priv), NULL, 4096 * 0x200, 0x1000,
+				 NVOBJ_FLAG_ZERO_ALLOC, &priv->user.mem);
+	if (ret)
+		return ret;
 
-	nv_wr32(priv, 0x002a00, 0xffffffff);
-	nv_wr32(priv, 0x002100, 0xffffffff);
-	nv_wr32(priv, 0x002140, 0x3fffffff);
+	ret = nouveau_gpuobj_map(priv->user.mem, NV_MEM_ACCESS_RW,
+				&priv->user.bar);
+	if (ret)
+		return ret;
+
+	priv->base.uevent->enable = nve0_fifo_uevent_enable;
+	priv->base.uevent->disable = nve0_fifo_uevent_disable;
+	priv->base.uevent->priv = priv;
+
+	nv_subdev(priv)->unit = 0x00000100;
+	nv_subdev(priv)->intr = nve0_fifo_intr;
+	nv_engine(priv)->cclass = &nve0_fifo_cclass;
+	nv_engine(priv)->sclass = nve0_fifo_sclass;
 	return 0;
 }
 
 struct nouveau_oclass *
-nve0_fifo_oclass = &(struct nouveau_oclass) {
-	.handle = NV_ENGINE(FIFO, 0xe0),
-	.ofuncs = &(struct nouveau_ofuncs) {
+nve0_fifo_oclass = &(struct nve0_fifo_impl) {
+	.base.handle = NV_ENGINE(FIFO, 0xe0),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nve0_fifo_ctor,
 		.dtor = nve0_fifo_dtor,
 		.init = nve0_fifo_init,
-		.fini = _nouveau_fifo_fini,
+		.fini = nve0_fifo_fini,
 	},
-};
+	.channels = 4096,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/fifo/nve0.h b/drivers/gpu/drm/nouveau/core/engine/fifo/nve0.h
new file mode 100644
index 000000000000..014344ebee66
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/fifo/nve0.h
@@ -0,0 +1,17 @@
+#ifndef __NVKM_FIFO_NVE0_H__
+#define __NVKM_FIFO_NVE0_H__
+
+#include <engine/fifo.h>
+
+int  nve0_fifo_ctor(struct nouveau_object *, struct nouveau_object *,
+		    struct nouveau_oclass *, void *, u32,
+		    struct nouveau_object **);
+void nve0_fifo_dtor(struct nouveau_object *);
+int  nve0_fifo_init(struct nouveau_object *);
+
+struct nve0_fifo_impl {
+	struct nouveau_oclass base;
+	u32 channels;
+};
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/ctxnv108.c b/drivers/gpu/drm/nouveau/core/engine/graph/ctxnv108.c
new file mode 100644
index 000000000000..a86bd3352bf8
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/ctxnv108.c
@@ -0,0 +1,1408 @@
+/*
+ * 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 <bskeggs@redhat.com>
+ */
+
+#include "nvc0.h"
+
+static struct nvc0_graph_init
+nv108_grctx_init_icmd[] = {
+	{ 0x001000,   1, 0x01, 0x00000004 },
+	{ 0x000039,   3, 0x01, 0x00000000 },
+	{ 0x0000a9,   1, 0x01, 0x0000ffff },
+	{ 0x000038,   1, 0x01, 0x0fac6881 },
+	{ 0x00003d,   1, 0x01, 0x00000001 },
+	{ 0x0000e8,   8, 0x01, 0x00000400 },
+	{ 0x000078,   8, 0x01, 0x00000300 },
+	{ 0x000050,   1, 0x01, 0x00000011 },
+	{ 0x000058,   8, 0x01, 0x00000008 },
+	{ 0x000208,   8, 0x01, 0x00000001 },
+	{ 0x000081,   1, 0x01, 0x00000001 },
+	{ 0x000085,   1, 0x01, 0x00000004 },
+	{ 0x000088,   1, 0x01, 0x00000400 },
+	{ 0x000090,   1, 0x01, 0x00000300 },
+	{ 0x000098,   1, 0x01, 0x00001001 },
+	{ 0x0000e3,   1, 0x01, 0x00000001 },
+	{ 0x0000da,   1, 0x01, 0x00000001 },
+	{ 0x0000f8,   1, 0x01, 0x00000003 },
+	{ 0x0000fa,   1, 0x01, 0x00000001 },
+	{ 0x00009f,   4, 0x01, 0x0000ffff },
+	{ 0x0000b1,   1, 0x01, 0x00000001 },
+	{ 0x0000ad,   1, 0x01, 0x0000013e },
+	{ 0x0000e1,   1, 0x01, 0x00000010 },
+	{ 0x000290,  16, 0x01, 0x00000000 },
+	{ 0x0003b0,  16, 0x01, 0x00000000 },
+	{ 0x0002a0,  16, 0x01, 0x00000000 },
+	{ 0x000420,  16, 0x01, 0x00000000 },
+	{ 0x0002b0,  16, 0x01, 0x00000000 },
+	{ 0x000430,  16, 0x01, 0x00000000 },
+	{ 0x0002c0,  16, 0x01, 0x00000000 },
+	{ 0x0004d0,  16, 0x01, 0x00000000 },
+	{ 0x000720,  16, 0x01, 0x00000000 },
+	{ 0x0008c0,  16, 0x01, 0x00000000 },
+	{ 0x000890,  16, 0x01, 0x00000000 },
+	{ 0x0008e0,  16, 0x01, 0x00000000 },
+	{ 0x0008a0,  16, 0x01, 0x00000000 },
+	{ 0x0008f0,  16, 0x01, 0x00000000 },
+	{ 0x00094c,   1, 0x01, 0x000000ff },
+	{ 0x00094d,   1, 0x01, 0xffffffff },
+	{ 0x00094e,   1, 0x01, 0x00000002 },
+	{ 0x0002ec,   1, 0x01, 0x00000001 },
+	{ 0x0002f2,   2, 0x01, 0x00000001 },
+	{ 0x0002f5,   1, 0x01, 0x00000001 },
+	{ 0x0002f7,   1, 0x01, 0x00000001 },
+	{ 0x000303,   1, 0x01, 0x00000001 },
+	{ 0x0002e6,   1, 0x01, 0x00000001 },
+	{ 0x000466,   1, 0x01, 0x00000052 },
+	{ 0x000301,   1, 0x01, 0x3f800000 },
+	{ 0x000304,   1, 0x01, 0x30201000 },
+	{ 0x000305,   1, 0x01, 0x70605040 },
+	{ 0x000306,   1, 0x01, 0xb8a89888 },
+	{ 0x000307,   1, 0x01, 0xf8e8d8c8 },
+	{ 0x00030a,   1, 0x01, 0x00ffff00 },
+	{ 0x00030b,   1, 0x01, 0x0000001a },
+	{ 0x00030c,   1, 0x01, 0x00000001 },
+	{ 0x000318,   1, 0x01, 0x00000001 },
+	{ 0x000340,   1, 0x01, 0x00000000 },
+	{ 0x000375,   1, 0x01, 0x00000001 },
+	{ 0x00037d,   1, 0x01, 0x00000006 },
+	{ 0x0003a0,   1, 0x01, 0x00000002 },
+	{ 0x0003aa,   1, 0x01, 0x00000001 },
+	{ 0x0003a9,   1, 0x01, 0x00000001 },
+	{ 0x000380,   1, 0x01, 0x00000001 },
+	{ 0x000383,   1, 0x01, 0x00000011 },
+	{ 0x000360,   1, 0x01, 0x00000040 },
+	{ 0x000366,   2, 0x01, 0x00000000 },
+	{ 0x000368,   1, 0x01, 0x00000fff },
+	{ 0x000370,   2, 0x01, 0x00000000 },
+	{ 0x000372,   1, 0x01, 0x000fffff },
+	{ 0x00037a,   1, 0x01, 0x00000012 },
+	{ 0x000619,   1, 0x01, 0x00000003 },
+	{ 0x000811,   1, 0x01, 0x00000003 },
+	{ 0x000812,   1, 0x01, 0x00000004 },
+	{ 0x000813,   1, 0x01, 0x00000006 },
+	{ 0x000814,   1, 0x01, 0x00000008 },
+	{ 0x000815,   1, 0x01, 0x0000000b },
+	{ 0x000800,   6, 0x01, 0x00000001 },
+	{ 0x000632,   1, 0x01, 0x00000001 },
+	{ 0x000633,   1, 0x01, 0x00000002 },
+	{ 0x000634,   1, 0x01, 0x00000003 },
+	{ 0x000635,   1, 0x01, 0x00000004 },
+	{ 0x000654,   1, 0x01, 0x3f800000 },
+	{ 0x000657,   1, 0x01, 0x3f800000 },
+	{ 0x000655,   2, 0x01, 0x3f800000 },
+	{ 0x0006cd,   1, 0x01, 0x3f800000 },
+	{ 0x0007f5,   1, 0x01, 0x3f800000 },
+	{ 0x0007dc,   1, 0x01, 0x39291909 },
+	{ 0x0007dd,   1, 0x01, 0x79695949 },
+	{ 0x0007de,   1, 0x01, 0xb9a99989 },
+	{ 0x0007df,   1, 0x01, 0xf9e9d9c9 },
+	{ 0x0007e8,   1, 0x01, 0x00003210 },
+	{ 0x0007e9,   1, 0x01, 0x00007654 },
+	{ 0x0007ea,   1, 0x01, 0x00000098 },
+	{ 0x0007ec,   1, 0x01, 0x39291909 },
+	{ 0x0007ed,   1, 0x01, 0x79695949 },
+	{ 0x0007ee,   1, 0x01, 0xb9a99989 },
+	{ 0x0007ef,   1, 0x01, 0xf9e9d9c9 },
+	{ 0x0007f0,   1, 0x01, 0x00003210 },
+	{ 0x0007f1,   1, 0x01, 0x00007654 },
+	{ 0x0007f2,   1, 0x01, 0x00000098 },
+	{ 0x0005a5,   1, 0x01, 0x00000001 },
+	{ 0x000980, 128, 0x01, 0x00000000 },
+	{ 0x000468,   1, 0x01, 0x00000004 },
+	{ 0x00046c,   1, 0x01, 0x00000001 },
+	{ 0x000470,  96, 0x01, 0x00000000 },
+	{ 0x000510,  16, 0x01, 0x3f800000 },
+	{ 0x000520,   1, 0x01, 0x000002b6 },
+	{ 0x000529,   1, 0x01, 0x00000001 },
+	{ 0x000530,  16, 0x01, 0xffff0000 },
+	{ 0x000585,   1, 0x01, 0x0000003f },
+	{ 0x000576,   1, 0x01, 0x00000003 },
+	{ 0x00057b,   1, 0x01, 0x00000059 },
+	{ 0x000586,   1, 0x01, 0x00000040 },
+	{ 0x000582,   2, 0x01, 0x00000080 },
+	{ 0x0005c2,   1, 0x01, 0x00000001 },
+	{ 0x000638,   2, 0x01, 0x00000001 },
+	{ 0x00063a,   1, 0x01, 0x00000002 },
+	{ 0x00063b,   2, 0x01, 0x00000001 },
+	{ 0x00063d,   1, 0x01, 0x00000002 },
+	{ 0x00063e,   1, 0x01, 0x00000001 },
+	{ 0x0008b8,   8, 0x01, 0x00000001 },
+	{ 0x000900,   8, 0x01, 0x00000001 },
+	{ 0x000908,   8, 0x01, 0x00000002 },
+	{ 0x000910,  16, 0x01, 0x00000001 },
+	{ 0x000920,   8, 0x01, 0x00000002 },
+	{ 0x000928,   8, 0x01, 0x00000001 },
+	{ 0x000662,   1, 0x01, 0x00000001 },
+	{ 0x000648,   9, 0x01, 0x00000001 },
+	{ 0x000658,   1, 0x01, 0x0000000f },
+	{ 0x0007ff,   1, 0x01, 0x0000000a },
+	{ 0x00066a,   1, 0x01, 0x40000000 },
+	{ 0x00066b,   1, 0x01, 0x10000000 },
+	{ 0x00066c,   2, 0x01, 0xffff0000 },
+	{ 0x0007af,   2, 0x01, 0x00000008 },
+	{ 0x0007f6,   1, 0x01, 0x00000001 },
+	{ 0x00080b,   1, 0x01, 0x00000002 },
+	{ 0x0006b2,   1, 0x01, 0x00000055 },
+	{ 0x0007ad,   1, 0x01, 0x00000003 },
+	{ 0x000937,   1, 0x01, 0x00000001 },
+	{ 0x000971,   1, 0x01, 0x00000008 },
+	{ 0x000972,   1, 0x01, 0x00000040 },
+	{ 0x000973,   1, 0x01, 0x0000012c },
+	{ 0x00097c,   1, 0x01, 0x00000040 },
+	{ 0x000979,   1, 0x01, 0x00000003 },
+	{ 0x000975,   1, 0x01, 0x00000020 },
+	{ 0x000976,   1, 0x01, 0x00000001 },
+	{ 0x000977,   1, 0x01, 0x00000020 },
+	{ 0x000978,   1, 0x01, 0x00000001 },
+	{ 0x000957,   1, 0x01, 0x00000003 },
+	{ 0x00095e,   1, 0x01, 0x20164010 },
+	{ 0x00095f,   1, 0x01, 0x00000020 },
+	{ 0x000a0d,   1, 0x01, 0x00000006 },
+	{ 0x00097d,   1, 0x01, 0x00000020 },
+	{ 0x000683,   1, 0x01, 0x00000006 },
+	{ 0x000685,   1, 0x01, 0x003fffff },
+	{ 0x000687,   1, 0x01, 0x003fffff },
+	{ 0x0006a0,   1, 0x01, 0x00000005 },
+	{ 0x000840,   1, 0x01, 0x00400008 },
+	{ 0x000841,   1, 0x01, 0x08000080 },
+	{ 0x000842,   1, 0x01, 0x00400008 },
+	{ 0x000843,   1, 0x01, 0x08000080 },
+	{ 0x0006aa,   1, 0x01, 0x00000001 },
+	{ 0x0006ab,   1, 0x01, 0x00000002 },
+	{ 0x0006ac,   1, 0x01, 0x00000080 },
+	{ 0x0006ad,   2, 0x01, 0x00000100 },
+	{ 0x0006b1,   1, 0x01, 0x00000011 },
+	{ 0x0006bb,   1, 0x01, 0x000000cf },
+	{ 0x0006ce,   1, 0x01, 0x2a712488 },
+	{ 0x000739,   1, 0x01, 0x4085c000 },
+	{ 0x00073a,   1, 0x01, 0x00000080 },
+	{ 0x000786,   1, 0x01, 0x80000100 },
+	{ 0x00073c,   1, 0x01, 0x00010100 },
+	{ 0x00073d,   1, 0x01, 0x02800000 },
+	{ 0x000787,   1, 0x01, 0x000000cf },
+	{ 0x00078c,   1, 0x01, 0x00000008 },
+	{ 0x000792,   1, 0x01, 0x00000001 },
+	{ 0x000794,   3, 0x01, 0x00000001 },
+	{ 0x000797,   1, 0x01, 0x000000cf },
+	{ 0x000836,   1, 0x01, 0x00000001 },
+	{ 0x00079a,   1, 0x01, 0x00000002 },
+	{ 0x000833,   1, 0x01, 0x04444480 },
+	{ 0x0007a1,   1, 0x01, 0x00000001 },
+	{ 0x0007a3,   3, 0x01, 0x00000001 },
+	{ 0x000831,   1, 0x01, 0x00000004 },
+	{ 0x000b07,   1, 0x01, 0x00000002 },
+	{ 0x000b08,   2, 0x01, 0x00000100 },
+	{ 0x000b0a,   1, 0x01, 0x00000001 },
+	{ 0x000a04,   1, 0x01, 0x000000ff },
+	{ 0x000a0b,   1, 0x01, 0x00000040 },
+	{ 0x00097f,   1, 0x01, 0x00000100 },
+	{ 0x000a02,   1, 0x01, 0x00000001 },
+	{ 0x000809,   1, 0x01, 0x00000007 },
+	{ 0x00c221,   1, 0x01, 0x00000040 },
+	{ 0x00c1b0,   8, 0x01, 0x0000000f },
+	{ 0x00c1b8,   1, 0x01, 0x0fac6881 },
+	{ 0x00c1b9,   1, 0x01, 0x00fac688 },
+	{ 0x00c401,   1, 0x01, 0x00000001 },
+	{ 0x00c402,   1, 0x01, 0x00010001 },
+	{ 0x00c403,   2, 0x01, 0x00000001 },
+	{ 0x00c40e,   1, 0x01, 0x00000020 },
+	{ 0x00c500,   1, 0x01, 0x00000003 },
+	{ 0x01e100,   1, 0x01, 0x00000001 },
+	{ 0x001000,   1, 0x01, 0x00000002 },
+	{ 0x0006aa,   1, 0x01, 0x00000001 },
+	{ 0x0006ad,   2, 0x01, 0x00000100 },
+	{ 0x0006b1,   1, 0x01, 0x00000011 },
+	{ 0x00078c,   1, 0x01, 0x00000008 },
+	{ 0x000792,   1, 0x01, 0x00000001 },
+	{ 0x000794,   3, 0x01, 0x00000001 },
+	{ 0x000797,   1, 0x01, 0x000000cf },
+	{ 0x00079a,   1, 0x01, 0x00000002 },
+	{ 0x0007a1,   1, 0x01, 0x00000001 },
+	{ 0x0007a3,   3, 0x01, 0x00000001 },
+	{ 0x000831,   1, 0x01, 0x00000004 },
+	{ 0x01e100,   1, 0x01, 0x00000001 },
+	{ 0x001000,   1, 0x01, 0x00000008 },
+	{ 0x000039,   3, 0x01, 0x00000000 },
+	{ 0x000380,   1, 0x01, 0x00000001 },
+	{ 0x000366,   2, 0x01, 0x00000000 },
+	{ 0x000368,   1, 0x01, 0x00000fff },
+	{ 0x000370,   2, 0x01, 0x00000000 },
+	{ 0x000372,   1, 0x01, 0x000fffff },
+	{ 0x000813,   1, 0x01, 0x00000006 },
+	{ 0x000814,   1, 0x01, 0x00000008 },
+	{ 0x000957,   1, 0x01, 0x00000003 },
+	{ 0x000b07,   1, 0x01, 0x00000002 },
+	{ 0x000b08,   2, 0x01, 0x00000100 },
+	{ 0x000b0a,   1, 0x01, 0x00000001 },
+	{ 0x000a04,   1, 0x01, 0x000000ff },
+	{ 0x000a0b,   1, 0x01, 0x00000040 },
+	{ 0x00097f,   1, 0x01, 0x00000100 },
+	{ 0x000a02,   1, 0x01, 0x00000001 },
+	{ 0x000809,   1, 0x01, 0x00000007 },
+	{ 0x00c221,   1, 0x01, 0x00000040 },
+	{ 0x00c401,   1, 0x01, 0x00000001 },
+	{ 0x00c402,   1, 0x01, 0x00010001 },
+	{ 0x00c403,   2, 0x01, 0x00000001 },
+	{ 0x00c40e,   1, 0x01, 0x00000020 },
+	{ 0x00c500,   1, 0x01, 0x00000003 },
+	{ 0x01e100,   1, 0x01, 0x00000001 },
+	{ 0x001000,   1, 0x01, 0x00000001 },
+	{ 0x000b07,   1, 0x01, 0x00000002 },
+	{ 0x000b08,   2, 0x01, 0x00000100 },
+	{ 0x000b0a,   1, 0x01, 0x00000001 },
+	{ 0x01e100,   1, 0x01, 0x00000001 },
+	{}
+};
+
+static struct nvc0_graph_init
+nv108_grctx_init_a197[] = {
+	{ 0x000800,   1, 0x04, 0x00000000 },
+	{ 0x000840,   1, 0x04, 0x00000000 },
+	{ 0x000880,   1, 0x04, 0x00000000 },
+	{ 0x0008c0,   1, 0x04, 0x00000000 },
+	{ 0x000900,   1, 0x04, 0x00000000 },
+	{ 0x000940,   1, 0x04, 0x00000000 },
+	{ 0x000980,   1, 0x04, 0x00000000 },
+	{ 0x0009c0,   1, 0x04, 0x00000000 },
+	{ 0x000804,   1, 0x04, 0x00000000 },
+	{ 0x000844,   1, 0x04, 0x00000000 },
+	{ 0x000884,   1, 0x04, 0x00000000 },
+	{ 0x0008c4,   1, 0x04, 0x00000000 },
+	{ 0x000904,   1, 0x04, 0x00000000 },
+	{ 0x000944,   1, 0x04, 0x00000000 },
+	{ 0x000984,   1, 0x04, 0x00000000 },
+	{ 0x0009c4,   1, 0x04, 0x00000000 },
+	{ 0x000808,   1, 0x04, 0x00000400 },
+	{ 0x000848,   1, 0x04, 0x00000400 },
+	{ 0x000888,   1, 0x04, 0x00000400 },
+	{ 0x0008c8,   1, 0x04, 0x00000400 },
+	{ 0x000908,   1, 0x04, 0x00000400 },
+	{ 0x000948,   1, 0x04, 0x00000400 },
+	{ 0x000988,   1, 0x04, 0x00000400 },
+	{ 0x0009c8,   1, 0x04, 0x00000400 },
+	{ 0x00080c,   1, 0x04, 0x00000300 },
+	{ 0x00084c,   1, 0x04, 0x00000300 },
+	{ 0x00088c,   1, 0x04, 0x00000300 },
+	{ 0x0008cc,   1, 0x04, 0x00000300 },
+	{ 0x00090c,   1, 0x04, 0x00000300 },
+	{ 0x00094c,   1, 0x04, 0x00000300 },
+	{ 0x00098c,   1, 0x04, 0x00000300 },
+	{ 0x0009cc,   1, 0x04, 0x00000300 },
+	{ 0x000810,   1, 0x04, 0x000000cf },
+	{ 0x000850,   1, 0x04, 0x00000000 },
+	{ 0x000890,   1, 0x04, 0x00000000 },
+	{ 0x0008d0,   1, 0x04, 0x00000000 },
+	{ 0x000910,   1, 0x04, 0x00000000 },
+	{ 0x000950,   1, 0x04, 0x00000000 },
+	{ 0x000990,   1, 0x04, 0x00000000 },
+	{ 0x0009d0,   1, 0x04, 0x00000000 },
+	{ 0x000814,   1, 0x04, 0x00000040 },
+	{ 0x000854,   1, 0x04, 0x00000040 },
+	{ 0x000894,   1, 0x04, 0x00000040 },
+	{ 0x0008d4,   1, 0x04, 0x00000040 },
+	{ 0x000914,   1, 0x04, 0x00000040 },
+	{ 0x000954,   1, 0x04, 0x00000040 },
+	{ 0x000994,   1, 0x04, 0x00000040 },
+	{ 0x0009d4,   1, 0x04, 0x00000040 },
+	{ 0x000818,   1, 0x04, 0x00000001 },
+	{ 0x000858,   1, 0x04, 0x00000001 },
+	{ 0x000898,   1, 0x04, 0x00000001 },
+	{ 0x0008d8,   1, 0x04, 0x00000001 },
+	{ 0x000918,   1, 0x04, 0x00000001 },
+	{ 0x000958,   1, 0x04, 0x00000001 },
+	{ 0x000998,   1, 0x04, 0x00000001 },
+	{ 0x0009d8,   1, 0x04, 0x00000001 },
+	{ 0x00081c,   1, 0x04, 0x00000000 },
+	{ 0x00085c,   1, 0x04, 0x00000000 },
+	{ 0x00089c,   1, 0x04, 0x00000000 },
+	{ 0x0008dc,   1, 0x04, 0x00000000 },
+	{ 0x00091c,   1, 0x04, 0x00000000 },
+	{ 0x00095c,   1, 0x04, 0x00000000 },
+	{ 0x00099c,   1, 0x04, 0x00000000 },
+	{ 0x0009dc,   1, 0x04, 0x00000000 },
+	{ 0x000820,   1, 0x04, 0x00000000 },
+	{ 0x000860,   1, 0x04, 0x00000000 },
+	{ 0x0008a0,   1, 0x04, 0x00000000 },
+	{ 0x0008e0,   1, 0x04, 0x00000000 },
+	{ 0x000920,   1, 0x04, 0x00000000 },
+	{ 0x000960,   1, 0x04, 0x00000000 },
+	{ 0x0009a0,   1, 0x04, 0x00000000 },
+	{ 0x0009e0,   1, 0x04, 0x00000000 },
+	{ 0x001c00,   1, 0x04, 0x00000000 },
+	{ 0x001c10,   1, 0x04, 0x00000000 },
+	{ 0x001c20,   1, 0x04, 0x00000000 },
+	{ 0x001c30,   1, 0x04, 0x00000000 },
+	{ 0x001c40,   1, 0x04, 0x00000000 },
+	{ 0x001c50,   1, 0x04, 0x00000000 },
+	{ 0x001c60,   1, 0x04, 0x00000000 },
+	{ 0x001c70,   1, 0x04, 0x00000000 },
+	{ 0x001c80,   1, 0x04, 0x00000000 },
+	{ 0x001c90,   1, 0x04, 0x00000000 },
+	{ 0x001ca0,   1, 0x04, 0x00000000 },
+	{ 0x001cb0,   1, 0x04, 0x00000000 },
+	{ 0x001cc0,   1, 0x04, 0x00000000 },
+	{ 0x001cd0,   1, 0x04, 0x00000000 },
+	{ 0x001ce0,   1, 0x04, 0x00000000 },
+	{ 0x001cf0,   1, 0x04, 0x00000000 },
+	{ 0x001c04,   1, 0x04, 0x00000000 },
+	{ 0x001c14,   1, 0x04, 0x00000000 },
+	{ 0x001c24,   1, 0x04, 0x00000000 },
+	{ 0x001c34,   1, 0x04, 0x00000000 },
+	{ 0x001c44,   1, 0x04, 0x00000000 },
+	{ 0x001c54,   1, 0x04, 0x00000000 },
+	{ 0x001c64,   1, 0x04, 0x00000000 },
+	{ 0x001c74,   1, 0x04, 0x00000000 },
+	{ 0x001c84,   1, 0x04, 0x00000000 },
+	{ 0x001c94,   1, 0x04, 0x00000000 },
+	{ 0x001ca4,   1, 0x04, 0x00000000 },
+	{ 0x001cb4,   1, 0x04, 0x00000000 },
+	{ 0x001cc4,   1, 0x04, 0x00000000 },
+	{ 0x001cd4,   1, 0x04, 0x00000000 },
+	{ 0x001ce4,   1, 0x04, 0x00000000 },
+	{ 0x001cf4,   1, 0x04, 0x00000000 },
+	{ 0x001c08,   1, 0x04, 0x00000000 },
+	{ 0x001c18,   1, 0x04, 0x00000000 },
+	{ 0x001c28,   1, 0x04, 0x00000000 },
+	{ 0x001c38,   1, 0x04, 0x00000000 },
+	{ 0x001c48,   1, 0x04, 0x00000000 },
+	{ 0x001c58,   1, 0x04, 0x00000000 },
+	{ 0x001c68,   1, 0x04, 0x00000000 },
+	{ 0x001c78,   1, 0x04, 0x00000000 },
+	{ 0x001c88,   1, 0x04, 0x00000000 },
+	{ 0x001c98,   1, 0x04, 0x00000000 },
+	{ 0x001ca8,   1, 0x04, 0x00000000 },
+	{ 0x001cb8,   1, 0x04, 0x00000000 },
+	{ 0x001cc8,   1, 0x04, 0x00000000 },
+	{ 0x001cd8,   1, 0x04, 0x00000000 },
+	{ 0x001ce8,   1, 0x04, 0x00000000 },
+	{ 0x001cf8,   1, 0x04, 0x00000000 },
+	{ 0x001c0c,   1, 0x04, 0x00000000 },
+	{ 0x001c1c,   1, 0x04, 0x00000000 },
+	{ 0x001c2c,   1, 0x04, 0x00000000 },
+	{ 0x001c3c,   1, 0x04, 0x00000000 },
+	{ 0x001c4c,   1, 0x04, 0x00000000 },
+	{ 0x001c5c,   1, 0x04, 0x00000000 },
+	{ 0x001c6c,   1, 0x04, 0x00000000 },
+	{ 0x001c7c,   1, 0x04, 0x00000000 },
+	{ 0x001c8c,   1, 0x04, 0x00000000 },
+	{ 0x001c9c,   1, 0x04, 0x00000000 },
+	{ 0x001cac,   1, 0x04, 0x00000000 },
+	{ 0x001cbc,   1, 0x04, 0x00000000 },
+	{ 0x001ccc,   1, 0x04, 0x00000000 },
+	{ 0x001cdc,   1, 0x04, 0x00000000 },
+	{ 0x001cec,   1, 0x04, 0x00000000 },
+	{ 0x001cfc,   2, 0x04, 0x00000000 },
+	{ 0x001d10,   1, 0x04, 0x00000000 },
+	{ 0x001d20,   1, 0x04, 0x00000000 },
+	{ 0x001d30,   1, 0x04, 0x00000000 },
+	{ 0x001d40,   1, 0x04, 0x00000000 },
+	{ 0x001d50,   1, 0x04, 0x00000000 },
+	{ 0x001d60,   1, 0x04, 0x00000000 },
+	{ 0x001d70,   1, 0x04, 0x00000000 },
+	{ 0x001d80,   1, 0x04, 0x00000000 },
+	{ 0x001d90,   1, 0x04, 0x00000000 },
+	{ 0x001da0,   1, 0x04, 0x00000000 },
+	{ 0x001db0,   1, 0x04, 0x00000000 },
+	{ 0x001dc0,   1, 0x04, 0x00000000 },
+	{ 0x001dd0,   1, 0x04, 0x00000000 },
+	{ 0x001de0,   1, 0x04, 0x00000000 },
+	{ 0x001df0,   1, 0x04, 0x00000000 },
+	{ 0x001d04,   1, 0x04, 0x00000000 },
+	{ 0x001d14,   1, 0x04, 0x00000000 },
+	{ 0x001d24,   1, 0x04, 0x00000000 },
+	{ 0x001d34,   1, 0x04, 0x00000000 },
+	{ 0x001d44,   1, 0x04, 0x00000000 },
+	{ 0x001d54,   1, 0x04, 0x00000000 },
+	{ 0x001d64,   1, 0x04, 0x00000000 },
+	{ 0x001d74,   1, 0x04, 0x00000000 },
+	{ 0x001d84,   1, 0x04, 0x00000000 },
+	{ 0x001d94,   1, 0x04, 0x00000000 },
+	{ 0x001da4,   1, 0x04, 0x00000000 },
+	{ 0x001db4,   1, 0x04, 0x00000000 },
+	{ 0x001dc4,   1, 0x04, 0x00000000 },
+	{ 0x001dd4,   1, 0x04, 0x00000000 },
+	{ 0x001de4,   1, 0x04, 0x00000000 },
+	{ 0x001df4,   1, 0x04, 0x00000000 },
+	{ 0x001d08,   1, 0x04, 0x00000000 },
+	{ 0x001d18,   1, 0x04, 0x00000000 },
+	{ 0x001d28,   1, 0x04, 0x00000000 },
+	{ 0x001d38,   1, 0x04, 0x00000000 },
+	{ 0x001d48,   1, 0x04, 0x00000000 },
+	{ 0x001d58,   1, 0x04, 0x00000000 },
+	{ 0x001d68,   1, 0x04, 0x00000000 },
+	{ 0x001d78,   1, 0x04, 0x00000000 },
+	{ 0x001d88,   1, 0x04, 0x00000000 },
+	{ 0x001d98,   1, 0x04, 0x00000000 },
+	{ 0x001da8,   1, 0x04, 0x00000000 },
+	{ 0x001db8,   1, 0x04, 0x00000000 },
+	{ 0x001dc8,   1, 0x04, 0x00000000 },
+	{ 0x001dd8,   1, 0x04, 0x00000000 },
+	{ 0x001de8,   1, 0x04, 0x00000000 },
+	{ 0x001df8,   1, 0x04, 0x00000000 },
+	{ 0x001d0c,   1, 0x04, 0x00000000 },
+	{ 0x001d1c,   1, 0x04, 0x00000000 },
+	{ 0x001d2c,   1, 0x04, 0x00000000 },
+	{ 0x001d3c,   1, 0x04, 0x00000000 },
+	{ 0x001d4c,   1, 0x04, 0x00000000 },
+	{ 0x001d5c,   1, 0x04, 0x00000000 },
+	{ 0x001d6c,   1, 0x04, 0x00000000 },
+	{ 0x001d7c,   1, 0x04, 0x00000000 },
+	{ 0x001d8c,   1, 0x04, 0x00000000 },
+	{ 0x001d9c,   1, 0x04, 0x00000000 },
+	{ 0x001dac,   1, 0x04, 0x00000000 },
+	{ 0x001dbc,   1, 0x04, 0x00000000 },
+	{ 0x001dcc,   1, 0x04, 0x00000000 },
+	{ 0x001ddc,   1, 0x04, 0x00000000 },
+	{ 0x001dec,   1, 0x04, 0x00000000 },
+	{ 0x001dfc,   1, 0x04, 0x00000000 },
+	{ 0x001f00,   1, 0x04, 0x00000000 },
+	{ 0x001f08,   1, 0x04, 0x00000000 },
+	{ 0x001f10,   1, 0x04, 0x00000000 },
+	{ 0x001f18,   1, 0x04, 0x00000000 },
+	{ 0x001f20,   1, 0x04, 0x00000000 },
+	{ 0x001f28,   1, 0x04, 0x00000000 },
+	{ 0x001f30,   1, 0x04, 0x00000000 },
+	{ 0x001f38,   1, 0x04, 0x00000000 },
+	{ 0x001f40,   1, 0x04, 0x00000000 },
+	{ 0x001f48,   1, 0x04, 0x00000000 },
+	{ 0x001f50,   1, 0x04, 0x00000000 },
+	{ 0x001f58,   1, 0x04, 0x00000000 },
+	{ 0x001f60,   1, 0x04, 0x00000000 },
+	{ 0x001f68,   1, 0x04, 0x00000000 },
+	{ 0x001f70,   1, 0x04, 0x00000000 },
+	{ 0x001f78,   1, 0x04, 0x00000000 },
+	{ 0x001f04,   1, 0x04, 0x00000000 },
+	{ 0x001f0c,   1, 0x04, 0x00000000 },
+	{ 0x001f14,   1, 0x04, 0x00000000 },
+	{ 0x001f1c,   1, 0x04, 0x00000000 },
+	{ 0x001f24,   1, 0x04, 0x00000000 },
+	{ 0x001f2c,   1, 0x04, 0x00000000 },
+	{ 0x001f34,   1, 0x04, 0x00000000 },
+	{ 0x001f3c,   1, 0x04, 0x00000000 },
+	{ 0x001f44,   1, 0x04, 0x00000000 },
+	{ 0x001f4c,   1, 0x04, 0x00000000 },
+	{ 0x001f54,   1, 0x04, 0x00000000 },
+	{ 0x001f5c,   1, 0x04, 0x00000000 },
+	{ 0x001f64,   1, 0x04, 0x00000000 },
+	{ 0x001f6c,   1, 0x04, 0x00000000 },
+	{ 0x001f74,   1, 0x04, 0x00000000 },
+	{ 0x001f7c,   2, 0x04, 0x00000000 },
+	{ 0x001f88,   1, 0x04, 0x00000000 },
+	{ 0x001f90,   1, 0x04, 0x00000000 },
+	{ 0x001f98,   1, 0x04, 0x00000000 },
+	{ 0x001fa0,   1, 0x04, 0x00000000 },
+	{ 0x001fa8,   1, 0x04, 0x00000000 },
+	{ 0x001fb0,   1, 0x04, 0x00000000 },
+	{ 0x001fb8,   1, 0x04, 0x00000000 },
+	{ 0x001fc0,   1, 0x04, 0x00000000 },
+	{ 0x001fc8,   1, 0x04, 0x00000000 },
+	{ 0x001fd0,   1, 0x04, 0x00000000 },
+	{ 0x001fd8,   1, 0x04, 0x00000000 },
+	{ 0x001fe0,   1, 0x04, 0x00000000 },
+	{ 0x001fe8,   1, 0x04, 0x00000000 },
+	{ 0x001ff0,   1, 0x04, 0x00000000 },
+	{ 0x001ff8,   1, 0x04, 0x00000000 },
+	{ 0x001f84,   1, 0x04, 0x00000000 },
+	{ 0x001f8c,   1, 0x04, 0x00000000 },
+	{ 0x001f94,   1, 0x04, 0x00000000 },
+	{ 0x001f9c,   1, 0x04, 0x00000000 },
+	{ 0x001fa4,   1, 0x04, 0x00000000 },
+	{ 0x001fac,   1, 0x04, 0x00000000 },
+	{ 0x001fb4,   1, 0x04, 0x00000000 },
+	{ 0x001fbc,   1, 0x04, 0x00000000 },
+	{ 0x001fc4,   1, 0x04, 0x00000000 },
+	{ 0x001fcc,   1, 0x04, 0x00000000 },
+	{ 0x001fd4,   1, 0x04, 0x00000000 },
+	{ 0x001fdc,   1, 0x04, 0x00000000 },
+	{ 0x001fe4,   1, 0x04, 0x00000000 },
+	{ 0x001fec,   1, 0x04, 0x00000000 },
+	{ 0x001ff4,   1, 0x04, 0x00000000 },
+	{ 0x001ffc,   2, 0x04, 0x00000000 },
+	{ 0x002040,   1, 0x04, 0x00000011 },
+	{ 0x002080,   1, 0x04, 0x00000020 },
+	{ 0x0020c0,   1, 0x04, 0x00000030 },
+	{ 0x002100,   1, 0x04, 0x00000040 },
+	{ 0x002140,   1, 0x04, 0x00000051 },
+	{ 0x00200c,   1, 0x04, 0x00000001 },
+	{ 0x00204c,   1, 0x04, 0x00000001 },
+	{ 0x00208c,   1, 0x04, 0x00000001 },
+	{ 0x0020cc,   1, 0x04, 0x00000001 },
+	{ 0x00210c,   1, 0x04, 0x00000001 },
+	{ 0x00214c,   1, 0x04, 0x00000001 },
+	{ 0x002010,   1, 0x04, 0x00000000 },
+	{ 0x002050,   1, 0x04, 0x00000000 },
+	{ 0x002090,   1, 0x04, 0x00000001 },
+	{ 0x0020d0,   1, 0x04, 0x00000002 },
+	{ 0x002110,   1, 0x04, 0x00000003 },
+	{ 0x002150,   1, 0x04, 0x00000004 },
+	{ 0x000380,   1, 0x04, 0x00000000 },
+	{ 0x0003a0,   1, 0x04, 0x00000000 },
+	{ 0x0003c0,   1, 0x04, 0x00000000 },
+	{ 0x0003e0,   1, 0x04, 0x00000000 },
+	{ 0x000384,   1, 0x04, 0x00000000 },
+	{ 0x0003a4,   1, 0x04, 0x00000000 },
+	{ 0x0003c4,   1, 0x04, 0x00000000 },
+	{ 0x0003e4,   1, 0x04, 0x00000000 },
+	{ 0x000388,   1, 0x04, 0x00000000 },
+	{ 0x0003a8,   1, 0x04, 0x00000000 },
+	{ 0x0003c8,   1, 0x04, 0x00000000 },
+	{ 0x0003e8,   1, 0x04, 0x00000000 },
+	{ 0x00038c,   1, 0x04, 0x00000000 },
+	{ 0x0003ac,   1, 0x04, 0x00000000 },
+	{ 0x0003cc,   1, 0x04, 0x00000000 },
+	{ 0x0003ec,   1, 0x04, 0x00000000 },
+	{ 0x000700,   1, 0x04, 0x00000000 },
+	{ 0x000710,   1, 0x04, 0x00000000 },
+	{ 0x000720,   1, 0x04, 0x00000000 },
+	{ 0x000730,   1, 0x04, 0x00000000 },
+	{ 0x000704,   1, 0x04, 0x00000000 },
+	{ 0x000714,   1, 0x04, 0x00000000 },
+	{ 0x000724,   1, 0x04, 0x00000000 },
+	{ 0x000734,   1, 0x04, 0x00000000 },
+	{ 0x000708,   1, 0x04, 0x00000000 },
+	{ 0x000718,   1, 0x04, 0x00000000 },
+	{ 0x000728,   1, 0x04, 0x00000000 },
+	{ 0x000738,   1, 0x04, 0x00000000 },
+	{ 0x002800, 128, 0x04, 0x00000000 },
+	{ 0x000a00,   1, 0x04, 0x00000000 },
+	{ 0x000a20,   1, 0x04, 0x00000000 },
+	{ 0x000a40,   1, 0x04, 0x00000000 },
+	{ 0x000a60,   1, 0x04, 0x00000000 },
+	{ 0x000a80,   1, 0x04, 0x00000000 },
+	{ 0x000aa0,   1, 0x04, 0x00000000 },
+	{ 0x000ac0,   1, 0x04, 0x00000000 },
+	{ 0x000ae0,   1, 0x04, 0x00000000 },
+	{ 0x000b00,   1, 0x04, 0x00000000 },
+	{ 0x000b20,   1, 0x04, 0x00000000 },
+	{ 0x000b40,   1, 0x04, 0x00000000 },
+	{ 0x000b60,   1, 0x04, 0x00000000 },
+	{ 0x000b80,   1, 0x04, 0x00000000 },
+	{ 0x000ba0,   1, 0x04, 0x00000000 },
+	{ 0x000bc0,   1, 0x04, 0x00000000 },
+	{ 0x000be0,   1, 0x04, 0x00000000 },
+	{ 0x000a04,   1, 0x04, 0x00000000 },
+	{ 0x000a24,   1, 0x04, 0x00000000 },
+	{ 0x000a44,   1, 0x04, 0x00000000 },
+	{ 0x000a64,   1, 0x04, 0x00000000 },
+	{ 0x000a84,   1, 0x04, 0x00000000 },
+	{ 0x000aa4,   1, 0x04, 0x00000000 },
+	{ 0x000ac4,   1, 0x04, 0x00000000 },
+	{ 0x000ae4,   1, 0x04, 0x00000000 },
+	{ 0x000b04,   1, 0x04, 0x00000000 },
+	{ 0x000b24,   1, 0x04, 0x00000000 },
+	{ 0x000b44,   1, 0x04, 0x00000000 },
+	{ 0x000b64,   1, 0x04, 0x00000000 },
+	{ 0x000b84,   1, 0x04, 0x00000000 },
+	{ 0x000ba4,   1, 0x04, 0x00000000 },
+	{ 0x000bc4,   1, 0x04, 0x00000000 },
+	{ 0x000be4,   1, 0x04, 0x00000000 },
+	{ 0x000a08,   1, 0x04, 0x00000000 },
+	{ 0x000a28,   1, 0x04, 0x00000000 },
+	{ 0x000a48,   1, 0x04, 0x00000000 },
+	{ 0x000a68,   1, 0x04, 0x00000000 },
+	{ 0x000a88,   1, 0x04, 0x00000000 },
+	{ 0x000aa8,   1, 0x04, 0x00000000 },
+	{ 0x000ac8,   1, 0x04, 0x00000000 },
+	{ 0x000ae8,   1, 0x04, 0x00000000 },
+	{ 0x000b08,   1, 0x04, 0x00000000 },
+	{ 0x000b28,   1, 0x04, 0x00000000 },
+	{ 0x000b48,   1, 0x04, 0x00000000 },
+	{ 0x000b68,   1, 0x04, 0x00000000 },
+	{ 0x000b88,   1, 0x04, 0x00000000 },
+	{ 0x000ba8,   1, 0x04, 0x00000000 },
+	{ 0x000bc8,   1, 0x04, 0x00000000 },
+	{ 0x000be8,   1, 0x04, 0x00000000 },
+	{ 0x000a0c,   1, 0x04, 0x00000000 },
+	{ 0x000a2c,   1, 0x04, 0x00000000 },
+	{ 0x000a4c,   1, 0x04, 0x00000000 },
+	{ 0x000a6c,   1, 0x04, 0x00000000 },
+	{ 0x000a8c,   1, 0x04, 0x00000000 },
+	{ 0x000aac,   1, 0x04, 0x00000000 },
+	{ 0x000acc,   1, 0x04, 0x00000000 },
+	{ 0x000aec,   1, 0x04, 0x00000000 },
+	{ 0x000b0c,   1, 0x04, 0x00000000 },
+	{ 0x000b2c,   1, 0x04, 0x00000000 },
+	{ 0x000b4c,   1, 0x04, 0x00000000 },
+	{ 0x000b6c,   1, 0x04, 0x00000000 },
+	{ 0x000b8c,   1, 0x04, 0x00000000 },
+	{ 0x000bac,   1, 0x04, 0x00000000 },
+	{ 0x000bcc,   1, 0x04, 0x00000000 },
+	{ 0x000bec,   1, 0x04, 0x00000000 },
+	{ 0x000a10,   1, 0x04, 0x00000000 },
+	{ 0x000a30,   1, 0x04, 0x00000000 },
+	{ 0x000a50,   1, 0x04, 0x00000000 },
+	{ 0x000a70,   1, 0x04, 0x00000000 },
+	{ 0x000a90,   1, 0x04, 0x00000000 },
+	{ 0x000ab0,   1, 0x04, 0x00000000 },
+	{ 0x000ad0,   1, 0x04, 0x00000000 },
+	{ 0x000af0,   1, 0x04, 0x00000000 },
+	{ 0x000b10,   1, 0x04, 0x00000000 },
+	{ 0x000b30,   1, 0x04, 0x00000000 },
+	{ 0x000b50,   1, 0x04, 0x00000000 },
+	{ 0x000b70,   1, 0x04, 0x00000000 },
+	{ 0x000b90,   1, 0x04, 0x00000000 },
+	{ 0x000bb0,   1, 0x04, 0x00000000 },
+	{ 0x000bd0,   1, 0x04, 0x00000000 },
+	{ 0x000bf0,   1, 0x04, 0x00000000 },
+	{ 0x000a14,   1, 0x04, 0x00000000 },
+	{ 0x000a34,   1, 0x04, 0x00000000 },
+	{ 0x000a54,   1, 0x04, 0x00000000 },
+	{ 0x000a74,   1, 0x04, 0x00000000 },
+	{ 0x000a94,   1, 0x04, 0x00000000 },
+	{ 0x000ab4,   1, 0x04, 0x00000000 },
+	{ 0x000ad4,   1, 0x04, 0x00000000 },
+	{ 0x000af4,   1, 0x04, 0x00000000 },
+	{ 0x000b14,   1, 0x04, 0x00000000 },
+	{ 0x000b34,   1, 0x04, 0x00000000 },
+	{ 0x000b54,   1, 0x04, 0x00000000 },
+	{ 0x000b74,   1, 0x04, 0x00000000 },
+	{ 0x000b94,   1, 0x04, 0x00000000 },
+	{ 0x000bb4,   1, 0x04, 0x00000000 },
+	{ 0x000bd4,   1, 0x04, 0x00000000 },
+	{ 0x000bf4,   1, 0x04, 0x00000000 },
+	{ 0x000c00,   1, 0x04, 0x00000000 },
+	{ 0x000c10,   1, 0x04, 0x00000000 },
+	{ 0x000c20,   1, 0x04, 0x00000000 },
+	{ 0x000c30,   1, 0x04, 0x00000000 },
+	{ 0x000c40,   1, 0x04, 0x00000000 },
+	{ 0x000c50,   1, 0x04, 0x00000000 },
+	{ 0x000c60,   1, 0x04, 0x00000000 },
+	{ 0x000c70,   1, 0x04, 0x00000000 },
+	{ 0x000c80,   1, 0x04, 0x00000000 },
+	{ 0x000c90,   1, 0x04, 0x00000000 },
+	{ 0x000ca0,   1, 0x04, 0x00000000 },
+	{ 0x000cb0,   1, 0x04, 0x00000000 },
+	{ 0x000cc0,   1, 0x04, 0x00000000 },
+	{ 0x000cd0,   1, 0x04, 0x00000000 },
+	{ 0x000ce0,   1, 0x04, 0x00000000 },
+	{ 0x000cf0,   1, 0x04, 0x00000000 },
+	{ 0x000c04,   1, 0x04, 0x00000000 },
+	{ 0x000c14,   1, 0x04, 0x00000000 },
+	{ 0x000c24,   1, 0x04, 0x00000000 },
+	{ 0x000c34,   1, 0x04, 0x00000000 },
+	{ 0x000c44,   1, 0x04, 0x00000000 },
+	{ 0x000c54,   1, 0x04, 0x00000000 },
+	{ 0x000c64,   1, 0x04, 0x00000000 },
+	{ 0x000c74,   1, 0x04, 0x00000000 },
+	{ 0x000c84,   1, 0x04, 0x00000000 },
+	{ 0x000c94,   1, 0x04, 0x00000000 },
+	{ 0x000ca4,   1, 0x04, 0x00000000 },
+	{ 0x000cb4,   1, 0x04, 0x00000000 },
+	{ 0x000cc4,   1, 0x04, 0x00000000 },
+	{ 0x000cd4,   1, 0x04, 0x00000000 },
+	{ 0x000ce4,   1, 0x04, 0x00000000 },
+	{ 0x000cf4,   1, 0x04, 0x00000000 },
+	{ 0x000c08,   1, 0x04, 0x00000000 },
+	{ 0x000c18,   1, 0x04, 0x00000000 },
+	{ 0x000c28,   1, 0x04, 0x00000000 },
+	{ 0x000c38,   1, 0x04, 0x00000000 },
+	{ 0x000c48,   1, 0x04, 0x00000000 },
+	{ 0x000c58,   1, 0x04, 0x00000000 },
+	{ 0x000c68,   1, 0x04, 0x00000000 },
+	{ 0x000c78,   1, 0x04, 0x00000000 },
+	{ 0x000c88,   1, 0x04, 0x00000000 },
+	{ 0x000c98,   1, 0x04, 0x00000000 },
+	{ 0x000ca8,   1, 0x04, 0x00000000 },
+	{ 0x000cb8,   1, 0x04, 0x00000000 },
+	{ 0x000cc8,   1, 0x04, 0x00000000 },
+	{ 0x000cd8,   1, 0x04, 0x00000000 },
+	{ 0x000ce8,   1, 0x04, 0x00000000 },
+	{ 0x000cf8,   1, 0x04, 0x00000000 },
+	{ 0x000c0c,   1, 0x04, 0x3f800000 },
+	{ 0x000c1c,   1, 0x04, 0x3f800000 },
+	{ 0x000c2c,   1, 0x04, 0x3f800000 },
+	{ 0x000c3c,   1, 0x04, 0x3f800000 },
+	{ 0x000c4c,   1, 0x04, 0x3f800000 },
+	{ 0x000c5c,   1, 0x04, 0x3f800000 },
+	{ 0x000c6c,   1, 0x04, 0x3f800000 },
+	{ 0x000c7c,   1, 0x04, 0x3f800000 },
+	{ 0x000c8c,   1, 0x04, 0x3f800000 },
+	{ 0x000c9c,   1, 0x04, 0x3f800000 },
+	{ 0x000cac,   1, 0x04, 0x3f800000 },
+	{ 0x000cbc,   1, 0x04, 0x3f800000 },
+	{ 0x000ccc,   1, 0x04, 0x3f800000 },
+	{ 0x000cdc,   1, 0x04, 0x3f800000 },
+	{ 0x000cec,   1, 0x04, 0x3f800000 },
+	{ 0x000cfc,   1, 0x04, 0x3f800000 },
+	{ 0x000d00,   1, 0x04, 0xffff0000 },
+	{ 0x000d08,   1, 0x04, 0xffff0000 },
+	{ 0x000d10,   1, 0x04, 0xffff0000 },
+	{ 0x000d18,   1, 0x04, 0xffff0000 },
+	{ 0x000d20,   1, 0x04, 0xffff0000 },
+	{ 0x000d28,   1, 0x04, 0xffff0000 },
+	{ 0x000d30,   1, 0x04, 0xffff0000 },
+	{ 0x000d38,   1, 0x04, 0xffff0000 },
+	{ 0x000d04,   1, 0x04, 0xffff0000 },
+	{ 0x000d0c,   1, 0x04, 0xffff0000 },
+	{ 0x000d14,   1, 0x04, 0xffff0000 },
+	{ 0x000d1c,   1, 0x04, 0xffff0000 },
+	{ 0x000d24,   1, 0x04, 0xffff0000 },
+	{ 0x000d2c,   1, 0x04, 0xffff0000 },
+	{ 0x000d34,   1, 0x04, 0xffff0000 },
+	{ 0x000d3c,   1, 0x04, 0xffff0000 },
+	{ 0x000e00,   1, 0x04, 0x00000000 },
+	{ 0x000e10,   1, 0x04, 0x00000000 },
+	{ 0x000e20,   1, 0x04, 0x00000000 },
+	{ 0x000e30,   1, 0x04, 0x00000000 },
+	{ 0x000e40,   1, 0x04, 0x00000000 },
+	{ 0x000e50,   1, 0x04, 0x00000000 },
+	{ 0x000e60,   1, 0x04, 0x00000000 },
+	{ 0x000e70,   1, 0x04, 0x00000000 },
+	{ 0x000e80,   1, 0x04, 0x00000000 },
+	{ 0x000e90,   1, 0x04, 0x00000000 },
+	{ 0x000ea0,   1, 0x04, 0x00000000 },
+	{ 0x000eb0,   1, 0x04, 0x00000000 },
+	{ 0x000ec0,   1, 0x04, 0x00000000 },
+	{ 0x000ed0,   1, 0x04, 0x00000000 },
+	{ 0x000ee0,   1, 0x04, 0x00000000 },
+	{ 0x000ef0,   1, 0x04, 0x00000000 },
+	{ 0x000e04,   1, 0x04, 0xffff0000 },
+	{ 0x000e14,   1, 0x04, 0xffff0000 },
+	{ 0x000e24,   1, 0x04, 0xffff0000 },
+	{ 0x000e34,   1, 0x04, 0xffff0000 },
+	{ 0x000e44,   1, 0x04, 0xffff0000 },
+	{ 0x000e54,   1, 0x04, 0xffff0000 },
+	{ 0x000e64,   1, 0x04, 0xffff0000 },
+	{ 0x000e74,   1, 0x04, 0xffff0000 },
+	{ 0x000e84,   1, 0x04, 0xffff0000 },
+	{ 0x000e94,   1, 0x04, 0xffff0000 },
+	{ 0x000ea4,   1, 0x04, 0xffff0000 },
+	{ 0x000eb4,   1, 0x04, 0xffff0000 },
+	{ 0x000ec4,   1, 0x04, 0xffff0000 },
+	{ 0x000ed4,   1, 0x04, 0xffff0000 },
+	{ 0x000ee4,   1, 0x04, 0xffff0000 },
+	{ 0x000ef4,   1, 0x04, 0xffff0000 },
+	{ 0x000e08,   1, 0x04, 0xffff0000 },
+	{ 0x000e18,   1, 0x04, 0xffff0000 },
+	{ 0x000e28,   1, 0x04, 0xffff0000 },
+	{ 0x000e38,   1, 0x04, 0xffff0000 },
+	{ 0x000e48,   1, 0x04, 0xffff0000 },
+	{ 0x000e58,   1, 0x04, 0xffff0000 },
+	{ 0x000e68,   1, 0x04, 0xffff0000 },
+	{ 0x000e78,   1, 0x04, 0xffff0000 },
+	{ 0x000e88,   1, 0x04, 0xffff0000 },
+	{ 0x000e98,   1, 0x04, 0xffff0000 },
+	{ 0x000ea8,   1, 0x04, 0xffff0000 },
+	{ 0x000eb8,   1, 0x04, 0xffff0000 },
+	{ 0x000ec8,   1, 0x04, 0xffff0000 },
+	{ 0x000ed8,   1, 0x04, 0xffff0000 },
+	{ 0x000ee8,   1, 0x04, 0xffff0000 },
+	{ 0x000ef8,   1, 0x04, 0xffff0000 },
+	{ 0x000d40,   1, 0x04, 0x00000000 },
+	{ 0x000d48,   1, 0x04, 0x00000000 },
+	{ 0x000d50,   1, 0x04, 0x00000000 },
+	{ 0x000d58,   1, 0x04, 0x00000000 },
+	{ 0x000d44,   1, 0x04, 0x00000000 },
+	{ 0x000d4c,   1, 0x04, 0x00000000 },
+	{ 0x000d54,   1, 0x04, 0x00000000 },
+	{ 0x000d5c,   1, 0x04, 0x00000000 },
+	{ 0x001e00,   1, 0x04, 0x00000001 },
+	{ 0x001e20,   1, 0x04, 0x00000001 },
+	{ 0x001e40,   1, 0x04, 0x00000001 },
+	{ 0x001e60,   1, 0x04, 0x00000001 },
+	{ 0x001e80,   1, 0x04, 0x00000001 },
+	{ 0x001ea0,   1, 0x04, 0x00000001 },
+	{ 0x001ec0,   1, 0x04, 0x00000001 },
+	{ 0x001ee0,   1, 0x04, 0x00000001 },
+	{ 0x001e04,   1, 0x04, 0x00000001 },
+	{ 0x001e24,   1, 0x04, 0x00000001 },
+	{ 0x001e44,   1, 0x04, 0x00000001 },
+	{ 0x001e64,   1, 0x04, 0x00000001 },
+	{ 0x001e84,   1, 0x04, 0x00000001 },
+	{ 0x001ea4,   1, 0x04, 0x00000001 },
+	{ 0x001ec4,   1, 0x04, 0x00000001 },
+	{ 0x001ee4,   1, 0x04, 0x00000001 },
+	{ 0x001e08,   1, 0x04, 0x00000002 },
+	{ 0x001e28,   1, 0x04, 0x00000002 },
+	{ 0x001e48,   1, 0x04, 0x00000002 },
+	{ 0x001e68,   1, 0x04, 0x00000002 },
+	{ 0x001e88,   1, 0x04, 0x00000002 },
+	{ 0x001ea8,   1, 0x04, 0x00000002 },
+	{ 0x001ec8,   1, 0x04, 0x00000002 },
+	{ 0x001ee8,   1, 0x04, 0x00000002 },
+	{ 0x001e0c,   1, 0x04, 0x00000001 },
+	{ 0x001e2c,   1, 0x04, 0x00000001 },
+	{ 0x001e4c,   1, 0x04, 0x00000001 },
+	{ 0x001e6c,   1, 0x04, 0x00000001 },
+	{ 0x001e8c,   1, 0x04, 0x00000001 },
+	{ 0x001eac,   1, 0x04, 0x00000001 },
+	{ 0x001ecc,   1, 0x04, 0x00000001 },
+	{ 0x001eec,   1, 0x04, 0x00000001 },
+	{ 0x001e10,   1, 0x04, 0x00000001 },
+	{ 0x001e30,   1, 0x04, 0x00000001 },
+	{ 0x001e50,   1, 0x04, 0x00000001 },
+	{ 0x001e70,   1, 0x04, 0x00000001 },
+	{ 0x001e90,   1, 0x04, 0x00000001 },
+	{ 0x001eb0,   1, 0x04, 0x00000001 },
+	{ 0x001ed0,   1, 0x04, 0x00000001 },
+	{ 0x001ef0,   1, 0x04, 0x00000001 },
+	{ 0x001e14,   1, 0x04, 0x00000002 },
+	{ 0x001e34,   1, 0x04, 0x00000002 },
+	{ 0x001e54,   1, 0x04, 0x00000002 },
+	{ 0x001e74,   1, 0x04, 0x00000002 },
+	{ 0x001e94,   1, 0x04, 0x00000002 },
+	{ 0x001eb4,   1, 0x04, 0x00000002 },
+	{ 0x001ed4,   1, 0x04, 0x00000002 },
+	{ 0x001ef4,   1, 0x04, 0x00000002 },
+	{ 0x001e18,   1, 0x04, 0x00000001 },
+	{ 0x001e38,   1, 0x04, 0x00000001 },
+	{ 0x001e58,   1, 0x04, 0x00000001 },
+	{ 0x001e78,   1, 0x04, 0x00000001 },
+	{ 0x001e98,   1, 0x04, 0x00000001 },
+	{ 0x001eb8,   1, 0x04, 0x00000001 },
+	{ 0x001ed8,   1, 0x04, 0x00000001 },
+	{ 0x001ef8,   1, 0x04, 0x00000001 },
+	{ 0x003400, 128, 0x04, 0x00000000 },
+	{ 0x00030c,   1, 0x04, 0x00000001 },
+	{ 0x001944,   1, 0x04, 0x00000000 },
+	{ 0x001514,   1, 0x04, 0x00000000 },
+	{ 0x000d68,   1, 0x04, 0x0000ffff },
+	{ 0x00121c,   1, 0x04, 0x0fac6881 },
+	{ 0x000fac,   1, 0x04, 0x00000001 },
+	{ 0x001538,   1, 0x04, 0x00000001 },
+	{ 0x000fe0,   2, 0x04, 0x00000000 },
+	{ 0x000fe8,   1, 0x04, 0x00000014 },
+	{ 0x000fec,   1, 0x04, 0x00000040 },
+	{ 0x000ff0,   1, 0x04, 0x00000000 },
+	{ 0x00179c,   1, 0x04, 0x00000000 },
+	{ 0x001228,   1, 0x04, 0x00000400 },
+	{ 0x00122c,   1, 0x04, 0x00000300 },
+	{ 0x001230,   1, 0x04, 0x00010001 },
+	{ 0x0007f8,   1, 0x04, 0x00000000 },
+	{ 0x0015b4,   1, 0x04, 0x00000001 },
+	{ 0x0015cc,   1, 0x04, 0x00000000 },
+	{ 0x001534,   1, 0x04, 0x00000000 },
+	{ 0x000fb0,   1, 0x04, 0x00000000 },
+	{ 0x0015d0,   1, 0x04, 0x00000000 },
+	{ 0x00153c,   1, 0x04, 0x00000000 },
+	{ 0x0016b4,   1, 0x04, 0x00000003 },
+	{ 0x000fbc,   4, 0x04, 0x0000ffff },
+	{ 0x000df8,   2, 0x04, 0x00000000 },
+	{ 0x001948,   1, 0x04, 0x00000000 },
+	{ 0x001970,   1, 0x04, 0x00000001 },
+	{ 0x00161c,   1, 0x04, 0x000009f0 },
+	{ 0x000dcc,   1, 0x04, 0x00000010 },
+	{ 0x00163c,   1, 0x04, 0x00000000 },
+	{ 0x0015e4,   1, 0x04, 0x00000000 },
+	{ 0x001160,  32, 0x04, 0x25e00040 },
+	{ 0x001880,  32, 0x04, 0x00000000 },
+	{ 0x000f84,   2, 0x04, 0x00000000 },
+	{ 0x0017c8,   2, 0x04, 0x00000000 },
+	{ 0x0017d0,   1, 0x04, 0x000000ff },
+	{ 0x0017d4,   1, 0x04, 0xffffffff },
+	{ 0x0017d8,   1, 0x04, 0x00000002 },
+	{ 0x0017dc,   1, 0x04, 0x00000000 },
+	{ 0x0015f4,   2, 0x04, 0x00000000 },
+	{ 0x001434,   2, 0x04, 0x00000000 },
+	{ 0x000d74,   1, 0x04, 0x00000000 },
+	{ 0x000dec,   1, 0x04, 0x00000001 },
+	{ 0x0013a4,   1, 0x04, 0x00000000 },
+	{ 0x001318,   1, 0x04, 0x00000001 },
+	{ 0x001644,   1, 0x04, 0x00000000 },
+	{ 0x000748,   1, 0x04, 0x00000000 },
+	{ 0x000de8,   1, 0x04, 0x00000000 },
+	{ 0x001648,   1, 0x04, 0x00000000 },
+	{ 0x0012a4,   1, 0x04, 0x00000000 },
+	{ 0x001120,   4, 0x04, 0x00000000 },
+	{ 0x001118,   1, 0x04, 0x00000000 },
+	{ 0x00164c,   1, 0x04, 0x00000000 },
+	{ 0x001658,   1, 0x04, 0x00000000 },
+	{ 0x001910,   1, 0x04, 0x00000290 },
+	{ 0x001518,   1, 0x04, 0x00000000 },
+	{ 0x00165c,   1, 0x04, 0x00000001 },
+	{ 0x001520,   1, 0x04, 0x00000000 },
+	{ 0x001604,   1, 0x04, 0x00000000 },
+	{ 0x001570,   1, 0x04, 0x00000000 },
+	{ 0x0013b0,   2, 0x04, 0x3f800000 },
+	{ 0x00020c,   1, 0x04, 0x00000000 },
+	{ 0x001670,   1, 0x04, 0x30201000 },
+	{ 0x001674,   1, 0x04, 0x70605040 },
+	{ 0x001678,   1, 0x04, 0xb8a89888 },
+	{ 0x00167c,   1, 0x04, 0xf8e8d8c8 },
+	{ 0x00166c,   1, 0x04, 0x00000000 },
+	{ 0x001680,   1, 0x04, 0x00ffff00 },
+	{ 0x0012d0,   1, 0x04, 0x00000003 },
+	{ 0x0012d4,   1, 0x04, 0x00000002 },
+	{ 0x001684,   2, 0x04, 0x00000000 },
+	{ 0x000dac,   2, 0x04, 0x00001b02 },
+	{ 0x000db4,   1, 0x04, 0x00000000 },
+	{ 0x00168c,   1, 0x04, 0x00000000 },
+	{ 0x0015bc,   1, 0x04, 0x00000000 },
+	{ 0x00156c,   1, 0x04, 0x00000000 },
+	{ 0x00187c,   1, 0x04, 0x00000000 },
+	{ 0x001110,   1, 0x04, 0x00000001 },
+	{ 0x000dc0,   3, 0x04, 0x00000000 },
+	{ 0x001234,   1, 0x04, 0x00000000 },
+	{ 0x001690,   1, 0x04, 0x00000000 },
+	{ 0x0012ac,   1, 0x04, 0x00000001 },
+	{ 0x0002c4,   1, 0x04, 0x00000000 },
+	{ 0x000790,   5, 0x04, 0x00000000 },
+	{ 0x00077c,   1, 0x04, 0x00000000 },
+	{ 0x001000,   1, 0x04, 0x00000010 },
+	{ 0x0010fc,   1, 0x04, 0x00000000 },
+	{ 0x001290,   1, 0x04, 0x00000000 },
+	{ 0x000218,   1, 0x04, 0x00000010 },
+	{ 0x0012d8,   1, 0x04, 0x00000000 },
+	{ 0x0012dc,   1, 0x04, 0x00000010 },
+	{ 0x000d94,   1, 0x04, 0x00000001 },
+	{ 0x00155c,   2, 0x04, 0x00000000 },
+	{ 0x001564,   1, 0x04, 0x00000fff },
+	{ 0x001574,   2, 0x04, 0x00000000 },
+	{ 0x00157c,   1, 0x04, 0x000fffff },
+	{ 0x001354,   1, 0x04, 0x00000000 },
+	{ 0x001610,   1, 0x04, 0x00000012 },
+	{ 0x001608,   2, 0x04, 0x00000000 },
+	{ 0x00260c,   1, 0x04, 0x00000000 },
+	{ 0x0007ac,   1, 0x04, 0x00000000 },
+	{ 0x00162c,   1, 0x04, 0x00000003 },
+	{ 0x000210,   1, 0x04, 0x00000000 },
+	{ 0x000320,   1, 0x04, 0x00000000 },
+	{ 0x000324,   6, 0x04, 0x3f800000 },
+	{ 0x000750,   1, 0x04, 0x00000000 },
+	{ 0x000760,   1, 0x04, 0x39291909 },
+	{ 0x000764,   1, 0x04, 0x79695949 },
+	{ 0x000768,   1, 0x04, 0xb9a99989 },
+	{ 0x00076c,   1, 0x04, 0xf9e9d9c9 },
+	{ 0x000770,   1, 0x04, 0x30201000 },
+	{ 0x000774,   1, 0x04, 0x70605040 },
+	{ 0x000778,   1, 0x04, 0x00009080 },
+	{ 0x000780,   1, 0x04, 0x39291909 },
+	{ 0x000784,   1, 0x04, 0x79695949 },
+	{ 0x000788,   1, 0x04, 0xb9a99989 },
+	{ 0x00078c,   1, 0x04, 0xf9e9d9c9 },
+	{ 0x0007d0,   1, 0x04, 0x30201000 },
+	{ 0x0007d4,   1, 0x04, 0x70605040 },
+	{ 0x0007d8,   1, 0x04, 0x00009080 },
+	{ 0x00037c,   1, 0x04, 0x00000001 },
+	{ 0x000740,   2, 0x04, 0x00000000 },
+	{ 0x002600,   1, 0x04, 0x00000000 },
+	{ 0x001918,   1, 0x04, 0x00000000 },
+	{ 0x00191c,   1, 0x04, 0x00000900 },
+	{ 0x001920,   1, 0x04, 0x00000405 },
+	{ 0x001308,   1, 0x04, 0x00000001 },
+	{ 0x001924,   1, 0x04, 0x00000000 },
+	{ 0x0013ac,   1, 0x04, 0x00000000 },
+	{ 0x00192c,   1, 0x04, 0x00000001 },
+	{ 0x00193c,   1, 0x04, 0x00002c1c },
+	{ 0x000d7c,   1, 0x04, 0x00000000 },
+	{ 0x000f8c,   1, 0x04, 0x00000000 },
+	{ 0x0002c0,   1, 0x04, 0x00000001 },
+	{ 0x001510,   1, 0x04, 0x00000000 },
+	{ 0x001940,   1, 0x04, 0x00000000 },
+	{ 0x000ff4,   2, 0x04, 0x00000000 },
+	{ 0x00194c,   2, 0x04, 0x00000000 },
+	{ 0x001968,   1, 0x04, 0x00000000 },
+	{ 0x001590,   1, 0x04, 0x0000003f },
+	{ 0x0007e8,   4, 0x04, 0x00000000 },
+	{ 0x00196c,   1, 0x04, 0x00000011 },
+	{ 0x0002e4,   1, 0x04, 0x0000b001 },
+	{ 0x00036c,   2, 0x04, 0x00000000 },
+	{ 0x00197c,   1, 0x04, 0x00000000 },
+	{ 0x000fcc,   2, 0x04, 0x00000000 },
+	{ 0x0002d8,   1, 0x04, 0x00000040 },
+	{ 0x001980,   1, 0x04, 0x00000080 },
+	{ 0x001504,   1, 0x04, 0x00000080 },
+	{ 0x001984,   1, 0x04, 0x00000000 },
+	{ 0x000300,   1, 0x04, 0x00000001 },
+	{ 0x0013a8,   1, 0x04, 0x00000000 },
+	{ 0x0012ec,   1, 0x04, 0x00000000 },
+	{ 0x001310,   1, 0x04, 0x00000000 },
+	{ 0x001314,   1, 0x04, 0x00000001 },
+	{ 0x001380,   1, 0x04, 0x00000000 },
+	{ 0x001384,   4, 0x04, 0x00000001 },
+	{ 0x001394,   1, 0x04, 0x00000000 },
+	{ 0x00139c,   1, 0x04, 0x00000000 },
+	{ 0x001398,   1, 0x04, 0x00000000 },
+	{ 0x001594,   1, 0x04, 0x00000000 },
+	{ 0x001598,   4, 0x04, 0x00000001 },
+	{ 0x000f54,   3, 0x04, 0x00000000 },
+	{ 0x0019bc,   1, 0x04, 0x00000000 },
+	{ 0x000f9c,   2, 0x04, 0x00000000 },
+	{ 0x0012cc,   1, 0x04, 0x00000000 },
+	{ 0x0012e8,   1, 0x04, 0x00000000 },
+	{ 0x00130c,   1, 0x04, 0x00000001 },
+	{ 0x001360,   8, 0x04, 0x00000000 },
+	{ 0x00133c,   2, 0x04, 0x00000001 },
+	{ 0x001344,   1, 0x04, 0x00000002 },
+	{ 0x001348,   2, 0x04, 0x00000001 },
+	{ 0x001350,   1, 0x04, 0x00000002 },
+	{ 0x001358,   1, 0x04, 0x00000001 },
+	{ 0x0012e4,   1, 0x04, 0x00000000 },
+	{ 0x00131c,   4, 0x04, 0x00000000 },
+	{ 0x0019c0,   1, 0x04, 0x00000000 },
+	{ 0x001140,   1, 0x04, 0x00000000 },
+	{ 0x0019c4,   1, 0x04, 0x00000000 },
+	{ 0x0019c8,   1, 0x04, 0x00001500 },
+	{ 0x00135c,   1, 0x04, 0x00000000 },
+	{ 0x000f90,   1, 0x04, 0x00000000 },
+	{ 0x0019e0,   8, 0x04, 0x00000001 },
+	{ 0x0019cc,   1, 0x04, 0x00000001 },
+	{ 0x0015b8,   1, 0x04, 0x00000000 },
+	{ 0x001a00,   1, 0x04, 0x00001111 },
+	{ 0x001a04,   7, 0x04, 0x00000000 },
+	{ 0x000d6c,   2, 0x04, 0xffff0000 },
+	{ 0x0010f8,   1, 0x04, 0x00001010 },
+	{ 0x000d80,   5, 0x04, 0x00000000 },
+	{ 0x000da0,   1, 0x04, 0x00000000 },
+	{ 0x0007a4,   2, 0x04, 0x00000000 },
+	{ 0x001508,   1, 0x04, 0x80000000 },
+	{ 0x00150c,   1, 0x04, 0x40000000 },
+	{ 0x001668,   1, 0x04, 0x00000000 },
+	{ 0x000318,   2, 0x04, 0x00000008 },
+	{ 0x000d9c,   1, 0x04, 0x00000001 },
+	{ 0x000ddc,   1, 0x04, 0x00000002 },
+	{ 0x000374,   1, 0x04, 0x00000000 },
+	{ 0x000378,   1, 0x04, 0x00000020 },
+	{ 0x0007dc,   1, 0x04, 0x00000000 },
+	{ 0x00074c,   1, 0x04, 0x00000055 },
+	{ 0x001420,   1, 0x04, 0x00000003 },
+	{ 0x0017bc,   2, 0x04, 0x00000000 },
+	{ 0x0017c4,   1, 0x04, 0x00000001 },
+	{ 0x001008,   1, 0x04, 0x00000008 },
+	{ 0x00100c,   1, 0x04, 0x00000040 },
+	{ 0x001010,   1, 0x04, 0x0000012c },
+	{ 0x000d60,   1, 0x04, 0x00000040 },
+	{ 0x00075c,   1, 0x04, 0x00000003 },
+	{ 0x001018,   1, 0x04, 0x00000020 },
+	{ 0x00101c,   1, 0x04, 0x00000001 },
+	{ 0x001020,   1, 0x04, 0x00000020 },
+	{ 0x001024,   1, 0x04, 0x00000001 },
+	{ 0x001444,   3, 0x04, 0x00000000 },
+	{ 0x000360,   1, 0x04, 0x20164010 },
+	{ 0x000364,   1, 0x04, 0x00000020 },
+	{ 0x000368,   1, 0x04, 0x00000000 },
+	{ 0x000de4,   1, 0x04, 0x00000000 },
+	{ 0x000204,   1, 0x04, 0x00000006 },
+	{ 0x000208,   1, 0x04, 0x00000000 },
+	{ 0x0002cc,   2, 0x04, 0x003fffff },
+	{ 0x001220,   1, 0x04, 0x00000005 },
+	{ 0x000fdc,   1, 0x04, 0x00000000 },
+	{ 0x000f98,   1, 0x04, 0x00400008 },
+	{ 0x001284,   1, 0x04, 0x08000080 },
+	{ 0x001450,   1, 0x04, 0x00400008 },
+	{ 0x001454,   1, 0x04, 0x08000080 },
+	{ 0x000214,   1, 0x04, 0x00000000 },
+	{}
+};
+
+static struct nvc0_graph_init
+nv108_grctx_init_unk40xx[] = {
+	{ 0x404004,   8, 0x04, 0x00000000 },
+	{ 0x404024,   1, 0x04, 0x0000e000 },
+	{ 0x404028,   8, 0x04, 0x00000000 },
+	{ 0x4040a8,   8, 0x04, 0x00000000 },
+	{ 0x4040c8,   1, 0x04, 0xf800008f },
+	{ 0x4040d0,   6, 0x04, 0x00000000 },
+	{ 0x4040e8,   1, 0x04, 0x00001000 },
+	{ 0x4040f8,   1, 0x04, 0x00000000 },
+	{ 0x404100,  10, 0x04, 0x00000000 },
+	{ 0x404130,   2, 0x04, 0x00000000 },
+	{ 0x404138,   1, 0x04, 0x20000040 },
+	{ 0x404150,   1, 0x04, 0x0000002e },
+	{ 0x404154,   1, 0x04, 0x00000400 },
+	{ 0x404158,   1, 0x04, 0x00000200 },
+	{ 0x404164,   1, 0x04, 0x00000055 },
+	{ 0x40417c,   2, 0x04, 0x00000000 },
+	{ 0x404194,   1, 0x04, 0x01000700 },
+	{ 0x4041a0,   4, 0x04, 0x00000000 },
+	{ 0x404200,   1, 0x04, 0x0000a197 },
+	{ 0x404204,   1, 0x04, 0x0000a1c0 },
+	{ 0x404208,   1, 0x04, 0x0000a140 },
+	{ 0x40420c,   1, 0x04, 0x0000902d },
+	{}
+};
+
+static struct nvc0_graph_init
+nv108_grctx_init_unk58xx[] = {
+	{ 0x405800,   1, 0x04, 0x0f8000bf },
+	{ 0x405830,   1, 0x04, 0x02180648 },
+	{ 0x405834,   1, 0x04, 0x08000000 },
+	{ 0x405838,   1, 0x04, 0x00000000 },
+	{ 0x405854,   1, 0x04, 0x00000000 },
+	{ 0x405870,   4, 0x04, 0x00000001 },
+	{ 0x405a00,   2, 0x04, 0x00000000 },
+	{ 0x405a18,   1, 0x04, 0x00000000 },
+	{ 0x405a1c,   1, 0x04, 0x000000ff },
+	{}
+};
+
+static struct nvc0_graph_init
+nv108_grctx_init_unk64xx[] = {
+	{ 0x4064a8,   1, 0x04, 0x00000000 },
+	{ 0x4064ac,   1, 0x04, 0x00003fff },
+	{ 0x4064b0,   3, 0x04, 0x00000000 },
+	{ 0x4064c0,   1, 0x04, 0x802000f0 },
+	{ 0x4064c4,   1, 0x04, 0x0192ffff },
+	{ 0x4064c8,   1, 0x04, 0x00c20200 },
+	{ 0x4064cc,   9, 0x04, 0x00000000 },
+	{ 0x4064fc,   1, 0x04, 0x0000022a },
+	{}
+};
+
+static struct nvc0_graph_init
+nv108_grctx_init_unk78xx[] = {
+	{ 0x407804,   1, 0x04, 0x00000063 },
+	{ 0x40780c,   1, 0x04, 0x0a418820 },
+	{ 0x407810,   1, 0x04, 0x062080e6 },
+	{ 0x407814,   1, 0x04, 0x020398a4 },
+	{ 0x407818,   1, 0x04, 0x0e629062 },
+	{ 0x40781c,   1, 0x04, 0x0a418820 },
+	{ 0x407820,   1, 0x04, 0x000000e6 },
+	{ 0x4078bc,   1, 0x04, 0x00000103 },
+	{}
+};
+
+static struct nvc0_graph_init
+nv108_grctx_init_unk88xx[] = {
+	{ 0x408800,   1, 0x04, 0x32802a3c },
+	{ 0x408804,   1, 0x04, 0x00000040 },
+	{ 0x408808,   1, 0x04, 0x1003e005 },
+	{ 0x408840,   1, 0x04, 0x0000000b },
+	{ 0x408900,   1, 0x04, 0xb080b801 },
+	{ 0x408904,   1, 0x04, 0x62000001 },
+	{ 0x408908,   1, 0x04, 0x02c8102f },
+	{ 0x408980,   1, 0x04, 0x0000011d },
+	{}
+};
+
+static struct nvc0_graph_init
+nv108_grctx_init_gpc_0[] = {
+	{ 0x418380,   1, 0x04, 0x00000016 },
+	{ 0x418400,   1, 0x04, 0x38005e00 },
+	{ 0x418404,   1, 0x04, 0x71e0ffff },
+	{ 0x41840c,   1, 0x04, 0x00001008 },
+	{ 0x418410,   1, 0x04, 0x0fff0fff },
+	{ 0x418414,   1, 0x04, 0x02200fff },
+	{ 0x418450,   6, 0x04, 0x00000000 },
+	{ 0x418468,   1, 0x04, 0x00000001 },
+	{ 0x41846c,   2, 0x04, 0x00000000 },
+	{ 0x418600,   1, 0x04, 0x0000007f },
+	{ 0x418684,   1, 0x04, 0x0000001f },
+	{ 0x418700,   1, 0x04, 0x00000002 },
+	{ 0x418704,   2, 0x04, 0x00000080 },
+	{ 0x41870c,   2, 0x04, 0x00000000 },
+	{ 0x418800,   1, 0x04, 0x7006863a },
+	{ 0x418808,   1, 0x04, 0x00000000 },
+	{ 0x41880c,   1, 0x04, 0x00000030 },
+	{ 0x418810,   1, 0x04, 0x00000000 },
+	{ 0x418828,   1, 0x04, 0x00000044 },
+	{ 0x418830,   1, 0x04, 0x10000001 },
+	{ 0x4188d8,   1, 0x04, 0x00000008 },
+	{ 0x4188e0,   1, 0x04, 0x01000000 },
+	{ 0x4188e8,   5, 0x04, 0x00000000 },
+	{ 0x4188fc,   1, 0x04, 0x20100058 },
+	{ 0x41891c,   1, 0x04, 0x00ff00ff },
+	{ 0x418924,   1, 0x04, 0x00000000 },
+	{ 0x418928,   1, 0x04, 0x00ffff00 },
+	{ 0x41892c,   1, 0x04, 0x0000ff00 },
+	{ 0x418b00,   1, 0x04, 0x0000001e },
+	{ 0x418b08,   1, 0x04, 0x0a418820 },
+	{ 0x418b0c,   1, 0x04, 0x062080e6 },
+	{ 0x418b10,   1, 0x04, 0x020398a4 },
+	{ 0x418b14,   1, 0x04, 0x0e629062 },
+	{ 0x418b18,   1, 0x04, 0x0a418820 },
+	{ 0x418b1c,   1, 0x04, 0x000000e6 },
+	{ 0x418bb8,   1, 0x04, 0x00000103 },
+	{ 0x418c08,   1, 0x04, 0x00000001 },
+	{ 0x418c10,   8, 0x04, 0x00000000 },
+	{ 0x418c40,   1, 0x04, 0xffffffff },
+	{ 0x418c6c,   1, 0x04, 0x00000001 },
+	{ 0x418c80,   1, 0x04, 0x2020000c },
+	{ 0x418c8c,   1, 0x04, 0x00000001 },
+	{ 0x418d24,   1, 0x04, 0x00000000 },
+	{ 0x419000,   1, 0x04, 0x00000780 },
+	{ 0x419004,   2, 0x04, 0x00000000 },
+	{ 0x419014,   1, 0x04, 0x00000004 },
+	{}
+};
+
+static struct nvc0_graph_init
+nv108_grctx_init_tpc[] = {
+	{ 0x419848,   1, 0x04, 0x00000000 },
+	{ 0x419864,   1, 0x04, 0x00000129 },
+	{ 0x419888,   1, 0x04, 0x00000000 },
+	{ 0x419a00,   1, 0x04, 0x000100f0 },
+	{ 0x419a04,   1, 0x04, 0x00000001 },
+	{ 0x419a08,   1, 0x04, 0x00000421 },
+	{ 0x419a0c,   1, 0x04, 0x00120000 },
+	{ 0x419a10,   1, 0x04, 0x00000000 },
+	{ 0x419a14,   1, 0x04, 0x00000200 },
+	{ 0x419a1c,   1, 0x04, 0x0000c000 },
+	{ 0x419a20,   1, 0x04, 0x00000800 },
+	{ 0x419a30,   1, 0x04, 0x00000001 },
+	{ 0x419ac4,   1, 0x04, 0x0037f440 },
+	{ 0x419c00,   1, 0x04, 0x0000001a },
+	{ 0x419c04,   1, 0x04, 0x80000006 },
+	{ 0x419c08,   1, 0x04, 0x00000002 },
+	{ 0x419c20,   1, 0x04, 0x00000000 },
+	{ 0x419c24,   1, 0x04, 0x00084210 },
+	{ 0x419c28,   1, 0x04, 0x3efbefbe },
+	{ 0x419ce8,   1, 0x04, 0x00000000 },
+	{ 0x419cf4,   1, 0x04, 0x00000203 },
+	{ 0x419e04,   1, 0x04, 0x00000000 },
+	{ 0x419e08,   1, 0x04, 0x0000001d },
+	{ 0x419e0c,   1, 0x04, 0x00000000 },
+	{ 0x419e10,   1, 0x04, 0x00001c02 },
+	{ 0x419e44,   1, 0x04, 0x0013eff2 },
+	{ 0x419e48,   1, 0x04, 0x00000000 },
+	{ 0x419e4c,   1, 0x04, 0x0000007f },
+	{ 0x419e50,   2, 0x04, 0x00000000 },
+	{ 0x419e58,   1, 0x04, 0x00000001 },
+	{ 0x419e5c,   3, 0x04, 0x00000000 },
+	{ 0x419e68,   1, 0x04, 0x00000002 },
+	{ 0x419e6c,  12, 0x04, 0x00000000 },
+	{ 0x419eac,   1, 0x04, 0x00001f8f },
+	{ 0x419eb0,   1, 0x04, 0x0db00da0 },
+	{ 0x419eb8,   1, 0x04, 0x00000000 },
+	{ 0x419ec8,   1, 0x04, 0x0001304f },
+	{ 0x419f30,   4, 0x04, 0x00000000 },
+	{ 0x419f40,   1, 0x04, 0x00000018 },
+	{ 0x419f44,   3, 0x04, 0x00000000 },
+	{ 0x419f58,   1, 0x04, 0x00000020 },
+	{ 0x419f70,   1, 0x04, 0x00000000 },
+	{ 0x419f78,   1, 0x04, 0x000001eb },
+	{ 0x419f7c,   1, 0x04, 0x00000404 },
+	{}
+};
+
+static struct nvc0_graph_init
+nv108_grctx_init_unk[] = {
+	{ 0x41be24,   1, 0x04, 0x00000006 },
+	{ 0x41bec0,   1, 0x04, 0x10000000 },
+	{ 0x41bec4,   1, 0x04, 0x00037f7f },
+	{ 0x41bee4,   1, 0x04, 0x00000000 },
+	{ 0x41bef0,   1, 0x04, 0x000003ff },
+	{ 0x41bf00,   1, 0x04, 0x0a418820 },
+	{ 0x41bf04,   1, 0x04, 0x062080e6 },
+	{ 0x41bf08,   1, 0x04, 0x020398a4 },
+	{ 0x41bf0c,   1, 0x04, 0x0e629062 },
+	{ 0x41bf10,   1, 0x04, 0x0a418820 },
+	{ 0x41bf14,   1, 0x04, 0x000000e6 },
+	{ 0x41bfd0,   1, 0x04, 0x00900103 },
+	{ 0x41bfe0,   1, 0x04, 0x00400001 },
+	{ 0x41bfe4,   1, 0x04, 0x00000000 },
+	{}
+};
+
+static void
+nv108_grctx_generate_mods(struct nvc0_graph_priv *priv, struct nvc0_grctx *info)
+{
+	u32 magic[GPC_MAX][2];
+	u32 offset;
+	int gpc;
+
+	mmio_data(0x003000, 0x0100, NV_MEM_ACCESS_RW | NV_MEM_ACCESS_SYS);
+	mmio_data(0x008000, 0x0100, NV_MEM_ACCESS_RW | NV_MEM_ACCESS_SYS);
+	mmio_data(0x060000, 0x1000, NV_MEM_ACCESS_RW);
+	mmio_list(0x40800c, 0x00000000,  8, 1);
+	mmio_list(0x408010, 0x80000000,  0, 0);
+	mmio_list(0x419004, 0x00000000,  8, 1);
+	mmio_list(0x419008, 0x00000000,  0, 0);
+	mmio_list(0x408004, 0x00000000,  8, 0);
+	mmio_list(0x408008, 0x80000030,  0, 0);
+	mmio_list(0x418808, 0x00000000,  8, 0);
+	mmio_list(0x41880c, 0x80000030,  0, 0);
+	mmio_list(0x418810, 0x80000000, 12, 2);
+	mmio_list(0x419848, 0x10000000, 12, 2);
+
+	mmio_list(0x405830, 0x02180648,  0, 0);
+	mmio_list(0x4064c4, 0x0192ffff,  0, 0);
+
+	for (gpc = 0, offset = 0; gpc < priv->gpc_nr; gpc++) {
+		u16 magic0 = 0x0218 * priv->tpc_nr[gpc];
+		u16 magic1 = 0x0648 * priv->tpc_nr[gpc];
+		magic[gpc][0]  = 0x10000000 | (magic0 << 16) | offset;
+		magic[gpc][1]  = 0x00000000 | (magic1 << 16);
+		offset += 0x0324 * priv->tpc_nr[gpc];
+	}
+
+	for (gpc = 0; gpc < priv->gpc_nr; gpc++) {
+		mmio_list(GPC_UNIT(gpc, 0x30c0), magic[gpc][0], 0, 0);
+		mmio_list(GPC_UNIT(gpc, 0x30e4), magic[gpc][1] | offset, 0, 0);
+		offset += 0x07ff * priv->tpc_nr[gpc];
+	}
+
+	mmio_list(0x17e91c, 0x0b040a0b, 0, 0);
+	mmio_list(0x17e920, 0x00090d08, 0, 0);
+}
+
+static struct nvc0_graph_init *
+nv108_grctx_init_hub[] = {
+	nvc0_grctx_init_base,
+	nv108_grctx_init_unk40xx,
+	nvf0_grctx_init_unk44xx,
+	nve4_grctx_init_unk46xx,
+	nve4_grctx_init_unk47xx,
+	nv108_grctx_init_unk58xx,
+	nvf0_grctx_init_unk5bxx,
+	nvf0_grctx_init_unk60xx,
+	nv108_grctx_init_unk64xx,
+	nv108_grctx_init_unk78xx,
+	nve4_grctx_init_unk80xx,
+	nv108_grctx_init_unk88xx,
+	NULL
+};
+
+struct nvc0_graph_init *
+nv108_grctx_init_gpc[] = {
+	nv108_grctx_init_gpc_0,
+	nvc0_grctx_init_gpc_1,
+	nv108_grctx_init_tpc,
+	nv108_grctx_init_unk,
+	NULL
+};
+
+struct nvc0_graph_init
+nv108_grctx_init_mthd_magic[] = {
+	{ 0x3410, 1, 0x04, 0x8e0e2006 },
+	{ 0x3414, 1, 0x04, 0x00000038 },
+	{}
+};
+
+static struct nvc0_graph_mthd
+nv108_grctx_init_mthd[] = {
+	{ 0xa197, nv108_grctx_init_a197, },
+	{ 0x902d, nvc0_grctx_init_902d, },
+	{ 0x902d, nv108_grctx_init_mthd_magic, },
+	{}
+};
+
+struct nouveau_oclass *
+nv108_grctx_oclass = &(struct nvc0_grctx_oclass) {
+	.base.handle = NV_ENGCTX(GR, 0x08),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nvc0_graph_context_ctor,
+		.dtor = nvc0_graph_context_dtor,
+		.init = _nouveau_graph_context_init,
+		.fini = _nouveau_graph_context_fini,
+		.rd32 = _nouveau_graph_context_rd32,
+		.wr32 = _nouveau_graph_context_wr32,
+	},
+	.main = nve4_grctx_generate_main,
+	.mods = nv108_grctx_generate_mods,
+	.unkn = nve4_grctx_generate_unkn,
+	.hub  = nv108_grctx_init_hub,
+	.gpc  = nv108_grctx_init_gpc,
+	.icmd = nv108_grctx_init_icmd,
+	.mthd = nv108_grctx_init_mthd,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvf0.c b/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvf0.c
index dcb2ebb8c29d..44012c3da538 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvf0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/ctxnvf0.c
@@ -50,7 +50,7 @@ nvf0_grctx_init_unk40xx[] = {
 	{}
 };
 
-static struct nvc0_graph_init
+struct nvc0_graph_init
 nvf0_grctx_init_unk44xx[] = {
 	{ 0x404404,  12, 0x04, 0x00000000 },
 	{ 0x404438,   1, 0x04, 0x00000000 },
@@ -62,7 +62,7 @@ nvf0_grctx_init_unk44xx[] = {
 	{}
 };
 
-static struct nvc0_graph_init
+struct nvc0_graph_init
 nvf0_grctx_init_unk5bxx[] = {
 	{ 0x405b00,   1, 0x04, 0x00000000 },
 	{ 0x405b10,   1, 0x04, 0x00001000 },
@@ -70,7 +70,7 @@ nvf0_grctx_init_unk5bxx[] = {
 	{}
 };
 
-static struct nvc0_graph_init
+struct nvc0_graph_init
 nvf0_grctx_init_unk60xx[] = {
 	{ 0x406020,   1, 0x04, 0x034103c1 },
 	{ 0x406028,   4, 0x04, 0x00000001 },
@@ -286,7 +286,6 @@ nvf0_grctx_init_hub[] = {
 	nvf0_grctx_init_unk64xx,
 	nve4_grctx_init_unk80xx,
 	nvf0_grctx_init_unk88xx,
-	nvd9_grctx_init_rop,
 	NULL
 };
 
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/com.fuc b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/com.fuc
index 5d24b6de16cc..e148961b8075 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/com.fuc
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/com.fuc
@@ -38,7 +38,7 @@ queue_put:
 	cmpu b32 $r8 $r9
 	bra ne #queue_put_next
 		mov $r15 E_CMD_OVERFLOW
-		call #error
+		call(error)
 		ret
 
 	// store cmd/data on queue
@@ -92,18 +92,16 @@ queue_get_done:
 // Out: $r15 value
 //
 nv_rd32:
-	mov $r11 0x728
-	shl b32 $r11 6
 	mov b32 $r12 $r14
 	bset $r12 31			// MMIO_CTRL_PENDING
-	iowr I[$r11 + 0x000] $r12	// MMIO_CTRL
+	nv_iowr(NV_PGRAPH_FECS_MMIO_CTRL, 0, $r12)
 	nv_rd32_wait:
-		iord $r12 I[$r11 + 0x000]
+		nv_iord($r12, NV_PGRAPH_FECS_MMIO_CTRL, 0)
 		xbit $r12 $r12 31
 		bra ne #nv_rd32_wait
 	mov $r10 6			// DONE_MMIO_RD
-	call #wait_doneo
-	iord $r15 I[$r11 + 0x100]	// MMIO_RDVAL
+	call(wait_doneo)
+	nv_iord($r15, NV_PGRAPH_FECS_MMIO_RDVAL, 0)
 	ret
 
 // nv_wr32 - write 32-bit value to nv register
@@ -112,37 +110,17 @@ nv_rd32:
 //      $r15 value
 //
 nv_wr32:
-	mov $r11 0x728
-	shl b32 $r11 6
-	iowr I[$r11 + 0x200] $r15	// MMIO_WRVAL
+	nv_iowr(NV_PGRAPH_FECS_MMIO_WRVAL, 0, $r15)
 	mov b32 $r12 $r14
 	bset $r12 31			// MMIO_CTRL_PENDING
 	bset $r12 30			// MMIO_CTRL_WRITE
-	iowr I[$r11 + 0x000] $r12	// MMIO_CTRL
+	nv_iowr(NV_PGRAPH_FECS_MMIO_CTRL, 0, $r12)
 	nv_wr32_wait:
-		iord $r12 I[$r11 + 0x000]
+		nv_iord($r12, NV_PGRAPH_FECS_MMIO_CTRL, 0)
 		xbit $r12 $r12 31
 		bra ne #nv_wr32_wait
 	ret
 
-// (re)set watchdog timer
-//
-// In : $r15 timeout
-//
-watchdog_reset:
-	mov $r8 0x430
-	shl b32 $r8 6
-	bset $r15 31
-	iowr I[$r8 + 0x000] $r15
-	ret
-
-// clear watchdog timer
-watchdog_clear:
-	mov $r8 0x430
-	shl b32 $r8 6
-	iowr I[$r8 + 0x000] $r0
-	ret
-
 // wait_donez - wait on FUC_DONE bit to become clear
 //
 // In : $r10 bit to wait on
@@ -163,13 +141,9 @@ wait_donez:
 //
 wait_doneo:
 	trace_set(T_WAIT);
-	mov $r8 0x818
-	shl b32 $r8 6
-	iowr I[$r8 + 0x000] $r10
+	nv_iowr(NV_PGRAPH_FECS_CC_SCRATCH_VAL(6), 0, $r10)
 	wait_doneo_e:
-		mov $r8 0x400
-		shl b32 $r8 6
-		iord $r8 I[$r8 + 0x000]
+		nv_iord($r8, NV_PGRAPH_FECS_SIGNAL, 0)
 		xbit $r8 $r8 $r10
 		bra e #wait_doneo_e
 	trace_clr(T_WAIT)
@@ -209,21 +183,18 @@ mmctx_size:
 //
 mmctx_xfer:
 	trace_set(T_MMCTX)
-	mov $r8 0x710
-	shl b32 $r8 6
 	clear b32 $r9
 	or $r11 $r11
 	bra e #mmctx_base_disabled
-		iowr I[$r8 + 0x000] $r11	// MMCTX_BASE
+		nv_iowr(NV_PGRAPH_FECS_MMCTX_BASE, 0, $r11)
 		bset $r9 0			// BASE_EN
 	mmctx_base_disabled:
 	or $r14 $r14
 	bra e #mmctx_multi_disabled
-		iowr I[$r8 + 0x200] $r14 	// MMCTX_MULTI_STRIDE
-		iowr I[$r8 + 0x300] $r15 	// MMCTX_MULTI_MASK
+		nv_iowr(NV_PGRAPH_FECS_MMCTX_MULTI_STRIDE, 0, $r14)
+		nv_iowr(NV_PGRAPH_FECS_MMCTX_MULTI_MASK, 0, $r15)
 		bset $r9 1			// MULTI_EN
 	mmctx_multi_disabled:
-	add b32 $r8 0x100
 
 	xbit $r11 $r10 0
 	shl b32 $r11 16			// DIR
@@ -231,20 +202,20 @@ mmctx_xfer:
 	xbit $r14 $r10 1
 	shl b32 $r14 17
 	or $r11 $r14			// START_TRIGGER
-	iowr I[$r8 + 0x000] $r11	// MMCTX_CTRL
+	nv_iowr(NV_PGRAPH_FECS_MMCTX_CTRL, 0, $r11)
 
 	// loop over the mmio list, and send requests to the hw
 	mmctx_exec_loop:
 		// wait for space in mmctx queue
 		mmctx_wait_free:
-			iord $r14 I[$r8 + 0x000] // MMCTX_CTRL
+			nv_iord($r14, NV_PGRAPH_FECS_MMCTX_CTRL, 0)
 			and $r14 0x1f
 			bra e #mmctx_wait_free
 
 		// queue up an entry
 		ld b32 $r14 D[$r12]
 		or $r14 $r9
-		iowr I[$r8 + 0x300] $r14
+		nv_iowr(NV_PGRAPH_FECS_MMCTX_QUEUE, 0, $r14)
 		add b32 $r12 4
 		cmpu b32 $r12 $r13
 		bra ne #mmctx_exec_loop
@@ -253,22 +224,22 @@ mmctx_xfer:
 	bra ne #mmctx_stop
 		// wait for queue to empty
 		mmctx_fini_wait:
-			iord $r11 I[$r8 + 0x000]	// MMCTX_CTRL
+			nv_iord($r11, NV_PGRAPH_FECS_MMCTX_CTRL, 0)
 			and $r11 0x1f
 			cmpu b32 $r11 0x10
 			bra ne #mmctx_fini_wait
 		mov $r10 2				// DONE_MMCTX
-		call #wait_donez
+		call(wait_donez)
 		bra #mmctx_done
 	mmctx_stop:
 		xbit $r11 $r10 0
 		shl b32 $r11 16			// DIR
 		bset $r11 12			// QLIMIT = 0x10
 		bset $r11 18			// STOP_TRIGGER
-		iowr I[$r8 + 0x000] $r11	// MMCTX_CTRL
+		nv_iowr(NV_PGRAPH_FECS_MMCTX_CTRL, 0, $r11)
 		mmctx_stop_wait:
 			// wait for STOP_TRIGGER to clear
-			iord $r11 I[$r8 + 0x000] // MMCTX_CTRL
+			nv_iord($r11, NV_PGRAPH_FECS_MMCTX_CTRL, 0)
 			xbit $r11 $r11 18
 			bra ne #mmctx_stop_wait
 	mmctx_done:
@@ -280,28 +251,24 @@ mmctx_xfer:
 strand_wait:
 	push $r10
 	mov $r10 2
-	call #wait_donez
+	call(wait_donez)
 	pop $r10
 	ret
 
 // unknown - call before issuing strand commands
 //
 strand_pre:
-	mov $r8 0x4afc
-	sethi $r8 0x20000
-	mov $r9 0xc
-	iowr I[$r8] $r9
-	call #strand_wait
+	mov $r9 NV_PGRAPH_FECS_STRAND_CMD_ENABLE
+	nv_iowr(NV_PGRAPH_FECS_STRAND_CMD, 0x3f, $r9)
+	call(strand_wait)
 	ret
 
 // unknown - call after issuing strand commands
 //
 strand_post:
-	mov $r8 0x4afc
-	sethi $r8 0x20000
-	mov $r9 0xd
-	iowr I[$r8] $r9
-	call #strand_wait
+	mov $r9 NV_PGRAPH_FECS_STRAND_CMD_DISABLE
+	nv_iowr(NV_PGRAPH_FECS_STRAND_CMD, 0x3f, $r9)
+	call(strand_wait)
 	ret
 
 // Selects strand set?!
@@ -309,18 +276,14 @@ strand_post:
 // In: $r14 id
 //
 strand_set:
-	mov $r10 0x4ffc
-	sethi $r10 0x20000
-	sub b32 $r11 $r10 0x500
 	mov $r12 0xf
-	iowr I[$r10 + 0x000] $r12		// 0x93c = 0xf
-	mov $r12 0xb
-	iowr I[$r11 + 0x000] $r12		// 0x928 = 0xb
-	call #strand_wait
-	iowr I[$r10 + 0x000] $r14		// 0x93c = <id>
-	mov $r12 0xa
-	iowr I[$r11 + 0x000] $r12		// 0x928 = 0xa
-	call #strand_wait
+	nv_iowr(NV_PGRAPH_FECS_STRAND_FILTER, 0x3f, $r12)
+	mov $r12 NV_PGRAPH_FECS_STRAND_CMD_DEACTIVATE_FILTER
+	nv_iowr(NV_PGRAPH_FECS_STRAND_CMD, 0x3f, $r12)
+	nv_iowr(NV_PGRAPH_FECS_STRAND_FILTER, 0x3f, $r14)
+	mov $r12 NV_PGRAPH_FECS_STRAND_CMD_ACTIVATE_FILTER
+	nv_iowr(NV_PGRAPH_FECS_STRAND_CMD, 0x3f, $r12)
+	call(strand_wait)
 	ret
 
 // Initialise strand context data
@@ -332,30 +295,27 @@ strand_set:
 //
 strand_ctx_init:
 	trace_set(T_STRINIT)
-	call #strand_pre
+	call(strand_pre)
 	mov $r14 3
-	call #strand_set
-	mov $r10 0x46fc
-	sethi $r10 0x20000
-	add b32 $r11 $r10 0x400
-	iowr I[$r10 + 0x100] $r0	// STRAND_FIRST_GENE = 0
-	mov $r12 1
-	iowr I[$r11 + 0x000] $r12	// STRAND_CMD = LATCH_FIRST_GENE
-	call #strand_wait
+	call(strand_set)
+
+	clear b32 $r12
+	nv_iowr(NV_PGRAPH_FECS_STRAND_SELECT, 0x3f, $r12)
+	mov $r12 NV_PGRAPH_FECS_STRAND_CMD_SEEK
+	nv_iowr(NV_PGRAPH_FECS_STRAND_CMD, 0x3f, $r12)
+	call(strand_wait)
 	sub b32 $r12 $r0 1
-	iowr I[$r10 + 0x000] $r12	// STRAND_GENE_CNT = 0xffffffff
-	mov $r12 2
-	iowr I[$r11 + 0x000] $r12	// STRAND_CMD = LATCH_GENE_CNT
-	call #strand_wait
-	call #strand_post
+	nv_iowr(NV_PGRAPH_FECS_STRAND_DATA, 0x3f, $r12)
+	mov $r12 NV_PGRAPH_FECS_STRAND_CMD_GET_INFO
+	nv_iowr(NV_PGRAPH_FECS_STRAND_CMD, 0x3f, $r12)
+	call(strand_wait)
+	call(strand_post)
 
 	// read the size of each strand, poke the context offset of
 	// each into STRAND_{SAVE,LOAD}_SWBASE now, no need to worry
 	// about it later then.
-	mov $r8 0x880
-	shl b32 $r8 6
-	iord $r9 I[$r8 + 0x000]		// STRANDS
-	add b32 $r8 0x2200
+	nv_mkio($r8, NV_PGRAPH_FECS_STRAND_SAVE_SWBASE, 0x00)
+	nv_iord($r9, NV_PGRAPH_FECS_STRANDS_CNT, 0x00)
 	shr b32 $r14 $r15 8
 	ctx_init_strand_loop:
 		iowr I[$r8 + 0x000] $r14	// STRAND_SAVE_SWBASE
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpc.fuc b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpc.fuc
index 5547c1b3f4f2..96cbcea3b2c9 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpc.fuc
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpc.fuc
@@ -58,12 +58,9 @@ mmio_list_base:
 //
 error:
 	push $r14
-	mov $r14 -0x67ec 	// 0x9814
-	sethi $r14 0x400000
-	call #nv_wr32		// HUB_CTXCTL_CC_SCRATCH[5] = error code
-	add b32 $r14 0x41c
+	nv_wr32(NV_PGRAPH_FECS_CC_SCRATCH_VAL(5), $r15)
 	mov $r15 1
-	call #nv_wr32		// HUB_CTXCTL_INTR_UP_SET
+	nv_wr32(NV_PGRAPH_FECS_INTR_UP_SET, $r15)
 	pop $r14
 	ret
 
@@ -84,46 +81,40 @@ init:
 	mov $sp $r0
 
 	// enable fifo access
-	mov $r1 0x1200
-	mov $r2 2
-	iowr I[$r1 + 0x000] $r2		// FIFO_ENABLE
+	mov $r2 NV_PGRAPH_GPCX_GPCCS_ACCESS_FIFO
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_ACCESS, 0, $r2)
 
 	// setup i0 handler, and route all interrupts to it
 	mov $r1 #ih
 	mov $iv0 $r1
-	mov $r1 0x400
-	iowr I[$r1 + 0x300] $r0		// INTR_DISPATCH
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_INTR_ROUTE, 0, $r0)
 
 	// enable fifo interrupt
-	mov $r2 4
-	iowr I[$r1 + 0x000] $r2		// INTR_EN_SET
+	mov $r2 NV_PGRAPH_GPCX_GPCCS_INTR_EN_SET_FIFO
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_INTR_EN_SET, 0, $r2)
 
 	// enable interrupts
 	bset $flags ie0
 
 	// figure out which GPC we are, and how many TPCs we have
-	mov $r1 0x608
-	shl b32 $r1 6
-	iord $r2 I[$r1 + 0x000]		// UNITS
+	nv_iord($r2, NV_PGRAPH_GPCX_GPCCS_UNITS, 0)
 	mov $r3 1
 	and $r2 0x1f
 	shl b32 $r3 $r2
 	sub b32 $r3 1
 	st b32 D[$r0 + #tpc_count] $r2
 	st b32 D[$r0 + #tpc_mask] $r3
-	add b32 $r1 0x400
-	iord $r2 I[$r1 + 0x000]		// MYINDEX
+	nv_iord($r2, NV_PGRAPH_GPCX_GPCCS_MYINDEX, 0)
 	st b32 D[$r0 + #gpc_id] $r2
 
 #if NV_PGRAPH_GPCX_UNK__SIZE > 0
 	// figure out which, and how many, UNKs are actually present
-	mov $r14 0x0c30
-	sethi $r14 0x500000
+	imm32($r14, 0x500c30)
 	clear b32 $r2
 	clear b32 $r3
 	clear b32 $r4
 	init_unk_loop:
-		call #nv_rd32
+		call(nv_rd32)
 		cmp b32 $r15 0
 		bra z #init_unk_next
 			mov $r15 1
@@ -146,23 +137,21 @@ init:
 
 	// set mmctx base addresses now so we don't have to do it later,
 	// they don't currently ever change
-	mov $r4 0x700
-	shl b32 $r4 6
 	shr b32 $r5 $r2 8
-	iowr I[$r4 + 0x000] $r5		// MMCTX_SAVE_SWBASE
-	iowr I[$r4 + 0x100] $r5		// MMCTX_LOAD_SWBASE
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_MMCTX_SAVE_SWBASE, 0, $r5)
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_MMCTX_LOAD_SWBASE, 0, $r5)
 
 	// calculate GPC mmio context size
 	ld b32 $r14 D[$r0 + #gpc_mmio_list_head]
 	ld b32 $r15 D[$r0 + #gpc_mmio_list_tail]
-	call #mmctx_size
+	call(mmctx_size)
 	add b32 $r2 $r15
 	add b32 $r3 $r15
 
 	// calculate per-TPC mmio context size
 	ld b32 $r14 D[$r0 + #tpc_mmio_list_head]
 	ld b32 $r15 D[$r0 + #tpc_mmio_list_tail]
-	call #mmctx_size
+	call(mmctx_size)
 	ld b32 $r14 D[$r0 + #tpc_count]
 	mulu $r14 $r15
 	add b32 $r2 $r14
@@ -172,7 +161,7 @@ init:
 	// calculate per-UNK mmio context size
 	ld b32 $r14 D[$r0 + #unk_mmio_list_head]
 	ld b32 $r15 D[$r0 + #unk_mmio_list_tail]
-	call #mmctx_size
+	call(mmctx_size)
 	ld b32 $r14 D[$r0 + #unk_count]
 	mulu $r14 $r15
 	add b32 $r2 $r14
@@ -180,9 +169,8 @@ init:
 #endif
 
 	// round up base/size to 256 byte boundary (for strand SWBASE)
-	add b32 $r4 0x1300
 	shr b32 $r3 2
-	iowr I[$r4 + 0x000] $r3		// MMCTX_LOAD_COUNT, wtf for?!?
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_MMCTX_LOAD_COUNT, 0, $r3) // wtf for?!
 	shr b32 $r2 8
 	shr b32 $r3 6
 	add b32 $r2 1
@@ -192,7 +180,7 @@ init:
 
 	// calculate size of strand context data
 	mov b32 $r15 $r2
-	call #strand_ctx_init
+	call(strand_ctx_init)
 	add b32 $r3 $r15
 
 	// save context size, and tell HUB we're done
@@ -208,7 +196,7 @@ main:
 	bset $flags $p0
 	sleep $p0
 	mov $r13 #cmd_queue
-	call #queue_get
+	call(queue_get)
 	bra $p1 #main
 
 	// 0x0000-0x0003 are all context transfers
@@ -224,13 +212,13 @@ main:
 		or $r1 $r14
 		mov $flags $r1
 		// transfer context data
-		call #ctx_xfer
+		call(ctx_xfer)
 		bra #main
 
 	main_not_ctx_xfer:
 	shl b32 $r15 $r14 16
 	or $r15 E_BAD_COMMAND
-	call #error
+	call(error)
 	bra #main
 
 // interrupt handler
@@ -247,22 +235,20 @@ ih:
 	clear b32 $r0
 
 	// incoming fifo command?
-	iord $r10 I[$r0 + 0x200]	// INTR
-	and $r11 $r10 0x00000004
+	nv_iord($r10, NV_PGRAPH_GPCX_GPCCS_INTR, 0)
+	and $r11 $r10 NV_PGRAPH_GPCX_GPCCS_INTR_FIFO
 	bra e #ih_no_fifo
 		// queue incoming fifo command for later processing
-		mov $r11 0x1900
 		mov $r13 #cmd_queue
-		iord $r14 I[$r11 + 0x100]	// FIFO_CMD
-		iord $r15 I[$r11 + 0x000]	// FIFO_DATA
-		call #queue_put
-		add b32 $r11 0x400
+		nv_iord($r14, NV_PGRAPH_GPCX_GPCCS_FIFO_CMD, 0)
+		nv_iord($r15, NV_PGRAPH_GPCX_GPCCS_FIFO_DATA, 0)
+		call(queue_put)
 		mov $r14 1
-		iowr I[$r11 + 0x000] $r14	// FIFO_ACK
+		nv_iowr(NV_PGRAPH_GPCX_GPCCS_FIFO_ACK, 0, $r14)
 
 	// ack, and wake up main()
 	ih_no_fifo:
-	iowr I[$r0 + 0x100] $r10	// INTR_ACK
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_INTR_ACK, 0, $r10)
 
 	pop $r15
 	pop $r14
@@ -283,9 +269,7 @@ hub_barrier_done:
 	mov $r15 1
 	ld b32 $r14 D[$r0 + #gpc_id]
 	shl b32 $r15 $r14
-	mov $r14 -0x6be8 	// 0x409418 - HUB_BAR_SET
-	sethi $r14 0x400000
-	call #nv_wr32
+	nv_wr32(0x409418, $r15)	// 0x409418 - HUB_BAR_SET
 	ret
 
 // Disables various things, waits a bit, and re-enables them..
@@ -295,16 +279,15 @@ hub_barrier_done:
 // funny things happen.
 //
 ctx_redswitch:
-	mov $r14 0x614
-	shl b32 $r14 6
-	mov $r15 0x020
-	iowr I[$r14] $r15	// GPC_RED_SWITCH = POWER
-	mov $r15 8
+	mov $r15 NV_PGRAPH_GPCX_GPCCS_RED_SWITCH_POWER
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_RED_SWITCH, 0, $r15)
+	mov $r14 8
 	ctx_redswitch_delay:
-		sub b32 $r15 1
+		sub b32 $r14 1
 		bra ne #ctx_redswitch_delay
-	mov $r15 0xa20
-	iowr I[$r14] $r15	// GPC_RED_SWITCH = UNK11, ENABLE, POWER
+	or $r15 NV_PGRAPH_GPCX_GPCCS_RED_SWITCH_UNK11
+	or $r15 NV_PGRAPH_GPCX_GPCCS_RED_SWITCH_ENABLE
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_RED_SWITCH, 0, $r15)
 	ret
 
 // Transfer GPC context data between GPU and storage area
@@ -317,46 +300,37 @@ ctx_redswitch:
 //
 ctx_xfer:
 	// set context base address
-	mov $r1 0xa04
-	shl b32 $r1 6
-	iowr I[$r1 + 0x000] $r15// MEM_BASE
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_MEM_BASE, 0, $r15)
 	bra not $p1 #ctx_xfer_not_load
-		call #ctx_redswitch
+		call(ctx_redswitch)
 	ctx_xfer_not_load:
 
 	// strands
-	mov $r1 0x4afc
-	sethi $r1 0x20000
-	mov $r2 0xc
-	iowr I[$r1] $r2		// STRAND_CMD(0x3f) = 0x0c
-	call #strand_wait
-	mov $r2 0x47fc
-	sethi $r2 0x20000
-	iowr I[$r2] $r0		// STRAND_FIRST_GENE(0x3f) = 0x00
-	xbit $r2 $flags $p1
-	add b32 $r2 3
-	iowr I[$r1] $r2		// STRAND_CMD(0x3f) = 0x03/0x04 (SAVE/LOAD)
+	call(strand_pre)
+	clear b32 $r2
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_STRAND_SELECT, 0x3f, $r2)
+	xbit $r2 $flags $p1	// SAVE/LOAD
+	add b32 $r2 NV_PGRAPH_GPCX_GPCCS_STRAND_CMD_SAVE
+	nv_iowr(NV_PGRAPH_GPCX_GPCCS_STRAND_CMD, 0x3f, $r2)
 
 	// mmio context
 	xbit $r10 $flags $p1	// direction
 	or $r10 2		// first
-	mov $r11 0x0000
-	sethi $r11 0x500000
+	imm32($r11,0x500000)
 	ld b32 $r12 D[$r0 + #gpc_id]
 	shl b32 $r12 15
 	add b32 $r11 $r12	// base = NV_PGRAPH_GPCn
 	ld b32 $r12 D[$r0 + #gpc_mmio_list_head]
 	ld b32 $r13 D[$r0 + #gpc_mmio_list_tail]
 	mov $r14 0		// not multi
-	call #mmctx_xfer
+	call(mmctx_xfer)
 
 	// per-TPC mmio context
 	xbit $r10 $flags $p1	// direction
 #if !NV_PGRAPH_GPCX_UNK__SIZE
 	or $r10 4		// last
 #endif
-	mov $r11 0x4000
-	sethi $r11 0x500000	// base = NV_PGRAPH_GPC0_TPC0
+	imm32($r11, 0x504000)
 	ld b32 $r12 D[$r0 + #gpc_id]
 	shl b32 $r12 15
 	add b32 $r11 $r12	// base = NV_PGRAPH_GPCn_TPC0
@@ -364,14 +338,13 @@ ctx_xfer:
 	ld b32 $r13 D[$r0 + #tpc_mmio_list_tail]
 	ld b32 $r15 D[$r0 + #tpc_mask]
 	mov $r14 0x800		// stride = 0x800
-	call #mmctx_xfer
+	call(mmctx_xfer)
 
 #if NV_PGRAPH_GPCX_UNK__SIZE > 0
 	// per-UNK mmio context
 	xbit $r10 $flags $p1	// direction
 	or $r10 4		// last
-	mov $r11 0x3000
-	sethi $r11 0x500000	// base = NV_PGRAPH_GPC0_UNK0
+	imm32($r11, 0x503000)
 	ld b32 $r12 D[$r0 + #gpc_id]
 	shl b32 $r12 15
 	add b32 $r11 $r12	// base = NV_PGRAPH_GPCn_UNK0
@@ -379,11 +352,11 @@ ctx_xfer:
 	ld b32 $r13 D[$r0 + #unk_mmio_list_tail]
 	ld b32 $r15 D[$r0 + #unk_mask]
 	mov $r14 0x200		// stride = 0x200
-	call #mmctx_xfer
+	call(mmctx_xfer)
 #endif
 
 	// wait for strands to finish
-	call #strand_wait
+	call(strand_wait)
 
 	// if load, or a save without a load following, do some
 	// unknown stuff that's done after finishing a block of
@@ -391,14 +364,10 @@ ctx_xfer:
 	bra $p1 #ctx_xfer_post
 	bra not $p2 #ctx_xfer_done
 	ctx_xfer_post:
-		mov $r1 0x4afc
-		sethi $r1 0x20000
-		mov $r2 0xd
-		iowr I[$r1] $r2		// STRAND_CMD(0x3f) = 0x0d
-		call #strand_wait
+		call(strand_post)
 
 	// mark completion in HUB's barrier
 	ctx_xfer_done:
-	call #hub_barrier_done
+	call(hub_barrier_done)
 	ret
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnv108.fuc5 b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnv108.fuc5
new file mode 100644
index 000000000000..bd30262d635b
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnv108.fuc5
@@ -0,0 +1,42 @@
+/*
+ * 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 <bskeggs@redhat.com>
+ */
+
+#define NV_PGRAPH_GPCX_UNK__SIZE                                     0x00000001
+
+#define CHIPSET GK208
+#include "macros.fuc"
+
+.section #nv108_grgpc_data
+#define INCLUDE_DATA
+#include "com.fuc"
+#include "gpc.fuc"
+#undef INCLUDE_DATA
+
+.section #nv108_grgpc_code
+#define INCLUDE_CODE
+bra #init
+#include "com.fuc"
+#include "gpc.fuc"
+.align 256
+#undef INCLUDE_CODE
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnv108.fuc5.h b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnv108.fuc5.h
new file mode 100644
index 000000000000..27dc1280dc10
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnv108.fuc5.h
@@ -0,0 +1,473 @@
+uint32_t nv108_grgpc_data[] = {
+/* 0x0000: gpc_mmio_list_head */
+	0x0000006c,
+/* 0x0004: gpc_mmio_list_tail */
+/* 0x0004: tpc_mmio_list_head */
+	0x0000006c,
+/* 0x0008: tpc_mmio_list_tail */
+/* 0x0008: unk_mmio_list_head */
+	0x0000006c,
+/* 0x000c: unk_mmio_list_tail */
+	0x0000006c,
+/* 0x0010: gpc_id */
+	0x00000000,
+/* 0x0014: tpc_count */
+	0x00000000,
+/* 0x0018: tpc_mask */
+	0x00000000,
+/* 0x001c: unk_count */
+	0x00000000,
+/* 0x0020: unk_mask */
+	0x00000000,
+/* 0x0024: cmd_queue */
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+};
+
+uint32_t nv108_grgpc_code[] = {
+	0x03140ef5,
+/* 0x0004: queue_put */
+	0x9800d898,
+	0x86f001d9,
+	0xf489a408,
+	0x020f0b1b,
+	0x0002f87e,
+/* 0x001a: queue_put_next */
+	0x98c400f8,
+	0x0384b607,
+	0xb6008dbb,
+	0x8eb50880,
+	0x018fb500,
+	0xf00190b6,
+	0xd9b50f94,
+/* 0x0037: queue_get */
+	0xf400f801,
+	0xd8980131,
+	0x01d99800,
+	0x0bf489a4,
+	0x0789c421,
+	0xbb0394b6,
+	0x90b6009d,
+	0x009e9808,
+	0xb6019f98,
+	0x84f00180,
+	0x00d8b50f,
+/* 0x0063: queue_get_done */
+	0xf80132f4,
+/* 0x0065: nv_rd32 */
+	0xf0ecb200,
+	0x00801fc9,
+	0x0cf601ca,
+/* 0x0073: nv_rd32_wait */
+	0x8c04bd00,
+	0xcf01ca00,
+	0xccc800cc,
+	0xf61bf41f,
+	0xec7e060a,
+	0x008f0000,
+	0xffcf01cb,
+/* 0x008f: nv_wr32 */
+	0x8000f800,
+	0xf601cc00,
+	0x04bd000f,
+	0xc9f0ecb2,
+	0x1ec9f01f,
+	0x01ca0080,
+	0xbd000cf6,
+/* 0x00a9: nv_wr32_wait */
+	0xca008c04,
+	0x00cccf01,
+	0xf41fccc8,
+	0x00f8f61b,
+/* 0x00b8: wait_donez */
+	0x99f094bd,
+	0x37008000,
+	0x0009f602,
+	0x008004bd,
+	0x0af60206,
+/* 0x00cf: wait_donez_ne */
+	0x8804bd00,
+	0xcf010000,
+	0x8aff0088,
+	0xf61bf488,
+	0x99f094bd,
+	0x17008000,
+	0x0009f602,
+	0x00f804bd,
+/* 0x00ec: wait_doneo */
+	0x99f094bd,
+	0x37008000,
+	0x0009f602,
+	0x008004bd,
+	0x0af60206,
+/* 0x0103: wait_doneo_e */
+	0x8804bd00,
+	0xcf010000,
+	0x8aff0088,
+	0xf60bf488,
+	0x99f094bd,
+	0x17008000,
+	0x0009f602,
+	0x00f804bd,
+/* 0x0120: mmctx_size */
+/* 0x0122: nv_mmctx_size_loop */
+	0xe89894bd,
+	0x1a85b600,
+	0xb60180b6,
+	0x98bb0284,
+	0x04e0b600,
+	0x1bf4efa4,
+	0xf89fb2ec,
+/* 0x013d: mmctx_xfer */
+	0xf094bd00,
+	0x00800199,
+	0x09f60237,
+	0xbd04bd00,
+	0x05bbfd94,
+	0x800f0bf4,
+	0xf601c400,
+	0x04bd000b,
+/* 0x015f: mmctx_base_disabled */
+	0xfd0099f0,
+	0x0bf405ee,
+	0xc6008018,
+	0x000ef601,
+	0x008004bd,
+	0x0ff601c7,
+	0xf004bd00,
+/* 0x017a: mmctx_multi_disabled */
+	0xabc80199,
+	0x10b4b600,
+	0xc80cb9f0,
+	0xe4b601ae,
+	0x05befd11,
+	0x01c50080,
+	0xbd000bf6,
+/* 0x0195: mmctx_exec_loop */
+/* 0x0195: mmctx_wait_free */
+	0xc5008e04,
+	0x00eecf01,
+	0xf41fe4f0,
+	0xce98f60b,
+	0x05e9fd00,
+	0x01c80080,
+	0xbd000ef6,
+	0x04c0b604,
+	0x1bf4cda4,
+	0x02abc8df,
+/* 0x01bf: mmctx_fini_wait */
+	0x8b1c1bf4,
+	0xcf01c500,
+	0xb4f000bb,
+	0x10b4b01f,
+	0x0af31bf4,
+	0x00b87e02,
+	0x250ef400,
+/* 0x01d8: mmctx_stop */
+	0xb600abc8,
+	0xb9f010b4,
+	0x12b9f00c,
+	0x01c50080,
+	0xbd000bf6,
+/* 0x01ed: mmctx_stop_wait */
+	0xc5008b04,
+	0x00bbcf01,
+	0xf412bbc8,
+/* 0x01fa: mmctx_done */
+	0x94bdf61b,
+	0x800199f0,
+	0xf6021700,
+	0x04bd0009,
+/* 0x020a: strand_wait */
+	0xa0f900f8,
+	0xb87e020a,
+	0xa0fc0000,
+/* 0x0216: strand_pre */
+	0x0c0900f8,
+	0x024afc80,
+	0xbd0009f6,
+	0x020a7e04,
+/* 0x0227: strand_post */
+	0x0900f800,
+	0x4afc800d,
+	0x0009f602,
+	0x0a7e04bd,
+	0x00f80002,
+/* 0x0238: strand_set */
+	0xfc800f0c,
+	0x0cf6024f,
+	0x0c04bd00,
+	0x4afc800b,
+	0x000cf602,
+	0xfc8004bd,
+	0x0ef6024f,
+	0x0c04bd00,
+	0x4afc800a,
+	0x000cf602,
+	0x0a7e04bd,
+	0x00f80002,
+/* 0x0268: strand_ctx_init */
+	0x99f094bd,
+	0x37008003,
+	0x0009f602,
+	0x167e04bd,
+	0x030e0002,
+	0x0002387e,
+	0xfc80c4bd,
+	0x0cf60247,
+	0x0c04bd00,
+	0x4afc8001,
+	0x000cf602,
+	0x0a7e04bd,
+	0x0c920002,
+	0x46fc8001,
+	0x000cf602,
+	0x020c04bd,
+	0x024afc80,
+	0xbd000cf6,
+	0x020a7e04,
+	0x02277e00,
+	0x42008800,
+	0x20008902,
+	0x0099cf02,
+/* 0x02c7: ctx_init_strand_loop */
+	0xf608fe95,
+	0x8ef6008e,
+	0x808acf40,
+	0xb606a5b6,
+	0xeabb01a0,
+	0x0480b600,
+	0xf40192b6,
+	0xe4b6e81b,
+	0xf2efbc08,
+	0x99f094bd,
+	0x17008003,
+	0x0009f602,
+	0x00f804bd,
+/* 0x02f8: error */
+	0xffb2e0f9,
+	0x4098148e,
+	0x00008f7e,
+	0xffb2010f,
+	0x409c1c8e,
+	0x00008f7e,
+	0x00f8e0fc,
+/* 0x0314: init */
+	0x04fe04bd,
+	0x40020200,
+	0x02f61200,
+	0x4104bd00,
+	0x10fe0465,
+	0x07004000,
+	0xbd0000f6,
+	0x40040204,
+	0x02f60400,
+	0xf404bd00,
+	0x00821031,
+	0x22cf0182,
+	0xf0010300,
+	0x32bb1f24,
+	0x0132b604,
+	0xb50502b5,
+	0x00820603,
+	0x22cf0186,
+	0x0402b500,
+	0x500c308e,
+	0x34bd24bd,
+/* 0x036a: init_unk_loop */
+	0x657e44bd,
+	0xf6b00000,
+	0x0e0bf400,
+	0xf2bb010f,
+	0x054ffd04,
+/* 0x037f: init_unk_next */
+	0xb60130b6,
+	0xe0b60120,
+	0x0126b004,
+/* 0x038b: init_unk_done */
+	0xb5e21bf4,
+	0x04b50703,
+	0x01008208,
+	0x0022cf02,
+	0x259534bd,
+	0xc0008008,
+	0x0005f601,
+	0x008004bd,
+	0x05f601c1,
+	0x9804bd00,
+	0x0f98000e,
+	0x01207e01,
+	0x002fbb00,
+	0x98003fbb,
+	0x0f98010e,
+	0x01207e02,
+	0x050e9800,
+	0xbb00effd,
+	0x3ebb002e,
+	0x020e9800,
+	0x7e030f98,
+	0x98000120,
+	0xeffd070e,
+	0x002ebb00,
+	0xb6003ebb,
+	0x00800235,
+	0x03f601d3,
+	0xb604bd00,
+	0x35b60825,
+	0x0120b606,
+	0xb60130b6,
+	0x34b60824,
+	0x7e2fb208,
+	0xbb000268,
+	0x0080003f,
+	0x03f60201,
+	0xbd04bd00,
+	0x1f29f024,
+	0x02300080,
+	0xbd0002f6,
+/* 0x0429: main */
+	0x0031f404,
+	0x0d0028f4,
+	0x00377e24,
+	0xf401f400,
+	0xf404e4b0,
+	0x81fe1d18,
+	0xbd060201,
+	0x0412fd20,
+	0xfd01e4b6,
+	0x18fe051e,
+	0x04fc7e00,
+	0xd40ef400,
+/* 0x0458: main_not_ctx_xfer */
+	0xf010ef94,
+	0xf87e01f5,
+	0x0ef40002,
+/* 0x0465: ih */
+	0xfe80f9c7,
+	0x80f90188,
+	0xa0f990f9,
+	0xd0f9b0f9,
+	0xf0f9e0f9,
+	0x004a04bd,
+	0x00aacf02,
+	0xf404abc4,
+	0x240d1f0b,
+	0xcf1a004e,
+	0x004f00ee,
+	0x00ffcf19,
+	0x0000047e,
+	0x0040010e,
+	0x000ef61d,
+/* 0x04a2: ih_no_fifo */
+	0x004004bd,
+	0x000af601,
+	0xf0fc04bd,
+	0xd0fce0fc,
+	0xa0fcb0fc,
+	0x80fc90fc,
+	0xfc0088fe,
+	0x0032f480,
+/* 0x04c2: hub_barrier_done */
+	0x010f01f8,
+	0xbb040e98,
+	0xffb204fe,
+	0x4094188e,
+	0x00008f7e,
+/* 0x04d6: ctx_redswitch */
+	0x200f00f8,
+	0x01850080,
+	0xbd000ff6,
+/* 0x04e3: ctx_redswitch_delay */
+	0xb6080e04,
+	0x1bf401e2,
+	0x00f5f1fd,
+	0x00f5f108,
+	0x85008002,
+	0x000ff601,
+	0x00f804bd,
+/* 0x04fc: ctx_xfer */
+	0x02810080,
+	0xbd000ff6,
+	0x0711f404,
+	0x0004d67e,
+/* 0x050c: ctx_xfer_not_load */
+	0x0002167e,
+	0xfc8024bd,
+	0x02f60247,
+	0xf004bd00,
+	0x20b6012c,
+	0x4afc8003,
+	0x0002f602,
+	0xacf004bd,
+	0x02a5f001,
+	0x5000008b,
+	0xb6040c98,
+	0xbcbb0fc4,
+	0x000c9800,
+	0x0e010d98,
+	0x013d7e00,
+	0x01acf000,
+	0x5040008b,
+	0xb6040c98,
+	0xbcbb0fc4,
+	0x010c9800,
+	0x98020d98,
+	0x004e060f,
+	0x013d7e08,
+	0x01acf000,
+	0x8b04a5f0,
+	0x98503000,
+	0xc4b6040c,
+	0x00bcbb0f,
+	0x98020c98,
+	0x0f98030d,
+	0x02004e08,
+	0x00013d7e,
+	0x00020a7e,
+	0xf40601f4,
+/* 0x0596: ctx_xfer_post */
+	0x277e0712,
+/* 0x059a: ctx_xfer_done */
+	0xc27e0002,
+	0x00f80004,
+	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/engine/graph/fuc/gpcnvc0.fuc.h b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnvc0.fuc.h
index f2b0dea80116..0e7b01efae8d 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnvc0.fuc.h
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnvc0.fuc.h
@@ -37,14 +37,14 @@ uint32_t nvc0_grgpc_data[] = {
 };
 
 uint32_t nvc0_grgpc_code[] = {
-	0x03180ef5,
+	0x03a10ef5,
 /* 0x0004: queue_put */
 	0x9800d898,
 	0x86f001d9,
 	0x0489b808,
 	0xf00c1bf4,
 	0x21f502f7,
-	0x00f802fe,
+	0x00f8037e,
 /* 0x001c: queue_put_next */
 	0xb60798c4,
 	0x8dbb0384,
@@ -68,184 +68,214 @@ uint32_t nvc0_grgpc_code[] = {
 /* 0x0066: queue_get_done */
 	0x00f80132,
 /* 0x0068: nv_rd32 */
-	0x0728b7f1,
-	0xb906b4b6,
-	0xc9f002ec,
-	0x00bcd01f,
-/* 0x0078: nv_rd32_wait */
-	0xc800bccf,
-	0x1bf41fcc,
-	0x06a7f0fa,
-	0x010921f5,
-	0xf840bfcf,
-/* 0x008d: nv_wr32 */
-	0x28b7f100,
-	0x06b4b607,
-	0xb980bfd0,
-	0xc9f002ec,
-	0x1ec9f01f,
-/* 0x00a3: nv_wr32_wait */
-	0xcf00bcd0,
-	0xccc800bc,
-	0xfa1bf41f,
-/* 0x00ae: watchdog_reset */
-	0x87f100f8,
-	0x84b60430,
-	0x1ff9f006,
-	0xf8008fd0,
-/* 0x00bd: watchdog_clear */
-	0x3087f100,
-	0x0684b604,
-	0xf80080d0,
-/* 0x00c9: wait_donez */
-	0xf094bd00,
-	0x07f10099,
-	0x03f00f00,
-	0x0009d002,
-	0x07f104bd,
-	0x03f00600,
-	0x000ad002,
-/* 0x00e6: wait_donez_ne */
-	0x87f104bd,
-	0x83f00000,
-	0x0088cf01,
-	0xf4888aff,
-	0x94bdf31b,
-	0xf10099f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0109: wait_doneo */
-	0xf094bd00,
+	0xf002ecb9,
+	0x07f11fc9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x007a: nv_rd32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0xa7f0f31b,
+	0x1021f506,
+	0x00f7f101,
+	0x01f3f0cb,
+	0xf800ffcf,
+/* 0x009d: nv_wr32 */
+	0x0007f100,
+	0x0103f0cc,
+	0xbd000fd0,
+	0x02ecb904,
+	0xf01fc9f0,
+	0x07f11ec9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x00be: nv_wr32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0x00f8f31b,
+/* 0x00d0: wait_donez */
+	0x99f094bd,
+	0x0007f100,
+	0x0203f00f,
+	0xbd0009d0,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x00ed: wait_donez_ne */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x1bf4888a,
+	0xf094bdf3,
 	0x07f10099,
-	0x03f00f00,
+	0x03f01700,
 	0x0009d002,
-	0x87f104bd,
-	0x84b60818,
-	0x008ad006,
-/* 0x0124: wait_doneo_e */
-	0x040087f1,
-	0xcf0684b6,
-	0x8aff0088,
-	0xf30bf488,
+	0x00f804bd,
+/* 0x0110: wait_doneo */
 	0x99f094bd,
 	0x0007f100,
-	0x0203f017,
+	0x0203f00f,
 	0xbd0009d0,
-/* 0x0147: mmctx_size */
-	0xbd00f804,
-/* 0x0149: nv_mmctx_size_loop */
-	0x00e89894,
-	0xb61a85b6,
-	0x84b60180,
-	0x0098bb02,
-	0xb804e0b6,
-	0x1bf404ef,
-	0x029fb9eb,
-/* 0x0166: mmctx_xfer */
-	0x94bd00f8,
-	0xf10199f0,
-	0xf00f0007,
-	0x09d00203,
-	0xf104bd00,
-	0xb6071087,
-	0x94bd0684,
-	0xf405bbfd,
-	0x8bd0090b,
-	0x0099f000,
-/* 0x018c: mmctx_base_disabled */
-	0xf405eefd,
-	0x8ed00c0b,
-	0xc08fd080,
-/* 0x019b: mmctx_multi_disabled */
-	0xb70199f0,
-	0xc8010080,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x012d: wait_doneo_e */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x0bf4888a,
+	0xf094bdf3,
+	0x07f10099,
+	0x03f01700,
+	0x0009d002,
+	0x00f804bd,
+/* 0x0150: mmctx_size */
+/* 0x0152: nv_mmctx_size_loop */
+	0xe89894bd,
+	0x1a85b600,
+	0xb60180b6,
+	0x98bb0284,
+	0x04e0b600,
+	0xf404efb8,
+	0x9fb9eb1b,
+/* 0x016f: mmctx_xfer */
+	0xbd00f802,
+	0x0199f094,
+	0x0f0007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0xbbfd94bd,
+	0x120bf405,
+	0xc40007f1,
+	0xd00103f0,
+	0x04bd000b,
+/* 0x0197: mmctx_base_disabled */
+	0xfd0099f0,
+	0x0bf405ee,
+	0x0007f11e,
+	0x0103f0c6,
+	0xbd000ed0,
+	0x0007f104,
+	0x0103f0c7,
+	0xbd000fd0,
+	0x0199f004,
+/* 0x01b8: mmctx_multi_disabled */
+	0xb600abc8,
+	0xb9f010b4,
+	0x01aec80c,
+	0xfd11e4b6,
+	0x07f105be,
+	0x03f0c500,
+	0x000bd001,
+/* 0x01d6: mmctx_exec_loop */
+/* 0x01d6: mmctx_wait_free */
+	0xe7f104bd,
+	0xe3f0c500,
+	0x00eecf01,
+	0xf41fe4f0,
+	0xce98f30b,
+	0x05e9fd00,
+	0xc80007f1,
+	0xd00103f0,
+	0x04bd000e,
+	0xb804c0b6,
+	0x1bf404cd,
+	0x02abc8d8,
+/* 0x0207: mmctx_fini_wait */
+	0xf11f1bf4,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x1fb4f000,
+	0xf410b4b0,
+	0xa7f0f01b,
+	0xd021f402,
+/* 0x0223: mmctx_stop */
+	0xc82b0ef4,
 	0xb4b600ab,
 	0x0cb9f010,
-	0xb601aec8,
-	0xbefd11e4,
-	0x008bd005,
-/* 0x01b4: mmctx_exec_loop */
-/* 0x01b4: mmctx_wait_free */
-	0xf0008ecf,
-	0x0bf41fe4,
-	0x00ce98fa,
-	0xd005e9fd,
-	0xc0b6c08e,
-	0x04cdb804,
-	0xc8e81bf4,
-	0x1bf402ab,
-/* 0x01d5: mmctx_fini_wait */
-	0x008bcf18,
-	0xb01fb4f0,
-	0x1bf410b4,
-	0x02a7f0f7,
-	0xf4c921f4,
-/* 0x01ea: mmctx_stop */
-	0xabc81b0e,
-	0x10b4b600,
-	0xf00cb9f0,
-	0x8bd012b9,
-/* 0x01f9: mmctx_stop_wait */
-	0x008bcf00,
-	0xf412bbc8,
-/* 0x0202: mmctx_done */
-	0x94bdfa1b,
-	0xf10199f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0215: strand_wait */
-	0xf0a0f900,
-	0x21f402a7,
-	0xf8a0fcc9,
-/* 0x0221: strand_pre */
-	0xfc87f100,
-	0x0283f04a,
-	0xd00c97f0,
-	0x21f50089,
-	0x00f80215,
-/* 0x0234: strand_post */
-	0x4afc87f1,
-	0xf00283f0,
-	0x89d00d97,
-	0x1521f500,
-/* 0x0247: strand_set */
-	0xf100f802,
-	0xf04ffca7,
-	0xaba202a3,
-	0xc7f00500,
-	0x00acd00f,
-	0xd00bc7f0,
-	0x21f500bc,
-	0xaed00215,
-	0x0ac7f000,
-	0xf500bcd0,
-	0xf8021521,
-/* 0x0271: strand_ctx_init */
-	0xf094bd00,
-	0x07f10399,
-	0x03f00f00,
+	0xf112b9f0,
+	0xf0c50007,
+	0x0bd00103,
+/* 0x023b: mmctx_stop_wait */
+	0xf104bd00,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x12bbc800,
+/* 0x024b: mmctx_done */
+	0xbdf31bf4,
+	0x0199f094,
+	0x170007f1,
+	0xd00203f0,
+	0x04bd0009,
+/* 0x025e: strand_wait */
+	0xa0f900f8,
+	0xf402a7f0,
+	0xa0fcd021,
+/* 0x026a: strand_pre */
+	0x97f000f8,
+	0xfc07f10c,
+	0x0203f04a,
+	0xbd0009d0,
+	0x5e21f504,
+/* 0x027f: strand_post */
+	0xf000f802,
+	0x07f10d97,
+	0x03f04afc,
 	0x0009d002,
 	0x21f504bd,
-	0xe7f00221,
-	0x4721f503,
-	0xfca7f102,
-	0x02a3f046,
-	0x0400aba0,
-	0xf040a0d0,
-	0xbcd001c7,
-	0x1521f500,
-	0x010c9202,
-	0xf000acd0,
-	0xbcd002c7,
-	0x1521f500,
-	0x3421f502,
-	0x8087f102,
-	0x0684b608,
-	0xb70089cf,
-	0x95220080,
-/* 0x02ca: ctx_init_strand_loop */
+	0x00f8025e,
+/* 0x0294: strand_set */
+	0xf10fc7f0,
+	0xf04ffc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f10bc7,
+	0x03f04afc,
+	0x000cd002,
+	0x07f104bd,
+	0x03f04ffc,
+	0x000ed002,
+	0xc7f004bd,
+	0xfc07f10a,
+	0x0203f04a,
+	0xbd000cd0,
+	0x5e21f504,
+/* 0x02d3: strand_ctx_init */
+	0xbd00f802,
+	0x0399f094,
+	0x0f0007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0x026a21f5,
+	0xf503e7f0,
+	0xbd029421,
+	0xfc07f1c4,
+	0x0203f047,
+	0xbd000cd0,
+	0x01c7f004,
+	0x4afc07f1,
+	0xd00203f0,
+	0x04bd000c,
+	0x025e21f5,
+	0xf1010c92,
+	0xf046fc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f102c7,
+	0x03f04afc,
+	0x000cd002,
+	0x21f504bd,
+	0x21f5025e,
+	0x87f1027f,
+	0x83f04200,
+	0x0097f102,
+	0x0293f020,
+	0x950099cf,
+/* 0x034a: ctx_init_strand_loop */
 	0x8ed008fe,
 	0x408ed000,
 	0xb6808acf,
@@ -259,167 +289,199 @@ uint32_t nvc0_grgpc_code[] = {
 	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-/* 0x02fe: error */
+/* 0x037e: error */
 	0xe0f900f8,
-	0x9814e7f1,
-	0xf440e3f0,
-	0xe0b78d21,
-	0xf7f0041c,
-	0x8d21f401,
-	0x00f8e0fc,
-/* 0x0318: init */
-	0x04fe04bd,
-	0x0017f100,
-	0x0227f012,
-	0xf10012d0,
-	0xfe042617,
-	0x17f10010,
-	0x10d00400,
-	0x0427f0c0,
-	0xf40012d0,
-	0x17f11031,
-	0x14b60608,
-	0x0012cf06,
+	0xf102ffb9,
+	0xf09814e7,
+	0x21f440e3,
+	0x01f7f09d,
+	0xf102ffb9,
+	0xf09c1ce7,
+	0x21f440e3,
+	0xf8e0fc9d,
+/* 0x03a1: init */
+	0xfe04bd00,
+	0x27f00004,
+	0x0007f102,
+	0x0003f012,
+	0xbd0002d0,
+	0xd517f104,
+	0x0010fe04,
+	0x070007f1,
+	0xd00003f0,
+	0x04bd0000,
+	0xf10427f0,
+	0xf0040007,
+	0x02d00003,
+	0xf404bd00,
+	0x27f11031,
+	0x23f08200,
+	0x0022cf01,
 	0xf00137f0,
 	0x32bb1f24,
 	0x0132b604,
 	0x80050280,
-	0x10b70603,
-	0x12cf0400,
-	0x04028000,
-	0x010027f1,
-	0xcf0223f0,
-	0x34bd0022,
-	0x070047f1,
-	0x950644b6,
-	0x45d00825,
-	0x4045d000,
-	0x98000e98,
-	0x21f5010f,
-	0x2fbb0147,
-	0x003fbb00,
-	0x98010e98,
-	0x21f5020f,
-	0x0e980147,
-	0x00effd05,
-	0xbb002ebb,
-	0x40b7003e,
-	0x35b61300,
-	0x0043d002,
-	0xb60825b6,
-	0x20b60635,
-	0x0130b601,
-	0xb60824b6,
-	0x2fb90834,
-	0x7121f502,
-	0x003fbb02,
-	0x010007f1,
+	0x27f10603,
+	0x23f08600,
+	0x0022cf01,
+	0xf1040280,
+	0xf0010027,
+	0x22cf0223,
+	0x9534bd00,
+	0x07f10825,
+	0x03f0c000,
+	0x0005d001,
+	0x07f104bd,
+	0x03f0c100,
+	0x0005d001,
+	0x0e9804bd,
+	0x010f9800,
+	0x015021f5,
+	0xbb002fbb,
+	0x0e98003f,
+	0x020f9801,
+	0x015021f5,
+	0xfd050e98,
+	0x2ebb00ef,
+	0x003ebb00,
+	0xf10235b6,
+	0xf0d30007,
+	0x03d00103,
+	0xb604bd00,
+	0x35b60825,
+	0x0120b606,
+	0xb60130b6,
+	0x34b60824,
+	0x022fb908,
+	0x02d321f5,
+	0xf1003fbb,
+	0xf0010007,
+	0x03d00203,
+	0xbd04bd00,
+	0x1f29f024,
+	0x080007f1,
 	0xd00203f0,
-	0x04bd0003,
-	0x29f024bd,
-	0x0007f11f,
-	0x0203f008,
-	0xbd0002d0,
-/* 0x03e9: main */
-	0x0031f404,
-	0xf00028f4,
-	0x21f41cd7,
-	0xf401f439,
-	0xf404e4b0,
-	0x81fe1e18,
-	0x0627f001,
-	0x12fd20bd,
-	0x01e4b604,
-	0xfe051efd,
-	0x21f50018,
-	0x0ef404ad,
-/* 0x0419: main_not_ctx_xfer */
-	0x10ef94d3,
-	0xf501f5f0,
-	0xf402fe21,
-/* 0x0426: ih */
-	0x80f9c60e,
-	0xf90188fe,
-	0xf990f980,
-	0xf9b0f9a0,
-	0xf9e0f9d0,
-	0xcf04bdf0,
-	0xabc4800a,
-	0x1d0bf404,
-	0x1900b7f1,
-	0xcf1cd7f0,
-	0xbfcf40be,
+	0x04bd0002,
+/* 0x0498: main */
+	0xf40031f4,
+	0xd7f00028,
+	0x3921f41c,
+	0xb0f401f4,
+	0x18f404e4,
+	0x0181fe1e,
+	0xbd0627f0,
+	0x0412fd20,
+	0xfd01e4b6,
+	0x18fe051e,
+	0x8d21f500,
+	0xd30ef405,
+/* 0x04c8: main_not_ctx_xfer */
+	0xf010ef94,
+	0x21f501f5,
+	0x0ef4037e,
+/* 0x04d5: ih */
+	0xfe80f9c6,
+	0x80f90188,
+	0xa0f990f9,
+	0xd0f9b0f9,
+	0xf0f9e0f9,
+	0xa7f104bd,
+	0xa3f00200,
+	0x00aacf00,
+	0xf404abc4,
+	0xd7f02c0b,
+	0x00e7f11c,
+	0x00e3f01a,
+	0xf100eecf,
+	0xf01900f7,
+	0xffcf00f3,
 	0x0421f400,
-	0x0400b0b7,
-	0xd001e7f0,
-/* 0x045e: ih_no_fifo */
-	0x0ad000be,
-	0xfcf0fc40,
-	0xfcd0fce0,
-	0xfca0fcb0,
-	0xfe80fc90,
-	0x80fc0088,
-	0xf80032f4,
-/* 0x0479: hub_barrier_done */
-	0x01f7f001,
-	0xbb040e98,
-	0xe7f104fe,
-	0xe3f09418,
-	0x8d21f440,
-/* 0x048e: ctx_redswitch */
-	0xe7f100f8,
-	0xe4b60614,
-	0x20f7f006,
-	0xf000efd0,
-/* 0x049e: ctx_redswitch_delay */
-	0xf2b608f7,
-	0xfd1bf401,
-	0x0a20f7f1,
-	0xf800efd0,
-/* 0x04ad: ctx_xfer */
-	0x0417f100,
-	0x0614b60a,
-	0xf4001fd0,
-	0x21f50711,
-/* 0x04be: ctx_xfer_not_load */
-	0x17f1048e,
-	0x13f04afc,
-	0x0c27f002,
-	0xf50012d0,
-	0xf1021521,
-	0xf047fc27,
-	0x20d00223,
-	0x012cf000,
-	0xd00320b6,
-	0xacf00012,
-	0x02a5f001,
-	0xf000b7f0,
-	0x0c9850b3,
-	0x0fc4b604,
-	0x9800bcbb,
-	0x0d98000c,
-	0x00e7f001,
-	0x016621f5,
+	0xf101e7f0,
+	0xf01d0007,
+	0x0ed00003,
+/* 0x0523: ih_no_fifo */
+	0xf104bd00,
+	0xf0010007,
+	0x0ad00003,
+	0xfc04bd00,
+	0xfce0fcf0,
+	0xfcb0fcd0,
+	0xfc90fca0,
+	0x0088fe80,
+	0x32f480fc,
+/* 0x0547: hub_barrier_done */
+	0xf001f800,
+	0x0e9801f7,
+	0x04febb04,
+	0xf102ffb9,
+	0xf09418e7,
+	0x21f440e3,
+/* 0x055f: ctx_redswitch */
+	0xf000f89d,
+	0x07f120f7,
+	0x03f08500,
+	0x000fd001,
+	0xe7f004bd,
+/* 0x0571: ctx_redswitch_delay */
+	0x01e2b608,
+	0xf1fd1bf4,
+	0xf10800f5,
+	0xf10200f5,
+	0xf0850007,
+	0x0fd00103,
+	0xf804bd00,
+/* 0x058d: ctx_xfer */
+	0x0007f100,
+	0x0203f081,
+	0xbd000fd0,
+	0x0711f404,
+	0x055f21f5,
+/* 0x05a0: ctx_xfer_not_load */
+	0x026a21f5,
+	0x07f124bd,
+	0x03f047fc,
+	0x0002d002,
+	0x2cf004bd,
+	0x0320b601,
+	0x4afc07f1,
+	0xd00203f0,
+	0x04bd0002,
 	0xf001acf0,
-	0xb7f104a5,
-	0xb3f04000,
+	0xb7f102a5,
+	0xb3f00000,
 	0x040c9850,
 	0xbb0fc4b6,
 	0x0c9800bc,
-	0x020d9801,
-	0xf1060f98,
-	0xf50800e7,
-	0xf5016621,
-	0xf4021521,
-	0x12f40601,
-/* 0x0535: ctx_xfer_post */
-	0xfc17f114,
-	0x0213f04a,
-	0xd00d27f0,
-	0x21f50012,
-/* 0x0546: ctx_xfer_done */
-	0x21f50215,
-	0x00f80479,
+	0x010d9800,
+	0xf500e7f0,
+	0xf0016f21,
+	0xa5f001ac,
+	0x00b7f104,
+	0x50b3f040,
+	0xb6040c98,
+	0xbcbb0fc4,
+	0x010c9800,
+	0x98020d98,
+	0xe7f1060f,
+	0x21f50800,
+	0x21f5016f,
+	0x01f4025e,
+	0x0712f406,
+/* 0x0618: ctx_xfer_post */
+	0x027f21f5,
+/* 0x061c: ctx_xfer_done */
+	0x054721f5,
+	0x000000f8,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnvd7.fuc.h b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnvd7.fuc.h
index dd346c2a1624..84dd32db28a0 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnvd7.fuc.h
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnvd7.fuc.h
@@ -41,14 +41,14 @@ uint32_t nvd7_grgpc_data[] = {
 };
 
 uint32_t nvd7_grgpc_code[] = {
-	0x03180ef5,
+	0x03a10ef5,
 /* 0x0004: queue_put */
 	0x9800d898,
 	0x86f001d9,
 	0x0489b808,
 	0xf00c1bf4,
 	0x21f502f7,
-	0x00f802fe,
+	0x00f8037e,
 /* 0x001c: queue_put_next */
 	0xb60798c4,
 	0x8dbb0384,
@@ -72,184 +72,214 @@ uint32_t nvd7_grgpc_code[] = {
 /* 0x0066: queue_get_done */
 	0x00f80132,
 /* 0x0068: nv_rd32 */
-	0x0728b7f1,
-	0xb906b4b6,
-	0xc9f002ec,
-	0x00bcd01f,
-/* 0x0078: nv_rd32_wait */
-	0xc800bccf,
-	0x1bf41fcc,
-	0x06a7f0fa,
-	0x010921f5,
-	0xf840bfcf,
-/* 0x008d: nv_wr32 */
-	0x28b7f100,
-	0x06b4b607,
-	0xb980bfd0,
-	0xc9f002ec,
-	0x1ec9f01f,
-/* 0x00a3: nv_wr32_wait */
-	0xcf00bcd0,
-	0xccc800bc,
-	0xfa1bf41f,
-/* 0x00ae: watchdog_reset */
-	0x87f100f8,
-	0x84b60430,
-	0x1ff9f006,
-	0xf8008fd0,
-/* 0x00bd: watchdog_clear */
-	0x3087f100,
-	0x0684b604,
-	0xf80080d0,
-/* 0x00c9: wait_donez */
-	0xf094bd00,
-	0x07f10099,
-	0x03f00f00,
-	0x0009d002,
-	0x07f104bd,
-	0x03f00600,
-	0x000ad002,
-/* 0x00e6: wait_donez_ne */
-	0x87f104bd,
-	0x83f00000,
-	0x0088cf01,
-	0xf4888aff,
-	0x94bdf31b,
-	0xf10099f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0109: wait_doneo */
-	0xf094bd00,
+	0xf002ecb9,
+	0x07f11fc9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x007a: nv_rd32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0xa7f0f31b,
+	0x1021f506,
+	0x00f7f101,
+	0x01f3f0cb,
+	0xf800ffcf,
+/* 0x009d: nv_wr32 */
+	0x0007f100,
+	0x0103f0cc,
+	0xbd000fd0,
+	0x02ecb904,
+	0xf01fc9f0,
+	0x07f11ec9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x00be: nv_wr32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0x00f8f31b,
+/* 0x00d0: wait_donez */
+	0x99f094bd,
+	0x0007f100,
+	0x0203f00f,
+	0xbd0009d0,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x00ed: wait_donez_ne */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x1bf4888a,
+	0xf094bdf3,
 	0x07f10099,
-	0x03f00f00,
+	0x03f01700,
 	0x0009d002,
-	0x87f104bd,
-	0x84b60818,
-	0x008ad006,
-/* 0x0124: wait_doneo_e */
-	0x040087f1,
-	0xcf0684b6,
-	0x8aff0088,
-	0xf30bf488,
+	0x00f804bd,
+/* 0x0110: wait_doneo */
 	0x99f094bd,
 	0x0007f100,
-	0x0203f017,
+	0x0203f00f,
 	0xbd0009d0,
-/* 0x0147: mmctx_size */
-	0xbd00f804,
-/* 0x0149: nv_mmctx_size_loop */
-	0x00e89894,
-	0xb61a85b6,
-	0x84b60180,
-	0x0098bb02,
-	0xb804e0b6,
-	0x1bf404ef,
-	0x029fb9eb,
-/* 0x0166: mmctx_xfer */
-	0x94bd00f8,
-	0xf10199f0,
-	0xf00f0007,
-	0x09d00203,
-	0xf104bd00,
-	0xb6071087,
-	0x94bd0684,
-	0xf405bbfd,
-	0x8bd0090b,
-	0x0099f000,
-/* 0x018c: mmctx_base_disabled */
-	0xf405eefd,
-	0x8ed00c0b,
-	0xc08fd080,
-/* 0x019b: mmctx_multi_disabled */
-	0xb70199f0,
-	0xc8010080,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x012d: wait_doneo_e */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x0bf4888a,
+	0xf094bdf3,
+	0x07f10099,
+	0x03f01700,
+	0x0009d002,
+	0x00f804bd,
+/* 0x0150: mmctx_size */
+/* 0x0152: nv_mmctx_size_loop */
+	0xe89894bd,
+	0x1a85b600,
+	0xb60180b6,
+	0x98bb0284,
+	0x04e0b600,
+	0xf404efb8,
+	0x9fb9eb1b,
+/* 0x016f: mmctx_xfer */
+	0xbd00f802,
+	0x0199f094,
+	0x0f0007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0xbbfd94bd,
+	0x120bf405,
+	0xc40007f1,
+	0xd00103f0,
+	0x04bd000b,
+/* 0x0197: mmctx_base_disabled */
+	0xfd0099f0,
+	0x0bf405ee,
+	0x0007f11e,
+	0x0103f0c6,
+	0xbd000ed0,
+	0x0007f104,
+	0x0103f0c7,
+	0xbd000fd0,
+	0x0199f004,
+/* 0x01b8: mmctx_multi_disabled */
+	0xb600abc8,
+	0xb9f010b4,
+	0x01aec80c,
+	0xfd11e4b6,
+	0x07f105be,
+	0x03f0c500,
+	0x000bd001,
+/* 0x01d6: mmctx_exec_loop */
+/* 0x01d6: mmctx_wait_free */
+	0xe7f104bd,
+	0xe3f0c500,
+	0x00eecf01,
+	0xf41fe4f0,
+	0xce98f30b,
+	0x05e9fd00,
+	0xc80007f1,
+	0xd00103f0,
+	0x04bd000e,
+	0xb804c0b6,
+	0x1bf404cd,
+	0x02abc8d8,
+/* 0x0207: mmctx_fini_wait */
+	0xf11f1bf4,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x1fb4f000,
+	0xf410b4b0,
+	0xa7f0f01b,
+	0xd021f402,
+/* 0x0223: mmctx_stop */
+	0xc82b0ef4,
 	0xb4b600ab,
 	0x0cb9f010,
-	0xb601aec8,
-	0xbefd11e4,
-	0x008bd005,
-/* 0x01b4: mmctx_exec_loop */
-/* 0x01b4: mmctx_wait_free */
-	0xf0008ecf,
-	0x0bf41fe4,
-	0x00ce98fa,
-	0xd005e9fd,
-	0xc0b6c08e,
-	0x04cdb804,
-	0xc8e81bf4,
-	0x1bf402ab,
-/* 0x01d5: mmctx_fini_wait */
-	0x008bcf18,
-	0xb01fb4f0,
-	0x1bf410b4,
-	0x02a7f0f7,
-	0xf4c921f4,
-/* 0x01ea: mmctx_stop */
-	0xabc81b0e,
-	0x10b4b600,
-	0xf00cb9f0,
-	0x8bd012b9,
-/* 0x01f9: mmctx_stop_wait */
-	0x008bcf00,
-	0xf412bbc8,
-/* 0x0202: mmctx_done */
-	0x94bdfa1b,
-	0xf10199f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0215: strand_wait */
-	0xf0a0f900,
-	0x21f402a7,
-	0xf8a0fcc9,
-/* 0x0221: strand_pre */
-	0xfc87f100,
-	0x0283f04a,
-	0xd00c97f0,
-	0x21f50089,
-	0x00f80215,
-/* 0x0234: strand_post */
-	0x4afc87f1,
-	0xf00283f0,
-	0x89d00d97,
-	0x1521f500,
-/* 0x0247: strand_set */
-	0xf100f802,
-	0xf04ffca7,
-	0xaba202a3,
-	0xc7f00500,
-	0x00acd00f,
-	0xd00bc7f0,
-	0x21f500bc,
-	0xaed00215,
-	0x0ac7f000,
-	0xf500bcd0,
-	0xf8021521,
-/* 0x0271: strand_ctx_init */
-	0xf094bd00,
-	0x07f10399,
-	0x03f00f00,
+	0xf112b9f0,
+	0xf0c50007,
+	0x0bd00103,
+/* 0x023b: mmctx_stop_wait */
+	0xf104bd00,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x12bbc800,
+/* 0x024b: mmctx_done */
+	0xbdf31bf4,
+	0x0199f094,
+	0x170007f1,
+	0xd00203f0,
+	0x04bd0009,
+/* 0x025e: strand_wait */
+	0xa0f900f8,
+	0xf402a7f0,
+	0xa0fcd021,
+/* 0x026a: strand_pre */
+	0x97f000f8,
+	0xfc07f10c,
+	0x0203f04a,
+	0xbd0009d0,
+	0x5e21f504,
+/* 0x027f: strand_post */
+	0xf000f802,
+	0x07f10d97,
+	0x03f04afc,
 	0x0009d002,
 	0x21f504bd,
-	0xe7f00221,
-	0x4721f503,
-	0xfca7f102,
-	0x02a3f046,
-	0x0400aba0,
-	0xf040a0d0,
-	0xbcd001c7,
-	0x1521f500,
-	0x010c9202,
-	0xf000acd0,
-	0xbcd002c7,
-	0x1521f500,
-	0x3421f502,
-	0x8087f102,
-	0x0684b608,
-	0xb70089cf,
-	0x95220080,
-/* 0x02ca: ctx_init_strand_loop */
+	0x00f8025e,
+/* 0x0294: strand_set */
+	0xf10fc7f0,
+	0xf04ffc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f10bc7,
+	0x03f04afc,
+	0x000cd002,
+	0x07f104bd,
+	0x03f04ffc,
+	0x000ed002,
+	0xc7f004bd,
+	0xfc07f10a,
+	0x0203f04a,
+	0xbd000cd0,
+	0x5e21f504,
+/* 0x02d3: strand_ctx_init */
+	0xbd00f802,
+	0x0399f094,
+	0x0f0007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0x026a21f5,
+	0xf503e7f0,
+	0xbd029421,
+	0xfc07f1c4,
+	0x0203f047,
+	0xbd000cd0,
+	0x01c7f004,
+	0x4afc07f1,
+	0xd00203f0,
+	0x04bd000c,
+	0x025e21f5,
+	0xf1010c92,
+	0xf046fc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f102c7,
+	0x03f04afc,
+	0x000cd002,
+	0x21f504bd,
+	0x21f5025e,
+	0x87f1027f,
+	0x83f04200,
+	0x0097f102,
+	0x0293f020,
+	0x950099cf,
+/* 0x034a: ctx_init_strand_loop */
 	0x8ed008fe,
 	0x408ed000,
 	0xb6808acf,
@@ -263,198 +293,230 @@ uint32_t nvd7_grgpc_code[] = {
 	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-/* 0x02fe: error */
+/* 0x037e: error */
 	0xe0f900f8,
-	0x9814e7f1,
-	0xf440e3f0,
-	0xe0b78d21,
-	0xf7f0041c,
-	0x8d21f401,
-	0x00f8e0fc,
-/* 0x0318: init */
-	0x04fe04bd,
-	0x0017f100,
-	0x0227f012,
-	0xf10012d0,
-	0xfe047017,
-	0x17f10010,
-	0x10d00400,
-	0x0427f0c0,
-	0xf40012d0,
-	0x17f11031,
-	0x14b60608,
-	0x0012cf06,
+	0xf102ffb9,
+	0xf09814e7,
+	0x21f440e3,
+	0x01f7f09d,
+	0xf102ffb9,
+	0xf09c1ce7,
+	0x21f440e3,
+	0xf8e0fc9d,
+/* 0x03a1: init */
+	0xfe04bd00,
+	0x27f00004,
+	0x0007f102,
+	0x0003f012,
+	0xbd0002d0,
+	0x1f17f104,
+	0x0010fe05,
+	0x070007f1,
+	0xd00003f0,
+	0x04bd0000,
+	0xf10427f0,
+	0xf0040007,
+	0x02d00003,
+	0xf404bd00,
+	0x27f11031,
+	0x23f08200,
+	0x0022cf01,
 	0xf00137f0,
 	0x32bb1f24,
 	0x0132b604,
 	0x80050280,
-	0x10b70603,
-	0x12cf0400,
-	0x04028000,
-	0x0c30e7f1,
-	0xbd50e3f0,
-	0xbd34bd24,
-/* 0x0371: init_unk_loop */
-	0x6821f444,
-	0xf400f6b0,
-	0xf7f00f0b,
-	0x04f2bb01,
-	0xb6054ffd,
-/* 0x0386: init_unk_next */
-	0x20b60130,
-	0x04e0b601,
-	0xf40126b0,
-/* 0x0392: init_unk_done */
-	0x0380e21b,
-	0x08048007,
-	0x010027f1,
-	0xcf0223f0,
-	0x34bd0022,
-	0x070047f1,
-	0x950644b6,
-	0x45d00825,
-	0x4045d000,
-	0x98000e98,
-	0x21f5010f,
-	0x2fbb0147,
-	0x003fbb00,
-	0x98010e98,
-	0x21f5020f,
-	0x0e980147,
-	0x00effd05,
-	0xbb002ebb,
-	0x0e98003e,
-	0x030f9802,
-	0x014721f5,
-	0xfd070e98,
+	0x27f10603,
+	0x23f08600,
+	0x0022cf01,
+	0xf1040280,
+	0xf00c30e7,
+	0x24bd50e3,
+	0x44bd34bd,
+/* 0x0410: init_unk_loop */
+	0xb06821f4,
+	0x0bf400f6,
+	0x01f7f00f,
+	0xfd04f2bb,
+	0x30b6054f,
+/* 0x0425: init_unk_next */
+	0x0120b601,
+	0xb004e0b6,
+	0x1bf40126,
+/* 0x0431: init_unk_done */
+	0x070380e2,
+	0xf1080480,
+	0xf0010027,
+	0x22cf0223,
+	0x9534bd00,
+	0x07f10825,
+	0x03f0c000,
+	0x0005d001,
+	0x07f104bd,
+	0x03f0c100,
+	0x0005d001,
+	0x0e9804bd,
+	0x010f9800,
+	0x015021f5,
+	0xbb002fbb,
+	0x0e98003f,
+	0x020f9801,
+	0x015021f5,
+	0xfd050e98,
 	0x2ebb00ef,
 	0x003ebb00,
-	0x130040b7,
-	0xd00235b6,
-	0x25b60043,
-	0x0635b608,
-	0xb60120b6,
-	0x24b60130,
-	0x0834b608,
-	0xf5022fb9,
-	0xbb027121,
-	0x07f1003f,
-	0x03f00100,
-	0x0003d002,
-	0x24bd04bd,
-	0xf11f29f0,
-	0xf0080007,
-	0x02d00203,
-/* 0x0433: main */
+	0x98020e98,
+	0x21f5030f,
+	0x0e980150,
+	0x00effd07,
+	0xbb002ebb,
+	0x35b6003e,
+	0x0007f102,
+	0x0103f0d3,
+	0xbd0003d0,
+	0x0825b604,
+	0xb60635b6,
+	0x30b60120,
+	0x0824b601,
+	0xb90834b6,
+	0x21f5022f,
+	0x3fbb02d3,
+	0x0007f100,
+	0x0203f001,
+	0xbd0003d0,
+	0xf024bd04,
+	0x07f11f29,
+	0x03f00800,
+	0x0002d002,
+/* 0x04e2: main */
+	0x31f404bd,
+	0x0028f400,
+	0xf424d7f0,
+	0x01f43921,
+	0x04e4b0f4,
+	0xfe1e18f4,
+	0x27f00181,
+	0xfd20bd06,
+	0xe4b60412,
+	0x051efd01,
+	0xf50018fe,
+	0xf405d721,
+/* 0x0512: main_not_ctx_xfer */
+	0xef94d30e,
+	0x01f5f010,
+	0x037e21f5,
+/* 0x051f: ih */
+	0xf9c60ef4,
+	0x0188fe80,
+	0x90f980f9,
+	0xb0f9a0f9,
+	0xe0f9d0f9,
+	0x04bdf0f9,
+	0x0200a7f1,
+	0xcf00a3f0,
+	0xabc400aa,
+	0x2c0bf404,
+	0xf124d7f0,
+	0xf01a00e7,
+	0xeecf00e3,
+	0x00f7f100,
+	0x00f3f019,
+	0xf400ffcf,
+	0xe7f00421,
+	0x0007f101,
+	0x0003f01d,
+	0xbd000ed0,
+/* 0x056d: ih_no_fifo */
+	0x0007f104,
+	0x0003f001,
+	0xbd000ad0,
+	0xfcf0fc04,
+	0xfcd0fce0,
+	0xfca0fcb0,
+	0xfe80fc90,
+	0x80fc0088,
+	0xf80032f4,
+/* 0x0591: hub_barrier_done */
+	0x01f7f001,
+	0xbb040e98,
+	0xffb904fe,
+	0x18e7f102,
+	0x40e3f094,
+	0xf89d21f4,
+/* 0x05a9: ctx_redswitch */
+	0x20f7f000,
+	0x850007f1,
+	0xd00103f0,
+	0x04bd000f,
+/* 0x05bb: ctx_redswitch_delay */
+	0xb608e7f0,
+	0x1bf401e2,
+	0x00f5f1fd,
+	0x00f5f108,
+	0x0007f102,
+	0x0103f085,
+	0xbd000fd0,
+/* 0x05d7: ctx_xfer */
+	0xf100f804,
+	0xf0810007,
+	0x0fd00203,
 	0xf404bd00,
-	0x28f40031,
-	0x24d7f000,
-	0xf43921f4,
-	0xe4b0f401,
-	0x1e18f404,
-	0xf00181fe,
-	0x20bd0627,
-	0xb60412fd,
-	0x1efd01e4,
-	0x0018fe05,
-	0x04f721f5,
-/* 0x0463: main_not_ctx_xfer */
-	0x94d30ef4,
-	0xf5f010ef,
-	0xfe21f501,
-	0xc60ef402,
-/* 0x0470: ih */
-	0x88fe80f9,
-	0xf980f901,
-	0xf9a0f990,
-	0xf9d0f9b0,
-	0xbdf0f9e0,
-	0x800acf04,
-	0xf404abc4,
-	0xb7f11d0b,
-	0xd7f01900,
-	0x40becf24,
-	0xf400bfcf,
-	0xb0b70421,
-	0xe7f00400,
-	0x00bed001,
-/* 0x04a8: ih_no_fifo */
-	0xfc400ad0,
-	0xfce0fcf0,
-	0xfcb0fcd0,
-	0xfc90fca0,
-	0x0088fe80,
-	0x32f480fc,
-/* 0x04c3: hub_barrier_done */
-	0xf001f800,
-	0x0e9801f7,
-	0x04febb04,
-	0x9418e7f1,
-	0xf440e3f0,
-	0x00f88d21,
-/* 0x04d8: ctx_redswitch */
-	0x0614e7f1,
-	0xf006e4b6,
-	0xefd020f7,
-	0x08f7f000,
-/* 0x04e8: ctx_redswitch_delay */
-	0xf401f2b6,
-	0xf7f1fd1b,
-	0xefd00a20,
-/* 0x04f7: ctx_xfer */
-	0xf100f800,
-	0xb60a0417,
-	0x1fd00614,
-	0x0711f400,
-	0x04d821f5,
-/* 0x0508: ctx_xfer_not_load */
-	0x4afc17f1,
-	0xf00213f0,
-	0x12d00c27,
-	0x1521f500,
-	0xfc27f102,
-	0x0223f047,
-	0xf00020d0,
-	0x20b6012c,
-	0x0012d003,
-	0xf001acf0,
-	0xb7f002a5,
-	0x50b3f000,
-	0xb6040c98,
-	0xbcbb0fc4,
-	0x000c9800,
-	0xf0010d98,
-	0x21f500e7,
-	0xacf00166,
-	0x00b7f101,
-	0x50b3f040,
-	0xb6040c98,
-	0xbcbb0fc4,
-	0x010c9800,
-	0x98020d98,
-	0xe7f1060f,
-	0x21f50800,
-	0xacf00166,
-	0x04a5f001,
-	0x3000b7f1,
+	0x21f50711,
+/* 0x05ea: ctx_xfer_not_load */
+	0x21f505a9,
+	0x24bd026a,
+	0x47fc07f1,
+	0xd00203f0,
+	0x04bd0002,
+	0xb6012cf0,
+	0x07f10320,
+	0x03f04afc,
+	0x0002d002,
+	0xacf004bd,
+	0x02a5f001,
+	0x0000b7f1,
 	0x9850b3f0,
 	0xc4b6040c,
 	0x00bcbb0f,
-	0x98020c98,
-	0x0f98030d,
-	0x00e7f108,
-	0x6621f502,
-	0x1521f501,
-	0x0601f402,
-/* 0x05a3: ctx_xfer_post */
-	0xf11412f4,
-	0xf04afc17,
-	0x27f00213,
-	0x0012d00d,
-	0x021521f5,
-/* 0x05b4: ctx_xfer_done */
-	0x04c321f5,
-	0x000000f8,
+	0x98000c98,
+	0xe7f0010d,
+	0x6f21f500,
+	0x01acf001,
+	0x4000b7f1,
+	0x9850b3f0,
+	0xc4b6040c,
+	0x00bcbb0f,
+	0x98010c98,
+	0x0f98020d,
+	0x00e7f106,
+	0x6f21f508,
+	0x01acf001,
+	0xf104a5f0,
+	0xf03000b7,
+	0x0c9850b3,
+	0x0fc4b604,
+	0x9800bcbb,
+	0x0d98020c,
+	0x080f9803,
+	0x0200e7f1,
+	0x016f21f5,
+	0x025e21f5,
+	0xf40601f4,
+/* 0x0686: ctx_xfer_post */
+	0x21f50712,
+/* 0x068a: ctx_xfer_done */
+	0x21f5027f,
+	0x00f80591,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnve0.fuc.h b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnve0.fuc.h
index 7ff5ef6b0804..b6da800ee9c2 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnve0.fuc.h
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnve0.fuc.h
@@ -41,14 +41,14 @@ uint32_t nve0_grgpc_data[] = {
 };
 
 uint32_t nve0_grgpc_code[] = {
-	0x03180ef5,
+	0x03a10ef5,
 /* 0x0004: queue_put */
 	0x9800d898,
 	0x86f001d9,
 	0x0489b808,
 	0xf00c1bf4,
 	0x21f502f7,
-	0x00f802fe,
+	0x00f8037e,
 /* 0x001c: queue_put_next */
 	0xb60798c4,
 	0x8dbb0384,
@@ -72,184 +72,214 @@ uint32_t nve0_grgpc_code[] = {
 /* 0x0066: queue_get_done */
 	0x00f80132,
 /* 0x0068: nv_rd32 */
-	0x0728b7f1,
-	0xb906b4b6,
-	0xc9f002ec,
-	0x00bcd01f,
-/* 0x0078: nv_rd32_wait */
-	0xc800bccf,
-	0x1bf41fcc,
-	0x06a7f0fa,
-	0x010921f5,
-	0xf840bfcf,
-/* 0x008d: nv_wr32 */
-	0x28b7f100,
-	0x06b4b607,
-	0xb980bfd0,
-	0xc9f002ec,
-	0x1ec9f01f,
-/* 0x00a3: nv_wr32_wait */
-	0xcf00bcd0,
-	0xccc800bc,
-	0xfa1bf41f,
-/* 0x00ae: watchdog_reset */
-	0x87f100f8,
-	0x84b60430,
-	0x1ff9f006,
-	0xf8008fd0,
-/* 0x00bd: watchdog_clear */
-	0x3087f100,
-	0x0684b604,
-	0xf80080d0,
-/* 0x00c9: wait_donez */
-	0xf094bd00,
-	0x07f10099,
-	0x03f00f00,
-	0x0009d002,
-	0x07f104bd,
-	0x03f00600,
-	0x000ad002,
-/* 0x00e6: wait_donez_ne */
-	0x87f104bd,
-	0x83f00000,
-	0x0088cf01,
-	0xf4888aff,
-	0x94bdf31b,
-	0xf10099f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0109: wait_doneo */
-	0xf094bd00,
+	0xf002ecb9,
+	0x07f11fc9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x007a: nv_rd32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0xa7f0f31b,
+	0x1021f506,
+	0x00f7f101,
+	0x01f3f0cb,
+	0xf800ffcf,
+/* 0x009d: nv_wr32 */
+	0x0007f100,
+	0x0103f0cc,
+	0xbd000fd0,
+	0x02ecb904,
+	0xf01fc9f0,
+	0x07f11ec9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x00be: nv_wr32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0x00f8f31b,
+/* 0x00d0: wait_donez */
+	0x99f094bd,
+	0x0007f100,
+	0x0203f00f,
+	0xbd0009d0,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x00ed: wait_donez_ne */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x1bf4888a,
+	0xf094bdf3,
 	0x07f10099,
-	0x03f00f00,
+	0x03f01700,
 	0x0009d002,
-	0x87f104bd,
-	0x84b60818,
-	0x008ad006,
-/* 0x0124: wait_doneo_e */
-	0x040087f1,
-	0xcf0684b6,
-	0x8aff0088,
-	0xf30bf488,
+	0x00f804bd,
+/* 0x0110: wait_doneo */
 	0x99f094bd,
 	0x0007f100,
-	0x0203f017,
+	0x0203f00f,
 	0xbd0009d0,
-/* 0x0147: mmctx_size */
-	0xbd00f804,
-/* 0x0149: nv_mmctx_size_loop */
-	0x00e89894,
-	0xb61a85b6,
-	0x84b60180,
-	0x0098bb02,
-	0xb804e0b6,
-	0x1bf404ef,
-	0x029fb9eb,
-/* 0x0166: mmctx_xfer */
-	0x94bd00f8,
-	0xf10199f0,
-	0xf00f0007,
-	0x09d00203,
-	0xf104bd00,
-	0xb6071087,
-	0x94bd0684,
-	0xf405bbfd,
-	0x8bd0090b,
-	0x0099f000,
-/* 0x018c: mmctx_base_disabled */
-	0xf405eefd,
-	0x8ed00c0b,
-	0xc08fd080,
-/* 0x019b: mmctx_multi_disabled */
-	0xb70199f0,
-	0xc8010080,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x012d: wait_doneo_e */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x0bf4888a,
+	0xf094bdf3,
+	0x07f10099,
+	0x03f01700,
+	0x0009d002,
+	0x00f804bd,
+/* 0x0150: mmctx_size */
+/* 0x0152: nv_mmctx_size_loop */
+	0xe89894bd,
+	0x1a85b600,
+	0xb60180b6,
+	0x98bb0284,
+	0x04e0b600,
+	0xf404efb8,
+	0x9fb9eb1b,
+/* 0x016f: mmctx_xfer */
+	0xbd00f802,
+	0x0199f094,
+	0x0f0007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0xbbfd94bd,
+	0x120bf405,
+	0xc40007f1,
+	0xd00103f0,
+	0x04bd000b,
+/* 0x0197: mmctx_base_disabled */
+	0xfd0099f0,
+	0x0bf405ee,
+	0x0007f11e,
+	0x0103f0c6,
+	0xbd000ed0,
+	0x0007f104,
+	0x0103f0c7,
+	0xbd000fd0,
+	0x0199f004,
+/* 0x01b8: mmctx_multi_disabled */
+	0xb600abc8,
+	0xb9f010b4,
+	0x01aec80c,
+	0xfd11e4b6,
+	0x07f105be,
+	0x03f0c500,
+	0x000bd001,
+/* 0x01d6: mmctx_exec_loop */
+/* 0x01d6: mmctx_wait_free */
+	0xe7f104bd,
+	0xe3f0c500,
+	0x00eecf01,
+	0xf41fe4f0,
+	0xce98f30b,
+	0x05e9fd00,
+	0xc80007f1,
+	0xd00103f0,
+	0x04bd000e,
+	0xb804c0b6,
+	0x1bf404cd,
+	0x02abc8d8,
+/* 0x0207: mmctx_fini_wait */
+	0xf11f1bf4,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x1fb4f000,
+	0xf410b4b0,
+	0xa7f0f01b,
+	0xd021f402,
+/* 0x0223: mmctx_stop */
+	0xc82b0ef4,
 	0xb4b600ab,
 	0x0cb9f010,
-	0xb601aec8,
-	0xbefd11e4,
-	0x008bd005,
-/* 0x01b4: mmctx_exec_loop */
-/* 0x01b4: mmctx_wait_free */
-	0xf0008ecf,
-	0x0bf41fe4,
-	0x00ce98fa,
-	0xd005e9fd,
-	0xc0b6c08e,
-	0x04cdb804,
-	0xc8e81bf4,
-	0x1bf402ab,
-/* 0x01d5: mmctx_fini_wait */
-	0x008bcf18,
-	0xb01fb4f0,
-	0x1bf410b4,
-	0x02a7f0f7,
-	0xf4c921f4,
-/* 0x01ea: mmctx_stop */
-	0xabc81b0e,
-	0x10b4b600,
-	0xf00cb9f0,
-	0x8bd012b9,
-/* 0x01f9: mmctx_stop_wait */
-	0x008bcf00,
-	0xf412bbc8,
-/* 0x0202: mmctx_done */
-	0x94bdfa1b,
-	0xf10199f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0215: strand_wait */
-	0xf0a0f900,
-	0x21f402a7,
-	0xf8a0fcc9,
-/* 0x0221: strand_pre */
-	0xfc87f100,
-	0x0283f04a,
-	0xd00c97f0,
-	0x21f50089,
-	0x00f80215,
-/* 0x0234: strand_post */
-	0x4afc87f1,
-	0xf00283f0,
-	0x89d00d97,
-	0x1521f500,
-/* 0x0247: strand_set */
-	0xf100f802,
-	0xf04ffca7,
-	0xaba202a3,
-	0xc7f00500,
-	0x00acd00f,
-	0xd00bc7f0,
-	0x21f500bc,
-	0xaed00215,
-	0x0ac7f000,
-	0xf500bcd0,
-	0xf8021521,
-/* 0x0271: strand_ctx_init */
-	0xf094bd00,
-	0x07f10399,
-	0x03f00f00,
+	0xf112b9f0,
+	0xf0c50007,
+	0x0bd00103,
+/* 0x023b: mmctx_stop_wait */
+	0xf104bd00,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x12bbc800,
+/* 0x024b: mmctx_done */
+	0xbdf31bf4,
+	0x0199f094,
+	0x170007f1,
+	0xd00203f0,
+	0x04bd0009,
+/* 0x025e: strand_wait */
+	0xa0f900f8,
+	0xf402a7f0,
+	0xa0fcd021,
+/* 0x026a: strand_pre */
+	0x97f000f8,
+	0xfc07f10c,
+	0x0203f04a,
+	0xbd0009d0,
+	0x5e21f504,
+/* 0x027f: strand_post */
+	0xf000f802,
+	0x07f10d97,
+	0x03f04afc,
 	0x0009d002,
 	0x21f504bd,
-	0xe7f00221,
-	0x4721f503,
-	0xfca7f102,
-	0x02a3f046,
-	0x0400aba0,
-	0xf040a0d0,
-	0xbcd001c7,
-	0x1521f500,
-	0x010c9202,
-	0xf000acd0,
-	0xbcd002c7,
-	0x1521f500,
-	0x3421f502,
-	0x8087f102,
-	0x0684b608,
-	0xb70089cf,
-	0x95220080,
-/* 0x02ca: ctx_init_strand_loop */
+	0x00f8025e,
+/* 0x0294: strand_set */
+	0xf10fc7f0,
+	0xf04ffc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f10bc7,
+	0x03f04afc,
+	0x000cd002,
+	0x07f104bd,
+	0x03f04ffc,
+	0x000ed002,
+	0xc7f004bd,
+	0xfc07f10a,
+	0x0203f04a,
+	0xbd000cd0,
+	0x5e21f504,
+/* 0x02d3: strand_ctx_init */
+	0xbd00f802,
+	0x0399f094,
+	0x0f0007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0x026a21f5,
+	0xf503e7f0,
+	0xbd029421,
+	0xfc07f1c4,
+	0x0203f047,
+	0xbd000cd0,
+	0x01c7f004,
+	0x4afc07f1,
+	0xd00203f0,
+	0x04bd000c,
+	0x025e21f5,
+	0xf1010c92,
+	0xf046fc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f102c7,
+	0x03f04afc,
+	0x000cd002,
+	0x21f504bd,
+	0x21f5025e,
+	0x87f1027f,
+	0x83f04200,
+	0x0097f102,
+	0x0293f020,
+	0x950099cf,
+/* 0x034a: ctx_init_strand_loop */
 	0x8ed008fe,
 	0x408ed000,
 	0xb6808acf,
@@ -263,198 +293,230 @@ uint32_t nve0_grgpc_code[] = {
 	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-/* 0x02fe: error */
+/* 0x037e: error */
 	0xe0f900f8,
-	0x9814e7f1,
-	0xf440e3f0,
-	0xe0b78d21,
-	0xf7f0041c,
-	0x8d21f401,
-	0x00f8e0fc,
-/* 0x0318: init */
-	0x04fe04bd,
-	0x0017f100,
-	0x0227f012,
-	0xf10012d0,
-	0xfe047017,
-	0x17f10010,
-	0x10d00400,
-	0x0427f0c0,
-	0xf40012d0,
-	0x17f11031,
-	0x14b60608,
-	0x0012cf06,
+	0xf102ffb9,
+	0xf09814e7,
+	0x21f440e3,
+	0x01f7f09d,
+	0xf102ffb9,
+	0xf09c1ce7,
+	0x21f440e3,
+	0xf8e0fc9d,
+/* 0x03a1: init */
+	0xfe04bd00,
+	0x27f00004,
+	0x0007f102,
+	0x0003f012,
+	0xbd0002d0,
+	0x1f17f104,
+	0x0010fe05,
+	0x070007f1,
+	0xd00003f0,
+	0x04bd0000,
+	0xf10427f0,
+	0xf0040007,
+	0x02d00003,
+	0xf404bd00,
+	0x27f11031,
+	0x23f08200,
+	0x0022cf01,
 	0xf00137f0,
 	0x32bb1f24,
 	0x0132b604,
 	0x80050280,
-	0x10b70603,
-	0x12cf0400,
-	0x04028000,
-	0x0c30e7f1,
-	0xbd50e3f0,
-	0xbd34bd24,
-/* 0x0371: init_unk_loop */
-	0x6821f444,
-	0xf400f6b0,
-	0xf7f00f0b,
-	0x04f2bb01,
-	0xb6054ffd,
-/* 0x0386: init_unk_next */
-	0x20b60130,
-	0x04e0b601,
-	0xf40126b0,
-/* 0x0392: init_unk_done */
-	0x0380e21b,
-	0x08048007,
-	0x010027f1,
-	0xcf0223f0,
-	0x34bd0022,
-	0x070047f1,
-	0x950644b6,
-	0x45d00825,
-	0x4045d000,
-	0x98000e98,
-	0x21f5010f,
-	0x2fbb0147,
-	0x003fbb00,
-	0x98010e98,
-	0x21f5020f,
-	0x0e980147,
-	0x00effd05,
-	0xbb002ebb,
-	0x0e98003e,
-	0x030f9802,
-	0x014721f5,
-	0xfd070e98,
+	0x27f10603,
+	0x23f08600,
+	0x0022cf01,
+	0xf1040280,
+	0xf00c30e7,
+	0x24bd50e3,
+	0x44bd34bd,
+/* 0x0410: init_unk_loop */
+	0xb06821f4,
+	0x0bf400f6,
+	0x01f7f00f,
+	0xfd04f2bb,
+	0x30b6054f,
+/* 0x0425: init_unk_next */
+	0x0120b601,
+	0xb004e0b6,
+	0x1bf40126,
+/* 0x0431: init_unk_done */
+	0x070380e2,
+	0xf1080480,
+	0xf0010027,
+	0x22cf0223,
+	0x9534bd00,
+	0x07f10825,
+	0x03f0c000,
+	0x0005d001,
+	0x07f104bd,
+	0x03f0c100,
+	0x0005d001,
+	0x0e9804bd,
+	0x010f9800,
+	0x015021f5,
+	0xbb002fbb,
+	0x0e98003f,
+	0x020f9801,
+	0x015021f5,
+	0xfd050e98,
 	0x2ebb00ef,
 	0x003ebb00,
-	0x130040b7,
-	0xd00235b6,
-	0x25b60043,
-	0x0635b608,
-	0xb60120b6,
-	0x24b60130,
-	0x0834b608,
-	0xf5022fb9,
-	0xbb027121,
-	0x07f1003f,
-	0x03f00100,
-	0x0003d002,
-	0x24bd04bd,
-	0xf11f29f0,
-	0xf0080007,
-	0x02d00203,
-/* 0x0433: main */
+	0x98020e98,
+	0x21f5030f,
+	0x0e980150,
+	0x00effd07,
+	0xbb002ebb,
+	0x35b6003e,
+	0x0007f102,
+	0x0103f0d3,
+	0xbd0003d0,
+	0x0825b604,
+	0xb60635b6,
+	0x30b60120,
+	0x0824b601,
+	0xb90834b6,
+	0x21f5022f,
+	0x3fbb02d3,
+	0x0007f100,
+	0x0203f001,
+	0xbd0003d0,
+	0xf024bd04,
+	0x07f11f29,
+	0x03f00800,
+	0x0002d002,
+/* 0x04e2: main */
+	0x31f404bd,
+	0x0028f400,
+	0xf424d7f0,
+	0x01f43921,
+	0x04e4b0f4,
+	0xfe1e18f4,
+	0x27f00181,
+	0xfd20bd06,
+	0xe4b60412,
+	0x051efd01,
+	0xf50018fe,
+	0xf405d721,
+/* 0x0512: main_not_ctx_xfer */
+	0xef94d30e,
+	0x01f5f010,
+	0x037e21f5,
+/* 0x051f: ih */
+	0xf9c60ef4,
+	0x0188fe80,
+	0x90f980f9,
+	0xb0f9a0f9,
+	0xe0f9d0f9,
+	0x04bdf0f9,
+	0x0200a7f1,
+	0xcf00a3f0,
+	0xabc400aa,
+	0x2c0bf404,
+	0xf124d7f0,
+	0xf01a00e7,
+	0xeecf00e3,
+	0x00f7f100,
+	0x00f3f019,
+	0xf400ffcf,
+	0xe7f00421,
+	0x0007f101,
+	0x0003f01d,
+	0xbd000ed0,
+/* 0x056d: ih_no_fifo */
+	0x0007f104,
+	0x0003f001,
+	0xbd000ad0,
+	0xfcf0fc04,
+	0xfcd0fce0,
+	0xfca0fcb0,
+	0xfe80fc90,
+	0x80fc0088,
+	0xf80032f4,
+/* 0x0591: hub_barrier_done */
+	0x01f7f001,
+	0xbb040e98,
+	0xffb904fe,
+	0x18e7f102,
+	0x40e3f094,
+	0xf89d21f4,
+/* 0x05a9: ctx_redswitch */
+	0x20f7f000,
+	0x850007f1,
+	0xd00103f0,
+	0x04bd000f,
+/* 0x05bb: ctx_redswitch_delay */
+	0xb608e7f0,
+	0x1bf401e2,
+	0x00f5f1fd,
+	0x00f5f108,
+	0x0007f102,
+	0x0103f085,
+	0xbd000fd0,
+/* 0x05d7: ctx_xfer */
+	0xf100f804,
+	0xf0810007,
+	0x0fd00203,
 	0xf404bd00,
-	0x28f40031,
-	0x24d7f000,
-	0xf43921f4,
-	0xe4b0f401,
-	0x1e18f404,
-	0xf00181fe,
-	0x20bd0627,
-	0xb60412fd,
-	0x1efd01e4,
-	0x0018fe05,
-	0x04f721f5,
-/* 0x0463: main_not_ctx_xfer */
-	0x94d30ef4,
-	0xf5f010ef,
-	0xfe21f501,
-	0xc60ef402,
-/* 0x0470: ih */
-	0x88fe80f9,
-	0xf980f901,
-	0xf9a0f990,
-	0xf9d0f9b0,
-	0xbdf0f9e0,
-	0x800acf04,
-	0xf404abc4,
-	0xb7f11d0b,
-	0xd7f01900,
-	0x40becf24,
-	0xf400bfcf,
-	0xb0b70421,
-	0xe7f00400,
-	0x00bed001,
-/* 0x04a8: ih_no_fifo */
-	0xfc400ad0,
-	0xfce0fcf0,
-	0xfcb0fcd0,
-	0xfc90fca0,
-	0x0088fe80,
-	0x32f480fc,
-/* 0x04c3: hub_barrier_done */
-	0xf001f800,
-	0x0e9801f7,
-	0x04febb04,
-	0x9418e7f1,
-	0xf440e3f0,
-	0x00f88d21,
-/* 0x04d8: ctx_redswitch */
-	0x0614e7f1,
-	0xf006e4b6,
-	0xefd020f7,
-	0x08f7f000,
-/* 0x04e8: ctx_redswitch_delay */
-	0xf401f2b6,
-	0xf7f1fd1b,
-	0xefd00a20,
-/* 0x04f7: ctx_xfer */
-	0xf100f800,
-	0xb60a0417,
-	0x1fd00614,
-	0x0711f400,
-	0x04d821f5,
-/* 0x0508: ctx_xfer_not_load */
-	0x4afc17f1,
-	0xf00213f0,
-	0x12d00c27,
-	0x1521f500,
-	0xfc27f102,
-	0x0223f047,
-	0xf00020d0,
-	0x20b6012c,
-	0x0012d003,
-	0xf001acf0,
-	0xb7f002a5,
-	0x50b3f000,
-	0xb6040c98,
-	0xbcbb0fc4,
-	0x000c9800,
-	0xf0010d98,
-	0x21f500e7,
-	0xacf00166,
-	0x00b7f101,
-	0x50b3f040,
-	0xb6040c98,
-	0xbcbb0fc4,
-	0x010c9800,
-	0x98020d98,
-	0xe7f1060f,
-	0x21f50800,
-	0xacf00166,
-	0x04a5f001,
-	0x3000b7f1,
+	0x21f50711,
+/* 0x05ea: ctx_xfer_not_load */
+	0x21f505a9,
+	0x24bd026a,
+	0x47fc07f1,
+	0xd00203f0,
+	0x04bd0002,
+	0xb6012cf0,
+	0x07f10320,
+	0x03f04afc,
+	0x0002d002,
+	0xacf004bd,
+	0x02a5f001,
+	0x0000b7f1,
 	0x9850b3f0,
 	0xc4b6040c,
 	0x00bcbb0f,
-	0x98020c98,
-	0x0f98030d,
-	0x00e7f108,
-	0x6621f502,
-	0x1521f501,
-	0x0601f402,
-/* 0x05a3: ctx_xfer_post */
-	0xf11412f4,
-	0xf04afc17,
-	0x27f00213,
-	0x0012d00d,
-	0x021521f5,
-/* 0x05b4: ctx_xfer_done */
-	0x04c321f5,
-	0x000000f8,
+	0x98000c98,
+	0xe7f0010d,
+	0x6f21f500,
+	0x01acf001,
+	0x4000b7f1,
+	0x9850b3f0,
+	0xc4b6040c,
+	0x00bcbb0f,
+	0x98010c98,
+	0x0f98020d,
+	0x00e7f106,
+	0x6f21f508,
+	0x01acf001,
+	0xf104a5f0,
+	0xf03000b7,
+	0x0c9850b3,
+	0x0fc4b604,
+	0x9800bcbb,
+	0x0d98020c,
+	0x080f9803,
+	0x0200e7f1,
+	0x016f21f5,
+	0x025e21f5,
+	0xf40601f4,
+/* 0x0686: ctx_xfer_post */
+	0x21f50712,
+/* 0x068a: ctx_xfer_done */
+	0x21f5027f,
+	0x00f80591,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnvf0.fuc.h b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnvf0.fuc.h
index f870507be880..6316ebaf5d9a 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnvf0.fuc.h
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/gpcnvf0.fuc.h
@@ -41,14 +41,14 @@ uint32_t nvf0_grgpc_data[] = {
 };
 
 uint32_t nvf0_grgpc_code[] = {
-	0x03180ef5,
+	0x03a10ef5,
 /* 0x0004: queue_put */
 	0x9800d898,
 	0x86f001d9,
 	0x0489b808,
 	0xf00c1bf4,
 	0x21f502f7,
-	0x00f802fe,
+	0x00f8037e,
 /* 0x001c: queue_put_next */
 	0xb60798c4,
 	0x8dbb0384,
@@ -72,184 +72,214 @@ uint32_t nvf0_grgpc_code[] = {
 /* 0x0066: queue_get_done */
 	0x00f80132,
 /* 0x0068: nv_rd32 */
-	0x0728b7f1,
-	0xb906b4b6,
-	0xc9f002ec,
-	0x00bcd01f,
-/* 0x0078: nv_rd32_wait */
-	0xc800bccf,
-	0x1bf41fcc,
-	0x06a7f0fa,
-	0x010921f5,
-	0xf840bfcf,
-/* 0x008d: nv_wr32 */
-	0x28b7f100,
-	0x06b4b607,
-	0xb980bfd0,
-	0xc9f002ec,
-	0x1ec9f01f,
-/* 0x00a3: nv_wr32_wait */
-	0xcf00bcd0,
-	0xccc800bc,
-	0xfa1bf41f,
-/* 0x00ae: watchdog_reset */
-	0x87f100f8,
-	0x84b60430,
-	0x1ff9f006,
-	0xf8008fd0,
-/* 0x00bd: watchdog_clear */
-	0x3087f100,
-	0x0684b604,
-	0xf80080d0,
-/* 0x00c9: wait_donez */
-	0xf094bd00,
-	0x07f10099,
-	0x03f03700,
-	0x0009d002,
-	0x07f104bd,
-	0x03f00600,
-	0x000ad002,
-/* 0x00e6: wait_donez_ne */
-	0x87f104bd,
-	0x83f00000,
-	0x0088cf01,
-	0xf4888aff,
-	0x94bdf31b,
-	0xf10099f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0109: wait_doneo */
-	0xf094bd00,
+	0xf002ecb9,
+	0x07f11fc9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x007a: nv_rd32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0xa7f0f31b,
+	0x1021f506,
+	0x00f7f101,
+	0x01f3f0cb,
+	0xf800ffcf,
+/* 0x009d: nv_wr32 */
+	0x0007f100,
+	0x0103f0cc,
+	0xbd000fd0,
+	0x02ecb904,
+	0xf01fc9f0,
+	0x07f11ec9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x00be: nv_wr32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0x00f8f31b,
+/* 0x00d0: wait_donez */
+	0x99f094bd,
+	0x0007f100,
+	0x0203f037,
+	0xbd0009d0,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x00ed: wait_donez_ne */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x1bf4888a,
+	0xf094bdf3,
 	0x07f10099,
-	0x03f03700,
+	0x03f01700,
 	0x0009d002,
-	0x87f104bd,
-	0x84b60818,
-	0x008ad006,
-/* 0x0124: wait_doneo_e */
-	0x040087f1,
-	0xcf0684b6,
-	0x8aff0088,
-	0xf30bf488,
+	0x00f804bd,
+/* 0x0110: wait_doneo */
 	0x99f094bd,
 	0x0007f100,
-	0x0203f017,
+	0x0203f037,
 	0xbd0009d0,
-/* 0x0147: mmctx_size */
-	0xbd00f804,
-/* 0x0149: nv_mmctx_size_loop */
-	0x00e89894,
-	0xb61a85b6,
-	0x84b60180,
-	0x0098bb02,
-	0xb804e0b6,
-	0x1bf404ef,
-	0x029fb9eb,
-/* 0x0166: mmctx_xfer */
-	0x94bd00f8,
-	0xf10199f0,
-	0xf0370007,
-	0x09d00203,
-	0xf104bd00,
-	0xb6071087,
-	0x94bd0684,
-	0xf405bbfd,
-	0x8bd0090b,
-	0x0099f000,
-/* 0x018c: mmctx_base_disabled */
-	0xf405eefd,
-	0x8ed00c0b,
-	0xc08fd080,
-/* 0x019b: mmctx_multi_disabled */
-	0xb70199f0,
-	0xc8010080,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x012d: wait_doneo_e */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x0bf4888a,
+	0xf094bdf3,
+	0x07f10099,
+	0x03f01700,
+	0x0009d002,
+	0x00f804bd,
+/* 0x0150: mmctx_size */
+/* 0x0152: nv_mmctx_size_loop */
+	0xe89894bd,
+	0x1a85b600,
+	0xb60180b6,
+	0x98bb0284,
+	0x04e0b600,
+	0xf404efb8,
+	0x9fb9eb1b,
+/* 0x016f: mmctx_xfer */
+	0xbd00f802,
+	0x0199f094,
+	0x370007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0xbbfd94bd,
+	0x120bf405,
+	0xc40007f1,
+	0xd00103f0,
+	0x04bd000b,
+/* 0x0197: mmctx_base_disabled */
+	0xfd0099f0,
+	0x0bf405ee,
+	0x0007f11e,
+	0x0103f0c6,
+	0xbd000ed0,
+	0x0007f104,
+	0x0103f0c7,
+	0xbd000fd0,
+	0x0199f004,
+/* 0x01b8: mmctx_multi_disabled */
+	0xb600abc8,
+	0xb9f010b4,
+	0x01aec80c,
+	0xfd11e4b6,
+	0x07f105be,
+	0x03f0c500,
+	0x000bd001,
+/* 0x01d6: mmctx_exec_loop */
+/* 0x01d6: mmctx_wait_free */
+	0xe7f104bd,
+	0xe3f0c500,
+	0x00eecf01,
+	0xf41fe4f0,
+	0xce98f30b,
+	0x05e9fd00,
+	0xc80007f1,
+	0xd00103f0,
+	0x04bd000e,
+	0xb804c0b6,
+	0x1bf404cd,
+	0x02abc8d8,
+/* 0x0207: mmctx_fini_wait */
+	0xf11f1bf4,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x1fb4f000,
+	0xf410b4b0,
+	0xa7f0f01b,
+	0xd021f402,
+/* 0x0223: mmctx_stop */
+	0xc82b0ef4,
 	0xb4b600ab,
 	0x0cb9f010,
-	0xb601aec8,
-	0xbefd11e4,
-	0x008bd005,
-/* 0x01b4: mmctx_exec_loop */
-/* 0x01b4: mmctx_wait_free */
-	0xf0008ecf,
-	0x0bf41fe4,
-	0x00ce98fa,
-	0xd005e9fd,
-	0xc0b6c08e,
-	0x04cdb804,
-	0xc8e81bf4,
-	0x1bf402ab,
-/* 0x01d5: mmctx_fini_wait */
-	0x008bcf18,
-	0xb01fb4f0,
-	0x1bf410b4,
-	0x02a7f0f7,
-	0xf4c921f4,
-/* 0x01ea: mmctx_stop */
-	0xabc81b0e,
-	0x10b4b600,
-	0xf00cb9f0,
-	0x8bd012b9,
-/* 0x01f9: mmctx_stop_wait */
-	0x008bcf00,
-	0xf412bbc8,
-/* 0x0202: mmctx_done */
-	0x94bdfa1b,
-	0xf10199f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0215: strand_wait */
-	0xf0a0f900,
-	0x21f402a7,
-	0xf8a0fcc9,
-/* 0x0221: strand_pre */
-	0xfc87f100,
-	0x0283f04a,
-	0xd00c97f0,
-	0x21f50089,
-	0x00f80215,
-/* 0x0234: strand_post */
-	0x4afc87f1,
-	0xf00283f0,
-	0x89d00d97,
-	0x1521f500,
-/* 0x0247: strand_set */
-	0xf100f802,
-	0xf04ffca7,
-	0xaba202a3,
-	0xc7f00500,
-	0x00acd00f,
-	0xd00bc7f0,
-	0x21f500bc,
-	0xaed00215,
-	0x0ac7f000,
-	0xf500bcd0,
-	0xf8021521,
-/* 0x0271: strand_ctx_init */
-	0xf094bd00,
-	0x07f10399,
-	0x03f03700,
+	0xf112b9f0,
+	0xf0c50007,
+	0x0bd00103,
+/* 0x023b: mmctx_stop_wait */
+	0xf104bd00,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x12bbc800,
+/* 0x024b: mmctx_done */
+	0xbdf31bf4,
+	0x0199f094,
+	0x170007f1,
+	0xd00203f0,
+	0x04bd0009,
+/* 0x025e: strand_wait */
+	0xa0f900f8,
+	0xf402a7f0,
+	0xa0fcd021,
+/* 0x026a: strand_pre */
+	0x97f000f8,
+	0xfc07f10c,
+	0x0203f04a,
+	0xbd0009d0,
+	0x5e21f504,
+/* 0x027f: strand_post */
+	0xf000f802,
+	0x07f10d97,
+	0x03f04afc,
 	0x0009d002,
 	0x21f504bd,
-	0xe7f00221,
-	0x4721f503,
-	0xfca7f102,
-	0x02a3f046,
-	0x0400aba0,
-	0xf040a0d0,
-	0xbcd001c7,
-	0x1521f500,
-	0x010c9202,
-	0xf000acd0,
-	0xbcd002c7,
-	0x1521f500,
-	0x3421f502,
-	0x8087f102,
-	0x0684b608,
-	0xb70089cf,
-	0x95220080,
-/* 0x02ca: ctx_init_strand_loop */
+	0x00f8025e,
+/* 0x0294: strand_set */
+	0xf10fc7f0,
+	0xf04ffc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f10bc7,
+	0x03f04afc,
+	0x000cd002,
+	0x07f104bd,
+	0x03f04ffc,
+	0x000ed002,
+	0xc7f004bd,
+	0xfc07f10a,
+	0x0203f04a,
+	0xbd000cd0,
+	0x5e21f504,
+/* 0x02d3: strand_ctx_init */
+	0xbd00f802,
+	0x0399f094,
+	0x370007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0x026a21f5,
+	0xf503e7f0,
+	0xbd029421,
+	0xfc07f1c4,
+	0x0203f047,
+	0xbd000cd0,
+	0x01c7f004,
+	0x4afc07f1,
+	0xd00203f0,
+	0x04bd000c,
+	0x025e21f5,
+	0xf1010c92,
+	0xf046fc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f102c7,
+	0x03f04afc,
+	0x000cd002,
+	0x21f504bd,
+	0x21f5025e,
+	0x87f1027f,
+	0x83f04200,
+	0x0097f102,
+	0x0293f020,
+	0x950099cf,
+/* 0x034a: ctx_init_strand_loop */
 	0x8ed008fe,
 	0x408ed000,
 	0xb6808acf,
@@ -263,198 +293,230 @@ uint32_t nvf0_grgpc_code[] = {
 	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-/* 0x02fe: error */
+/* 0x037e: error */
 	0xe0f900f8,
-	0x9814e7f1,
-	0xf440e3f0,
-	0xe0b78d21,
-	0xf7f0041c,
-	0x8d21f401,
-	0x00f8e0fc,
-/* 0x0318: init */
-	0x04fe04bd,
-	0x0017f100,
-	0x0227f012,
-	0xf10012d0,
-	0xfe047017,
-	0x17f10010,
-	0x10d00400,
-	0x0427f0c0,
-	0xf40012d0,
-	0x17f11031,
-	0x14b60608,
-	0x0012cf06,
+	0xf102ffb9,
+	0xf09814e7,
+	0x21f440e3,
+	0x01f7f09d,
+	0xf102ffb9,
+	0xf09c1ce7,
+	0x21f440e3,
+	0xf8e0fc9d,
+/* 0x03a1: init */
+	0xfe04bd00,
+	0x27f00004,
+	0x0007f102,
+	0x0003f012,
+	0xbd0002d0,
+	0x1f17f104,
+	0x0010fe05,
+	0x070007f1,
+	0xd00003f0,
+	0x04bd0000,
+	0xf10427f0,
+	0xf0040007,
+	0x02d00003,
+	0xf404bd00,
+	0x27f11031,
+	0x23f08200,
+	0x0022cf01,
 	0xf00137f0,
 	0x32bb1f24,
 	0x0132b604,
 	0x80050280,
-	0x10b70603,
-	0x12cf0400,
-	0x04028000,
-	0x0c30e7f1,
-	0xbd50e3f0,
-	0xbd34bd24,
-/* 0x0371: init_unk_loop */
-	0x6821f444,
-	0xf400f6b0,
-	0xf7f00f0b,
-	0x04f2bb01,
-	0xb6054ffd,
-/* 0x0386: init_unk_next */
-	0x20b60130,
-	0x04e0b601,
-	0xf40226b0,
-/* 0x0392: init_unk_done */
-	0x0380e21b,
-	0x08048007,
-	0x010027f1,
-	0xcf0223f0,
-	0x34bd0022,
-	0x070047f1,
-	0x950644b6,
-	0x45d00825,
-	0x4045d000,
-	0x98000e98,
-	0x21f5010f,
-	0x2fbb0147,
-	0x003fbb00,
-	0x98010e98,
-	0x21f5020f,
-	0x0e980147,
-	0x00effd05,
-	0xbb002ebb,
-	0x0e98003e,
-	0x030f9802,
-	0x014721f5,
-	0xfd070e98,
+	0x27f10603,
+	0x23f08600,
+	0x0022cf01,
+	0xf1040280,
+	0xf00c30e7,
+	0x24bd50e3,
+	0x44bd34bd,
+/* 0x0410: init_unk_loop */
+	0xb06821f4,
+	0x0bf400f6,
+	0x01f7f00f,
+	0xfd04f2bb,
+	0x30b6054f,
+/* 0x0425: init_unk_next */
+	0x0120b601,
+	0xb004e0b6,
+	0x1bf40226,
+/* 0x0431: init_unk_done */
+	0x070380e2,
+	0xf1080480,
+	0xf0010027,
+	0x22cf0223,
+	0x9534bd00,
+	0x07f10825,
+	0x03f0c000,
+	0x0005d001,
+	0x07f104bd,
+	0x03f0c100,
+	0x0005d001,
+	0x0e9804bd,
+	0x010f9800,
+	0x015021f5,
+	0xbb002fbb,
+	0x0e98003f,
+	0x020f9801,
+	0x015021f5,
+	0xfd050e98,
 	0x2ebb00ef,
 	0x003ebb00,
-	0x130040b7,
-	0xd00235b6,
-	0x25b60043,
-	0x0635b608,
-	0xb60120b6,
-	0x24b60130,
-	0x0834b608,
-	0xf5022fb9,
-	0xbb027121,
-	0x07f1003f,
-	0x03f00100,
-	0x0003d002,
-	0x24bd04bd,
-	0xf11f29f0,
-	0xf0300007,
-	0x02d00203,
-/* 0x0433: main */
+	0x98020e98,
+	0x21f5030f,
+	0x0e980150,
+	0x00effd07,
+	0xbb002ebb,
+	0x35b6003e,
+	0x0007f102,
+	0x0103f0d3,
+	0xbd0003d0,
+	0x0825b604,
+	0xb60635b6,
+	0x30b60120,
+	0x0824b601,
+	0xb90834b6,
+	0x21f5022f,
+	0x3fbb02d3,
+	0x0007f100,
+	0x0203f001,
+	0xbd0003d0,
+	0xf024bd04,
+	0x07f11f29,
+	0x03f03000,
+	0x0002d002,
+/* 0x04e2: main */
+	0x31f404bd,
+	0x0028f400,
+	0xf424d7f0,
+	0x01f43921,
+	0x04e4b0f4,
+	0xfe1e18f4,
+	0x27f00181,
+	0xfd20bd06,
+	0xe4b60412,
+	0x051efd01,
+	0xf50018fe,
+	0xf405d721,
+/* 0x0512: main_not_ctx_xfer */
+	0xef94d30e,
+	0x01f5f010,
+	0x037e21f5,
+/* 0x051f: ih */
+	0xf9c60ef4,
+	0x0188fe80,
+	0x90f980f9,
+	0xb0f9a0f9,
+	0xe0f9d0f9,
+	0x04bdf0f9,
+	0x0200a7f1,
+	0xcf00a3f0,
+	0xabc400aa,
+	0x2c0bf404,
+	0xf124d7f0,
+	0xf01a00e7,
+	0xeecf00e3,
+	0x00f7f100,
+	0x00f3f019,
+	0xf400ffcf,
+	0xe7f00421,
+	0x0007f101,
+	0x0003f01d,
+	0xbd000ed0,
+/* 0x056d: ih_no_fifo */
+	0x0007f104,
+	0x0003f001,
+	0xbd000ad0,
+	0xfcf0fc04,
+	0xfcd0fce0,
+	0xfca0fcb0,
+	0xfe80fc90,
+	0x80fc0088,
+	0xf80032f4,
+/* 0x0591: hub_barrier_done */
+	0x01f7f001,
+	0xbb040e98,
+	0xffb904fe,
+	0x18e7f102,
+	0x40e3f094,
+	0xf89d21f4,
+/* 0x05a9: ctx_redswitch */
+	0x20f7f000,
+	0x850007f1,
+	0xd00103f0,
+	0x04bd000f,
+/* 0x05bb: ctx_redswitch_delay */
+	0xb608e7f0,
+	0x1bf401e2,
+	0x00f5f1fd,
+	0x00f5f108,
+	0x0007f102,
+	0x0103f085,
+	0xbd000fd0,
+/* 0x05d7: ctx_xfer */
+	0xf100f804,
+	0xf0810007,
+	0x0fd00203,
 	0xf404bd00,
-	0x28f40031,
-	0x24d7f000,
-	0xf43921f4,
-	0xe4b0f401,
-	0x1e18f404,
-	0xf00181fe,
-	0x20bd0627,
-	0xb60412fd,
-	0x1efd01e4,
-	0x0018fe05,
-	0x04f721f5,
-/* 0x0463: main_not_ctx_xfer */
-	0x94d30ef4,
-	0xf5f010ef,
-	0xfe21f501,
-	0xc60ef402,
-/* 0x0470: ih */
-	0x88fe80f9,
-	0xf980f901,
-	0xf9a0f990,
-	0xf9d0f9b0,
-	0xbdf0f9e0,
-	0x800acf04,
-	0xf404abc4,
-	0xb7f11d0b,
-	0xd7f01900,
-	0x40becf24,
-	0xf400bfcf,
-	0xb0b70421,
-	0xe7f00400,
-	0x00bed001,
-/* 0x04a8: ih_no_fifo */
-	0xfc400ad0,
-	0xfce0fcf0,
-	0xfcb0fcd0,
-	0xfc90fca0,
-	0x0088fe80,
-	0x32f480fc,
-/* 0x04c3: hub_barrier_done */
-	0xf001f800,
-	0x0e9801f7,
-	0x04febb04,
-	0x9418e7f1,
-	0xf440e3f0,
-	0x00f88d21,
-/* 0x04d8: ctx_redswitch */
-	0x0614e7f1,
-	0xf006e4b6,
-	0xefd020f7,
-	0x08f7f000,
-/* 0x04e8: ctx_redswitch_delay */
-	0xf401f2b6,
-	0xf7f1fd1b,
-	0xefd00a20,
-/* 0x04f7: ctx_xfer */
-	0xf100f800,
-	0xb60a0417,
-	0x1fd00614,
-	0x0711f400,
-	0x04d821f5,
-/* 0x0508: ctx_xfer_not_load */
-	0x4afc17f1,
-	0xf00213f0,
-	0x12d00c27,
-	0x1521f500,
-	0xfc27f102,
-	0x0223f047,
-	0xf00020d0,
-	0x20b6012c,
-	0x0012d003,
-	0xf001acf0,
-	0xb7f002a5,
-	0x50b3f000,
-	0xb6040c98,
-	0xbcbb0fc4,
-	0x000c9800,
-	0xf0010d98,
-	0x21f500e7,
-	0xacf00166,
-	0x00b7f101,
-	0x50b3f040,
-	0xb6040c98,
-	0xbcbb0fc4,
-	0x010c9800,
-	0x98020d98,
-	0xe7f1060f,
-	0x21f50800,
-	0xacf00166,
-	0x04a5f001,
-	0x3000b7f1,
+	0x21f50711,
+/* 0x05ea: ctx_xfer_not_load */
+	0x21f505a9,
+	0x24bd026a,
+	0x47fc07f1,
+	0xd00203f0,
+	0x04bd0002,
+	0xb6012cf0,
+	0x07f10320,
+	0x03f04afc,
+	0x0002d002,
+	0xacf004bd,
+	0x02a5f001,
+	0x0000b7f1,
 	0x9850b3f0,
 	0xc4b6040c,
 	0x00bcbb0f,
-	0x98020c98,
-	0x0f98030d,
-	0x00e7f108,
-	0x6621f502,
-	0x1521f501,
-	0x0601f402,
-/* 0x05a3: ctx_xfer_post */
-	0xf11412f4,
-	0xf04afc17,
-	0x27f00213,
-	0x0012d00d,
-	0x021521f5,
-/* 0x05b4: ctx_xfer_done */
-	0x04c321f5,
-	0x000000f8,
+	0x98000c98,
+	0xe7f0010d,
+	0x6f21f500,
+	0x01acf001,
+	0x4000b7f1,
+	0x9850b3f0,
+	0xc4b6040c,
+	0x00bcbb0f,
+	0x98010c98,
+	0x0f98020d,
+	0x00e7f106,
+	0x6f21f508,
+	0x01acf001,
+	0xf104a5f0,
+	0xf03000b7,
+	0x0c9850b3,
+	0x0fc4b604,
+	0x9800bcbb,
+	0x0d98020c,
+	0x080f9803,
+	0x0200e7f1,
+	0x016f21f5,
+	0x025e21f5,
+	0xf40601f4,
+/* 0x0686: ctx_xfer_post */
+	0x21f50712,
+/* 0x068a: ctx_xfer_done */
+	0x21f5027f,
+	0x00f80591,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hub.fuc b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hub.fuc
index b82d2ae89917..c8ddb8d71b91 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hub.fuc
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hub.fuc
@@ -68,60 +68,57 @@ error:
 //
 init:
 	clear b32 $r0
-	mov $sp $r0
 	mov $xdbase $r0
 
+	// setup stack
+	nv_iord($r1, NV_PGRAPH_FECS_CAPS, 0)
+	extr $r1 $r1 9:17
+	shl b32 $r1 8
+	mov $sp $r1
+
 	// enable fifo access
-	mov $r1 0x1200
-	mov $r2 2
-	iowr I[$r1 + 0x000] $r2	// FIFO_ENABLE
+	mov $r2 NV_PGRAPH_FECS_ACCESS_FIFO
+	nv_iowr(NV_PGRAPH_FECS_ACCESS, 0, $r2)
 
 	// setup i0 handler, and route all interrupts to it
 	mov $r1 #ih
 	mov $iv0 $r1
-	mov $r1 0x400
-	iowr I[$r1 + 0x300] $r0	// INTR_DISPATCH
 
-	// route HUB_CHANNEL_SWITCH to fuc interrupt 8
-	mov $r3 0x404
-	shl b32 $r3 6
-	mov $r2 0x2003		// { HUB_CHANNEL_SWITCH, ZERO } -> intr 8
-	iowr I[$r3 + 0x000] $r2
+	clear b32 $r2
+	nv_iowr(NV_PGRAPH_FECS_INTR_ROUTE, 0, $r2)
+
+	// route HUB_CHSW_PULSE to fuc interrupt 8
+	mov $r2 0x2003		// { HUB_CHSW_PULSE, ZERO } -> intr 8
+	nv_iowr(NV_PGRAPH_FECS_IROUTE, 0, $r2)
 
 	// not sure what these are, route them because NVIDIA does, and
 	// the IRQ handler will signal the host if we ever get one.. we
 	// may find out if/why we need to handle these if so..
 	//
-	mov $r2 0x2004
-	iowr I[$r3 + 0x004] $r2 // { 0x04, ZERO } -> intr 9
-	mov $r2 0x200b
-	iowr I[$r3 + 0x008] $r2 // { 0x0b, ZERO } -> intr 10
-	mov $r2 0x200c
-	iowr I[$r3 + 0x01c] $r2 // { 0x0c, ZERO } -> intr 15
+	mov $r2 0x2004		// { 0x04, ZERO } -> intr 9
+	nv_iowr(NV_PGRAPH_FECS_IROUTE, 1, $r2)
+	mov $r2 0x200b		// { HUB_FIRMWARE_MTHD, ZERO } -> intr 10
+	nv_iowr(NV_PGRAPH_FECS_IROUTE, 2, $r2)
+	mov $r2 0x200c		// { 0x0c, ZERO } -> intr 15
+	nv_iowr(NV_PGRAPH_FECS_IROUTE, 7, $r2)
 
 	// enable all INTR_UP interrupts
-	mov $r2 0xc24
-	shl b32 $r2 6
-	not b32 $r3 $r0
-	iowr I[$r2] $r3
+	sub b32 $r3 $r0 1
+	nv_iowr(NV_PGRAPH_FECS_INTR_UP_EN, 0, $r3)
 
-	// enable fifo, ctxsw, 9, 10, 15 interrupts
-	mov $r2 -0x78fc		// 0x8704
-	sethi $r2 0
-	iowr I[$r1 + 0x000] $r2	// INTR_EN_SET
+	// enable fifo, ctxsw, 9, fwmthd, 15 interrupts
+	imm32($r2, 0x8704)
+	nv_iowr(NV_PGRAPH_FECS_INTR_EN_SET, 0, $r2)
 
 	// fifo level triggered, rest edge
-	sub b32 $r1 0x100
-	mov $r2 4
-	iowr I[$r1] $r2
+	mov $r2 NV_PGRAPH_FECS_INTR_MODE_FIFO_LEVEL
+	nv_iowr(NV_PGRAPH_FECS_INTR_MODE, 0, $r2)
 
 	// enable interrupts
 	bset $flags ie0
 
 	// fetch enabled GPC/ROP counts
-	mov $r14 -0x69fc	// 0x409604
-	sethi $r14 0x400000
-	call #nv_rd32
+	nv_rd32($r14, 0x409604)
 	extr $r1 $r15 16:20
 	st b32 D[$r0 + #rop_count] $r1
 	and $r15 0x1f
@@ -131,37 +128,40 @@ init:
 	mov $r1 1
 	shl b32 $r1 $r15
 	sub b32 $r1 1
-	mov $r2 0x40c
-	shl b32 $r2 6
-	iowr I[$r2 + 0x000] $r1
-	iowr I[$r2 + 0x100] $r1
+	nv_iowr(NV_PGRAPH_FECS_BAR_MASK0, 0, $r1)
+	nv_iowr(NV_PGRAPH_FECS_BAR_MASK1, 0, $r1)
 
 	// context size calculation, reserve first 256 bytes for use by fuc
 	mov $r1 256
 
+	//
+	mov $r15 2
+	call(ctx_4170s)
+	call(ctx_4170w)
+	mov $r15 0x10
+	call(ctx_86c)
+
 	// calculate size of mmio context data
 	ld b32 $r14 D[$r0 + #hub_mmio_list_head]
 	ld b32 $r15 D[$r0 + #hub_mmio_list_tail]
-	call #mmctx_size
+	call(mmctx_size)
 
 	// set mmctx base addresses now so we don't have to do it later,
 	// they don't (currently) ever change
-	mov $r3 0x700
-	shl b32 $r3 6
 	shr b32 $r4 $r1 8
-	iowr I[$r3 + 0x000] $r4		// MMCTX_SAVE_SWBASE
-	iowr I[$r3 + 0x100] $r4		// MMCTX_LOAD_SWBASE
+	nv_iowr(NV_PGRAPH_FECS_MMCTX_SAVE_SWBASE, 0, $r4)
+	nv_iowr(NV_PGRAPH_FECS_MMCTX_LOAD_SWBASE, 0, $r4)
 	add b32 $r3 0x1300
 	add b32 $r1 $r15
 	shr b32 $r15 2
-	iowr I[$r3 + 0x000] $r15	// MMCTX_LOAD_COUNT, wtf for?!?
+	nv_iowr(NV_PGRAPH_FECS_MMCTX_LOAD_COUNT, 0, $r15) // wtf??
 
 	// strands, base offset needs to be aligned to 256 bytes
 	shr b32 $r1 8
 	add b32 $r1 1
 	shl b32 $r1 8
 	mov b32 $r15 $r1
-	call #strand_ctx_init
+	call(strand_ctx_init)
 	add b32 $r1 $r15
 
 	// initialise each GPC in sequence by passing in the offset of its
@@ -173,30 +173,29 @@ init:
 	// in GPCn_CC_SCRATCH[1]
 	//
 	ld b32 $r3 D[$r0 + #gpc_count]
-	mov $r4 0x2000
-	sethi $r4 0x500000
+	imm32($r4, 0x502000)
 	init_gpc:
 		// setup, and start GPC ucode running
 		add b32 $r14 $r4 0x804
 		mov b32 $r15 $r1
-		call #nv_wr32			// CC_SCRATCH[1] = ctx offset
+		call(nv_wr32)			// CC_SCRATCH[1] = ctx offset
 		add b32 $r14 $r4 0x10c
 		clear b32 $r15
-		call #nv_wr32
+		call(nv_wr32)
 		add b32 $r14 $r4 0x104
-		call #nv_wr32			// ENTRY
+		call(nv_wr32)			// ENTRY
 		add b32 $r14 $r4 0x100
 		mov $r15 2			// CTRL_START_TRIGGER
-		call #nv_wr32			// CTRL
+		call(nv_wr32)			// CTRL
 
 		// wait for it to complete, and adjust context size
 		add b32 $r14 $r4 0x800
 		init_gpc_wait:
-			call #nv_rd32
+			call(nv_rd32)
 			xbit $r15 $r15 31
 			bra e #init_gpc_wait
 		add b32 $r14 $r4 0x804
-		call #nv_rd32
+		call(nv_rd32)
 		add b32 $r1 $r15
 
 		// next!
@@ -204,6 +203,12 @@ init:
 		sub b32 $r3 1
 		bra ne #init_gpc
 
+	//
+	mov $r15 0
+	call(ctx_86c)
+	mov $r15 0
+	call(ctx_4170s)
+
 	// save context size, and tell host we're ready
 	nv_iowr(NV_PGRAPH_FECS_CC_SCRATCH_VAL(1), 0, $r1)
 	clear b32 $r1
@@ -218,17 +223,15 @@ main:
 	bset $flags $p0
 	sleep $p0
 	mov $r13 #cmd_queue
-	call #queue_get
+	call(queue_get)
 	bra $p1 #main
 
 	// context switch, requested by GPU?
 	cmpu b32 $r14 0x4001
 	bra ne #main_not_ctx_switch
 		trace_set(T_AUTO)
-		mov $r1 0xb00
-		shl b32 $r1 6
-		iord $r2 I[$r1 + 0x100]		// CHAN_NEXT
-		iord $r1 I[$r1 + 0x000]		// CHAN_CUR
+		nv_iord($r1, NV_PGRAPH_FECS_CHAN_ADDR, 0)
+		nv_iord($r2, NV_PGRAPH_FECS_CHAN_NEXT, 0)
 
 		xbit $r3 $r1 31
 		bra e #chsw_no_prev
@@ -239,12 +242,12 @@ main:
 				trace_set(T_SAVE)
 				bclr $flags $p1
 				bset $flags $p2
-				call #ctx_xfer
+				call(ctx_xfer)
 				trace_clr(T_SAVE);
 				pop $r2
 				trace_set(T_LOAD);
 				bset $flags $p1
-				call #ctx_xfer
+				call(ctx_xfer)
 				trace_clr(T_LOAD);
 				bra #chsw_done
 			chsw_prev_no_next:
@@ -252,25 +255,21 @@ main:
 				mov b32 $r2 $r1
 				bclr $flags $p1
 				bclr $flags $p2
-				call #ctx_xfer
+				call(ctx_xfer)
 				pop $r2
-				mov $r1 0xb00
-				shl b32 $r1 6
-				iowr I[$r1] $r2
+				nv_iowr(NV_PGRAPH_FECS_CHAN_ADDR, 0, $r2)
 				bra #chsw_done
 		chsw_no_prev:
 			xbit $r3 $r2 31
 			bra e #chsw_done
 				bset $flags $p1
 				bclr $flags $p2
-				call #ctx_xfer
+				call(ctx_xfer)
 
 		// ack the context switch request
 		chsw_done:
-		mov $r1 0xb0c
-		shl b32 $r1 6
-		mov $r2 1
-		iowr I[$r1 + 0x000] $r2		// 0x409b0c
+		mov $r2 NV_PGRAPH_FECS_CHSW_ACK
+		nv_iowr(NV_PGRAPH_FECS_CHSW, 0, $r2)
 		trace_clr(T_AUTO)
 		bra #main
 
@@ -279,7 +278,7 @@ main:
 	cmpu b32 $r14 0x0001
 	bra ne #main_not_ctx_chan
 		mov b32 $r2 $r15
-		call #ctx_chan
+		call(ctx_chan)
 		bra #main_done
 
 	// request to store current channel context?
@@ -289,14 +288,14 @@ main:
 		trace_set(T_SAVE)
 		bclr $flags $p1
 		bclr $flags $p2
-		call #ctx_xfer
+		call(ctx_xfer)
 		trace_clr(T_SAVE)
 		bra #main_done
 
 	main_not_ctx_save:
 		shl b32 $r15 $r14 16
 		or $r15 E_BAD_COMMAND
-		call #error
+		call(error)
 		bra #main
 
 	main_done:
@@ -319,41 +318,46 @@ ih:
 	clear b32 $r0
 
 	// incoming fifo command?
-	iord $r10 I[$r0 + 0x200]	// INTR
-	and $r11 $r10 0x00000004
+	nv_iord($r10, NV_PGRAPH_FECS_INTR, 0)
+	and $r11 $r10 NV_PGRAPH_FECS_INTR_FIFO
 	bra e #ih_no_fifo
 		// queue incoming fifo command for later processing
-		mov $r11 0x1900
 		mov $r13 #cmd_queue
-		iord $r14 I[$r11 + 0x100]	// FIFO_CMD
-		iord $r15 I[$r11 + 0x000]	// FIFO_DATA
-		call #queue_put
+		nv_iord($r14, NV_PGRAPH_FECS_FIFO_CMD, 0)
+		nv_iord($r15, NV_PGRAPH_FECS_FIFO_DATA, 0)
+		call(queue_put)
 		add b32 $r11 0x400
 		mov $r14 1
-		iowr I[$r11 + 0x000] $r14	// FIFO_ACK
+		nv_iowr(NV_PGRAPH_FECS_FIFO_ACK, 0, $r14)
 
 	// context switch request?
 	ih_no_fifo:
-	and $r11 $r10 0x00000100
+	and $r11 $r10 NV_PGRAPH_FECS_INTR_CHSW
 	bra e #ih_no_ctxsw
 		// enqueue a context switch for later processing
 		mov $r13 #cmd_queue
 		mov $r14 0x4001
-		call #queue_put
+		call(queue_put)
 
-	// anything we didn't handle, bring it to the host's attention
+	// firmware method?
 	ih_no_ctxsw:
-	mov $r11 0x104
+	and $r11 $r10 NV_PGRAPH_FECS_INTR_FWMTHD
+	bra e #ih_no_fwmthd
+		// none we handle, ack, and fall-through to unhandled
+		mov $r11 0x100
+		nv_wr32(0x400144, $r11)
+
+	// anything we didn't handle, bring it to the host's attention
+	ih_no_fwmthd:
+	mov $r11 0x104 // FIFO | CHSW
 	not b32 $r11
 	and $r11 $r10 $r11
 	bra e #ih_no_other
-		mov $r10 0xc1c
-		shl b32 $r10 6
-		iowr I[$r10] $r11	// INTR_UP_SET
+		nv_iowr(NV_PGRAPH_FECS_INTR_UP_SET, 0, $r11)
 
 	// ack, and wake up main()
 	ih_no_other:
-	iowr I[$r0 + 0x100] $r10	// INTR_ACK
+	nv_iowr(NV_PGRAPH_FECS_INTR_ACK, 0, $r10)
 
 	pop $r15
 	pop $r14
@@ -370,12 +374,10 @@ ih:
 #if CHIPSET < GK100
 // Not real sure, but, MEM_CMD 7 will hang forever if this isn't done
 ctx_4160s:
-	mov $r14 0x4160
-	sethi $r14 0x400000
 	mov $r15 1
-	call #nv_wr32
+	nv_wr32(0x404160, $r15)
 	ctx_4160s_wait:
-		call #nv_rd32
+		nv_rd32($r15, 0x404160)
 		xbit $r15 $r15 4
 		bra e #ctx_4160s_wait
 	ret
@@ -384,10 +386,8 @@ ctx_4160s:
 // to hang with STATUS=0x00000007 until it's cleared.. fbcon can
 // still function with it set however...
 ctx_4160c:
-	mov $r14 0x4160
-	sethi $r14 0x400000
 	clear b32 $r15
-	call #nv_wr32
+	nv_wr32(0x404160, $r15)
 	ret
 #endif
 
@@ -396,18 +396,14 @@ ctx_4160c:
 // In: $r15 value to set 0x404170 to
 //
 ctx_4170s:
-	mov $r14 0x4170
-	sethi $r14 0x400000
 	or $r15 0x10
-	call #nv_wr32
+	nv_wr32(0x404170, $r15)
 	ret
 
 // Waits for a ctx_4170s() call to complete
 //
 ctx_4170w:
-	mov $r14 0x4170
-	sethi $r14 0x400000
-	call #nv_rd32
+	nv_rd32($r15, 0x404170)
 	and $r15 0x10
 	bra ne #ctx_4170w
 	ret
@@ -419,16 +415,18 @@ ctx_4170w:
 // funny things happen.
 //
 ctx_redswitch:
-	mov $r14 0x614
-	shl b32 $r14 6
-	mov $r15 0x270
-	iowr I[$r14] $r15	// HUB_RED_SWITCH = ENABLE_GPC, POWER_ALL
+	mov $r14 NV_PGRAPH_FECS_RED_SWITCH_ENABLE_GPC
+	or  $r14 NV_PGRAPH_FECS_RED_SWITCH_POWER_ROP
+	or  $r14 NV_PGRAPH_FECS_RED_SWITCH_POWER_GPC
+	or  $r14 NV_PGRAPH_FECS_RED_SWITCH_POWER_MAIN
+	nv_iowr(NV_PGRAPH_FECS_RED_SWITCH, 0, $r14)
 	mov $r15 8
 	ctx_redswitch_delay:
 		sub b32 $r15 1
 		bra ne #ctx_redswitch_delay
-	mov $r15 0x770
-	iowr I[$r14] $r15	// HUB_RED_SWITCH = ENABLE_ALL, POWER_ALL
+	or  $r14 NV_PGRAPH_FECS_RED_SWITCH_ENABLE_ROP
+	or  $r14 NV_PGRAPH_FECS_RED_SWITCH_ENABLE_MAIN
+	nv_iowr(NV_PGRAPH_FECS_RED_SWITCH, 0, $r14)
 	ret
 
 // Not a clue what this is for, except that unless the value is 0x10, the
@@ -437,15 +435,18 @@ ctx_redswitch:
 // In: $r15 value to set to (0x00/0x10 are used)
 //
 ctx_86c:
-	mov $r14 0x86c
-	shl b32 $r14 6
-	iowr I[$r14] $r15	// HUB(0x86c) = val
-	mov $r14 -0x75ec
-	sethi $r14 0x400000
-	call #nv_wr32		// ROP(0xa14) = val
-	mov $r14 -0x5794
-	sethi $r14 0x410000
-	call #nv_wr32		// GPC(0x86c) = val
+	nv_iowr(NV_PGRAPH_FECS_UNK86C, 0, $r15)
+	nv_wr32(0x408a14, $r15)
+	nv_wr32(NV_PGRAPH_GPCX_GPCCS_UNK86C, $r15)
+	ret
+
+// In: $r15 NV_PGRAPH_FECS_MEM_CMD_*
+ctx_mem:
+	nv_iowr(NV_PGRAPH_FECS_MEM_CMD, 0, $r15)
+	ctx_mem_wait:
+		nv_iord($r15, NV_PGRAPH_FECS_MEM_CMD, 0)
+		or $r15 $r15
+		bra ne #ctx_mem_wait
 	ret
 
 // ctx_load - load's a channel's ctxctl data, and selects its vm
@@ -457,23 +458,14 @@ ctx_load:
 
 	// switch to channel, somewhat magic in parts..
 	mov $r10 12		// DONE_UNK12
-	call #wait_donez
-	mov $r1 0xa24
-	shl b32 $r1 6
-	iowr I[$r1 + 0x000] $r0	// 0x409a24
-	mov $r3 0xb00
-	shl b32 $r3 6
-	iowr I[$r3 + 0x100] $r2	// CHAN_NEXT
-	mov $r1 0xa0c
-	shl b32 $r1 6
-	mov $r4 7
-	iowr I[$r1 + 0x000] $r2 // MEM_CHAN
-	iowr I[$r1 + 0x100] $r4	// MEM_CMD
-	ctx_chan_wait_0:
-		iord $r4 I[$r1 + 0x100]
-		and $r4 0x1f
-		bra ne #ctx_chan_wait_0
-	iowr I[$r3 + 0x000] $r2	// CHAN_CUR
+	call(wait_donez)
+	clear b32 $r15
+	nv_iowr(0x409a24, 0, $r15)
+	nv_iowr(NV_PGRAPH_FECS_CHAN_NEXT, 0, $r2)
+	nv_iowr(NV_PGRAPH_FECS_MEM_CHAN, 0, $r2)
+	mov $r15 NV_PGRAPH_FECS_MEM_CMD_LOAD_CHAN
+	call(ctx_mem)
+	nv_iowr(NV_PGRAPH_FECS_CHAN_ADDR, 0, $r2)
 
 	// load channel header, fetch PGRAPH context pointer
 	mov $xtargets $r0
@@ -482,14 +474,10 @@ ctx_load:
 	add b32 $r2 2
 
 	trace_set(T_LCHAN)
-	mov $r1 0xa04
-	shl b32 $r1 6
-	iowr I[$r1 + 0x000] $r2		// MEM_BASE
-	mov $r1 0xa20
-	shl b32 $r1 6
-	mov $r2 0x0002
-	sethi $r2 0x80000000
-	iowr I[$r1 + 0x000] $r2		// MEM_TARGET = vram
+	nv_iowr(NV_PGRAPH_FECS_MEM_BASE, 0, $r2)
+	imm32($r2, NV_PGRAPH_FECS_MEM_TARGET_UNK31)
+	or  $r2 NV_PGRAPH_FECS_MEM_TARGET_AS_VRAM
+	nv_iowr(NV_PGRAPH_FECS_MEM_TARGET, 0, $r2)
 	mov $r1 0x10			// chan + 0x0210
 	mov $r2 #xfer_data
 	sethi $r2 0x00020000		// 16 bytes
@@ -507,13 +495,9 @@ ctx_load:
 
 	// set transfer base to start of context, and fetch context header
 	trace_set(T_LCTXH)
-	mov $r2 0xa04
-	shl b32 $r2 6
-	iowr I[$r2 + 0x000] $r1		// MEM_BASE
-	mov $r2 1
-	mov $r1 0xa20
-	shl b32 $r1 6
-	iowr I[$r1 + 0x000] $r2		// MEM_TARGET = vm
+	nv_iowr(NV_PGRAPH_FECS_MEM_BASE, 0, $r1)
+	mov $r2 NV_PGRAPH_FECS_MEM_TARGET_AS_VM
+	nv_iowr(NV_PGRAPH_FECS_MEM_TARGET, 0, $r2)
 	mov $r1 #chan_data
 	sethi $r1 0x00060000		// 256 bytes
 	xdld $r0 $r1
@@ -532,21 +516,15 @@ ctx_load:
 //
 ctx_chan:
 #if CHIPSET < GK100
-	call #ctx_4160s
+	call(ctx_4160s)
 #endif
-	call #ctx_load
+	call(ctx_load)
 	mov $r10 12			// DONE_UNK12
-	call #wait_donez
-	mov $r1 0xa10
-	shl b32 $r1 6
-	mov $r2 5
-	iowr I[$r1 + 0x000] $r2		// MEM_CMD = 5 (???)
-	ctx_chan_wait:
-		iord $r2 I[$r1 + 0x000]
-		or $r2 $r2
-		bra ne #ctx_chan_wait
+	call(wait_donez)
+	mov $r15 5 // MEM_CMD 5 ???
+	call(ctx_mem)
 #if CHIPSET < GK100
-	call #ctx_4160c
+	call(ctx_4160c)
 #endif
 	ret
 
@@ -562,9 +540,7 @@ ctx_chan:
 ctx_mmio_exec:
 	// set transfer base to be the mmio list
 	ld b32 $r3 D[$r0 + #chan_mmio_address]
-	mov $r2 0xa04
-	shl b32 $r2 6
-	iowr I[$r2 + 0x000] $r3		// MEM_BASE
+	nv_iowr(NV_PGRAPH_FECS_MEM_BASE, 0, $r3)
 
 	clear b32 $r3
 	ctx_mmio_loop:
@@ -580,7 +556,7 @@ ctx_mmio_exec:
 		ctx_mmio_pull:
 		ld b32 $r14 D[$r4 + #xfer_data + 0x00]
 		ld b32 $r15 D[$r4 + #xfer_data + 0x04]
-		call #nv_wr32
+		call(nv_wr32)
 
 		// next!
 		add b32 $r3 8
@@ -590,7 +566,7 @@ ctx_mmio_exec:
 	// set transfer base back to the current context
 	ctx_mmio_done:
 	ld b32 $r3 D[$r0 + #ctx_current]
-	iowr I[$r2 + 0x000] $r3		// MEM_BASE
+	nv_iowr(NV_PGRAPH_FECS_MEM_BASE, 0, $r3)
 
 	// disable the mmio list now, we don't need/want to execute it again
 	st b32 D[$r0 + #chan_mmio_count] $r0
@@ -610,12 +586,10 @@ ctx_mmio_exec:
 //
 ctx_xfer:
 	// according to mwk, some kind of wait for idle
-	mov $r15 0xc00
-	shl b32 $r15 6
 	mov $r14 4
-	iowr I[$r15 + 0x200] $r14
+	nv_iowr(0x409c08, 0, $r14)
 	ctx_xfer_idle:
-		iord $r14 I[$r15 + 0x000]
+		nv_iord($r14, 0x409c00, 0)
 		and $r14 0x2000
 		bra ne #ctx_xfer_idle
 
@@ -623,50 +597,42 @@ ctx_xfer:
 	bra $p2 #ctx_xfer_pre_load
 	ctx_xfer_pre:
 		mov $r15 0x10
-		call #ctx_86c
+		call(ctx_86c)
 #if CHIPSET < GK100
-		call #ctx_4160s
+		call(ctx_4160s)
 #endif
 		bra not $p1 #ctx_xfer_exec
 
 	ctx_xfer_pre_load:
 		mov $r15 2
-		call #ctx_4170s
-		call #ctx_4170w
-		call #ctx_redswitch
+		call(ctx_4170s)
+		call(ctx_4170w)
+		call(ctx_redswitch)
 		clear b32 $r15
-		call #ctx_4170s
-		call #ctx_load
+		call(ctx_4170s)
+		call(ctx_load)
 
 	// fetch context pointer, and initiate xfer on all GPCs
 	ctx_xfer_exec:
 	ld b32 $r1 D[$r0 + #ctx_current]
-	mov $r2 0x414
-	shl b32 $r2 6
-	iowr I[$r2 + 0x000] $r0	// BAR_STATUS = reset
-	mov $r14 -0x5b00
-	sethi $r14 0x410000
-	mov b32 $r15 $r1
-	call #nv_wr32		// GPC_BCAST_WRCMD_DATA = ctx pointer
-	add b32 $r14 4
+
+	clear b32 $r2
+	nv_iowr(NV_PGRAPH_FECS_BAR, 0, $r2)
+
+	nv_wr32(0x41a500, $r1)	// GPC_BCAST_WRCMD_DATA = ctx pointer
 	xbit $r15 $flags $p1
 	xbit $r2 $flags $p2
 	shl b32 $r2 1
 	or $r15 $r2
-	call #nv_wr32		// GPC_BCAST_WRCMD_CMD = GPC_XFER(type)
+	nv_wr32(0x41a504, $r15)	// GPC_BCAST_WRCMD_CMD = GPC_XFER(type)
 
 	// strands
-	mov $r1 0x4afc
-	sethi $r1 0x20000
-	mov $r2 0xc
-	iowr I[$r1] $r2		// STRAND_CMD(0x3f) = 0x0c
-	call #strand_wait
-	mov $r2 0x47fc
-	sethi $r2 0x20000
-	iowr I[$r2] $r0		// STRAND_FIRST_GENE(0x3f) = 0x00
-	xbit $r2 $flags $p1
-	add b32 $r2 3
-	iowr I[$r1] $r2		// STRAND_CMD(0x3f) = 0x03/0x04 (SAVE/LOAD)
+	call(strand_pre)
+	clear b32 $r2
+	nv_iowr(NV_PGRAPH_FECS_STRAND_SELECT, 0x3f, $r2)
+	xbit $r2 $flags $p1	// SAVE/LOAD
+	add b32 $r2 NV_PGRAPH_FECS_STRAND_CMD_SAVE
+	nv_iowr(NV_PGRAPH_FECS_STRAND_CMD, 0x3f, $r2)
 
 	// mmio context
 	xbit $r10 $flags $p1	// direction
@@ -675,48 +641,42 @@ ctx_xfer:
 	ld b32 $r12 D[$r0 + #hub_mmio_list_head]
 	ld b32 $r13 D[$r0 + #hub_mmio_list_tail]
 	mov $r14 0		// not multi
-	call #mmctx_xfer
+	call(mmctx_xfer)
 
 	// wait for GPCs to all complete
 	mov $r10 8		// DONE_BAR
-	call #wait_doneo
+	call(wait_doneo)
 
 	// wait for strand xfer to complete
-	call #strand_wait
+	call(strand_wait)
 
 	// post-op
 	bra $p1 #ctx_xfer_post
 		mov $r10 12		// DONE_UNK12
-		call #wait_donez
-		mov $r1 0xa10
-		shl b32 $r1 6
-		mov $r2 5
-		iowr I[$r1] $r2		// MEM_CMD
-		ctx_xfer_post_save_wait:
-			iord $r2 I[$r1]
-			or $r2 $r2
-			bra ne #ctx_xfer_post_save_wait
+		call(wait_donez)
+		mov $r15 5 // MEM_CMD 5 ???
+		call(ctx_mem)
 
 	bra $p2 #ctx_xfer_done
 	ctx_xfer_post:
 		mov $r15 2
-		call #ctx_4170s
+		call(ctx_4170s)
 		clear b32 $r15
-		call #ctx_86c
-		call #strand_post
-		call #ctx_4170w
+		call(ctx_86c)
+		call(strand_post)
+		call(ctx_4170w)
 		clear b32 $r15
-		call #ctx_4170s
+		call(ctx_4170s)
 
 		bra not $p1 #ctx_xfer_no_post_mmio
 		ld b32 $r1 D[$r0 + #chan_mmio_count]
 		or $r1 $r1
 		bra e #ctx_xfer_no_post_mmio
-			call #ctx_mmio_exec
+			call(ctx_mmio_exec)
 
 		ctx_xfer_no_post_mmio:
 #if CHIPSET < GK100
-		call #ctx_4160c
+		call(ctx_4160c)
 #endif
 
 	ctx_xfer_done:
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnv108.fuc5 b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnv108.fuc5
new file mode 100644
index 000000000000..7c5d25630fa8
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnv108.fuc5
@@ -0,0 +1,40 @@
+/*
+ * 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 <bskeggs@redhat.com>
+ */
+
+#define CHIPSET GK208
+#include "macros.fuc"
+
+.section #nv108_grhub_data
+#define INCLUDE_DATA
+#include "com.fuc"
+#include "hub.fuc"
+#undef INCLUDE_DATA
+
+.section #nv108_grhub_code
+#define INCLUDE_CODE
+bra #init
+#include "com.fuc"
+#include "hub.fuc"
+.align 256
+#undef INCLUDE_CODE
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnv108.fuc5.h b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnv108.fuc5.h
new file mode 100644
index 000000000000..4750984bf380
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnv108.fuc5.h
@@ -0,0 +1,916 @@
+uint32_t nv108_grhub_data[] = {
+/* 0x0000: hub_mmio_list_head */
+	0x00000300,
+/* 0x0004: hub_mmio_list_tail */
+	0x00000304,
+/* 0x0008: gpc_count */
+	0x00000000,
+/* 0x000c: rop_count */
+	0x00000000,
+/* 0x0010: cmd_queue */
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+/* 0x0058: ctx_current */
+	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,
+/* 0x0100: chan_data */
+/* 0x0100: chan_mmio_count */
+	0x00000000,
+/* 0x0104: chan_mmio_address */
+	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,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+/* 0x0200: xfer_data */
+	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,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+/* 0x0300: hub_mmio_list_base */
+	0x0417e91c,
+};
+
+uint32_t nv108_grhub_code[] = {
+	0x030e0ef5,
+/* 0x0004: queue_put */
+	0x9800d898,
+	0x86f001d9,
+	0xf489a408,
+	0x020f0b1b,
+	0x0002f87e,
+/* 0x001a: queue_put_next */
+	0x98c400f8,
+	0x0384b607,
+	0xb6008dbb,
+	0x8eb50880,
+	0x018fb500,
+	0xf00190b6,
+	0xd9b50f94,
+/* 0x0037: queue_get */
+	0xf400f801,
+	0xd8980131,
+	0x01d99800,
+	0x0bf489a4,
+	0x0789c421,
+	0xbb0394b6,
+	0x90b6009d,
+	0x009e9808,
+	0xb6019f98,
+	0x84f00180,
+	0x00d8b50f,
+/* 0x0063: queue_get_done */
+	0xf80132f4,
+/* 0x0065: nv_rd32 */
+	0xf0ecb200,
+	0x00801fc9,
+	0x0cf601ca,
+/* 0x0073: nv_rd32_wait */
+	0x8c04bd00,
+	0xcf01ca00,
+	0xccc800cc,
+	0xf61bf41f,
+	0xec7e060a,
+	0x008f0000,
+	0xffcf01cb,
+/* 0x008f: nv_wr32 */
+	0x8000f800,
+	0xf601cc00,
+	0x04bd000f,
+	0xc9f0ecb2,
+	0x1ec9f01f,
+	0x01ca0080,
+	0xbd000cf6,
+/* 0x00a9: nv_wr32_wait */
+	0xca008c04,
+	0x00cccf01,
+	0xf41fccc8,
+	0x00f8f61b,
+/* 0x00b8: wait_donez */
+	0x99f094bd,
+	0x37008000,
+	0x0009f602,
+	0x008004bd,
+	0x0af60206,
+/* 0x00cf: wait_donez_ne */
+	0x8804bd00,
+	0xcf010000,
+	0x8aff0088,
+	0xf61bf488,
+	0x99f094bd,
+	0x17008000,
+	0x0009f602,
+	0x00f804bd,
+/* 0x00ec: wait_doneo */
+	0x99f094bd,
+	0x37008000,
+	0x0009f602,
+	0x008004bd,
+	0x0af60206,
+/* 0x0103: wait_doneo_e */
+	0x8804bd00,
+	0xcf010000,
+	0x8aff0088,
+	0xf60bf488,
+	0x99f094bd,
+	0x17008000,
+	0x0009f602,
+	0x00f804bd,
+/* 0x0120: mmctx_size */
+/* 0x0122: nv_mmctx_size_loop */
+	0xe89894bd,
+	0x1a85b600,
+	0xb60180b6,
+	0x98bb0284,
+	0x04e0b600,
+	0x1bf4efa4,
+	0xf89fb2ec,
+/* 0x013d: mmctx_xfer */
+	0xf094bd00,
+	0x00800199,
+	0x09f60237,
+	0xbd04bd00,
+	0x05bbfd94,
+	0x800f0bf4,
+	0xf601c400,
+	0x04bd000b,
+/* 0x015f: mmctx_base_disabled */
+	0xfd0099f0,
+	0x0bf405ee,
+	0xc6008018,
+	0x000ef601,
+	0x008004bd,
+	0x0ff601c7,
+	0xf004bd00,
+/* 0x017a: mmctx_multi_disabled */
+	0xabc80199,
+	0x10b4b600,
+	0xc80cb9f0,
+	0xe4b601ae,
+	0x05befd11,
+	0x01c50080,
+	0xbd000bf6,
+/* 0x0195: mmctx_exec_loop */
+/* 0x0195: mmctx_wait_free */
+	0xc5008e04,
+	0x00eecf01,
+	0xf41fe4f0,
+	0xce98f60b,
+	0x05e9fd00,
+	0x01c80080,
+	0xbd000ef6,
+	0x04c0b604,
+	0x1bf4cda4,
+	0x02abc8df,
+/* 0x01bf: mmctx_fini_wait */
+	0x8b1c1bf4,
+	0xcf01c500,
+	0xb4f000bb,
+	0x10b4b01f,
+	0x0af31bf4,
+	0x00b87e02,
+	0x250ef400,
+/* 0x01d8: mmctx_stop */
+	0xb600abc8,
+	0xb9f010b4,
+	0x12b9f00c,
+	0x01c50080,
+	0xbd000bf6,
+/* 0x01ed: mmctx_stop_wait */
+	0xc5008b04,
+	0x00bbcf01,
+	0xf412bbc8,
+/* 0x01fa: mmctx_done */
+	0x94bdf61b,
+	0x800199f0,
+	0xf6021700,
+	0x04bd0009,
+/* 0x020a: strand_wait */
+	0xa0f900f8,
+	0xb87e020a,
+	0xa0fc0000,
+/* 0x0216: strand_pre */
+	0x0c0900f8,
+	0x024afc80,
+	0xbd0009f6,
+	0x020a7e04,
+/* 0x0227: strand_post */
+	0x0900f800,
+	0x4afc800d,
+	0x0009f602,
+	0x0a7e04bd,
+	0x00f80002,
+/* 0x0238: strand_set */
+	0xfc800f0c,
+	0x0cf6024f,
+	0x0c04bd00,
+	0x4afc800b,
+	0x000cf602,
+	0xfc8004bd,
+	0x0ef6024f,
+	0x0c04bd00,
+	0x4afc800a,
+	0x000cf602,
+	0x0a7e04bd,
+	0x00f80002,
+/* 0x0268: strand_ctx_init */
+	0x99f094bd,
+	0x37008003,
+	0x0009f602,
+	0x167e04bd,
+	0x030e0002,
+	0x0002387e,
+	0xfc80c4bd,
+	0x0cf60247,
+	0x0c04bd00,
+	0x4afc8001,
+	0x000cf602,
+	0x0a7e04bd,
+	0x0c920002,
+	0x46fc8001,
+	0x000cf602,
+	0x020c04bd,
+	0x024afc80,
+	0xbd000cf6,
+	0x020a7e04,
+	0x02277e00,
+	0x42008800,
+	0x20008902,
+	0x0099cf02,
+/* 0x02c7: ctx_init_strand_loop */
+	0xf608fe95,
+	0x8ef6008e,
+	0x808acf40,
+	0xb606a5b6,
+	0xeabb01a0,
+	0x0480b600,
+	0xf40192b6,
+	0xe4b6e81b,
+	0xf2efbc08,
+	0x99f094bd,
+	0x17008003,
+	0x0009f602,
+	0x00f804bd,
+/* 0x02f8: error */
+	0x02050080,
+	0xbd000ff6,
+	0x80010f04,
+	0xf6030700,
+	0x04bd000f,
+/* 0x030e: init */
+	0x04bd00f8,
+	0x410007fe,
+	0x11cf4200,
+	0x0911e700,
+	0x0814b601,
+	0x020014fe,
+	0x12004002,
+	0xbd0002f6,
+	0x05c94104,
+	0xbd0010fe,
+	0x07004024,
+	0xbd0002f6,
+	0x20034204,
+	0x01010080,
+	0xbd0002f6,
+	0x20044204,
+	0x01010480,
+	0xbd0002f6,
+	0x200b4204,
+	0x01010880,
+	0xbd0002f6,
+	0x200c4204,
+	0x01011c80,
+	0xbd0002f6,
+	0x01039204,
+	0x03090080,
+	0xbd0003f6,
+	0x87044204,
+	0xf6040040,
+	0x04bd0002,
+	0x00400402,
+	0x0002f603,
+	0x31f404bd,
+	0x96048e10,
+	0x00657e40,
+	0xc7feb200,
+	0x01b590f1,
+	0x1ff4f003,
+	0x01020fb5,
+	0x041fbb01,
+	0x800112b6,
+	0xf6010300,
+	0x04bd0001,
+	0x01040080,
+	0xbd0001f6,
+	0x01004104,
+	0x627e020f,
+	0x717e0006,
+	0x100f0006,
+	0x0006b37e,
+	0x98000e98,
+	0x207e010f,
+	0x14950001,
+	0xc0008008,
+	0x0004f601,
+	0x008004bd,
+	0x04f601c1,
+	0xb704bd00,
+	0xbb130030,
+	0xf5b6001f,
+	0xd3008002,
+	0x000ff601,
+	0x15b604bd,
+	0x0110b608,
+	0xb20814b6,
+	0x02687e1f,
+	0x001fbb00,
+	0x84020398,
+/* 0x041f: init_gpc */
+	0xb8502000,
+	0x0008044e,
+	0x8f7e1fb2,
+	0x4eb80000,
+	0xbd00010c,
+	0x008f7ef4,
+	0x044eb800,
+	0x8f7e0001,
+	0x4eb80000,
+	0x0f000100,
+	0x008f7e02,
+	0x004eb800,
+/* 0x044e: init_gpc_wait */
+	0x657e0008,
+	0xffc80000,
+	0xf90bf41f,
+	0x08044eb8,
+	0x00657e00,
+	0x001fbb00,
+	0x800040b7,
+	0xf40132b6,
+	0x000fb41b,
+	0x0006b37e,
+	0x627e000f,
+	0x00800006,
+	0x01f60201,
+	0xbd04bd00,
+	0x1f19f014,
+	0x02300080,
+	0xbd0001f6,
+/* 0x0491: main */
+	0x0031f404,
+	0x0d0028f4,
+	0x00377e10,
+	0xf401f400,
+	0x4001e4b1,
+	0x00c71bf5,
+	0x99f094bd,
+	0x37008004,
+	0x0009f602,
+	0x008104bd,
+	0x11cf02c0,
+	0xc1008200,
+	0x0022cf02,
+	0xf41f13c8,
+	0x23c8770b,
+	0x550bf41f,
+	0x12b220f9,
+	0x99f094bd,
+	0x37008007,
+	0x0009f602,
+	0x32f404bd,
+	0x0231f401,
+	0x0008367e,
+	0x99f094bd,
+	0x17008007,
+	0x0009f602,
+	0x20fc04bd,
+	0x99f094bd,
+	0x37008006,
+	0x0009f602,
+	0x31f404bd,
+	0x08367e01,
+	0xf094bd00,
+	0x00800699,
+	0x09f60217,
+	0xf404bd00,
+/* 0x0522: chsw_prev_no_next */
+	0x20f92f0e,
+	0x32f412b2,
+	0x0232f401,
+	0x0008367e,
+	0x008020fc,
+	0x02f602c0,
+	0xf404bd00,
+/* 0x053e: chsw_no_prev */
+	0x23c8130e,
+	0x0d0bf41f,
+	0xf40131f4,
+	0x367e0232,
+/* 0x054e: chsw_done */
+	0x01020008,
+	0x02c30080,
+	0xbd0002f6,
+	0xf094bd04,
+	0x00800499,
+	0x09f60217,
+	0xf504bd00,
+/* 0x056b: main_not_ctx_switch */
+	0xb0ff2a0e,
+	0x1bf401e4,
+	0x7ef2b20c,
+	0xf40007d6,
+/* 0x057a: main_not_ctx_chan */
+	0xe4b0400e,
+	0x2c1bf402,
+	0x99f094bd,
+	0x37008007,
+	0x0009f602,
+	0x32f404bd,
+	0x0232f401,
+	0x0008367e,
+	0x99f094bd,
+	0x17008007,
+	0x0009f602,
+	0x0ef404bd,
+/* 0x05a9: main_not_ctx_save */
+	0x10ef9411,
+	0x7e01f5f0,
+	0xf50002f8,
+/* 0x05b7: main_done */
+	0xbdfede0e,
+	0x1f29f024,
+	0x02300080,
+	0xbd0002f6,
+	0xcc0ef504,
+/* 0x05c9: ih */
+	0xfe80f9fe,
+	0x80f90188,
+	0xa0f990f9,
+	0xd0f9b0f9,
+	0xf0f9e0f9,
+	0x004a04bd,
+	0x00aacf02,
+	0xf404abc4,
+	0x100d230b,
+	0xcf1a004e,
+	0x004f00ee,
+	0x00ffcf19,
+	0x0000047e,
+	0x0400b0b7,
+	0x0040010e,
+	0x000ef61d,
+/* 0x060a: ih_no_fifo */
+	0xabe404bd,
+	0x0bf40100,
+	0x4e100d0c,
+	0x047e4001,
+/* 0x061a: ih_no_ctxsw */
+	0xabe40000,
+	0x0bf40400,
+	0x01004b10,
+	0x448ebfb2,
+	0x8f7e4001,
+/* 0x062e: ih_no_fwmthd */
+	0x044b0000,
+	0xffb0bd01,
+	0x0bf4b4ab,
+	0x0700800c,
+	0x000bf603,
+/* 0x0642: ih_no_other */
+	0x004004bd,
+	0x000af601,
+	0xf0fc04bd,
+	0xd0fce0fc,
+	0xa0fcb0fc,
+	0x80fc90fc,
+	0xfc0088fe,
+	0x0032f480,
+/* 0x0662: ctx_4170s */
+	0xf5f001f8,
+	0x8effb210,
+	0x7e404170,
+	0xf800008f,
+/* 0x0671: ctx_4170w */
+	0x41708e00,
+	0x00657e40,
+	0xf0ffb200,
+	0x1bf410f4,
+/* 0x0683: ctx_redswitch */
+	0x4e00f8f3,
+	0xe5f00200,
+	0x20e5f040,
+	0x8010e5f0,
+	0xf6018500,
+	0x04bd000e,
+/* 0x069a: ctx_redswitch_delay */
+	0xf2b6080f,
+	0xfd1bf401,
+	0x0400e5f1,
+	0x0100e5f1,
+	0x01850080,
+	0xbd000ef6,
+/* 0x06b3: ctx_86c */
+	0x8000f804,
+	0xf6022300,
+	0x04bd000f,
+	0x148effb2,
+	0x8f7e408a,
+	0xffb20000,
+	0x41a88c8e,
+	0x00008f7e,
+/* 0x06d2: ctx_mem */
+	0x008000f8,
+	0x0ff60284,
+/* 0x06db: ctx_mem_wait */
+	0x8f04bd00,
+	0xcf028400,
+	0xfffd00ff,
+	0xf61bf405,
+/* 0x06ea: ctx_load */
+	0x94bd00f8,
+	0x800599f0,
+	0xf6023700,
+	0x04bd0009,
+	0xb87e0c0a,
+	0xf4bd0000,
+	0x02890080,
+	0xbd000ff6,
+	0xc1008004,
+	0x0002f602,
+	0x008004bd,
+	0x02f60283,
+	0x0f04bd00,
+	0x06d27e07,
+	0xc0008000,
+	0x0002f602,
+	0x0bfe04bd,
+	0x1f2af000,
+	0xb60424b6,
+	0x94bd0220,
+	0x800899f0,
+	0xf6023700,
+	0x04bd0009,
+	0x02810080,
+	0xbd0002f6,
+	0x0000d204,
+	0x25f08000,
+	0x88008002,
+	0x0002f602,
+	0x100104bd,
+	0xf0020042,
+	0x12fa0223,
+	0xbd03f805,
+	0x0899f094,
+	0x02170080,
+	0xbd0009f6,
+	0x81019804,
+	0x981814b6,
+	0x25b68002,
+	0x0512fd08,
+	0xbd1601b5,
+	0x0999f094,
+	0x02370080,
+	0xbd0009f6,
+	0x81008004,
+	0x0001f602,
+	0x010204bd,
+	0x02880080,
+	0xbd0002f6,
+	0x01004104,
+	0xfa0613f0,
+	0x03f80501,
+	0x99f094bd,
+	0x17008009,
+	0x0009f602,
+	0x94bd04bd,
+	0x800599f0,
+	0xf6021700,
+	0x04bd0009,
+/* 0x07d6: ctx_chan */
+	0xea7e00f8,
+	0x0c0a0006,
+	0x0000b87e,
+	0xd27e050f,
+	0x00f80006,
+/* 0x07e8: ctx_mmio_exec */
+	0x80410398,
+	0xf6028100,
+	0x04bd0003,
+/* 0x07f6: ctx_mmio_loop */
+	0x34c434bd,
+	0x0e1bf4ff,
+	0xf0020045,
+	0x35fa0653,
+/* 0x0807: ctx_mmio_pull */
+	0x9803f805,
+	0x4f98804e,
+	0x008f7e81,
+	0x0830b600,
+	0xf40112b6,
+/* 0x081a: ctx_mmio_done */
+	0x0398df1b,
+	0x81008016,
+	0x0003f602,
+	0x00b504bd,
+	0x01004140,
+	0xfa0613f0,
+	0x03f80601,
+/* 0x0836: ctx_xfer */
+	0x040e00f8,
+	0x03020080,
+	0xbd000ef6,
+/* 0x0841: ctx_xfer_idle */
+	0x00008e04,
+	0x00eecf03,
+	0x2000e4f1,
+	0xf4f51bf4,
+	0x02f40611,
+/* 0x0855: ctx_xfer_pre */
+	0x7e100f0c,
+	0xf40006b3,
+/* 0x085e: ctx_xfer_pre_load */
+	0x020f1b11,
+	0x0006627e,
+	0x0006717e,
+	0x0006837e,
+	0x627ef4bd,
+	0xea7e0006,
+/* 0x0876: ctx_xfer_exec */
+	0x01980006,
+	0x8024bd16,
+	0xf6010500,
+	0x04bd0002,
+	0x008e1fb2,
+	0x8f7e41a5,
+	0xfcf00000,
+	0x022cf001,
+	0xfd0124b6,
+	0xffb205f2,
+	0x41a5048e,
+	0x00008f7e,
+	0x0002167e,
+	0xfc8024bd,
+	0x02f60247,
+	0xf004bd00,
+	0x20b6012c,
+	0x4afc8003,
+	0x0002f602,
+	0xacf004bd,
+	0x06a5f001,
+	0x0c98000b,
+	0x010d9800,
+	0x3d7e000e,
+	0x080a0001,
+	0x0000ec7e,
+	0x00020a7e,
+	0x0a1201f4,
+	0x00b87e0c,
+	0x7e050f00,
+	0xf40006d2,
+/* 0x08f2: ctx_xfer_post */
+	0x020f2d02,
+	0x0006627e,
+	0xb37ef4bd,
+	0x277e0006,
+	0x717e0002,
+	0xf4bd0006,
+	0x0006627e,
+	0x981011f4,
+	0x11fd4001,
+	0x070bf405,
+	0x0007e87e,
+/* 0x091c: ctx_xfer_no_post_mmio */
+/* 0x091c: ctx_xfer_done */
+	0x000000f8,
+	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,
+};
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnvc0.fuc.h b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnvc0.fuc.h
index b59f694c0423..132f684b1946 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnvc0.fuc.h
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnvc0.fuc.h
@@ -206,14 +206,14 @@ uint32_t nvc0_grhub_data[] = {
 };
 
 uint32_t nvc0_grhub_code[] = {
-	0x031b0ef5,
+	0x039b0ef5,
 /* 0x0004: queue_put */
 	0x9800d898,
 	0x86f001d9,
 	0x0489b808,
 	0xf00c1bf4,
 	0x21f502f7,
-	0x00f802fe,
+	0x00f8037e,
 /* 0x001c: queue_put_next */
 	0xb60798c4,
 	0x8dbb0384,
@@ -237,184 +237,214 @@ uint32_t nvc0_grhub_code[] = {
 /* 0x0066: queue_get_done */
 	0x00f80132,
 /* 0x0068: nv_rd32 */
-	0x0728b7f1,
-	0xb906b4b6,
-	0xc9f002ec,
-	0x00bcd01f,
-/* 0x0078: nv_rd32_wait */
-	0xc800bccf,
-	0x1bf41fcc,
-	0x06a7f0fa,
-	0x010921f5,
-	0xf840bfcf,
-/* 0x008d: nv_wr32 */
-	0x28b7f100,
-	0x06b4b607,
-	0xb980bfd0,
-	0xc9f002ec,
-	0x1ec9f01f,
-/* 0x00a3: nv_wr32_wait */
-	0xcf00bcd0,
-	0xccc800bc,
-	0xfa1bf41f,
-/* 0x00ae: watchdog_reset */
-	0x87f100f8,
-	0x84b60430,
-	0x1ff9f006,
-	0xf8008fd0,
-/* 0x00bd: watchdog_clear */
-	0x3087f100,
-	0x0684b604,
-	0xf80080d0,
-/* 0x00c9: wait_donez */
-	0xf094bd00,
-	0x07f10099,
-	0x03f00f00,
-	0x0009d002,
-	0x07f104bd,
-	0x03f00600,
-	0x000ad002,
-/* 0x00e6: wait_donez_ne */
-	0x87f104bd,
-	0x83f00000,
-	0x0088cf01,
-	0xf4888aff,
-	0x94bdf31b,
-	0xf10099f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0109: wait_doneo */
-	0xf094bd00,
+	0xf002ecb9,
+	0x07f11fc9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x007a: nv_rd32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0xa7f0f31b,
+	0x1021f506,
+	0x00f7f101,
+	0x01f3f0cb,
+	0xf800ffcf,
+/* 0x009d: nv_wr32 */
+	0x0007f100,
+	0x0103f0cc,
+	0xbd000fd0,
+	0x02ecb904,
+	0xf01fc9f0,
+	0x07f11ec9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x00be: nv_wr32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0x00f8f31b,
+/* 0x00d0: wait_donez */
+	0x99f094bd,
+	0x0007f100,
+	0x0203f00f,
+	0xbd0009d0,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x00ed: wait_donez_ne */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x1bf4888a,
+	0xf094bdf3,
 	0x07f10099,
-	0x03f00f00,
+	0x03f01700,
 	0x0009d002,
-	0x87f104bd,
-	0x84b60818,
-	0x008ad006,
-/* 0x0124: wait_doneo_e */
-	0x040087f1,
-	0xcf0684b6,
-	0x8aff0088,
-	0xf30bf488,
+	0x00f804bd,
+/* 0x0110: wait_doneo */
 	0x99f094bd,
 	0x0007f100,
-	0x0203f017,
+	0x0203f00f,
 	0xbd0009d0,
-/* 0x0147: mmctx_size */
-	0xbd00f804,
-/* 0x0149: nv_mmctx_size_loop */
-	0x00e89894,
-	0xb61a85b6,
-	0x84b60180,
-	0x0098bb02,
-	0xb804e0b6,
-	0x1bf404ef,
-	0x029fb9eb,
-/* 0x0166: mmctx_xfer */
-	0x94bd00f8,
-	0xf10199f0,
-	0xf00f0007,
-	0x09d00203,
-	0xf104bd00,
-	0xb6071087,
-	0x94bd0684,
-	0xf405bbfd,
-	0x8bd0090b,
-	0x0099f000,
-/* 0x018c: mmctx_base_disabled */
-	0xf405eefd,
-	0x8ed00c0b,
-	0xc08fd080,
-/* 0x019b: mmctx_multi_disabled */
-	0xb70199f0,
-	0xc8010080,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x012d: wait_doneo_e */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x0bf4888a,
+	0xf094bdf3,
+	0x07f10099,
+	0x03f01700,
+	0x0009d002,
+	0x00f804bd,
+/* 0x0150: mmctx_size */
+/* 0x0152: nv_mmctx_size_loop */
+	0xe89894bd,
+	0x1a85b600,
+	0xb60180b6,
+	0x98bb0284,
+	0x04e0b600,
+	0xf404efb8,
+	0x9fb9eb1b,
+/* 0x016f: mmctx_xfer */
+	0xbd00f802,
+	0x0199f094,
+	0x0f0007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0xbbfd94bd,
+	0x120bf405,
+	0xc40007f1,
+	0xd00103f0,
+	0x04bd000b,
+/* 0x0197: mmctx_base_disabled */
+	0xfd0099f0,
+	0x0bf405ee,
+	0x0007f11e,
+	0x0103f0c6,
+	0xbd000ed0,
+	0x0007f104,
+	0x0103f0c7,
+	0xbd000fd0,
+	0x0199f004,
+/* 0x01b8: mmctx_multi_disabled */
+	0xb600abc8,
+	0xb9f010b4,
+	0x01aec80c,
+	0xfd11e4b6,
+	0x07f105be,
+	0x03f0c500,
+	0x000bd001,
+/* 0x01d6: mmctx_exec_loop */
+/* 0x01d6: mmctx_wait_free */
+	0xe7f104bd,
+	0xe3f0c500,
+	0x00eecf01,
+	0xf41fe4f0,
+	0xce98f30b,
+	0x05e9fd00,
+	0xc80007f1,
+	0xd00103f0,
+	0x04bd000e,
+	0xb804c0b6,
+	0x1bf404cd,
+	0x02abc8d8,
+/* 0x0207: mmctx_fini_wait */
+	0xf11f1bf4,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x1fb4f000,
+	0xf410b4b0,
+	0xa7f0f01b,
+	0xd021f402,
+/* 0x0223: mmctx_stop */
+	0xc82b0ef4,
 	0xb4b600ab,
 	0x0cb9f010,
-	0xb601aec8,
-	0xbefd11e4,
-	0x008bd005,
-/* 0x01b4: mmctx_exec_loop */
-/* 0x01b4: mmctx_wait_free */
-	0xf0008ecf,
-	0x0bf41fe4,
-	0x00ce98fa,
-	0xd005e9fd,
-	0xc0b6c08e,
-	0x04cdb804,
-	0xc8e81bf4,
-	0x1bf402ab,
-/* 0x01d5: mmctx_fini_wait */
-	0x008bcf18,
-	0xb01fb4f0,
-	0x1bf410b4,
-	0x02a7f0f7,
-	0xf4c921f4,
-/* 0x01ea: mmctx_stop */
-	0xabc81b0e,
-	0x10b4b600,
-	0xf00cb9f0,
-	0x8bd012b9,
-/* 0x01f9: mmctx_stop_wait */
-	0x008bcf00,
-	0xf412bbc8,
-/* 0x0202: mmctx_done */
-	0x94bdfa1b,
-	0xf10199f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0215: strand_wait */
-	0xf0a0f900,
-	0x21f402a7,
-	0xf8a0fcc9,
-/* 0x0221: strand_pre */
-	0xfc87f100,
-	0x0283f04a,
-	0xd00c97f0,
-	0x21f50089,
-	0x00f80215,
-/* 0x0234: strand_post */
-	0x4afc87f1,
-	0xf00283f0,
-	0x89d00d97,
-	0x1521f500,
-/* 0x0247: strand_set */
-	0xf100f802,
-	0xf04ffca7,
-	0xaba202a3,
-	0xc7f00500,
-	0x00acd00f,
-	0xd00bc7f0,
-	0x21f500bc,
-	0xaed00215,
-	0x0ac7f000,
-	0xf500bcd0,
-	0xf8021521,
-/* 0x0271: strand_ctx_init */
-	0xf094bd00,
-	0x07f10399,
-	0x03f00f00,
+	0xf112b9f0,
+	0xf0c50007,
+	0x0bd00103,
+/* 0x023b: mmctx_stop_wait */
+	0xf104bd00,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x12bbc800,
+/* 0x024b: mmctx_done */
+	0xbdf31bf4,
+	0x0199f094,
+	0x170007f1,
+	0xd00203f0,
+	0x04bd0009,
+/* 0x025e: strand_wait */
+	0xa0f900f8,
+	0xf402a7f0,
+	0xa0fcd021,
+/* 0x026a: strand_pre */
+	0x97f000f8,
+	0xfc07f10c,
+	0x0203f04a,
+	0xbd0009d0,
+	0x5e21f504,
+/* 0x027f: strand_post */
+	0xf000f802,
+	0x07f10d97,
+	0x03f04afc,
 	0x0009d002,
 	0x21f504bd,
-	0xe7f00221,
-	0x4721f503,
-	0xfca7f102,
-	0x02a3f046,
-	0x0400aba0,
-	0xf040a0d0,
-	0xbcd001c7,
-	0x1521f500,
-	0x010c9202,
-	0xf000acd0,
-	0xbcd002c7,
-	0x1521f500,
-	0x3421f502,
-	0x8087f102,
-	0x0684b608,
-	0xb70089cf,
-	0x95220080,
-/* 0x02ca: ctx_init_strand_loop */
+	0x00f8025e,
+/* 0x0294: strand_set */
+	0xf10fc7f0,
+	0xf04ffc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f10bc7,
+	0x03f04afc,
+	0x000cd002,
+	0x07f104bd,
+	0x03f04ffc,
+	0x000ed002,
+	0xc7f004bd,
+	0xfc07f10a,
+	0x0203f04a,
+	0xbd000cd0,
+	0x5e21f504,
+/* 0x02d3: strand_ctx_init */
+	0xbd00f802,
+	0x0399f094,
+	0x0f0007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0x026a21f5,
+	0xf503e7f0,
+	0xbd029421,
+	0xfc07f1c4,
+	0x0203f047,
+	0xbd000cd0,
+	0x01c7f004,
+	0x4afc07f1,
+	0xd00203f0,
+	0x04bd000c,
+	0x025e21f5,
+	0xf1010c92,
+	0xf046fc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f102c7,
+	0x03f04afc,
+	0x000cd002,
+	0x21f504bd,
+	0x21f5025e,
+	0x87f1027f,
+	0x83f04200,
+	0x0097f102,
+	0x0293f020,
+	0x950099cf,
+/* 0x034a: ctx_init_strand_loop */
 	0x8ed008fe,
 	0x408ed000,
 	0xb6808acf,
@@ -428,7 +458,7 @@ uint32_t nvc0_grhub_code[] = {
 	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-/* 0x02fe: error */
+/* 0x037e: error */
 	0x07f100f8,
 	0x03f00500,
 	0x000fd002,
@@ -436,82 +466,117 @@ uint32_t nvc0_grhub_code[] = {
 	0x0007f101,
 	0x0303f007,
 	0xbd000fd0,
-/* 0x031b: init */
+/* 0x039b: init */
 	0xbd00f804,
-	0x0004fe04,
-	0xf10007fe,
-	0xf0120017,
-	0x12d00227,
-	0xb117f100,
-	0x0010fe05,
-	0x040017f1,
-	0xf1c010d0,
-	0xb6040437,
-	0x27f10634,
-	0x32d02003,
-	0x0427f100,
-	0x0132d020,
+	0x0007fe04,
+	0x420017f1,
+	0xcf0013f0,
+	0x11e70011,
+	0x14b60109,
+	0x0014fe08,
+	0xf10227f0,
+	0xf0120007,
+	0x02d00003,
+	0xf104bd00,
+	0xfe06c817,
+	0x24bd0010,
+	0x070007f1,
+	0xd00003f0,
+	0x04bd0002,
+	0x200327f1,
+	0x010007f1,
+	0xd00103f0,
+	0x04bd0002,
+	0x200427f1,
+	0x010407f1,
+	0xd00103f0,
+	0x04bd0002,
 	0x200b27f1,
-	0xf10232d0,
-	0xd0200c27,
-	0x27f10732,
-	0x24b60c24,
-	0x0003b906,
-	0xf10023d0,
+	0x010807f1,
+	0xd00103f0,
+	0x04bd0002,
+	0x200c27f1,
+	0x011c07f1,
+	0xd00103f0,
+	0x04bd0002,
+	0xf1010392,
+	0xf0090007,
+	0x03d00303,
+	0xf104bd00,
 	0xf0870427,
-	0x12d00023,
-	0x0012b700,
-	0x0427f001,
-	0xf40012d0,
-	0xe7f11031,
-	0xe3f09604,
-	0x6821f440,
-	0x8090f1c7,
-	0xf4f00301,
-	0x020f801f,
-	0xbb0117f0,
-	0x12b6041f,
-	0x0c27f101,
-	0x0624b604,
-	0xd00021d0,
-	0x17f14021,
-	0x0e980100,
-	0x010f9800,
-	0x014721f5,
-	0x070037f1,
-	0x950634b6,
-	0x34d00814,
-	0x4034d000,
-	0x130030b7,
-	0xb6001fbb,
-	0x3fd002f5,
-	0x0815b600,
-	0xb60110b6,
-	0x1fb90814,
-	0x7121f502,
-	0x001fbb02,
-	0xf1020398,
-	0xf0200047,
-/* 0x03f6: init_gpc */
-	0x4ea05043,
-	0x1fb90804,
-	0x8d21f402,
-	0x010c4ea0,
-	0x21f4f4bd,
-	0x044ea08d,
-	0x8d21f401,
-	0x01004ea0,
-	0xf402f7f0,
-	0x4ea08d21,
-/* 0x041e: init_gpc_wait */
-	0x21f40800,
-	0x1fffc868,
-	0xa0fa0bf4,
-	0xf408044e,
-	0x1fbb6821,
-	0x0040b700,
-	0x0132b680,
-	0xf1be1bf4,
+	0x07f10023,
+	0x03f00400,
+	0x0002d000,
+	0x27f004bd,
+	0x0007f104,
+	0x0003f003,
+	0xbd0002d0,
+	0x1031f404,
+	0x9604e7f1,
+	0xf440e3f0,
+	0xfeb96821,
+	0x90f1c702,
+	0xf0030180,
+	0x0f801ff4,
+	0x0117f002,
+	0xb6041fbb,
+	0x07f10112,
+	0x03f00300,
+	0x0001d001,
+	0x07f104bd,
+	0x03f00400,
+	0x0001d001,
+	0x17f104bd,
+	0xf7f00100,
+	0xb521f502,
+	0xc721f507,
+	0x10f7f007,
+	0x081421f5,
+	0x98000e98,
+	0x21f5010f,
+	0x14950150,
+	0x0007f108,
+	0x0103f0c0,
+	0xbd0004d0,
+	0x0007f104,
+	0x0103f0c1,
+	0xbd0004d0,
+	0x0030b704,
+	0x001fbb13,
+	0xf102f5b6,
+	0xf0d30007,
+	0x0fd00103,
+	0xb604bd00,
+	0x10b60815,
+	0x0814b601,
+	0xf5021fb9,
+	0xbb02d321,
+	0x0398001f,
+	0x0047f102,
+	0x5043f020,
+/* 0x04f4: init_gpc */
+	0x08044ea0,
+	0xf4021fb9,
+	0x4ea09d21,
+	0xf4bd010c,
+	0xa09d21f4,
+	0xf401044e,
+	0x4ea09d21,
+	0xf7f00100,
+	0x9d21f402,
+	0x08004ea0,
+/* 0x051c: init_gpc_wait */
+	0xc86821f4,
+	0x0bf41fff,
+	0x044ea0fa,
+	0x6821f408,
+	0xb7001fbb,
+	0xb6800040,
+	0x1bf40132,
+	0x00f7f0be,
+	0x081421f5,
+	0xf500f7f0,
+	0xf107b521,
 	0xf0010007,
 	0x01d00203,
 	0xbd04bd00,
@@ -519,402 +584,399 @@ uint32_t nvc0_grhub_code[] = {
 	0x080007f1,
 	0xd00203f0,
 	0x04bd0001,
-/* 0x0458: main */
+/* 0x0564: main */
 	0xf40031f4,
 	0xd7f00028,
 	0x3921f410,
 	0xb1f401f4,
 	0xf54001e4,
-	0xbd00de1b,
+	0xbd00e91b,
 	0x0499f094,
 	0x0f0007f1,
 	0xd00203f0,
 	0x04bd0009,
-	0x0b0017f1,
-	0xcf0614b6,
-	0x11cf4012,
-	0x1f13c800,
-	0x00870bf5,
-	0xf41f23c8,
-	0x20f9620b,
-	0xbd0212b9,
-	0x0799f094,
-	0x0f0007f1,
-	0xd00203f0,
-	0x04bd0009,
-	0xf40132f4,
-	0x21f50231,
-	0x94bd082f,
+	0xc00017f1,
+	0xcf0213f0,
+	0x27f10011,
+	0x23f0c100,
+	0x0022cf02,
+	0xf51f13c8,
+	0xc800890b,
+	0x0bf41f23,
+	0xb920f962,
+	0x94bd0212,
 	0xf10799f0,
-	0xf0170007,
+	0xf00f0007,
 	0x09d00203,
-	0xfc04bd00,
-	0xf094bd20,
-	0x07f10699,
-	0x03f00f00,
-	0x0009d002,
-	0x31f404bd,
-	0x2f21f501,
-	0xf094bd08,
-	0x07f10699,
+	0xf404bd00,
+	0x31f40132,
+	0xe821f502,
+	0xf094bd09,
+	0x07f10799,
 	0x03f01700,
 	0x0009d002,
-	0x0ef404bd,
-/* 0x04f9: chsw_prev_no_next */
-	0xb920f931,
-	0x32f40212,
-	0x0232f401,
-	0x082f21f5,
-	0x17f120fc,
-	0x14b60b00,
-	0x0012d006,
-/* 0x0517: chsw_no_prev */
-	0xc8130ef4,
-	0x0bf41f23,
-	0x0131f40d,
-	0xf50232f4,
-/* 0x0527: chsw_done */
-	0xf1082f21,
-	0xb60b0c17,
-	0x27f00614,
-	0x0012d001,
+	0x20fc04bd,
 	0x99f094bd,
-	0x0007f104,
+	0x0007f106,
+	0x0203f00f,
+	0xbd0009d0,
+	0x0131f404,
+	0x09e821f5,
+	0x99f094bd,
+	0x0007f106,
 	0x0203f017,
 	0xbd0009d0,
-	0x130ef504,
-/* 0x0549: main_not_ctx_switch */
-	0x01e4b0ff,
-	0xb90d1bf4,
-	0x21f502f2,
-	0x0ef407bb,
-/* 0x0559: main_not_ctx_chan */
-	0x02e4b046,
-	0xbd321bf4,
-	0x0799f094,
-	0x0f0007f1,
+	0x330ef404,
+/* 0x060c: chsw_prev_no_next */
+	0x12b920f9,
+	0x0132f402,
+	0xf50232f4,
+	0xfc09e821,
+	0x0007f120,
+	0x0203f0c0,
+	0xbd0002d0,
+	0x130ef404,
+/* 0x062c: chsw_no_prev */
+	0xf41f23c8,
+	0x31f40d0b,
+	0x0232f401,
+	0x09e821f5,
+/* 0x063c: chsw_done */
+	0xf10127f0,
+	0xf0c30007,
+	0x02d00203,
+	0xbd04bd00,
+	0x0499f094,
+	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-	0xf40132f4,
-	0x21f50232,
-	0x94bd082f,
+	0xff080ef5,
+/* 0x0660: main_not_ctx_switch */
+	0xf401e4b0,
+	0xf2b90d1b,
+	0x7821f502,
+	0x460ef409,
+/* 0x0670: main_not_ctx_chan */
+	0xf402e4b0,
+	0x94bd321b,
 	0xf10799f0,
-	0xf0170007,
+	0xf00f0007,
 	0x09d00203,
 	0xf404bd00,
-/* 0x058e: main_not_ctx_save */
-	0xef94110e,
-	0x01f5f010,
-	0x02fe21f5,
-	0xfec00ef5,
-/* 0x059c: main_done */
-	0x29f024bd,
-	0x0007f11f,
-	0x0203f008,
-	0xbd0002d0,
-	0xab0ef504,
-/* 0x05b1: ih */
-	0xfe80f9fe,
-	0x80f90188,
-	0xa0f990f9,
-	0xd0f9b0f9,
-	0xf0f9e0f9,
-	0x0acf04bd,
-	0x04abc480,
-	0xf11d0bf4,
-	0xf01900b7,
-	0xbecf10d7,
-	0x00bfcf40,
+	0x32f40132,
+	0xe821f502,
+	0xf094bd09,
+	0x07f10799,
+	0x03f01700,
+	0x0009d002,
+	0x0ef404bd,
+/* 0x06a5: main_not_ctx_save */
+	0x10ef9411,
+	0xf501f5f0,
+	0xf5037e21,
+/* 0x06b3: main_done */
+	0xbdfeb50e,
+	0x1f29f024,
+	0x080007f1,
+	0xd00203f0,
+	0x04bd0002,
+	0xfea00ef5,
+/* 0x06c8: ih */
+	0x88fe80f9,
+	0xf980f901,
+	0xf9a0f990,
+	0xf9d0f9b0,
+	0xbdf0f9e0,
+	0x00a7f104,
+	0x00a3f002,
+	0xc400aacf,
+	0x0bf404ab,
+	0x10d7f030,
+	0x1a00e7f1,
+	0xcf00e3f0,
+	0xf7f100ee,
+	0xf3f01900,
+	0x00ffcf00,
 	0xb70421f4,
 	0xf00400b0,
-	0xbed001e7,
-/* 0x05e9: ih_no_fifo */
-	0x00abe400,
-	0x0d0bf401,
-	0xf110d7f0,
-	0xf44001e7,
-/* 0x05fa: ih_no_ctxsw */
-	0xb7f10421,
-	0xb0bd0104,
-	0xf4b4abff,
-	0xa7f10d0b,
-	0xa4b60c1c,
-	0x00abd006,
-/* 0x0610: ih_no_other */
-	0xfc400ad0,
+	0x07f101e7,
+	0x03f01d00,
+	0x000ed000,
+/* 0x071a: ih_no_fifo */
+	0xabe404bd,
+	0x0bf40100,
+	0x10d7f00d,
+	0x4001e7f1,
+/* 0x072b: ih_no_ctxsw */
+	0xe40421f4,
+	0xf40400ab,
+	0xb7f1140b,
+	0xbfb90100,
+	0x44e7f102,
+	0x40e3f001,
+/* 0x0743: ih_no_fwmthd */
+	0xf19d21f4,
+	0xbd0104b7,
+	0xb4abffb0,
+	0xf10f0bf4,
+	0xf0070007,
+	0x0bd00303,
+/* 0x075b: ih_no_other */
+	0xf104bd00,
+	0xf0010007,
+	0x0ad00003,
+	0xfc04bd00,
 	0xfce0fcf0,
 	0xfcb0fcd0,
 	0xfc90fca0,
 	0x0088fe80,
 	0x32f480fc,
-/* 0x062b: ctx_4160s */
-	0xf101f800,
-	0xf04160e7,
-	0xf7f040e3,
-	0x8d21f401,
-/* 0x0638: ctx_4160s_wait */
-	0xc86821f4,
-	0x0bf404ff,
-/* 0x0643: ctx_4160c */
-	0xf100f8fa,
+/* 0x077f: ctx_4160s */
+	0xf001f800,
+	0xffb901f7,
+	0x60e7f102,
+	0x40e3f041,
+/* 0x078f: ctx_4160s_wait */
+	0xf19d21f4,
 	0xf04160e7,
-	0xf4bd40e3,
-	0xf88d21f4,
-/* 0x0651: ctx_4170s */
-	0x70e7f100,
+	0x21f440e3,
+	0x02ffb968,
+	0xf404ffc8,
+	0x00f8f00b,
+/* 0x07a4: ctx_4160c */
+	0xffb9f4bd,
+	0x60e7f102,
 	0x40e3f041,
-	0xf410f5f0,
-	0x00f88d21,
-/* 0x0660: ctx_4170w */
-	0x4170e7f1,
-	0xf440e3f0,
-	0xf4f06821,
-	0xf31bf410,
-/* 0x0672: ctx_redswitch */
-	0xe7f100f8,
-	0xe4b60614,
-	0x70f7f106,
-	0x00efd002,
-/* 0x0683: ctx_redswitch_delay */
-	0xb608f7f0,
-	0x1bf401f2,
-	0x70f7f1fd,
-	0x00efd007,
-/* 0x0692: ctx_86c */
-	0xe7f100f8,
-	0xe4b6086c,
-	0x00efd006,
-	0x8a14e7f1,
-	0xf440e3f0,
-	0xe7f18d21,
-	0xe3f0a86c,
-	0x8d21f441,
-/* 0x06b2: ctx_load */
+	0xf89d21f4,
+/* 0x07b5: ctx_4170s */
+	0x10f5f000,
+	0xf102ffb9,
+	0xf04170e7,
+	0x21f440e3,
+/* 0x07c7: ctx_4170w */
+	0xf100f89d,
+	0xf04170e7,
+	0x21f440e3,
+	0x02ffb968,
+	0xf410f4f0,
+	0x00f8f01b,
+/* 0x07dc: ctx_redswitch */
+	0x0200e7f1,
+	0xf040e5f0,
+	0xe5f020e5,
+	0x0007f110,
+	0x0103f085,
+	0xbd000ed0,
+	0x08f7f004,
+/* 0x07f8: ctx_redswitch_delay */
+	0xf401f2b6,
+	0xe5f1fd1b,
+	0xe5f10400,
+	0x07f10100,
+	0x03f08500,
+	0x000ed001,
+	0x00f804bd,
+/* 0x0814: ctx_86c */
+	0x1b0007f1,
+	0xd00203f0,
+	0x04bd000f,
+	0xf102ffb9,
+	0xf08a14e7,
+	0x21f440e3,
+	0x02ffb99d,
+	0xa86ce7f1,
+	0xf441e3f0,
+	0x00f89d21,
+/* 0x083c: ctx_mem */
+	0x840007f1,
+	0xd00203f0,
+	0x04bd000f,
+/* 0x0848: ctx_mem_wait */
+	0x8400f7f1,
+	0xcf02f3f0,
+	0xfffd00ff,
+	0xf31bf405,
+/* 0x085a: ctx_load */
 	0x94bd00f8,
 	0xf10599f0,
 	0xf00f0007,
 	0x09d00203,
 	0xf004bd00,
 	0x21f40ca7,
-	0x2417f1c9,
-	0x0614b60a,
-	0xf10010d0,
-	0xb60b0037,
-	0x32d00634,
-	0x0c17f140,
-	0x0614b60a,
-	0xd00747f0,
-	0x14d00012,
-/* 0x06ed: ctx_chan_wait_0 */
-	0x4014cf40,
-	0xf41f44f0,
-	0x32d0fa1b,
-	0x000bfe00,
-	0xb61f2af0,
-	0x20b60424,
-	0xf094bd02,
+	0xf1f4bdd0,
+	0xf0890007,
+	0x0fd00203,
+	0xf104bd00,
+	0xf0c10007,
+	0x02d00203,
+	0xf104bd00,
+	0xf0830007,
+	0x02d00203,
+	0xf004bd00,
+	0x21f507f7,
+	0x07f1083c,
+	0x03f0c000,
+	0x0002d002,
+	0x0bfe04bd,
+	0x1f2af000,
+	0xb60424b6,
+	0x94bd0220,
+	0xf10899f0,
+	0xf00f0007,
+	0x09d00203,
+	0xf104bd00,
+	0xf0810007,
+	0x02d00203,
+	0xf104bd00,
+	0xf1000027,
+	0xf0800023,
+	0x07f10225,
+	0x03f08800,
+	0x0002d002,
+	0x17f004bd,
+	0x0027f110,
+	0x0223f002,
+	0xf80512fa,
+	0xf094bd03,
 	0x07f10899,
-	0x03f00f00,
+	0x03f01700,
 	0x0009d002,
-	0x17f104bd,
-	0x14b60a04,
-	0x0012d006,
-	0x0a2017f1,
-	0xf00614b6,
-	0x23f10227,
-	0x12d08000,
-	0x1017f000,
-	0x020027f1,
-	0xfa0223f0,
-	0x03f80512,
+	0x019804bd,
+	0x1814b681,
+	0xb6800298,
+	0x12fd0825,
+	0x16018005,
 	0x99f094bd,
-	0x0007f108,
-	0x0203f017,
+	0x0007f109,
+	0x0203f00f,
 	0xbd0009d0,
-	0x81019804,
-	0x981814b6,
-	0x25b68002,
-	0x0512fd08,
-	0xbd160180,
-	0x0999f094,
-	0x0f0007f1,
-	0xd00203f0,
-	0x04bd0009,
-	0x0a0427f1,
-	0xd00624b6,
-	0x27f00021,
-	0x2017f101,
-	0x0614b60a,
-	0xf10012d0,
-	0xf0010017,
-	0x01fa0613,
-	0xbd03f805,
-	0x0999f094,
-	0x170007f1,
+	0x0007f104,
+	0x0203f081,
+	0xbd0001d0,
+	0x0127f004,
+	0x880007f1,
 	0xd00203f0,
-	0x04bd0009,
+	0x04bd0002,
+	0x010017f1,
+	0xfa0613f0,
+	0x03f80501,
 	0x99f094bd,
-	0x0007f105,
+	0x0007f109,
 	0x0203f017,
 	0xbd0009d0,
-/* 0x07bb: ctx_chan */
-	0xf500f804,
-	0xf5062b21,
-	0xf006b221,
-	0x21f40ca7,
-	0x1017f1c9,
-	0x0614b60a,
-	0xd00527f0,
-/* 0x07d6: ctx_chan_wait */
-	0x12cf0012,
-	0x0522fd00,
-	0xf5fa1bf4,
-	0xf8064321,
-/* 0x07e5: ctx_mmio_exec */
-	0x41039800,
-	0x0a0427f1,
-	0xd00624b6,
-	0x34bd0023,
-/* 0x07f4: ctx_mmio_loop */
+	0xf094bd04,
+	0x07f10599,
+	0x03f01700,
+	0x0009d002,
+	0x00f804bd,
+/* 0x0978: ctx_chan */
+	0x077f21f5,
+	0x085a21f5,
+	0xf40ca7f0,
+	0xf7f0d021,
+	0x3c21f505,
+	0xa421f508,
+/* 0x0993: ctx_mmio_exec */
+	0x9800f807,
+	0x07f14103,
+	0x03f08100,
+	0x0003d002,
+	0x34bd04bd,
+/* 0x09a4: ctx_mmio_loop */
 	0xf4ff34c4,
 	0x57f10f1b,
 	0x53f00200,
 	0x0535fa06,
-/* 0x0806: ctx_mmio_pull */
+/* 0x09b6: ctx_mmio_pull */
 	0x4e9803f8,
 	0x814f9880,
-	0xb68d21f4,
+	0xb69d21f4,
 	0x12b60830,
 	0xdf1bf401,
-/* 0x0818: ctx_mmio_done */
-	0xd0160398,
-	0x00800023,
-	0x0017f140,
-	0x0613f001,
-	0xf80601fa,
-/* 0x082f: ctx_xfer */
-	0xf100f803,
-	0xb60c00f7,
-	0xe7f006f4,
-	0x80fed004,
-/* 0x083c: ctx_xfer_idle */
-	0xf100fecf,
-	0xf42000e4,
-	0x11f4f91b,
-	0x1102f406,
-/* 0x084c: ctx_xfer_pre */
-	0xf510f7f0,
-	0xf5069221,
-	0xf4062b21,
-/* 0x085a: ctx_xfer_pre_load */
-	0xf7f01c11,
-	0x5121f502,
-	0x6021f506,
-	0x7221f506,
-	0xf5f4bd06,
-	0xf5065121,
-/* 0x0873: ctx_xfer_exec */
-	0x9806b221,
-	0x27f11601,
-	0x24b60414,
-	0x0020d006,
-	0xa500e7f1,
-	0xb941e3f0,
-	0x21f4021f,
-	0x04e0b68d,
-	0xf001fcf0,
-	0x24b6022c,
-	0x05f2fd01,
-	0xf18d21f4,
-	0xf04afc17,
-	0x27f00213,
-	0x0012d00c,
-	0x021521f5,
-	0x47fc27f1,
-	0xd00223f0,
-	0x2cf00020,
+/* 0x09c8: ctx_mmio_done */
+	0xf1160398,
+	0xf0810007,
+	0x03d00203,
+	0x8004bd00,
+	0x17f14000,
+	0x13f00100,
+	0x0601fa06,
+	0x00f803f8,
+/* 0x09e8: ctx_xfer */
+	0xf104e7f0,
+	0xf0020007,
+	0x0ed00303,
+/* 0x09f7: ctx_xfer_idle */
+	0xf104bd00,
+	0xf00000e7,
+	0xeecf03e3,
+	0x00e4f100,
+	0xf21bf420,
+	0xf40611f4,
+/* 0x0a0e: ctx_xfer_pre */
+	0xf7f01102,
+	0x1421f510,
+	0x7f21f508,
+	0x1c11f407,
+/* 0x0a1c: ctx_xfer_pre_load */
+	0xf502f7f0,
+	0xf507b521,
+	0xf507c721,
+	0xbd07dc21,
+	0xb521f5f4,
+	0x5a21f507,
+/* 0x0a35: ctx_xfer_exec */
+	0x16019808,
+	0x07f124bd,
+	0x03f00500,
+	0x0002d001,
+	0x1fb904bd,
+	0x00e7f102,
+	0x41e3f0a5,
+	0xf09d21f4,
+	0x2cf001fc,
+	0x0124b602,
+	0xb905f2fd,
+	0xe7f102ff,
+	0xe3f0a504,
+	0x9d21f441,
+	0x026a21f5,
+	0x07f124bd,
+	0x03f047fc,
+	0x0002d002,
+	0x2cf004bd,
 	0x0320b601,
-	0xf00012d0,
-	0xa5f001ac,
-	0x00b7f006,
-	0x98000c98,
-	0xe7f0010d,
-	0x6621f500,
-	0x08a7f001,
-	0x010921f5,
-	0x021521f5,
-	0xf02201f4,
-	0x21f40ca7,
-	0x1017f1c9,
-	0x0614b60a,
-	0xd00527f0,
-/* 0x08fa: ctx_xfer_post_save_wait */
-	0x12cf0012,
-	0x0522fd00,
-	0xf4fa1bf4,
-/* 0x0906: ctx_xfer_post */
-	0xf7f03202,
-	0x5121f502,
-	0xf5f4bd06,
-	0xf5069221,
-	0xf5023421,
-	0xbd066021,
-	0x5121f5f4,
-	0x1011f406,
-	0xfd400198,
-	0x0bf40511,
-	0xe521f507,
-/* 0x0931: ctx_xfer_no_post_mmio */
-	0x4321f507,
-/* 0x0935: ctx_xfer_done */
-	0x0000f806,
-	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,
+	0x4afc07f1,
+	0xd00203f0,
+	0x04bd0002,
+	0xf001acf0,
+	0xb7f006a5,
+	0x000c9800,
+	0xf0010d98,
+	0x21f500e7,
+	0xa7f0016f,
+	0x1021f508,
+	0x5e21f501,
+	0x1301f402,
+	0xf40ca7f0,
+	0xf7f0d021,
+	0x3c21f505,
+	0x3202f408,
+/* 0x0ac4: ctx_xfer_post */
+	0xf502f7f0,
+	0xbd07b521,
+	0x1421f5f4,
+	0x7f21f508,
+	0xc721f502,
+	0xf5f4bd07,
+	0xf407b521,
+	0x01981011,
+	0x0511fd40,
+	0xf5070bf4,
+/* 0x0aef: ctx_xfer_no_post_mmio */
+	0xf5099321,
+/* 0x0af3: ctx_xfer_done */
+	0xf807a421,
 	0x00000000,
 	0x00000000,
 	0x00000000,
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnvd7.fuc.h b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnvd7.fuc.h
index a1b9f763996a..84af82418987 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnvd7.fuc.h
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnvd7.fuc.h
@@ -206,14 +206,14 @@ uint32_t nvd7_grhub_data[] = {
 };
 
 uint32_t nvd7_grhub_code[] = {
-	0x031b0ef5,
+	0x039b0ef5,
 /* 0x0004: queue_put */
 	0x9800d898,
 	0x86f001d9,
 	0x0489b808,
 	0xf00c1bf4,
 	0x21f502f7,
-	0x00f802fe,
+	0x00f8037e,
 /* 0x001c: queue_put_next */
 	0xb60798c4,
 	0x8dbb0384,
@@ -237,184 +237,214 @@ uint32_t nvd7_grhub_code[] = {
 /* 0x0066: queue_get_done */
 	0x00f80132,
 /* 0x0068: nv_rd32 */
-	0x0728b7f1,
-	0xb906b4b6,
-	0xc9f002ec,
-	0x00bcd01f,
-/* 0x0078: nv_rd32_wait */
-	0xc800bccf,
-	0x1bf41fcc,
-	0x06a7f0fa,
-	0x010921f5,
-	0xf840bfcf,
-/* 0x008d: nv_wr32 */
-	0x28b7f100,
-	0x06b4b607,
-	0xb980bfd0,
-	0xc9f002ec,
-	0x1ec9f01f,
-/* 0x00a3: nv_wr32_wait */
-	0xcf00bcd0,
-	0xccc800bc,
-	0xfa1bf41f,
-/* 0x00ae: watchdog_reset */
-	0x87f100f8,
-	0x84b60430,
-	0x1ff9f006,
-	0xf8008fd0,
-/* 0x00bd: watchdog_clear */
-	0x3087f100,
-	0x0684b604,
-	0xf80080d0,
-/* 0x00c9: wait_donez */
-	0xf094bd00,
-	0x07f10099,
-	0x03f00f00,
-	0x0009d002,
-	0x07f104bd,
-	0x03f00600,
-	0x000ad002,
-/* 0x00e6: wait_donez_ne */
-	0x87f104bd,
-	0x83f00000,
-	0x0088cf01,
-	0xf4888aff,
-	0x94bdf31b,
-	0xf10099f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0109: wait_doneo */
-	0xf094bd00,
+	0xf002ecb9,
+	0x07f11fc9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x007a: nv_rd32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0xa7f0f31b,
+	0x1021f506,
+	0x00f7f101,
+	0x01f3f0cb,
+	0xf800ffcf,
+/* 0x009d: nv_wr32 */
+	0x0007f100,
+	0x0103f0cc,
+	0xbd000fd0,
+	0x02ecb904,
+	0xf01fc9f0,
+	0x07f11ec9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x00be: nv_wr32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0x00f8f31b,
+/* 0x00d0: wait_donez */
+	0x99f094bd,
+	0x0007f100,
+	0x0203f00f,
+	0xbd0009d0,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x00ed: wait_donez_ne */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x1bf4888a,
+	0xf094bdf3,
 	0x07f10099,
-	0x03f00f00,
+	0x03f01700,
 	0x0009d002,
-	0x87f104bd,
-	0x84b60818,
-	0x008ad006,
-/* 0x0124: wait_doneo_e */
-	0x040087f1,
-	0xcf0684b6,
-	0x8aff0088,
-	0xf30bf488,
+	0x00f804bd,
+/* 0x0110: wait_doneo */
 	0x99f094bd,
 	0x0007f100,
-	0x0203f017,
+	0x0203f00f,
 	0xbd0009d0,
-/* 0x0147: mmctx_size */
-	0xbd00f804,
-/* 0x0149: nv_mmctx_size_loop */
-	0x00e89894,
-	0xb61a85b6,
-	0x84b60180,
-	0x0098bb02,
-	0xb804e0b6,
-	0x1bf404ef,
-	0x029fb9eb,
-/* 0x0166: mmctx_xfer */
-	0x94bd00f8,
-	0xf10199f0,
-	0xf00f0007,
-	0x09d00203,
-	0xf104bd00,
-	0xb6071087,
-	0x94bd0684,
-	0xf405bbfd,
-	0x8bd0090b,
-	0x0099f000,
-/* 0x018c: mmctx_base_disabled */
-	0xf405eefd,
-	0x8ed00c0b,
-	0xc08fd080,
-/* 0x019b: mmctx_multi_disabled */
-	0xb70199f0,
-	0xc8010080,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x012d: wait_doneo_e */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x0bf4888a,
+	0xf094bdf3,
+	0x07f10099,
+	0x03f01700,
+	0x0009d002,
+	0x00f804bd,
+/* 0x0150: mmctx_size */
+/* 0x0152: nv_mmctx_size_loop */
+	0xe89894bd,
+	0x1a85b600,
+	0xb60180b6,
+	0x98bb0284,
+	0x04e0b600,
+	0xf404efb8,
+	0x9fb9eb1b,
+/* 0x016f: mmctx_xfer */
+	0xbd00f802,
+	0x0199f094,
+	0x0f0007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0xbbfd94bd,
+	0x120bf405,
+	0xc40007f1,
+	0xd00103f0,
+	0x04bd000b,
+/* 0x0197: mmctx_base_disabled */
+	0xfd0099f0,
+	0x0bf405ee,
+	0x0007f11e,
+	0x0103f0c6,
+	0xbd000ed0,
+	0x0007f104,
+	0x0103f0c7,
+	0xbd000fd0,
+	0x0199f004,
+/* 0x01b8: mmctx_multi_disabled */
+	0xb600abc8,
+	0xb9f010b4,
+	0x01aec80c,
+	0xfd11e4b6,
+	0x07f105be,
+	0x03f0c500,
+	0x000bd001,
+/* 0x01d6: mmctx_exec_loop */
+/* 0x01d6: mmctx_wait_free */
+	0xe7f104bd,
+	0xe3f0c500,
+	0x00eecf01,
+	0xf41fe4f0,
+	0xce98f30b,
+	0x05e9fd00,
+	0xc80007f1,
+	0xd00103f0,
+	0x04bd000e,
+	0xb804c0b6,
+	0x1bf404cd,
+	0x02abc8d8,
+/* 0x0207: mmctx_fini_wait */
+	0xf11f1bf4,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x1fb4f000,
+	0xf410b4b0,
+	0xa7f0f01b,
+	0xd021f402,
+/* 0x0223: mmctx_stop */
+	0xc82b0ef4,
 	0xb4b600ab,
 	0x0cb9f010,
-	0xb601aec8,
-	0xbefd11e4,
-	0x008bd005,
-/* 0x01b4: mmctx_exec_loop */
-/* 0x01b4: mmctx_wait_free */
-	0xf0008ecf,
-	0x0bf41fe4,
-	0x00ce98fa,
-	0xd005e9fd,
-	0xc0b6c08e,
-	0x04cdb804,
-	0xc8e81bf4,
-	0x1bf402ab,
-/* 0x01d5: mmctx_fini_wait */
-	0x008bcf18,
-	0xb01fb4f0,
-	0x1bf410b4,
-	0x02a7f0f7,
-	0xf4c921f4,
-/* 0x01ea: mmctx_stop */
-	0xabc81b0e,
-	0x10b4b600,
-	0xf00cb9f0,
-	0x8bd012b9,
-/* 0x01f9: mmctx_stop_wait */
-	0x008bcf00,
-	0xf412bbc8,
-/* 0x0202: mmctx_done */
-	0x94bdfa1b,
-	0xf10199f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0215: strand_wait */
-	0xf0a0f900,
-	0x21f402a7,
-	0xf8a0fcc9,
-/* 0x0221: strand_pre */
-	0xfc87f100,
-	0x0283f04a,
-	0xd00c97f0,
-	0x21f50089,
-	0x00f80215,
-/* 0x0234: strand_post */
-	0x4afc87f1,
-	0xf00283f0,
-	0x89d00d97,
-	0x1521f500,
-/* 0x0247: strand_set */
-	0xf100f802,
-	0xf04ffca7,
-	0xaba202a3,
-	0xc7f00500,
-	0x00acd00f,
-	0xd00bc7f0,
-	0x21f500bc,
-	0xaed00215,
-	0x0ac7f000,
-	0xf500bcd0,
-	0xf8021521,
-/* 0x0271: strand_ctx_init */
-	0xf094bd00,
-	0x07f10399,
-	0x03f00f00,
+	0xf112b9f0,
+	0xf0c50007,
+	0x0bd00103,
+/* 0x023b: mmctx_stop_wait */
+	0xf104bd00,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x12bbc800,
+/* 0x024b: mmctx_done */
+	0xbdf31bf4,
+	0x0199f094,
+	0x170007f1,
+	0xd00203f0,
+	0x04bd0009,
+/* 0x025e: strand_wait */
+	0xa0f900f8,
+	0xf402a7f0,
+	0xa0fcd021,
+/* 0x026a: strand_pre */
+	0x97f000f8,
+	0xfc07f10c,
+	0x0203f04a,
+	0xbd0009d0,
+	0x5e21f504,
+/* 0x027f: strand_post */
+	0xf000f802,
+	0x07f10d97,
+	0x03f04afc,
 	0x0009d002,
 	0x21f504bd,
-	0xe7f00221,
-	0x4721f503,
-	0xfca7f102,
-	0x02a3f046,
-	0x0400aba0,
-	0xf040a0d0,
-	0xbcd001c7,
-	0x1521f500,
-	0x010c9202,
-	0xf000acd0,
-	0xbcd002c7,
-	0x1521f500,
-	0x3421f502,
-	0x8087f102,
-	0x0684b608,
-	0xb70089cf,
-	0x95220080,
-/* 0x02ca: ctx_init_strand_loop */
+	0x00f8025e,
+/* 0x0294: strand_set */
+	0xf10fc7f0,
+	0xf04ffc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f10bc7,
+	0x03f04afc,
+	0x000cd002,
+	0x07f104bd,
+	0x03f04ffc,
+	0x000ed002,
+	0xc7f004bd,
+	0xfc07f10a,
+	0x0203f04a,
+	0xbd000cd0,
+	0x5e21f504,
+/* 0x02d3: strand_ctx_init */
+	0xbd00f802,
+	0x0399f094,
+	0x0f0007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0x026a21f5,
+	0xf503e7f0,
+	0xbd029421,
+	0xfc07f1c4,
+	0x0203f047,
+	0xbd000cd0,
+	0x01c7f004,
+	0x4afc07f1,
+	0xd00203f0,
+	0x04bd000c,
+	0x025e21f5,
+	0xf1010c92,
+	0xf046fc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f102c7,
+	0x03f04afc,
+	0x000cd002,
+	0x21f504bd,
+	0x21f5025e,
+	0x87f1027f,
+	0x83f04200,
+	0x0097f102,
+	0x0293f020,
+	0x950099cf,
+/* 0x034a: ctx_init_strand_loop */
 	0x8ed008fe,
 	0x408ed000,
 	0xb6808acf,
@@ -428,7 +458,7 @@ uint32_t nvd7_grhub_code[] = {
 	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-/* 0x02fe: error */
+/* 0x037e: error */
 	0x07f100f8,
 	0x03f00500,
 	0x000fd002,
@@ -436,82 +466,117 @@ uint32_t nvd7_grhub_code[] = {
 	0x0007f101,
 	0x0303f007,
 	0xbd000fd0,
-/* 0x031b: init */
+/* 0x039b: init */
 	0xbd00f804,
-	0x0004fe04,
-	0xf10007fe,
-	0xf0120017,
-	0x12d00227,
-	0xb117f100,
-	0x0010fe05,
-	0x040017f1,
-	0xf1c010d0,
-	0xb6040437,
-	0x27f10634,
-	0x32d02003,
-	0x0427f100,
-	0x0132d020,
+	0x0007fe04,
+	0x420017f1,
+	0xcf0013f0,
+	0x11e70011,
+	0x14b60109,
+	0x0014fe08,
+	0xf10227f0,
+	0xf0120007,
+	0x02d00003,
+	0xf104bd00,
+	0xfe06c817,
+	0x24bd0010,
+	0x070007f1,
+	0xd00003f0,
+	0x04bd0002,
+	0x200327f1,
+	0x010007f1,
+	0xd00103f0,
+	0x04bd0002,
+	0x200427f1,
+	0x010407f1,
+	0xd00103f0,
+	0x04bd0002,
 	0x200b27f1,
-	0xf10232d0,
-	0xd0200c27,
-	0x27f10732,
-	0x24b60c24,
-	0x0003b906,
-	0xf10023d0,
+	0x010807f1,
+	0xd00103f0,
+	0x04bd0002,
+	0x200c27f1,
+	0x011c07f1,
+	0xd00103f0,
+	0x04bd0002,
+	0xf1010392,
+	0xf0090007,
+	0x03d00303,
+	0xf104bd00,
 	0xf0870427,
-	0x12d00023,
-	0x0012b700,
-	0x0427f001,
-	0xf40012d0,
-	0xe7f11031,
-	0xe3f09604,
-	0x6821f440,
-	0x8090f1c7,
-	0xf4f00301,
-	0x020f801f,
-	0xbb0117f0,
-	0x12b6041f,
-	0x0c27f101,
-	0x0624b604,
-	0xd00021d0,
-	0x17f14021,
-	0x0e980100,
-	0x010f9800,
-	0x014721f5,
-	0x070037f1,
-	0x950634b6,
-	0x34d00814,
-	0x4034d000,
-	0x130030b7,
-	0xb6001fbb,
-	0x3fd002f5,
-	0x0815b600,
-	0xb60110b6,
-	0x1fb90814,
-	0x7121f502,
-	0x001fbb02,
-	0xf1020398,
-	0xf0200047,
-/* 0x03f6: init_gpc */
-	0x4ea05043,
-	0x1fb90804,
-	0x8d21f402,
-	0x010c4ea0,
-	0x21f4f4bd,
-	0x044ea08d,
-	0x8d21f401,
-	0x01004ea0,
-	0xf402f7f0,
-	0x4ea08d21,
-/* 0x041e: init_gpc_wait */
-	0x21f40800,
-	0x1fffc868,
-	0xa0fa0bf4,
-	0xf408044e,
-	0x1fbb6821,
-	0x0040b700,
-	0x0132b680,
-	0xf1be1bf4,
+	0x07f10023,
+	0x03f00400,
+	0x0002d000,
+	0x27f004bd,
+	0x0007f104,
+	0x0003f003,
+	0xbd0002d0,
+	0x1031f404,
+	0x9604e7f1,
+	0xf440e3f0,
+	0xfeb96821,
+	0x90f1c702,
+	0xf0030180,
+	0x0f801ff4,
+	0x0117f002,
+	0xb6041fbb,
+	0x07f10112,
+	0x03f00300,
+	0x0001d001,
+	0x07f104bd,
+	0x03f00400,
+	0x0001d001,
+	0x17f104bd,
+	0xf7f00100,
+	0xb521f502,
+	0xc721f507,
+	0x10f7f007,
+	0x081421f5,
+	0x98000e98,
+	0x21f5010f,
+	0x14950150,
+	0x0007f108,
+	0x0103f0c0,
+	0xbd0004d0,
+	0x0007f104,
+	0x0103f0c1,
+	0xbd0004d0,
+	0x0030b704,
+	0x001fbb13,
+	0xf102f5b6,
+	0xf0d30007,
+	0x0fd00103,
+	0xb604bd00,
+	0x10b60815,
+	0x0814b601,
+	0xf5021fb9,
+	0xbb02d321,
+	0x0398001f,
+	0x0047f102,
+	0x5043f020,
+/* 0x04f4: init_gpc */
+	0x08044ea0,
+	0xf4021fb9,
+	0x4ea09d21,
+	0xf4bd010c,
+	0xa09d21f4,
+	0xf401044e,
+	0x4ea09d21,
+	0xf7f00100,
+	0x9d21f402,
+	0x08004ea0,
+/* 0x051c: init_gpc_wait */
+	0xc86821f4,
+	0x0bf41fff,
+	0x044ea0fa,
+	0x6821f408,
+	0xb7001fbb,
+	0xb6800040,
+	0x1bf40132,
+	0x00f7f0be,
+	0x081421f5,
+	0xf500f7f0,
+	0xf107b521,
 	0xf0010007,
 	0x01d00203,
 	0xbd04bd00,
@@ -519,402 +584,399 @@ uint32_t nvd7_grhub_code[] = {
 	0x080007f1,
 	0xd00203f0,
 	0x04bd0001,
-/* 0x0458: main */
+/* 0x0564: main */
 	0xf40031f4,
 	0xd7f00028,
 	0x3921f410,
 	0xb1f401f4,
 	0xf54001e4,
-	0xbd00de1b,
+	0xbd00e91b,
 	0x0499f094,
 	0x0f0007f1,
 	0xd00203f0,
 	0x04bd0009,
-	0x0b0017f1,
-	0xcf0614b6,
-	0x11cf4012,
-	0x1f13c800,
-	0x00870bf5,
-	0xf41f23c8,
-	0x20f9620b,
-	0xbd0212b9,
-	0x0799f094,
-	0x0f0007f1,
-	0xd00203f0,
-	0x04bd0009,
-	0xf40132f4,
-	0x21f50231,
-	0x94bd082f,
+	0xc00017f1,
+	0xcf0213f0,
+	0x27f10011,
+	0x23f0c100,
+	0x0022cf02,
+	0xf51f13c8,
+	0xc800890b,
+	0x0bf41f23,
+	0xb920f962,
+	0x94bd0212,
 	0xf10799f0,
-	0xf0170007,
+	0xf00f0007,
 	0x09d00203,
-	0xfc04bd00,
-	0xf094bd20,
-	0x07f10699,
-	0x03f00f00,
-	0x0009d002,
-	0x31f404bd,
-	0x2f21f501,
-	0xf094bd08,
-	0x07f10699,
+	0xf404bd00,
+	0x31f40132,
+	0xe821f502,
+	0xf094bd09,
+	0x07f10799,
 	0x03f01700,
 	0x0009d002,
-	0x0ef404bd,
-/* 0x04f9: chsw_prev_no_next */
-	0xb920f931,
-	0x32f40212,
-	0x0232f401,
-	0x082f21f5,
-	0x17f120fc,
-	0x14b60b00,
-	0x0012d006,
-/* 0x0517: chsw_no_prev */
-	0xc8130ef4,
-	0x0bf41f23,
-	0x0131f40d,
-	0xf50232f4,
-/* 0x0527: chsw_done */
-	0xf1082f21,
-	0xb60b0c17,
-	0x27f00614,
-	0x0012d001,
+	0x20fc04bd,
 	0x99f094bd,
-	0x0007f104,
+	0x0007f106,
+	0x0203f00f,
+	0xbd0009d0,
+	0x0131f404,
+	0x09e821f5,
+	0x99f094bd,
+	0x0007f106,
 	0x0203f017,
 	0xbd0009d0,
-	0x130ef504,
-/* 0x0549: main_not_ctx_switch */
-	0x01e4b0ff,
-	0xb90d1bf4,
-	0x21f502f2,
-	0x0ef407bb,
-/* 0x0559: main_not_ctx_chan */
-	0x02e4b046,
-	0xbd321bf4,
-	0x0799f094,
-	0x0f0007f1,
+	0x330ef404,
+/* 0x060c: chsw_prev_no_next */
+	0x12b920f9,
+	0x0132f402,
+	0xf50232f4,
+	0xfc09e821,
+	0x0007f120,
+	0x0203f0c0,
+	0xbd0002d0,
+	0x130ef404,
+/* 0x062c: chsw_no_prev */
+	0xf41f23c8,
+	0x31f40d0b,
+	0x0232f401,
+	0x09e821f5,
+/* 0x063c: chsw_done */
+	0xf10127f0,
+	0xf0c30007,
+	0x02d00203,
+	0xbd04bd00,
+	0x0499f094,
+	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-	0xf40132f4,
-	0x21f50232,
-	0x94bd082f,
+	0xff080ef5,
+/* 0x0660: main_not_ctx_switch */
+	0xf401e4b0,
+	0xf2b90d1b,
+	0x7821f502,
+	0x460ef409,
+/* 0x0670: main_not_ctx_chan */
+	0xf402e4b0,
+	0x94bd321b,
 	0xf10799f0,
-	0xf0170007,
+	0xf00f0007,
 	0x09d00203,
 	0xf404bd00,
-/* 0x058e: main_not_ctx_save */
-	0xef94110e,
-	0x01f5f010,
-	0x02fe21f5,
-	0xfec00ef5,
-/* 0x059c: main_done */
-	0x29f024bd,
-	0x0007f11f,
-	0x0203f008,
-	0xbd0002d0,
-	0xab0ef504,
-/* 0x05b1: ih */
-	0xfe80f9fe,
-	0x80f90188,
-	0xa0f990f9,
-	0xd0f9b0f9,
-	0xf0f9e0f9,
-	0x0acf04bd,
-	0x04abc480,
-	0xf11d0bf4,
-	0xf01900b7,
-	0xbecf10d7,
-	0x00bfcf40,
+	0x32f40132,
+	0xe821f502,
+	0xf094bd09,
+	0x07f10799,
+	0x03f01700,
+	0x0009d002,
+	0x0ef404bd,
+/* 0x06a5: main_not_ctx_save */
+	0x10ef9411,
+	0xf501f5f0,
+	0xf5037e21,
+/* 0x06b3: main_done */
+	0xbdfeb50e,
+	0x1f29f024,
+	0x080007f1,
+	0xd00203f0,
+	0x04bd0002,
+	0xfea00ef5,
+/* 0x06c8: ih */
+	0x88fe80f9,
+	0xf980f901,
+	0xf9a0f990,
+	0xf9d0f9b0,
+	0xbdf0f9e0,
+	0x00a7f104,
+	0x00a3f002,
+	0xc400aacf,
+	0x0bf404ab,
+	0x10d7f030,
+	0x1a00e7f1,
+	0xcf00e3f0,
+	0xf7f100ee,
+	0xf3f01900,
+	0x00ffcf00,
 	0xb70421f4,
 	0xf00400b0,
-	0xbed001e7,
-/* 0x05e9: ih_no_fifo */
-	0x00abe400,
-	0x0d0bf401,
-	0xf110d7f0,
-	0xf44001e7,
-/* 0x05fa: ih_no_ctxsw */
-	0xb7f10421,
-	0xb0bd0104,
-	0xf4b4abff,
-	0xa7f10d0b,
-	0xa4b60c1c,
-	0x00abd006,
-/* 0x0610: ih_no_other */
-	0xfc400ad0,
+	0x07f101e7,
+	0x03f01d00,
+	0x000ed000,
+/* 0x071a: ih_no_fifo */
+	0xabe404bd,
+	0x0bf40100,
+	0x10d7f00d,
+	0x4001e7f1,
+/* 0x072b: ih_no_ctxsw */
+	0xe40421f4,
+	0xf40400ab,
+	0xb7f1140b,
+	0xbfb90100,
+	0x44e7f102,
+	0x40e3f001,
+/* 0x0743: ih_no_fwmthd */
+	0xf19d21f4,
+	0xbd0104b7,
+	0xb4abffb0,
+	0xf10f0bf4,
+	0xf0070007,
+	0x0bd00303,
+/* 0x075b: ih_no_other */
+	0xf104bd00,
+	0xf0010007,
+	0x0ad00003,
+	0xfc04bd00,
 	0xfce0fcf0,
 	0xfcb0fcd0,
 	0xfc90fca0,
 	0x0088fe80,
 	0x32f480fc,
-/* 0x062b: ctx_4160s */
-	0xf101f800,
-	0xf04160e7,
-	0xf7f040e3,
-	0x8d21f401,
-/* 0x0638: ctx_4160s_wait */
-	0xc86821f4,
-	0x0bf404ff,
-/* 0x0643: ctx_4160c */
-	0xf100f8fa,
+/* 0x077f: ctx_4160s */
+	0xf001f800,
+	0xffb901f7,
+	0x60e7f102,
+	0x40e3f041,
+/* 0x078f: ctx_4160s_wait */
+	0xf19d21f4,
 	0xf04160e7,
-	0xf4bd40e3,
-	0xf88d21f4,
-/* 0x0651: ctx_4170s */
-	0x70e7f100,
+	0x21f440e3,
+	0x02ffb968,
+	0xf404ffc8,
+	0x00f8f00b,
+/* 0x07a4: ctx_4160c */
+	0xffb9f4bd,
+	0x60e7f102,
 	0x40e3f041,
-	0xf410f5f0,
-	0x00f88d21,
-/* 0x0660: ctx_4170w */
-	0x4170e7f1,
-	0xf440e3f0,
-	0xf4f06821,
-	0xf31bf410,
-/* 0x0672: ctx_redswitch */
-	0xe7f100f8,
-	0xe4b60614,
-	0x70f7f106,
-	0x00efd002,
-/* 0x0683: ctx_redswitch_delay */
-	0xb608f7f0,
-	0x1bf401f2,
-	0x70f7f1fd,
-	0x00efd007,
-/* 0x0692: ctx_86c */
-	0xe7f100f8,
-	0xe4b6086c,
-	0x00efd006,
-	0x8a14e7f1,
-	0xf440e3f0,
-	0xe7f18d21,
-	0xe3f0a86c,
-	0x8d21f441,
-/* 0x06b2: ctx_load */
+	0xf89d21f4,
+/* 0x07b5: ctx_4170s */
+	0x10f5f000,
+	0xf102ffb9,
+	0xf04170e7,
+	0x21f440e3,
+/* 0x07c7: ctx_4170w */
+	0xf100f89d,
+	0xf04170e7,
+	0x21f440e3,
+	0x02ffb968,
+	0xf410f4f0,
+	0x00f8f01b,
+/* 0x07dc: ctx_redswitch */
+	0x0200e7f1,
+	0xf040e5f0,
+	0xe5f020e5,
+	0x0007f110,
+	0x0103f085,
+	0xbd000ed0,
+	0x08f7f004,
+/* 0x07f8: ctx_redswitch_delay */
+	0xf401f2b6,
+	0xe5f1fd1b,
+	0xe5f10400,
+	0x07f10100,
+	0x03f08500,
+	0x000ed001,
+	0x00f804bd,
+/* 0x0814: ctx_86c */
+	0x1b0007f1,
+	0xd00203f0,
+	0x04bd000f,
+	0xf102ffb9,
+	0xf08a14e7,
+	0x21f440e3,
+	0x02ffb99d,
+	0xa86ce7f1,
+	0xf441e3f0,
+	0x00f89d21,
+/* 0x083c: ctx_mem */
+	0x840007f1,
+	0xd00203f0,
+	0x04bd000f,
+/* 0x0848: ctx_mem_wait */
+	0x8400f7f1,
+	0xcf02f3f0,
+	0xfffd00ff,
+	0xf31bf405,
+/* 0x085a: ctx_load */
 	0x94bd00f8,
 	0xf10599f0,
 	0xf00f0007,
 	0x09d00203,
 	0xf004bd00,
 	0x21f40ca7,
-	0x2417f1c9,
-	0x0614b60a,
-	0xf10010d0,
-	0xb60b0037,
-	0x32d00634,
-	0x0c17f140,
-	0x0614b60a,
-	0xd00747f0,
-	0x14d00012,
-/* 0x06ed: ctx_chan_wait_0 */
-	0x4014cf40,
-	0xf41f44f0,
-	0x32d0fa1b,
-	0x000bfe00,
-	0xb61f2af0,
-	0x20b60424,
-	0xf094bd02,
+	0xf1f4bdd0,
+	0xf0890007,
+	0x0fd00203,
+	0xf104bd00,
+	0xf0c10007,
+	0x02d00203,
+	0xf104bd00,
+	0xf0830007,
+	0x02d00203,
+	0xf004bd00,
+	0x21f507f7,
+	0x07f1083c,
+	0x03f0c000,
+	0x0002d002,
+	0x0bfe04bd,
+	0x1f2af000,
+	0xb60424b6,
+	0x94bd0220,
+	0xf10899f0,
+	0xf00f0007,
+	0x09d00203,
+	0xf104bd00,
+	0xf0810007,
+	0x02d00203,
+	0xf104bd00,
+	0xf1000027,
+	0xf0800023,
+	0x07f10225,
+	0x03f08800,
+	0x0002d002,
+	0x17f004bd,
+	0x0027f110,
+	0x0223f002,
+	0xf80512fa,
+	0xf094bd03,
 	0x07f10899,
-	0x03f00f00,
+	0x03f01700,
 	0x0009d002,
-	0x17f104bd,
-	0x14b60a04,
-	0x0012d006,
-	0x0a2017f1,
-	0xf00614b6,
-	0x23f10227,
-	0x12d08000,
-	0x1017f000,
-	0x020027f1,
-	0xfa0223f0,
-	0x03f80512,
+	0x019804bd,
+	0x1814b681,
+	0xb6800298,
+	0x12fd0825,
+	0x16018005,
 	0x99f094bd,
-	0x0007f108,
-	0x0203f017,
+	0x0007f109,
+	0x0203f00f,
 	0xbd0009d0,
-	0x81019804,
-	0x981814b6,
-	0x25b68002,
-	0x0512fd08,
-	0xbd160180,
-	0x0999f094,
-	0x0f0007f1,
-	0xd00203f0,
-	0x04bd0009,
-	0x0a0427f1,
-	0xd00624b6,
-	0x27f00021,
-	0x2017f101,
-	0x0614b60a,
-	0xf10012d0,
-	0xf0010017,
-	0x01fa0613,
-	0xbd03f805,
-	0x0999f094,
-	0x170007f1,
+	0x0007f104,
+	0x0203f081,
+	0xbd0001d0,
+	0x0127f004,
+	0x880007f1,
 	0xd00203f0,
-	0x04bd0009,
+	0x04bd0002,
+	0x010017f1,
+	0xfa0613f0,
+	0x03f80501,
 	0x99f094bd,
-	0x0007f105,
+	0x0007f109,
 	0x0203f017,
 	0xbd0009d0,
-/* 0x07bb: ctx_chan */
-	0xf500f804,
-	0xf5062b21,
-	0xf006b221,
-	0x21f40ca7,
-	0x1017f1c9,
-	0x0614b60a,
-	0xd00527f0,
-/* 0x07d6: ctx_chan_wait */
-	0x12cf0012,
-	0x0522fd00,
-	0xf5fa1bf4,
-	0xf8064321,
-/* 0x07e5: ctx_mmio_exec */
-	0x41039800,
-	0x0a0427f1,
-	0xd00624b6,
-	0x34bd0023,
-/* 0x07f4: ctx_mmio_loop */
+	0xf094bd04,
+	0x07f10599,
+	0x03f01700,
+	0x0009d002,
+	0x00f804bd,
+/* 0x0978: ctx_chan */
+	0x077f21f5,
+	0x085a21f5,
+	0xf40ca7f0,
+	0xf7f0d021,
+	0x3c21f505,
+	0xa421f508,
+/* 0x0993: ctx_mmio_exec */
+	0x9800f807,
+	0x07f14103,
+	0x03f08100,
+	0x0003d002,
+	0x34bd04bd,
+/* 0x09a4: ctx_mmio_loop */
 	0xf4ff34c4,
 	0x57f10f1b,
 	0x53f00200,
 	0x0535fa06,
-/* 0x0806: ctx_mmio_pull */
+/* 0x09b6: ctx_mmio_pull */
 	0x4e9803f8,
 	0x814f9880,
-	0xb68d21f4,
+	0xb69d21f4,
 	0x12b60830,
 	0xdf1bf401,
-/* 0x0818: ctx_mmio_done */
-	0xd0160398,
-	0x00800023,
-	0x0017f140,
-	0x0613f001,
-	0xf80601fa,
-/* 0x082f: ctx_xfer */
-	0xf100f803,
-	0xb60c00f7,
-	0xe7f006f4,
-	0x80fed004,
-/* 0x083c: ctx_xfer_idle */
-	0xf100fecf,
-	0xf42000e4,
-	0x11f4f91b,
-	0x1102f406,
-/* 0x084c: ctx_xfer_pre */
-	0xf510f7f0,
-	0xf5069221,
-	0xf4062b21,
-/* 0x085a: ctx_xfer_pre_load */
-	0xf7f01c11,
-	0x5121f502,
-	0x6021f506,
-	0x7221f506,
-	0xf5f4bd06,
-	0xf5065121,
-/* 0x0873: ctx_xfer_exec */
-	0x9806b221,
-	0x27f11601,
-	0x24b60414,
-	0x0020d006,
-	0xa500e7f1,
-	0xb941e3f0,
-	0x21f4021f,
-	0x04e0b68d,
-	0xf001fcf0,
-	0x24b6022c,
-	0x05f2fd01,
-	0xf18d21f4,
-	0xf04afc17,
-	0x27f00213,
-	0x0012d00c,
-	0x021521f5,
-	0x47fc27f1,
-	0xd00223f0,
-	0x2cf00020,
+/* 0x09c8: ctx_mmio_done */
+	0xf1160398,
+	0xf0810007,
+	0x03d00203,
+	0x8004bd00,
+	0x17f14000,
+	0x13f00100,
+	0x0601fa06,
+	0x00f803f8,
+/* 0x09e8: ctx_xfer */
+	0xf104e7f0,
+	0xf0020007,
+	0x0ed00303,
+/* 0x09f7: ctx_xfer_idle */
+	0xf104bd00,
+	0xf00000e7,
+	0xeecf03e3,
+	0x00e4f100,
+	0xf21bf420,
+	0xf40611f4,
+/* 0x0a0e: ctx_xfer_pre */
+	0xf7f01102,
+	0x1421f510,
+	0x7f21f508,
+	0x1c11f407,
+/* 0x0a1c: ctx_xfer_pre_load */
+	0xf502f7f0,
+	0xf507b521,
+	0xf507c721,
+	0xbd07dc21,
+	0xb521f5f4,
+	0x5a21f507,
+/* 0x0a35: ctx_xfer_exec */
+	0x16019808,
+	0x07f124bd,
+	0x03f00500,
+	0x0002d001,
+	0x1fb904bd,
+	0x00e7f102,
+	0x41e3f0a5,
+	0xf09d21f4,
+	0x2cf001fc,
+	0x0124b602,
+	0xb905f2fd,
+	0xe7f102ff,
+	0xe3f0a504,
+	0x9d21f441,
+	0x026a21f5,
+	0x07f124bd,
+	0x03f047fc,
+	0x0002d002,
+	0x2cf004bd,
 	0x0320b601,
-	0xf00012d0,
-	0xa5f001ac,
-	0x00b7f006,
-	0x98000c98,
-	0xe7f0010d,
-	0x6621f500,
-	0x08a7f001,
-	0x010921f5,
-	0x021521f5,
-	0xf02201f4,
-	0x21f40ca7,
-	0x1017f1c9,
-	0x0614b60a,
-	0xd00527f0,
-/* 0x08fa: ctx_xfer_post_save_wait */
-	0x12cf0012,
-	0x0522fd00,
-	0xf4fa1bf4,
-/* 0x0906: ctx_xfer_post */
-	0xf7f03202,
-	0x5121f502,
-	0xf5f4bd06,
-	0xf5069221,
-	0xf5023421,
-	0xbd066021,
-	0x5121f5f4,
-	0x1011f406,
-	0xfd400198,
-	0x0bf40511,
-	0xe521f507,
-/* 0x0931: ctx_xfer_no_post_mmio */
-	0x4321f507,
-/* 0x0935: ctx_xfer_done */
-	0x0000f806,
-	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,
+	0x4afc07f1,
+	0xd00203f0,
+	0x04bd0002,
+	0xf001acf0,
+	0xb7f006a5,
+	0x000c9800,
+	0xf0010d98,
+	0x21f500e7,
+	0xa7f0016f,
+	0x1021f508,
+	0x5e21f501,
+	0x1301f402,
+	0xf40ca7f0,
+	0xf7f0d021,
+	0x3c21f505,
+	0x3202f408,
+/* 0x0ac4: ctx_xfer_post */
+	0xf502f7f0,
+	0xbd07b521,
+	0x1421f5f4,
+	0x7f21f508,
+	0xc721f502,
+	0xf5f4bd07,
+	0xf407b521,
+	0x01981011,
+	0x0511fd40,
+	0xf5070bf4,
+/* 0x0aef: ctx_xfer_no_post_mmio */
+	0xf5099321,
+/* 0x0af3: ctx_xfer_done */
+	0xf807a421,
 	0x00000000,
 	0x00000000,
 	0x00000000,
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnve0.fuc.h b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnve0.fuc.h
index eb7bc0e9576e..1c179bdd48cc 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnve0.fuc.h
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnve0.fuc.h
@@ -206,14 +206,14 @@ uint32_t nve0_grhub_data[] = {
 };
 
 uint32_t nve0_grhub_code[] = {
-	0x031b0ef5,
+	0x039b0ef5,
 /* 0x0004: queue_put */
 	0x9800d898,
 	0x86f001d9,
 	0x0489b808,
 	0xf00c1bf4,
 	0x21f502f7,
-	0x00f802fe,
+	0x00f8037e,
 /* 0x001c: queue_put_next */
 	0xb60798c4,
 	0x8dbb0384,
@@ -237,184 +237,214 @@ uint32_t nve0_grhub_code[] = {
 /* 0x0066: queue_get_done */
 	0x00f80132,
 /* 0x0068: nv_rd32 */
-	0x0728b7f1,
-	0xb906b4b6,
-	0xc9f002ec,
-	0x00bcd01f,
-/* 0x0078: nv_rd32_wait */
-	0xc800bccf,
-	0x1bf41fcc,
-	0x06a7f0fa,
-	0x010921f5,
-	0xf840bfcf,
-/* 0x008d: nv_wr32 */
-	0x28b7f100,
-	0x06b4b607,
-	0xb980bfd0,
-	0xc9f002ec,
-	0x1ec9f01f,
-/* 0x00a3: nv_wr32_wait */
-	0xcf00bcd0,
-	0xccc800bc,
-	0xfa1bf41f,
-/* 0x00ae: watchdog_reset */
-	0x87f100f8,
-	0x84b60430,
-	0x1ff9f006,
-	0xf8008fd0,
-/* 0x00bd: watchdog_clear */
-	0x3087f100,
-	0x0684b604,
-	0xf80080d0,
-/* 0x00c9: wait_donez */
-	0xf094bd00,
-	0x07f10099,
-	0x03f00f00,
-	0x0009d002,
-	0x07f104bd,
-	0x03f00600,
-	0x000ad002,
-/* 0x00e6: wait_donez_ne */
-	0x87f104bd,
-	0x83f00000,
-	0x0088cf01,
-	0xf4888aff,
-	0x94bdf31b,
-	0xf10099f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0109: wait_doneo */
-	0xf094bd00,
+	0xf002ecb9,
+	0x07f11fc9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x007a: nv_rd32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0xa7f0f31b,
+	0x1021f506,
+	0x00f7f101,
+	0x01f3f0cb,
+	0xf800ffcf,
+/* 0x009d: nv_wr32 */
+	0x0007f100,
+	0x0103f0cc,
+	0xbd000fd0,
+	0x02ecb904,
+	0xf01fc9f0,
+	0x07f11ec9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x00be: nv_wr32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0x00f8f31b,
+/* 0x00d0: wait_donez */
+	0x99f094bd,
+	0x0007f100,
+	0x0203f00f,
+	0xbd0009d0,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x00ed: wait_donez_ne */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x1bf4888a,
+	0xf094bdf3,
 	0x07f10099,
-	0x03f00f00,
+	0x03f01700,
 	0x0009d002,
-	0x87f104bd,
-	0x84b60818,
-	0x008ad006,
-/* 0x0124: wait_doneo_e */
-	0x040087f1,
-	0xcf0684b6,
-	0x8aff0088,
-	0xf30bf488,
+	0x00f804bd,
+/* 0x0110: wait_doneo */
 	0x99f094bd,
 	0x0007f100,
-	0x0203f017,
+	0x0203f00f,
 	0xbd0009d0,
-/* 0x0147: mmctx_size */
-	0xbd00f804,
-/* 0x0149: nv_mmctx_size_loop */
-	0x00e89894,
-	0xb61a85b6,
-	0x84b60180,
-	0x0098bb02,
-	0xb804e0b6,
-	0x1bf404ef,
-	0x029fb9eb,
-/* 0x0166: mmctx_xfer */
-	0x94bd00f8,
-	0xf10199f0,
-	0xf00f0007,
-	0x09d00203,
-	0xf104bd00,
-	0xb6071087,
-	0x94bd0684,
-	0xf405bbfd,
-	0x8bd0090b,
-	0x0099f000,
-/* 0x018c: mmctx_base_disabled */
-	0xf405eefd,
-	0x8ed00c0b,
-	0xc08fd080,
-/* 0x019b: mmctx_multi_disabled */
-	0xb70199f0,
-	0xc8010080,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x012d: wait_doneo_e */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x0bf4888a,
+	0xf094bdf3,
+	0x07f10099,
+	0x03f01700,
+	0x0009d002,
+	0x00f804bd,
+/* 0x0150: mmctx_size */
+/* 0x0152: nv_mmctx_size_loop */
+	0xe89894bd,
+	0x1a85b600,
+	0xb60180b6,
+	0x98bb0284,
+	0x04e0b600,
+	0xf404efb8,
+	0x9fb9eb1b,
+/* 0x016f: mmctx_xfer */
+	0xbd00f802,
+	0x0199f094,
+	0x0f0007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0xbbfd94bd,
+	0x120bf405,
+	0xc40007f1,
+	0xd00103f0,
+	0x04bd000b,
+/* 0x0197: mmctx_base_disabled */
+	0xfd0099f0,
+	0x0bf405ee,
+	0x0007f11e,
+	0x0103f0c6,
+	0xbd000ed0,
+	0x0007f104,
+	0x0103f0c7,
+	0xbd000fd0,
+	0x0199f004,
+/* 0x01b8: mmctx_multi_disabled */
+	0xb600abc8,
+	0xb9f010b4,
+	0x01aec80c,
+	0xfd11e4b6,
+	0x07f105be,
+	0x03f0c500,
+	0x000bd001,
+/* 0x01d6: mmctx_exec_loop */
+/* 0x01d6: mmctx_wait_free */
+	0xe7f104bd,
+	0xe3f0c500,
+	0x00eecf01,
+	0xf41fe4f0,
+	0xce98f30b,
+	0x05e9fd00,
+	0xc80007f1,
+	0xd00103f0,
+	0x04bd000e,
+	0xb804c0b6,
+	0x1bf404cd,
+	0x02abc8d8,
+/* 0x0207: mmctx_fini_wait */
+	0xf11f1bf4,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x1fb4f000,
+	0xf410b4b0,
+	0xa7f0f01b,
+	0xd021f402,
+/* 0x0223: mmctx_stop */
+	0xc82b0ef4,
 	0xb4b600ab,
 	0x0cb9f010,
-	0xb601aec8,
-	0xbefd11e4,
-	0x008bd005,
-/* 0x01b4: mmctx_exec_loop */
-/* 0x01b4: mmctx_wait_free */
-	0xf0008ecf,
-	0x0bf41fe4,
-	0x00ce98fa,
-	0xd005e9fd,
-	0xc0b6c08e,
-	0x04cdb804,
-	0xc8e81bf4,
-	0x1bf402ab,
-/* 0x01d5: mmctx_fini_wait */
-	0x008bcf18,
-	0xb01fb4f0,
-	0x1bf410b4,
-	0x02a7f0f7,
-	0xf4c921f4,
-/* 0x01ea: mmctx_stop */
-	0xabc81b0e,
-	0x10b4b600,
-	0xf00cb9f0,
-	0x8bd012b9,
-/* 0x01f9: mmctx_stop_wait */
-	0x008bcf00,
-	0xf412bbc8,
-/* 0x0202: mmctx_done */
-	0x94bdfa1b,
-	0xf10199f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0215: strand_wait */
-	0xf0a0f900,
-	0x21f402a7,
-	0xf8a0fcc9,
-/* 0x0221: strand_pre */
-	0xfc87f100,
-	0x0283f04a,
-	0xd00c97f0,
-	0x21f50089,
-	0x00f80215,
-/* 0x0234: strand_post */
-	0x4afc87f1,
-	0xf00283f0,
-	0x89d00d97,
-	0x1521f500,
-/* 0x0247: strand_set */
-	0xf100f802,
-	0xf04ffca7,
-	0xaba202a3,
-	0xc7f00500,
-	0x00acd00f,
-	0xd00bc7f0,
-	0x21f500bc,
-	0xaed00215,
-	0x0ac7f000,
-	0xf500bcd0,
-	0xf8021521,
-/* 0x0271: strand_ctx_init */
-	0xf094bd00,
-	0x07f10399,
-	0x03f00f00,
+	0xf112b9f0,
+	0xf0c50007,
+	0x0bd00103,
+/* 0x023b: mmctx_stop_wait */
+	0xf104bd00,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x12bbc800,
+/* 0x024b: mmctx_done */
+	0xbdf31bf4,
+	0x0199f094,
+	0x170007f1,
+	0xd00203f0,
+	0x04bd0009,
+/* 0x025e: strand_wait */
+	0xa0f900f8,
+	0xf402a7f0,
+	0xa0fcd021,
+/* 0x026a: strand_pre */
+	0x97f000f8,
+	0xfc07f10c,
+	0x0203f04a,
+	0xbd0009d0,
+	0x5e21f504,
+/* 0x027f: strand_post */
+	0xf000f802,
+	0x07f10d97,
+	0x03f04afc,
 	0x0009d002,
 	0x21f504bd,
-	0xe7f00221,
-	0x4721f503,
-	0xfca7f102,
-	0x02a3f046,
-	0x0400aba0,
-	0xf040a0d0,
-	0xbcd001c7,
-	0x1521f500,
-	0x010c9202,
-	0xf000acd0,
-	0xbcd002c7,
-	0x1521f500,
-	0x3421f502,
-	0x8087f102,
-	0x0684b608,
-	0xb70089cf,
-	0x95220080,
-/* 0x02ca: ctx_init_strand_loop */
+	0x00f8025e,
+/* 0x0294: strand_set */
+	0xf10fc7f0,
+	0xf04ffc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f10bc7,
+	0x03f04afc,
+	0x000cd002,
+	0x07f104bd,
+	0x03f04ffc,
+	0x000ed002,
+	0xc7f004bd,
+	0xfc07f10a,
+	0x0203f04a,
+	0xbd000cd0,
+	0x5e21f504,
+/* 0x02d3: strand_ctx_init */
+	0xbd00f802,
+	0x0399f094,
+	0x0f0007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0x026a21f5,
+	0xf503e7f0,
+	0xbd029421,
+	0xfc07f1c4,
+	0x0203f047,
+	0xbd000cd0,
+	0x01c7f004,
+	0x4afc07f1,
+	0xd00203f0,
+	0x04bd000c,
+	0x025e21f5,
+	0xf1010c92,
+	0xf046fc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f102c7,
+	0x03f04afc,
+	0x000cd002,
+	0x21f504bd,
+	0x21f5025e,
+	0x87f1027f,
+	0x83f04200,
+	0x0097f102,
+	0x0293f020,
+	0x950099cf,
+/* 0x034a: ctx_init_strand_loop */
 	0x8ed008fe,
 	0x408ed000,
 	0xb6808acf,
@@ -428,7 +458,7 @@ uint32_t nve0_grhub_code[] = {
 	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-/* 0x02fe: error */
+/* 0x037e: error */
 	0x07f100f8,
 	0x03f00500,
 	0x000fd002,
@@ -436,82 +466,117 @@ uint32_t nve0_grhub_code[] = {
 	0x0007f101,
 	0x0303f007,
 	0xbd000fd0,
-/* 0x031b: init */
+/* 0x039b: init */
 	0xbd00f804,
-	0x0004fe04,
-	0xf10007fe,
-	0xf0120017,
-	0x12d00227,
-	0xb117f100,
-	0x0010fe05,
-	0x040017f1,
-	0xf1c010d0,
-	0xb6040437,
-	0x27f10634,
-	0x32d02003,
-	0x0427f100,
-	0x0132d020,
+	0x0007fe04,
+	0x420017f1,
+	0xcf0013f0,
+	0x11e70011,
+	0x14b60109,
+	0x0014fe08,
+	0xf10227f0,
+	0xf0120007,
+	0x02d00003,
+	0xf104bd00,
+	0xfe06c817,
+	0x24bd0010,
+	0x070007f1,
+	0xd00003f0,
+	0x04bd0002,
+	0x200327f1,
+	0x010007f1,
+	0xd00103f0,
+	0x04bd0002,
+	0x200427f1,
+	0x010407f1,
+	0xd00103f0,
+	0x04bd0002,
 	0x200b27f1,
-	0xf10232d0,
-	0xd0200c27,
-	0x27f10732,
-	0x24b60c24,
-	0x0003b906,
-	0xf10023d0,
+	0x010807f1,
+	0xd00103f0,
+	0x04bd0002,
+	0x200c27f1,
+	0x011c07f1,
+	0xd00103f0,
+	0x04bd0002,
+	0xf1010392,
+	0xf0090007,
+	0x03d00303,
+	0xf104bd00,
 	0xf0870427,
-	0x12d00023,
-	0x0012b700,
-	0x0427f001,
-	0xf40012d0,
-	0xe7f11031,
-	0xe3f09604,
-	0x6821f440,
-	0x8090f1c7,
-	0xf4f00301,
-	0x020f801f,
-	0xbb0117f0,
-	0x12b6041f,
-	0x0c27f101,
-	0x0624b604,
-	0xd00021d0,
-	0x17f14021,
-	0x0e980100,
-	0x010f9800,
-	0x014721f5,
-	0x070037f1,
-	0x950634b6,
-	0x34d00814,
-	0x4034d000,
-	0x130030b7,
-	0xb6001fbb,
-	0x3fd002f5,
-	0x0815b600,
-	0xb60110b6,
-	0x1fb90814,
-	0x7121f502,
-	0x001fbb02,
-	0xf1020398,
-	0xf0200047,
-/* 0x03f6: init_gpc */
-	0x4ea05043,
-	0x1fb90804,
-	0x8d21f402,
-	0x010c4ea0,
-	0x21f4f4bd,
-	0x044ea08d,
-	0x8d21f401,
-	0x01004ea0,
-	0xf402f7f0,
-	0x4ea08d21,
-/* 0x041e: init_gpc_wait */
-	0x21f40800,
-	0x1fffc868,
-	0xa0fa0bf4,
-	0xf408044e,
-	0x1fbb6821,
-	0x0040b700,
-	0x0132b680,
-	0xf1be1bf4,
+	0x07f10023,
+	0x03f00400,
+	0x0002d000,
+	0x27f004bd,
+	0x0007f104,
+	0x0003f003,
+	0xbd0002d0,
+	0x1031f404,
+	0x9604e7f1,
+	0xf440e3f0,
+	0xfeb96821,
+	0x90f1c702,
+	0xf0030180,
+	0x0f801ff4,
+	0x0117f002,
+	0xb6041fbb,
+	0x07f10112,
+	0x03f00300,
+	0x0001d001,
+	0x07f104bd,
+	0x03f00400,
+	0x0001d001,
+	0x17f104bd,
+	0xf7f00100,
+	0x7f21f502,
+	0x9121f507,
+	0x10f7f007,
+	0x07de21f5,
+	0x98000e98,
+	0x21f5010f,
+	0x14950150,
+	0x0007f108,
+	0x0103f0c0,
+	0xbd0004d0,
+	0x0007f104,
+	0x0103f0c1,
+	0xbd0004d0,
+	0x0030b704,
+	0x001fbb13,
+	0xf102f5b6,
+	0xf0d30007,
+	0x0fd00103,
+	0xb604bd00,
+	0x10b60815,
+	0x0814b601,
+	0xf5021fb9,
+	0xbb02d321,
+	0x0398001f,
+	0x0047f102,
+	0x5043f020,
+/* 0x04f4: init_gpc */
+	0x08044ea0,
+	0xf4021fb9,
+	0x4ea09d21,
+	0xf4bd010c,
+	0xa09d21f4,
+	0xf401044e,
+	0x4ea09d21,
+	0xf7f00100,
+	0x9d21f402,
+	0x08004ea0,
+/* 0x051c: init_gpc_wait */
+	0xc86821f4,
+	0x0bf41fff,
+	0x044ea0fa,
+	0x6821f408,
+	0xb7001fbb,
+	0xb6800040,
+	0x1bf40132,
+	0x00f7f0be,
+	0x07de21f5,
+	0xf500f7f0,
+	0xf1077f21,
 	0xf0010007,
 	0x01d00203,
 	0xbd04bd00,
@@ -519,382 +584,379 @@ uint32_t nve0_grhub_code[] = {
 	0x080007f1,
 	0xd00203f0,
 	0x04bd0001,
-/* 0x0458: main */
+/* 0x0564: main */
 	0xf40031f4,
 	0xd7f00028,
 	0x3921f410,
 	0xb1f401f4,
 	0xf54001e4,
-	0xbd00de1b,
+	0xbd00e91b,
 	0x0499f094,
 	0x0f0007f1,
 	0xd00203f0,
 	0x04bd0009,
-	0x0b0017f1,
-	0xcf0614b6,
-	0x11cf4012,
-	0x1f13c800,
-	0x00870bf5,
-	0xf41f23c8,
-	0x20f9620b,
-	0xbd0212b9,
-	0x0799f094,
-	0x0f0007f1,
-	0xd00203f0,
-	0x04bd0009,
-	0xf40132f4,
-	0x21f50231,
-	0x94bd0801,
+	0xc00017f1,
+	0xcf0213f0,
+	0x27f10011,
+	0x23f0c100,
+	0x0022cf02,
+	0xf51f13c8,
+	0xc800890b,
+	0x0bf41f23,
+	0xb920f962,
+	0x94bd0212,
 	0xf10799f0,
-	0xf0170007,
+	0xf00f0007,
 	0x09d00203,
-	0xfc04bd00,
-	0xf094bd20,
-	0x07f10699,
-	0x03f00f00,
-	0x0009d002,
-	0x31f404bd,
-	0x0121f501,
-	0xf094bd08,
-	0x07f10699,
+	0xf404bd00,
+	0x31f40132,
+	0xaa21f502,
+	0xf094bd09,
+	0x07f10799,
 	0x03f01700,
 	0x0009d002,
-	0x0ef404bd,
-/* 0x04f9: chsw_prev_no_next */
-	0xb920f931,
-	0x32f40212,
-	0x0232f401,
-	0x080121f5,
-	0x17f120fc,
-	0x14b60b00,
-	0x0012d006,
-/* 0x0517: chsw_no_prev */
-	0xc8130ef4,
-	0x0bf41f23,
-	0x0131f40d,
-	0xf50232f4,
-/* 0x0527: chsw_done */
-	0xf1080121,
-	0xb60b0c17,
-	0x27f00614,
-	0x0012d001,
+	0x20fc04bd,
 	0x99f094bd,
-	0x0007f104,
+	0x0007f106,
+	0x0203f00f,
+	0xbd0009d0,
+	0x0131f404,
+	0x09aa21f5,
+	0x99f094bd,
+	0x0007f106,
 	0x0203f017,
 	0xbd0009d0,
-	0x130ef504,
-/* 0x0549: main_not_ctx_switch */
-	0x01e4b0ff,
-	0xb90d1bf4,
-	0x21f502f2,
-	0x0ef40795,
-/* 0x0559: main_not_ctx_chan */
-	0x02e4b046,
-	0xbd321bf4,
-	0x0799f094,
-	0x0f0007f1,
+	0x330ef404,
+/* 0x060c: chsw_prev_no_next */
+	0x12b920f9,
+	0x0132f402,
+	0xf50232f4,
+	0xfc09aa21,
+	0x0007f120,
+	0x0203f0c0,
+	0xbd0002d0,
+	0x130ef404,
+/* 0x062c: chsw_no_prev */
+	0xf41f23c8,
+	0x31f40d0b,
+	0x0232f401,
+	0x09aa21f5,
+/* 0x063c: chsw_done */
+	0xf10127f0,
+	0xf0c30007,
+	0x02d00203,
+	0xbd04bd00,
+	0x0499f094,
+	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-	0xf40132f4,
-	0x21f50232,
-	0x94bd0801,
+	0xff080ef5,
+/* 0x0660: main_not_ctx_switch */
+	0xf401e4b0,
+	0xf2b90d1b,
+	0x4221f502,
+	0x460ef409,
+/* 0x0670: main_not_ctx_chan */
+	0xf402e4b0,
+	0x94bd321b,
 	0xf10799f0,
-	0xf0170007,
+	0xf00f0007,
 	0x09d00203,
 	0xf404bd00,
-/* 0x058e: main_not_ctx_save */
-	0xef94110e,
-	0x01f5f010,
-	0x02fe21f5,
-	0xfec00ef5,
-/* 0x059c: main_done */
-	0x29f024bd,
-	0x0007f11f,
-	0x0203f008,
-	0xbd0002d0,
-	0xab0ef504,
-/* 0x05b1: ih */
-	0xfe80f9fe,
-	0x80f90188,
-	0xa0f990f9,
-	0xd0f9b0f9,
-	0xf0f9e0f9,
-	0x0acf04bd,
-	0x04abc480,
-	0xf11d0bf4,
-	0xf01900b7,
-	0xbecf10d7,
-	0x00bfcf40,
+	0x32f40132,
+	0xaa21f502,
+	0xf094bd09,
+	0x07f10799,
+	0x03f01700,
+	0x0009d002,
+	0x0ef404bd,
+/* 0x06a5: main_not_ctx_save */
+	0x10ef9411,
+	0xf501f5f0,
+	0xf5037e21,
+/* 0x06b3: main_done */
+	0xbdfeb50e,
+	0x1f29f024,
+	0x080007f1,
+	0xd00203f0,
+	0x04bd0002,
+	0xfea00ef5,
+/* 0x06c8: ih */
+	0x88fe80f9,
+	0xf980f901,
+	0xf9a0f990,
+	0xf9d0f9b0,
+	0xbdf0f9e0,
+	0x00a7f104,
+	0x00a3f002,
+	0xc400aacf,
+	0x0bf404ab,
+	0x10d7f030,
+	0x1a00e7f1,
+	0xcf00e3f0,
+	0xf7f100ee,
+	0xf3f01900,
+	0x00ffcf00,
 	0xb70421f4,
 	0xf00400b0,
-	0xbed001e7,
-/* 0x05e9: ih_no_fifo */
-	0x00abe400,
-	0x0d0bf401,
-	0xf110d7f0,
-	0xf44001e7,
-/* 0x05fa: ih_no_ctxsw */
-	0xb7f10421,
-	0xb0bd0104,
-	0xf4b4abff,
-	0xa7f10d0b,
-	0xa4b60c1c,
-	0x00abd006,
-/* 0x0610: ih_no_other */
-	0xfc400ad0,
+	0x07f101e7,
+	0x03f01d00,
+	0x000ed000,
+/* 0x071a: ih_no_fifo */
+	0xabe404bd,
+	0x0bf40100,
+	0x10d7f00d,
+	0x4001e7f1,
+/* 0x072b: ih_no_ctxsw */
+	0xe40421f4,
+	0xf40400ab,
+	0xb7f1140b,
+	0xbfb90100,
+	0x44e7f102,
+	0x40e3f001,
+/* 0x0743: ih_no_fwmthd */
+	0xf19d21f4,
+	0xbd0104b7,
+	0xb4abffb0,
+	0xf10f0bf4,
+	0xf0070007,
+	0x0bd00303,
+/* 0x075b: ih_no_other */
+	0xf104bd00,
+	0xf0010007,
+	0x0ad00003,
+	0xfc04bd00,
 	0xfce0fcf0,
 	0xfcb0fcd0,
 	0xfc90fca0,
 	0x0088fe80,
 	0x32f480fc,
-/* 0x062b: ctx_4170s */
-	0xf101f800,
-	0xf04170e7,
-	0xf5f040e3,
-	0x8d21f410,
-/* 0x063a: ctx_4170w */
+/* 0x077f: ctx_4170s */
+	0xf001f800,
+	0xffb910f5,
+	0x70e7f102,
+	0x40e3f041,
+	0xf89d21f4,
+/* 0x0791: ctx_4170w */
+	0x70e7f100,
+	0x40e3f041,
+	0xb96821f4,
+	0xf4f002ff,
+	0xf01bf410,
+/* 0x07a6: ctx_redswitch */
 	0xe7f100f8,
-	0xe3f04170,
-	0x6821f440,
-	0xf410f4f0,
+	0xe5f00200,
+	0x20e5f040,
+	0xf110e5f0,
+	0xf0850007,
+	0x0ed00103,
+	0xf004bd00,
+/* 0x07c2: ctx_redswitch_delay */
+	0xf2b608f7,
+	0xfd1bf401,
+	0x0400e5f1,
+	0x0100e5f1,
+	0x850007f1,
+	0xd00103f0,
+	0x04bd000e,
+/* 0x07de: ctx_86c */
+	0x07f100f8,
+	0x03f01b00,
+	0x000fd002,
+	0xffb904bd,
+	0x14e7f102,
+	0x40e3f08a,
+	0xb99d21f4,
+	0xe7f102ff,
+	0xe3f0a86c,
+	0x9d21f441,
+/* 0x0806: ctx_mem */
+	0x07f100f8,
+	0x03f08400,
+	0x000fd002,
+/* 0x0812: ctx_mem_wait */
+	0xf7f104bd,
+	0xf3f08400,
+	0x00ffcf02,
+	0xf405fffd,
 	0x00f8f31b,
-/* 0x064c: ctx_redswitch */
-	0x0614e7f1,
-	0xf106e4b6,
-	0xd00270f7,
-	0xf7f000ef,
-/* 0x065d: ctx_redswitch_delay */
-	0x01f2b608,
-	0xf1fd1bf4,
-	0xd00770f7,
-	0x00f800ef,
-/* 0x066c: ctx_86c */
-	0x086ce7f1,
-	0xd006e4b6,
-	0xe7f100ef,
-	0xe3f08a14,
-	0x8d21f440,
-	0xa86ce7f1,
-	0xf441e3f0,
-	0x00f88d21,
-/* 0x068c: ctx_load */
+/* 0x0824: ctx_load */
 	0x99f094bd,
 	0x0007f105,
 	0x0203f00f,
 	0xbd0009d0,
 	0x0ca7f004,
-	0xf1c921f4,
-	0xb60a2417,
-	0x10d00614,
-	0x0037f100,
-	0x0634b60b,
-	0xf14032d0,
-	0xb60a0c17,
-	0x47f00614,
-	0x0012d007,
-/* 0x06c7: ctx_chan_wait_0 */
-	0xcf4014d0,
-	0x44f04014,
-	0xfa1bf41f,
-	0xfe0032d0,
-	0x2af0000b,
-	0x0424b61f,
-	0xbd0220b6,
+	0xbdd021f4,
+	0x0007f1f4,
+	0x0203f089,
+	0xbd000fd0,
+	0x0007f104,
+	0x0203f0c1,
+	0xbd0002d0,
+	0x0007f104,
+	0x0203f083,
+	0xbd0002d0,
+	0x07f7f004,
+	0x080621f5,
+	0xc00007f1,
+	0xd00203f0,
+	0x04bd0002,
+	0xf0000bfe,
+	0x24b61f2a,
+	0x0220b604,
+	0x99f094bd,
+	0x0007f108,
+	0x0203f00f,
+	0xbd0009d0,
+	0x0007f104,
+	0x0203f081,
+	0xbd0002d0,
+	0x0027f104,
+	0x0023f100,
+	0x0225f080,
+	0x880007f1,
+	0xd00203f0,
+	0x04bd0002,
+	0xf11017f0,
+	0xf0020027,
+	0x12fa0223,
+	0xbd03f805,
 	0x0899f094,
-	0x0f0007f1,
+	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-	0x0a0417f1,
-	0xd00614b6,
-	0x17f10012,
-	0x14b60a20,
-	0x0227f006,
-	0x800023f1,
-	0xf00012d0,
-	0x27f11017,
-	0x23f00200,
-	0x0512fa02,
-	0x94bd03f8,
-	0xf10899f0,
-	0xf0170007,
+	0xb6810198,
+	0x02981814,
+	0x0825b680,
+	0x800512fd,
+	0x94bd1601,
+	0xf10999f0,
+	0xf00f0007,
 	0x09d00203,
-	0x9804bd00,
-	0x14b68101,
-	0x80029818,
-	0xfd0825b6,
-	0x01800512,
-	0xf094bd16,
-	0x07f10999,
-	0x03f00f00,
-	0x0009d002,
-	0x27f104bd,
-	0x24b60a04,
-	0x0021d006,
-	0xf10127f0,
-	0xb60a2017,
-	0x12d00614,
-	0x0017f100,
-	0x0613f001,
-	0xf80501fa,
-	0xf094bd03,
-	0x07f10999,
-	0x03f01700,
-	0x0009d002,
-	0x94bd04bd,
-	0xf10599f0,
+	0xf104bd00,
+	0xf0810007,
+	0x01d00203,
+	0xf004bd00,
+	0x07f10127,
+	0x03f08800,
+	0x0002d002,
+	0x17f104bd,
+	0x13f00100,
+	0x0501fa06,
+	0x94bd03f8,
+	0xf10999f0,
 	0xf0170007,
 	0x09d00203,
-	0xf804bd00,
-/* 0x0795: ctx_chan */
-	0x8c21f500,
-	0x0ca7f006,
-	0xf1c921f4,
-	0xb60a1017,
-	0x27f00614,
-	0x0012d005,
-/* 0x07ac: ctx_chan_wait */
-	0xfd0012cf,
-	0x1bf40522,
-/* 0x07b7: ctx_mmio_exec */
-	0x9800f8fa,
-	0x27f14103,
-	0x24b60a04,
-	0x0023d006,
-/* 0x07c6: ctx_mmio_loop */
+	0xbd04bd00,
+	0x0599f094,
+	0x170007f1,
+	0xd00203f0,
+	0x04bd0009,
+/* 0x0942: ctx_chan */
+	0x21f500f8,
+	0xa7f00824,
+	0xd021f40c,
+	0xf505f7f0,
+	0xf8080621,
+/* 0x0955: ctx_mmio_exec */
+	0x41039800,
+	0x810007f1,
+	0xd00203f0,
+	0x04bd0003,
+/* 0x0966: ctx_mmio_loop */
 	0x34c434bd,
 	0x0f1bf4ff,
 	0x020057f1,
 	0xfa0653f0,
 	0x03f80535,
-/* 0x07d8: ctx_mmio_pull */
+/* 0x0978: ctx_mmio_pull */
 	0x98804e98,
 	0x21f4814f,
-	0x0830b68d,
+	0x0830b69d,
 	0xf40112b6,
-/* 0x07ea: ctx_mmio_done */
+/* 0x098a: ctx_mmio_done */
 	0x0398df1b,
-	0x0023d016,
-	0xf1400080,
-	0xf0010017,
-	0x01fa0613,
-	0xf803f806,
-/* 0x0801: ctx_xfer */
-	0x00f7f100,
-	0x06f4b60c,
-	0xd004e7f0,
-/* 0x080e: ctx_xfer_idle */
-	0xfecf80fe,
-	0x00e4f100,
-	0xf91bf420,
-	0xf40611f4,
-/* 0x081e: ctx_xfer_pre */
-	0xf7f00d02,
-	0x6c21f510,
-	0x1c11f406,
-/* 0x0828: ctx_xfer_pre_load */
-	0xf502f7f0,
-	0xf5062b21,
-	0xf5063a21,
-	0xbd064c21,
-	0x2b21f5f4,
-	0x8c21f506,
-/* 0x0841: ctx_xfer_exec */
-	0x16019806,
-	0x041427f1,
-	0xd00624b6,
-	0xe7f10020,
-	0xe3f0a500,
-	0x021fb941,
-	0xb68d21f4,
-	0xfcf004e0,
-	0x022cf001,
-	0xfd0124b6,
-	0x21f405f2,
-	0xfc17f18d,
-	0x0213f04a,
-	0xd00c27f0,
-	0x21f50012,
-	0x27f10215,
-	0x23f047fc,
-	0x0020d002,
+	0x0007f116,
+	0x0203f081,
+	0xbd0003d0,
+	0x40008004,
+	0x010017f1,
+	0xfa0613f0,
+	0x03f80601,
+/* 0x09aa: ctx_xfer */
+	0xe7f000f8,
+	0x0007f104,
+	0x0303f002,
+	0xbd000ed0,
+/* 0x09b9: ctx_xfer_idle */
+	0x00e7f104,
+	0x03e3f000,
+	0xf100eecf,
+	0xf42000e4,
+	0x11f4f21b,
+	0x0d02f406,
+/* 0x09d0: ctx_xfer_pre */
+	0xf510f7f0,
+	0xf407de21,
+/* 0x09da: ctx_xfer_pre_load */
+	0xf7f01c11,
+	0x7f21f502,
+	0x9121f507,
+	0xa621f507,
+	0xf5f4bd07,
+	0xf5077f21,
+/* 0x09f3: ctx_xfer_exec */
+	0x98082421,
+	0x24bd1601,
+	0x050007f1,
+	0xd00103f0,
+	0x04bd0002,
+	0xf1021fb9,
+	0xf0a500e7,
+	0x21f441e3,
+	0x01fcf09d,
+	0xb6022cf0,
+	0xf2fd0124,
+	0x02ffb905,
+	0xa504e7f1,
+	0xf441e3f0,
+	0x21f59d21,
+	0x24bd026a,
+	0x47fc07f1,
+	0xd00203f0,
+	0x04bd0002,
 	0xb6012cf0,
-	0x12d00320,
-	0x01acf000,
-	0xf006a5f0,
-	0x0c9800b7,
-	0x010d9800,
-	0xf500e7f0,
-	0xf0016621,
-	0x21f508a7,
-	0x21f50109,
-	0x01f40215,
-	0x0ca7f022,
-	0xf1c921f4,
-	0xb60a1017,
-	0x27f00614,
-	0x0012d005,
-/* 0x08c8: ctx_xfer_post_save_wait */
-	0xfd0012cf,
-	0x1bf40522,
-	0x2e02f4fa,
-/* 0x08d4: ctx_xfer_post */
-	0xf502f7f0,
-	0xbd062b21,
-	0x6c21f5f4,
-	0x3421f506,
-	0x3a21f502,
-	0xf5f4bd06,
-	0xf4062b21,
-	0x01981011,
-	0x0511fd40,
-	0xf5070bf4,
-/* 0x08ff: ctx_xfer_no_post_mmio */
-/* 0x08ff: ctx_xfer_done */
-	0xf807b721,
-	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,
+	0x07f10320,
+	0x03f04afc,
+	0x0002d002,
+	0xacf004bd,
+	0x06a5f001,
+	0x9800b7f0,
+	0x0d98000c,
+	0x00e7f001,
+	0x016f21f5,
+	0xf508a7f0,
+	0xf5011021,
+	0xf4025e21,
+	0xa7f01301,
+	0xd021f40c,
+	0xf505f7f0,
+	0xf4080621,
+/* 0x0a82: ctx_xfer_post */
+	0xf7f02e02,
+	0x7f21f502,
+	0xf5f4bd07,
+	0xf507de21,
+	0xf5027f21,
+	0xbd079121,
+	0x7f21f5f4,
+	0x1011f407,
+	0xfd400198,
+	0x0bf40511,
+	0x5521f507,
+/* 0x0aad: ctx_xfer_no_post_mmio */
+/* 0x0aad: ctx_xfer_done */
+	0x0000f809,
 	0x00000000,
 	0x00000000,
 	0x00000000,
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnvf0.fuc.h b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnvf0.fuc.h
index 438506d14749..229c0ae37228 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnvf0.fuc.h
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/hubnvf0.fuc.h
@@ -206,14 +206,14 @@ uint32_t nvf0_grhub_data[] = {
 };
 
 uint32_t nvf0_grhub_code[] = {
-	0x031b0ef5,
+	0x039b0ef5,
 /* 0x0004: queue_put */
 	0x9800d898,
 	0x86f001d9,
 	0x0489b808,
 	0xf00c1bf4,
 	0x21f502f7,
-	0x00f802fe,
+	0x00f8037e,
 /* 0x001c: queue_put_next */
 	0xb60798c4,
 	0x8dbb0384,
@@ -237,184 +237,214 @@ uint32_t nvf0_grhub_code[] = {
 /* 0x0066: queue_get_done */
 	0x00f80132,
 /* 0x0068: nv_rd32 */
-	0x0728b7f1,
-	0xb906b4b6,
-	0xc9f002ec,
-	0x00bcd01f,
-/* 0x0078: nv_rd32_wait */
-	0xc800bccf,
-	0x1bf41fcc,
-	0x06a7f0fa,
-	0x010921f5,
-	0xf840bfcf,
-/* 0x008d: nv_wr32 */
-	0x28b7f100,
-	0x06b4b607,
-	0xb980bfd0,
-	0xc9f002ec,
-	0x1ec9f01f,
-/* 0x00a3: nv_wr32_wait */
-	0xcf00bcd0,
-	0xccc800bc,
-	0xfa1bf41f,
-/* 0x00ae: watchdog_reset */
-	0x87f100f8,
-	0x84b60430,
-	0x1ff9f006,
-	0xf8008fd0,
-/* 0x00bd: watchdog_clear */
-	0x3087f100,
-	0x0684b604,
-	0xf80080d0,
-/* 0x00c9: wait_donez */
-	0xf094bd00,
-	0x07f10099,
-	0x03f03700,
-	0x0009d002,
-	0x07f104bd,
-	0x03f00600,
-	0x000ad002,
-/* 0x00e6: wait_donez_ne */
-	0x87f104bd,
-	0x83f00000,
-	0x0088cf01,
-	0xf4888aff,
-	0x94bdf31b,
-	0xf10099f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0109: wait_doneo */
-	0xf094bd00,
+	0xf002ecb9,
+	0x07f11fc9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x007a: nv_rd32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0xa7f0f31b,
+	0x1021f506,
+	0x00f7f101,
+	0x01f3f0cb,
+	0xf800ffcf,
+/* 0x009d: nv_wr32 */
+	0x0007f100,
+	0x0103f0cc,
+	0xbd000fd0,
+	0x02ecb904,
+	0xf01fc9f0,
+	0x07f11ec9,
+	0x03f0ca00,
+	0x000cd001,
+/* 0x00be: nv_wr32_wait */
+	0xc7f104bd,
+	0xc3f0ca00,
+	0x00cccf01,
+	0xf41fccc8,
+	0x00f8f31b,
+/* 0x00d0: wait_donez */
+	0x99f094bd,
+	0x0007f100,
+	0x0203f037,
+	0xbd0009d0,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x00ed: wait_donez_ne */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x1bf4888a,
+	0xf094bdf3,
 	0x07f10099,
-	0x03f03700,
+	0x03f01700,
 	0x0009d002,
-	0x87f104bd,
-	0x84b60818,
-	0x008ad006,
-/* 0x0124: wait_doneo_e */
-	0x040087f1,
-	0xcf0684b6,
-	0x8aff0088,
-	0xf30bf488,
+	0x00f804bd,
+/* 0x0110: wait_doneo */
 	0x99f094bd,
 	0x0007f100,
-	0x0203f017,
+	0x0203f037,
 	0xbd0009d0,
-/* 0x0147: mmctx_size */
-	0xbd00f804,
-/* 0x0149: nv_mmctx_size_loop */
-	0x00e89894,
-	0xb61a85b6,
-	0x84b60180,
-	0x0098bb02,
-	0xb804e0b6,
-	0x1bf404ef,
-	0x029fb9eb,
-/* 0x0166: mmctx_xfer */
-	0x94bd00f8,
-	0xf10199f0,
-	0xf0370007,
-	0x09d00203,
-	0xf104bd00,
-	0xb6071087,
-	0x94bd0684,
-	0xf405bbfd,
-	0x8bd0090b,
-	0x0099f000,
-/* 0x018c: mmctx_base_disabled */
-	0xf405eefd,
-	0x8ed00c0b,
-	0xc08fd080,
-/* 0x019b: mmctx_multi_disabled */
-	0xb70199f0,
-	0xc8010080,
+	0x0007f104,
+	0x0203f006,
+	0xbd000ad0,
+/* 0x012d: wait_doneo_e */
+	0x0087f104,
+	0x0183f000,
+	0xff0088cf,
+	0x0bf4888a,
+	0xf094bdf3,
+	0x07f10099,
+	0x03f01700,
+	0x0009d002,
+	0x00f804bd,
+/* 0x0150: mmctx_size */
+/* 0x0152: nv_mmctx_size_loop */
+	0xe89894bd,
+	0x1a85b600,
+	0xb60180b6,
+	0x98bb0284,
+	0x04e0b600,
+	0xf404efb8,
+	0x9fb9eb1b,
+/* 0x016f: mmctx_xfer */
+	0xbd00f802,
+	0x0199f094,
+	0x370007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0xbbfd94bd,
+	0x120bf405,
+	0xc40007f1,
+	0xd00103f0,
+	0x04bd000b,
+/* 0x0197: mmctx_base_disabled */
+	0xfd0099f0,
+	0x0bf405ee,
+	0x0007f11e,
+	0x0103f0c6,
+	0xbd000ed0,
+	0x0007f104,
+	0x0103f0c7,
+	0xbd000fd0,
+	0x0199f004,
+/* 0x01b8: mmctx_multi_disabled */
+	0xb600abc8,
+	0xb9f010b4,
+	0x01aec80c,
+	0xfd11e4b6,
+	0x07f105be,
+	0x03f0c500,
+	0x000bd001,
+/* 0x01d6: mmctx_exec_loop */
+/* 0x01d6: mmctx_wait_free */
+	0xe7f104bd,
+	0xe3f0c500,
+	0x00eecf01,
+	0xf41fe4f0,
+	0xce98f30b,
+	0x05e9fd00,
+	0xc80007f1,
+	0xd00103f0,
+	0x04bd000e,
+	0xb804c0b6,
+	0x1bf404cd,
+	0x02abc8d8,
+/* 0x0207: mmctx_fini_wait */
+	0xf11f1bf4,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x1fb4f000,
+	0xf410b4b0,
+	0xa7f0f01b,
+	0xd021f402,
+/* 0x0223: mmctx_stop */
+	0xc82b0ef4,
 	0xb4b600ab,
 	0x0cb9f010,
-	0xb601aec8,
-	0xbefd11e4,
-	0x008bd005,
-/* 0x01b4: mmctx_exec_loop */
-/* 0x01b4: mmctx_wait_free */
-	0xf0008ecf,
-	0x0bf41fe4,
-	0x00ce98fa,
-	0xd005e9fd,
-	0xc0b6c08e,
-	0x04cdb804,
-	0xc8e81bf4,
-	0x1bf402ab,
-/* 0x01d5: mmctx_fini_wait */
-	0x008bcf18,
-	0xb01fb4f0,
-	0x1bf410b4,
-	0x02a7f0f7,
-	0xf4c921f4,
-/* 0x01ea: mmctx_stop */
-	0xabc81b0e,
-	0x10b4b600,
-	0xf00cb9f0,
-	0x8bd012b9,
-/* 0x01f9: mmctx_stop_wait */
-	0x008bcf00,
-	0xf412bbc8,
-/* 0x0202: mmctx_done */
-	0x94bdfa1b,
-	0xf10199f0,
-	0xf0170007,
-	0x09d00203,
-	0xf804bd00,
-/* 0x0215: strand_wait */
-	0xf0a0f900,
-	0x21f402a7,
-	0xf8a0fcc9,
-/* 0x0221: strand_pre */
-	0xfc87f100,
-	0x0283f04a,
-	0xd00c97f0,
-	0x21f50089,
-	0x00f80215,
-/* 0x0234: strand_post */
-	0x4afc87f1,
-	0xf00283f0,
-	0x89d00d97,
-	0x1521f500,
-/* 0x0247: strand_set */
-	0xf100f802,
-	0xf04ffca7,
-	0xaba202a3,
-	0xc7f00500,
-	0x00acd00f,
-	0xd00bc7f0,
-	0x21f500bc,
-	0xaed00215,
-	0x0ac7f000,
-	0xf500bcd0,
-	0xf8021521,
-/* 0x0271: strand_ctx_init */
-	0xf094bd00,
-	0x07f10399,
-	0x03f03700,
+	0xf112b9f0,
+	0xf0c50007,
+	0x0bd00103,
+/* 0x023b: mmctx_stop_wait */
+	0xf104bd00,
+	0xf0c500b7,
+	0xbbcf01b3,
+	0x12bbc800,
+/* 0x024b: mmctx_done */
+	0xbdf31bf4,
+	0x0199f094,
+	0x170007f1,
+	0xd00203f0,
+	0x04bd0009,
+/* 0x025e: strand_wait */
+	0xa0f900f8,
+	0xf402a7f0,
+	0xa0fcd021,
+/* 0x026a: strand_pre */
+	0x97f000f8,
+	0xfc07f10c,
+	0x0203f04a,
+	0xbd0009d0,
+	0x5e21f504,
+/* 0x027f: strand_post */
+	0xf000f802,
+	0x07f10d97,
+	0x03f04afc,
 	0x0009d002,
 	0x21f504bd,
-	0xe7f00221,
-	0x4721f503,
-	0xfca7f102,
-	0x02a3f046,
-	0x0400aba0,
-	0xf040a0d0,
-	0xbcd001c7,
-	0x1521f500,
-	0x010c9202,
-	0xf000acd0,
-	0xbcd002c7,
-	0x1521f500,
-	0x3421f502,
-	0x8087f102,
-	0x0684b608,
-	0xb70089cf,
-	0x95220080,
-/* 0x02ca: ctx_init_strand_loop */
+	0x00f8025e,
+/* 0x0294: strand_set */
+	0xf10fc7f0,
+	0xf04ffc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f10bc7,
+	0x03f04afc,
+	0x000cd002,
+	0x07f104bd,
+	0x03f04ffc,
+	0x000ed002,
+	0xc7f004bd,
+	0xfc07f10a,
+	0x0203f04a,
+	0xbd000cd0,
+	0x5e21f504,
+/* 0x02d3: strand_ctx_init */
+	0xbd00f802,
+	0x0399f094,
+	0x370007f1,
+	0xd00203f0,
+	0x04bd0009,
+	0x026a21f5,
+	0xf503e7f0,
+	0xbd029421,
+	0xfc07f1c4,
+	0x0203f047,
+	0xbd000cd0,
+	0x01c7f004,
+	0x4afc07f1,
+	0xd00203f0,
+	0x04bd000c,
+	0x025e21f5,
+	0xf1010c92,
+	0xf046fc07,
+	0x0cd00203,
+	0xf004bd00,
+	0x07f102c7,
+	0x03f04afc,
+	0x000cd002,
+	0x21f504bd,
+	0x21f5025e,
+	0x87f1027f,
+	0x83f04200,
+	0x0097f102,
+	0x0293f020,
+	0x950099cf,
+/* 0x034a: ctx_init_strand_loop */
 	0x8ed008fe,
 	0x408ed000,
 	0xb6808acf,
@@ -428,7 +458,7 @@ uint32_t nvf0_grhub_code[] = {
 	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-/* 0x02fe: error */
+/* 0x037e: error */
 	0x07f100f8,
 	0x03f00500,
 	0x000fd002,
@@ -436,82 +466,117 @@ uint32_t nvf0_grhub_code[] = {
 	0x0007f101,
 	0x0303f007,
 	0xbd000fd0,
-/* 0x031b: init */
+/* 0x039b: init */
 	0xbd00f804,
-	0x0004fe04,
-	0xf10007fe,
-	0xf0120017,
-	0x12d00227,
-	0xb117f100,
-	0x0010fe05,
-	0x040017f1,
-	0xf1c010d0,
-	0xb6040437,
-	0x27f10634,
-	0x32d02003,
-	0x0427f100,
-	0x0132d020,
+	0x0007fe04,
+	0x420017f1,
+	0xcf0013f0,
+	0x11e70011,
+	0x14b60109,
+	0x0014fe08,
+	0xf10227f0,
+	0xf0120007,
+	0x02d00003,
+	0xf104bd00,
+	0xfe06c817,
+	0x24bd0010,
+	0x070007f1,
+	0xd00003f0,
+	0x04bd0002,
+	0x200327f1,
+	0x010007f1,
+	0xd00103f0,
+	0x04bd0002,
+	0x200427f1,
+	0x010407f1,
+	0xd00103f0,
+	0x04bd0002,
 	0x200b27f1,
-	0xf10232d0,
-	0xd0200c27,
-	0x27f10732,
-	0x24b60c24,
-	0x0003b906,
-	0xf10023d0,
+	0x010807f1,
+	0xd00103f0,
+	0x04bd0002,
+	0x200c27f1,
+	0x011c07f1,
+	0xd00103f0,
+	0x04bd0002,
+	0xf1010392,
+	0xf0090007,
+	0x03d00303,
+	0xf104bd00,
 	0xf0870427,
-	0x12d00023,
-	0x0012b700,
-	0x0427f001,
-	0xf40012d0,
-	0xe7f11031,
-	0xe3f09604,
-	0x6821f440,
-	0x8090f1c7,
-	0xf4f00301,
-	0x020f801f,
-	0xbb0117f0,
-	0x12b6041f,
-	0x0c27f101,
-	0x0624b604,
-	0xd00021d0,
-	0x17f14021,
-	0x0e980100,
-	0x010f9800,
-	0x014721f5,
-	0x070037f1,
-	0x950634b6,
-	0x34d00814,
-	0x4034d000,
-	0x130030b7,
-	0xb6001fbb,
-	0x3fd002f5,
-	0x0815b600,
-	0xb60110b6,
-	0x1fb90814,
-	0x7121f502,
-	0x001fbb02,
-	0xf1020398,
-	0xf0200047,
-/* 0x03f6: init_gpc */
-	0x4ea05043,
-	0x1fb90804,
-	0x8d21f402,
-	0x010c4ea0,
-	0x21f4f4bd,
-	0x044ea08d,
-	0x8d21f401,
-	0x01004ea0,
-	0xf402f7f0,
-	0x4ea08d21,
-/* 0x041e: init_gpc_wait */
-	0x21f40800,
-	0x1fffc868,
-	0xa0fa0bf4,
-	0xf408044e,
-	0x1fbb6821,
-	0x0040b700,
-	0x0132b680,
-	0xf1be1bf4,
+	0x07f10023,
+	0x03f00400,
+	0x0002d000,
+	0x27f004bd,
+	0x0007f104,
+	0x0003f003,
+	0xbd0002d0,
+	0x1031f404,
+	0x9604e7f1,
+	0xf440e3f0,
+	0xfeb96821,
+	0x90f1c702,
+	0xf0030180,
+	0x0f801ff4,
+	0x0117f002,
+	0xb6041fbb,
+	0x07f10112,
+	0x03f00300,
+	0x0001d001,
+	0x07f104bd,
+	0x03f00400,
+	0x0001d001,
+	0x17f104bd,
+	0xf7f00100,
+	0x7f21f502,
+	0x9121f507,
+	0x10f7f007,
+	0x07de21f5,
+	0x98000e98,
+	0x21f5010f,
+	0x14950150,
+	0x0007f108,
+	0x0103f0c0,
+	0xbd0004d0,
+	0x0007f104,
+	0x0103f0c1,
+	0xbd0004d0,
+	0x0030b704,
+	0x001fbb13,
+	0xf102f5b6,
+	0xf0d30007,
+	0x0fd00103,
+	0xb604bd00,
+	0x10b60815,
+	0x0814b601,
+	0xf5021fb9,
+	0xbb02d321,
+	0x0398001f,
+	0x0047f102,
+	0x5043f020,
+/* 0x04f4: init_gpc */
+	0x08044ea0,
+	0xf4021fb9,
+	0x4ea09d21,
+	0xf4bd010c,
+	0xa09d21f4,
+	0xf401044e,
+	0x4ea09d21,
+	0xf7f00100,
+	0x9d21f402,
+	0x08004ea0,
+/* 0x051c: init_gpc_wait */
+	0xc86821f4,
+	0x0bf41fff,
+	0x044ea0fa,
+	0x6821f408,
+	0xb7001fbb,
+	0xb6800040,
+	0x1bf40132,
+	0x00f7f0be,
+	0x07de21f5,
+	0xf500f7f0,
+	0xf1077f21,
 	0xf0010007,
 	0x01d00203,
 	0xbd04bd00,
@@ -519,382 +584,379 @@ uint32_t nvf0_grhub_code[] = {
 	0x300007f1,
 	0xd00203f0,
 	0x04bd0001,
-/* 0x0458: main */
+/* 0x0564: main */
 	0xf40031f4,
 	0xd7f00028,
 	0x3921f410,
 	0xb1f401f4,
 	0xf54001e4,
-	0xbd00de1b,
+	0xbd00e91b,
 	0x0499f094,
 	0x370007f1,
 	0xd00203f0,
 	0x04bd0009,
-	0x0b0017f1,
-	0xcf0614b6,
-	0x11cf4012,
-	0x1f13c800,
-	0x00870bf5,
-	0xf41f23c8,
-	0x20f9620b,
-	0xbd0212b9,
-	0x0799f094,
-	0x370007f1,
-	0xd00203f0,
-	0x04bd0009,
-	0xf40132f4,
-	0x21f50231,
-	0x94bd0801,
+	0xc00017f1,
+	0xcf0213f0,
+	0x27f10011,
+	0x23f0c100,
+	0x0022cf02,
+	0xf51f13c8,
+	0xc800890b,
+	0x0bf41f23,
+	0xb920f962,
+	0x94bd0212,
 	0xf10799f0,
-	0xf0170007,
+	0xf0370007,
 	0x09d00203,
-	0xfc04bd00,
-	0xf094bd20,
-	0x07f10699,
-	0x03f03700,
-	0x0009d002,
-	0x31f404bd,
-	0x0121f501,
-	0xf094bd08,
-	0x07f10699,
+	0xf404bd00,
+	0x31f40132,
+	0xaa21f502,
+	0xf094bd09,
+	0x07f10799,
 	0x03f01700,
 	0x0009d002,
-	0x0ef404bd,
-/* 0x04f9: chsw_prev_no_next */
-	0xb920f931,
-	0x32f40212,
-	0x0232f401,
-	0x080121f5,
-	0x17f120fc,
-	0x14b60b00,
-	0x0012d006,
-/* 0x0517: chsw_no_prev */
-	0xc8130ef4,
-	0x0bf41f23,
-	0x0131f40d,
-	0xf50232f4,
-/* 0x0527: chsw_done */
-	0xf1080121,
-	0xb60b0c17,
-	0x27f00614,
-	0x0012d001,
+	0x20fc04bd,
 	0x99f094bd,
-	0x0007f104,
+	0x0007f106,
+	0x0203f037,
+	0xbd0009d0,
+	0x0131f404,
+	0x09aa21f5,
+	0x99f094bd,
+	0x0007f106,
 	0x0203f017,
 	0xbd0009d0,
-	0x130ef504,
-/* 0x0549: main_not_ctx_switch */
-	0x01e4b0ff,
-	0xb90d1bf4,
-	0x21f502f2,
-	0x0ef40795,
-/* 0x0559: main_not_ctx_chan */
-	0x02e4b046,
-	0xbd321bf4,
-	0x0799f094,
-	0x370007f1,
+	0x330ef404,
+/* 0x060c: chsw_prev_no_next */
+	0x12b920f9,
+	0x0132f402,
+	0xf50232f4,
+	0xfc09aa21,
+	0x0007f120,
+	0x0203f0c0,
+	0xbd0002d0,
+	0x130ef404,
+/* 0x062c: chsw_no_prev */
+	0xf41f23c8,
+	0x31f40d0b,
+	0x0232f401,
+	0x09aa21f5,
+/* 0x063c: chsw_done */
+	0xf10127f0,
+	0xf0c30007,
+	0x02d00203,
+	0xbd04bd00,
+	0x0499f094,
+	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-	0xf40132f4,
-	0x21f50232,
-	0x94bd0801,
+	0xff080ef5,
+/* 0x0660: main_not_ctx_switch */
+	0xf401e4b0,
+	0xf2b90d1b,
+	0x4221f502,
+	0x460ef409,
+/* 0x0670: main_not_ctx_chan */
+	0xf402e4b0,
+	0x94bd321b,
 	0xf10799f0,
-	0xf0170007,
+	0xf0370007,
 	0x09d00203,
 	0xf404bd00,
-/* 0x058e: main_not_ctx_save */
-	0xef94110e,
-	0x01f5f010,
-	0x02fe21f5,
-	0xfec00ef5,
-/* 0x059c: main_done */
-	0x29f024bd,
-	0x0007f11f,
-	0x0203f030,
-	0xbd0002d0,
-	0xab0ef504,
-/* 0x05b1: ih */
-	0xfe80f9fe,
-	0x80f90188,
-	0xa0f990f9,
-	0xd0f9b0f9,
-	0xf0f9e0f9,
-	0x0acf04bd,
-	0x04abc480,
-	0xf11d0bf4,
-	0xf01900b7,
-	0xbecf10d7,
-	0x00bfcf40,
+	0x32f40132,
+	0xaa21f502,
+	0xf094bd09,
+	0x07f10799,
+	0x03f01700,
+	0x0009d002,
+	0x0ef404bd,
+/* 0x06a5: main_not_ctx_save */
+	0x10ef9411,
+	0xf501f5f0,
+	0xf5037e21,
+/* 0x06b3: main_done */
+	0xbdfeb50e,
+	0x1f29f024,
+	0x300007f1,
+	0xd00203f0,
+	0x04bd0002,
+	0xfea00ef5,
+/* 0x06c8: ih */
+	0x88fe80f9,
+	0xf980f901,
+	0xf9a0f990,
+	0xf9d0f9b0,
+	0xbdf0f9e0,
+	0x00a7f104,
+	0x00a3f002,
+	0xc400aacf,
+	0x0bf404ab,
+	0x10d7f030,
+	0x1a00e7f1,
+	0xcf00e3f0,
+	0xf7f100ee,
+	0xf3f01900,
+	0x00ffcf00,
 	0xb70421f4,
 	0xf00400b0,
-	0xbed001e7,
-/* 0x05e9: ih_no_fifo */
-	0x00abe400,
-	0x0d0bf401,
-	0xf110d7f0,
-	0xf44001e7,
-/* 0x05fa: ih_no_ctxsw */
-	0xb7f10421,
-	0xb0bd0104,
-	0xf4b4abff,
-	0xa7f10d0b,
-	0xa4b60c1c,
-	0x00abd006,
-/* 0x0610: ih_no_other */
-	0xfc400ad0,
+	0x07f101e7,
+	0x03f01d00,
+	0x000ed000,
+/* 0x071a: ih_no_fifo */
+	0xabe404bd,
+	0x0bf40100,
+	0x10d7f00d,
+	0x4001e7f1,
+/* 0x072b: ih_no_ctxsw */
+	0xe40421f4,
+	0xf40400ab,
+	0xb7f1140b,
+	0xbfb90100,
+	0x44e7f102,
+	0x40e3f001,
+/* 0x0743: ih_no_fwmthd */
+	0xf19d21f4,
+	0xbd0104b7,
+	0xb4abffb0,
+	0xf10f0bf4,
+	0xf0070007,
+	0x0bd00303,
+/* 0x075b: ih_no_other */
+	0xf104bd00,
+	0xf0010007,
+	0x0ad00003,
+	0xfc04bd00,
 	0xfce0fcf0,
 	0xfcb0fcd0,
 	0xfc90fca0,
 	0x0088fe80,
 	0x32f480fc,
-/* 0x062b: ctx_4170s */
-	0xf101f800,
-	0xf04170e7,
-	0xf5f040e3,
-	0x8d21f410,
-/* 0x063a: ctx_4170w */
+/* 0x077f: ctx_4170s */
+	0xf001f800,
+	0xffb910f5,
+	0x70e7f102,
+	0x40e3f041,
+	0xf89d21f4,
+/* 0x0791: ctx_4170w */
+	0x70e7f100,
+	0x40e3f041,
+	0xb96821f4,
+	0xf4f002ff,
+	0xf01bf410,
+/* 0x07a6: ctx_redswitch */
 	0xe7f100f8,
-	0xe3f04170,
-	0x6821f440,
-	0xf410f4f0,
+	0xe5f00200,
+	0x20e5f040,
+	0xf110e5f0,
+	0xf0850007,
+	0x0ed00103,
+	0xf004bd00,
+/* 0x07c2: ctx_redswitch_delay */
+	0xf2b608f7,
+	0xfd1bf401,
+	0x0400e5f1,
+	0x0100e5f1,
+	0x850007f1,
+	0xd00103f0,
+	0x04bd000e,
+/* 0x07de: ctx_86c */
+	0x07f100f8,
+	0x03f02300,
+	0x000fd002,
+	0xffb904bd,
+	0x14e7f102,
+	0x40e3f08a,
+	0xb99d21f4,
+	0xe7f102ff,
+	0xe3f0a88c,
+	0x9d21f441,
+/* 0x0806: ctx_mem */
+	0x07f100f8,
+	0x03f08400,
+	0x000fd002,
+/* 0x0812: ctx_mem_wait */
+	0xf7f104bd,
+	0xf3f08400,
+	0x00ffcf02,
+	0xf405fffd,
 	0x00f8f31b,
-/* 0x064c: ctx_redswitch */
-	0x0614e7f1,
-	0xf106e4b6,
-	0xd00270f7,
-	0xf7f000ef,
-/* 0x065d: ctx_redswitch_delay */
-	0x01f2b608,
-	0xf1fd1bf4,
-	0xd00770f7,
-	0x00f800ef,
-/* 0x066c: ctx_86c */
-	0x086ce7f1,
-	0xd006e4b6,
-	0xe7f100ef,
-	0xe3f08a14,
-	0x8d21f440,
-	0xa86ce7f1,
-	0xf441e3f0,
-	0x00f88d21,
-/* 0x068c: ctx_load */
+/* 0x0824: ctx_load */
 	0x99f094bd,
 	0x0007f105,
 	0x0203f037,
 	0xbd0009d0,
 	0x0ca7f004,
-	0xf1c921f4,
-	0xb60a2417,
-	0x10d00614,
-	0x0037f100,
-	0x0634b60b,
-	0xf14032d0,
-	0xb60a0c17,
-	0x47f00614,
-	0x0012d007,
-/* 0x06c7: ctx_chan_wait_0 */
-	0xcf4014d0,
-	0x44f04014,
-	0xfa1bf41f,
-	0xfe0032d0,
-	0x2af0000b,
-	0x0424b61f,
-	0xbd0220b6,
+	0xbdd021f4,
+	0x0007f1f4,
+	0x0203f089,
+	0xbd000fd0,
+	0x0007f104,
+	0x0203f0c1,
+	0xbd0002d0,
+	0x0007f104,
+	0x0203f083,
+	0xbd0002d0,
+	0x07f7f004,
+	0x080621f5,
+	0xc00007f1,
+	0xd00203f0,
+	0x04bd0002,
+	0xf0000bfe,
+	0x24b61f2a,
+	0x0220b604,
+	0x99f094bd,
+	0x0007f108,
+	0x0203f037,
+	0xbd0009d0,
+	0x0007f104,
+	0x0203f081,
+	0xbd0002d0,
+	0x0027f104,
+	0x0023f100,
+	0x0225f080,
+	0x880007f1,
+	0xd00203f0,
+	0x04bd0002,
+	0xf11017f0,
+	0xf0020027,
+	0x12fa0223,
+	0xbd03f805,
 	0x0899f094,
-	0x370007f1,
+	0x170007f1,
 	0xd00203f0,
 	0x04bd0009,
-	0x0a0417f1,
-	0xd00614b6,
-	0x17f10012,
-	0x14b60a20,
-	0x0227f006,
-	0x800023f1,
-	0xf00012d0,
-	0x27f11017,
-	0x23f00200,
-	0x0512fa02,
-	0x94bd03f8,
-	0xf10899f0,
-	0xf0170007,
+	0xb6810198,
+	0x02981814,
+	0x0825b680,
+	0x800512fd,
+	0x94bd1601,
+	0xf10999f0,
+	0xf0370007,
 	0x09d00203,
-	0x9804bd00,
-	0x14b68101,
-	0x80029818,
-	0xfd0825b6,
-	0x01800512,
-	0xf094bd16,
-	0x07f10999,
-	0x03f03700,
-	0x0009d002,
-	0x27f104bd,
-	0x24b60a04,
-	0x0021d006,
-	0xf10127f0,
-	0xb60a2017,
-	0x12d00614,
-	0x0017f100,
-	0x0613f001,
-	0xf80501fa,
-	0xf094bd03,
-	0x07f10999,
-	0x03f01700,
-	0x0009d002,
-	0x94bd04bd,
-	0xf10599f0,
+	0xf104bd00,
+	0xf0810007,
+	0x01d00203,
+	0xf004bd00,
+	0x07f10127,
+	0x03f08800,
+	0x0002d002,
+	0x17f104bd,
+	0x13f00100,
+	0x0501fa06,
+	0x94bd03f8,
+	0xf10999f0,
 	0xf0170007,
 	0x09d00203,
-	0xf804bd00,
-/* 0x0795: ctx_chan */
-	0x8c21f500,
-	0x0ca7f006,
-	0xf1c921f4,
-	0xb60a1017,
-	0x27f00614,
-	0x0012d005,
-/* 0x07ac: ctx_chan_wait */
-	0xfd0012cf,
-	0x1bf40522,
-/* 0x07b7: ctx_mmio_exec */
-	0x9800f8fa,
-	0x27f14103,
-	0x24b60a04,
-	0x0023d006,
-/* 0x07c6: ctx_mmio_loop */
+	0xbd04bd00,
+	0x0599f094,
+	0x170007f1,
+	0xd00203f0,
+	0x04bd0009,
+/* 0x0942: ctx_chan */
+	0x21f500f8,
+	0xa7f00824,
+	0xd021f40c,
+	0xf505f7f0,
+	0xf8080621,
+/* 0x0955: ctx_mmio_exec */
+	0x41039800,
+	0x810007f1,
+	0xd00203f0,
+	0x04bd0003,
+/* 0x0966: ctx_mmio_loop */
 	0x34c434bd,
 	0x0f1bf4ff,
 	0x020057f1,
 	0xfa0653f0,
 	0x03f80535,
-/* 0x07d8: ctx_mmio_pull */
+/* 0x0978: ctx_mmio_pull */
 	0x98804e98,
 	0x21f4814f,
-	0x0830b68d,
+	0x0830b69d,
 	0xf40112b6,
-/* 0x07ea: ctx_mmio_done */
+/* 0x098a: ctx_mmio_done */
 	0x0398df1b,
-	0x0023d016,
-	0xf1400080,
-	0xf0010017,
-	0x01fa0613,
-	0xf803f806,
-/* 0x0801: ctx_xfer */
-	0x00f7f100,
-	0x06f4b60c,
-	0xd004e7f0,
-/* 0x080e: ctx_xfer_idle */
-	0xfecf80fe,
-	0x00e4f100,
-	0xf91bf420,
-	0xf40611f4,
-/* 0x081e: ctx_xfer_pre */
-	0xf7f00d02,
-	0x6c21f510,
-	0x1c11f406,
-/* 0x0828: ctx_xfer_pre_load */
-	0xf502f7f0,
-	0xf5062b21,
-	0xf5063a21,
-	0xbd064c21,
-	0x2b21f5f4,
-	0x8c21f506,
-/* 0x0841: ctx_xfer_exec */
-	0x16019806,
-	0x041427f1,
-	0xd00624b6,
-	0xe7f10020,
-	0xe3f0a500,
-	0x021fb941,
-	0xb68d21f4,
-	0xfcf004e0,
-	0x022cf001,
-	0xfd0124b6,
-	0x21f405f2,
-	0xfc17f18d,
-	0x0213f04a,
-	0xd00c27f0,
-	0x21f50012,
-	0x27f10215,
-	0x23f047fc,
-	0x0020d002,
+	0x0007f116,
+	0x0203f081,
+	0xbd0003d0,
+	0x40008004,
+	0x010017f1,
+	0xfa0613f0,
+	0x03f80601,
+/* 0x09aa: ctx_xfer */
+	0xe7f000f8,
+	0x0007f104,
+	0x0303f002,
+	0xbd000ed0,
+/* 0x09b9: ctx_xfer_idle */
+	0x00e7f104,
+	0x03e3f000,
+	0xf100eecf,
+	0xf42000e4,
+	0x11f4f21b,
+	0x0d02f406,
+/* 0x09d0: ctx_xfer_pre */
+	0xf510f7f0,
+	0xf407de21,
+/* 0x09da: ctx_xfer_pre_load */
+	0xf7f01c11,
+	0x7f21f502,
+	0x9121f507,
+	0xa621f507,
+	0xf5f4bd07,
+	0xf5077f21,
+/* 0x09f3: ctx_xfer_exec */
+	0x98082421,
+	0x24bd1601,
+	0x050007f1,
+	0xd00103f0,
+	0x04bd0002,
+	0xf1021fb9,
+	0xf0a500e7,
+	0x21f441e3,
+	0x01fcf09d,
+	0xb6022cf0,
+	0xf2fd0124,
+	0x02ffb905,
+	0xa504e7f1,
+	0xf441e3f0,
+	0x21f59d21,
+	0x24bd026a,
+	0x47fc07f1,
+	0xd00203f0,
+	0x04bd0002,
 	0xb6012cf0,
-	0x12d00320,
-	0x01acf000,
-	0xf006a5f0,
-	0x0c9800b7,
-	0x010d9800,
-	0xf500e7f0,
-	0xf0016621,
-	0x21f508a7,
-	0x21f50109,
-	0x01f40215,
-	0x0ca7f022,
-	0xf1c921f4,
-	0xb60a1017,
-	0x27f00614,
-	0x0012d005,
-/* 0x08c8: ctx_xfer_post_save_wait */
-	0xfd0012cf,
-	0x1bf40522,
-	0x2e02f4fa,
-/* 0x08d4: ctx_xfer_post */
-	0xf502f7f0,
-	0xbd062b21,
-	0x6c21f5f4,
-	0x3421f506,
-	0x3a21f502,
-	0xf5f4bd06,
-	0xf4062b21,
-	0x01981011,
-	0x0511fd40,
-	0xf5070bf4,
-/* 0x08ff: ctx_xfer_no_post_mmio */
-/* 0x08ff: ctx_xfer_done */
-	0xf807b721,
-	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,
+	0x07f10320,
+	0x03f04afc,
+	0x0002d002,
+	0xacf004bd,
+	0x06a5f001,
+	0x9800b7f0,
+	0x0d98000c,
+	0x00e7f001,
+	0x016f21f5,
+	0xf508a7f0,
+	0xf5011021,
+	0xf4025e21,
+	0xa7f01301,
+	0xd021f40c,
+	0xf505f7f0,
+	0xf4080621,
+/* 0x0a82: ctx_xfer_post */
+	0xf7f02e02,
+	0x7f21f502,
+	0xf5f4bd07,
+	0xf507de21,
+	0xf5027f21,
+	0xbd079121,
+	0x7f21f5f4,
+	0x1011f407,
+	0xfd400198,
+	0x0bf40511,
+	0x5521f507,
+/* 0x0aad: ctx_xfer_no_post_mmio */
+/* 0x0aad: ctx_xfer_done */
+	0x0000f809,
 	0x00000000,
 	0x00000000,
 	0x00000000,
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/macros.fuc b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/macros.fuc
index 33a5a82eccbd..6ffe28307dbd 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/fuc/macros.fuc
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/fuc/macros.fuc
@@ -28,28 +28,135 @@
 #define GF117 0xd7
 #define GK100 0xe0
 #define GK110 0xf0
+#define GK208 0x108
 
+#define NV_PGRAPH_FECS_INTR_ACK                                        0x409004
+#define NV_PGRAPH_FECS_INTR                                            0x409008
+#define NV_PGRAPH_FECS_INTR_FWMTHD                                   0x00000400
+#define NV_PGRAPH_FECS_INTR_CHSW                                     0x00000100
+#define NV_PGRAPH_FECS_INTR_FIFO                                     0x00000004
+#define NV_PGRAPH_FECS_INTR_MODE                                       0x40900c
+#define NV_PGRAPH_FECS_INTR_MODE_FIFO                                0x00000004
+#define NV_PGRAPH_FECS_INTR_MODE_FIFO_LEVEL                          0x00000004
+#define NV_PGRAPH_FECS_INTR_MODE_FIFO_EDGE                           0x00000000
+#define NV_PGRAPH_FECS_INTR_EN_SET                                     0x409010
+#define NV_PGRAPH_FECS_INTR_EN_SET_FIFO                              0x00000004
+#define NV_PGRAPH_FECS_INTR_ROUTE                                      0x40901c
+#define NV_PGRAPH_FECS_ACCESS                                          0x409048
+#define NV_PGRAPH_FECS_ACCESS_FIFO                                   0x00000002
+#define NV_PGRAPH_FECS_FIFO_DATA                                       0x409064
+#define NV_PGRAPH_FECS_FIFO_CMD                                        0x409068
+#define NV_PGRAPH_FECS_FIFO_ACK                                        0x409074
+#define NV_PGRAPH_FECS_CAPS                                            0x409108
 #define NV_PGRAPH_FECS_SIGNAL                                          0x409400
+#define NV_PGRAPH_FECS_IROUTE                                          0x409404
+#define NV_PGRAPH_FECS_BAR_MASK0                                       0x40940c
+#define NV_PGRAPH_FECS_BAR_MASK1                                       0x409410
+#define NV_PGRAPH_FECS_BAR                                             0x409414
+#define NV_PGRAPH_FECS_BAR_SET                                         0x409418
+#define NV_PGRAPH_FECS_RED_SWITCH                                      0x409614
+#define NV_PGRAPH_FECS_RED_SWITCH_ENABLE_ROP                         0x00000400
+#define NV_PGRAPH_FECS_RED_SWITCH_ENABLE_GPC                         0x00000200
+#define NV_PGRAPH_FECS_RED_SWITCH_ENABLE_MAIN                        0x00000100
+#define NV_PGRAPH_FECS_RED_SWITCH_POWER_ROP                          0x00000040
+#define NV_PGRAPH_FECS_RED_SWITCH_POWER_GPC                          0x00000020
+#define NV_PGRAPH_FECS_RED_SWITCH_POWER_MAIN                         0x00000010
+#define NV_PGRAPH_FECS_RED_SWITCH_PAUSE_GPC                          0x00000002
+#define NV_PGRAPH_FECS_RED_SWITCH_PAUSE_MAIN                         0x00000001
+#define NV_PGRAPH_FECS_MMCTX_SAVE_SWBASE                               0x409700
+#define NV_PGRAPH_FECS_MMCTX_LOAD_SWBASE                               0x409704
+#define NV_PGRAPH_FECS_MMCTX_LOAD_COUNT                                0x40974c
+#define NV_PGRAPH_FECS_MMCTX_SAVE_SWBASE                               0x409700
+#define NV_PGRAPH_FECS_MMCTX_LOAD_SWBASE                               0x409704
+#define NV_PGRAPH_FECS_MMCTX_BASE                                      0x409710
+#define NV_PGRAPH_FECS_MMCTX_CTRL                                      0x409714
+#define NV_PGRAPH_FECS_MMCTX_MULTI_STRIDE                              0x409718
+#define NV_PGRAPH_FECS_MMCTX_MULTI_MASK                                0x40971c
+#define NV_PGRAPH_FECS_MMCTX_QUEUE                                     0x409720
+#define NV_PGRAPH_FECS_MMIO_CTRL                                       0x409728
+#define NV_PGRAPH_FECS_MMIO_RDVAL                                      0x40972c
+#define NV_PGRAPH_FECS_MMIO_WRVAL                                      0x409730
+#define NV_PGRAPH_FECS_MMCTX_LOAD_COUNT                                0x40974c
 #if CHIPSET < GK110
 #define NV_PGRAPH_FECS_CC_SCRATCH_VAL(n)                    ((n) * 4 + 0x409800)
 #define NV_PGRAPH_FECS_CC_SCRATCH_SET(n)                    ((n) * 4 + 0x409820)
 #define NV_PGRAPH_FECS_CC_SCRATCH_CLR(n)                    ((n) * 4 + 0x409840)
+#define NV_PGRAPH_FECS_UNK86C                                          0x40986c
 #else
 #define NV_PGRAPH_FECS_CC_SCRATCH_VAL(n)                    ((n) * 4 + 0x409800)
 #define NV_PGRAPH_FECS_CC_SCRATCH_CLR(n)                    ((n) * 4 + 0x409840)
+#define NV_PGRAPH_FECS_UNK86C                                          0x40988c
 #define NV_PGRAPH_FECS_CC_SCRATCH_SET(n)                    ((n) * 4 + 0x4098c0)
 #endif
+#define NV_PGRAPH_FECS_STRANDS_CNT                                     0x409880
+#define NV_PGRAPH_FECS_STRAND_SAVE_SWBASE                              0x409908
+#define NV_PGRAPH_FECS_STRAND_LOAD_SWBASE                              0x40990c
+#define NV_PGRAPH_FECS_STRAND_WORDS                                    0x409910
+#define NV_PGRAPH_FECS_STRAND_DATA                                     0x409918
+#define NV_PGRAPH_FECS_STRAND_SELECT                                   0x40991c
+#define NV_PGRAPH_FECS_STRAND_CMD                                      0x409928
+#define NV_PGRAPH_FECS_STRAND_CMD_SEEK                               0x00000001
+#define NV_PGRAPH_FECS_STRAND_CMD_GET_INFO                           0x00000002
+#define NV_PGRAPH_FECS_STRAND_CMD_SAVE                               0x00000003
+#define NV_PGRAPH_FECS_STRAND_CMD_LOAD                               0x00000004
+#define NV_PGRAPH_FECS_STRAND_CMD_ACTIVATE_FILTER                    0x0000000a
+#define NV_PGRAPH_FECS_STRAND_CMD_DEACTIVATE_FILTER                  0x0000000b
+#define NV_PGRAPH_FECS_STRAND_CMD_ENABLE                             0x0000000c
+#define NV_PGRAPH_FECS_STRAND_CMD_DISABLE                            0x0000000d
+#define NV_PGRAPH_FECS_STRAND_FILTER                                   0x40993c
+#define NV_PGRAPH_FECS_MEM_BASE                                        0x409a04
+#define NV_PGRAPH_FECS_MEM_CHAN                                        0x409a0c
+#define NV_PGRAPH_FECS_MEM_CMD                                         0x409a10
+#define NV_PGRAPH_FECS_MEM_CMD_LOAD_CHAN                             0x00000007
+#define NV_PGRAPH_FECS_MEM_TARGET                                      0x409a20
+#define NV_PGRAPH_FECS_MEM_TARGET_UNK31                              0x80000000
+#define NV_PGRAPH_FECS_MEM_TARGET_AS                                 0x0000001f
+#define NV_PGRAPH_FECS_MEM_TARGET_AS_VM                              0x00000001
+#define NV_PGRAPH_FECS_MEM_TARGET_AS_VRAM                            0x00000002
+#define NV_PGRAPH_FECS_CHAN_ADDR                                       0x409b00
+#define NV_PGRAPH_FECS_CHAN_NEXT                                       0x409b04
+#define NV_PGRAPH_FECS_CHSW                                            0x409b0c
+#define NV_PGRAPH_FECS_CHSW_ACK                                      0x00000001
 #define NV_PGRAPH_FECS_INTR_UP_SET                                     0x409c1c
+#define NV_PGRAPH_FECS_INTR_UP_EN                                      0x409c24
 
+#define NV_PGRAPH_GPCX_GPCCS_INTR_ACK                                  0x41a004
+#define NV_PGRAPH_GPCX_GPCCS_INTR                                      0x41a008
+#define NV_PGRAPH_GPCX_GPCCS_INTR_FIFO                               0x00000004
+#define NV_PGRAPH_GPCX_GPCCS_INTR_EN_SET                               0x41a010
+#define NV_PGRAPH_GPCX_GPCCS_INTR_EN_SET_FIFO                        0x00000004
+#define NV_PGRAPH_GPCX_GPCCS_INTR_ROUTE                                0x41a01c
+#define NV_PGRAPH_GPCX_GPCCS_ACCESS                                    0x41a048
+#define NV_PGRAPH_GPCX_GPCCS_ACCESS_FIFO                             0x00000002
+#define NV_PGRAPH_GPCX_GPCCS_FIFO_DATA                                 0x41a064
+#define NV_PGRAPH_GPCX_GPCCS_FIFO_CMD                                  0x41a068
+#define NV_PGRAPH_GPCX_GPCCS_FIFO_ACK                                  0x41a074
+#define NV_PGRAPH_GPCX_GPCCS_UNITS                                     0x41a608
+#define NV_PGRAPH_GPCX_GPCCS_RED_SWITCH                                0x41a614
+#define NV_PGRAPH_GPCX_GPCCS_RED_SWITCH_UNK11                        0x00000800
+#define NV_PGRAPH_GPCX_GPCCS_RED_SWITCH_ENABLE                       0x00000200
+#define NV_PGRAPH_GPCX_GPCCS_RED_SWITCH_POWER                        0x00000020
+#define NV_PGRAPH_GPCX_GPCCS_RED_SWITCH_PAUSE                        0x00000002
+#define NV_PGRAPH_GPCX_GPCCS_MYINDEX                                   0x41a618
+#define NV_PGRAPH_GPCX_GPCCS_MMCTX_SAVE_SWBASE                         0x41a700
+#define NV_PGRAPH_GPCX_GPCCS_MMCTX_LOAD_SWBASE                         0x41a704
+#define NV_PGRAPH_GPCX_GPCCS_MMCTX_LOAD_COUNT                          0x41a74c
 #if CHIPSET < GK110
 #define NV_PGRAPH_GPCX_GPCCS_CC_SCRATCH_VAL(n)              ((n) * 4 + 0x41a800)
 #define NV_PGRAPH_GPCX_GPCCS_CC_SCRATCH_SET(n)              ((n) * 4 + 0x41a820)
 #define NV_PGRAPH_GPCX_GPCCS_CC_SCRATCH_CLR(n)              ((n) * 4 + 0x41a840)
+#define NV_PGRAPH_GPCX_GPCCS_UNK86C                                    0x41a86c
 #else
 #define NV_PGRAPH_GPCX_GPCCS_CC_SCRATCH_VAL(n)              ((n) * 4 + 0x41a800)
 #define NV_PGRAPH_GPCX_GPCCS_CC_SCRATCH_CLR(n)              ((n) * 4 + 0x41a840)
+#define NV_PGRAPH_GPCX_GPCCS_UNK86C                                    0x41a88c
 #define NV_PGRAPH_GPCX_GPCCS_CC_SCRATCH_SET(n)              ((n) * 4 + 0x41a8c0)
 #endif
+#define NV_PGRAPH_GPCX_GPCCS_STRAND_SELECT                             0x41a91c
+#define NV_PGRAPH_GPCX_GPCCS_STRAND_CMD                                0x41a928
+#define NV_PGRAPH_GPCX_GPCCS_STRAND_CMD_SAVE                         0x00000003
+#define NV_PGRAPH_GPCX_GPCCS_STRAND_CMD_LOAD                         0x00000004
+#define NV_PGRAPH_GPCX_GPCCS_MEM_BASE                                  0x41aa04
 
 #define mmctx_data(r,c) .b32 (((c - 1) << 26) | r)
 #define queue_init      .skip 72 // (2 * 4) + ((8 * 4) * 2)
@@ -65,24 +172,50 @@
 #define T_LCHAN   8
 #define T_LCTXH   9
 
-#define nv_mkmm(rv,r) /*
-*/	movw rv  ((r) & 0x0000fffc) /*
-*/	sethi rv ((r) & 0x00ff0000)
+#if CHIPSET < GK208
+#define imm32(reg,val) /*
+*/	movw reg  ((val) & 0x0000ffff) /*
+*/	sethi reg ((val) & 0xffff0000)
+#else
+#define imm32(reg,val) /*
+*/	mov reg (val)
+#endif
+
 #define nv_mkio(rv,r,i) /*
-*/	nv_mkmm(rv, (((r) & 0xffc) << 6) | ((i) << 2))
+*/	imm32(rv, (((r) & 0xffc) << 6) | ((i) << 2))
+
+#define hash #
+#define fn(a) a
+#if CHIPSET < GK208
+#define call(a) call fn(hash)a
+#else
+#define call(a) lcall fn(hash)a
+#endif
 
 #define nv_iord(rv,r,i) /*
 */	nv_mkio(rv,r,i) /*
 */	iord rv I[rv]
+
 #define nv_iowr(r,i,rv) /*
 */	nv_mkio($r0,r,i) /*
 */	iowr I[$r0] rv /*
 */	clear b32 $r0
 
+#define nv_rd32(reg,addr) /*
+*/	imm32($r14, addr) /*
+*/	call(nv_rd32) /*
+*/	mov b32 reg $r15
+
+#define nv_wr32(addr,reg) /*
+*/	mov b32 $r15 reg /*
+*/	imm32($r14, addr) /*
+*/	call(nv_wr32)
+
 #define trace_set(bit) /*
 */	clear b32 $r9 /*
 */	bset $r9 bit /*
 */	nv_iowr(NV_PGRAPH_FECS_CC_SCRATCH_SET(7), 0, $r9)
+
 #define trace_clr(bit) /*
 */	clear b32 $r9 /*
 */	bset $r9 bit /*
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/nv108.c b/drivers/gpu/drm/nouveau/core/engine/graph/nv108.c
new file mode 100644
index 000000000000..e1af65ead379
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/nv108.c
@@ -0,0 +1,236 @@
+/*
+ * 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 <bskeggs@redhat.com>
+ */
+
+#include "nvc0.h"
+
+/*******************************************************************************
+ * Graphics object classes
+ ******************************************************************************/
+
+static struct nouveau_oclass
+nv108_graph_sclass[] = {
+	{ 0x902d, &nouveau_object_ofuncs },
+	{ 0xa140, &nouveau_object_ofuncs },
+	{ 0xa197, &nouveau_object_ofuncs },
+	{ 0xa1c0, &nouveau_object_ofuncs },
+	{}
+};
+
+/*******************************************************************************
+ * PGRAPH engine/subdev functions
+ ******************************************************************************/
+
+static struct nvc0_graph_init
+nv108_graph_init_regs[] = {
+	{ 0x400080,   1, 0x04, 0x003083c2 },
+	{ 0x400088,   1, 0x04, 0x0001bfe7 },
+	{ 0x40008c,   1, 0x04, 0x00000000 },
+	{ 0x400090,   1, 0x04, 0x00000030 },
+	{ 0x40013c,   1, 0x04, 0x003901f7 },
+	{ 0x400140,   1, 0x04, 0x00000100 },
+	{ 0x400144,   1, 0x04, 0x00000000 },
+	{ 0x400148,   1, 0x04, 0x00000110 },
+	{ 0x400138,   1, 0x04, 0x00000000 },
+	{ 0x400130,   2, 0x04, 0x00000000 },
+	{ 0x400124,   1, 0x04, 0x00000002 },
+	{}
+};
+
+struct nvc0_graph_init
+nv108_graph_init_unk58xx[] = {
+	{ 0x405844,   1, 0x04, 0x00ffffff },
+	{ 0x405850,   1, 0x04, 0x00000000 },
+	{ 0x405900,   1, 0x04, 0x00000000 },
+	{ 0x405908,   1, 0x04, 0x00000000 },
+	{ 0x405928,   1, 0x04, 0x00000000 },
+	{ 0x40592c,   1, 0x04, 0x00000000 },
+	{}
+};
+
+static struct nvc0_graph_init
+nv108_graph_init_gpc[] = {
+	{ 0x418408,   1, 0x04, 0x00000000 },
+	{ 0x4184a0,   3, 0x04, 0x00000000 },
+	{ 0x418604,   1, 0x04, 0x00000000 },
+	{ 0x418680,   1, 0x04, 0x00000000 },
+	{ 0x418714,   1, 0x04, 0x00000000 },
+	{ 0x418384,   2, 0x04, 0x00000000 },
+	{ 0x418814,   3, 0x04, 0x00000000 },
+	{ 0x418b04,   1, 0x04, 0x00000000 },
+	{ 0x4188c8,   2, 0x04, 0x00000000 },
+	{ 0x4188d0,   1, 0x04, 0x00010000 },
+	{ 0x4188d4,   1, 0x04, 0x00000201 },
+	{ 0x418910,   1, 0x04, 0x00010001 },
+	{ 0x418914,   1, 0x04, 0x00000301 },
+	{ 0x418918,   1, 0x04, 0x00800000 },
+	{ 0x418980,   1, 0x04, 0x77777770 },
+	{ 0x418984,   3, 0x04, 0x77777777 },
+	{ 0x418c04,   1, 0x04, 0x00000000 },
+	{ 0x418c64,   2, 0x04, 0x00000000 },
+	{ 0x418c88,   1, 0x04, 0x00000000 },
+	{ 0x418cb4,   2, 0x04, 0x00000000 },
+	{ 0x418d00,   1, 0x04, 0x00000000 },
+	{ 0x418d28,   2, 0x04, 0x00000000 },
+	{ 0x418f00,   1, 0x04, 0x00000400 },
+	{ 0x418f08,   1, 0x04, 0x00000000 },
+	{ 0x418f20,   2, 0x04, 0x00000000 },
+	{ 0x418e00,   1, 0x04, 0x00000000 },
+	{ 0x418e08,   1, 0x04, 0x00000000 },
+	{ 0x418e1c,   2, 0x04, 0x00000000 },
+	{ 0x41900c,   1, 0x04, 0x00000000 },
+	{ 0x419018,   1, 0x04, 0x00000000 },
+	{}
+};
+
+static struct nvc0_graph_init
+nv108_graph_init_tpc[] = {
+	{ 0x419d0c,   1, 0x04, 0x00000000 },
+	{ 0x419d10,   1, 0x04, 0x00000014 },
+	{ 0x419ab0,   1, 0x04, 0x00000000 },
+	{ 0x419ac8,   1, 0x04, 0x00000000 },
+	{ 0x419ab8,   1, 0x04, 0x000000e7 },
+	{ 0x419abc,   2, 0x04, 0x00000000 },
+	{ 0x419ab4,   1, 0x04, 0x00000000 },
+	{ 0x419aa8,   2, 0x04, 0x00000000 },
+	{ 0x41980c,   1, 0x04, 0x00000010 },
+	{ 0x419844,   1, 0x04, 0x00000000 },
+	{ 0x419850,   1, 0x04, 0x00000004 },
+	{ 0x419854,   2, 0x04, 0x00000000 },
+	{ 0x419c98,   1, 0x04, 0x00000000 },
+	{ 0x419ca8,   1, 0x04, 0x00000000 },
+	{ 0x419cb0,   1, 0x04, 0x01000000 },
+	{ 0x419cb4,   1, 0x04, 0x00000000 },
+	{ 0x419cb8,   1, 0x04, 0x00b08bea },
+	{ 0x419c84,   1, 0x04, 0x00010384 },
+	{ 0x419cbc,   1, 0x04, 0x281b3646 },
+	{ 0x419cc0,   2, 0x04, 0x00000000 },
+	{ 0x419c80,   1, 0x04, 0x00000230 },
+	{ 0x419ccc,   2, 0x04, 0x00000000 },
+	{ 0x419c0c,   1, 0x04, 0x00000000 },
+	{ 0x419e00,   1, 0x04, 0x00000080 },
+	{ 0x419ea0,   1, 0x04, 0x00000000 },
+	{ 0x419ee4,   1, 0x04, 0x00000000 },
+	{ 0x419ea4,   1, 0x04, 0x00000100 },
+	{ 0x419ea8,   1, 0x04, 0x00000000 },
+	{ 0x419eb4,   1, 0x04, 0x00000000 },
+	{ 0x419ebc,   2, 0x04, 0x00000000 },
+	{ 0x419edc,   1, 0x04, 0x00000000 },
+	{ 0x419f00,   1, 0x04, 0x00000000 },
+	{ 0x419ed0,   1, 0x04, 0x00003234 },
+	{ 0x419f74,   1, 0x04, 0x00015555 },
+	{ 0x419f80,   4, 0x04, 0x00000000 },
+	{}
+};
+
+static int
+nv108_graph_fini(struct nouveau_object *object, bool suspend)
+{
+	struct nvc0_graph_priv *priv = (void *)object;
+	static const struct {
+		u32 addr;
+		u32 data;
+	} magic[] = {
+		{ 0x020520, 0xfffffffc },
+		{ 0x020524, 0xfffffffe },
+		{ 0x020524, 0xfffffffc },
+		{ 0x020524, 0xfffffff8 },
+		{ 0x020524, 0xffffffe0 },
+		{ 0x020530, 0xfffffffe },
+		{ 0x02052c, 0xfffffffa },
+		{ 0x02052c, 0xfffffff0 },
+		{ 0x02052c, 0xffffffc0 },
+		{ 0x02052c, 0xffffff00 },
+		{ 0x02052c, 0xfffffc00 },
+		{ 0x02052c, 0xfffcfc00 },
+		{ 0x02052c, 0xfff0fc00 },
+		{ 0x02052c, 0xff80fc00 },
+		{ 0x020528, 0xfffffffe },
+		{ 0x020528, 0xfffffffc },
+	};
+	int i;
+
+	nv_mask(priv, 0x000200, 0x08001000, 0x00000000);
+	nv_mask(priv, 0x0206b4, 0x00000000, 0x00000000);
+	for (i = 0; i < ARRAY_SIZE(magic); i++) {
+		nv_wr32(priv, magic[i].addr, magic[i].data);
+		nv_wait(priv, magic[i].addr, 0x80000000, 0x00000000);
+	}
+
+	return nouveau_graph_fini(&priv->base, suspend);
+}
+
+static struct nvc0_graph_init *
+nv108_graph_init_mmio[] = {
+	nv108_graph_init_regs,
+	nvf0_graph_init_unk40xx,
+	nvc0_graph_init_unk44xx,
+	nvc0_graph_init_unk78xx,
+	nvc0_graph_init_unk60xx,
+	nvd9_graph_init_unk64xx,
+	nv108_graph_init_unk58xx,
+	nvc0_graph_init_unk80xx,
+	nvf0_graph_init_unk70xx,
+	nvf0_graph_init_unk5bxx,
+	nv108_graph_init_gpc,
+	nv108_graph_init_tpc,
+	nve4_graph_init_unk,
+	nve4_graph_init_unk88xx,
+	NULL
+};
+
+#include "fuc/hubnv108.fuc5.h"
+
+static struct nvc0_graph_ucode
+nv108_graph_fecs_ucode = {
+	.code.data = nv108_grhub_code,
+	.code.size = sizeof(nv108_grhub_code),
+	.data.data = nv108_grhub_data,
+	.data.size = sizeof(nv108_grhub_data),
+};
+
+#include "fuc/gpcnv108.fuc5.h"
+
+static struct nvc0_graph_ucode
+nv108_graph_gpccs_ucode = {
+	.code.data = nv108_grgpc_code,
+	.code.size = sizeof(nv108_grgpc_code),
+	.data.data = nv108_grgpc_data,
+	.data.size = sizeof(nv108_grgpc_data),
+};
+
+struct nouveau_oclass *
+nv108_graph_oclass = &(struct nvc0_graph_oclass) {
+	.base.handle = NV_ENGINE(GR, 0x08),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nvc0_graph_ctor,
+		.dtor = nvc0_graph_dtor,
+		.init = nve4_graph_init,
+		.fini = nv108_graph_fini,
+	},
+	.cclass = &nv108_grctx_oclass,
+	.sclass =  nv108_graph_sclass,
+	.mmio = nv108_graph_init_mmio,
+	.fecs.ucode = &nv108_graph_fecs_ucode,
+	.gpccs.ucode = &nv108_graph_gpccs_ucode,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/nv50.c b/drivers/gpu/drm/nouveau/core/engine/graph/nv50.c
index 03de5175dd9f..30ed19c52e05 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/nv50.c
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/nv50.c
@@ -304,12 +304,28 @@ nv84_graph_tlb_flush(struct nouveau_engine *engine)
 	return timeout ? -EBUSY : 0;
 }
 
-static const struct nouveau_enum nv50_mp_exec_error_names[] = {
-	{ 3, "STACK_UNDERFLOW", NULL },
-	{ 4, "QUADON_ACTIVE", NULL },
-	{ 8, "TIMEOUT", NULL },
-	{ 0x10, "INVALID_OPCODE", NULL },
-	{ 0x40, "BREAKPOINT", NULL },
+static const struct nouveau_bitfield nv50_mp_exec_errors[] = {
+	{ 0x01, "STACK_UNDERFLOW" },
+	{ 0x02, "STACK_MISMATCH" },
+	{ 0x04, "QUADON_ACTIVE" },
+	{ 0x08, "TIMEOUT" },
+	{ 0x10, "INVALID_OPCODE" },
+	{ 0x20, "PM_OVERFLOW" },
+	{ 0x40, "BREAKPOINT" },
+	{}
+};
+
+static const struct nouveau_bitfield nv50_mpc_traps[] = {
+	{ 0x0000001, "LOCAL_LIMIT_READ" },
+	{ 0x0000010, "LOCAL_LIMIT_WRITE" },
+	{ 0x0000040, "STACK_LIMIT" },
+	{ 0x0000100, "GLOBAL_LIMIT_READ" },
+	{ 0x0001000, "GLOBAL_LIMIT_WRITE" },
+	{ 0x0010000, "MP0" },
+	{ 0x0020000, "MP1" },
+	{ 0x0040000, "GLOBAL_LIMIT_RED" },
+	{ 0x0400000, "GLOBAL_LIMIT_ATOM" },
+	{ 0x4000000, "MP2" },
 	{}
 };
 
@@ -396,6 +412,60 @@ static const struct nouveau_bitfield nv50_graph_intr_name[] = {
 	{}
 };
 
+static const struct nouveau_bitfield nv50_graph_trap_prop[] = {
+	{ 0x00000004, "SURF_WIDTH_OVERRUN" },
+	{ 0x00000008, "SURF_HEIGHT_OVERRUN" },
+	{ 0x00000010, "DST2D_FAULT" },
+	{ 0x00000020, "ZETA_FAULT" },
+	{ 0x00000040, "RT_FAULT" },
+	{ 0x00000080, "CUDA_FAULT" },
+	{ 0x00000100, "DST2D_STORAGE_TYPE_MISMATCH" },
+	{ 0x00000200, "ZETA_STORAGE_TYPE_MISMATCH" },
+	{ 0x00000400, "RT_STORAGE_TYPE_MISMATCH" },
+	{ 0x00000800, "DST2D_LINEAR_MISMATCH" },
+	{ 0x00001000, "RT_LINEAR_MISMATCH" },
+	{}
+};
+
+static void
+nv50_priv_prop_trap(struct nv50_graph_priv *priv,
+		    u32 ustatus_addr, u32 ustatus, u32 tp)
+{
+	u32 e0c = nv_rd32(priv, ustatus_addr + 0x04);
+	u32 e10 = nv_rd32(priv, ustatus_addr + 0x08);
+	u32 e14 = nv_rd32(priv, ustatus_addr + 0x0c);
+	u32 e18 = nv_rd32(priv, ustatus_addr + 0x10);
+	u32 e1c = nv_rd32(priv, ustatus_addr + 0x14);
+	u32 e20 = nv_rd32(priv, ustatus_addr + 0x18);
+	u32 e24 = nv_rd32(priv, ustatus_addr + 0x1c);
+
+	/* CUDA memory: l[], g[] or stack. */
+	if (ustatus & 0x00000080) {
+		if (e18 & 0x80000000) {
+			/* g[] read fault? */
+			nv_error(priv, "TRAP_PROP - TP %d - CUDA_FAULT - Global read fault at address %02x%08x\n",
+					 tp, e14, e10 | ((e18 >> 24) & 0x1f));
+			e18 &= ~0x1f000000;
+		} else if (e18 & 0xc) {
+			/* g[] write fault? */
+			nv_error(priv, "TRAP_PROP - TP %d - CUDA_FAULT - Global write fault at address %02x%08x\n",
+				 tp, e14, e10 | ((e18 >> 7) & 0x1f));
+			e18 &= ~0x00000f80;
+		} else {
+			nv_error(priv, "TRAP_PROP - TP %d - Unknown CUDA fault at address %02x%08x\n",
+				 tp, e14, e10);
+		}
+		ustatus &= ~0x00000080;
+	}
+	if (ustatus) {
+		nv_error(priv, "TRAP_PROP - TP %d -", tp);
+		nouveau_bitfield_print(nv50_graph_trap_prop, ustatus);
+		pr_cont(" - Address %02x%08x\n", e14, e10);
+	}
+	nv_error(priv, "TRAP_PROP - TP %d - e0c: %08x, e18: %08x, e1c: %08x, e20: %08x, e24: %08x\n",
+		 tp, e0c, e18, e1c, e20, e24);
+}
+
 static void
 nv50_priv_mp_trap(struct nv50_graph_priv *priv, int tpid, int display)
 {
@@ -420,8 +490,8 @@ nv50_priv_mp_trap(struct nv50_graph_priv *priv, int tpid, int display)
 			oplow = nv_rd32(priv, addr + 0x70);
 			ophigh = nv_rd32(priv, addr + 0x74);
 			nv_error(priv, "TRAP_MP_EXEC - "
-					"TP %d MP %d: ", tpid, i);
-			nouveau_enum_print(nv50_mp_exec_error_names, status);
+					"TP %d MP %d:", tpid, i);
+			nouveau_bitfield_print(nv50_mp_exec_errors, status);
 			pr_cont(" at %06x warp %d, opcode %08x %08x\n",
 					pc&0xffffff, pc >> 24,
 					oplow, ophigh);
@@ -468,60 +538,19 @@ nv50_priv_tp_trap(struct nv50_graph_priv *priv, int type, u32 ustatus_old,
 				nv50_priv_mp_trap(priv, i, display);
 				ustatus &= ~0x04030000;
 			}
-			break;
-		case 8: /* TPDMA error */
-			{
-			u32 e0c = nv_rd32(priv, ustatus_addr + 4);
-			u32 e10 = nv_rd32(priv, ustatus_addr + 8);
-			u32 e14 = nv_rd32(priv, ustatus_addr + 0xc);
-			u32 e18 = nv_rd32(priv, ustatus_addr + 0x10);
-			u32 e1c = nv_rd32(priv, ustatus_addr + 0x14);
-			u32 e20 = nv_rd32(priv, ustatus_addr + 0x18);
-			u32 e24 = nv_rd32(priv, ustatus_addr + 0x1c);
-			/* 2d engine destination */
-			if (ustatus & 0x00000010) {
-				if (display) {
-					nv_error(priv, "TRAP_TPDMA_2D - TP %d - Unknown fault at address %02x%08x\n",
-							i, e14, e10);
-					nv_error(priv, "TRAP_TPDMA_2D - TP %d - e0c: %08x, e18: %08x, e1c: %08x, e20: %08x, e24: %08x\n",
-							i, e0c, e18, e1c, e20, e24);
-				}
-				ustatus &= ~0x00000010;
-			}
-			/* Render target */
-			if (ustatus & 0x00000040) {
-				if (display) {
-					nv_error(priv, "TRAP_TPDMA_RT - TP %d - Unknown fault at address %02x%08x\n",
-							i, e14, e10);
-					nv_error(priv, "TRAP_TPDMA_RT - TP %d - e0c: %08x, e18: %08x, e1c: %08x, e20: %08x, e24: %08x\n",
-							i, e0c, e18, e1c, e20, e24);
-				}
-				ustatus &= ~0x00000040;
-			}
-			/* CUDA memory: l[], g[] or stack. */
-			if (ustatus & 0x00000080) {
-				if (display) {
-					if (e18 & 0x80000000) {
-						/* g[] read fault? */
-						nv_error(priv, "TRAP_TPDMA - TP %d - Global read fault at address %02x%08x\n",
-								i, e14, e10 | ((e18 >> 24) & 0x1f));
-						e18 &= ~0x1f000000;
-					} else if (e18 & 0xc) {
-						/* g[] write fault? */
-						nv_error(priv, "TRAP_TPDMA - TP %d - Global write fault at address %02x%08x\n",
-								i, e14, e10 | ((e18 >> 7) & 0x1f));
-						e18 &= ~0x00000f80;
-					} else {
-						nv_error(priv, "TRAP_TPDMA - TP %d - Unknown CUDA fault at address %02x%08x\n",
-								i, e14, e10);
-					}
-					nv_error(priv, "TRAP_TPDMA - TP %d - e0c: %08x, e18: %08x, e1c: %08x, e20: %08x, e24: %08x\n",
-							i, e0c, e18, e1c, e20, e24);
-				}
-				ustatus &= ~0x00000080;
-			}
+			if (ustatus && display) {
+				nv_error("%s - TP%d:", name, i);
+				nouveau_bitfield_print(nv50_mpc_traps, ustatus);
+				pr_cont("\n");
+				ustatus = 0;
 			}
 			break;
+		case 8: /* PROP error */
+			if (display)
+				nv50_priv_prop_trap(
+						priv, ustatus_addr, ustatus, i);
+			ustatus = 0;
+			break;
 		}
 		if (ustatus) {
 			if (display)
@@ -727,11 +756,11 @@ nv50_graph_trap_handler(struct nv50_graph_priv *priv, u32 display,
 		status &= ~0x080;
 	}
 
-	/* TPDMA:  Handles TP-initiated uncached memory accesses:
+	/* PROP:  Handles TP-initiated uncached memory accesses:
 	 * l[], g[], stack, 2d surfaces, render targets. */
 	if (status & 0x100) {
 		nv50_priv_tp_trap(priv, 8, 0x408e08, 0x408708, display,
-				    "TRAP_TPDMA");
+				    "TRAP_PROP");
 		nv_wr32(priv, 0x400108, 0x100);
 		status &= ~0x100;
 	}
@@ -760,7 +789,7 @@ nv50_graph_intr(struct nouveau_subdev *subdev)
 	u32 mthd = (addr & 0x00001ffc);
 	u32 data = nv_rd32(priv, 0x400708);
 	u32 class = nv_rd32(priv, 0x400814);
-	u32 show = stat;
+	u32 show = stat, show_bitfield = stat;
 	int chid;
 
 	engctx = nouveau_engctx_get(engine, inst);
@@ -778,21 +807,26 @@ nv50_graph_intr(struct nouveau_subdev *subdev)
 		nv_error(priv, "DATA_ERROR ");
 		nouveau_enum_print(nv50_data_error_names, ecode);
 		pr_cont("\n");
+		show_bitfield &= ~0x00100000;
 	}
 
 	if (stat & 0x00200000) {
 		if (!nv50_graph_trap_handler(priv, show, chid, (u64)inst << 12,
 				engctx))
 			show &= ~0x00200000;
+		show_bitfield &= ~0x00200000;
 	}
 
 	nv_wr32(priv, 0x400100, stat);
 	nv_wr32(priv, 0x400500, 0x00010001);
 
 	if (show) {
-		nv_error(priv, "%s", "");
-		nouveau_bitfield_print(nv50_graph_intr_name, show);
-		pr_cont("\n");
+		show &= show_bitfield;
+		if (show) {
+			nv_error(priv, "%s", "");
+			nouveau_bitfield_print(nv50_graph_intr_name, show);
+			pr_cont("\n");
+		}
 		nv_error(priv,
 			 "ch %d [0x%010llx %s] subc %d class 0x%04x mthd 0x%04x data 0x%08x\n",
 			 chid, (u64)inst << 12, nouveau_client_name(engctx),
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/nvc0.c b/drivers/gpu/drm/nouveau/core/engine/graph/nvc0.c
index 5c8a63dc506a..a73ab209ea88 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/nvc0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/nvc0.c
@@ -901,6 +901,9 @@ nvc0_graph_init_ctxctl(struct nvc0_graph_priv *priv)
 		}
 
 		return 0;
+	} else
+	if (!oclass->fecs.ucode) {
+		return -ENOSYS;
 	}
 
 	/* load HUB microcode */
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/nvc0.h b/drivers/gpu/drm/nouveau/core/engine/graph/nvc0.h
index ea17a80ad7fc..b0ab6de270b2 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/nvc0.h
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/nvc0.h
@@ -205,6 +205,11 @@ extern struct nvc0_graph_init nve4_graph_init_regs[];
 extern struct nvc0_graph_init nve4_graph_init_unk[];
 extern struct nvc0_graph_init nve4_graph_init_unk88xx[];
 
+extern struct nvc0_graph_init nvf0_graph_init_unk40xx[];
+extern struct nvc0_graph_init nvf0_graph_init_unk70xx[];
+extern struct nvc0_graph_init nvf0_graph_init_unk5bxx[];
+extern struct nvc0_graph_init nvf0_graph_init_tpc[];
+
 int  nvc0_grctx_generate(struct nvc0_graph_priv *);
 void nvc0_grctx_generate_main(struct nvc0_graph_priv *, struct nvc0_grctx *);
 void nvc0_grctx_generate_mods(struct nvc0_graph_priv *, struct nvc0_grctx *);
@@ -266,6 +271,11 @@ extern struct nvc0_graph_init nve4_grctx_init_unk80xx[];
 extern struct nvc0_graph_init nve4_grctx_init_unk90xx[];
 
 extern struct nouveau_oclass *nvf0_grctx_oclass;
+extern struct nvc0_graph_init nvf0_grctx_init_unk44xx[];
+extern struct nvc0_graph_init nvf0_grctx_init_unk5bxx[];
+extern struct nvc0_graph_init nvf0_grctx_init_unk60xx[];
+
+extern struct nouveau_oclass *nv108_grctx_oclass;
 
 #define mmio_data(s,a,p) do {                                                  \
 	info->buffer[info->buffer_nr] = round_up(info->addr, (a));             \
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/nvf0.c b/drivers/gpu/drm/nouveau/core/engine/graph/nvf0.c
index 2f0ac7832234..b1acb9939d95 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/nvf0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/nvf0.c
@@ -41,7 +41,7 @@ nvf0_graph_sclass[] = {
  * PGRAPH engine/subdev functions
  ******************************************************************************/
 
-static struct nvc0_graph_init
+struct nvc0_graph_init
 nvf0_graph_init_unk40xx[] = {
 	{ 0x40415c,   1, 0x04, 0x00000000 },
 	{ 0x404170,   1, 0x04, 0x00000000 },
@@ -60,7 +60,7 @@ nvf0_graph_init_unk58xx[] = {
 	{}
 };
 
-static struct nvc0_graph_init
+struct nvc0_graph_init
 nvf0_graph_init_unk70xx[] = {
 	{ 0x407010,   1, 0x04, 0x00000000 },
 	{ 0x407040,   1, 0x04, 0x80440424 },
@@ -68,7 +68,7 @@ nvf0_graph_init_unk70xx[] = {
 	{}
 };
 
-static struct nvc0_graph_init
+struct nvc0_graph_init
 nvf0_graph_init_unk5bxx[] = {
 	{ 0x405b44,   1, 0x04, 0x00000000 },
 	{ 0x405b50,   1, 0x04, 0x00000000 },
@@ -114,7 +114,7 @@ nvf0_graph_init_gpc[] = {
 	{}
 };
 
-static struct nvc0_graph_init
+struct nvc0_graph_init
 nvf0_graph_init_tpc[] = {
 	{ 0x419d0c,   1, 0x04, 0x00000000 },
 	{ 0x419d10,   1, 0x04, 0x00000014 },
@@ -243,6 +243,6 @@ nvf0_graph_oclass = &(struct nvc0_graph_oclass) {
 	.cclass = &nvf0_grctx_oclass,
 	.sclass =  nvf0_graph_sclass,
 	.mmio = nvf0_graph_init_mmio,
-	.fecs.ucode = 0 ? &nvf0_graph_fecs_ucode : NULL,
+	.fecs.ucode = &nvf0_graph_fecs_ucode,
 	.gpccs.ucode = &nvf0_graph_gpccs_ucode,
 }.base;
diff --git a/drivers/gpu/drm/nouveau/core/include/core/class.h b/drivers/gpu/drm/nouveau/core/include/core/class.h
index 560c3593dae7..e71a4325e670 100644
--- a/drivers/gpu/drm/nouveau/core/include/core/class.h
+++ b/drivers/gpu/drm/nouveau/core/include/core/class.h
@@ -230,9 +230,26 @@ struct nve0_channel_ind_class {
 
 #define NV04_DISP_CLASS                                              0x00000046
 
+#define NV04_DISP_MTHD                                               0x00000000
+#define NV04_DISP_MTHD_HEAD                                          0x00000001
+
+#define NV04_DISP_SCANOUTPOS                                         0x00000000
+
 struct nv04_display_class {
 };
 
+struct nv04_display_scanoutpos {
+	s64 time[2];
+	u32 vblanks;
+	u32 vblanke;
+	u32 vtotal;
+	u32 vline;
+	u32 hblanks;
+	u32 hblanke;
+	u32 htotal;
+	u32 hline;
+};
+
 /* 5070: NV50_DISP
  * 8270: NV84_DISP
  * 8370: NVA0_DISP
@@ -252,6 +269,11 @@ struct nv04_display_class {
 #define NVE0_DISP_CLASS                                              0x00009170
 #define NVF0_DISP_CLASS                                              0x00009270
 
+#define NV50_DISP_MTHD                                               0x00000000
+#define NV50_DISP_MTHD_HEAD                                          0x00000003
+
+#define NV50_DISP_SCANOUTPOS                                         0x00000000
+
 #define NV50_DISP_SOR_MTHD                                           0x00010000
 #define NV50_DISP_SOR_MTHD_TYPE                                      0x0000f000
 #define NV50_DISP_SOR_MTHD_HEAD                                      0x00000018
diff --git a/drivers/gpu/drm/nouveau/core/include/core/device.h b/drivers/gpu/drm/nouveau/core/include/core/device.h
index ac2881d1776a..7b8ea221b00d 100644
--- a/drivers/gpu/drm/nouveau/core/include/core/device.h
+++ b/drivers/gpu/drm/nouveau/core/include/core/device.h
@@ -38,7 +38,8 @@ enum nv_subdev_type {
 	NVDEV_SUBDEV_THERM,
 	NVDEV_SUBDEV_CLOCK,
 
-	NVDEV_ENGINE_DMAOBJ,
+	NVDEV_ENGINE_FIRST,
+	NVDEV_ENGINE_DMAOBJ = NVDEV_ENGINE_FIRST,
 	NVDEV_ENGINE_FIFO,
 	NVDEV_ENGINE_SW,
 	NVDEV_ENGINE_GR,
@@ -70,6 +71,7 @@ struct nouveau_device {
 	const char *dbgopt;
 	const char *name;
 	const char *cname;
+	u64 disable_mask;
 
 	enum {
 		NV_04    = 0x04,
diff --git a/drivers/gpu/drm/nouveau/core/include/engine/fifo.h b/drivers/gpu/drm/nouveau/core/include/engine/fifo.h
index 8c32cf4d83c7..26b6b2bb1112 100644
--- a/drivers/gpu/drm/nouveau/core/include/engine/fifo.h
+++ b/drivers/gpu/drm/nouveau/core/include/engine/fifo.h
@@ -109,6 +109,7 @@ extern struct nouveau_oclass *nv50_fifo_oclass;
 extern struct nouveau_oclass *nv84_fifo_oclass;
 extern struct nouveau_oclass *nvc0_fifo_oclass;
 extern struct nouveau_oclass *nve0_fifo_oclass;
+extern struct nouveau_oclass *nv108_fifo_oclass;
 
 void nv04_fifo_intr(struct nouveau_subdev *);
 int  nv04_fifo_context_attach(struct nouveau_object *, struct nouveau_object *);
diff --git a/drivers/gpu/drm/nouveau/core/include/engine/graph.h b/drivers/gpu/drm/nouveau/core/include/engine/graph.h
index 8e1b52312ddc..97705618de97 100644
--- a/drivers/gpu/drm/nouveau/core/include/engine/graph.h
+++ b/drivers/gpu/drm/nouveau/core/include/engine/graph.h
@@ -69,6 +69,7 @@ extern struct nouveau_oclass *nvd7_graph_oclass;
 extern struct nouveau_oclass *nvd9_graph_oclass;
 extern struct nouveau_oclass *nve4_graph_oclass;
 extern struct nouveau_oclass *nvf0_graph_oclass;
+extern struct nouveau_oclass *nv108_graph_oclass;
 
 extern const struct nouveau_bitfield nv04_graph_nsource[];
 extern struct nouveau_ofuncs nv04_graph_ofuncs;
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/bar.h b/drivers/gpu/drm/nouveau/core/include/subdev/bar.h
index 4f4ff4502c3d..9faa98e67ad8 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/bar.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bar.h
@@ -4,8 +4,7 @@
 #include <core/subdev.h>
 #include <core/device.h>
 
-#include <subdev/fb.h>
-
+struct nouveau_mem;
 struct nouveau_vma;
 
 struct nouveau_bar {
@@ -29,27 +28,7 @@ nouveau_bar(void *obj)
 	return (void *)nv_device(obj)->subdev[NVDEV_SUBDEV_BAR];
 }
 
-#define nouveau_bar_create(p,e,o,d)                                            \
-	nouveau_bar_create_((p), (e), (o), sizeof(**d), (void **)d)
-#define nouveau_bar_init(p)                                                    \
-	nouveau_subdev_init(&(p)->base)
-#define nouveau_bar_fini(p,s)                                                  \
-	nouveau_subdev_fini(&(p)->base, (s))
-
-int nouveau_bar_create_(struct nouveau_object *, struct nouveau_object *,
-			struct nouveau_oclass *, int, void **);
-void nouveau_bar_destroy(struct nouveau_bar *);
-
-void _nouveau_bar_dtor(struct nouveau_object *);
-#define _nouveau_bar_init _nouveau_subdev_init
-#define _nouveau_bar_fini _nouveau_subdev_fini
-
 extern struct nouveau_oclass nv50_bar_oclass;
 extern struct nouveau_oclass nvc0_bar_oclass;
 
-int nouveau_bar_alloc(struct nouveau_bar *, struct nouveau_object *,
-		      struct nouveau_mem *, struct nouveau_object **);
-
-void nv84_bar_flush(struct nouveau_bar *);
-
 #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
new file mode 100644
index 000000000000..c5e6d1e6ac1d
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bios/ramcfg.h
@@ -0,0 +1,66 @@
+#ifndef __NVBIOS_RAMCFG_H__
+#define __NVBIOS_RAMCFG_H__
+
+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 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[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;
+};
+
+u8 nvbios_ramcfg_count(struct nouveau_bios *);
+u8 nvbios_ramcfg_index(struct nouveau_bios *);
+
+#endif
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 bc15e0320877..5bdf8e4db40a 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/bios/rammap.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bios/rammap.h
@@ -1,11 +1,25 @@
 #ifndef __NVBIOS_RAMMAP_H__
 #define __NVBIOS_RAMMAP_H__
 
-u16 nvbios_rammap_table(struct nouveau_bios *, u8 *ver, u8 *hdr,
-			u8 *cnt, u8 *len, u8 *snr, u8 *ssz);
-u16 nvbios_rammap_entry(struct nouveau_bios *, int idx,
-			u8 *ver, u8 *hdr, u8 *cnt, u8 *len);
-u16 nvbios_rammap_match(struct nouveau_bios *, u16 khz,
-			u8 *ver, u8 *hdr, u8 *cnt, u8 *len);
+struct nvbios_ramcfg;
+
+u32 nvbios_rammapTe(struct nouveau_bios *, u8 *ver, u8 *hdr,
+		    u8 *cnt, u8 *len, u8 *snr, u8 *ssz);
+
+u32 nvbios_rammapEe(struct nouveau_bios *, int idx,
+		    u8 *ver, u8 *hdr, u8 *cnt, u8 *len);
+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 *);
+
+u32 nvbios_rammapSe(struct nouveau_bios *, u32 data,
+		    u8 ever, u8 ehdr, u8 ecnt, u8 elen, int idx,
+		    u8 *ver, u8 *hdr);
+u32 nvbios_rammapSp(struct nouveau_bios *, u32 data,
+		    u8 ever, u8 ehdr, u8 ecnt, u8 elen, int idx,
+		    u8 *ver, u8 *hdr,
+		    struct nvbios_ramcfg *);
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/bios/timing.h b/drivers/gpu/drm/nouveau/core/include/subdev/bios/timing.h
index 963694b54224..76d914b67ab5 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/bios/timing.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/bios/timing.h
@@ -1,8 +1,14 @@
 #ifndef __NVBIOS_TIMING_H__
 #define __NVBIOS_TIMING_H__
 
-u16 nvbios_timing_table(struct nouveau_bios *,
-			u8 *ver, u8 *hdr, u8 *cnt, u8 *len);
-u16 nvbios_timing_entry(struct nouveau_bios *, int idx, u8 *ver, u8 *hdr);
+struct nvbios_ramcfg;
+
+u16 nvbios_timingTe(struct nouveau_bios *,
+		    u8 *ver, u8 *hdr, u8 *cnt, u8 *len, u8 *snr, u8 *ssz);
+u16 nvbios_timingEe(struct nouveau_bios *, int idx,
+		    u8 *ver, u8 *hdr, u8 *cnt, u8 *len);
+u16 nvbios_timingEp(struct nouveau_bios *, int idx,
+		    u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
+		    struct nvbios_ramcfg *);
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/devinit.h b/drivers/gpu/drm/nouveau/core/include/subdev/devinit.h
index 685c9b12ee4c..ed1ac68c38b3 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/devinit.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/devinit.h
@@ -9,7 +9,6 @@ struct nouveau_devinit {
 	bool post;
 	void (*meminit)(struct nouveau_devinit *);
 	int  (*pll_set)(struct nouveau_devinit *, u32 type, u32 freq);
-
 };
 
 static inline struct nouveau_devinit *
@@ -18,32 +17,16 @@ nouveau_devinit(void *obj)
 	return (void *)nv_device(obj)->subdev[NVDEV_SUBDEV_DEVINIT];
 }
 
-#define nouveau_devinit_create(p,e,o,d)                                        \
-	nouveau_devinit_create_((p), (e), (o), sizeof(**d), (void **)d)
-#define nouveau_devinit_destroy(p)                                             \
-	nouveau_subdev_destroy(&(p)->base)
-#define nouveau_devinit_init(p) ({                                             \
-	struct nouveau_devinit *d = (p);                                       \
-	_nouveau_devinit_init(nv_object(d));                                   \
-})
-#define nouveau_devinit_fini(p,s) ({                                           \
-	struct nouveau_devinit *d = (p);                                       \
-	_nouveau_devinit_fini(nv_object(d), (s));                              \
-})
-
-int nouveau_devinit_create_(struct nouveau_object *, struct nouveau_object *,
-			    struct nouveau_oclass *, int, void **);
-#define _nouveau_devinit_dtor _nouveau_subdev_dtor
-int _nouveau_devinit_init(struct nouveau_object *);
-int _nouveau_devinit_fini(struct nouveau_object *, bool suspend);
-
-extern struct nouveau_oclass nv04_devinit_oclass;
-extern struct nouveau_oclass nv05_devinit_oclass;
-extern struct nouveau_oclass nv10_devinit_oclass;
-extern struct nouveau_oclass nv1a_devinit_oclass;
-extern struct nouveau_oclass nv20_devinit_oclass;
-extern struct nouveau_oclass nv50_devinit_oclass;
-extern struct nouveau_oclass nva3_devinit_oclass;
-extern struct nouveau_oclass nvc0_devinit_oclass;
+extern struct nouveau_oclass *nv04_devinit_oclass;
+extern struct nouveau_oclass *nv05_devinit_oclass;
+extern struct nouveau_oclass *nv10_devinit_oclass;
+extern struct nouveau_oclass *nv1a_devinit_oclass;
+extern struct nouveau_oclass *nv20_devinit_oclass;
+extern struct nouveau_oclass *nv50_devinit_oclass;
+extern struct nouveau_oclass *nv84_devinit_oclass;
+extern struct nouveau_oclass *nv98_devinit_oclass;
+extern struct nouveau_oclass *nva3_devinit_oclass;
+extern struct nouveau_oclass *nvaf_devinit_oclass;
+extern struct nouveau_oclass *nvc0_devinit_oclass;
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/fb.h b/drivers/gpu/drm/nouveau/core/include/subdev/fb.h
index d89dbdf39b0d..d7ecafbae1ca 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/fb.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/fb.h
@@ -106,6 +106,13 @@ extern struct nouveau_oclass *nvaf_fb_oclass;
 extern struct nouveau_oclass *nvc0_fb_oclass;
 extern struct nouveau_oclass *nve0_fb_oclass;
 
+#include <subdev/bios/ramcfg.h>
+
+struct nouveau_ram_data {
+	struct nvbios_ramcfg bios;
+	u32 freq;
+};
+
 struct nouveau_ram {
 	struct nouveau_object base;
 	enum {
@@ -142,6 +149,12 @@ struct nouveau_ram {
 	} rammap, ramcfg, timing;
 	u32 freq;
 	u32 mr[16];
+	u32 mr1_nuts;
+
+	struct nouveau_ram_data *next;
+	struct nouveau_ram_data former;
+	struct nouveau_ram_data xition;
+	struct nouveau_ram_data target;
 };
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/instmem.h b/drivers/gpu/drm/nouveau/core/include/subdev/instmem.h
index 4aca33887aaa..c1df26f3230c 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/instmem.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/instmem.h
@@ -23,21 +23,6 @@ nv_memobj(void *obj)
 	return obj;
 }
 
-#define nouveau_instobj_create(p,e,o,d)                                        \
-	nouveau_instobj_create_((p), (e), (o), sizeof(**d), (void **)d)
-#define nouveau_instobj_init(p)                                                \
-	nouveau_object_init(&(p)->base)
-#define nouveau_instobj_fini(p,s)                                              \
-	nouveau_object_fini(&(p)->base, (s))
-
-int  nouveau_instobj_create_(struct nouveau_object *, struct nouveau_object *,
-			     struct nouveau_oclass *, int, void **);
-void nouveau_instobj_destroy(struct nouveau_instobj *);
-
-void _nouveau_instobj_dtor(struct nouveau_object *);
-#define _nouveau_instobj_init nouveau_object_init
-#define _nouveau_instobj_fini nouveau_object_fini
-
 struct nouveau_instmem {
 	struct nouveau_subdev base;
 	struct list_head list;
@@ -60,21 +45,8 @@ nouveau_instmem(void *obj)
 	return (void *)nv_device(obj)->subdev[NVDEV_SUBDEV_INSTMEM];
 }
 
-#define nouveau_instmem_create(p,e,o,d)                                        \
-	nouveau_instmem_create_((p), (e), (o), sizeof(**d), (void **)d)
-#define nouveau_instmem_destroy(p)                                             \
-	nouveau_subdev_destroy(&(p)->base)
-int nouveau_instmem_create_(struct nouveau_object *, struct nouveau_object *,
-			    struct nouveau_oclass *, int, void **);
-int nouveau_instmem_init(struct nouveau_instmem *);
-int nouveau_instmem_fini(struct nouveau_instmem *, bool);
-
-#define _nouveau_instmem_dtor _nouveau_subdev_dtor
-int _nouveau_instmem_init(struct nouveau_object *);
-int _nouveau_instmem_fini(struct nouveau_object *, bool);
-
-extern struct nouveau_oclass nv04_instmem_oclass;
-extern struct nouveau_oclass nv40_instmem_oclass;
-extern struct nouveau_oclass nv50_instmem_oclass;
+extern struct nouveau_oclass *nv04_instmem_oclass;
+extern struct nouveau_oclass *nv40_instmem_oclass;
+extern struct nouveau_oclass *nv50_instmem_oclass;
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/include/subdev/vm.h b/drivers/gpu/drm/nouveau/core/include/subdev/vm.h
index fcf57fa309bf..c9509039f94b 100644
--- a/drivers/gpu/drm/nouveau/core/include/subdev/vm.h
+++ b/drivers/gpu/drm/nouveau/core/include/subdev/vm.h
@@ -131,9 +131,5 @@ void nouveau_vm_map(struct nouveau_vma *, struct nouveau_mem *);
 void nouveau_vm_map_at(struct nouveau_vma *, u64 offset, struct nouveau_mem *);
 void nouveau_vm_unmap(struct nouveau_vma *);
 void nouveau_vm_unmap_at(struct nouveau_vma *, u64 offset, u64 length);
-void nouveau_vm_map_sg(struct nouveau_vma *, u64 offset, u64 length,
-		       struct nouveau_mem *);
-void nouveau_vm_map_sg_table(struct nouveau_vma *vma, u64 delta, u64 length,
-		     struct nouveau_mem *mem);
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bar/base.c b/drivers/gpu/drm/nouveau/core/subdev/bar/base.c
index d70ba342aa2e..7098ddd54678 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bar/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bar/base.c
@@ -23,7 +23,11 @@
  */
 
 #include <core/object.h>
-#include <subdev/bar.h>
+
+#include <subdev/fb.h>
+#include <subdev/vm.h>
+
+#include "priv.h"
 
 struct nouveau_barobj {
 	struct nouveau_object base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bar/nv50.c b/drivers/gpu/drm/nouveau/core/subdev/bar/nv50.c
index 160d27f3c7b4..090d594a21b3 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bar/nv50.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bar/nv50.c
@@ -25,10 +25,11 @@
 #include <core/gpuobj.h>
 
 #include <subdev/timer.h>
-#include <subdev/bar.h>
 #include <subdev/fb.h>
 #include <subdev/vm.h>
 
+#include "priv.h"
+
 struct nv50_bar_priv {
 	struct nouveau_bar base;
 	spinlock_t lock;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bar/nvc0.c b/drivers/gpu/drm/nouveau/core/subdev/bar/nvc0.c
index b2ec7411eb2e..bac5e754de35 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bar/nvc0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bar/nvc0.c
@@ -25,10 +25,11 @@
 #include <core/gpuobj.h>
 
 #include <subdev/timer.h>
-#include <subdev/bar.h>
 #include <subdev/fb.h>
 #include <subdev/vm.h>
 
+#include "priv.h"
+
 struct nvc0_bar_priv {
 	struct nouveau_bar base;
 	spinlock_t lock;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bar/priv.h b/drivers/gpu/drm/nouveau/core/subdev/bar/priv.h
new file mode 100644
index 000000000000..ffad8f337ead
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/bar/priv.h
@@ -0,0 +1,26 @@
+#ifndef __NVKM_BAR_PRIV_H__
+#define __NVKM_BAR_PRIV_H__
+
+#include <subdev/bar.h>
+
+#define nouveau_bar_create(p,e,o,d)                                            \
+	nouveau_bar_create_((p), (e), (o), sizeof(**d), (void **)d)
+#define nouveau_bar_init(p)                                                    \
+	nouveau_subdev_init(&(p)->base)
+#define nouveau_bar_fini(p,s)                                                  \
+	nouveau_subdev_fini(&(p)->base, (s))
+
+int nouveau_bar_create_(struct nouveau_object *, struct nouveau_object *,
+			struct nouveau_oclass *, int, void **);
+void nouveau_bar_destroy(struct nouveau_bar *);
+
+void _nouveau_bar_dtor(struct nouveau_object *);
+#define _nouveau_bar_init _nouveau_subdev_init
+#define _nouveau_bar_fini _nouveau_subdev_fini
+
+int  nouveau_bar_alloc(struct nouveau_bar *, struct nouveau_object *,
+		       struct nouveau_mem *, struct nouveau_object **);
+
+void nv84_bar_flush(struct nouveau_bar *);
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/init.c b/drivers/gpu/drm/nouveau/core/subdev/bios/init.c
index df1b1b423093..de201baeb053 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bios/init.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/init.c
@@ -9,6 +9,7 @@
 #include <subdev/bios/dp.h>
 #include <subdev/bios/gpio.h>
 #include <subdev/bios/init.h>
+#include <subdev/bios/ramcfg.h>
 #include <subdev/devinit.h>
 #include <subdev/i2c.h>
 #include <subdev/vga.h>
@@ -391,43 +392,14 @@ init_unknown_script(struct nouveau_bios *bios)
 	return 0x0000;
 }
 
-static u16
-init_ram_restrict_table(struct nvbios_init *init)
-{
-	struct nouveau_bios *bios = init->bios;
-	struct bit_entry bit_M;
-	u16 data = 0x0000;
-
-	if (!bit_entry(bios, 'M', &bit_M)) {
-		if (bit_M.version == 1 && bit_M.length >= 5)
-			data = nv_ro16(bios, bit_M.offset + 3);
-		if (bit_M.version == 2 && bit_M.length >= 3)
-			data = nv_ro16(bios, bit_M.offset + 1);
-	}
-
-	if (data == 0x0000)
-		warn("ram restrict table not found\n");
-	return data;
-}
-
 static u8
 init_ram_restrict_group_count(struct nvbios_init *init)
 {
-	struct nouveau_bios *bios = init->bios;
-	struct bit_entry bit_M;
-
-	if (!bit_entry(bios, 'M', &bit_M)) {
-		if (bit_M.version == 1 && bit_M.length >= 5)
-			return nv_ro08(bios, bit_M.offset + 2);
-		if (bit_M.version == 2 && bit_M.length >= 3)
-			return nv_ro08(bios, bit_M.offset + 0);
-	}
-
-	return 0x00;
+	return nvbios_ramcfg_count(init->bios);
 }
 
 static u8
-init_ram_restrict_strap(struct nvbios_init *init)
+init_ram_restrict(struct nvbios_init *init)
 {
 	/* This appears to be the behaviour of the VBIOS parser, and *is*
 	 * important to cache the NV_PEXTDEV_BOOT0 on later chipsets to
@@ -438,18 +410,8 @@ init_ram_restrict_strap(struct nvbios_init *init)
 	 * in case *not* re-reading the strap causes similar breakage.
 	 */
 	if (!init->ramcfg || init->bios->version.major < 0x70)
-		init->ramcfg = init_rd32(init, 0x101000);
-	return (init->ramcfg & 0x00000003c) >> 2;
-}
-
-static u8
-init_ram_restrict(struct nvbios_init *init)
-{
-	u8  strap = init_ram_restrict_strap(init);
-	u16 table = init_ram_restrict_table(init);
-	if (table)
-		return nv_ro08(init->bios, table + strap);
-	return 0x00;
+		init->ramcfg = 0x80000000 | nvbios_ramcfg_index(init->bios);
+	return (init->ramcfg & 0x7fffffff);
 }
 
 static u8
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/ramcfg.c b/drivers/gpu/drm/nouveau/core/subdev/bios/ramcfg.c
new file mode 100644
index 000000000000..991aedda999b
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/ramcfg.c
@@ -0,0 +1,67 @@
+/*
+ * 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 <bskeggs@redhat.com>
+ */
+
+#include <subdev/bios.h>
+#include <subdev/bios/bit.h>
+#include <subdev/bios/ramcfg.h>
+
+static u8
+nvbios_ramcfg_strap(struct nouveau_bios *bios)
+{
+	return (nv_rd32(bios, 0x101000) & 0x0000003c) >> 2;
+}
+
+u8
+nvbios_ramcfg_count(struct nouveau_bios *bios)
+{
+	struct bit_entry bit_M;
+
+	if (!bit_entry(bios, 'M', &bit_M)) {
+		if (bit_M.version == 1 && bit_M.length >= 5)
+			return nv_ro08(bios, bit_M.offset + 2);
+		if (bit_M.version == 2 && bit_M.length >= 3)
+			return nv_ro08(bios, bit_M.offset + 0);
+	}
+
+	return 0x00;
+}
+
+u8
+nvbios_ramcfg_index(struct nouveau_bios *bios)
+{
+	u8 strap = nvbios_ramcfg_strap(bios);
+	u32 xlat = 0x00000000;
+	struct bit_entry bit_M;
+
+	if (!bit_entry(bios, 'M', &bit_M)) {
+		if (bit_M.version == 1 && bit_M.length >= 5)
+			xlat = nv_ro16(bios, bit_M.offset + 3);
+		if (bit_M.version == 2 && bit_M.length >= 3)
+			xlat = nv_ro16(bios, bit_M.offset + 1);
+	}
+
+	if (xlat)
+		strap = nv_ro08(bios, xlat + strap);
+	return strap;
+}
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/rammap.c b/drivers/gpu/drm/nouveau/core/subdev/bios/rammap.c
index 916fa9d302b7..1811b2cb0472 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bios/rammap.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/rammap.c
@@ -24,11 +24,12 @@
 
 #include <subdev/bios.h>
 #include <subdev/bios/bit.h>
+#include <subdev/bios/ramcfg.h>
 #include <subdev/bios/rammap.h>
 
-u16
-nvbios_rammap_table(struct nouveau_bios *bios, u8 *ver, u8 *hdr,
-		    u8 *cnt, u8 *len, u8 *snr, u8 *ssz)
+u32
+nvbios_rammapTe(struct nouveau_bios *bios, u8 *ver, u8 *hdr,
+		u8 *cnt, u8 *len, u8 *snr, u8 *ssz)
 {
 	struct bit_entry bit_P;
 	u16 rammap = 0x0000;
@@ -57,12 +58,12 @@ nvbios_rammap_table(struct nouveau_bios *bios, u8 *ver, u8 *hdr,
 	return 0x0000;
 }
 
-u16
-nvbios_rammap_entry(struct nouveau_bios *bios, int idx,
-		    u8 *ver, u8 *hdr, u8 *cnt, u8 *len)
+u32
+nvbios_rammapEe(struct nouveau_bios *bios, int idx,
+		u8 *ver, u8 *hdr, u8 *cnt, u8 *len)
 {
 	u8  snr, ssz;
-	u16 rammap = nvbios_rammap_table(bios, ver, hdr, cnt, len, &snr, &ssz);
+	u16 rammap = nvbios_rammapTe(bios, ver, hdr, cnt, len, &snr, &ssz);
 	if (rammap && idx < *cnt) {
 		rammap = rammap + *hdr + (idx * (*len + (snr * ssz)));
 		*hdr = *len;
@@ -73,16 +74,100 @@ nvbios_rammap_entry(struct nouveau_bios *bios, int idx,
 	return 0x0000;
 }
 
-u16
-nvbios_rammap_match(struct nouveau_bios *bios, u16 khz,
-		    u8 *ver, u8 *hdr, u8 *cnt, u8 *len)
+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_rammap_entry(bios, idx++, ver, hdr, cnt, len))) {
+	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,
+		u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
+		struct nvbios_ramcfg *p)
+{
+	u32 data = nvbios_rammapEm(bios, khz, ver, hdr, cnt, len);
+	memset(p, 0x00, sizeof(*p));
+	switch (!!data * *ver) {
+	case 0x11:
+		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;
+		p->rammap_11_11_0c = (nv_ro08(bios, data + 0x11) & 0x0c) >> 2;
+		break;
+	default:
+		data = 0;
+		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)
+{
+	if (idx < ecnt) {
+		data = data + ehdr + (idx * elen);
+		*ver = ever;
+		*hdr = elen;
+		return data;
+	}
+	return 0;
+}
+
+u32
+nvbios_rammapSp(struct nouveau_bios *bios, u32 data,
+		u8 ever, u8 ehdr, u8 ecnt, u8 elen, int idx,
+		u8 *ver, u8 *hdr, struct nvbios_ramcfg *p)
+{
+	data = nvbios_rammapSe(bios, data, ever, ehdr, ecnt, elen, idx, ver, hdr);
+	switch (!!data * *ver) {
+	case 0x11:
+		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;
+		p->ramcfg_11_01_08 = (nv_ro08(bios, data + 0x01) & 0x08) >> 3;
+		p->ramcfg_11_01_10 = (nv_ro08(bios, data + 0x01) & 0x10) >> 4;
+		p->ramcfg_11_01_20 = (nv_ro08(bios, data + 0x01) & 0x20) >> 5;
+		p->ramcfg_11_01_40 = (nv_ro08(bios, data + 0x01) & 0x40) >> 6;
+		p->ramcfg_11_01_80 = (nv_ro08(bios, data + 0x01) & 0x80) >> 7;
+		p->ramcfg_11_02_03 = (nv_ro08(bios, data + 0x02) & 0x03) >> 0;
+		p->ramcfg_11_02_04 = (nv_ro08(bios, data + 0x02) & 0x04) >> 2;
+		p->ramcfg_11_02_08 = (nv_ro08(bios, data + 0x02) & 0x08) >> 3;
+		p->ramcfg_11_02_10 = (nv_ro08(bios, data + 0x02) & 0x10) >> 4;
+		p->ramcfg_11_02_40 = (nv_ro08(bios, data + 0x02) & 0x40) >> 6;
+		p->ramcfg_11_02_80 = (nv_ro08(bios, data + 0x02) & 0x80) >> 7;
+		p->ramcfg_11_03_0f = (nv_ro08(bios, data + 0x03) & 0x0f) >> 0;
+		p->ramcfg_11_03_30 = (nv_ro08(bios, data + 0x03) & 0x30) >> 4;
+		p->ramcfg_11_03_c0 = (nv_ro08(bios, data + 0x03) & 0xc0) >> 6;
+		p->ramcfg_11_03_f0 = (nv_ro08(bios, data + 0x03) & 0xf0) >> 4;
+		p->ramcfg_11_04    = (nv_ro08(bios, data + 0x04) & 0xff) >> 0;
+		p->ramcfg_11_06    = (nv_ro08(bios, data + 0x06) & 0xff) >> 0;
+		p->ramcfg_11_07_02 = (nv_ro08(bios, data + 0x07) & 0x02) >> 1;
+		p->ramcfg_11_07_04 = (nv_ro08(bios, data + 0x07) & 0x04) >> 2;
+		p->ramcfg_11_07_08 = (nv_ro08(bios, data + 0x07) & 0x08) >> 3;
+		p->ramcfg_11_07_10 = (nv_ro08(bios, data + 0x07) & 0x10) >> 4;
+		p->ramcfg_11_07_40 = (nv_ro08(bios, data + 0x07) & 0x40) >> 6;
+		p->ramcfg_11_07_80 = (nv_ro08(bios, data + 0x07) & 0x80) >> 7;
+		p->ramcfg_11_08_01 = (nv_ro08(bios, data + 0x08) & 0x01) >> 0;
+		p->ramcfg_11_08_02 = (nv_ro08(bios, data + 0x08) & 0x02) >> 1;
+		p->ramcfg_11_08_04 = (nv_ro08(bios, data + 0x08) & 0x04) >> 2;
+		p->ramcfg_11_08_08 = (nv_ro08(bios, data + 0x08) & 0x08) >> 3;
+		p->ramcfg_11_08_10 = (nv_ro08(bios, data + 0x08) & 0x10) >> 4;
+		p->ramcfg_11_08_20 = (nv_ro08(bios, data + 0x08) & 0x20) >> 5;
+		p->ramcfg_11_09    = (nv_ro08(bios, data + 0x09) & 0xff) >> 0;
+		break;
+	default:
+		data = 0;
+		break;
+	}
+	return data;
+}
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/timing.c b/drivers/gpu/drm/nouveau/core/subdev/bios/timing.c
index 151c2d6aaee8..350d44ab2ba2 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bios/timing.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/timing.c
@@ -24,11 +24,12 @@
 
 #include <subdev/bios.h>
 #include <subdev/bios/bit.h>
+#include <subdev/bios/ramcfg.h>
 #include <subdev/bios/timing.h>
 
 u16
-nvbios_timing_table(struct nouveau_bios *bios,
-		    u8 *ver, u8 *hdr, u8 *cnt, u8 *len)
+nvbios_timingTe(struct nouveau_bios *bios,
+		u8 *ver, u8 *hdr, u8 *cnt, u8 *len, u8 *snr, u8 *ssz)
 {
 	struct bit_entry bit_P;
 	u16 timing = 0x0000;
@@ -47,11 +48,15 @@ nvbios_timing_table(struct nouveau_bios *bios,
 				*hdr = nv_ro08(bios, timing + 1);
 				*cnt = nv_ro08(bios, timing + 2);
 				*len = nv_ro08(bios, timing + 3);
+				*snr = 0;
+				*ssz = 0;
 				return timing;
 			case 0x20:
 				*hdr = nv_ro08(bios, timing + 1);
-				*cnt = nv_ro08(bios, timing + 3);
+				*cnt = nv_ro08(bios, timing + 5);
 				*len = nv_ro08(bios, timing + 2);
+				*snr = nv_ro08(bios, timing + 4);
+				*ssz = nv_ro08(bios, timing + 3);
 				return timing;
 			default:
 				break;
@@ -63,11 +68,60 @@ nvbios_timing_table(struct nouveau_bios *bios,
 }
 
 u16
-nvbios_timing_entry(struct nouveau_bios *bios, int idx, u8 *ver, u8 *len)
+nvbios_timingEe(struct nouveau_bios *bios, int idx,
+		u8 *ver, u8 *hdr, u8 *cnt, u8 *len)
 {
-	u8  hdr, cnt;
-	u16 timing = nvbios_timing_table(bios, ver, &hdr, &cnt, len);
-	if (timing && idx < cnt)
-		return timing + hdr + (idx * *len);
+	u8  snr, ssz;
+	u16 timing = nvbios_timingTe(bios, ver, hdr, cnt, len, &snr, &ssz);
+	if (timing && idx < *cnt) {
+		timing += *hdr + idx * (*len + (snr * ssz));
+		*hdr = *len;
+		*cnt = snr;
+		*len = ssz;
+		return timing;
+	}
 	return 0x0000;
 }
+
+u16
+nvbios_timingEp(struct nouveau_bios *bios, int idx,
+		u8 *ver, u8 *hdr, u8 *cnt, u8 *len,
+		struct nvbios_ramcfg *p)
+{
+	u16 data = nvbios_timingEe(bios, idx, ver, hdr, cnt, len), temp;
+	switch (!!data * *ver) {
+	case 0x20:
+		p->timing[0] = nv_ro32(bios, data + 0x00);
+		p->timing[1] = nv_ro32(bios, data + 0x04);
+		p->timing[2] = nv_ro32(bios, data + 0x08);
+		p->timing[3] = nv_ro32(bios, data + 0x0c);
+		p->timing[4] = nv_ro32(bios, data + 0x10);
+		p->timing[5] = nv_ro32(bios, data + 0x14);
+		p->timing[6] = nv_ro32(bios, data + 0x18);
+		p->timing[7] = nv_ro32(bios, data + 0x1c);
+		p->timing[8] = nv_ro32(bios, data + 0x20);
+		p->timing[9] = nv_ro32(bios, data + 0x24);
+		p->timing[10] = nv_ro32(bios, data + 0x28);
+		p->timing_20_2e_03 = (nv_ro08(bios, data + 0x2e) & 0x03) >> 0;
+		p->timing_20_2e_30 = (nv_ro08(bios, data + 0x2e) & 0x30) >> 4;
+		p->timing_20_2e_c0 = (nv_ro08(bios, data + 0x2e) & 0xc0) >> 6;
+		p->timing_20_2f_03 = (nv_ro08(bios, data + 0x2f) & 0x03) >> 0;
+		temp = nv_ro16(bios, data + 0x2c);
+		p->timing_20_2c_003f = (temp & 0x003f) >> 0;
+		p->timing_20_2c_1fc0 = (temp & 0x1fc0) >> 6;
+		p->timing_20_30_07 = (nv_ro08(bios, data + 0x30) & 0x07) >> 0;
+		p->timing_20_30_f8 = (nv_ro08(bios, data + 0x30) & 0xf8) >> 3;
+		temp = nv_ro16(bios, data + 0x31);
+		p->timing_20_31_0007 = (temp & 0x0007) >> 0;
+		p->timing_20_31_0078 = (temp & 0x0078) >> 3;
+		p->timing_20_31_0780 = (temp & 0x0780) >> 7;
+		p->timing_20_31_0800 = (temp & 0x0800) >> 11;
+		p->timing_20_31_7000 = (temp & 0x7000) >> 12;
+		p->timing_20_31_8000 = (temp & 0x8000) >> 15;
+		break;
+	default:
+		data = 0;
+		break;
+	}
+	return data;
+}
diff --git a/drivers/gpu/drm/nouveau/core/subdev/clock/base.c b/drivers/gpu/drm/nouveau/core/subdev/clock/base.c
index e2938a21b06f..dd62baead39c 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/clock/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/clock/base.c
@@ -182,9 +182,12 @@ nouveau_pstate_prog(struct nouveau_clock *clk, int pstatei)
 	clk->pstate = pstatei;
 
 	if (pfb->ram->calc) {
-		ret = pfb->ram->calc(pfb, pstate->base.domain[nv_clk_src_mem]);
-		if (ret == 0)
-			ret = pfb->ram->prog(pfb);
+		int khz = pstate->base.domain[nv_clk_src_mem];
+		do {
+			ret = pfb->ram->calc(pfb, khz);
+			if (ret == 0)
+				ret = pfb->ram->prog(pfb);
+		} while (ret > 0);
 		pfb->ram->tidy(pfb);
 	}
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/clock/nv04.c b/drivers/gpu/drm/nouveau/core/subdev/clock/nv04.c
index 30c1f3a4158e..b74db6cfc4e2 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/clock/nv04.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/clock/nv04.c
@@ -25,7 +25,7 @@
 #include <subdev/bios.h>
 #include <subdev/bios/pll.h>
 #include <subdev/clock.h>
-#include <subdev/devinit/priv.h>
+#include <subdev/devinit/nv04.h>
 
 #include "pll.h"
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/clock/nve0.c b/drivers/gpu/drm/nouveau/core/subdev/clock/nve0.c
index 4c62e84b96f5..d3c37c96f0e7 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/clock/nve0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/clock/nve0.c
@@ -457,7 +457,7 @@ nve0_domain[] = {
 	{ nv_clk_src_gpc    , 0x00, NVKM_CLK_DOM_FLAG_CORE, "core", 2000 },
 	{ nv_clk_src_hubk07 , 0x01, NVKM_CLK_DOM_FLAG_CORE },
 	{ nv_clk_src_rop    , 0x02, NVKM_CLK_DOM_FLAG_CORE },
-	{ nv_clk_src_mem    , 0x03, 0, "memory", 1000 },
+	{ nv_clk_src_mem    , 0x03, 0, "memory", 500 },
 	{ nv_clk_src_hubk06 , 0x04, NVKM_CLK_DOM_FLAG_CORE },
 	{ nv_clk_src_hubk01 , 0x05 },
 	{ nv_clk_src_vdec   , 0x06 },
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/base.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/base.c
index 79c81d3d9bac..8fa34e8152c2 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/base.c
@@ -24,9 +24,11 @@
 
 #include <core/option.h>
 
-#include <subdev/devinit.h>
 #include <subdev/bios.h>
 #include <subdev/bios/init.h>
+#include <subdev/vga.h>
+
+#include "priv.h"
 
 int
 _nouveau_devinit_fini(struct nouveau_object *object, bool suspend)
@@ -37,18 +39,41 @@ _nouveau_devinit_fini(struct nouveau_object *object, bool suspend)
 	if (suspend)
 		devinit->post = true;
 
+	/* unlock the extended vga crtc regs */
+	nv_lockvgac(devinit, false);
+
 	return nouveau_subdev_fini(&devinit->base, suspend);
 }
 
 int
 _nouveau_devinit_init(struct nouveau_object *object)
 {
+	struct nouveau_devinit_impl *impl = (void *)object->oclass;
 	struct nouveau_devinit *devinit = (void *)object;
-	int ret = nouveau_subdev_init(&devinit->base);
+	int ret;
+
+	ret = nouveau_subdev_init(&devinit->base);
+	if (ret)
+		return ret;
+
+	ret = nvbios_init(&devinit->base, devinit->post);
 	if (ret)
 		return ret;
 
-	return nvbios_init(&devinit->base, devinit->post);
+	if (impl->disable)
+		nv_device(devinit)->disable_mask |= impl->disable(devinit);
+	return 0;
+}
+
+void
+_nouveau_devinit_dtor(struct nouveau_object *object)
+{
+	struct nouveau_devinit *devinit = (void *)object;
+
+	/* lock crtc regs */
+	nv_lockvgac(devinit, true);
+
+	nouveau_subdev_destroy(&devinit->base);
 }
 
 int
@@ -57,6 +82,7 @@ nouveau_devinit_create_(struct nouveau_object *parent,
 			struct nouveau_oclass *oclass,
 			int size, void **pobject)
 {
+	struct nouveau_devinit_impl *impl = (void *)oclass;
 	struct nouveau_device *device = nv_device(parent);
 	struct nouveau_devinit *devinit;
 	int ret;
@@ -68,5 +94,7 @@ nouveau_devinit_create_(struct nouveau_object *parent,
 		return ret;
 
 	devinit->post = nouveau_boolopt(device->cfgopt, "NvForcePost", false);
+	devinit->meminit = impl->meminit;
+	devinit->pll_set = impl->pll_set;
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv04.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv04.c
index 27c8235f1a85..7037eae46e44 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv04.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv04.c
@@ -27,12 +27,7 @@
 #include <subdev/vga.h>
 
 #include "fbmem.h"
-#include "priv.h"
-
-struct nv04_devinit_priv {
-	struct nouveau_devinit base;
-	int owner;
-};
+#include "nv04.h"
 
 static void
 nv04_devinit_meminit(struct nouveau_devinit *devinit)
@@ -393,17 +388,21 @@ int
 nv04_devinit_fini(struct nouveau_object *object, bool suspend)
 {
 	struct nv04_devinit_priv *priv = (void *)object;
+	int ret;
 
 	/* make i2c busses accessible */
 	nv_mask(priv, 0x000200, 0x00000001, 0x00000001);
 
-	/* unlock extended vga crtc regs, and unslave crtcs */
-	nv_lockvgac(priv, false);
+	ret = nouveau_devinit_fini(&priv->base, suspend);
+	if (ret)
+		return ret;
+
+	/* unslave crtcs */
 	if (priv->owner < 0)
 		priv->owner = nv_rdvgaowner(priv);
 	nv_wrvgaowner(priv, 0);
 
-	return nouveau_devinit_fini(&priv->base, suspend);
+	return 0;
 }
 
 int
@@ -431,14 +430,13 @@ nv04_devinit_dtor(struct nouveau_object *object)
 {
 	struct nv04_devinit_priv *priv = (void *)object;
 
-	/* restore vga owner saved at first init, and lock crtc regs  */
+	/* restore vga owner saved at first init */
 	nv_wrvgaowner(priv, priv->owner);
-	nv_lockvgac(priv, true);
 
 	nouveau_devinit_destroy(&priv->base);
 }
 
-static int
+int
 nv04_devinit_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 		  struct nouveau_oclass *oclass, void *data, u32 size,
 		  struct nouveau_object **pobject)
@@ -451,19 +449,19 @@ nv04_devinit_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
-	priv->base.meminit = nv04_devinit_meminit;
-	priv->base.pll_set = nv04_devinit_pll_set;
 	priv->owner = -1;
 	return 0;
 }
 
-struct nouveau_oclass
-nv04_devinit_oclass = {
-	.handle = NV_SUBDEV(DEVINIT, 0x04),
-	.ofuncs = &(struct nouveau_ofuncs) {
+struct nouveau_oclass *
+nv04_devinit_oclass = &(struct nouveau_devinit_impl) {
+	.base.handle = NV_SUBDEV(DEVINIT, 0x04),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nv04_devinit_ctor,
 		.dtor = nv04_devinit_dtor,
 		.init = nv04_devinit_init,
 		.fini = nv04_devinit_fini,
 	},
-};
+	.meminit = nv04_devinit_meminit,
+	.pll_set = nv04_devinit_pll_set,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv04.h b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv04.h
new file mode 100644
index 000000000000..23470a57510c
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv04.h
@@ -0,0 +1,23 @@
+#ifndef __NVKM_DEVINIT_NV04_H__
+#define __NVKM_DEVINIT_NV04_H__
+
+#include "priv.h"
+
+struct nv04_devinit_priv {
+	struct nouveau_devinit base;
+	u8 owner;
+};
+
+int  nv04_devinit_ctor(struct nouveau_object *, struct nouveau_object *,
+		       struct nouveau_oclass *, void *, u32,
+		       struct nouveau_object **);
+void nv04_devinit_dtor(struct nouveau_object *);
+int  nv04_devinit_init(struct nouveau_object *);
+int  nv04_devinit_fini(struct nouveau_object *, bool);
+int  nv04_devinit_pll_set(struct nouveau_devinit *, u32, u32);
+
+void setPLL_single(struct nouveau_devinit *, u32, struct nouveau_pll_vals *);
+void setPLL_double_highregs(struct nouveau_devinit *, u32, struct nouveau_pll_vals *);
+void setPLL_double_lowregs(struct nouveau_devinit *, u32, struct nouveau_pll_vals *);
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv05.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv05.c
index b1912a8a8942..98b7e6780dc7 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv05.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv05.c
@@ -29,12 +29,7 @@
 #include <subdev/vga.h>
 
 #include "fbmem.h"
-#include "priv.h"
-
-struct nv05_devinit_priv {
-	struct nouveau_devinit base;
-	u8 owner;
-};
+#include "nv04.h"
 
 static void
 nv05_devinit_meminit(struct nouveau_devinit *devinit)
@@ -49,7 +44,7 @@ nv05_devinit_meminit(struct nouveau_devinit *devinit)
 		{ 0x06, 0x00 },
 		{ 0x00, 0x00 }
 	};
-	struct nv05_devinit_priv *priv = (void *)devinit;
+	struct nv04_devinit_priv *priv = (void *)devinit;
 	struct nouveau_bios *bios = nouveau_bios(priv);
 	struct io_mapping *fb;
 	u32 patt = 0xdeadbeef;
@@ -130,31 +125,15 @@ out:
 	fbmem_fini(fb);
 }
 
-static int
-nv05_devinit_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
-		  struct nouveau_oclass *oclass, void *data, u32 size,
-		  struct nouveau_object **pobject)
-{
-	struct nv05_devinit_priv *priv;
-	int ret;
-
-	ret = nouveau_devinit_create(parent, engine, oclass, &priv);
-	*pobject = nv_object(priv);
-	if (ret)
-		return ret;
-
-	priv->base.meminit = nv05_devinit_meminit;
-	priv->base.pll_set = nv04_devinit_pll_set;
-	return 0;
-}
-
-struct nouveau_oclass
-nv05_devinit_oclass = {
-	.handle = NV_SUBDEV(DEVINIT, 0x05),
-	.ofuncs = &(struct nouveau_ofuncs) {
-		.ctor = nv05_devinit_ctor,
+struct nouveau_oclass *
+nv05_devinit_oclass = &(struct nouveau_devinit_impl) {
+	.base.handle = NV_SUBDEV(DEVINIT, 0x05),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nv04_devinit_ctor,
 		.dtor = nv04_devinit_dtor,
 		.init = nv04_devinit_init,
 		.fini = nv04_devinit_fini,
 	},
-};
+	.meminit = nv05_devinit_meminit,
+	.pll_set = nv04_devinit_pll_set,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv10.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv10.c
index 8d274dba1ef1..32b3d2131a7f 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv10.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv10.c
@@ -27,17 +27,12 @@
 #include <subdev/vga.h>
 
 #include "fbmem.h"
-#include "priv.h"
-
-struct nv10_devinit_priv {
-	struct nouveau_devinit base;
-	u8 owner;
-};
+#include "nv04.h"
 
 static void
 nv10_devinit_meminit(struct nouveau_devinit *devinit)
 {
-	struct nv10_devinit_priv *priv = (void *)devinit;
+	struct nv04_devinit_priv *priv = (void *)devinit;
 	static const int mem_width[] = { 0x10, 0x00, 0x20 };
 	int mem_width_count;
 	uint32_t patt = 0xdeadbeef;
@@ -101,31 +96,15 @@ amount_found:
 	fbmem_fini(fb);
 }
 
-static int
-nv10_devinit_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
-		  struct nouveau_oclass *oclass, void *data, u32 size,
-		  struct nouveau_object **pobject)
-{
-	struct nv10_devinit_priv *priv;
-	int ret;
-
-	ret = nouveau_devinit_create(parent, engine, oclass, &priv);
-	*pobject = nv_object(priv);
-	if (ret)
-		return ret;
-
-	priv->base.meminit = nv10_devinit_meminit;
-	priv->base.pll_set = nv04_devinit_pll_set;
-	return 0;
-}
-
-struct nouveau_oclass
-nv10_devinit_oclass = {
-	.handle = NV_SUBDEV(DEVINIT, 0x10),
-	.ofuncs = &(struct nouveau_ofuncs) {
-		.ctor = nv10_devinit_ctor,
+struct nouveau_oclass *
+nv10_devinit_oclass = &(struct nouveau_devinit_impl) {
+	.base.handle = NV_SUBDEV(DEVINIT, 0x10),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nv04_devinit_ctor,
 		.dtor = nv04_devinit_dtor,
 		.init = nv04_devinit_init,
 		.fini = nv04_devinit_fini,
 	},
-};
+	.meminit = nv10_devinit_meminit,
+	.pll_set = nv04_devinit_pll_set,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv1a.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv1a.c
index e9743cdabe75..526d0c6faacd 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv1a.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv1a.c
@@ -22,37 +22,16 @@
  * Authors: Ben Skeggs
  */
 
-#include "priv.h"
+#include "nv04.h"
 
-struct nv1a_devinit_priv {
-	struct nouveau_devinit base;
-	u8 owner;
-};
-
-static int
-nv1a_devinit_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
-		  struct nouveau_oclass *oclass, void *data, u32 size,
-		  struct nouveau_object **pobject)
-{
-	struct nv1a_devinit_priv *priv;
-	int ret;
-
-	ret = nouveau_devinit_create(parent, engine, oclass, &priv);
-	*pobject = nv_object(priv);
-	if (ret)
-		return ret;
-
-	priv->base.pll_set = nv04_devinit_pll_set;
-	return 0;
-}
-
-struct nouveau_oclass
-nv1a_devinit_oclass = {
-	.handle = NV_SUBDEV(DEVINIT, 0x1a),
-	.ofuncs = &(struct nouveau_ofuncs) {
-		.ctor = nv1a_devinit_ctor,
+struct nouveau_oclass *
+nv1a_devinit_oclass = &(struct nouveau_devinit_impl) {
+	.base.handle = NV_SUBDEV(DEVINIT, 0x1a),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nv04_devinit_ctor,
 		.dtor = nv04_devinit_dtor,
 		.init = nv04_devinit_init,
 		.fini = nv04_devinit_fini,
 	},
-};
+	.pll_set = nv04_devinit_pll_set,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv20.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv20.c
index 6cc6080d3bc0..4689ba303b0b 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv20.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv20.c
@@ -24,18 +24,13 @@
  *
  */
 
-#include "priv.h"
+#include "nv04.h"
 #include "fbmem.h"
 
-struct nv20_devinit_priv {
-	struct nouveau_devinit base;
-	u8 owner;
-};
-
 static void
 nv20_devinit_meminit(struct nouveau_devinit *devinit)
 {
-	struct nv20_devinit_priv *priv = (void *)devinit;
+	struct nv04_devinit_priv *priv = (void *)devinit;
 	struct nouveau_device *device = nv_device(priv);
 	uint32_t mask = (device->chipset >= 0x25 ? 0x300 : 0x900);
 	uint32_t amount, off;
@@ -65,31 +60,15 @@ nv20_devinit_meminit(struct nouveau_devinit *devinit)
 	fbmem_fini(fb);
 }
 
-static int
-nv20_devinit_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
-		  struct nouveau_oclass *oclass, void *data, u32 size,
-		  struct nouveau_object **pobject)
-{
-	struct nv20_devinit_priv *priv;
-	int ret;
-
-	ret = nouveau_devinit_create(parent, engine, oclass, &priv);
-	*pobject = nv_object(priv);
-	if (ret)
-		return ret;
-
-	priv->base.meminit = nv20_devinit_meminit;
-	priv->base.pll_set = nv04_devinit_pll_set;
-	return 0;
-}
-
-struct nouveau_oclass
-nv20_devinit_oclass = {
-	.handle = NV_SUBDEV(DEVINIT, 0x20),
-	.ofuncs = &(struct nouveau_ofuncs) {
-		.ctor = nv20_devinit_ctor,
+struct nouveau_oclass *
+nv20_devinit_oclass = &(struct nouveau_devinit_impl) {
+	.base.handle = NV_SUBDEV(DEVINIT, 0x20),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nv04_devinit_ctor,
 		.dtor = nv04_devinit_dtor,
 		.init = nv04_devinit_init,
 		.fini = nv04_devinit_fini,
 	},
-};
+	.meminit = nv20_devinit_meminit,
+	.pll_set = nv04_devinit_pll_set,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv50.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv50.c
index 6df72247c477..b46c62a1d5d8 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv50.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv50.c
@@ -28,9 +28,9 @@
 #include <subdev/bios/init.h>
 #include <subdev/vga.h>
 
-#include "priv.h"
+#include "nv50.h"
 
-static int
+int
 nv50_devinit_pll_set(struct nouveau_devinit *devinit, u32 type, u32 freq)
 {
 	struct nv50_devinit_priv *priv = (void *)devinit;
@@ -74,6 +74,19 @@ nv50_devinit_pll_set(struct nouveau_devinit *devinit, u32 type, u32 freq)
 	return 0;
 }
 
+static u64
+nv50_devinit_disable(struct nouveau_devinit *devinit)
+{
+	struct nv50_devinit_priv *priv = (void *)devinit;
+	u32 r001540 = nv_rd32(priv, 0x001540);
+	u64 disable = 0ULL;
+
+	if (!(r001540 & 0x40000000))
+		disable |= (1ULL << NVDEV_ENGINE_MPEG);
+
+	return disable;
+}
+
 int
 nv50_devinit_init(struct nouveau_object *object)
 {
@@ -120,7 +133,7 @@ nv50_devinit_init(struct nouveau_object *object)
 	return 0;
 }
 
-static int
+int
 nv50_devinit_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 		  struct nouveau_oclass *oclass, void *data, u32 size,
 		  struct nouveau_object **pobject)
@@ -133,17 +146,18 @@ nv50_devinit_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
-	priv->base.pll_set = nv50_devinit_pll_set;
 	return 0;
 }
 
-struct nouveau_oclass
-nv50_devinit_oclass = {
-	.handle = NV_SUBDEV(DEVINIT, 0x50),
-	.ofuncs = &(struct nouveau_ofuncs) {
+struct nouveau_oclass *
+nv50_devinit_oclass = &(struct nouveau_devinit_impl) {
+	.base.handle = NV_SUBDEV(DEVINIT, 0x50),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nv50_devinit_ctor,
 		.dtor = _nouveau_devinit_dtor,
 		.init = nv50_devinit_init,
 		.fini = _nouveau_devinit_fini,
 	},
-};
+	.pll_set = nv50_devinit_pll_set,
+	.disable = nv50_devinit_disable,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv50.h b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv50.h
new file mode 100644
index 000000000000..141c27e9f182
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv50.h
@@ -0,0 +1,18 @@
+#ifndef __NVKM_DEVINIT_NV50_H__
+#define __NVKM_DEVINIT_NV50_H__
+
+#include "priv.h"
+
+struct nv50_devinit_priv {
+	struct nouveau_devinit base;
+};
+
+int  nv50_devinit_ctor(struct nouveau_object *, struct nouveau_object *,
+		       struct nouveau_oclass *, void *, u32,
+		       struct nouveau_object **);
+int  nv50_devinit_init(struct nouveau_object *);
+int  nv50_devinit_pll_set(struct nouveau_devinit *, u32, u32);
+
+int  nva3_devinit_pll_set(struct nouveau_devinit *, u32, u32);
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv84.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv84.c
new file mode 100644
index 000000000000..787422505d87
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv84.c
@@ -0,0 +1,63 @@
+/*
+ * 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 "nv50.h"
+
+static u64
+nv84_devinit_disable(struct nouveau_devinit *devinit)
+{
+	struct nv50_devinit_priv *priv = (void *)devinit;
+	u32 r001540 = nv_rd32(priv, 0x001540);
+	u32 r00154c = nv_rd32(priv, 0x00154c);
+	u64 disable = 0ULL;
+
+	if (!(r001540 & 0x40000000)) {
+		disable |= (1ULL << NVDEV_ENGINE_MPEG);
+		disable |= (1ULL << NVDEV_ENGINE_VP);
+		disable |= (1ULL << NVDEV_ENGINE_BSP);
+		disable |= (1ULL << NVDEV_ENGINE_CRYPT);
+	}
+
+	if (!(r00154c & 0x00000004))
+		disable |= (1ULL << NVDEV_ENGINE_DISP);
+	if (!(r00154c & 0x00000020))
+		disable |= (1ULL << NVDEV_ENGINE_BSP);
+	if (!(r00154c & 0x00000040))
+		disable |= (1ULL << NVDEV_ENGINE_CRYPT);
+
+	return disable;
+}
+
+struct nouveau_oclass *
+nv84_devinit_oclass = &(struct nouveau_devinit_impl) {
+	.base.handle = NV_SUBDEV(DEVINIT, 0x84),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nv50_devinit_ctor,
+		.dtor = _nouveau_devinit_dtor,
+		.init = nv50_devinit_init,
+		.fini = _nouveau_devinit_fini,
+	},
+	.pll_set = nv50_devinit_pll_set,
+	.disable = nv84_devinit_disable,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nv98.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv98.c
new file mode 100644
index 000000000000..2b0e963fc6f0
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nv98.c
@@ -0,0 +1,62 @@
+/*
+ * 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 "nv50.h"
+
+static u64
+nv98_devinit_disable(struct nouveau_devinit *devinit)
+{
+	struct nv50_devinit_priv *priv = (void *)devinit;
+	u32 r001540 = nv_rd32(priv, 0x001540);
+	u32 r00154c = nv_rd32(priv, 0x00154c);
+	u64 disable = 0ULL;
+
+	if (!(r001540 & 0x40000000)) {
+		disable |= (1ULL << NVDEV_ENGINE_VP);
+		disable |= (1ULL << NVDEV_ENGINE_BSP);
+		disable |= (1ULL << NVDEV_ENGINE_PPP);
+	}
+
+	if (!(r00154c & 0x00000004))
+		disable |= (1ULL << NVDEV_ENGINE_DISP);
+	if (!(r00154c & 0x00000020))
+		disable |= (1ULL << NVDEV_ENGINE_BSP);
+	if (!(r00154c & 0x00000040))
+		disable |= (1ULL << NVDEV_ENGINE_CRYPT);
+
+	return disable;
+}
+
+struct nouveau_oclass *
+nv98_devinit_oclass = &(struct nouveau_devinit_impl) {
+	.base.handle = NV_SUBDEV(DEVINIT, 0x98),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nv50_devinit_ctor,
+		.dtor = _nouveau_devinit_dtor,
+		.init = nv50_devinit_init,
+		.fini = _nouveau_devinit_fini,
+	},
+	.pll_set = nv50_devinit_pll_set,
+	.disable = nv98_devinit_disable,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nva3.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nva3.c
index 76a68b290141..6dedf1dad7f7 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/nva3.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nva3.c
@@ -22,12 +22,12 @@
  * Authors: Ben Skeggs
  */
 
-#include "priv.h"
+#include "nv50.h"
 
-static int
+int
 nva3_devinit_pll_set(struct nouveau_devinit *devinit, u32 type, u32 freq)
 {
-	struct nva3_devinit_priv *priv = (void *)devinit;
+	struct nv50_devinit_priv *priv = (void *)devinit;
 	struct nouveau_bios *bios = nouveau_bios(priv);
 	struct nvbios_pll info;
 	int N, fN, M, P;
@@ -58,30 +58,38 @@ nva3_devinit_pll_set(struct nouveau_devinit *devinit, u32 type, u32 freq)
 	return ret;
 }
 
-static int
-nva3_devinit_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
-		  struct nouveau_oclass *oclass, void *data, u32 size,
-		  struct nouveau_object **pobject)
+static u64
+nva3_devinit_disable(struct nouveau_devinit *devinit)
 {
-	struct nv50_devinit_priv *priv;
-	int ret;
+	struct nv50_devinit_priv *priv = (void *)devinit;
+	u32 r001540 = nv_rd32(priv, 0x001540);
+	u32 r00154c = nv_rd32(priv, 0x00154c);
+	u64 disable = 0ULL;
 
-	ret = nouveau_devinit_create(parent, engine, oclass, &priv);
-	*pobject = nv_object(priv);
-	if (ret)
-		return ret;
+	if (!(r001540 & 0x40000000)) {
+		disable |= (1ULL << NVDEV_ENGINE_VP);
+		disable |= (1ULL << NVDEV_ENGINE_PPP);
+	}
+
+	if (!(r00154c & 0x00000004))
+		disable |= (1ULL << NVDEV_ENGINE_DISP);
+	if (!(r00154c & 0x00000020))
+		disable |= (1ULL << NVDEV_ENGINE_BSP);
+	if (!(r00154c & 0x00000200))
+		disable |= (1ULL << NVDEV_ENGINE_COPY0);
 
-	priv->base.pll_set = nva3_devinit_pll_set;
-	return 0;
+	return disable;
 }
 
-struct nouveau_oclass
-nva3_devinit_oclass = {
-	.handle = NV_SUBDEV(DEVINIT, 0xa3),
-	.ofuncs = &(struct nouveau_ofuncs) {
-		.ctor = nva3_devinit_ctor,
+struct nouveau_oclass *
+nva3_devinit_oclass = &(struct nouveau_devinit_impl) {
+	.base.handle = NV_SUBDEV(DEVINIT, 0xa3),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nv50_devinit_ctor,
 		.dtor = _nouveau_devinit_dtor,
 		.init = nv50_devinit_init,
 		.fini = _nouveau_devinit_fini,
 	},
-};
+	.pll_set = nva3_devinit_pll_set,
+	.disable = nva3_devinit_disable,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nvaf.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nvaf.c
new file mode 100644
index 000000000000..4fc68d27eff3
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nvaf.c
@@ -0,0 +1,63 @@
+/*
+ * 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 "nv50.h"
+
+static u64
+nvaf_devinit_disable(struct nouveau_devinit *devinit)
+{
+	struct nv50_devinit_priv *priv = (void *)devinit;
+	u32 r001540 = nv_rd32(priv, 0x001540);
+	u32 r00154c = nv_rd32(priv, 0x00154c);
+	u64 disable = 0;
+
+	if (!(r001540 & 0x40000000)) {
+		disable |= (1ULL << NVDEV_ENGINE_VP);
+		disable |= (1ULL << NVDEV_ENGINE_PPP);
+	}
+
+	if (!(r00154c & 0x00000004))
+		disable |= (1ULL << NVDEV_ENGINE_DISP);
+	if (!(r00154c & 0x00000020))
+		disable |= (1ULL << NVDEV_ENGINE_BSP);
+	if (!(r00154c & 0x00000040))
+		disable |= (1ULL << NVDEV_ENGINE_VIC);
+	if (!(r00154c & 0x00000200))
+		disable |= (1ULL << NVDEV_ENGINE_COPY0);
+
+	return disable;
+}
+
+struct nouveau_oclass *
+nvaf_devinit_oclass = &(struct nouveau_devinit_impl) {
+	.base.handle = NV_SUBDEV(DEVINIT, 0xaf),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
+		.ctor = nv50_devinit_ctor,
+		.dtor = _nouveau_devinit_dtor,
+		.init = nv50_devinit_init,
+		.fini = _nouveau_devinit_fini,
+	},
+	.pll_set = nva3_devinit_pll_set,
+	.disable = nvaf_devinit_disable,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/nvc0.c b/drivers/gpu/drm/nouveau/core/subdev/devinit/nvc0.c
index 19e265bf4574..fa7e63766b1b 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/nvc0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/nvc0.c
@@ -22,12 +22,12 @@
  * Authors: Ben Skeggs
  */
 
-#include "priv.h"
+#include "nv50.h"
 
 static int
 nvc0_devinit_pll_set(struct nouveau_devinit *devinit, u32 type, u32 freq)
 {
-	struct nvc0_devinit_priv *priv = (void *)devinit;
+	struct nv50_devinit_priv *priv = (void *)devinit;
 	struct nouveau_bios *bios = nouveau_bios(priv);
 	struct nvbios_pll info;
 	int N, fN, M, P;
@@ -59,6 +59,33 @@ nvc0_devinit_pll_set(struct nouveau_devinit *devinit, u32 type, u32 freq)
 	return ret;
 }
 
+static u64
+nvc0_devinit_disable(struct nouveau_devinit *devinit)
+{
+	struct nv50_devinit_priv *priv = (void *)devinit;
+	u32 r022500 = nv_rd32(priv, 0x022500);
+	u64 disable = 0ULL;
+
+	if (r022500 & 0x00000001)
+		disable |= (1ULL << NVDEV_ENGINE_DISP);
+
+	if (r022500 & 0x00000002) {
+		disable |= (1ULL << NVDEV_ENGINE_VP);
+		disable |= (1ULL << NVDEV_ENGINE_PPP);
+	}
+
+	if (r022500 & 0x00000004)
+		disable |= (1ULL << NVDEV_ENGINE_BSP);
+	if (r022500 & 0x00000008)
+		disable |= (1ULL << NVDEV_ENGINE_VENC);
+	if (r022500 & 0x00000100)
+		disable |= (1ULL << NVDEV_ENGINE_COPY0);
+	if (r022500 & 0x00000200)
+		disable |= (1ULL << NVDEV_ENGINE_COPY1);
+
+	return disable;
+}
+
 static int
 nvc0_devinit_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 		  struct nouveau_oclass *oclass, void *data, u32 size,
@@ -72,19 +99,20 @@ nvc0_devinit_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	if (ret)
 		return ret;
 
-	priv->base.pll_set = nvc0_devinit_pll_set;
 	if (nv_rd32(priv, 0x022500) & 0x00000001)
 		priv->base.post = true;
 	return 0;
 }
 
-struct nouveau_oclass
-nvc0_devinit_oclass = {
-	.handle = NV_SUBDEV(DEVINIT, 0xc0),
-	.ofuncs = &(struct nouveau_ofuncs) {
+struct nouveau_oclass *
+nvc0_devinit_oclass = &(struct nouveau_devinit_impl) {
+	.base.handle = NV_SUBDEV(DEVINIT, 0xc0),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nvc0_devinit_ctor,
 		.dtor = _nouveau_devinit_dtor,
 		.init = nv50_devinit_init,
 		.fini = _nouveau_devinit_fini,
 	},
-};
+	.pll_set = nvc0_devinit_pll_set,
+	.disable = nvc0_devinit_disable,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/devinit/priv.h b/drivers/gpu/drm/nouveau/core/subdev/devinit/priv.h
index 7d622e2b0171..822a2fbf44a5 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/devinit/priv.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/devinit/priv.h
@@ -6,20 +6,32 @@
 #include <subdev/clock/pll.h>
 #include <subdev/devinit.h>
 
-void nv04_devinit_dtor(struct nouveau_object *);
-int  nv04_devinit_init(struct nouveau_object *);
-int  nv04_devinit_fini(struct nouveau_object *, bool);
-int  nv04_devinit_pll_set(struct nouveau_devinit *, u32, u32);
-
-void setPLL_single(struct nouveau_devinit *, u32, struct nouveau_pll_vals *);
-void setPLL_double_highregs(struct nouveau_devinit *, u32, struct nouveau_pll_vals *);
-void setPLL_double_lowregs(struct nouveau_devinit *, u32, struct nouveau_pll_vals *);
-
-
-struct nv50_devinit_priv {
-	struct nouveau_devinit base;
+struct nouveau_devinit_impl {
+	struct nouveau_oclass base;
+	void (*meminit)(struct nouveau_devinit *);
+	int  (*pll_set)(struct nouveau_devinit *, u32 type, u32 freq);
+	u64  (*disable)(struct nouveau_devinit *);
 };
 
-int  nv50_devinit_init(struct nouveau_object *);
+#define nouveau_devinit_create(p,e,o,d)                                        \
+	nouveau_devinit_create_((p), (e), (o), sizeof(**d), (void **)d)
+#define nouveau_devinit_destroy(p) ({                                          \
+	struct nouveau_devinit *d = (p);                                       \
+	_nouveau_devinit_dtor(nv_object(d));                                   \
+})
+#define nouveau_devinit_init(p) ({                                             \
+	struct nouveau_devinit *d = (p);                                       \
+	_nouveau_devinit_init(nv_object(d));                                   \
+})
+#define nouveau_devinit_fini(p,s) ({                                           \
+	struct nouveau_devinit *d = (p);                                       \
+	_nouveau_devinit_fini(nv_object(d), (s));                              \
+})
+
+int nouveau_devinit_create_(struct nouveau_object *, struct nouveau_object *,
+			    struct nouveau_oclass *, int, void **);
+void _nouveau_devinit_dtor(struct nouveau_object *);
+int _nouveau_devinit_init(struct nouveau_object *);
+int _nouveau_devinit_fini(struct nouveau_object *, bool suspend);
 
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/gddr5.c b/drivers/gpu/drm/nouveau/core/subdev/fb/gddr5.c
index 34f9605ffee6..66fe959b4f74 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/gddr5.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/gddr5.c
@@ -25,35 +25,44 @@
 #include <subdev/bios.h>
 #include "priv.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
+
 int
-nouveau_gddr5_calc(struct nouveau_ram *ram)
+nouveau_gddr5_calc(struct nouveau_ram *ram, bool nuts)
 {
-	struct nouveau_bios *bios = nouveau_bios(ram);
-	int pd, lf, xd, vh, vr, vo;
-	int WL, CL, WR, at, dt, ds;
+	int pd, lf, xd, vh, vr, vo, l3;
+	int WL, CL, WR, at[2], dt, ds;
 	int rq = ram->freq < 1000000; /* XXX */
 
-	switch (!!ram->ramcfg.data * ram->ramcfg.version) {
+	switch (ram->ramcfg.version) {
 	case 0x11:
-		pd =  (nv_ro08(bios, ram->ramcfg.data + 0x01) & 0x80) >> 7;
-		lf =  (nv_ro08(bios, ram->ramcfg.data + 0x01) & 0x40) >> 6;
-		xd = !(nv_ro08(bios, ram->ramcfg.data + 0x01) & 0x20);
-		vh =  (nv_ro08(bios, ram->ramcfg.data + 0x02) & 0x10) >> 4;
-		vr =  (nv_ro08(bios, ram->ramcfg.data + 0x02) & 0x04) >> 2;
-		vo =   nv_ro08(bios, ram->ramcfg.data + 0x06) & 0xff;
+		pd =  ram->next->bios.ramcfg_11_01_80;
+		lf =  ram->next->bios.ramcfg_11_01_40;
+		xd = !ram->next->bios.ramcfg_11_01_20;
+		vh =  ram->next->bios.ramcfg_11_02_10;
+		vr =  ram->next->bios.ramcfg_11_02_04;
+		vo =  ram->next->bios.ramcfg_11_06;
+		l3 = !ram->next->bios.ramcfg_11_07_02;
 		break;
 	default:
 		return -ENOSYS;
 	}
 
-	switch (!!ram->timing.data * ram->timing.version) {
+	switch (ram->timing.version) {
 	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;
-		at = (nv_ro08(bios, ram->timing.data + 0x2e) & 0xc0) >> 6;
-		dt =  nv_ro08(bios, ram->timing.data + 0x2e) & 0x03;
-		ds =  nv_ro08(bios, ram->timing.data + 0x2f) & 0x03;
+		WL = (ram->next->bios.timing[1] & 0x00000f80) >> 7;
+		CL = (ram->next->bios.timing[1] & 0x0000001f);
+		WR = (ram->next->bios.timing[2] & 0x007f0000) >> 16;
+		at[0] = ram->next->bios.timing_20_2e_c0;
+		at[1] = ram->next->bios.timing_20_2e_30;
+		dt =  ram->next->bios.timing_20_2e_03;
+		ds =  ram->next->bios.timing_20_2f_03;
 		break;
 	default:
 		return -ENOSYS;
@@ -71,13 +80,25 @@ nouveau_gddr5_calc(struct nouveau_ram *ram)
 
 	ram->mr[1] &= ~0x0bf;
 	ram->mr[1] |= (xd & 0x01) << 7;
-	ram->mr[1] |= (at & 0x03) << 4;
+	ram->mr[1] |= (at[0] & 0x03) << 4;
 	ram->mr[1] |= (dt & 0x03) << 2;
 	ram->mr[1] |= (ds & 0x03) << 0;
 
+	/* this seems wrong, alternate field used for the broadcast
+	 * on nuts vs non-nuts configs..  meh, it matches for now.
+	 */
+	ram->mr1_nuts = ram->mr[1];
+	if (nuts) {
+		ram->mr[1] &= ~0x030;
+		ram->mr[1] |= (at[1] & 0x03) << 4;
+	}
+
 	ram->mr[3] &= ~0x020;
 	ram->mr[3] |= (rq & 0x01) << 5;
 
+	ram->mr[5] &= ~0x004;
+	ram->mr[5] |= (l3 << 2);
+
 	if (!vo)
 		vo = (ram->mr[6] & 0xff0) >> 4;
 	if (ram->mr[6] & 0x001)
@@ -86,11 +107,16 @@ nouveau_gddr5_calc(struct nouveau_ram *ram)
 	ram->mr[6] |= (vo & 0xff) << 4;
 	ram->mr[6] |= (pd & 0x01) << 0;
 
-	if (!(ram->mr[7] & 0x100))
-		vr = 0; /* binary driver does this.. bug? */
-	ram->mr[7] &= ~0x188;
-	ram->mr[7] |= (vr & 0x01) << 8;
+	if (NOTE00(vr)) {
+		ram->mr[7] &= ~0x300;
+		ram->mr[7] |= (vr & 0x03) << 8;
+	}
+	ram->mr[7] &= ~0x088;
 	ram->mr[7] |= (vh & 0x01) << 7;
 	ram->mr[7] |= (lf & 0x01) << 3;
+
+	ram->mr[8] &= ~0x003;
+	ram->mr[8] |= (WR & 0x10) >> 3;
+	ram->mr[8] |= (CL & 0x10) >> 4;
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/nvc0.c b/drivers/gpu/drm/nouveau/core/subdev/fb/nvc0.c
index e5fc37c4caac..45470e1f0385 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/nvc0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/nvc0.c
@@ -33,6 +33,21 @@ nvc0_fb_memtype_valid(struct nouveau_fb *pfb, u32 tile_flags)
 	return likely((nvc0_pte_storage_type_map[memtype] != 0xff));
 }
 
+static void
+nvc0_fb_intr(struct nouveau_subdev *subdev)
+{
+	struct nvc0_fb_priv *priv = (void *)subdev;
+	u32 intr = nv_rd32(priv, 0x000100);
+	if (intr & 0x08000000) {
+		nv_debug(priv, "PFFB intr\n");
+		intr &= ~0x08000000;
+	}
+	if (intr & 0x00002000) {
+		nv_debug(priv, "PBFB intr\n");
+		intr &= ~0x00002000;
+	}
+}
+
 int
 nvc0_fb_init(struct nouveau_object *object)
 {
@@ -86,6 +101,7 @@ nvc0_fb_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 			return -EFAULT;
 	}
 
+	nv_subdev(priv)->intr = nvc0_fb_intr;
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/priv.h b/drivers/gpu/drm/nouveau/core/subdev/fb/priv.h
index 493125214e88..edaf95dee612 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/priv.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/priv.h
@@ -34,7 +34,7 @@ extern struct nouveau_oclass nvc0_ram_oclass;
 extern struct nouveau_oclass nve0_ram_oclass;
 
 int nouveau_sddr3_calc(struct nouveau_ram *ram);
-int nouveau_gddr5_calc(struct nouveau_ram *ram);
+int nouveau_gddr5_calc(struct nouveau_ram *ram, bool nuts);
 
 #define nouveau_fb_create(p,e,c,d)                                             \
 	nouveau_fb_create_((p), (e), (c), sizeof(**d), (void **)d)
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnv50.c b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnv50.c
index 76762a17d89c..c7fdb3a9e88b 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnv50.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnv50.c
@@ -70,13 +70,11 @@ nv50_ram_calc(struct nouveau_fb *pfb, u32 freq)
 	struct nv50_ramseq *hwsq = &ram->hwsq;
 	struct nvbios_perfE perfE;
 	struct nvbios_pll mpll;
-	struct bit_entry M;
 	struct {
 		u32 data;
 		u8  size;
 	} ramcfg, timing;
-	u8  ver, hdr, cnt, strap;
-	u32 data;
+	u8  ver, hdr, cnt, len, strap;
 	int N1, M1, N2, M2, P;
 	int ret, i;
 
@@ -93,16 +91,7 @@ nv50_ram_calc(struct nouveau_fb *pfb, u32 freq)
 	} while (perfE.memory < freq);
 
 	/* locate specific data set for the attached memory */
-	if (bit_entry(bios, 'M', &M) || M.version != 1 || M.length < 5) {
-		nv_error(pfb, "invalid/missing memory table\n");
-		return -EINVAL;
-	}
-
-	strap = (nv_rd32(pfb, 0x101000) & 0x0000003c) >> 2;
-	data = nv_ro16(bios, M.offset + 3);
-	if (data)
-		strap = nv_ro08(bios, data + strap);
-
+	strap = nvbios_ramcfg_index(bios);
 	if (strap >= cnt) {
 		nv_error(pfb, "invalid ramcfg strap\n");
 		return -EINVAL;
@@ -113,7 +102,8 @@ nv50_ram_calc(struct nouveau_fb *pfb, u32 freq)
 	/* lookup memory timings, if bios says they're present */
 	strap = nv_ro08(bios, ramcfg.data + 0x01);
 	if (strap != 0xff) {
-		timing.data = nvbios_timing_entry(bios, strap, &ver, &hdr);
+		timing.data = nvbios_timingEe(bios, strap, &ver, &hdr,
+					     &cnt, &len);
 		if (!timing.data || ver != 0x10 || hdr < 0x12) {
 			nv_error(pfb, "invalid/missing timing entry "
 				 "%02x %04x %02x %02x\n",
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnva3.c b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnva3.c
index f6292cd9207c..f4ae8aa46a25 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnva3.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnva3.c
@@ -79,8 +79,7 @@ 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;
-	struct bit_entry M;
-	u8  ver, cnt, strap;
+	u8  ver, cnt, len, strap;
 	u32 data;
 	struct {
 		u32 data;
@@ -91,24 +90,15 @@ nva3_ram_calc(struct nouveau_fb *pfb, u32 freq)
 	int ret;
 
 	/* lookup memory config data relevant to the target frequency */
-	rammap.data = nvbios_rammap_match(bios, freq / 1000, &ver, &rammap.size,
-					 &cnt, &ramcfg.size);
+	rammap.data = nvbios_rammapEm(bios, freq / 1000, &ver, &rammap.size,
+				     &cnt, &ramcfg.size);
 	if (!rammap.data || ver != 0x10 || rammap.size < 0x0e) {
 		nv_error(pfb, "invalid/missing rammap entry\n");
 		return -EINVAL;
 	}
 
 	/* locate specific data set for the attached memory */
-	if (bit_entry(bios, 'M', &M) || M.version != 2 || M.length < 3) {
-		nv_error(pfb, "invalid/missing memory table\n");
-		return -EINVAL;
-	}
-
-	strap = (nv_rd32(pfb, 0x101000) & 0x0000003c) >> 2;
-	data = nv_ro16(bios, M.offset + 1);
-	if (data)
-		strap = nv_ro08(bios, data + strap);
-
+	strap = nvbios_ramcfg_index(bios);
 	if (strap >= cnt) {
 		nv_error(pfb, "invalid ramcfg strap\n");
 		return -EINVAL;
@@ -123,8 +113,8 @@ nva3_ram_calc(struct nouveau_fb *pfb, u32 freq)
 	/* lookup memory timings, if bios says they're present */
 	strap = nv_ro08(bios, ramcfg.data + 0x01);
 	if (strap != 0xff) {
-		timing.data = nvbios_timing_entry(bios, strap, &ver,
-						 &timing.size);
+		timing.data = nvbios_timingEe(bios, strap, &ver, &timing.size,
+					     &cnt, &len);
 		if (!timing.data || ver != 0x10 || timing.size < 0x19) {
 			nv_error(pfb, "invalid/missing timing entry\n");
 			return -EINVAL;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnvc0.c b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnvc0.c
index f464547c6bab..0391b824ee76 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnvc0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnvc0.c
@@ -23,7 +23,6 @@
  */
 
 #include <subdev/bios.h>
-#include <subdev/bios/bit.h>
 #include <subdev/bios/pll.h>
 #include <subdev/bios/rammap.h>
 #include <subdev/bios/timing.h>
@@ -134,9 +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 bit_entry M;
-	u8  ver, cnt, strap;
-	u32 data;
+	u8  ver, cnt, len, strap;
 	struct {
 		u32 data;
 		u8  size;
@@ -147,24 +144,15 @@ nvc0_ram_calc(struct nouveau_fb *pfb, u32 freq)
 	int ret;
 
 	/* lookup memory config data relevant to the target frequency */
-	rammap.data = nvbios_rammap_match(bios, freq / 1000, &ver, &rammap.size,
-					 &cnt, &ramcfg.size);
+	rammap.data = nvbios_rammapEm(bios, freq / 1000, &ver, &rammap.size,
+				     &cnt, &ramcfg.size);
 	if (!rammap.data || ver != 0x10 || rammap.size < 0x0e) {
 		nv_error(pfb, "invalid/missing rammap entry\n");
 		return -EINVAL;
 	}
 
 	/* locate specific data set for the attached memory */
-	if (bit_entry(bios, 'M', &M) || M.version != 2 || M.length < 3) {
-		nv_error(pfb, "invalid/missing memory table\n");
-		return -EINVAL;
-	}
-
-	strap = (nv_rd32(pfb, 0x101000) & 0x0000003c) >> 2;
-	data = nv_ro16(bios, M.offset + 1);
-	if (data)
-		strap = nv_ro08(bios, data + strap);
-
+	strap = nvbios_ramcfg_index(bios);
 	if (strap >= cnt) {
 		nv_error(pfb, "invalid ramcfg strap\n");
 		return -EINVAL;
@@ -179,8 +167,8 @@ nvc0_ram_calc(struct nouveau_fb *pfb, u32 freq)
 	/* lookup memory timings, if bios says they're present */
 	strap = nv_ro08(bios, ramcfg.data + 0x01);
 	if (strap != 0xff) {
-		timing.data = nvbios_timing_entry(bios, strap, &ver,
-						 &timing.size);
+		timing.data = nvbios_timingEe(bios, strap, &ver, &timing.size,
+					     &cnt, &len);
 		if (!timing.data || ver != 0x10 || timing.size < 0x19) {
 			nv_error(pfb, "invalid/missing timing entry\n");
 			return -EINVAL;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnve0.c b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnve0.c
index bc86cfd084f6..3257c522a021 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/fb/ramnve0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/fb/ramnve0.c
@@ -25,7 +25,6 @@
 #include <subdev/gpio.h>
 
 #include <subdev/bios.h>
-#include <subdev/bios/bit.h>
 #include <subdev/bios/pll.h>
 #include <subdev/bios/init.h>
 #include <subdev/bios/rammap.h>
@@ -42,6 +41,14 @@
 
 #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;
 
@@ -104,7 +111,9 @@ struct nve0_ramfuc {
 	struct ramfuc_reg r_mr[16]; /* MR0 - MR8, MR15 */
 
 	struct ramfuc_reg r_0x62c000;
+
 	struct ramfuc_reg r_0x10f200;
+
 	struct ramfuc_reg r_0x10f210;
 	struct ramfuc_reg r_0x10f310;
 	struct ramfuc_reg r_0x10f314;
@@ -118,12 +127,17 @@ struct nve0_ramfuc {
 	struct ramfuc_reg r_0x10f65c;
 	struct ramfuc_reg r_0x10f6bc;
 	struct ramfuc_reg r_0x100710;
-	struct ramfuc_reg r_0x10f750;
+	struct ramfuc_reg r_0x100750;
 };
 
 struct nve0_ram {
 	struct nouveau_ram base;
 	struct nve0_ramfuc fuc;
+
+	u32 parts;
+	u32 pmask;
+	u32 pnuts;
+
 	int from;
 	int mode;
 	int N1, fN1, M1, P1;
@@ -134,17 +148,17 @@ struct nve0_ram {
  * GDDR5
  ******************************************************************************/
 static void
-train(struct nve0_ramfuc *fuc, u32 magic)
+nve0_ram_train(struct nve0_ramfuc *fuc, u32 mask, u32 data)
 {
 	struct nve0_ram *ram = container_of(fuc, typeof(*ram), fuc);
-	struct nouveau_fb *pfb = nouveau_fb(ram);
-	const int mc = nv_rd32(pfb, 0x02243c);
-	int i;
-
-	ram_mask(fuc, 0x10f910, 0xbc0e0000, magic);
-	ram_mask(fuc, 0x10f914, 0xbc0e0000, magic);
-	for (i = 0; i < mc; i++) {
-		const u32 addr = 0x110974 + (i * 0x1000);
+	u32 addr = 0x110974, i;
+
+	ram_mask(fuc, 0x10f910, mask, data);
+	ram_mask(fuc, 0x10f914, mask, data);
+
+	for (i = 0; (data & 0x80000000) && i < ram->parts; addr += 0x1000, i++) {
+		if (ram->pmask & (1 << i))
+			continue;
 		ram_wait(fuc, addr, 0x0000000f, 0x00000000, 500000);
 	}
 }
@@ -199,12 +213,12 @@ r1373f4_init(struct nve0_ramfuc *fuc)
 }
 
 static void
-r1373f4_fini(struct nve0_ramfuc *fuc, u32 ramcfg)
+r1373f4_fini(struct nve0_ramfuc *fuc)
 {
 	struct nve0_ram *ram = container_of(fuc, typeof(*ram), fuc);
-	struct nouveau_bios *bios = nouveau_bios(ram);
-	u8 v0 = (nv_ro08(bios, ramcfg + 0x03) & 0xc0) >> 6;
-	u8 v1 = (nv_ro08(bios, ramcfg + 0x03) & 0x30) >> 4;
+	struct nouveau_ram_data *next = ram->base.next;
+	u8 v0 = next->bios.ramcfg_11_03_c0;
+	u8 v1 = next->bios.ramcfg_11_03_30;
 	u32 tmp;
 
 	tmp = ram_rd32(fuc, 0x1373ec) & ~0x00030000;
@@ -220,25 +234,46 @@ r1373f4_fini(struct nve0_ramfuc *fuc, u32 ramcfg)
 	ram_mask(fuc, 0x10f800, 0x00000030, (v0 ^ v1) << 4);
 }
 
+static void
+nve0_ram_nuts(struct nve0_ram *ram, struct ramfuc_reg *reg,
+	      u32 _mask, u32 _data, u32 _copy)
+{
+	struct nve0_fb_priv *priv = (void *)nouveau_fb(ram);
+	struct ramfuc *fuc = &ram->fuc.base;
+	u32 addr = 0x110000 + (reg->addr[0] & 0xfff);
+	u32 mask = _mask | _copy;
+	u32 data = (_data & _mask) | (reg->data & _copy);
+	u32 i;
+
+	for (i = 0; i < 16; i++, addr += 0x1000) {
+		if (ram->pnuts & (1 << i)) {
+			u32 prev = nv_rd32(priv, addr);
+			u32 next = (prev & ~mask) | data;
+			nouveau_memx_wr32(fuc->memx, addr, next);
+		}
+	}
+}
+#define ram_nuts(s,r,m,d,c)                                                    \
+	nve0_ram_nuts((s), &(s)->fuc.r_##r, (m), (d), (c))
+
 static int
 nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 {
-	struct nouveau_bios *bios = nouveau_bios(pfb);
 	struct nve0_ram *ram = (void *)pfb->ram;
 	struct nve0_ramfuc *fuc = &ram->fuc;
-	const u32 rammap = ram->base.rammap.data;
-	const u32 ramcfg = ram->base.ramcfg.data;
-	const u32 timing = ram->base.timing.data;
-	int vc = !(nv_ro08(bios, ramcfg + 0x02) & 0x08);
-	int mv = 1; /*XXX*/
+	struct nouveau_ram_data *next = ram->base.next;
+	int vc = !(next->bios.ramcfg_11_02_08);
+	int mv = !(next->bios.ramcfg_11_02_04);
 	u32 mask, data;
 
 	ram_mask(fuc, 0x10f808, 0x40000000, 0x40000000);
 	ram_wr32(fuc, 0x62c000, 0x0f0f0000);
 
 	/* MR1: turn termination on early, for some reason.. */
-	if ((ram->base.mr[1] & 0x03c) != 0x030)
+	if ((ram->base.mr[1] & 0x03c) != 0x030) {
 		ram_mask(fuc, mr[1], 0x03c, ram->base.mr[1] & 0x03c);
+		ram_nuts(ram, mr[1], 0x03c, ram->base.mr1_nuts & 0x03c, 0x000);
+	}
 
 	if (vc == 1 && ram_have(fuc, gpio2E)) {
 		u32 temp  = ram_mask(fuc, gpio2E, 0x3000, fuc->r_func2E[1]);
@@ -250,8 +285,7 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 
 	ram_mask(fuc, 0x10f200, 0x00000800, 0x00000000);
 
-	ram_mask(fuc, 0x10f914, 0x01020000, 0x000c0000);
-	ram_mask(fuc, 0x10f910, 0x01020000, 0x000c0000);
+	nve0_ram_train(fuc, 0x01020000, 0x000c0000);
 
 	ram_wr32(fuc, 0x10f210, 0x00000000); /* REFRESH_AUTO = 0 */
 	ram_nsec(fuc, 1000);
@@ -280,28 +314,28 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 
 	if (1) {
 		data |= 0x800807e0;
-		switch (nv_ro08(bios, ramcfg + 0x03) & 0xc0) {
-		case 0xc0: data &= ~0x00000040; break;
-		case 0x80: data &= ~0x00000100; break;
-		case 0x40: data &= ~0x80000000; break;
-		case 0x00: data &= ~0x00000400; break;
+		switch (next->bios.ramcfg_11_03_c0) {
+		case 3: data &= ~0x00000040; break;
+		case 2: data &= ~0x00000100; break;
+		case 1: data &= ~0x80000000; break;
+		case 0: data &= ~0x00000400; break;
 		}
 
-		switch (nv_ro08(bios, ramcfg + 0x03) & 0x30) {
-		case 0x30: data &= ~0x00000020; break;
-		case 0x20: data &= ~0x00000080; break;
-		case 0x10: data &= ~0x00080000; break;
-		case 0x00: data &= ~0x00000200; break;
+		switch (next->bios.ramcfg_11_03_30) {
+		case 3: data &= ~0x00000020; break;
+		case 2: data &= ~0x00000080; break;
+		case 1: data &= ~0x00080000; break;
+		case 0: data &= ~0x00000200; break;
 		}
 	}
 
-	if (nv_ro08(bios, ramcfg + 0x02) & 0x80)
+	if (next->bios.ramcfg_11_02_80)
 		mask |= 0x03000000;
-	if (nv_ro08(bios, ramcfg + 0x02) & 0x40)
+	if (next->bios.ramcfg_11_02_40)
 		mask |= 0x00002000;
-	if (nv_ro08(bios, ramcfg + 0x07) & 0x10)
+	if (next->bios.ramcfg_11_07_10)
 		mask |= 0x00004000;
-	if (nv_ro08(bios, ramcfg + 0x07) & 0x08)
+	if (next->bios.ramcfg_11_07_08)
 		mask |= 0x00000003;
 	else {
 		mask |= 0x34000000;
@@ -314,18 +348,18 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 
 	if (ram->from == 2 && ram->mode != 2) {
 		ram_mask(fuc, 0x10f808, 0x00080000, 0x00000000);
-		ram_mask(fuc, 0x10f200, 0x00008000, 0x00008000);
+		ram_mask(fuc, 0x10f200, 0x18008000, 0x00008000);
 		ram_mask(fuc, 0x10f800, 0x00000000, 0x00000004);
 		ram_mask(fuc, 0x10f830, 0x00008000, 0x01040010);
 		ram_mask(fuc, 0x10f830, 0x01000000, 0x00000000);
 		r1373f4_init(fuc);
 		ram_mask(fuc, 0x1373f0, 0x00000002, 0x00000001);
-		r1373f4_fini(fuc, ramcfg);
+		r1373f4_fini(fuc);
 		ram_mask(fuc, 0x10f830, 0x00c00000, 0x00240001);
 	} else
 	if (ram->from != 2 && ram->mode != 2) {
 		r1373f4_init(fuc);
-		r1373f4_fini(fuc, ramcfg);
+		r1373f4_fini(fuc);
 	}
 
 	if (ram_have(fuc, gpioMV)) {
@@ -336,49 +370,54 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 		}
 	}
 
-	if ( (nv_ro08(bios, ramcfg + 0x02) & 0x40) ||
-	     (nv_ro08(bios, ramcfg + 0x07) & 0x10)) {
+	if ( (next->bios.ramcfg_11_02_40) ||
+	     (next->bios.ramcfg_11_07_10)) {
 		ram_mask(fuc, 0x132040, 0x00010000, 0x00010000);
 		ram_nsec(fuc, 20000);
 	}
 
 	if (ram->from != 2 && ram->mode == 2) {
+		if (0 /*XXX: Titan */)
+			ram_mask(fuc, 0x10f200, 0x18000000, 0x18000000);
 		ram_mask(fuc, 0x10f800, 0x00000004, 0x00000000);
 		ram_mask(fuc, 0x1373f0, 0x00000000, 0x00000002);
 		ram_mask(fuc, 0x10f830, 0x00800001, 0x00408010);
 		r1373f4_init(fuc);
-		r1373f4_fini(fuc, ramcfg);
+		r1373f4_fini(fuc);
 		ram_mask(fuc, 0x10f808, 0x00000000, 0x00080000);
 		ram_mask(fuc, 0x10f200, 0x00808000, 0x00800000);
 	} else
 	if (ram->from == 2 && ram->mode == 2) {
 		ram_mask(fuc, 0x10f800, 0x00000004, 0x00000000);
 		r1373f4_init(fuc);
-		r1373f4_fini(fuc, ramcfg);
+		r1373f4_fini(fuc);
 	}
 
 	if (ram->mode != 2) /*XXX*/ {
-		if (nv_ro08(bios, ramcfg + 0x07) & 0x40)
+		if (next->bios.ramcfg_11_07_40)
 			ram_mask(fuc, 0x10f670, 0x80000000, 0x80000000);
 	}
 
-	data = (nv_ro08(bios, rammap + 0x11) & 0x0c) >> 2;
-	ram_wr32(fuc, 0x10f65c, 0x00000011 * data);
-	ram_wr32(fuc, 0x10f6b8, 0x01010101 * nv_ro08(bios, ramcfg + 0x09));
-	ram_wr32(fuc, 0x10f6bc, 0x01010101 * nv_ro08(bios, ramcfg + 0x09));
+	ram_wr32(fuc, 0x10f65c, 0x00000011 * next->bios.rammap_11_11_0c);
+	ram_wr32(fuc, 0x10f6b8, 0x01010101 * next->bios.ramcfg_11_09);
+	ram_wr32(fuc, 0x10f6bc, 0x01010101 * next->bios.ramcfg_11_09);
 
-	data = nv_ro08(bios, ramcfg + 0x04);
-	if (!(nv_ro08(bios, ramcfg + 0x07) & 0x08)) {
-		ram_wr32(fuc, 0x10f698, 0x01010101 * data);
-		ram_wr32(fuc, 0x10f69c, 0x01010101 * data);
+	if (!next->bios.ramcfg_11_07_08 && !next->bios.ramcfg_11_07_04) {
+		ram_wr32(fuc, 0x10f698, 0x01010101 * next->bios.ramcfg_11_04);
+		ram_wr32(fuc, 0x10f69c, 0x01010101 * next->bios.ramcfg_11_04);
+	} else
+	if (!next->bios.ramcfg_11_07_08) {
+		ram_wr32(fuc, 0x10f698, 0x00000000);
+		ram_wr32(fuc, 0x10f69c, 0x00000000);
 	}
 
 	if (ram->mode != 2) {
-		u32 temp = ram_rd32(fuc, 0x10f694) & ~0xff00ff00;
-		ram_wr32(fuc, 0x10f694, temp | (0x01000100 * data));
+		u32 data = 0x01000100 * next->bios.ramcfg_11_04;
+		ram_nuke(fuc, 0x10f694);
+		ram_mask(fuc, 0x10f694, 0xff00ff00, data);
 	}
 
-	if (ram->mode == 2 && (nv_ro08(bios, ramcfg + 0x08) & 0x10))
+	if (ram->mode == 2 && (next->bios.ramcfg_11_08_10))
 		data = 0x00000080;
 	else
 		data = 0x00000000;
@@ -386,19 +425,19 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 
 	mask = 0x00070000;
 	data = 0x00000000;
-	if (!(nv_ro08(bios, ramcfg + 0x02) & 0x80))
+	if (!(next->bios.ramcfg_11_02_80))
 		data |= 0x03000000;
-	if (!(nv_ro08(bios, ramcfg + 0x02) & 0x40))
+	if (!(next->bios.ramcfg_11_02_40))
 		data |= 0x00002000;
-	if (!(nv_ro08(bios, ramcfg + 0x07) & 0x10))
+	if (!(next->bios.ramcfg_11_07_10))
 		data |= 0x00004000;
-	if (!(nv_ro08(bios, ramcfg + 0x07) & 0x08))
+	if (!(next->bios.ramcfg_11_07_08))
 		data |= 0x00000003;
 	else
 		data |= 0x74000000;
 	ram_mask(fuc, 0x10f824, mask, data);
 
-	if (nv_ro08(bios, ramcfg + 0x01) & 0x08)
+	if (next->bios.ramcfg_11_01_08)
 		data = 0x00000000;
 	else
 		data = 0x00001000;
@@ -409,61 +448,90 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 		ram_mask(fuc, 0x10f670, 0x80000000, 0x00000000);
 	}
 
-	if (nv_ro08(bios, ramcfg + 0x08) & 0x01)
+	if (next->bios.ramcfg_11_08_01)
 		data = 0x00100000;
 	else
 		data = 0x00000000;
 	ram_mask(fuc, 0x10f82c, 0x00100000, data);
 
 	data = 0x00000000;
-	if (nv_ro08(bios, ramcfg + 0x08) & 0x08)
+	if (next->bios.ramcfg_11_08_08)
 		data |= 0x00002000;
-	if (nv_ro08(bios, ramcfg + 0x08) & 0x04)
+	if (next->bios.ramcfg_11_08_04)
 		data |= 0x00001000;
-	if (nv_ro08(bios, ramcfg + 0x08) & 0x02)
+	if (next->bios.ramcfg_11_08_02)
 		data |= 0x00004000;
 	ram_mask(fuc, 0x10f830, 0x00007000, data);
 
 	/* PFB timing */
-	ram_mask(fuc, 0x10f248, 0xffffffff, nv_ro32(bios, timing + 0x28));
-	ram_mask(fuc, 0x10f290, 0xffffffff, nv_ro32(bios, timing + 0x00));
-	ram_mask(fuc, 0x10f294, 0xffffffff, nv_ro32(bios, timing + 0x04));
-	ram_mask(fuc, 0x10f298, 0xffffffff, nv_ro32(bios, timing + 0x08));
-	ram_mask(fuc, 0x10f29c, 0xffffffff, nv_ro32(bios, timing + 0x0c));
-	ram_mask(fuc, 0x10f2a0, 0xffffffff, nv_ro32(bios, timing + 0x10));
-	ram_mask(fuc, 0x10f2a4, 0xffffffff, nv_ro32(bios, timing + 0x14));
-	ram_mask(fuc, 0x10f2a8, 0xffffffff, nv_ro32(bios, timing + 0x18));
-	ram_mask(fuc, 0x10f2ac, 0xffffffff, nv_ro32(bios, timing + 0x1c));
-	ram_mask(fuc, 0x10f2cc, 0xffffffff, nv_ro32(bios, timing + 0x20));
-	ram_mask(fuc, 0x10f2e8, 0xffffffff, nv_ro32(bios, timing + 0x24));
-
-	data = (nv_ro08(bios, ramcfg + 0x02) & 0x03) << 8;
-	if (nv_ro08(bios, ramcfg + 0x01) & 0x10)
-		data |= 0x70000000;
-	ram_mask(fuc, 0x10f604, 0x70000300, data);
-
-	data = (nv_ro08(bios, timing + 0x30) & 0x07) << 28;
-	if (nv_ro08(bios, ramcfg + 0x01) & 0x01)
-		data |= 0x00000100;
-	ram_mask(fuc, 0x10f614, 0x70000000, data);
-
-	data = (nv_ro08(bios, timing + 0x30) & 0x07) << 28;
-	if (nv_ro08(bios, ramcfg + 0x01) & 0x02)
-		data |= 0x00000100;
-	ram_mask(fuc, 0x10f610, 0x70000000, data);
+	ram_mask(fuc, 0x10f248, 0xffffffff, next->bios.timing[10]);
+	ram_mask(fuc, 0x10f290, 0xffffffff, next->bios.timing[0]);
+	ram_mask(fuc, 0x10f294, 0xffffffff, next->bios.timing[1]);
+	ram_mask(fuc, 0x10f298, 0xffffffff, next->bios.timing[2]);
+	ram_mask(fuc, 0x10f29c, 0xffffffff, next->bios.timing[3]);
+	ram_mask(fuc, 0x10f2a0, 0xffffffff, next->bios.timing[4]);
+	ram_mask(fuc, 0x10f2a4, 0xffffffff, next->bios.timing[5]);
+	ram_mask(fuc, 0x10f2a8, 0xffffffff, next->bios.timing[6]);
+	ram_mask(fuc, 0x10f2ac, 0xffffffff, next->bios.timing[7]);
+	ram_mask(fuc, 0x10f2cc, 0xffffffff, next->bios.timing[8]);
+	ram_mask(fuc, 0x10f2e8, 0xffffffff, next->bios.timing[9]);
+
+	data = mask = 0x00000000;
+	if (NOTE00(ramcfg_08_20)) {
+		if (next->bios.ramcfg_11_08_20)
+			data |= 0x01000000;
+		mask |= 0x01000000;
+	}
+	ram_mask(fuc, 0x10f200, mask, data);
+
+	data = mask = 0x00000000;
+	if (NOTE00(ramcfg_02_03 != 0)) {
+		data |= (next->bios.ramcfg_11_02_03) << 8;
+		mask |= 0x00000300;
+	}
+	if (NOTE00(ramcfg_01_10)) {
+		if (next->bios.ramcfg_11_01_10)
+			data |= 0x70000000;
+		mask |= 0x70000000;
+	}
+	ram_mask(fuc, 0x10f604, mask, data);
+
+	data = mask = 0x00000000;
+	if (NOTE00(timing_30_07 != 0)) {
+		data |= (next->bios.timing_20_30_07) << 28;
+		mask |= 0x70000000;
+	}
+	if (NOTE00(ramcfg_01_01)) {
+		if (next->bios.ramcfg_11_01_01)
+			data |= 0x00000100;
+		mask |= 0x00000100;
+	}
+	ram_mask(fuc, 0x10f614, mask, data);
+
+	data = mask = 0x00000000;
+	if (NOTE00(timing_30_07 != 0)) {
+		data |= (next->bios.timing_20_30_07) << 28;
+		mask |= 0x70000000;
+	}
+	if (NOTE00(ramcfg_01_02)) {
+		if (next->bios.ramcfg_11_01_02)
+			data |= 0x00000100;
+		mask |= 0x00000100;
+	}
+	ram_mask(fuc, 0x10f610, mask, data);
 
 	mask = 0x33f00000;
 	data = 0x00000000;
-	if (!(nv_ro08(bios, ramcfg + 0x01) & 0x04))
+	if (!(next->bios.ramcfg_11_01_04))
 		data |= 0x20200000;
-	if (!(nv_ro08(bios, ramcfg + 0x07) & 0x80))
+	if (!(next->bios.ramcfg_11_07_80))
 		data |= 0x12800000;
 	/*XXX: see note above about there probably being some condition
 	 *     for the 10f824 stuff that uses ramcfg 3...
 	 */
-	if ( (nv_ro08(bios, ramcfg + 0x03) & 0xf0)) {
-		if (nv_ro08(bios, rammap + 0x08) & 0x0c) {
-			if (!(nv_ro08(bios, ramcfg + 0x07) & 0x80))
+	if ( (next->bios.ramcfg_11_03_f0)) {
+		if (next->bios.rammap_11_08_0c) {
+			if (!(next->bios.ramcfg_11_07_80))
 				mask |= 0x00000020;
 			else
 				data |= 0x00000020;
@@ -476,49 +544,53 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 
 	ram_mask(fuc, 0x10f808, mask, data);
 
-	data = nv_ro08(bios, ramcfg + 0x03) & 0x0f;
-	ram_wr32(fuc, 0x10f870, 0x11111111 * data);
+	ram_wr32(fuc, 0x10f870, 0x11111111 * next->bios.ramcfg_11_03_0f);
 
-	data = nv_ro08(bios, ramcfg + 0x02) & 0x03;
-	if (nv_ro08(bios, ramcfg + 0x01) & 0x10)
-		data |= 0x00000004;
-	if ((nv_rd32(bios, 0x100770) & 0x00000004) != (data & 0x00000004)) {
-		ram_wr32(fuc, 0x10f750, 0x04000009);
+	data = mask = 0x00000000;
+	if (NOTE00(ramcfg_02_03 != 0)) {
+		data |= next->bios.ramcfg_11_02_03;
+		mask |= 0x00000003;
+	}
+	if (NOTE00(ramcfg_01_10)) {
+		if (next->bios.ramcfg_11_01_10)
+			data |= 0x00000004;
+		mask |= 0x00000004;
+	}
+
+	if ((ram_mask(fuc, 0x100770, mask, data) & mask & 4) != (data & 4)) {
+		ram_mask(fuc, 0x100750, 0x00000008, 0x00000008);
 		ram_wr32(fuc, 0x100710, 0x00000000);
 		ram_wait(fuc, 0x100710, 0x80000000, 0x80000000, 200000);
 	}
-	ram_mask(fuc, 0x100770, 0x00000007, data);
 
-	data = (nv_ro08(bios, timing + 0x30) & 0x07) << 8;
-	if (nv_ro08(bios, ramcfg + 0x01) & 0x01)
+	data = (next->bios.timing_20_30_07) << 8;
+	if (next->bios.ramcfg_11_01_01)
 		data |= 0x80000000;
 	ram_mask(fuc, 0x100778, 0x00000700, data);
 
-	data = nv_ro16(bios, timing + 0x2c);
-	ram_mask(fuc, 0x10f250, 0x000003f0, (data & 0x003f) <<  4);
-	ram_mask(fuc, 0x10f24c, 0x7f000000, (data & 0x1fc0) << 18);
-
-	data = nv_ro08(bios, timing + 0x30);
-	ram_mask(fuc, 0x10f224, 0x001f0000, (data & 0xf8) << 13);
+	ram_mask(fuc, 0x10f250, 0x000003f0, next->bios.timing_20_2c_003f << 4);
+	data = (next->bios.timing[10] & 0x7f000000) >> 24;
+	if (data < next->bios.timing_20_2c_1fc0)
+		data = next->bios.timing_20_2c_1fc0;
+	ram_mask(fuc, 0x10f24c, 0x7f000000, data << 24);
+	ram_mask(fuc, 0x10f224, 0x001f0000, next->bios.timing_20_30_f8 << 16);
 
-	data = nv_ro16(bios, timing + 0x31);
-	ram_mask(fuc, 0x10fec4, 0x041e0f07, (data & 0x0800) << 15 |
-					    (data & 0x0780) << 10 |
-					    (data & 0x0078) <<  5 |
-					    (data & 0x0007));
-	ram_mask(fuc, 0x10fec8, 0x00000027, (data & 0x8000) >> 10 |
-					    (data & 0x7000) >> 12);
+	ram_mask(fuc, 0x10fec4, 0x041e0f07, next->bios.timing_20_31_0800 << 26 |
+					    next->bios.timing_20_31_0780 << 17 |
+					    next->bios.timing_20_31_0078 << 8 |
+					    next->bios.timing_20_31_0007);
+	ram_mask(fuc, 0x10fec8, 0x00000027, next->bios.timing_20_31_8000 << 5 |
+					    next->bios.timing_20_31_7000);
 
 	ram_wr32(fuc, 0x10f090, 0x4000007e);
-	ram_nsec(fuc, 1000);
+	ram_nsec(fuc, 2000);
 	ram_wr32(fuc, 0x10f314, 0x00000001); /* PRECHARGE */
 	ram_wr32(fuc, 0x10f310, 0x00000001); /* REFRESH */
-	ram_nsec(fuc, 2000);
 	ram_wr32(fuc, 0x10f210, 0x80000000); /* REFRESH_AUTO = 1 */
 
-	if ((nv_ro08(bios, ramcfg + 0x08) & 0x10) && (ram->mode == 2) /*XXX*/) {
+	if ((next->bios.ramcfg_11_08_10) && (ram->mode == 2) /*XXX*/) {
 		u32 temp = ram_mask(fuc, 0x10f294, 0xff000000, 0x24000000);
-		train(fuc, 0xa4010000); /*XXX*/
+		nve0_ram_train(fuc, 0xbc0e0000, 0xa4010000); /*XXX*/
 		ram_nsec(fuc, 1000);
 		ram_wr32(fuc, 0x10f294, temp);
 	}
@@ -528,7 +600,7 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 	ram_mask(fuc, mr[8], 0xfff, ram->base.mr[8]);
 	ram_nsec(fuc, 1000);
 	ram_mask(fuc, mr[1], 0xfff, ram->base.mr[1]);
-	ram_mask(fuc, mr[5], 0xfff, ram->base.mr[5]);
+	ram_mask(fuc, mr[5], 0xfff, ram->base.mr[5] & ~0x004); /* LP3 later */
 	ram_mask(fuc, mr[6], 0xfff, ram->base.mr[6]);
 	ram_mask(fuc, mr[7], 0xfff, ram->base.mr[7]);
 
@@ -544,12 +616,13 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 	ram_wr32(fuc, 0x10f318, 0x00000001); /* NOP? */
 	ram_mask(fuc, 0x10f200, 0x80000000, 0x00000000);
 	ram_nsec(fuc, 1000);
+	ram_nuts(ram, 0x10f200, 0x18808800, 0x00000000, 0x18808800);
 
 	data  = ram_rd32(fuc, 0x10f978);
 	data &= ~0x00046144;
 	data |=  0x0000000b;
-	if (!(nv_ro08(bios, ramcfg + 0x07) & 0x08)) {
-		if (!(nv_ro08(bios, ramcfg + 0x07) & 0x04))
+	if (!(next->bios.ramcfg_11_07_08)) {
+		if (!(next->bios.ramcfg_11_07_04))
 			data |= 0x0000200c;
 		else
 			data |= 0x00000000;
@@ -563,44 +636,43 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 		ram_wr32(fuc, 0x10f830, data);
 	}
 
-	if (!(nv_ro08(bios, ramcfg + 0x07) & 0x08)) {
+	if (!(next->bios.ramcfg_11_07_08)) {
 		data = 0x88020000;
-		if ( (nv_ro08(bios, ramcfg + 0x07) & 0x04))
+		if ( (next->bios.ramcfg_11_07_04))
 			data |= 0x10000000;
-		if (!(nv_ro08(bios, rammap + 0x08) & 0x10))
+		if (!(next->bios.rammap_11_08_10))
 			data |= 0x00080000;
 	} else {
 		data = 0xa40e0000;
 	}
-	train(fuc, data);
-	ram_nsec(fuc, 1000);
+	nve0_ram_train(fuc, 0xbc0f0000, data);
+	if (1) /* XXX: not always? */
+		ram_nsec(fuc, 1000);
 
 	if (ram->mode == 2) { /*XXX*/
 		ram_mask(fuc, 0x10f800, 0x00000004, 0x00000004);
 	}
 
-	/* MR5: (re)enable LP3 if necessary
-	 * XXX: need to find the switch, keeping off for now
-	 */
-	ram_mask(fuc, mr[5], 0x00000004, 0x00000000);
+	/* LP3 */
+	if (ram_mask(fuc, mr[5], 0x004, ram->base.mr[5]) != ram->base.mr[5])
+		ram_nsec(fuc, 1000);
 
 	if (ram->mode != 2) {
 		ram_mask(fuc, 0x10f830, 0x01000000, 0x01000000);
 		ram_mask(fuc, 0x10f830, 0x01000000, 0x00000000);
 	}
 
-	if (nv_ro08(bios, ramcfg + 0x07) & 0x02) {
-		ram_mask(fuc, 0x10f910, 0x80020000, 0x01000000);
-		ram_mask(fuc, 0x10f914, 0x80020000, 0x01000000);
-	}
+	if (next->bios.ramcfg_11_07_02)
+		nve0_ram_train(fuc, 0x80020000, 0x01000000);
 
 	ram_wr32(fuc, 0x62c000, 0x0f0f0f00);
 
-	if (nv_ro08(bios, rammap + 0x08) & 0x01)
+	if (next->bios.rammap_11_08_01)
 		data = 0x00000800;
 	else
 		data = 0x00000000;
 	ram_mask(fuc, 0x10f200, 0x00000800, data);
+	ram_nuts(ram, 0x10f200, 0x18808800, data, 0x18808800);
 	return 0;
 }
 
@@ -611,17 +683,14 @@ nve0_ram_calc_gddr5(struct nouveau_fb *pfb, u32 freq)
 static int
 nve0_ram_calc_sddr3(struct nouveau_fb *pfb, u32 freq)
 {
-	struct nouveau_bios *bios = nouveau_bios(pfb);
 	struct nve0_ram *ram = (void *)pfb->ram;
 	struct nve0_ramfuc *fuc = &ram->fuc;
 	const u32 rcoef = ((  ram->P1 << 16) | (ram->N1 << 8) | ram->M1);
 	const u32 runk0 = ram->fN1 << 16;
 	const u32 runk1 = ram->fN1;
-	const u32 rammap = ram->base.rammap.data;
-	const u32 ramcfg = ram->base.ramcfg.data;
-	const u32 timing = ram->base.timing.data;
-	int vc = !(nv_ro08(bios, ramcfg + 0x02) & 0x08);
-	int mv = 1; /*XXX*/
+	struct nouveau_ram_data *next = ram->base.next;
+	int vc = !(next->bios.ramcfg_11_02_08);
+	int mv = !(next->bios.ramcfg_11_02_04);
 	u32 mask, data;
 
 	ram_mask(fuc, 0x10f808, 0x40000000, 0x40000000);
@@ -636,7 +705,7 @@ nve0_ram_calc_sddr3(struct nouveau_fb *pfb, u32 freq)
 	}
 
 	ram_mask(fuc, 0x10f200, 0x00000800, 0x00000000);
-	if ((nv_ro08(bios, ramcfg + 0x03) & 0xf0))
+	if ((next->bios.ramcfg_11_03_f0))
 		ram_mask(fuc, 0x10f808, 0x04000000, 0x04000000);
 
 	ram_wr32(fuc, 0x10f314, 0x00000001); /* PRECHARGE */
@@ -661,28 +730,28 @@ nve0_ram_calc_sddr3(struct nouveau_fb *pfb, u32 freq)
 	if (1) {
 		mask |= 0x800807e0;
 		data |= 0x800807e0;
-		switch (nv_ro08(bios, ramcfg + 0x03) & 0xc0) {
-		case 0xc0: data &= ~0x00000040; break;
-		case 0x80: data &= ~0x00000100; break;
-		case 0x40: data &= ~0x80000000; break;
-		case 0x00: data &= ~0x00000400; break;
+		switch (next->bios.ramcfg_11_03_c0) {
+		case 3: data &= ~0x00000040; break;
+		case 2: data &= ~0x00000100; break;
+		case 1: data &= ~0x80000000; break;
+		case 0: data &= ~0x00000400; break;
 		}
 
-		switch (nv_ro08(bios, ramcfg + 0x03) & 0x30) {
-		case 0x30: data &= ~0x00000020; break;
-		case 0x20: data &= ~0x00000080; break;
-		case 0x10: data &= ~0x00080000; break;
-		case 0x00: data &= ~0x00000200; break;
+		switch (next->bios.ramcfg_11_03_30) {
+		case 3: data &= ~0x00000020; break;
+		case 2: data &= ~0x00000080; break;
+		case 1: data &= ~0x00080000; break;
+		case 0: data &= ~0x00000200; break;
 		}
 	}
 
-	if (nv_ro08(bios, ramcfg + 0x02) & 0x80)
+	if (next->bios.ramcfg_11_02_80)
 		mask |= 0x03000000;
-	if (nv_ro08(bios, ramcfg + 0x02) & 0x40)
+	if (next->bios.ramcfg_11_02_40)
 		mask |= 0x00002000;
-	if (nv_ro08(bios, ramcfg + 0x07) & 0x10)
+	if (next->bios.ramcfg_11_07_10)
 		mask |= 0x00004000;
-	if (nv_ro08(bios, ramcfg + 0x07) & 0x08)
+	if (next->bios.ramcfg_11_07_08)
 		mask |= 0x00000003;
 	else
 		mask |= 0x14000000;
@@ -692,7 +761,7 @@ nve0_ram_calc_sddr3(struct nouveau_fb *pfb, u32 freq)
 
 	ram_mask(fuc, 0x1373f4, 0x00000000, 0x00010010);
 	data  = ram_rd32(fuc, 0x1373ec) & ~0x00030000;
-	data |= (nv_ro08(bios, ramcfg + 0x03) & 0x30) << 12;
+	data |= (next->bios.ramcfg_11_03_30) << 12;
 	ram_wr32(fuc, 0x1373ec, data);
 	ram_mask(fuc, 0x1373f4, 0x00000003, 0x00000000);
 	ram_mask(fuc, 0x1373f4, 0x00000010, 0x00000000);
@@ -724,68 +793,67 @@ nve0_ram_calc_sddr3(struct nouveau_fb *pfb, u32 freq)
 		}
 	}
 
-	if ( (nv_ro08(bios, ramcfg + 0x02) & 0x40) ||
-	     (nv_ro08(bios, ramcfg + 0x07) & 0x10)) {
+	if ( (next->bios.ramcfg_11_02_40) ||
+	     (next->bios.ramcfg_11_07_10)) {
 		ram_mask(fuc, 0x132040, 0x00010000, 0x00010000);
 		ram_nsec(fuc, 20000);
 	}
 
 	if (ram->mode != 2) /*XXX*/ {
-		if (nv_ro08(bios, ramcfg + 0x07) & 0x40)
+		if (next->bios.ramcfg_11_07_40)
 			ram_mask(fuc, 0x10f670, 0x80000000, 0x80000000);
 	}
 
-	data = (nv_ro08(bios, rammap + 0x11) & 0x0c) >> 2;
-	ram_wr32(fuc, 0x10f65c, 0x00000011 * data);
-	ram_wr32(fuc, 0x10f6b8, 0x01010101 * nv_ro08(bios, ramcfg + 0x09));
-	ram_wr32(fuc, 0x10f6bc, 0x01010101 * nv_ro08(bios, ramcfg + 0x09));
+	ram_wr32(fuc, 0x10f65c, 0x00000011 * next->bios.rammap_11_11_0c);
+	ram_wr32(fuc, 0x10f6b8, 0x01010101 * next->bios.ramcfg_11_09);
+	ram_wr32(fuc, 0x10f6bc, 0x01010101 * next->bios.ramcfg_11_09);
 
 	mask = 0x00010000;
 	data = 0x00000000;
-	if (!(nv_ro08(bios, ramcfg + 0x02) & 0x80))
+	if (!(next->bios.ramcfg_11_02_80))
 		data |= 0x03000000;
-	if (!(nv_ro08(bios, ramcfg + 0x02) & 0x40))
+	if (!(next->bios.ramcfg_11_02_40))
 		data |= 0x00002000;
-	if (!(nv_ro08(bios, ramcfg + 0x07) & 0x10))
+	if (!(next->bios.ramcfg_11_07_10))
 		data |= 0x00004000;
-	if (!(nv_ro08(bios, ramcfg + 0x07) & 0x08))
+	if (!(next->bios.ramcfg_11_07_08))
 		data |= 0x00000003;
 	else
 		data |= 0x14000000;
 	ram_mask(fuc, 0x10f824, mask, data);
 	ram_nsec(fuc, 1000);
 
-	if (nv_ro08(bios, ramcfg + 0x08) & 0x01)
+	if (next->bios.ramcfg_11_08_01)
 		data = 0x00100000;
 	else
 		data = 0x00000000;
 	ram_mask(fuc, 0x10f82c, 0x00100000, data);
 
 	/* PFB timing */
-	ram_mask(fuc, 0x10f248, 0xffffffff, nv_ro32(bios, timing + 0x28));
-	ram_mask(fuc, 0x10f290, 0xffffffff, nv_ro32(bios, timing + 0x00));
-	ram_mask(fuc, 0x10f294, 0xffffffff, nv_ro32(bios, timing + 0x04));
-	ram_mask(fuc, 0x10f298, 0xffffffff, nv_ro32(bios, timing + 0x08));
-	ram_mask(fuc, 0x10f29c, 0xffffffff, nv_ro32(bios, timing + 0x0c));
-	ram_mask(fuc, 0x10f2a0, 0xffffffff, nv_ro32(bios, timing + 0x10));
-	ram_mask(fuc, 0x10f2a4, 0xffffffff, nv_ro32(bios, timing + 0x14));
-	ram_mask(fuc, 0x10f2a8, 0xffffffff, nv_ro32(bios, timing + 0x18));
-	ram_mask(fuc, 0x10f2ac, 0xffffffff, nv_ro32(bios, timing + 0x1c));
-	ram_mask(fuc, 0x10f2cc, 0xffffffff, nv_ro32(bios, timing + 0x20));
-	ram_mask(fuc, 0x10f2e8, 0xffffffff, nv_ro32(bios, timing + 0x24));
+	ram_mask(fuc, 0x10f248, 0xffffffff, next->bios.timing[10]);
+	ram_mask(fuc, 0x10f290, 0xffffffff, next->bios.timing[0]);
+	ram_mask(fuc, 0x10f294, 0xffffffff, next->bios.timing[1]);
+	ram_mask(fuc, 0x10f298, 0xffffffff, next->bios.timing[2]);
+	ram_mask(fuc, 0x10f29c, 0xffffffff, next->bios.timing[3]);
+	ram_mask(fuc, 0x10f2a0, 0xffffffff, next->bios.timing[4]);
+	ram_mask(fuc, 0x10f2a4, 0xffffffff, next->bios.timing[5]);
+	ram_mask(fuc, 0x10f2a8, 0xffffffff, next->bios.timing[6]);
+	ram_mask(fuc, 0x10f2ac, 0xffffffff, next->bios.timing[7]);
+	ram_mask(fuc, 0x10f2cc, 0xffffffff, next->bios.timing[8]);
+	ram_mask(fuc, 0x10f2e8, 0xffffffff, next->bios.timing[9]);
 
 	mask = 0x33f00000;
 	data = 0x00000000;
-	if (!(nv_ro08(bios, ramcfg + 0x01) & 0x04))
+	if (!(next->bios.ramcfg_11_01_04))
 		data |= 0x20200000;
-	if (!(nv_ro08(bios, ramcfg + 0x07) & 0x80))
+	if (!(next->bios.ramcfg_11_07_80))
 		data |= 0x12800000;
 	/*XXX: see note above about there probably being some condition
 	 *     for the 10f824 stuff that uses ramcfg 3...
 	 */
-	if ( (nv_ro08(bios, ramcfg + 0x03) & 0xf0)) {
-		if (nv_ro08(bios, rammap + 0x08) & 0x0c) {
-			if (!(nv_ro08(bios, ramcfg + 0x07) & 0x80))
+	if ( (next->bios.ramcfg_11_03_f0)) {
+		if (next->bios.rammap_11_08_0c) {
+			if (!(next->bios.ramcfg_11_07_80))
 				mask |= 0x00000020;
 			else
 				data |= 0x00000020;
@@ -799,21 +867,16 @@ nve0_ram_calc_sddr3(struct nouveau_fb *pfb, u32 freq)
 
 	ram_mask(fuc, 0x10f808, mask, data);
 
-	data = nv_ro08(bios, ramcfg + 0x03) & 0x0f;
-	ram_wr32(fuc, 0x10f870, 0x11111111 * data);
+	ram_wr32(fuc, 0x10f870, 0x11111111 * next->bios.ramcfg_11_03_0f);
 
-	data = nv_ro16(bios, timing + 0x2c);
-	ram_mask(fuc, 0x10f250, 0x000003f0, (data & 0x003f) <<  4);
+	ram_mask(fuc, 0x10f250, 0x000003f0, next->bios.timing_20_2c_003f << 4);
 
-	if (((nv_ro32(bios, timing + 0x2c) & 0x00001fc0) >>  6) >
-	    ((nv_ro32(bios, timing + 0x28) & 0x7f000000) >> 24))
-		data = (nv_ro32(bios, timing + 0x2c) & 0x00001fc0) >>  6;
-	else
-		data = (nv_ro32(bios, timing + 0x28) & 0x1f000000) >> 24;
+	data = (next->bios.timing[10] & 0x7f000000) >> 24;
+	if (data < next->bios.timing_20_2c_1fc0)
+		data = next->bios.timing_20_2c_1fc0;
 	ram_mask(fuc, 0x10f24c, 0x7f000000, data << 24);
 
-	data = nv_ro08(bios, timing + 0x30);
-	ram_mask(fuc, 0x10f224, 0x001f0000, (data & 0xf8) << 13);
+	ram_mask(fuc, 0x10f224, 0x001f0000, next->bios.timing_20_30_f8);
 
 	ram_wr32(fuc, 0x10f090, 0x4000007f);
 	ram_nsec(fuc, 1000);
@@ -855,7 +918,7 @@ nve0_ram_calc_sddr3(struct nouveau_fb *pfb, u32 freq)
 
 	ram_wr32(fuc, 0x62c000, 0x0f0f0f00);
 
-	if (nv_ro08(bios, rammap + 0x08) & 0x01)
+	if (next->bios.rammap_11_08_01)
 		data = 0x00000800;
 	else
 		data = 0x00000000;
@@ -868,21 +931,18 @@ nve0_ram_calc_sddr3(struct nouveau_fb *pfb, u32 freq)
  ******************************************************************************/
 
 static int
-nve0_ram_calc(struct nouveau_fb *pfb, u32 freq)
+nve0_ram_calc_data(struct nouveau_fb *pfb, u32 freq,
+		   struct nouveau_ram_data *data)
 {
 	struct nouveau_bios *bios = nouveau_bios(pfb);
 	struct nve0_ram *ram = (void *)pfb->ram;
-	struct nve0_ramfuc *fuc = &ram->fuc;
-	struct bit_entry M;
-	int ret, refclk, strap, i;
-	u32 data;
-	u8  cnt;
+	u8 strap, cnt, len;
 
 	/* lookup memory config data relevant to the target frequency */
-	ram->base.rammap.data = nvbios_rammap_match(bios, freq / 1000,
-						   &ram->base.rammap.version,
-						   &ram->base.rammap.size, &cnt,
-						   &ram->base.ramcfg.size);
+	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");
@@ -890,24 +950,13 @@ nve0_ram_calc(struct nouveau_fb *pfb, u32 freq)
 	}
 
 	/* locate specific data set for the attached memory */
-	if (bit_entry(bios, 'M', &M) || M.version != 2 || M.length < 3) {
-		nv_error(pfb, "invalid/missing memory table\n");
-		return -EINVAL;
-	}
-
-	strap = (nv_rd32(pfb, 0x101000) & 0x0000003c) >> 2;
-	data = nv_ro16(bios, M.offset + 1);
-	if (data)
-		strap = nv_ro08(bios, data + strap);
-
-	if (strap >= cnt) {
-		nv_error(pfb, "invalid ramcfg strap\n");
-		return -EINVAL;
-	}
-
-	ram->base.ramcfg.version = ram->base.rammap.version;
-	ram->base.ramcfg.data = ram->base.rammap.data + ram->base.rammap.size +
-			       (ram->base.ramcfg.size * strap);
+	ram->base.ramcfg.data = nvbios_rammapSp(bios, ram->base.rammap.data,
+						ram->base.rammap.version,
+						ram->base.rammap.size, cnt, len,
+						nvbios_ramcfg_index(bios),
+						&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");
@@ -918,9 +967,9 @@ nve0_ram_calc(struct nouveau_fb *pfb, u32 freq)
 	strap = nv_ro08(bios, ram->base.ramcfg.data + 0x00);
 	if (strap != 0xff) {
 		ram->base.timing.data =
-			nvbios_timing_entry(bios, strap,
-					   &ram->base.timing.version,
-					   &ram->base.timing.size);
+			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) {
@@ -931,11 +980,23 @@ nve0_ram_calc(struct nouveau_fb *pfb, u32 freq)
 		ram->base.timing.data = 0;
 	}
 
+	data->freq = freq;
+	return 0;
+}
+
+static int
+nve0_ram_calc_xits(struct nouveau_fb *pfb, struct nouveau_ram_data *next)
+{
+	struct nve0_ram *ram = (void *)pfb->ram;
+	struct nve0_ramfuc *fuc = &ram->fuc;
+	int refclk, i;
+	int ret;
+
 	ret = ram_init(fuc, pfb);
 	if (ret)
 		return ret;
 
-	ram->mode = (freq > fuc->refpll.vco1.max_freq) ? 2 : 1;
+	ram->mode = (next->freq > fuc->refpll.vco1.max_freq) ? 2 : 1;
 	ram->from = ram_rd32(fuc, 0x1373f4) & 0x0000000f;
 
 	/* XXX: this is *not* what nvidia do.  on fermi nvidia generally
@@ -946,7 +1007,7 @@ nve0_ram_calc(struct nouveau_fb *pfb, u32 freq)
 	 * so far, i've seen very weird values being chosen by nvidia on
 	 * kepler boards, no idea how/why they're chosen.
 	 */
-	refclk = freq;
+	refclk = next->freq;
 	if (ram->mode == 2)
 		refclk = fuc->mempll.refclk;
 
@@ -968,7 +1029,7 @@ nve0_ram_calc(struct nouveau_fb *pfb, u32 freq)
 		fuc->mempll.min_p = 1;
 		fuc->mempll.max_p = 2;
 
-		ret = nva3_pll_calc(nv_subdev(pfb), &fuc->mempll, freq,
+		ret = nva3_pll_calc(nv_subdev(pfb), &fuc->mempll, next->freq,
 				   &ram->N2, NULL, &ram->M2, &ram->P2);
 		if (ret <= 0) {
 			nv_error(pfb, "unable to calc mempll\n");
@@ -980,17 +1041,18 @@ nve0_ram_calc(struct nouveau_fb *pfb, u32 freq)
 		if (ram_have(fuc, mr[i]))
 			ram->base.mr[i] = ram_rd32(fuc, mr[i]);
 	}
+	ram->base.freq = next->freq;
 
 	switch (ram->base.type) {
 	case NV_MEM_TYPE_DDR3:
 		ret = nouveau_sddr3_calc(&ram->base);
 		if (ret == 0)
-			ret = nve0_ram_calc_sddr3(pfb, freq);
+			ret = nve0_ram_calc_sddr3(pfb, next->freq);
 		break;
 	case NV_MEM_TYPE_GDDR5:
-		ret = nouveau_gddr5_calc(&ram->base);
+		ret = nouveau_gddr5_calc(&ram->base, ram->pnuts != 0);
 		if (ret == 0)
-			ret = nve0_ram_calc_gddr5(pfb, freq);
+			ret = nve0_ram_calc_gddr5(pfb, next->freq);
 		break;
 	default:
 		ret = -ENOSYS;
@@ -1001,13 +1063,55 @@ nve0_ram_calc(struct nouveau_fb *pfb, u32 freq)
 }
 
 static int
+nve0_ram_calc(struct nouveau_fb *pfb, u32 freq)
+{
+	struct nouveau_clock *clk = nouveau_clock(pfb);
+	struct nve0_ram *ram = (void *)pfb->ram;
+	struct nouveau_ram_data *xits = &ram->base.xition;
+	struct nouveau_ram_data *copy;
+	int ret;
+
+	if (ram->base.next == NULL) {
+		ret = nve0_ram_calc_data(pfb, clk->read(clk, nv_clk_src_mem),
+					&ram->base.former);
+		if (ret)
+			return ret;
+
+		ret = nve0_ram_calc_data(pfb, freq, &ram->base.target);
+		if (ret)
+			return ret;
+
+		if (ram->base.target.freq < ram->base.former.freq) {
+			*xits = ram->base.target;
+			copy = &ram->base.former;
+		} else {
+			*xits = ram->base.former;
+			copy = &ram->base.target;
+		}
+
+		xits->bios.ramcfg_11_02_04 = copy->bios.ramcfg_11_02_04;
+		xits->bios.ramcfg_11_02_03 = copy->bios.ramcfg_11_02_03;
+		xits->bios.timing_20_30_07 = copy->bios.timing_20_30_07;
+
+		ram->base.next = &ram->base.target;
+		if (memcmp(xits, &ram->base.former, sizeof(xits->bios)))
+			ram->base.next = &ram->base.xition;
+	} else {
+		BUG_ON(ram->base.next != &ram->base.xition);
+		ram->base.next = &ram->base.target;
+	}
+
+	return nve0_ram_calc_xits(pfb, ram->base.next);
+}
+
+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", false));
-	return 0;
+	return (ram->base.next == &ram->base.xition);
 }
 
 static void
@@ -1015,6 +1119,7 @@ nve0_ram_tidy(struct nouveau_fb *pfb)
 {
 	struct nve0_ram *ram = (void *)pfb->ram;
 	struct nve0_ramfuc *fuc = &ram->fuc;
+	ram->base.next = NULL;
 	ram_exec(fuc, false);
 }
 
@@ -1055,7 +1160,7 @@ nve0_ram_init(struct nouveau_object *object)
 	 * binary driver skips the one that's already been setup by
 	 * the init tables.
 	 */
-	data = nvbios_rammap_table(bios, &ver, &hdr, &cnt, &len, &snr, &ssz);
+	data = nvbios_rammapTe(bios, &ver, &hdr, &cnt, &len, &snr, &ssz);
 	if (!data || hdr < 0x15)
 		return -EINVAL;
 
@@ -1073,6 +1178,7 @@ nve0_ram_init(struct nouveau_object *object)
 		data += 4;
 	}
 	nv_wr32(pfb, 0x10f65c, save);
+	nv_mask(pfb, 0x10f584, 0x11000000, 0x00000000);
 
 	switch (ram->base.type) {
 	case NV_MEM_TYPE_GDDR5:
@@ -1117,7 +1223,8 @@ nve0_ram_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	struct nouveau_gpio *gpio = nouveau_gpio(pfb);
 	struct dcb_gpio_func func;
 	struct nve0_ram *ram;
-	int ret;
+	int ret, i;
+	u32 tmp;
 
 	ret = nvc0_ram_create(parent, engine, oclass, &ram);
 	*pobject = nv_object(ram);
@@ -1136,6 +1243,25 @@ nve0_ram_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 		break;
 	}
 
+	/* calculate a mask of differently configured memory partitions,
+	 * because, of course reclocking wasn't complicated enough
+	 * already without having to treat some of them differently to
+	 * the others....
+	 */
+	ram->parts = nv_rd32(pfb, 0x022438);
+	ram->pmask = nv_rd32(pfb, 0x022554);
+	ram->pnuts = 0;
+	for (i = 0, tmp = 0; i < ram->parts; i++) {
+		if (!(ram->pmask & (1 << i))) {
+			u32 cfg1 = nv_rd32(pfb, 0x110204 + (i * 0x1000));
+			if (tmp && tmp != cfg1) {
+				ram->pnuts |= (1 << i);
+				continue;
+			}
+			tmp = cfg1;
+		}
+	}
+
 	// parse bios data for both pll's
 	ret = nvbios_pll_parse(bios, 0x0c, &ram->fuc.refpll);
 	if (ret) {
@@ -1248,7 +1374,7 @@ nve0_ram_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	ram->fuc.r_0x10f65c = ramfuc_reg(0x10f65c);
 	ram->fuc.r_0x10f6bc = ramfuc_reg(0x10f6bc);
 	ram->fuc.r_0x100710 = ramfuc_reg(0x100710);
-	ram->fuc.r_0x10f750 = ramfuc_reg(0x10f750);
+	ram->fuc.r_0x100750 = ramfuc_reg(0x100750);
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/nouveau/core/subdev/instmem/base.c b/drivers/gpu/drm/nouveau/core/subdev/instmem/base.c
index 6565f3dbbe04..14706d9842ca 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/instmem/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/instmem/base.c
@@ -22,7 +22,24 @@
  * Authors: Ben Skeggs
  */
 
-#include <subdev/instmem.h>
+#include "priv.h"
+
+/******************************************************************************
+ * instmem object base implementation
+ *****************************************************************************/
+
+void
+_nouveau_instobj_dtor(struct nouveau_object *object)
+{
+	struct nouveau_instmem *imem = (void *)object->engine;
+	struct nouveau_instobj *iobj = (void *)object;
+
+	mutex_lock(&nv_subdev(imem)->mutex);
+	list_del(&iobj->head);
+	mutex_unlock(&nv_subdev(imem)->mutex);
+
+	return nouveau_object_destroy(&iobj->base);
+}
 
 int
 nouveau_instobj_create_(struct nouveau_object *parent,
@@ -46,73 +63,26 @@ nouveau_instobj_create_(struct nouveau_object *parent,
 	return 0;
 }
 
-void
-nouveau_instobj_destroy(struct nouveau_instobj *iobj)
-{
-	struct nouveau_subdev *subdev = nv_subdev(iobj->base.engine);
+/******************************************************************************
+ * instmem subdev base implementation
+ *****************************************************************************/
 
-	mutex_lock(&subdev->mutex);
-	list_del(&iobj->head);
-	mutex_unlock(&subdev->mutex);
-
-	return nouveau_object_destroy(&iobj->base);
-}
-
-void
-_nouveau_instobj_dtor(struct nouveau_object *object)
+static int
+nouveau_instmem_alloc(struct nouveau_instmem *imem,
+		      struct nouveau_object *parent, u32 size, u32 align,
+		      struct nouveau_object **pobject)
 {
-	struct nouveau_instobj *iobj = (void *)object;
-	return nouveau_instobj_destroy(iobj);
+	struct nouveau_object *engine = nv_object(imem);
+	struct nouveau_instmem_impl *impl = (void *)engine->oclass;
+	struct nouveau_instobj_args args = { .size = size, .align = align };
+	return nouveau_object_ctor(parent, engine, impl->instobj, &args,
+				   sizeof(args), pobject);
 }
 
 int
-nouveau_instmem_create_(struct nouveau_object *parent,
-			struct nouveau_object *engine,
-			struct nouveau_oclass *oclass,
-			int length, void **pobject)
-{
-	struct nouveau_instmem *imem;
-	int ret;
-
-	ret = nouveau_subdev_create_(parent, engine, oclass, 0,
-				     "INSTMEM", "instmem", length, pobject);
-	imem = *pobject;
-	if (ret)
-		return ret;
-
-	INIT_LIST_HEAD(&imem->list);
-	return 0;
-}
-
-int
-nouveau_instmem_init(struct nouveau_instmem *imem)
-{
-	struct nouveau_instobj *iobj;
-	int ret, i;
-
-	ret = nouveau_subdev_init(&imem->base);
-	if (ret)
-		return ret;
-
-	mutex_lock(&imem->base.mutex);
-
-	list_for_each_entry(iobj, &imem->list, head) {
-		if (iobj->suspend) {
-			for (i = 0; i < iobj->size; i += 4)
-				nv_wo32(iobj, i, iobj->suspend[i / 4]);
-			vfree(iobj->suspend);
-			iobj->suspend = NULL;
-		}
-	}
-
-	mutex_unlock(&imem->base.mutex);
-
-	return 0;
-}
-
-int
-nouveau_instmem_fini(struct nouveau_instmem *imem, bool suspend)
+_nouveau_instmem_fini(struct nouveau_object *object, bool suspend)
 {
+	struct nouveau_instmem *imem = (void *)object;
 	struct nouveau_instobj *iobj;
 	int i, ret = 0;
 
@@ -143,12 +113,45 @@ int
 _nouveau_instmem_init(struct nouveau_object *object)
 {
 	struct nouveau_instmem *imem = (void *)object;
-	return nouveau_instmem_init(imem);
+	struct nouveau_instobj *iobj;
+	int ret, i;
+
+	ret = nouveau_subdev_init(&imem->base);
+	if (ret)
+		return ret;
+
+	mutex_lock(&imem->base.mutex);
+
+	list_for_each_entry(iobj, &imem->list, head) {
+		if (iobj->suspend) {
+			for (i = 0; i < iobj->size; i += 4)
+				nv_wo32(iobj, i, iobj->suspend[i / 4]);
+			vfree(iobj->suspend);
+			iobj->suspend = NULL;
+		}
+	}
+
+	mutex_unlock(&imem->base.mutex);
+
+	return 0;
 }
 
 int
-_nouveau_instmem_fini(struct nouveau_object *object, bool suspend)
+nouveau_instmem_create_(struct nouveau_object *parent,
+			struct nouveau_object *engine,
+			struct nouveau_oclass *oclass,
+			int length, void **pobject)
 {
-	struct nouveau_instmem *imem = (void *)object;
-	return nouveau_instmem_fini(imem, suspend);
+	struct nouveau_instmem *imem;
+	int ret;
+
+	ret = nouveau_subdev_create_(parent, engine, oclass, 0,
+				     "INSTMEM", "instmem", length, pobject);
+	imem = *pobject;
+	if (ret)
+		return ret;
+
+	INIT_LIST_HEAD(&imem->list);
+	imem->alloc = nouveau_instmem_alloc;
+	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/core/subdev/instmem/nv04.c b/drivers/gpu/drm/nouveau/core/subdev/instmem/nv04.c
index 795393d7b2f5..7b64befee48f 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/instmem/nv04.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/instmem/nv04.c
@@ -22,10 +22,35 @@
  * Authors: Ben Skeggs
  */
 
-#include <subdev/fb.h>
-
 #include "nv04.h"
 
+/******************************************************************************
+ * instmem object implementation
+ *****************************************************************************/
+
+static u32
+nv04_instobj_rd32(struct nouveau_object *object, u64 addr)
+{
+	struct nv04_instobj_priv *node = (void *)object;
+	return nv_ro32(object->engine, node->mem->offset + addr);
+}
+
+static void
+nv04_instobj_wr32(struct nouveau_object *object, u64 addr, u32 data)
+{
+	struct nv04_instobj_priv *node = (void *)object;
+	nv_wo32(object->engine, node->mem->offset + addr, data);
+}
+
+static void
+nv04_instobj_dtor(struct nouveau_object *object)
+{
+	struct nv04_instmem_priv *priv = (void *)object->engine;
+	struct nv04_instobj_priv *node = (void *)object;
+	nouveau_mm_free(&priv->heap, &node->mem);
+	nouveau_instobj_destroy(&node->base);
+}
+
 static int
 nv04_instobj_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 		  struct nouveau_oclass *oclass, void *data, u32 size,
@@ -33,18 +58,19 @@ nv04_instobj_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 {
 	struct nv04_instmem_priv *priv = (void *)engine;
 	struct nv04_instobj_priv *node;
-	int ret, align;
+	struct nouveau_instobj_args *args = data;
+	int ret;
 
-	align = (unsigned long)data;
-	if (!align)
-		align = 1;
+	if (!args->align)
+		args->align = 1;
 
 	ret = nouveau_instobj_create(parent, engine, oclass, &node);
 	*pobject = nv_object(node);
 	if (ret)
 		return ret;
 
-	ret = nouveau_mm_head(&priv->heap, 1, size, size, align, &node->mem);
+	ret = nouveau_mm_head(&priv->heap, 1, args->size, args->size,
+			      args->align, &node->mem);
 	if (ret)
 		return ret;
 
@@ -53,32 +79,9 @@ nv04_instobj_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	return 0;
 }
 
-static void
-nv04_instobj_dtor(struct nouveau_object *object)
-{
-	struct nv04_instmem_priv *priv = (void *)object->engine;
-	struct nv04_instobj_priv *node = (void *)object;
-	nouveau_mm_free(&priv->heap, &node->mem);
-	nouveau_instobj_destroy(&node->base);
-}
-
-static u32
-nv04_instobj_rd32(struct nouveau_object *object, u64 addr)
-{
-	struct nv04_instobj_priv *node = (void *)object;
-	return nv_ro32(object->engine, node->mem->offset + addr);
-}
-
-static void
-nv04_instobj_wr32(struct nouveau_object *object, u64 addr, u32 data)
-{
-	struct nv04_instobj_priv *node = (void *)object;
-	nv_wo32(object->engine, node->mem->offset + addr, data);
-}
-
-static struct nouveau_oclass
+struct nouveau_instobj_impl
 nv04_instobj_oclass = {
-	.ofuncs = &(struct nouveau_ofuncs) {
+	.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nv04_instobj_ctor,
 		.dtor = nv04_instobj_dtor,
 		.init = _nouveau_instobj_init,
@@ -88,19 +91,34 @@ nv04_instobj_oclass = {
 	},
 };
 
-int
-nv04_instmem_alloc(struct nouveau_instmem *imem, struct nouveau_object *parent,
-		   u32 size, u32 align, struct nouveau_object **pobject)
+/******************************************************************************
+ * instmem subdev implementation
+ *****************************************************************************/
+
+static u32
+nv04_instmem_rd32(struct nouveau_object *object, u64 addr)
 {
-	struct nouveau_object *engine = nv_object(imem);
-	int ret;
+	return nv_rd32(object, 0x700000 + addr);
+}
 
-	ret = nouveau_object_ctor(parent, engine, &nv04_instobj_oclass,
-				  (void *)(unsigned long)align, size, pobject);
-	if (ret)
-		return ret;
+static void
+nv04_instmem_wr32(struct nouveau_object *object, u64 addr, u32 data)
+{
+	return nv_wr32(object, 0x700000 + addr, data);
+}
 
-	return 0;
+void
+nv04_instmem_dtor(struct nouveau_object *object)
+{
+	struct nv04_instmem_priv *priv = (void *)object;
+	nouveau_gpuobj_ref(NULL, &priv->ramfc);
+	nouveau_gpuobj_ref(NULL, &priv->ramro);
+	nouveau_ramht_ref(NULL, &priv->ramht);
+	nouveau_gpuobj_ref(NULL, &priv->vbios);
+	nouveau_mm_fini(&priv->heap);
+	if (priv->iomem)
+		iounmap(priv->iomem);
+	nouveau_instmem_destroy(&priv->base);
 }
 
 static int
@@ -118,7 +136,6 @@ nv04_instmem_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 
 	/* PRAMIN aperture maps over the end of VRAM, reserve it */
 	priv->base.reserved = 512 * 1024;
-	priv->base.alloc    = nv04_instmem_alloc;
 
 	ret = nouveau_mm_init(&priv->heap, 0, priv->base.reserved, 1);
 	if (ret)
@@ -150,36 +167,10 @@ nv04_instmem_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	return 0;
 }
 
-void
-nv04_instmem_dtor(struct nouveau_object *object)
-{
-	struct nv04_instmem_priv *priv = (void *)object;
-	nouveau_gpuobj_ref(NULL, &priv->ramfc);
-	nouveau_gpuobj_ref(NULL, &priv->ramro);
-	nouveau_ramht_ref(NULL, &priv->ramht);
-	nouveau_gpuobj_ref(NULL, &priv->vbios);
-	nouveau_mm_fini(&priv->heap);
-	if (priv->iomem)
-		iounmap(priv->iomem);
-	nouveau_instmem_destroy(&priv->base);
-}
-
-static u32
-nv04_instmem_rd32(struct nouveau_object *object, u64 addr)
-{
-	return nv_rd32(object, 0x700000 + addr);
-}
-
-static void
-nv04_instmem_wr32(struct nouveau_object *object, u64 addr, u32 data)
-{
-	return nv_wr32(object, 0x700000 + addr, data);
-}
-
-struct nouveau_oclass
-nv04_instmem_oclass = {
-	.handle = NV_SUBDEV(INSTMEM, 0x04),
-	.ofuncs = &(struct nouveau_ofuncs) {
+struct nouveau_oclass *
+nv04_instmem_oclass = &(struct nouveau_instmem_impl) {
+	.base.handle = NV_SUBDEV(INSTMEM, 0x04),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nv04_instmem_ctor,
 		.dtor = nv04_instmem_dtor,
 		.init = _nouveau_instmem_init,
@@ -187,4 +178,5 @@ nv04_instmem_oclass = {
 		.rd32 = nv04_instmem_rd32,
 		.wr32 = nv04_instmem_wr32,
 	},
-};
+	.instobj = &nv04_instobj_oclass.base,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/instmem/nv04.h b/drivers/gpu/drm/nouveau/core/subdev/instmem/nv04.h
index b15b61310236..095fbc6fc099 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/instmem/nv04.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/instmem/nv04.h
@@ -5,7 +5,9 @@
 #include <core/ramht.h>
 #include <core/mm.h>
 
-#include <subdev/instmem.h>
+#include "priv.h"
+
+extern struct nouveau_instobj_impl nv04_instobj_oclass;
 
 struct nv04_instmem_priv {
 	struct nouveau_instmem base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/instmem/nv40.c b/drivers/gpu/drm/nouveau/core/subdev/instmem/nv40.c
index b10a143787a7..ec0b9661d614 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/instmem/nv40.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/instmem/nv40.c
@@ -26,6 +26,24 @@
 
 #include "nv04.h"
 
+/******************************************************************************
+ * instmem subdev implementation
+ *****************************************************************************/
+
+static u32
+nv40_instmem_rd32(struct nouveau_object *object, u64 addr)
+{
+	struct nv04_instmem_priv *priv = (void *)object;
+	return ioread32_native(priv->iomem + addr);
+}
+
+static void
+nv40_instmem_wr32(struct nouveau_object *object, u64 addr, u32 data)
+{
+	struct nv04_instmem_priv *priv = (void *)object;
+	iowrite32_native(data, priv->iomem + addr);
+}
+
 static int
 nv40_instmem_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 		  struct nouveau_oclass *oclass, void *data, u32 size,
@@ -69,7 +87,6 @@ nv40_instmem_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	priv->base.reserved += 512 * 1024;	/* object storage */
 
 	priv->base.reserved = round_up(priv->base.reserved, 4096);
-	priv->base.alloc    = nv04_instmem_alloc;
 
 	ret = nouveau_mm_init(&priv->heap, 0, priv->base.reserved, 1);
 	if (ret)
@@ -106,24 +123,10 @@ nv40_instmem_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 	return 0;
 }
 
-static u32
-nv40_instmem_rd32(struct nouveau_object *object, u64 addr)
-{
-	struct nv04_instmem_priv *priv = (void *)object;
-	return ioread32_native(priv->iomem + addr);
-}
-
-static void
-nv40_instmem_wr32(struct nouveau_object *object, u64 addr, u32 data)
-{
-	struct nv04_instmem_priv *priv = (void *)object;
-	iowrite32_native(data, priv->iomem + addr);
-}
-
-struct nouveau_oclass
-nv40_instmem_oclass = {
-	.handle = NV_SUBDEV(INSTMEM, 0x40),
-	.ofuncs = &(struct nouveau_ofuncs) {
+struct nouveau_oclass *
+nv40_instmem_oclass = &(struct nouveau_instmem_impl) {
+	.base.handle = NV_SUBDEV(INSTMEM, 0x40),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nv40_instmem_ctor,
 		.dtor = nv04_instmem_dtor,
 		.init = _nouveau_instmem_init,
@@ -131,4 +134,5 @@ nv40_instmem_oclass = {
 		.rd32 = nv40_instmem_rd32,
 		.wr32 = nv40_instmem_wr32,
 	},
-};
+	.instobj = &nv04_instobj_oclass.base,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/instmem/nv50.c b/drivers/gpu/drm/nouveau/core/subdev/instmem/nv50.c
index 97bc5dff93e7..7cb3b098a08d 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/instmem/nv50.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/instmem/nv50.c
@@ -22,11 +22,11 @@
  * Authors: Ben Skeggs
  */
 
-#include <subdev/instmem.h>
 #include <subdev/fb.h>
-
 #include <core/mm.h>
 
+#include "priv.h"
+
 struct nv50_instmem_priv {
 	struct nouveau_instmem base;
 	spinlock_t lock;
@@ -38,42 +38,9 @@ struct nv50_instobj_priv {
 	struct nouveau_mem *mem;
 };
 
-static int
-nv50_instobj_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
-		  struct nouveau_oclass *oclass, void *data, u32 size,
-		  struct nouveau_object **pobject)
-{
-	struct nouveau_fb *pfb = nouveau_fb(parent);
-	struct nv50_instobj_priv *node;
-	u32 align = (unsigned long)data;
-	int ret;
-
-	size  = max((size  + 4095) & ~4095, (u32)4096);
-	align = max((align + 4095) & ~4095, (u32)4096);
-
-	ret = nouveau_instobj_create(parent, engine, oclass, &node);
-	*pobject = nv_object(node);
-	if (ret)
-		return ret;
-
-	ret = pfb->ram->get(pfb, size, align, 0, 0x800, &node->mem);
-	if (ret)
-		return ret;
-
-	node->base.addr = node->mem->offset;
-	node->base.size = node->mem->size << 12;
-	node->mem->page_shift = 12;
-	return 0;
-}
-
-static void
-nv50_instobj_dtor(struct nouveau_object *object)
-{
-	struct nv50_instobj_priv *node = (void *)object;
-	struct nouveau_fb *pfb = nouveau_fb(object);
-	pfb->ram->put(pfb, &node->mem);
-	nouveau_instobj_destroy(&node->base);
-}
+/******************************************************************************
+ * instmem object implementation
+ *****************************************************************************/
 
 static u32
 nv50_instobj_rd32(struct nouveau_object *object, u64 offset)
@@ -113,9 +80,46 @@ nv50_instobj_wr32(struct nouveau_object *object, u64 offset, u32 data)
 	spin_unlock_irqrestore(&priv->lock, flags);
 }
 
-static struct nouveau_oclass
+static void
+nv50_instobj_dtor(struct nouveau_object *object)
+{
+	struct nv50_instobj_priv *node = (void *)object;
+	struct nouveau_fb *pfb = nouveau_fb(object);
+	pfb->ram->put(pfb, &node->mem);
+	nouveau_instobj_destroy(&node->base);
+}
+
+static int
+nv50_instobj_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
+		  struct nouveau_oclass *oclass, void *data, u32 size,
+		  struct nouveau_object **pobject)
+{
+	struct nouveau_fb *pfb = nouveau_fb(parent);
+	struct nouveau_instobj_args *args = data;
+	struct nv50_instobj_priv *node;
+	int ret;
+
+	args->size  = max((args->size  + 4095) & ~4095, (u32)4096);
+	args->align = max((args->align + 4095) & ~4095, (u32)4096);
+
+	ret = nouveau_instobj_create(parent, engine, oclass, &node);
+	*pobject = nv_object(node);
+	if (ret)
+		return ret;
+
+	ret = pfb->ram->get(pfb, args->size, args->align, 0, 0x800, &node->mem);
+	if (ret)
+		return ret;
+
+	node->base.addr = node->mem->offset;
+	node->base.size = node->mem->size << 12;
+	node->mem->page_shift = 12;
+	return 0;
+}
+
+static struct nouveau_instobj_impl
 nv50_instobj_oclass = {
-	.ofuncs = &(struct nouveau_ofuncs) {
+	.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nv50_instobj_ctor,
 		.dtor = nv50_instobj_dtor,
 		.init = _nouveau_instobj_init,
@@ -125,13 +129,16 @@ nv50_instobj_oclass = {
 	},
 };
 
+/******************************************************************************
+ * instmem subdev implementation
+ *****************************************************************************/
+
 static int
-nv50_instmem_alloc(struct nouveau_instmem *imem, struct nouveau_object *parent,
-		   u32 size, u32 align, struct nouveau_object **pobject)
+nv50_instmem_fini(struct nouveau_object *object, bool suspend)
 {
-	struct nouveau_object *engine = nv_object(imem);
-	return nouveau_object_ctor(parent, engine, &nv50_instobj_oclass,
-				   (void *)(unsigned long)align, size, pobject);
+	struct nv50_instmem_priv *priv = (void *)object;
+	priv->addr = ~0ULL;
+	return nouveau_instmem_fini(&priv->base, suspend);
 }
 
 static int
@@ -148,25 +155,17 @@ nv50_instmem_ctor(struct nouveau_object *parent, struct nouveau_object *engine,
 		return ret;
 
 	spin_lock_init(&priv->lock);
-	priv->base.alloc = nv50_instmem_alloc;
 	return 0;
 }
 
-static int
-nv50_instmem_fini(struct nouveau_object *object, bool suspend)
-{
-	struct nv50_instmem_priv *priv = (void *)object;
-	priv->addr = ~0ULL;
-	return nouveau_instmem_fini(&priv->base, suspend);
-}
-
-struct nouveau_oclass
-nv50_instmem_oclass = {
-	.handle = NV_SUBDEV(INSTMEM, 0x50),
-	.ofuncs = &(struct nouveau_ofuncs) {
+struct nouveau_oclass *
+nv50_instmem_oclass = &(struct nouveau_instmem_impl) {
+	.base.handle = NV_SUBDEV(INSTMEM, 0x50),
+	.base.ofuncs = &(struct nouveau_ofuncs) {
 		.ctor = nv50_instmem_ctor,
 		.dtor = _nouveau_instmem_dtor,
 		.init = _nouveau_instmem_init,
 		.fini = nv50_instmem_fini,
 	},
-};
+	.instobj = &nv50_instobj_oclass.base,
+}.base;
diff --git a/drivers/gpu/drm/nouveau/core/subdev/instmem/priv.h b/drivers/gpu/drm/nouveau/core/subdev/instmem/priv.h
new file mode 100644
index 000000000000..8d67dedc5bb2
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/instmem/priv.h
@@ -0,0 +1,56 @@
+#ifndef __NVKM_INSTMEM_PRIV_H__
+#define __NVKM_INSTMEM_PRIV_H__
+
+#include <subdev/instmem.h>
+
+struct nouveau_instobj_impl {
+	struct nouveau_oclass base;
+};
+
+struct nouveau_instobj_args {
+	u32 size;
+	u32 align;
+};
+
+#define nouveau_instobj_create(p,e,o,d)                                        \
+	nouveau_instobj_create_((p), (e), (o), sizeof(**d), (void **)d)
+#define nouveau_instobj_destroy(p) ({                                          \
+	struct nouveau_instobj *iobj = (p);                                    \
+	_nouveau_instobj_dtor(nv_object(iobj));                                \
+})
+#define nouveau_instobj_init(p)                                                \
+	nouveau_object_init(&(p)->base)
+#define nouveau_instobj_fini(p,s)                                              \
+	nouveau_object_fini(&(p)->base, (s))
+
+int  nouveau_instobj_create_(struct nouveau_object *, struct nouveau_object *,
+			     struct nouveau_oclass *, int, void **);
+void _nouveau_instobj_dtor(struct nouveau_object *);
+#define _nouveau_instobj_init nouveau_object_init
+#define _nouveau_instobj_fini nouveau_object_fini
+
+struct nouveau_instmem_impl {
+	struct nouveau_oclass base;
+	struct nouveau_oclass *instobj;
+};
+
+#define nouveau_instmem_create(p,e,o,d)                                        \
+	nouveau_instmem_create_((p), (e), (o), sizeof(**d), (void **)d)
+#define nouveau_instmem_destroy(p)                                             \
+	nouveau_subdev_destroy(&(p)->base)
+#define nouveau_instmem_init(p) ({                                             \
+	struct nouveau_instmem *imem = (p);                                    \
+	_nouveau_instmem_init(nv_object(imem));                                \
+})
+#define nouveau_instmem_fini(p,s) ({                                           \
+	struct nouveau_instmem *imem = (p);                                    \
+	_nouveau_instmem_fini(nv_object(imem), (s));                           \
+})
+
+int nouveau_instmem_create_(struct nouveau_object *, struct nouveau_object *,
+			    struct nouveau_oclass *, int, void **);
+#define _nouveau_instmem_dtor _nouveau_subdev_dtor
+int _nouveau_instmem_init(struct nouveau_object *);
+int _nouveau_instmem_fini(struct nouveau_object *, bool);
+
+#endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/mc/nvc0.c b/drivers/gpu/drm/nouveau/core/subdev/mc/nvc0.c
index c02b4763a2d5..34472d317097 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/mc/nvc0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/mc/nvc0.c
@@ -32,6 +32,7 @@ nvc0_mc_intr[] = {
 	{ 0x00000080, NVDEV_ENGINE_COPY2 },
 	{ 0x00000100, NVDEV_ENGINE_FIFO },
 	{ 0x00001000, NVDEV_ENGINE_GR },
+	{ 0x00002000, NVDEV_SUBDEV_FB },
 	{ 0x00008000, NVDEV_ENGINE_BSP },
 	{ 0x00040000, NVDEV_SUBDEV_THERM },
 	{ 0x00020000, NVDEV_ENGINE_VP },
@@ -40,6 +41,7 @@ nvc0_mc_intr[] = {
 	{ 0x01000000, NVDEV_SUBDEV_PWR },
 	{ 0x02000000, NVDEV_SUBDEV_LTCG },
 	{ 0x04000000, NVDEV_ENGINE_DISP },
+	{ 0x08000000, NVDEV_SUBDEV_FB },
 	{ 0x10000000, NVDEV_SUBDEV_BUS },
 	{ 0x40000000, NVDEV_SUBDEV_IBUS },
 	{ 0x80000000, NVDEV_ENGINE_SW },
diff --git a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/i2c_.fuc b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/i2c_.fuc
new file mode 100644
index 000000000000..757dda700024
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/i2c_.fuc
@@ -0,0 +1,393 @@
+/*
+ * 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
+ */
+
+#define T_TIMEOUT  2200000
+#define T_RISEFALL 1000
+#define T_HOLD     5000
+
+#ifdef INCLUDE_PROC
+process(PROC_I2C_, #i2c_init, #i2c_recv)
+#endif
+
+/******************************************************************************
+ * I2C_ data segment
+ *****************************************************************************/
+#ifdef INCLUDE_DATA
+i2c_scl_map:
+.b32 NV_PPWR_OUTPUT_I2C_0_SCL
+.b32 NV_PPWR_OUTPUT_I2C_1_SCL
+.b32 NV_PPWR_OUTPUT_I2C_2_SCL
+.b32 NV_PPWR_OUTPUT_I2C_3_SCL
+.b32 NV_PPWR_OUTPUT_I2C_4_SCL
+.b32 NV_PPWR_OUTPUT_I2C_5_SCL
+.b32 NV_PPWR_OUTPUT_I2C_6_SCL
+.b32 NV_PPWR_OUTPUT_I2C_7_SCL
+.b32 NV_PPWR_OUTPUT_I2C_8_SCL
+.b32 NV_PPWR_OUTPUT_I2C_9_SCL
+i2c_sda_map:
+.b32 NV_PPWR_OUTPUT_I2C_0_SDA
+.b32 NV_PPWR_OUTPUT_I2C_1_SDA
+.b32 NV_PPWR_OUTPUT_I2C_2_SDA
+.b32 NV_PPWR_OUTPUT_I2C_3_SDA
+.b32 NV_PPWR_OUTPUT_I2C_4_SDA
+.b32 NV_PPWR_OUTPUT_I2C_5_SDA
+.b32 NV_PPWR_OUTPUT_I2C_6_SDA
+.b32 NV_PPWR_OUTPUT_I2C_7_SDA
+.b32 NV_PPWR_OUTPUT_I2C_8_SDA
+.b32 NV_PPWR_OUTPUT_I2C_9_SDA
+#if NVKM_PPWR_CHIPSET < GF119
+i2c_ctrl:
+.b32 0x00e138
+.b32 0x00e150
+.b32 0x00e168
+.b32 0x00e180
+.b32 0x00e254
+.b32 0x00e274
+.b32 0x00e764
+.b32 0x00e780
+.b32 0x00e79c
+.b32 0x00e7b8
+#endif
+#endif
+
+/******************************************************************************
+ * I2C_ code segment
+ *****************************************************************************/
+#ifdef INCLUDE_CODE
+
+// $r3  - value
+// $r2  - sda line
+// $r1  - scl line
+// $r0  - zero
+i2c_drive_scl:
+	cmp b32 $r3 0
+	bra e #i2c_drive_scl_lo
+	nv_iowr(NV_PPWR_OUTPUT_SET, $r1)
+	ret
+	i2c_drive_scl_lo:
+	nv_iowr(NV_PPWR_OUTPUT_CLR, $r1)
+	ret
+
+i2c_drive_sda:
+	cmp b32 $r3 0
+	bra e #i2c_drive_sda_lo
+	nv_iowr(NV_PPWR_OUTPUT_SET, $r2)
+	ret
+	i2c_drive_sda_lo:
+	nv_iowr(NV_PPWR_OUTPUT_CLR, $r2)
+	ret
+
+i2c_sense_scl:
+	bclr $flags $p1
+	nv_iord($r3, NV_PPWR_INPUT)
+	and $r3 $r1
+	bra z #i2c_sense_scl_done
+		bset $flags $p1
+	i2c_sense_scl_done:
+	ret
+
+i2c_sense_sda:
+	bclr $flags $p1
+	nv_iord($r3, NV_PPWR_INPUT)
+	and $r3 $r2
+	bra z #i2c_sense_sda_done
+		bset $flags $p1
+	i2c_sense_sda_done:
+	ret
+
+#define i2c_drive_scl(v) /*
+*/	mov $r3 (v) /*
+*/	call(i2c_drive_scl)
+#define i2c_drive_sda(v) /*
+*/	mov $r3 (v) /*
+*/	call(i2c_drive_sda)
+#define i2c_sense_scl() /*
+*/	call(i2c_sense_scl)
+#define i2c_sense_sda() /*
+*/	call(i2c_sense_sda)
+#define i2c_delay(v) /*
+*/	mov $r14 (v) /*
+*/	call(nsec)
+
+#define i2c_trace_init() /*
+*/	imm32($r6, 0x10000000) /*
+*/	sub b32 $r7 $r6 1 /*
+*/
+#define i2c_trace_down() /*
+*/	shr b32 $r6 4 /*
+*/	push $r5 /*
+*/	shl b32 $r5 $r6 4 /*
+*/	sub b32 $r5 $r6 /*
+*/	not b32 $r5 /*
+*/	and $r7 $r5 /*
+*/	pop $r5 /*
+*/
+#define i2c_trace_exit() /*
+*/	shl b32 $r6 4 /*
+*/
+#define i2c_trace_next() /*
+*/	add b32 $r7 $r6 /*
+*/
+#define i2c_trace_call(func) /*
+*/	i2c_trace_next() /*
+*/	i2c_trace_down() /*
+*/	call(func) /*
+*/	i2c_trace_exit() /*
+*/
+
+i2c_raise_scl:
+	push $r4
+	mov $r4 (T_TIMEOUT / T_RISEFALL)
+	i2c_drive_scl(1)
+	i2c_raise_scl_wait:
+		i2c_delay(T_RISEFALL)
+		i2c_sense_scl()
+		bra $p1 #i2c_raise_scl_done
+		sub b32 $r4 1
+		bra nz #i2c_raise_scl_wait
+	i2c_raise_scl_done:
+	pop $r4
+	ret
+
+i2c_start:
+	i2c_sense_scl()
+	bra not $p1 #i2c_start_rep
+	i2c_sense_sda()
+	bra not $p1 #i2c_start_rep
+	bra #i2c_start_send
+	i2c_start_rep:
+		i2c_drive_scl(0)
+		i2c_drive_sda(1)
+		i2c_trace_call(i2c_raise_scl)
+		bra not $p1 #i2c_start_out
+	i2c_start_send:
+	i2c_drive_sda(0)
+	i2c_delay(T_HOLD)
+	i2c_drive_scl(0)
+	i2c_delay(T_HOLD)
+	i2c_start_out:
+	ret
+
+i2c_stop:
+	i2c_drive_scl(0)
+	i2c_drive_sda(0)
+	i2c_delay(T_RISEFALL)
+	i2c_drive_scl(1)
+	i2c_delay(T_HOLD)
+	i2c_drive_sda(1)
+	i2c_delay(T_HOLD)
+	ret
+
+// $r3  - value
+// $r2  - sda line
+// $r1  - scl line
+// $r0  - zero
+i2c_bitw:
+	call(i2c_drive_sda)
+	i2c_delay(T_RISEFALL)
+	i2c_trace_call(i2c_raise_scl)
+	bra not $p1 #i2c_bitw_out
+	i2c_delay(T_HOLD)
+	i2c_drive_scl(0)
+	i2c_delay(T_HOLD)
+	i2c_bitw_out:
+	ret
+
+// $r3  - value (out)
+// $r2  - sda line
+// $r1  - scl line
+// $r0  - zero
+i2c_bitr:
+	i2c_drive_sda(1)
+	i2c_delay(T_RISEFALL)
+	i2c_trace_call(i2c_raise_scl)
+	bra not $p1 #i2c_bitr_done
+	i2c_sense_sda()
+	i2c_drive_scl(0)
+	i2c_delay(T_HOLD)
+	xbit $r3 $flags $p1
+	bset $flags $p1
+	i2c_bitr_done:
+	ret
+
+i2c_get_byte:
+	mov $r5 0
+	mov $r4 8
+	i2c_get_byte_next:
+		shl b32 $r5 1
+		i2c_trace_call(i2c_bitr)
+		bra not $p1 #i2c_get_byte_done
+		or $r5 $r3
+		sub b32 $r4 1
+		bra nz #i2c_get_byte_next
+	mov $r3 1
+	i2c_trace_call(i2c_bitw)
+	i2c_get_byte_done:
+	ret
+
+i2c_put_byte:
+	mov $r4 8
+	i2c_put_byte_next:
+		sub b32 $r4 1
+		xbit $r3 $r5 $r4
+		i2c_trace_call(i2c_bitw)
+		bra not $p1 #i2c_put_byte_done
+		cmp b32 $r4 0
+		bra ne #i2c_put_byte_next
+	i2c_trace_call(i2c_bitr)
+	bra not $p1 #i2c_put_byte_done
+	i2c_trace_next()
+	cmp b32 $r3 1
+	bra ne #i2c_put_byte_done
+	bclr $flags $p1	// nack
+	i2c_put_byte_done:
+	ret
+
+i2c_addr:
+	i2c_trace_call(i2c_start)
+	bra not $p1 #i2c_addr_done
+	extr $r3 $r12 I2C__MSG_DATA0_ADDR
+	shl b32 $r3 1
+	or $r5 $r3
+	i2c_trace_call(i2c_put_byte)
+	i2c_addr_done:
+	ret
+
+i2c_acquire_addr:
+	extr $r14 $r12 I2C__MSG_DATA0_PORT
+#if NVKM_PPWR_CHIPSET < GF119
+	shl b32 $r14 2
+	add b32 $r14 #i2c_ctrl
+	ld b32 $r14 D[$r14]
+#else
+	shl b32 $r14 5
+	add b32 $r14 0x00d014
+#endif
+	ret
+
+i2c_acquire:
+	call(i2c_acquire_addr)
+	call(rd32)
+	bset $r13 3
+	call(wr32)
+	ret
+
+i2c_release:
+	call(i2c_acquire_addr)
+	call(rd32)
+	bclr $r13 3
+	call(wr32)
+	ret
+
+// description
+//
+// $r15 - current (i2c)
+// $r14 - sender process name
+// $r13 - message
+// $r12 - data0
+// $r11 - data1
+// $r0  - zero
+i2c_recv:
+	bclr $flags $p1
+	extr $r1 $r12 I2C__MSG_DATA0_PORT
+	shl b32 $r1 2
+	cmp b32 $r1 (#i2c_sda_map - #i2c_scl_map)
+	bra ge #i2c_recv_done
+	add b32 $r3 $r1 #i2c_sda_map
+	ld b32 $r2 D[$r3]
+	add b32 $r3 $r1 #i2c_scl_map
+	ld b32 $r1 D[$r3]
+
+	bset $flags $p2
+	push $r13
+	push $r14
+
+	push $r13
+	i2c_trace_init()
+	i2c_trace_call(i2c_acquire)
+	pop $r13
+
+	cmp b32 $r13 I2C__MSG_RD08
+	bra ne #i2c_recv_not_rd08
+		mov $r5 0
+		i2c_trace_call(i2c_addr)
+		bra not $p1 #i2c_recv_done
+		extr $r5 $r12 I2C__MSG_DATA0_RD08_REG
+		i2c_trace_call(i2c_put_byte)
+		bra not $p1 #i2c_recv_done
+		mov $r5 1
+		i2c_trace_call(i2c_addr)
+		bra not $p1 #i2c_recv_done
+		i2c_trace_call(i2c_get_byte)
+		bra not $p1 #i2c_recv_done
+		ins $r11 $r5 I2C__MSG_DATA1_RD08_VAL
+		i2c_trace_call(i2c_stop)
+		mov b32 $r11 $r5
+		clear b32 $r7
+		bra #i2c_recv_done
+
+	i2c_recv_not_rd08:
+	cmp b32 $r13 I2C__MSG_WR08
+	bra ne #i2c_recv_not_wr08
+		mov $r5 0
+		call(i2c_addr)
+		bra not $p1 #i2c_recv_done
+		extr $r5 $r12 I2C__MSG_DATA0_WR08_REG
+		call(i2c_put_byte)
+		bra not $p1 #i2c_recv_done
+		mov $r5 0
+		call(i2c_addr)
+		bra not $p1 #i2c_recv_done
+		extr $r5 $r11 I2C__MSG_DATA1_WR08_VAL
+		call(i2c_put_byte)
+		bra not $p1 #i2c_recv_done
+		call(i2c_stop)
+		clear b32 $r7
+		extr $r5 $r12 I2C__MSG_DATA0_WR08_SYNC
+		bra nz #i2c_recv_done
+		bclr $flags $p2
+		bra #i2c_recv_done
+
+	i2c_recv_not_wr08:
+
+	i2c_recv_done:
+	extr $r14 $r12 I2C__MSG_DATA0_PORT
+	call(i2c_release)
+
+	pop $r14
+	pop $r13
+	bra not $p2 #i2c_recv_exit
+	mov b32 $r12 $r7
+	call(send)
+
+	i2c_recv_exit:
+	ret
+
+// description
+//
+// $r15 - current (i2c)
+// $r0  - zero
+i2c_init:
+	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 0a7b05fa5c11..8f29badd785f 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/kernel.fuc
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/kernel.fuc
@@ -51,12 +51,12 @@ time_next: .b32 0
 // $r0  - zero
 rd32:
 	nv_iowr(NV_PPWR_MMIO_ADDR, $r14)
-	mov $r14 NV_PPWR_MMIO_CTRL_OP_RD
-	sethi $r14 NV_PPWR_MMIO_CTRL_TRIGGER
-	nv_iowr(NV_PPWR_MMIO_CTRL, $r14)
+	mov $r13 NV_PPWR_MMIO_CTRL_OP_RD
+	sethi $r13 NV_PPWR_MMIO_CTRL_TRIGGER
+	nv_iowr(NV_PPWR_MMIO_CTRL, $r13)
 	rd32_wait:
-		nv_iord($r14, NV_PPWR_MMIO_CTRL)
-		and $r14 NV_PPWR_MMIO_CTRL_STATUS
+		nv_iord($r13, NV_PPWR_MMIO_CTRL)
+		and $r13 NV_PPWR_MMIO_CTRL_STATUS
 		bra nz #rd32_wait
 	nv_iord($r13, NV_PPWR_MMIO_DATA)
 	ret
@@ -70,23 +70,25 @@ rd32:
 wr32:
 	nv_iowr(NV_PPWR_MMIO_ADDR, $r14)
 	nv_iowr(NV_PPWR_MMIO_DATA, $r13)
-	mov $r14 NV_PPWR_MMIO_CTRL_OP_WR
-	or $r14 NV_PPWR_MMIO_CTRL_MASK_B32_0
-	sethi $r14 NV_PPWR_MMIO_CTRL_TRIGGER
+	mov $r13 NV_PPWR_MMIO_CTRL_OP_WR
+	or $r13 NV_PPWR_MMIO_CTRL_MASK_B32_0
+	sethi $r13 NV_PPWR_MMIO_CTRL_TRIGGER
 
 #ifdef NVKM_FALCON_MMIO_TRAP
-	mov $r8 NV_PPWR_INTR_TRIGGER_USER1
-	nv_iowr(NV_PPWR_INTR_TRIGGER, $r8)
+	push $r13
+	mov $r13 NV_PPWR_INTR_TRIGGER_USER1
+	nv_iowr(NV_PPWR_INTR_TRIGGER, $r13)
 	wr32_host:
-		nv_iord($r8, NV_PPWR_INTR)
-		and $r8 NV_PPWR_INTR_USER1
+		nv_iord($r13, NV_PPWR_INTR)
+		and $r13 NV_PPWR_INTR_USER1
 		bra nz #wr32_host
+	pop $r13
 #endif
 
-	nv_iowr(NV_PPWR_MMIO_CTRL, $r14)
+	nv_iowr(NV_PPWR_MMIO_CTRL, $r13)
 	wr32_wait:
-		nv_iord($r14, NV_PPWR_MMIO_CTRL)
-		and $r14 NV_PPWR_MMIO_CTRL_STATUS
+		nv_iord($r13, NV_PPWR_MMIO_CTRL)
+		and $r13 NV_PPWR_MMIO_CTRL_STATUS
 		bra nz #wr32_wait
 	ret
 
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 2a74ea907604..e2a63ac5422b 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/macros.fuc
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/macros.fuc
@@ -83,6 +83,50 @@
 #define NV_PPWR_MMIO_CTRL_OP_WR                                      0x00000002
 #define NV_PPWR_OUTPUT                                                   0x07c0
 #define NV_PPWR_OUTPUT_FB_PAUSE                                      0x00000004
+#if NVKM_PPWR_CHIPSET < GF119
+#define NV_PPWR_OUTPUT_I2C_3_SCL                                     0x00000100
+#define NV_PPWR_OUTPUT_I2C_3_SDA                                     0x00000200
+#define NV_PPWR_OUTPUT_I2C_0_SCL                                     0x00001000
+#define NV_PPWR_OUTPUT_I2C_0_SDA                                     0x00002000
+#define NV_PPWR_OUTPUT_I2C_1_SCL                                     0x00004000
+#define NV_PPWR_OUTPUT_I2C_1_SDA                                     0x00008000
+#define NV_PPWR_OUTPUT_I2C_2_SCL                                     0x00010000
+#define NV_PPWR_OUTPUT_I2C_2_SDA                                     0x00020000
+#define NV_PPWR_OUTPUT_I2C_4_SCL                                     0x00040000
+#define NV_PPWR_OUTPUT_I2C_4_SDA                                     0x00080000
+#define NV_PPWR_OUTPUT_I2C_5_SCL                                     0x00100000
+#define NV_PPWR_OUTPUT_I2C_5_SDA                                     0x00200000
+#define NV_PPWR_OUTPUT_I2C_6_SCL                                     0x00400000
+#define NV_PPWR_OUTPUT_I2C_6_SDA                                     0x00800000
+#define NV_PPWR_OUTPUT_I2C_7_SCL                                     0x01000000
+#define NV_PPWR_OUTPUT_I2C_7_SDA                                     0x02000000
+#define NV_PPWR_OUTPUT_I2C_8_SCL                                     0x04000000
+#define NV_PPWR_OUTPUT_I2C_8_SDA                                     0x08000000
+#define NV_PPWR_OUTPUT_I2C_9_SCL                                     0x10000000
+#define NV_PPWR_OUTPUT_I2C_9_SDA                                     0x20000000
+#else
+#define NV_PPWR_OUTPUT_I2C_0_SCL                                     0x00000400
+#define NV_PPWR_OUTPUT_I2C_1_SCL                                     0x00000800
+#define NV_PPWR_OUTPUT_I2C_2_SCL                                     0x00001000
+#define NV_PPWR_OUTPUT_I2C_3_SCL                                     0x00002000
+#define NV_PPWR_OUTPUT_I2C_4_SCL                                     0x00004000
+#define NV_PPWR_OUTPUT_I2C_5_SCL                                     0x00008000
+#define NV_PPWR_OUTPUT_I2C_6_SCL                                     0x00010000
+#define NV_PPWR_OUTPUT_I2C_7_SCL                                     0x00020000
+#define NV_PPWR_OUTPUT_I2C_8_SCL                                     0x00040000
+#define NV_PPWR_OUTPUT_I2C_9_SCL                                     0x00080000
+#define NV_PPWR_OUTPUT_I2C_0_SDA                                     0x00100000
+#define NV_PPWR_OUTPUT_I2C_1_SDA                                     0x00200000
+#define NV_PPWR_OUTPUT_I2C_2_SDA                                     0x00400000
+#define NV_PPWR_OUTPUT_I2C_3_SDA                                     0x00800000
+#define NV_PPWR_OUTPUT_I2C_4_SDA                                     0x01000000
+#define NV_PPWR_OUTPUT_I2C_5_SDA                                     0x02000000
+#define NV_PPWR_OUTPUT_I2C_6_SDA                                     0x04000000
+#define NV_PPWR_OUTPUT_I2C_7_SDA                                     0x08000000
+#define NV_PPWR_OUTPUT_I2C_8_SDA                                     0x10000000
+#define NV_PPWR_OUTPUT_I2C_9_SDA                                     0x20000000
+#endif
+#define NV_PPWR_INPUT                                                    0x07c4
 #define NV_PPWR_OUTPUT_SET                                               0x07e0
 #define NV_PPWR_OUTPUT_SET_FB_PAUSE                                  0x00000004
 #define NV_PPWR_OUTPUT_CLR                                               0x07e4
@@ -125,6 +169,15 @@
 */	.b32 0 /*
 */	.skip 64
 
+#if NV_PPWR_CHIPSET < GK208
+#define imm32(reg,val) /*
+*/	movw reg  ((val) & 0x0000ffff) /*
+*/	sethi reg ((val) & 0xffff0000)
+#else
+#define imm32(reg,val) /*
+*/	mov reg (val)
+#endif
+
 #ifndef NVKM_FALCON_UNSHIFTED_IO
 #define nv_iord(reg,ior) /*
 */	mov reg ior /*
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 947be536daef..17a8a383d91a 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nv108.fuc
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nv108.fuc
@@ -37,6 +37,7 @@
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
+#include "i2c_.fuc"
 #include "test.fuc"
 #include "idle.fuc"
 #undef INCLUDE_PROC
@@ -46,6 +47,7 @@
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
+#include "i2c_.fuc"
 #include "test.fuc"
 #include "idle.fuc"
 #undef INCLUDE_DATA
@@ -57,6 +59,7 @@
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
+#include "i2c_.fuc"
 #include "test.fuc"
 #include "idle.fuc"
 #undef INCLUDE_CODE
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 9342e2d7d3b7..4bd43a99fdcc 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
@@ -89,16 +89,9 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-	0x54534554,
-	0x00000494,
-	0x00000475,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
+	0x5f433249,
+	0x00000877,
+	0x0000071e,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -111,16 +104,6 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-	0x454c4449,
-	0x0000049f,
-	0x0000049d,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -128,17 +111,16 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+	0x54534554,
+	0x00000898,
+	0x00000879,
 	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0210: proc_list_tail */
-/* 0x0210: time_prev */
 	0x00000000,
-/* 0x0214: time_next */
 	0x00000000,
-/* 0x0218: fifo_queue */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -151,6 +133,9 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+	0x454c4449,
+	0x000008a3,
+	0x000008a1,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -170,9 +155,12 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+/* 0x0268: proc_list_tail */
+/* 0x0268: time_prev */
 	0x00000000,
-/* 0x0298: rfifo_queue */
+/* 0x026c: time_next */
 	0x00000000,
+/* 0x0270: fifo_queue */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -204,31 +192,8 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0318: memx_func_head */
-	0x00010000,
-	0x00000000,
-	0x000003a9,
-/* 0x0324: memx_func_next */
-	0x00000001,
-	0x00000000,
-	0x000003c7,
-	0x00000002,
-	0x00000002,
-	0x000003df,
-	0x00040003,
-	0x00000000,
-	0x00000407,
-	0x00010004,
-	0x00000000,
-	0x00000421,
-/* 0x0354: memx_func_tail */
-/* 0x0354: memx_data_head */
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
 	0x00000000,
+/* 0x02f0: rfifo_queue */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -261,10 +226,25 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+/* 0x0370: memx_func_head */
+	0x00010000,
 	0x00000000,
+	0x000003a9,
+/* 0x037c: memx_func_next */
+	0x00000001,
 	0x00000000,
+	0x000003c7,
+	0x00000002,
+	0x00000002,
+	0x000003df,
+	0x00040003,
 	0x00000000,
+	0x00000407,
+	0x00010004,
 	0x00000000,
+	0x00000421,
+/* 0x03ac: memx_func_tail */
+/* 0x03ac: memx_data_head */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -735,7 +715,6 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0b54: memx_data_tail */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -778,6 +757,29 @@ uint32_t nv108_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+/* 0x0bac: memx_data_tail */
+/* 0x0bac: i2c_scl_map */
+	0x00000400,
+	0x00000800,
+	0x00001000,
+	0x00002000,
+	0x00004000,
+	0x00008000,
+	0x00010000,
+	0x00020000,
+	0x00040000,
+	0x00080000,
+/* 0x0bd4: i2c_sda_map */
+	0x00100000,
+	0x00200000,
+	0x00400000,
+	0x00800000,
+	0x01000000,
+	0x02000000,
+	0x04000000,
+	0x08000000,
+	0x10000000,
+	0x20000000,
 	0x00000000,
 };
 
@@ -786,13 +788,13 @@ uint32_t nv108_pwr_code[] = {
 /* 0x0004: rd32 */
 	0xf607a040,
 	0x04bd000e,
-	0xe3f0010e,
+	0xd3f0010d,
 	0x07ac4001,
-	0xbd000ef6,
+	0xbd000df6,
 /* 0x0019: rd32_wait */
-	0x07ac4e04,
-	0xf100eecf,
-	0xf47000e4,
+	0x07ac4d04,
+	0xf100ddcf,
+	0xf47000d4,
 	0xa44df61b,
 	0x00ddcf07,
 /* 0x002e: wr32 */
@@ -800,14 +802,14 @@ uint32_t nv108_pwr_code[] = {
 	0x000ef607,
 	0xa44004bd,
 	0x000df607,
-	0x020e04bd,
-	0xf0f0e5f0,
-	0xac4001e3,
-	0x000ef607,
+	0x020d04bd,
+	0xf0f0d5f0,
+	0xac4001d3,
+	0x000df607,
 /* 0x004e: wr32_wait */
-	0xac4e04bd,
-	0x00eecf07,
-	0x7000e4f1,
+	0xac4d04bd,
+	0x00ddcf07,
+	0x7000d4f1,
 	0xf8f61bf4,
 /* 0x005d: nsec */
 	0xcf2c0800,
@@ -832,20 +834,20 @@ uint32_t nv108_pwr_code[] = {
 	0x03e99800,
 	0xf40096b0,
 	0x0a98280b,
-	0x029abb84,
+	0x029abb9a,
 	0x0d0e1cf4,
 	0x01de7e01,
 	0xf494bd00,
 /* 0x00b2: intr_watchdog_next_time */
 	0x0a98140e,
-	0x00a6b085,
+	0x00a6b09b,
 	0xa6080bf4,
 	0x061cf49a,
 /* 0x00c0: intr_watchdog_next_time_set */
 /* 0x00c3: intr_watchdog_next_proc */
-	0xb58509b5,
+	0xb59b09b5,
 	0xe0b603e9,
-	0x10e6b158,
+	0x68e6b158,
 	0xc81bf402,
 /* 0x00d2: intr */
 	0x00f900f8,
@@ -862,15 +864,15 @@ uint32_t nv108_pwr_code[] = {
 	0x080804bd,
 	0xc40088cf,
 	0x0bf40289,
-	0x8500b51f,
+	0x9b00b51f,
 	0x957e580e,
 	0x09980000,
-	0x0096b085,
+	0x0096b09b,
 	0x000d0bf4,
 	0x0009f634,
 	0x09b504bd,
 /* 0x0125: intr_skip_watchdog */
-	0x0089e484,
+	0x0089e49a,
 	0x360bf408,
 	0xcf068849,
 	0x9ac40099,
@@ -918,7 +920,7 @@ uint32_t nv108_pwr_code[] = {
 /* 0x01c6: timer_reset */
 	0x3400161e,
 	0xbd000ef6,
-	0x840eb504,
+	0x9a0eb504,
 /* 0x01d0: timer_enable */
 	0x38000108,
 	0xbd0008f6,
@@ -949,7 +951,7 @@ uint32_t nv108_pwr_code[] = {
 	0xa6008a98,
 	0x100bf4ae,
 	0xb15880b6,
-	0xf4021086,
+	0xf4026886,
 	0x32f4f11b,
 /* 0x0239: find_done */
 	0xfc8eb201,
@@ -1009,7 +1011,7 @@ uint32_t nv108_pwr_code[] = {
 	0x0bf412a6,
 	0x071ec42e,
 	0xb704ee94,
-	0x980218e0,
+	0x980270e0,
 	0xec9803eb,
 	0x01ed9802,
 	0x7e00ee98,
@@ -1031,7 +1033,7 @@ uint32_t nv108_pwr_code[] = {
 	0xf412a608,
 	0x23c4ef0b,
 	0x0434b607,
-	0x029830b7,
+	0x02f030b7,
 	0xb5033bb5,
 	0x3db5023c,
 	0x003eb501,
@@ -1044,11 +1046,11 @@ uint32_t nv108_pwr_code[] = {
 /* 0x0379: host_init */
 	0x00804100,
 	0xf11014b6,
-	0x40021815,
+	0x40027015,
 	0x01f604d0,
 	0x4104bd00,
 	0x14b60080,
-	0x9815f110,
+	0xf015f110,
 	0x04dc4002,
 	0xbd0001f6,
 	0x40010104,
@@ -1101,13 +1103,13 @@ uint32_t nv108_pwr_code[] = {
 	0x001398b2,
 	0x950410b6,
 	0x30f01034,
-	0xc835980c,
+	0xde35980c,
 	0x12a655f9,
 	0xfced1ef4,
 	0x7ee0fcd0,
 	0xf800023f,
 /* 0x0455: memx_info */
-	0x03544c00,
+	0x03ac4c00,
 	0x7e08004b,
 	0xf800023f,
 /* 0x0461: memx_recv */
@@ -1119,7 +1121,301 @@ uint32_t nv108_pwr_code[] = {
 /* 0x0471: perf_recv */
 /* 0x0473: perf_init */
 	0xf800f800,
-/* 0x0475: test_recv */
+/* 0x0475: i2c_drive_scl */
+	0x0036b000,
+	0x400d0bf4,
+	0x01f607e0,
+	0xf804bd00,
+/* 0x0485: i2c_drive_scl_lo */
+	0x07e44000,
+	0xbd0001f6,
+/* 0x048f: i2c_drive_sda */
+	0xb000f804,
+	0x0bf40036,
+	0x07e0400d,
+	0xbd0002f6,
+/* 0x049f: i2c_drive_sda_lo */
+	0x4000f804,
+	0x02f607e4,
+	0xf804bd00,
+/* 0x04a9: i2c_sense_scl */
+	0x0132f400,
+	0xcf07c443,
+	0x31fd0033,
+	0x060bf404,
+/* 0x04bb: i2c_sense_scl_done */
+	0xf80131f4,
+/* 0x04bd: i2c_sense_sda */
+	0x0132f400,
+	0xcf07c443,
+	0x32fd0033,
+	0x060bf404,
+/* 0x04cf: i2c_sense_sda_done */
+	0xf80131f4,
+/* 0x04d1: i2c_raise_scl */
+	0x4440f900,
+	0x01030898,
+	0x0004757e,
+/* 0x04dc: i2c_raise_scl_wait */
+	0x7e03e84e,
+	0x7e00005d,
+	0xf40004a9,
+	0x42b60901,
+	0xef1bf401,
+/* 0x04f0: i2c_raise_scl_done */
+	0x00f840fc,
+/* 0x04f4: i2c_start */
+	0x0004a97e,
+	0x7e0d11f4,
+	0xf40004bd,
+	0x0ef40611,
+/* 0x0505: i2c_start_rep */
+	0x7e00032e,
+	0x03000475,
+	0x048f7e01,
+	0x0076bb00,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0xd17e50fc,
+	0x64b60004,
+	0x1d11f404,
+/* 0x0530: i2c_start_send */
+	0x8f7e0003,
+	0x884e0004,
+	0x005d7e13,
+	0x7e000300,
+	0x4e000475,
+	0x5d7e1388,
+/* 0x054a: i2c_start_out */
+	0x00f80000,
+/* 0x054c: i2c_stop */
+	0x757e0003,
+	0x00030004,
+	0x00048f7e,
+	0x7e03e84e,
+	0x0300005d,
+	0x04757e01,
+	0x13884e00,
+	0x00005d7e,
+	0x8f7e0103,
+	0x884e0004,
+	0x005d7e13,
+/* 0x057b: i2c_bitw */
+	0x7e00f800,
+	0x4e00048f,
+	0x5d7e03e8,
+	0x76bb0000,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0x7e50fc04,
+	0xb60004d1,
+	0x11f40464,
+	0x13884e17,
+	0x00005d7e,
+	0x757e0003,
+	0x884e0004,
+	0x005d7e13,
+/* 0x05b9: i2c_bitw_out */
+/* 0x05bb: i2c_bitr */
+	0x0300f800,
+	0x048f7e01,
+	0x03e84e00,
+	0x00005d7e,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0x04d17e50,
+	0x0464b600,
+	0x7e1a11f4,
+	0x030004bd,
+	0x04757e00,
+	0x13884e00,
+	0x00005d7e,
+	0xf4013cf0,
+/* 0x05fe: i2c_bitr_done */
+	0x00f80131,
+/* 0x0600: i2c_get_byte */
+	0x08040005,
+/* 0x0604: i2c_get_byte_next */
+	0xbb0154b6,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x0005bb7e,
+	0xf40464b6,
+	0x53fd2a11,
+	0x0142b605,
+	0x03d81bf4,
+	0x0076bb01,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0x7b7e50fc,
+	0x64b60005,
+/* 0x064d: i2c_get_byte_done */
+/* 0x064f: i2c_put_byte */
+	0x0400f804,
+/* 0x0651: i2c_put_byte_next */
+	0x0142b608,
+	0xbb3854ff,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x00057b7e,
+	0xf40464b6,
+	0x46b03411,
+	0xd81bf400,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0x05bb7e50,
+	0x0464b600,
+	0xbb0f11f4,
+	0x36b00076,
+	0x061bf401,
+/* 0x06a7: i2c_put_byte_done */
+	0xf80132f4,
+/* 0x06a9: i2c_addr */
+	0x0076bb00,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0xf47e50fc,
+	0x64b60004,
+	0x2911f404,
+	0x012ec3e7,
+	0xfd0134b6,
+	0x76bb0553,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0x7e50fc04,
+	0xb600064f,
+/* 0x06ee: i2c_addr_done */
+	0x00f80464,
+/* 0x06f0: i2c_acquire_addr */
+	0xb6f8cec7,
+	0xe0b705e4,
+	0x00f8d014,
+/* 0x06fc: i2c_acquire */
+	0x0006f07e,
+	0x0000047e,
+	0x7e03d9f0,
+	0xf800002e,
+/* 0x070d: i2c_release */
+	0x06f07e00,
+	0x00047e00,
+	0x03daf000,
+	0x00002e7e,
+/* 0x071e: i2c_recv */
+	0x32f400f8,
+	0xf8c1c701,
+	0xb00214b6,
+	0x1ff52816,
+	0x13b80137,
+	0x98000bd4,
+	0x13b80032,
+	0x98000bac,
+	0x31f40031,
+	0xf9d0f902,
+	0xf1d0f9e0,
+	0xf1000067,
+	0x92100063,
+	0x76bb0167,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0x7e50fc04,
+	0xb60006fc,
+	0xd0fc0464,
+	0xf500d6b0,
+	0x0500b01b,
+	0x0076bb00,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0xa97e50fc,
+	0x64b60006,
+	0xcc11f504,
+	0xe0c5c700,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0x064f7e50,
+	0x0464b600,
+	0x00a911f5,
+	0x76bb0105,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0x7e50fc04,
+	0xb60006a9,
+	0x11f50464,
+	0x76bb0087,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0x7e50fc04,
+	0xb6000600,
+	0x11f40464,
+	0xe05bcb67,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0x054c7e50,
+	0x0464b600,
+	0x74bd5bb2,
+/* 0x0823: i2c_recv_not_rd08 */
+	0xb0410ef4,
+	0x1bf401d6,
+	0x7e00053b,
+	0xf40006a9,
+	0xc5c73211,
+	0x064f7ee0,
+	0x2811f400,
+	0xa97e0005,
+	0x11f40006,
+	0xe0b5c71f,
+	0x00064f7e,
+	0x7e1511f4,
+	0xbd00054c,
+	0x08c5c774,
+	0xf4091bf4,
+	0x0ef40232,
+/* 0x0861: i2c_recv_not_wr08 */
+/* 0x0861: i2c_recv_done */
+	0xf8cec703,
+	0x00070d7e,
+	0xd0fce0fc,
+	0xb20912f4,
+	0x023f7e7c,
+/* 0x0875: i2c_recv_exit */
+/* 0x0877: i2c_init */
+	0xf800f800,
+/* 0x0879: test_recv */
 	0x04584100,
 	0xb60011cf,
 	0x58400110,
@@ -1128,26 +1424,26 @@ uint32_t nv108_pwr_code[] = {
 	0xe3f1d900,
 	0x967e134f,
 	0x00f80001,
-/* 0x0494: test_init */
+/* 0x0898: test_init */
 	0x7e08004e,
 	0xf8000196,
-/* 0x049d: idle_recv */
-/* 0x049f: idle */
+/* 0x08a1: idle_recv */
+/* 0x08a3: idle */
 	0xf400f800,
 	0x54410031,
 	0x0011cf04,
 	0x400110b6,
 	0x01f60454,
-/* 0x04b3: idle_loop */
+/* 0x08b7: idle_loop */
 	0x0104bd00,
 	0x0232f458,
-/* 0x04b8: idle_proc */
-/* 0x04b8: idle_proc_exec */
+/* 0x08bc: idle_proc */
+/* 0x08bc: idle_proc_exec */
 	0x1eb210f9,
 	0x0002487e,
 	0x11f410fc,
 	0x0231f409,
-/* 0x04cb: idle_proc_next */
+/* 0x08cf: idle_proc_next */
 	0xb6f00ef4,
 	0x1fa65810,
 	0xf4e81bf4,
@@ -1161,5 +1457,4 @@ uint32_t nv108_pwr_code[] = {
 	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 6fde0b89e5aa..6744fcc06151 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nva3.fuc
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nva3.fuc
@@ -37,6 +37,7 @@
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
+#include "i2c_.fuc"
 #include "test.fuc"
 #include "idle.fuc"
 #undef INCLUDE_PROC
@@ -46,6 +47,7 @@
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
+#include "i2c_.fuc"
 #include "test.fuc"
 #include "idle.fuc"
 #undef INCLUDE_DATA
@@ -57,6 +59,7 @@
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
+#include "i2c_.fuc"
 #include "test.fuc"
 #include "idle.fuc"
 #undef INCLUDE_CODE
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 0fa4d7dcd407..5a73fa620978 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
@@ -89,9 +89,31 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+	0x5f433249,
+	0x00000982,
+	0x00000825,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 	0x54534554,
-	0x0000057b,
-	0x00000554,
+	0x000009ab,
+	0x00000984,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -112,8 +134,8 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x454c4449,
-	0x00000587,
-	0x00000585,
+	0x000009b7,
+	0x000009b5,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -133,12 +155,12 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0210: proc_list_tail */
-/* 0x0210: time_prev */
+/* 0x0268: proc_list_tail */
+/* 0x0268: time_prev */
 	0x00000000,
-/* 0x0214: time_next */
+/* 0x026c: time_next */
 	0x00000000,
-/* 0x0218: fifo_queue */
+/* 0x0270: fifo_queue */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -171,7 +193,7 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0298: rfifo_queue */
+/* 0x02f0: rfifo_queue */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -204,11 +226,11 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0318: memx_func_head */
+/* 0x0370: memx_func_head */
 	0x00010000,
 	0x00000000,
 	0x0000046f,
-/* 0x0324: memx_func_next */
+/* 0x037c: memx_func_next */
 	0x00000001,
 	0x00000000,
 	0x00000496,
@@ -221,8 +243,18 @@ uint32_t nva3_pwr_data[] = {
 	0x00010004,
 	0x00000000,
 	0x000004fc,
-/* 0x0354: memx_func_tail */
-/* 0x0354: memx_data_head */
+/* 0x03ac: memx_func_tail */
+/* 0x03ac: memx_data_head */
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -725,6 +757,42 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+/* 0x0bac: memx_data_tail */
+/* 0x0bac: i2c_scl_map */
+	0x00001000,
+	0x00004000,
+	0x00010000,
+	0x00000100,
+	0x00040000,
+	0x00100000,
+	0x00400000,
+	0x01000000,
+	0x04000000,
+	0x10000000,
+/* 0x0bd4: i2c_sda_map */
+	0x00002000,
+	0x00008000,
+	0x00020000,
+	0x00000200,
+	0x00080000,
+	0x00200000,
+	0x00800000,
+	0x02000000,
+	0x08000000,
+	0x20000000,
+/* 0x0bfc: i2c_ctrl */
+	0x0000e138,
+	0x0000e150,
+	0x0000e168,
+	0x0000e180,
+	0x0000e254,
+	0x0000e274,
+	0x0000e764,
+	0x0000e780,
+	0x0000e79c,
+	0x0000e7b8,
+	0x00000000,
+	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -735,7 +803,6 @@ uint32_t nva3_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0b54: memx_data_tail */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -787,15 +854,15 @@ uint32_t nva3_pwr_code[] = {
 	0x07a007f1,
 	0xd00604b6,
 	0x04bd000e,
-	0xf001e7f0,
-	0x07f101e3,
+	0xf001d7f0,
+	0x07f101d3,
 	0x04b607ac,
-	0x000ed006,
+	0x000dd006,
 /* 0x0022: rd32_wait */
-	0xe7f104bd,
-	0xe4b607ac,
-	0x00eecf06,
-	0x7000e4f1,
+	0xd7f104bd,
+	0xd4b607ac,
+	0x00ddcf06,
+	0x7000d4f1,
 	0xf1f21bf4,
 	0xb607a4d7,
 	0xddcf06d4,
@@ -807,15 +874,15 @@ uint32_t nva3_pwr_code[] = {
 	0xb607a407,
 	0x0dd00604,
 	0xf004bd00,
-	0xe5f002e7,
-	0x01e3f0f0,
+	0xd5f002d7,
+	0x01d3f0f0,
 	0x07ac07f1,
 	0xd00604b6,
-	0x04bd000e,
+	0x04bd000d,
 /* 0x006c: wr32_wait */
-	0x07ace7f1,
-	0xcf06e4b6,
-	0xe4f100ee,
+	0x07acd7f1,
+	0xcf06d4b6,
+	0xd4f100dd,
 	0x1bf47000,
 /* 0x007f: nsec */
 	0xf000f8f2,
@@ -845,21 +912,21 @@ uint32_t nva3_pwr_code[] = {
 	0x9800f8df,
 	0x96b003e9,
 	0x2a0bf400,
-	0xbb840a98,
+	0xbb9a0a98,
 	0x1cf4029a,
 	0x01d7f00f,
 	0x025421f5,
 	0x0ef494bd,
 /* 0x00e9: intr_watchdog_next_time */
-	0x850a9815,
+	0x9b0a9815,
 	0xf400a6b0,
 	0x9ab8090b,
 	0x061cf406,
 /* 0x00f8: intr_watchdog_next_time_set */
 /* 0x00fb: intr_watchdog_next_proc */
-	0x80850980,
+	0x809b0980,
 	0xe0b603e9,
-	0x10e6b158,
+	0x68e6b158,
 	0xc61bf402,
 /* 0x010a: intr */
 	0x00f900f8,
@@ -880,15 +947,15 @@ uint32_t nva3_pwr_code[] = {
 	0x0088cf06,
 	0xf40289c4,
 	0x0080230b,
-	0x58e7f085,
+	0x58e7f09b,
 	0x98cb21f4,
-	0x96b08509,
+	0x96b09b09,
 	0x110bf400,
 	0xb63407f0,
 	0x09d00604,
 	0x8004bd00,
 /* 0x016e: intr_skip_watchdog */
-	0x89e48409,
+	0x89e49a09,
 	0x0bf40800,
 	0x8897f148,
 	0x0694b606,
@@ -948,7 +1015,7 @@ uint32_t nva3_pwr_code[] = {
 	0x000ed006,
 	0x0e8004bd,
 /* 0x0241: timer_enable */
-	0x0187f084,
+	0x0187f09a,
 	0xb63807f0,
 	0x08d00604,
 /* 0x024f: timer_done */
@@ -979,7 +1046,7 @@ uint32_t nva3_pwr_code[] = {
 	0xb8008a98,
 	0x0bf406ae,
 	0x5880b610,
-	0x021086b1,
+	0x026886b1,
 	0xf4f01bf4,
 /* 0x02b2: find_done */
 	0x8eb90132,
@@ -1049,7 +1116,7 @@ uint32_t nva3_pwr_code[] = {
 	0x320bf406,
 	0x94071ec4,
 	0xe0b704ee,
-	0xeb980218,
+	0xeb980270,
 	0x02ec9803,
 	0x9801ed98,
 	0x21f500ee,
@@ -1075,7 +1142,7 @@ uint32_t nva3_pwr_code[] = {
 	0xe60bf406,
 	0xb60723c4,
 	0x30b70434,
-	0x3b800298,
+	0x3b8002f0,
 	0x023c8003,
 	0x80013d80,
 	0x20b6003e,
@@ -1090,13 +1157,13 @@ uint32_t nva3_pwr_code[] = {
 /* 0x0430: host_init */
 	0x008017f1,
 	0xf11014b6,
-	0xf1021815,
+	0xf1027015,
 	0xb604d007,
 	0x01d00604,
 	0xf104bd00,
 	0xb6008017,
 	0x15f11014,
-	0x07f10298,
+	0x07f102f0,
 	0x04b604dc,
 	0x0001d006,
 	0x17f004bd,
@@ -1156,14 +1223,14 @@ uint32_t nva3_pwr_code[] = {
 	0x00139802,
 	0x950410b6,
 	0x30f01034,
-	0xc835980c,
+	0xde35980c,
 	0x12b855f9,
 	0xec1ef406,
 	0xe0fcd0fc,
 	0x02b921f5,
 /* 0x0532: memx_info */
 	0xc7f100f8,
-	0xb7f10354,
+	0xb7f103ac,
 	0x21f50800,
 	0x00f802b9,
 /* 0x0540: memx_recv */
@@ -1175,7 +1242,312 @@ uint32_t nva3_pwr_code[] = {
 /* 0x0550: perf_recv */
 /* 0x0552: perf_init */
 	0x00f800f8,
-/* 0x0554: test_recv */
+/* 0x0554: i2c_drive_scl */
+	0xf40036b0,
+	0x07f1110b,
+	0x04b607e0,
+	0x0001d006,
+	0x00f804bd,
+/* 0x0568: i2c_drive_scl_lo */
+	0x07e407f1,
+	0xd00604b6,
+	0x04bd0001,
+/* 0x0576: i2c_drive_sda */
+	0x36b000f8,
+	0x110bf400,
+	0x07e007f1,
+	0xd00604b6,
+	0x04bd0002,
+/* 0x058a: i2c_drive_sda_lo */
+	0x07f100f8,
+	0x04b607e4,
+	0x0002d006,
+	0x00f804bd,
+/* 0x0598: i2c_sense_scl */
+	0xf10132f4,
+	0xb607c437,
+	0x33cf0634,
+	0x0431fd00,
+	0xf4060bf4,
+/* 0x05ae: i2c_sense_scl_done */
+	0x00f80131,
+/* 0x05b0: i2c_sense_sda */
+	0xf10132f4,
+	0xb607c437,
+	0x33cf0634,
+	0x0432fd00,
+	0xf4060bf4,
+/* 0x05c6: i2c_sense_sda_done */
+	0x00f80131,
+/* 0x05c8: i2c_raise_scl */
+	0x47f140f9,
+	0x37f00898,
+	0x5421f501,
+/* 0x05d5: i2c_raise_scl_wait */
+	0xe8e7f105,
+	0x7f21f403,
+	0x059821f5,
+	0xb60901f4,
+	0x1bf40142,
+/* 0x05e9: i2c_raise_scl_done */
+	0xf840fcef,
+/* 0x05ed: i2c_start */
+	0x9821f500,
+	0x0d11f405,
+	0x05b021f5,
+	0xf40611f4,
+/* 0x05fe: i2c_start_rep */
+	0x37f0300e,
+	0x5421f500,
+	0x0137f005,
+	0x057621f5,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0xc821f550,
+	0x0464b605,
+/* 0x062b: i2c_start_send */
+	0xf01f11f4,
+	0x21f50037,
+	0xe7f10576,
+	0x21f41388,
+	0x0037f07f,
+	0x055421f5,
+	0x1388e7f1,
+/* 0x0647: i2c_start_out */
+	0xf87f21f4,
+/* 0x0649: i2c_stop */
+	0x0037f000,
+	0x055421f5,
+	0xf50037f0,
+	0xf1057621,
+	0xf403e8e7,
+	0x37f07f21,
+	0x5421f501,
+	0x88e7f105,
+	0x7f21f413,
+	0xf50137f0,
+	0xf1057621,
+	0xf41388e7,
+	0x00f87f21,
+/* 0x067c: i2c_bitw */
+	0x057621f5,
+	0x03e8e7f1,
+	0xbb7f21f4,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x05c821f5,
+	0xf40464b6,
+	0xe7f11811,
+	0x21f41388,
+	0x0037f07f,
+	0x055421f5,
+	0x1388e7f1,
+/* 0x06bb: i2c_bitw_out */
+	0xf87f21f4,
+/* 0x06bd: i2c_bitr */
+	0x0137f000,
+	0x057621f5,
+	0x03e8e7f1,
+	0xbb7f21f4,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x05c821f5,
+	0xf40464b6,
+	0x21f51b11,
+	0x37f005b0,
+	0x5421f500,
+	0x88e7f105,
+	0x7f21f413,
+	0xf4013cf0,
+/* 0x0702: i2c_bitr_done */
+	0x00f80131,
+/* 0x0704: i2c_get_byte */
+	0xf00057f0,
+/* 0x070a: i2c_get_byte_next */
+	0x54b60847,
+	0x0076bb01,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0x21f550fc,
+	0x64b606bd,
+	0x2b11f404,
+	0xb60553fd,
+	0x1bf40142,
+	0x0137f0d8,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0x7c21f550,
+	0x0464b606,
+/* 0x0754: i2c_get_byte_done */
+/* 0x0756: i2c_put_byte */
+	0x47f000f8,
+/* 0x0759: i2c_put_byte_next */
+	0x0142b608,
+	0xbb3854ff,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x067c21f5,
+	0xf40464b6,
+	0x46b03411,
+	0xd81bf400,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0xbd21f550,
+	0x0464b606,
+	0xbb0f11f4,
+	0x36b00076,
+	0x061bf401,
+/* 0x07af: i2c_put_byte_done */
+	0xf80132f4,
+/* 0x07b1: i2c_addr */
+	0x0076bb00,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0x21f550fc,
+	0x64b605ed,
+	0x2911f404,
+	0x012ec3e7,
+	0xfd0134b6,
+	0x76bb0553,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0xf550fc04,
+	0xb6075621,
+/* 0x07f6: i2c_addr_done */
+	0x00f80464,
+/* 0x07f8: i2c_acquire_addr */
+	0xb6f8cec7,
+	0xe0b702e4,
+	0xee980bfc,
+/* 0x0807: i2c_acquire */
+	0xf500f800,
+	0xf407f821,
+	0xd9f00421,
+	0x3f21f403,
+/* 0x0816: i2c_release */
+	0x21f500f8,
+	0x21f407f8,
+	0x03daf004,
+	0xf83f21f4,
+/* 0x0825: i2c_recv */
+	0x0132f400,
+	0xb6f8c1c7,
+	0x16b00214,
+	0x3a1ff528,
+	0xd413a001,
+	0x0032980b,
+	0x0bac13a0,
+	0xf4003198,
+	0xd0f90231,
+	0xd0f9e0f9,
+	0x000067f1,
+	0x100063f1,
+	0xbb016792,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x080721f5,
+	0xfc0464b6,
+	0x00d6b0d0,
+	0x00b31bf5,
+	0xbb0057f0,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x07b121f5,
+	0xf50464b6,
+	0xc700d011,
+	0x76bbe0c5,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0xf550fc04,
+	0xb6075621,
+	0x11f50464,
+	0x57f000ad,
+	0x0076bb01,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0x21f550fc,
+	0x64b607b1,
+	0x8a11f504,
+	0x0076bb00,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0x21f550fc,
+	0x64b60704,
+	0x6a11f404,
+	0xbbe05bcb,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x064921f5,
+	0xb90464b6,
+	0x74bd025b,
+/* 0x092b: i2c_recv_not_rd08 */
+	0xb0430ef4,
+	0x1bf401d6,
+	0x0057f03d,
+	0x07b121f5,
+	0xc73311f4,
+	0x21f5e0c5,
+	0x11f40756,
+	0x0057f029,
+	0x07b121f5,
+	0xc71f11f4,
+	0x21f5e0b5,
+	0x11f40756,
+	0x4921f515,
+	0xc774bd06,
+	0x1bf408c5,
+	0x0232f409,
+/* 0x096b: i2c_recv_not_wr08 */
+/* 0x096b: i2c_recv_done */
+	0xc7030ef4,
+	0x21f5f8ce,
+	0xe0fc0816,
+	0x12f4d0fc,
+	0x027cb90a,
+	0x02b921f5,
+/* 0x0980: i2c_recv_exit */
+/* 0x0982: i2c_init */
+	0x00f800f8,
+/* 0x0984: test_recv */
 	0x05d817f1,
 	0xcf0614b6,
 	0x10b60011,
@@ -1185,12 +1557,12 @@ uint32_t nva3_pwr_code[] = {
 	0x00e7f104,
 	0x4fe3f1d9,
 	0xf521f513,
-/* 0x057b: test_init */
+/* 0x09ab: test_init */
 	0xf100f801,
 	0xf50800e7,
 	0xf801f521,
-/* 0x0585: idle_recv */
-/* 0x0587: idle */
+/* 0x09b5: idle_recv */
+/* 0x09b7: idle */
 	0xf400f800,
 	0x17f10031,
 	0x14b605d4,
@@ -1198,32 +1570,20 @@ uint32_t nva3_pwr_code[] = {
 	0xf10110b6,
 	0xb605d407,
 	0x01d00604,
-/* 0x05a3: idle_loop */
+/* 0x09d3: idle_loop */
 	0xf004bd00,
 	0x32f45817,
-/* 0x05a9: idle_proc */
-/* 0x05a9: idle_proc_exec */
+/* 0x09d9: idle_proc */
+/* 0x09d9: idle_proc_exec */
 	0xb910f902,
 	0x21f5021e,
 	0x10fc02c2,
 	0xf40911f4,
 	0x0ef40231,
-/* 0x05bd: idle_proc_next */
+/* 0x09ed: idle_proc_next */
 	0x5810b6ef,
 	0xf4061fb8,
 	0x02f4e61b,
 	0x0028f4dd,
 	0x00bb0ef4,
-	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 eaa64da68e36..48f79434a449 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvc0.fuc
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvc0.fuc
@@ -37,6 +37,7 @@
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
+#include "i2c_.fuc"
 #include "test.fuc"
 #include "idle.fuc"
 #undef INCLUDE_PROC
@@ -46,6 +47,7 @@
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
+#include "i2c_.fuc"
 #include "test.fuc"
 #include "idle.fuc"
 #undef INCLUDE_DATA
@@ -57,6 +59,7 @@
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
+#include "i2c_.fuc"
 #include "test.fuc"
 #include "idle.fuc"
 #undef INCLUDE_CODE
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 82c8e8b88917..4dba00d2dd1a 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
@@ -89,9 +89,31 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+	0x5f433249,
+	0x00000982,
+	0x00000825,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 	0x54534554,
-	0x0000057b,
-	0x00000554,
+	0x000009ab,
+	0x00000984,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -112,8 +134,8 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x454c4449,
-	0x00000587,
-	0x00000585,
+	0x000009b7,
+	0x000009b5,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -133,12 +155,12 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0210: proc_list_tail */
-/* 0x0210: time_prev */
+/* 0x0268: proc_list_tail */
+/* 0x0268: time_prev */
 	0x00000000,
-/* 0x0214: time_next */
+/* 0x026c: time_next */
 	0x00000000,
-/* 0x0218: fifo_queue */
+/* 0x0270: fifo_queue */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -171,7 +193,7 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0298: rfifo_queue */
+/* 0x02f0: rfifo_queue */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -204,11 +226,11 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0318: memx_func_head */
+/* 0x0370: memx_func_head */
 	0x00010000,
 	0x00000000,
 	0x0000046f,
-/* 0x0324: memx_func_next */
+/* 0x037c: memx_func_next */
 	0x00000001,
 	0x00000000,
 	0x00000496,
@@ -221,8 +243,18 @@ uint32_t nvc0_pwr_data[] = {
 	0x00010004,
 	0x00000000,
 	0x000004fc,
-/* 0x0354: memx_func_tail */
-/* 0x0354: memx_data_head */
+/* 0x03ac: memx_func_tail */
+/* 0x03ac: memx_data_head */
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
+	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -725,6 +757,42 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+/* 0x0bac: memx_data_tail */
+/* 0x0bac: i2c_scl_map */
+	0x00001000,
+	0x00004000,
+	0x00010000,
+	0x00000100,
+	0x00040000,
+	0x00100000,
+	0x00400000,
+	0x01000000,
+	0x04000000,
+	0x10000000,
+/* 0x0bd4: i2c_sda_map */
+	0x00002000,
+	0x00008000,
+	0x00020000,
+	0x00000200,
+	0x00080000,
+	0x00200000,
+	0x00800000,
+	0x02000000,
+	0x08000000,
+	0x20000000,
+/* 0x0bfc: i2c_ctrl */
+	0x0000e138,
+	0x0000e150,
+	0x0000e168,
+	0x0000e180,
+	0x0000e254,
+	0x0000e274,
+	0x0000e764,
+	0x0000e780,
+	0x0000e79c,
+	0x0000e7b8,
+	0x00000000,
+	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -735,7 +803,6 @@ uint32_t nvc0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0b54: memx_data_tail */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -787,15 +854,15 @@ uint32_t nvc0_pwr_code[] = {
 	0x07a007f1,
 	0xd00604b6,
 	0x04bd000e,
-	0xf001e7f0,
-	0x07f101e3,
+	0xf001d7f0,
+	0x07f101d3,
 	0x04b607ac,
-	0x000ed006,
+	0x000dd006,
 /* 0x0022: rd32_wait */
-	0xe7f104bd,
-	0xe4b607ac,
-	0x00eecf06,
-	0x7000e4f1,
+	0xd7f104bd,
+	0xd4b607ac,
+	0x00ddcf06,
+	0x7000d4f1,
 	0xf1f21bf4,
 	0xb607a4d7,
 	0xddcf06d4,
@@ -807,15 +874,15 @@ uint32_t nvc0_pwr_code[] = {
 	0xb607a407,
 	0x0dd00604,
 	0xf004bd00,
-	0xe5f002e7,
-	0x01e3f0f0,
+	0xd5f002d7,
+	0x01d3f0f0,
 	0x07ac07f1,
 	0xd00604b6,
-	0x04bd000e,
+	0x04bd000d,
 /* 0x006c: wr32_wait */
-	0x07ace7f1,
-	0xcf06e4b6,
-	0xe4f100ee,
+	0x07acd7f1,
+	0xcf06d4b6,
+	0xd4f100dd,
 	0x1bf47000,
 /* 0x007f: nsec */
 	0xf000f8f2,
@@ -845,21 +912,21 @@ uint32_t nvc0_pwr_code[] = {
 	0x9800f8df,
 	0x96b003e9,
 	0x2a0bf400,
-	0xbb840a98,
+	0xbb9a0a98,
 	0x1cf4029a,
 	0x01d7f00f,
 	0x025421f5,
 	0x0ef494bd,
 /* 0x00e9: intr_watchdog_next_time */
-	0x850a9815,
+	0x9b0a9815,
 	0xf400a6b0,
 	0x9ab8090b,
 	0x061cf406,
 /* 0x00f8: intr_watchdog_next_time_set */
 /* 0x00fb: intr_watchdog_next_proc */
-	0x80850980,
+	0x809b0980,
 	0xe0b603e9,
-	0x10e6b158,
+	0x68e6b158,
 	0xc61bf402,
 /* 0x010a: intr */
 	0x00f900f8,
@@ -880,15 +947,15 @@ uint32_t nvc0_pwr_code[] = {
 	0x0088cf06,
 	0xf40289c4,
 	0x0080230b,
-	0x58e7f085,
+	0x58e7f09b,
 	0x98cb21f4,
-	0x96b08509,
+	0x96b09b09,
 	0x110bf400,
 	0xb63407f0,
 	0x09d00604,
 	0x8004bd00,
 /* 0x016e: intr_skip_watchdog */
-	0x89e48409,
+	0x89e49a09,
 	0x0bf40800,
 	0x8897f148,
 	0x0694b606,
@@ -948,7 +1015,7 @@ uint32_t nvc0_pwr_code[] = {
 	0x000ed006,
 	0x0e8004bd,
 /* 0x0241: timer_enable */
-	0x0187f084,
+	0x0187f09a,
 	0xb63807f0,
 	0x08d00604,
 /* 0x024f: timer_done */
@@ -979,7 +1046,7 @@ uint32_t nvc0_pwr_code[] = {
 	0xb8008a98,
 	0x0bf406ae,
 	0x5880b610,
-	0x021086b1,
+	0x026886b1,
 	0xf4f01bf4,
 /* 0x02b2: find_done */
 	0x8eb90132,
@@ -1049,7 +1116,7 @@ uint32_t nvc0_pwr_code[] = {
 	0x320bf406,
 	0x94071ec4,
 	0xe0b704ee,
-	0xeb980218,
+	0xeb980270,
 	0x02ec9803,
 	0x9801ed98,
 	0x21f500ee,
@@ -1075,7 +1142,7 @@ uint32_t nvc0_pwr_code[] = {
 	0xe60bf406,
 	0xb60723c4,
 	0x30b70434,
-	0x3b800298,
+	0x3b8002f0,
 	0x023c8003,
 	0x80013d80,
 	0x20b6003e,
@@ -1090,13 +1157,13 @@ uint32_t nvc0_pwr_code[] = {
 /* 0x0430: host_init */
 	0x008017f1,
 	0xf11014b6,
-	0xf1021815,
+	0xf1027015,
 	0xb604d007,
 	0x01d00604,
 	0xf104bd00,
 	0xb6008017,
 	0x15f11014,
-	0x07f10298,
+	0x07f102f0,
 	0x04b604dc,
 	0x0001d006,
 	0x17f004bd,
@@ -1156,14 +1223,14 @@ uint32_t nvc0_pwr_code[] = {
 	0x00139802,
 	0x950410b6,
 	0x30f01034,
-	0xc835980c,
+	0xde35980c,
 	0x12b855f9,
 	0xec1ef406,
 	0xe0fcd0fc,
 	0x02b921f5,
 /* 0x0532: memx_info */
 	0xc7f100f8,
-	0xb7f10354,
+	0xb7f103ac,
 	0x21f50800,
 	0x00f802b9,
 /* 0x0540: memx_recv */
@@ -1175,7 +1242,312 @@ uint32_t nvc0_pwr_code[] = {
 /* 0x0550: perf_recv */
 /* 0x0552: perf_init */
 	0x00f800f8,
-/* 0x0554: test_recv */
+/* 0x0554: i2c_drive_scl */
+	0xf40036b0,
+	0x07f1110b,
+	0x04b607e0,
+	0x0001d006,
+	0x00f804bd,
+/* 0x0568: i2c_drive_scl_lo */
+	0x07e407f1,
+	0xd00604b6,
+	0x04bd0001,
+/* 0x0576: i2c_drive_sda */
+	0x36b000f8,
+	0x110bf400,
+	0x07e007f1,
+	0xd00604b6,
+	0x04bd0002,
+/* 0x058a: i2c_drive_sda_lo */
+	0x07f100f8,
+	0x04b607e4,
+	0x0002d006,
+	0x00f804bd,
+/* 0x0598: i2c_sense_scl */
+	0xf10132f4,
+	0xb607c437,
+	0x33cf0634,
+	0x0431fd00,
+	0xf4060bf4,
+/* 0x05ae: i2c_sense_scl_done */
+	0x00f80131,
+/* 0x05b0: i2c_sense_sda */
+	0xf10132f4,
+	0xb607c437,
+	0x33cf0634,
+	0x0432fd00,
+	0xf4060bf4,
+/* 0x05c6: i2c_sense_sda_done */
+	0x00f80131,
+/* 0x05c8: i2c_raise_scl */
+	0x47f140f9,
+	0x37f00898,
+	0x5421f501,
+/* 0x05d5: i2c_raise_scl_wait */
+	0xe8e7f105,
+	0x7f21f403,
+	0x059821f5,
+	0xb60901f4,
+	0x1bf40142,
+/* 0x05e9: i2c_raise_scl_done */
+	0xf840fcef,
+/* 0x05ed: i2c_start */
+	0x9821f500,
+	0x0d11f405,
+	0x05b021f5,
+	0xf40611f4,
+/* 0x05fe: i2c_start_rep */
+	0x37f0300e,
+	0x5421f500,
+	0x0137f005,
+	0x057621f5,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0xc821f550,
+	0x0464b605,
+/* 0x062b: i2c_start_send */
+	0xf01f11f4,
+	0x21f50037,
+	0xe7f10576,
+	0x21f41388,
+	0x0037f07f,
+	0x055421f5,
+	0x1388e7f1,
+/* 0x0647: i2c_start_out */
+	0xf87f21f4,
+/* 0x0649: i2c_stop */
+	0x0037f000,
+	0x055421f5,
+	0xf50037f0,
+	0xf1057621,
+	0xf403e8e7,
+	0x37f07f21,
+	0x5421f501,
+	0x88e7f105,
+	0x7f21f413,
+	0xf50137f0,
+	0xf1057621,
+	0xf41388e7,
+	0x00f87f21,
+/* 0x067c: i2c_bitw */
+	0x057621f5,
+	0x03e8e7f1,
+	0xbb7f21f4,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x05c821f5,
+	0xf40464b6,
+	0xe7f11811,
+	0x21f41388,
+	0x0037f07f,
+	0x055421f5,
+	0x1388e7f1,
+/* 0x06bb: i2c_bitw_out */
+	0xf87f21f4,
+/* 0x06bd: i2c_bitr */
+	0x0137f000,
+	0x057621f5,
+	0x03e8e7f1,
+	0xbb7f21f4,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x05c821f5,
+	0xf40464b6,
+	0x21f51b11,
+	0x37f005b0,
+	0x5421f500,
+	0x88e7f105,
+	0x7f21f413,
+	0xf4013cf0,
+/* 0x0702: i2c_bitr_done */
+	0x00f80131,
+/* 0x0704: i2c_get_byte */
+	0xf00057f0,
+/* 0x070a: i2c_get_byte_next */
+	0x54b60847,
+	0x0076bb01,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0x21f550fc,
+	0x64b606bd,
+	0x2b11f404,
+	0xb60553fd,
+	0x1bf40142,
+	0x0137f0d8,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0x7c21f550,
+	0x0464b606,
+/* 0x0754: i2c_get_byte_done */
+/* 0x0756: i2c_put_byte */
+	0x47f000f8,
+/* 0x0759: i2c_put_byte_next */
+	0x0142b608,
+	0xbb3854ff,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x067c21f5,
+	0xf40464b6,
+	0x46b03411,
+	0xd81bf400,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0xbd21f550,
+	0x0464b606,
+	0xbb0f11f4,
+	0x36b00076,
+	0x061bf401,
+/* 0x07af: i2c_put_byte_done */
+	0xf80132f4,
+/* 0x07b1: i2c_addr */
+	0x0076bb00,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0x21f550fc,
+	0x64b605ed,
+	0x2911f404,
+	0x012ec3e7,
+	0xfd0134b6,
+	0x76bb0553,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0xf550fc04,
+	0xb6075621,
+/* 0x07f6: i2c_addr_done */
+	0x00f80464,
+/* 0x07f8: i2c_acquire_addr */
+	0xb6f8cec7,
+	0xe0b702e4,
+	0xee980bfc,
+/* 0x0807: i2c_acquire */
+	0xf500f800,
+	0xf407f821,
+	0xd9f00421,
+	0x3f21f403,
+/* 0x0816: i2c_release */
+	0x21f500f8,
+	0x21f407f8,
+	0x03daf004,
+	0xf83f21f4,
+/* 0x0825: i2c_recv */
+	0x0132f400,
+	0xb6f8c1c7,
+	0x16b00214,
+	0x3a1ff528,
+	0xd413a001,
+	0x0032980b,
+	0x0bac13a0,
+	0xf4003198,
+	0xd0f90231,
+	0xd0f9e0f9,
+	0x000067f1,
+	0x100063f1,
+	0xbb016792,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x080721f5,
+	0xfc0464b6,
+	0x00d6b0d0,
+	0x00b31bf5,
+	0xbb0057f0,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x07b121f5,
+	0xf50464b6,
+	0xc700d011,
+	0x76bbe0c5,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0xf550fc04,
+	0xb6075621,
+	0x11f50464,
+	0x57f000ad,
+	0x0076bb01,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0x21f550fc,
+	0x64b607b1,
+	0x8a11f504,
+	0x0076bb00,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0x21f550fc,
+	0x64b60704,
+	0x6a11f404,
+	0xbbe05bcb,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x064921f5,
+	0xb90464b6,
+	0x74bd025b,
+/* 0x092b: i2c_recv_not_rd08 */
+	0xb0430ef4,
+	0x1bf401d6,
+	0x0057f03d,
+	0x07b121f5,
+	0xc73311f4,
+	0x21f5e0c5,
+	0x11f40756,
+	0x0057f029,
+	0x07b121f5,
+	0xc71f11f4,
+	0x21f5e0b5,
+	0x11f40756,
+	0x4921f515,
+	0xc774bd06,
+	0x1bf408c5,
+	0x0232f409,
+/* 0x096b: i2c_recv_not_wr08 */
+/* 0x096b: i2c_recv_done */
+	0xc7030ef4,
+	0x21f5f8ce,
+	0xe0fc0816,
+	0x12f4d0fc,
+	0x027cb90a,
+	0x02b921f5,
+/* 0x0980: i2c_recv_exit */
+/* 0x0982: i2c_init */
+	0x00f800f8,
+/* 0x0984: test_recv */
 	0x05d817f1,
 	0xcf0614b6,
 	0x10b60011,
@@ -1185,12 +1557,12 @@ uint32_t nvc0_pwr_code[] = {
 	0x00e7f104,
 	0x4fe3f1d9,
 	0xf521f513,
-/* 0x057b: test_init */
+/* 0x09ab: test_init */
 	0xf100f801,
 	0xf50800e7,
 	0xf801f521,
-/* 0x0585: idle_recv */
-/* 0x0587: idle */
+/* 0x09b5: idle_recv */
+/* 0x09b7: idle */
 	0xf400f800,
 	0x17f10031,
 	0x14b605d4,
@@ -1198,32 +1570,20 @@ uint32_t nvc0_pwr_code[] = {
 	0xf10110b6,
 	0xb605d407,
 	0x01d00604,
-/* 0x05a3: idle_loop */
+/* 0x09d3: idle_loop */
 	0xf004bd00,
 	0x32f45817,
-/* 0x05a9: idle_proc */
-/* 0x05a9: idle_proc_exec */
+/* 0x09d9: idle_proc */
+/* 0x09d9: idle_proc_exec */
 	0xb910f902,
 	0x21f5021e,
 	0x10fc02c2,
 	0xf40911f4,
 	0x0ef40231,
-/* 0x05bd: idle_proc_next */
+/* 0x09ed: idle_proc_next */
 	0x5810b6ef,
 	0xf4061fb8,
 	0x02f4e61b,
 	0x0028f4dd,
 	0x00bb0ef4,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	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 32d65ea254dd..8a89dfe41ce1 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvd0.fuc
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/nvd0.fuc
@@ -37,6 +37,7 @@
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
+#include "i2c_.fuc"
 #include "test.fuc"
 #include "idle.fuc"
 #undef INCLUDE_PROC
@@ -46,6 +47,7 @@
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
+#include "i2c_.fuc"
 #include "test.fuc"
 #include "idle.fuc"
 #undef INCLUDE_DATA
@@ -57,6 +59,7 @@
 #include "host.fuc"
 #include "memx.fuc"
 #include "perf.fuc"
+#include "i2c_.fuc"
 #include "test.fuc"
 #include "idle.fuc"
 #undef INCLUDE_CODE
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 ce65e2a4b789..5e24c6bc041d 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
@@ -89,33 +89,13 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-	0x54534554,
-	0x000004eb,
-	0x000004ca,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
+	0x5f433249,
+	0x000008e3,
+	0x00000786,
 	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
-	0x454c4449,
-	0x000004f7,
-	0x000004f5,
-	0x00000000,
-	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -131,14 +111,13 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+	0x54534554,
+	0x00000906,
+	0x000008e5,
 	0x00000000,
 	0x00000000,
-/* 0x0210: proc_list_tail */
-/* 0x0210: time_prev */
 	0x00000000,
-/* 0x0214: time_next */
 	0x00000000,
-/* 0x0218: fifo_queue */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -154,6 +133,9 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+	0x454c4449,
+	0x00000912,
+	0x00000910,
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -171,11 +153,14 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0298: rfifo_queue */
 	0x00000000,
 	0x00000000,
+/* 0x0268: proc_list_tail */
+/* 0x0268: time_prev */
 	0x00000000,
+/* 0x026c: time_next */
 	0x00000000,
+/* 0x0270: fifo_queue */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -204,31 +189,11 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0318: memx_func_head */
-	0x00010000,
-	0x00000000,
-	0x000003f4,
-/* 0x0324: memx_func_next */
-	0x00000001,
-	0x00000000,
-	0x00000415,
-	0x00000002,
-	0x00000002,
-	0x00000430,
-	0x00040003,
-	0x00000000,
-	0x00000458,
-	0x00010004,
-	0x00000000,
-	0x00000472,
-/* 0x0354: memx_func_tail */
-/* 0x0354: memx_data_head */
-	0x00000000,
-	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
 	0x00000000,
+/* 0x02f0: rfifo_queue */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -261,10 +226,25 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+/* 0x0370: memx_func_head */
+	0x00010000,
 	0x00000000,
+	0x000003f4,
+/* 0x037c: memx_func_next */
+	0x00000001,
 	0x00000000,
+	0x00000415,
+	0x00000002,
+	0x00000002,
+	0x00000430,
+	0x00040003,
 	0x00000000,
+	0x00000458,
+	0x00010004,
 	0x00000000,
+	0x00000472,
+/* 0x03ac: memx_func_tail */
+/* 0x03ac: memx_data_head */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -735,7 +715,6 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
-/* 0x0b54: memx_data_tail */
 	0x00000000,
 	0x00000000,
 	0x00000000,
@@ -778,6 +757,29 @@ uint32_t nvd0_pwr_data[] = {
 	0x00000000,
 	0x00000000,
 	0x00000000,
+/* 0x0bac: memx_data_tail */
+/* 0x0bac: i2c_scl_map */
+	0x00000400,
+	0x00000800,
+	0x00001000,
+	0x00002000,
+	0x00004000,
+	0x00008000,
+	0x00010000,
+	0x00020000,
+	0x00040000,
+	0x00080000,
+/* 0x0bd4: i2c_sda_map */
+	0x00100000,
+	0x00200000,
+	0x00400000,
+	0x00800000,
+	0x01000000,
+	0x02000000,
+	0x04000000,
+	0x08000000,
+	0x10000000,
+	0x20000000,
 	0x00000000,
 };
 
@@ -786,14 +788,14 @@ uint32_t nvd0_pwr_code[] = {
 /* 0x0004: rd32 */
 	0x07a007f1,
 	0xbd000ed0,
-	0x01e7f004,
-	0xf101e3f0,
+	0x01d7f004,
+	0xf101d3f0,
 	0xd007ac07,
-	0x04bd000e,
+	0x04bd000d,
 /* 0x001c: rd32_wait */
-	0x07ace7f1,
-	0xf100eecf,
-	0xf47000e4,
+	0x07acd7f1,
+	0xf100ddcf,
+	0xf47000d4,
 	0xd7f1f51b,
 	0xddcf07a4,
 /* 0x0033: wr32 */
@@ -802,14 +804,14 @@ uint32_t nvd0_pwr_code[] = {
 	0x04bd000e,
 	0x07a407f1,
 	0xbd000dd0,
-	0x02e7f004,
-	0xf0f0e5f0,
-	0x07f101e3,
-	0x0ed007ac,
+	0x02d7f004,
+	0xf0f0d5f0,
+	0x07f101d3,
+	0x0dd007ac,
 /* 0x0057: wr32_wait */
 	0xf104bd00,
-	0xcf07ace7,
-	0xe4f100ee,
+	0xcf07acd7,
+	0xd4f100dd,
 	0x1bf47000,
 /* 0x0067: nsec */
 	0xf000f8f5,
@@ -836,21 +838,21 @@ uint32_t nvd0_pwr_code[] = {
 	0x9800f8e2,
 	0x96b003e9,
 	0x2a0bf400,
-	0xbb840a98,
+	0xbb9a0a98,
 	0x1cf4029a,
 	0x01d7f00f,
 	0x020621f5,
 	0x0ef494bd,
 /* 0x00c5: intr_watchdog_next_time */
-	0x850a9815,
+	0x9b0a9815,
 	0xf400a6b0,
 	0x9ab8090b,
 	0x061cf406,
 /* 0x00d4: intr_watchdog_next_time_set */
 /* 0x00d7: intr_watchdog_next_proc */
-	0x80850980,
+	0x809b0980,
 	0xe0b603e9,
-	0x10e6b158,
+	0x68e6b158,
 	0xc61bf402,
 /* 0x00e6: intr */
 	0x00f900f8,
@@ -868,15 +870,15 @@ uint32_t nvd0_pwr_code[] = {
 	0x0887f004,
 	0xc40088cf,
 	0x0bf40289,
-	0x85008020,
+	0x9b008020,
 	0xf458e7f0,
 	0x0998a721,
-	0x0096b085,
+	0x0096b09b,
 	0xf00e0bf4,
 	0x09d03407,
 	0x8004bd00,
 /* 0x013e: intr_skip_watchdog */
-	0x89e48409,
+	0x89e49a09,
 	0x0bf40800,
 	0x8897f13c,
 	0x0099cf06,
@@ -929,7 +931,7 @@ uint32_t nvd0_pwr_code[] = {
 	0x0ed03407,
 	0x8004bd00,
 /* 0x01f6: timer_enable */
-	0x87f0840e,
+	0x87f09a0e,
 	0x3807f001,
 	0xbd0008d0,
 /* 0x0201: timer_done */
@@ -960,7 +962,7 @@ uint32_t nvd0_pwr_code[] = {
 	0x06aeb800,
 	0xb6100bf4,
 	0x86b15880,
-	0x1bf40210,
+	0x1bf40268,
 	0x0132f4f0,
 /* 0x0264: find_done */
 	0xfc028eb9,
@@ -1024,7 +1026,7 @@ uint32_t nvd0_pwr_code[] = {
 	0x0bf40612,
 	0x071ec42f,
 	0xb704ee94,
-	0x980218e0,
+	0x980270e0,
 	0xec9803eb,
 	0x01ed9802,
 	0xf500ee98,
@@ -1048,7 +1050,7 @@ uint32_t nvd0_pwr_code[] = {
 	0xec0bf406,
 	0xb60723c4,
 	0x30b70434,
-	0x3b800298,
+	0x3b8002f0,
 	0x023c8003,
 	0x80013d80,
 	0x20b6003e,
@@ -1061,12 +1063,12 @@ uint32_t nvd0_pwr_code[] = {
 /* 0x03be: host_init */
 	0x17f100f8,
 	0x14b60080,
-	0x1815f110,
+	0x7015f110,
 	0xd007f102,
 	0x0001d004,
 	0x17f104bd,
 	0x14b60080,
-	0x9815f110,
+	0xf015f110,
 	0xdc07f102,
 	0x0001d004,
 	0x17f004bd,
@@ -1122,13 +1124,13 @@ uint32_t nvd0_pwr_code[] = {
 	0x10b60013,
 	0x10349504,
 	0x980c30f0,
-	0x55f9c835,
+	0x55f9de35,
 	0xf40612b8,
 	0xd0fcec1e,
 	0x21f5e0fc,
 	0x00f8026b,
 /* 0x04a8: memx_info */
-	0x0354c7f1,
+	0x03acc7f1,
 	0x0800b7f1,
 	0x026b21f5,
 /* 0x04b6: memx_recv */
@@ -1140,49 +1142,342 @@ uint32_t nvd0_pwr_code[] = {
 /* 0x04c6: perf_recv */
 	0x00f800f8,
 /* 0x04c8: perf_init */
-/* 0x04ca: test_recv */
-	0x17f100f8,
-	0x11cf05d8,
-	0x0110b600,
-	0x05d807f1,
+/* 0x04ca: i2c_drive_scl */
+	0x36b000f8,
+	0x0e0bf400,
+	0x07e007f1,
 	0xbd0001d0,
-	0x00e7f104,
-	0x4fe3f1d9,
-	0xb621f513,
-/* 0x04eb: test_init */
-	0xf100f801,
-	0xf50800e7,
-	0xf801b621,
-/* 0x04f5: idle_recv */
-/* 0x04f7: idle */
-	0xf400f800,
-	0x17f10031,
-	0x11cf05d4,
-	0x0110b600,
-	0x05d407f1,
-	0xbd0001d0,
-/* 0x050d: idle_loop */
-	0x5817f004,
-/* 0x0513: idle_proc */
-/* 0x0513: idle_proc_exec */
-	0xf90232f4,
-	0x021eb910,
-	0x027421f5,
-	0x11f410fc,
-	0x0231f409,
-/* 0x0527: idle_proc_next */
-	0xb6ef0ef4,
-	0x1fb85810,
-	0xe61bf406,
-	0xf4dd02f4,
-	0x0ef40028,
-	0x000000c1,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
-	0x00000000,
+/* 0x04db: i2c_drive_scl_lo */
+	0xf100f804,
+	0xd007e407,
+	0x04bd0001,
+/* 0x04e6: i2c_drive_sda */
+	0x36b000f8,
+	0x0e0bf400,
+	0x07e007f1,
+	0xbd0002d0,
+/* 0x04f7: i2c_drive_sda_lo */
+	0xf100f804,
+	0xd007e407,
+	0x04bd0002,
+/* 0x0502: i2c_sense_scl */
+	0x32f400f8,
+	0xc437f101,
+	0x0033cf07,
+	0xf40431fd,
+	0x31f4060b,
+/* 0x0515: i2c_sense_scl_done */
+/* 0x0517: i2c_sense_sda */
+	0xf400f801,
+	0x37f10132,
+	0x33cf07c4,
+	0x0432fd00,
+	0xf4060bf4,
+/* 0x052a: i2c_sense_sda_done */
+	0x00f80131,
+/* 0x052c: i2c_raise_scl */
+	0x47f140f9,
+	0x37f00898,
+	0xca21f501,
+/* 0x0539: i2c_raise_scl_wait */
+	0xe8e7f104,
+	0x6721f403,
+	0x050221f5,
+	0xb60901f4,
+	0x1bf40142,
+/* 0x054d: i2c_raise_scl_done */
+	0xf840fcef,
+/* 0x0551: i2c_start */
+	0x0221f500,
+	0x0d11f405,
+	0x051721f5,
+	0xf40611f4,
+/* 0x0562: i2c_start_rep */
+	0x37f0300e,
+	0xca21f500,
+	0x0137f004,
+	0x04e621f5,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0x2c21f550,
+	0x0464b605,
+/* 0x058f: i2c_start_send */
+	0xf01f11f4,
+	0x21f50037,
+	0xe7f104e6,
+	0x21f41388,
+	0x0037f067,
+	0x04ca21f5,
+	0x1388e7f1,
+/* 0x05ab: i2c_start_out */
+	0xf86721f4,
+/* 0x05ad: i2c_stop */
+	0x0037f000,
+	0x04ca21f5,
+	0xf50037f0,
+	0xf104e621,
+	0xf403e8e7,
+	0x37f06721,
+	0xca21f501,
+	0x88e7f104,
+	0x6721f413,
+	0xf50137f0,
+	0xf104e621,
+	0xf41388e7,
+	0x00f86721,
+/* 0x05e0: i2c_bitw */
+	0x04e621f5,
+	0x03e8e7f1,
+	0xbb6721f4,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x052c21f5,
+	0xf40464b6,
+	0xe7f11811,
+	0x21f41388,
+	0x0037f067,
+	0x04ca21f5,
+	0x1388e7f1,
+/* 0x061f: i2c_bitw_out */
+	0xf86721f4,
+/* 0x0621: i2c_bitr */
+	0x0137f000,
+	0x04e621f5,
+	0x03e8e7f1,
+	0xbb6721f4,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x052c21f5,
+	0xf40464b6,
+	0x21f51b11,
+	0x37f00517,
+	0xca21f500,
+	0x88e7f104,
+	0x6721f413,
+	0xf4013cf0,
+/* 0x0666: i2c_bitr_done */
+	0x00f80131,
+/* 0x0668: i2c_get_byte */
+	0xf00057f0,
+/* 0x066e: i2c_get_byte_next */
+	0x54b60847,
+	0x0076bb01,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0x21f550fc,
+	0x64b60621,
+	0x2b11f404,
+	0xb60553fd,
+	0x1bf40142,
+	0x0137f0d8,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0xe021f550,
+	0x0464b605,
+/* 0x06b8: i2c_get_byte_done */
+/* 0x06ba: i2c_put_byte */
+	0x47f000f8,
+/* 0x06bd: i2c_put_byte_next */
+	0x0142b608,
+	0xbb3854ff,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x05e021f5,
+	0xf40464b6,
+	0x46b03411,
+	0xd81bf400,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0x2121f550,
+	0x0464b606,
+	0xbb0f11f4,
+	0x36b00076,
+	0x061bf401,
+/* 0x0713: i2c_put_byte_done */
+	0xf80132f4,
+/* 0x0715: i2c_addr */
+	0x0076bb00,
+	0xf90465b6,
+	0x04659450,
+	0xbd0256bb,
+	0x0475fd50,
+	0x21f550fc,
+	0x64b60551,
+	0x2911f404,
+	0x012ec3e7,
+	0xfd0134b6,
+	0x76bb0553,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0xf550fc04,
+	0xb606ba21,
+/* 0x075a: i2c_addr_done */
+	0x00f80464,
+/* 0x075c: i2c_acquire_addr */
+	0xb6f8cec7,
+	0xe0b705e4,
+	0x00f8d014,
+/* 0x0768: i2c_acquire */
+	0x075c21f5,
+	0xf00421f4,
+	0x21f403d9,
+/* 0x0777: i2c_release */
+	0xf500f833,
+	0xf4075c21,
+	0xdaf00421,
+	0x3321f403,
+/* 0x0786: i2c_recv */
+	0x32f400f8,
+	0xf8c1c701,
+	0xb00214b6,
+	0x1ff52816,
+	0x13a0013a,
+	0x32980bd4,
+	0xac13a000,
+	0x0031980b,
+	0xf90231f4,
+	0xf9e0f9d0,
+	0x0067f1d0,
+	0x0063f100,
+	0x01679210,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0x6821f550,
+	0x0464b607,
+	0xd6b0d0fc,
+	0xb31bf500,
+	0x0057f000,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0x1521f550,
+	0x0464b607,
+	0x00d011f5,
+	0xbbe0c5c7,
+	0x65b60076,
+	0x9450f904,
+	0x56bb0465,
+	0xfd50bd02,
+	0x50fc0475,
+	0x06ba21f5,
+	0xf50464b6,
+	0xf000ad11,
+	0x76bb0157,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0xf550fc04,
+	0xb6071521,
+	0x11f50464,
+	0x76bb008a,
+	0x0465b600,
+	0x659450f9,
+	0x0256bb04,
+	0x75fd50bd,
+	0xf550fc04,
+	0xb6066821,
+	0x11f40464,
+	0xe05bcb6a,
+	0xb60076bb,
+	0x50f90465,
+	0xbb046594,
+	0x50bd0256,
+	0xfc0475fd,
+	0xad21f550,
+	0x0464b605,
+	0xbd025bb9,
+	0x430ef474,
+/* 0x088c: i2c_recv_not_rd08 */
+	0xf401d6b0,
+	0x57f03d1b,
+	0x1521f500,
+	0x3311f407,
+	0xf5e0c5c7,
+	0xf406ba21,
+	0x57f02911,
+	0x1521f500,
+	0x1f11f407,
+	0xf5e0b5c7,
+	0xf406ba21,
+	0x21f51511,
+	0x74bd05ad,
+	0xf408c5c7,
+	0x32f4091b,
+	0x030ef402,
+/* 0x08cc: i2c_recv_not_wr08 */
+/* 0x08cc: i2c_recv_done */
+	0xf5f8cec7,
+	0xfc077721,
+	0xf4d0fce0,
+	0x7cb90a12,
+	0x6b21f502,
+/* 0x08e1: i2c_recv_exit */
+/* 0x08e3: i2c_init */
+	0xf800f802,
+/* 0x08e5: test_recv */
+	0xd817f100,
+	0x0011cf05,
+	0xf10110b6,
+	0xd005d807,
+	0x04bd0001,
+	0xd900e7f1,
+	0x134fe3f1,
+	0x01b621f5,
+/* 0x0906: test_init */
+	0xe7f100f8,
+	0x21f50800,
+	0x00f801b6,
+/* 0x0910: idle_recv */
+/* 0x0912: idle */
+	0x31f400f8,
+	0xd417f100,
+	0x0011cf05,
+	0xf10110b6,
+	0xd005d407,
+	0x04bd0001,
+/* 0x0928: idle_loop */
+	0xf45817f0,
+/* 0x092e: idle_proc */
+/* 0x092e: idle_proc_exec */
+	0x10f90232,
+	0xf5021eb9,
+	0xfc027421,
+	0x0911f410,
+	0xf40231f4,
+/* 0x0942: idle_proc_next */
+	0x10b6ef0e,
+	0x061fb858,
+	0xf4e61bf4,
+	0x28f4dd02,
+	0xc10ef400,
 	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 5fb0cccc6c64..574acfa44c8c 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/os.h
+++ b/drivers/gpu/drm/nouveau/core/subdev/pwr/fuc/os.h
@@ -7,6 +7,7 @@
 #define PROC_HOST 0x54534f48
 #define PROC_MEMX 0x584d454d
 #define PROC_PERF 0x46524550
+#define PROC_I2C_ 0x5f433249
 #define PROC_TEST 0x54534554
 
 /* KERN: message identifiers */
@@ -24,4 +25,22 @@
 #define MEMX_WAIT   3
 #define MEMX_DELAY  4
 
+/* I2C_: message identifiers */
+#define I2C__MSG_RD08 0
+#define I2C__MSG_WR08 1
+
+#define I2C__MSG_DATA0_PORT 24:31
+#define I2C__MSG_DATA0_ADDR 14:23
+
+#define I2C__MSG_DATA0_RD08_PORT I2C__MSG_DATA0_PORT
+#define I2C__MSG_DATA0_RD08_ADDR I2C__MSG_DATA0_ADDR
+#define I2C__MSG_DATA0_RD08_REG 0:7
+#define I2C__MSG_DATA1_RD08_VAL 0:7
+
+#define I2C__MSG_DATA0_WR08_PORT I2C__MSG_DATA0_PORT
+#define I2C__MSG_DATA0_WR08_ADDR I2C__MSG_DATA0_ADDR
+#define I2C__MSG_DATA0_WR08_SYNC 8:8
+#define I2C__MSG_DATA0_WR08_REG 0:7
+#define I2C__MSG_DATA1_WR08_VAL 0:7
+
 #endif
diff --git a/drivers/gpu/drm/nouveau/core/subdev/vm/base.c b/drivers/gpu/drm/nouveau/core/subdev/vm/base.c
index ef3133e7575c..7dd680ff2f6f 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/vm/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/vm/base.c
@@ -72,13 +72,7 @@ nouveau_vm_map_at(struct nouveau_vma *vma, u64 delta, struct nouveau_mem *node)
 	vmm->flush(vm);
 }
 
-void
-nouveau_vm_map(struct nouveau_vma *vma, struct nouveau_mem *node)
-{
-	nouveau_vm_map_at(vma, 0, node);
-}
-
-void
+static void
 nouveau_vm_map_sg_table(struct nouveau_vma *vma, u64 delta, u64 length,
 			struct nouveau_mem *mem)
 {
@@ -136,7 +130,7 @@ finish:
 	vmm->flush(vm);
 }
 
-void
+static void
 nouveau_vm_map_sg(struct nouveau_vma *vma, u64 delta, u64 length,
 		  struct nouveau_mem *mem)
 {
@@ -175,6 +169,18 @@ nouveau_vm_map_sg(struct nouveau_vma *vma, u64 delta, u64 length,
 }
 
 void
+nouveau_vm_map(struct nouveau_vma *vma, struct nouveau_mem *node)
+{
+	if (node->sg)
+		nouveau_vm_map_sg_table(vma, 0, node->size << 12, node);
+	else
+	if (node->pages)
+		nouveau_vm_map_sg(vma, 0, node->size << 12, node);
+	else
+		nouveau_vm_map_at(vma, 0, node);
+}
+
+void
 nouveau_vm_unmap_at(struct nouveau_vma *vma, u64 delta, u64 length)
 {
 	struct nouveau_vm *vm = vma->vm;
diff --git a/drivers/gpu/drm/nouveau/dispnv04/disp.c b/drivers/gpu/drm/nouveau/dispnv04/disp.c
index b13ff0fc42de..2f1ed61f7c8c 100644
--- a/drivers/gpu/drm/nouveau/dispnv04/disp.c
+++ b/drivers/gpu/drm/nouveau/dispnv04/disp.c
@@ -77,11 +77,6 @@ nv04_display_create(struct drm_device *dev)
 
 	nouveau_hw_save_vga_fonts(dev, 1);
 
-	ret = nouveau_object_new(nv_object(drm), NVDRM_DEVICE, 0xd1500000,
-				 NV04_DISP_CLASS, NULL, 0, &disp->core);
-	if (ret)
-		return ret;
-
 	nv04_crtc_create(dev, 0);
 	if (nv_two_heads(dev))
 		nv04_crtc_create(dev, 1);
diff --git a/drivers/gpu/drm/nouveau/dispnv04/disp.h b/drivers/gpu/drm/nouveau/dispnv04/disp.h
index 56a28db04000..4245fc3dab70 100644
--- a/drivers/gpu/drm/nouveau/dispnv04/disp.h
+++ b/drivers/gpu/drm/nouveau/dispnv04/disp.h
@@ -80,7 +80,6 @@ struct nv04_display {
 	struct nv04_mode_state saved_reg;
 	uint32_t saved_vga_font[4][16384];
 	uint32_t dac_users[4];
-	struct nouveau_object *core;
 	struct nouveau_bo *image[2];
 };
 
diff --git a/drivers/gpu/drm/nouveau/dispnv04/overlay.c b/drivers/gpu/drm/nouveau/dispnv04/overlay.c
index 32e7064b819b..ab03f7719d2d 100644
--- a/drivers/gpu/drm/nouveau/dispnv04/overlay.c
+++ b/drivers/gpu/drm/nouveau/dispnv04/overlay.c
@@ -55,9 +55,12 @@ struct nouveau_plane {
 	int hue;
 	int saturation;
 	int iturbt_709;
+
+	void (*set_params)(struct nouveau_plane *);
 };
 
 static uint32_t formats[] = {
+	DRM_FORMAT_YUYV,
 	DRM_FORMAT_UYVY,
 	DRM_FORMAT_NV12,
 };
@@ -140,10 +143,10 @@ nv10_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
 	nv_wr32(dev, NV_PVIDEO_POINT_OUT(flip), crtc_y << 16 | crtc_x);
 	nv_wr32(dev, NV_PVIDEO_SIZE_OUT(flip), crtc_h << 16 | crtc_w);
 
-	if (fb->pixel_format == DRM_FORMAT_NV12) {
+	if (fb->pixel_format != DRM_FORMAT_UYVY)
 		format |= NV_PVIDEO_FORMAT_COLOR_LE_CR8YB8CB8YA8;
+	if (fb->pixel_format == DRM_FORMAT_NV12)
 		format |= NV_PVIDEO_FORMAT_PLANAR;
-	}
 	if (nv_plane->iturbt_709)
 		format |= NV_PVIDEO_FORMAT_MATRIX_ITURBT709;
 	if (nv_plane->colorkey & (1 << 24))
@@ -182,9 +185,9 @@ nv10_disable_plane(struct drm_plane *plane)
 }
 
 static void
-nv10_destroy_plane(struct drm_plane *plane)
+nv_destroy_plane(struct drm_plane *plane)
 {
-	nv10_disable_plane(plane);
+	plane->funcs->disable_plane(plane);
 	drm_plane_cleanup(plane);
 	kfree(plane);
 }
@@ -217,9 +220,9 @@ nv10_set_params(struct nouveau_plane *plane)
 }
 
 static int
-nv10_set_property(struct drm_plane *plane,
-		  struct drm_property *property,
-		  uint64_t value)
+nv_set_property(struct drm_plane *plane,
+		struct drm_property *property,
+		uint64_t value)
 {
 	struct nouveau_plane *nv_plane = (struct nouveau_plane *)plane;
 
@@ -238,15 +241,16 @@ nv10_set_property(struct drm_plane *plane,
 	else
 		return -EINVAL;
 
-	nv10_set_params(nv_plane);
+	if (nv_plane->set_params)
+		nv_plane->set_params(nv_plane);
 	return 0;
 }
 
 static const struct drm_plane_funcs nv10_plane_funcs = {
 	.update_plane = nv10_update_plane,
 	.disable_plane = nv10_disable_plane,
-	.set_property = nv10_set_property,
-	.destroy = nv10_destroy_plane,
+	.set_property = nv_set_property,
+	.destroy = nv_destroy_plane,
 };
 
 static void
@@ -266,7 +270,7 @@ nv10_overlay_init(struct drm_device *device)
 	case 0x15:
 	case 0x1a:
 	case 0x20:
-		num_formats = 1;
+		num_formats = 2;
 		break;
 	}
 
@@ -321,8 +325,159 @@ nv10_overlay_init(struct drm_device *device)
 	drm_object_attach_property(&plane->base.base,
 				   plane->props.iturbt_709, plane->iturbt_709);
 
+	plane->set_params = nv10_set_params;
 	nv10_set_params(plane);
-	nv_wr32(dev, NV_PVIDEO_STOP, 1);
+	nv10_disable_plane(&plane->base);
+	return;
+cleanup:
+	drm_plane_cleanup(&plane->base);
+err:
+	kfree(plane);
+	nv_error(dev, "Failed to create plane\n");
+}
+
+static int
+nv04_update_plane(struct drm_plane *plane, struct drm_crtc *crtc,
+		  struct drm_framebuffer *fb, int crtc_x, int crtc_y,
+		  unsigned int crtc_w, unsigned int crtc_h,
+		  uint32_t src_x, uint32_t src_y,
+		  uint32_t src_w, uint32_t src_h)
+{
+	struct nouveau_device *dev = nouveau_dev(plane->dev);
+	struct nouveau_plane *nv_plane = (struct nouveau_plane *)plane;
+	struct nouveau_framebuffer *nv_fb = nouveau_framebuffer(fb);
+	struct nouveau_bo *cur = nv_plane->cur;
+	uint32_t overlay = 1;
+	int brightness = (nv_plane->brightness - 512) * 62 / 512;
+	int pitch, ret, i;
+
+	/* Source parameters given in 16.16 fixed point, ignore fractional. */
+	src_x >>= 16;
+	src_y >>= 16;
+	src_w >>= 16;
+	src_h >>= 16;
+
+	pitch = ALIGN(src_w * 4, 0x100);
+
+	if (pitch > 0xffff)
+		return -ERANGE;
+
+	/* TODO: Compute an offset? Not sure how to do this for YUYV. */
+	if (src_x != 0 || src_y != 0)
+		return -ERANGE;
+
+	if (crtc_w < src_w || crtc_h < src_h)
+		return -ERANGE;
+
+	ret = nouveau_bo_pin(nv_fb->nvbo, TTM_PL_FLAG_VRAM);
+	if (ret)
+		return ret;
+
+	nv_plane->cur = nv_fb->nvbo;
+
+	nv_wr32(dev, NV_PVIDEO_OE_STATE, 0);
+	nv_wr32(dev, NV_PVIDEO_SU_STATE, 0);
+	nv_wr32(dev, NV_PVIDEO_RM_STATE, 0);
+
+	for (i = 0; i < 2; i++) {
+		nv_wr32(dev, NV_PVIDEO_BUFF0_START_ADDRESS + 4 * i,
+			nv_fb->nvbo->bo.offset);
+		nv_wr32(dev, NV_PVIDEO_BUFF0_PITCH_LENGTH + 4 * i, pitch);
+		nv_wr32(dev, NV_PVIDEO_BUFF0_OFFSET + 4 * i, 0);
+	}
+	nv_wr32(dev, NV_PVIDEO_WINDOW_START, crtc_y << 16 | crtc_x);
+	nv_wr32(dev, NV_PVIDEO_WINDOW_SIZE, crtc_h << 16 | crtc_w);
+	nv_wr32(dev, NV_PVIDEO_STEP_SIZE,
+		(uint32_t)(((src_h - 1) << 11) / (crtc_h - 1)) << 16 | (uint32_t)(((src_w - 1) << 11) / (crtc_w - 1)));
+
+	/* It should be possible to convert hue/contrast to this */
+	nv_wr32(dev, NV_PVIDEO_RED_CSC_OFFSET, 0x69 - brightness);
+	nv_wr32(dev, NV_PVIDEO_GREEN_CSC_OFFSET, 0x3e + brightness);
+	nv_wr32(dev, NV_PVIDEO_BLUE_CSC_OFFSET, 0x89 - brightness);
+	nv_wr32(dev, NV_PVIDEO_CSC_ADJUST, 0);
+
+	nv_wr32(dev, NV_PVIDEO_CONTROL_Y, 0x001); /* (BLUR_ON, LINE_HALF) */
+	nv_wr32(dev, NV_PVIDEO_CONTROL_X, 0x111); /* (WEIGHT_HEAVY, SHARPENING_ON, SMOOTHING_ON) */
+
+	nv_wr32(dev, NV_PVIDEO_FIFO_BURST_LENGTH, 0x03);
+	nv_wr32(dev, NV_PVIDEO_FIFO_THRES_SIZE, 0x38);
+
+	nv_wr32(dev, NV_PVIDEO_KEY, nv_plane->colorkey);
+
+	if (nv_plane->colorkey & (1 << 24))
+		overlay |= 0x10;
+	if (fb->pixel_format == DRM_FORMAT_YUYV)
+		overlay |= 0x100;
+
+	nv_wr32(dev, NV_PVIDEO_OVERLAY, overlay);
+
+	nv_wr32(dev, NV_PVIDEO_SU_STATE, nv_rd32(dev, NV_PVIDEO_SU_STATE) ^ (1 << 16));
+
+	if (cur)
+		nouveau_bo_unpin(cur);
+
+	return 0;
+}
+
+static int
+nv04_disable_plane(struct drm_plane *plane)
+{
+	struct nouveau_device *dev = nouveau_dev(plane->dev);
+	struct nouveau_plane *nv_plane = (struct nouveau_plane *)plane;
+
+	nv_mask(dev, NV_PVIDEO_OVERLAY, 1, 0);
+	nv_wr32(dev, NV_PVIDEO_OE_STATE, 0);
+	nv_wr32(dev, NV_PVIDEO_SU_STATE, 0);
+	nv_wr32(dev, NV_PVIDEO_RM_STATE, 0);
+	if (nv_plane->cur) {
+		nouveau_bo_unpin(nv_plane->cur);
+		nv_plane->cur = NULL;
+	}
+
+	return 0;
+}
+
+static const struct drm_plane_funcs nv04_plane_funcs = {
+	.update_plane = nv04_update_plane,
+	.disable_plane = nv04_disable_plane,
+	.set_property = nv_set_property,
+	.destroy = nv_destroy_plane,
+};
+
+static void
+nv04_overlay_init(struct drm_device *device)
+{
+	struct nouveau_device *dev = nouveau_dev(device);
+	struct nouveau_plane *plane = kzalloc(sizeof(struct nouveau_plane), GFP_KERNEL);
+	int ret;
+
+	if (!plane)
+		return;
+
+	ret = drm_plane_init(device, &plane->base, 1 /* single crtc */,
+			     &nv04_plane_funcs,
+			     formats, 2, false);
+	if (ret)
+		goto err;
+
+	/* Set up the plane properties */
+	plane->props.colorkey = drm_property_create_range(
+			device, 0, "colorkey", 0, 0x01ffffff);
+	plane->props.brightness = drm_property_create_range(
+			device, 0, "brightness", 0, 1024);
+	if (!plane->props.colorkey ||
+	    !plane->props.brightness)
+		goto cleanup;
+
+	plane->colorkey = 0;
+	drm_object_attach_property(&plane->base.base,
+				   plane->props.colorkey, plane->colorkey);
+
+	plane->brightness = 512;
+	drm_object_attach_property(&plane->base.base,
+				   plane->props.brightness, plane->brightness);
+
+	nv04_disable_plane(&plane->base);
 	return;
 cleanup:
 	drm_plane_cleanup(&plane->base);
@@ -335,6 +490,8 @@ void
 nouveau_overlay_init(struct drm_device *device)
 {
 	struct nouveau_device *dev = nouveau_dev(device);
-	if (dev->chipset >= 0x10 && dev->chipset <= 0x40)
+	if (dev->chipset < 0x10)
+		nv04_overlay_init(device);
+	else if (dev->chipset <= 0x40)
 		nv10_overlay_init(device);
 }
diff --git a/drivers/gpu/drm/nouveau/nouveau_acpi.c b/drivers/gpu/drm/nouveau/nouveau_acpi.c
index 3c149617cfcb..4ef83df2b246 100644
--- a/drivers/gpu/drm/nouveau/nouveau_acpi.c
+++ b/drivers/gpu/drm/nouveau/nouveau_acpi.c
@@ -61,6 +61,7 @@ bool nouveau_is_v1_dsm(void) {
 #define NOUVEAU_DSM_HAS_MUX 0x1
 #define NOUVEAU_DSM_HAS_OPT 0x2
 
+#ifdef CONFIG_VGA_SWITCHEROO
 static const char nouveau_dsm_muid[] = {
 	0xA0, 0xA0, 0x95, 0x9D, 0x60, 0x00, 0x48, 0x4D,
 	0xB3, 0x4D, 0x7E, 0x5F, 0xEA, 0x12, 0x9F, 0xD4,
@@ -326,6 +327,11 @@ void nouveau_unregister_dsm_handler(void)
 	if (nouveau_dsm_priv.optimus_detected || nouveau_dsm_priv.dsm_detected)
 		vga_switcheroo_unregister_handler();
 }
+#else
+void nouveau_register_dsm_handler(void) {}
+void nouveau_unregister_dsm_handler(void) {}
+void nouveau_switcheroo_optimus_dsm(void) {}
+#endif
 
 /* retrieve the ROM in 4k blocks */
 static int nouveau_rom_call(acpi_handle rom_handle, uint8_t *bios,
diff --git a/drivers/gpu/drm/nouveau/nouveau_bo.c b/drivers/gpu/drm/nouveau/nouveau_bo.c
index c0fde6b9393c..488686d490c0 100644
--- a/drivers/gpu/drm/nouveau/nouveau_bo.c
+++ b/drivers/gpu/drm/nouveau/nouveau_bo.c
@@ -560,28 +560,6 @@ nouveau_bo_evict_flags(struct ttm_buffer_object *bo, struct ttm_placement *pl)
 }
 
 
-/* GPU-assisted copy using NV_MEMORY_TO_MEMORY_FORMAT, can access
- * TTM_PL_{VRAM,TT} directly.
- */
-
-static int
-nouveau_bo_move_accel_cleanup(struct nouveau_channel *chan,
-			      struct nouveau_bo *nvbo, bool evict,
-			      bool no_wait_gpu, struct ttm_mem_reg *new_mem)
-{
-	struct nouveau_fence *fence = NULL;
-	int ret;
-
-	ret = nouveau_fence_new(chan, false, &fence);
-	if (ret)
-		return ret;
-
-	ret = ttm_bo_move_accel_cleanup(&nvbo->bo, fence, evict,
-					no_wait_gpu, new_mem);
-	nouveau_fence_unref(&fence);
-	return ret;
-}
-
 static int
 nve0_bo_move_init(struct nouveau_channel *chan, u32 handle)
 {
@@ -798,25 +776,25 @@ nv50_bo_move_m2mf(struct nouveau_channel *chan, struct ttm_buffer_object *bo,
 		  struct ttm_mem_reg *old_mem, struct ttm_mem_reg *new_mem)
 {
 	struct nouveau_mem *node = old_mem->mm_node;
-	struct nouveau_bo *nvbo = nouveau_bo(bo);
 	u64 length = (new_mem->num_pages << PAGE_SHIFT);
 	u64 src_offset = node->vma[0].offset;
 	u64 dst_offset = node->vma[1].offset;
+	int src_tiled = !!node->memtype;
+	int dst_tiled = !!((struct nouveau_mem *)new_mem->mm_node)->memtype;
 	int ret;
 
 	while (length) {
 		u32 amount, stride, height;
 
+		ret = RING_SPACE(chan, 18 + 6 * (src_tiled + dst_tiled));
+		if (ret)
+			return ret;
+
 		amount  = min(length, (u64)(4 * 1024 * 1024));
 		stride  = 16 * 4;
 		height  = amount / stride;
 
-		if (old_mem->mem_type == TTM_PL_VRAM &&
-		    nouveau_bo_tile_layout(nvbo)) {
-			ret = RING_SPACE(chan, 8);
-			if (ret)
-				return ret;
-
+		if (src_tiled) {
 			BEGIN_NV04(chan, NvSubCopy, 0x0200, 7);
 			OUT_RING  (chan, 0);
 			OUT_RING  (chan, 0);
@@ -826,19 +804,10 @@ nv50_bo_move_m2mf(struct nouveau_channel *chan, struct ttm_buffer_object *bo,
 			OUT_RING  (chan, 0);
 			OUT_RING  (chan, 0);
 		} else {
-			ret = RING_SPACE(chan, 2);
-			if (ret)
-				return ret;
-
 			BEGIN_NV04(chan, NvSubCopy, 0x0200, 1);
 			OUT_RING  (chan, 1);
 		}
-		if (new_mem->mem_type == TTM_PL_VRAM &&
-		    nouveau_bo_tile_layout(nvbo)) {
-			ret = RING_SPACE(chan, 8);
-			if (ret)
-				return ret;
-
+		if (dst_tiled) {
 			BEGIN_NV04(chan, NvSubCopy, 0x021c, 7);
 			OUT_RING  (chan, 0);
 			OUT_RING  (chan, 0);
@@ -848,18 +817,10 @@ nv50_bo_move_m2mf(struct nouveau_channel *chan, struct ttm_buffer_object *bo,
 			OUT_RING  (chan, 0);
 			OUT_RING  (chan, 0);
 		} else {
-			ret = RING_SPACE(chan, 2);
-			if (ret)
-				return ret;
-
 			BEGIN_NV04(chan, NvSubCopy, 0x021c, 1);
 			OUT_RING  (chan, 1);
 		}
 
-		ret = RING_SPACE(chan, 14);
-		if (ret)
-			return ret;
-
 		BEGIN_NV04(chan, NvSubCopy, 0x0238, 2);
 		OUT_RING  (chan, upper_32_bits(src_offset));
 		OUT_RING  (chan, upper_32_bits(dst_offset));
@@ -953,23 +914,28 @@ nv04_bo_move_m2mf(struct nouveau_channel *chan, struct ttm_buffer_object *bo,
 }
 
 static int
-nouveau_vma_getmap(struct nouveau_channel *chan, struct nouveau_bo *nvbo,
-		   struct ttm_mem_reg *mem, struct nouveau_vma *vma)
+nouveau_bo_move_prep(struct nouveau_drm *drm, struct ttm_buffer_object *bo,
+		     struct ttm_mem_reg *mem)
 {
-	struct nouveau_mem *node = mem->mm_node;
+	struct nouveau_mem *old_node = bo->mem.mm_node;
+	struct nouveau_mem *new_node = mem->mm_node;
+	u64 size = (u64)mem->num_pages << PAGE_SHIFT;
 	int ret;
 
-	ret = nouveau_vm_get(nv_client(chan->cli)->vm, mem->num_pages <<
-			     PAGE_SHIFT, node->page_shift,
-			     NV_MEM_ACCESS_RW, vma);
+	ret = nouveau_vm_get(nv_client(drm)->vm, size, old_node->page_shift,
+			     NV_MEM_ACCESS_RW, &old_node->vma[0]);
 	if (ret)
 		return ret;
 
-	if (mem->mem_type == TTM_PL_VRAM)
-		nouveau_vm_map(vma, node);
-	else
-		nouveau_vm_map_sg(vma, 0, mem->num_pages << PAGE_SHIFT, node);
+	ret = nouveau_vm_get(nv_client(drm)->vm, size, new_node->page_shift,
+			     NV_MEM_ACCESS_RW, &old_node->vma[1]);
+	if (ret) {
+		nouveau_vm_put(&old_node->vma[0]);
+		return ret;
+	}
 
+	nouveau_vm_map(&old_node->vma[0], old_node);
+	nouveau_vm_map(&old_node->vma[1], new_node);
 	return 0;
 }
 
@@ -979,35 +945,34 @@ nouveau_bo_move_m2mf(struct ttm_buffer_object *bo, int evict, bool intr,
 {
 	struct nouveau_drm *drm = nouveau_bdev(bo->bdev);
 	struct nouveau_channel *chan = drm->ttm.chan;
-	struct nouveau_bo *nvbo = nouveau_bo(bo);
-	struct ttm_mem_reg *old_mem = &bo->mem;
+	struct nouveau_fence *fence;
 	int ret;
 
-	mutex_lock_nested(&chan->cli->mutex, SINGLE_DEPTH_NESTING);
-
 	/* create temporary vmas for the transfer and attach them to the
 	 * old nouveau_mem node, these will get cleaned up after ttm has
 	 * destroyed the ttm_mem_reg
 	 */
 	if (nv_device(drm->device)->card_type >= NV_50) {
-		struct nouveau_mem *node = old_mem->mm_node;
-
-		ret = nouveau_vma_getmap(chan, nvbo, old_mem, &node->vma[0]);
-		if (ret)
-			goto out;
-
-		ret = nouveau_vma_getmap(chan, nvbo, new_mem, &node->vma[1]);
+		ret = nouveau_bo_move_prep(drm, bo, new_mem);
 		if (ret)
-			goto out;
+			return ret;
 	}
 
-	ret = drm->ttm.move(chan, bo, &bo->mem, new_mem);
+	mutex_lock_nested(&chan->cli->mutex, SINGLE_DEPTH_NESTING);
+	ret = nouveau_fence_sync(bo->sync_obj, chan);
 	if (ret == 0) {
-		ret = nouveau_bo_move_accel_cleanup(chan, nvbo, evict,
-						    no_wait_gpu, new_mem);
+		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,
+								evict,
+								no_wait_gpu,
+								new_mem);
+				nouveau_fence_unref(&fence);
+			}
+		}
 	}
-
-out:
 	mutex_unlock(&chan->cli->mutex);
 	return ret;
 }
@@ -1147,19 +1112,10 @@ nouveau_bo_move_ntfy(struct ttm_buffer_object *bo, struct ttm_mem_reg *new_mem)
 		return;
 
 	list_for_each_entry(vma, &nvbo->vma_list, head) {
-		if (new_mem && new_mem->mem_type == TTM_PL_VRAM) {
+		if (new_mem && new_mem->mem_type != TTM_PL_SYSTEM &&
+			      (new_mem->mem_type == TTM_PL_VRAM ||
+			       nvbo->page_shift != vma->vm->vmm->lpg_shift)) {
 			nouveau_vm_map(vma, new_mem->mm_node);
-		} else
-		if (new_mem && new_mem->mem_type == TTM_PL_TT &&
-		    nvbo->page_shift == vma->vm->vmm->spg_shift) {
-			if (((struct nouveau_mem *)new_mem->mm_node)->sg)
-				nouveau_vm_map_sg_table(vma, 0, new_mem->
-						  num_pages << PAGE_SHIFT,
-						  new_mem->mm_node);
-			else
-				nouveau_vm_map_sg(vma, 0, new_mem->
-						  num_pages << PAGE_SHIFT,
-						  new_mem->mm_node);
 		} else {
 			nouveau_vm_unmap(vma);
 		}
@@ -1224,28 +1180,27 @@ nouveau_bo_move(struct ttm_buffer_object *bo, bool evict, bool intr,
 		goto out;
 	}
 
-	/* CPU copy if we have no accelerated method available */
-	if (!drm->ttm.move) {
-		ret = ttm_bo_move_memcpy(bo, evict, no_wait_gpu, new_mem);
-		goto out;
-	}
-
 	/* Hardware assisted copy. */
-	if (new_mem->mem_type == TTM_PL_SYSTEM)
-		ret = nouveau_bo_move_flipd(bo, evict, intr,
-					    no_wait_gpu, new_mem);
-	else if (old_mem->mem_type == TTM_PL_SYSTEM)
-		ret = nouveau_bo_move_flips(bo, evict, intr,
-					    no_wait_gpu, new_mem);
-	else
-		ret = nouveau_bo_move_m2mf(bo, evict, intr,
-					   no_wait_gpu, new_mem);
-
-	if (!ret)
-		goto out;
+	if (drm->ttm.move) {
+		if (new_mem->mem_type == TTM_PL_SYSTEM)
+			ret = nouveau_bo_move_flipd(bo, evict, intr,
+						    no_wait_gpu, new_mem);
+		else if (old_mem->mem_type == TTM_PL_SYSTEM)
+			ret = nouveau_bo_move_flips(bo, evict, intr,
+						    no_wait_gpu, new_mem);
+		else
+			ret = nouveau_bo_move_m2mf(bo, evict, intr,
+						   no_wait_gpu, new_mem);
+		if (!ret)
+			goto out;
+	}
 
 	/* Fallback to software copy. */
-	ret = ttm_bo_move_memcpy(bo, evict, no_wait_gpu, new_mem);
+	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);
 
 out:
 	if (nv_device(drm->device)->card_type < NV_50) {
@@ -1271,6 +1226,7 @@ nouveau_ttm_io_mem_reserve(struct ttm_bo_device *bdev, struct ttm_mem_reg *mem)
 {
 	struct ttm_mem_type_manager *man = &bdev->man[mem->mem_type];
 	struct nouveau_drm *drm = nouveau_bdev(bdev);
+	struct nouveau_mem *node = mem->mm_node;
 	struct drm_device *dev = drm->dev;
 	int ret;
 
@@ -1293,14 +1249,16 @@ nouveau_ttm_io_mem_reserve(struct ttm_bo_device *bdev, struct ttm_mem_reg *mem)
 			mem->bus.is_iomem = !dev->agp->cant_use_aperture;
 		}
 #endif
-		break;
+		if (!node->memtype)
+			/* untiled */
+			break;
+		/* fallthrough, tiled memory */
 	case TTM_PL_VRAM:
 		mem->bus.offset = mem->start << PAGE_SHIFT;
 		mem->bus.base = pci_resource_start(dev->pdev, 1);
 		mem->bus.is_iomem = true;
 		if (nv_device(drm->device)->card_type >= NV_50) {
 			struct nouveau_bar *bar = nouveau_bar(drm->device);
-			struct nouveau_mem *node = mem->mm_node;
 
 			ret = bar->umap(bar, node, NV_MEM_ACCESS_RW,
 					&node->bar_vma);
@@ -1336,6 +1294,7 @@ nouveau_ttm_fault_reserve_notify(struct ttm_buffer_object *bo)
 	struct nouveau_bo *nvbo = nouveau_bo(bo);
 	struct nouveau_device *device = nv_device(drm->device);
 	u32 mappable = pci_resource_len(device->pdev, 1) >> PAGE_SHIFT;
+	int ret;
 
 	/* as long as the bo isn't in vram, and isn't tiled, we've got
 	 * nothing to do here.
@@ -1344,10 +1303,20 @@ nouveau_ttm_fault_reserve_notify(struct ttm_buffer_object *bo)
 		if (nv_device(drm->device)->card_type < NV_50 ||
 		    !nouveau_bo_tile_layout(nvbo))
 			return 0;
+
+		if (bo->mem.mem_type == TTM_PL_SYSTEM) {
+			nouveau_bo_placement_set(nvbo, TTM_PL_TT, 0);
+
+			ret = nouveau_bo_validate(nvbo, false, false);
+			if (ret)
+				return ret;
+		}
+		return 0;
 	}
 
 	/* make sure bo is in mappable vram */
-	if (bo->mem.start + bo->mem.num_pages < mappable)
+	if (nv_device(drm->device)->card_type >= NV_50 ||
+	    bo->mem.start + bo->mem.num_pages < mappable)
 		return 0;
 
 
@@ -1535,7 +1504,6 @@ nouveau_bo_vma_add(struct nouveau_bo *nvbo, struct nouveau_vm *vm,
 		   struct nouveau_vma *vma)
 {
 	const u32 size = nvbo->bo.mem.num_pages << PAGE_SHIFT;
-	struct nouveau_mem *node = nvbo->bo.mem.mm_node;
 	int ret;
 
 	ret = nouveau_vm_get(vm, size, nvbo->page_shift,
@@ -1543,15 +1511,10 @@ nouveau_bo_vma_add(struct nouveau_bo *nvbo, struct nouveau_vm *vm,
 	if (ret)
 		return ret;
 
-	if (nvbo->bo.mem.mem_type == TTM_PL_VRAM)
+	if ( nvbo->bo.mem.mem_type != TTM_PL_SYSTEM &&
+	    (nvbo->bo.mem.mem_type == TTM_PL_VRAM ||
+	     nvbo->page_shift != vma->vm->vmm->lpg_shift))
 		nouveau_vm_map(vma, nvbo->bo.mem.mm_node);
-	else if (nvbo->bo.mem.mem_type == TTM_PL_TT &&
-		 nvbo->page_shift == vma->vm->vmm->spg_shift) {
-		if (node->sg)
-			nouveau_vm_map_sg_table(vma, 0, size, node);
-		else
-			nouveau_vm_map_sg(vma, 0, size, node);
-	}
 
 	list_add_tail(&vma->head, &nvbo->vma_list);
 	vma->refcount = 1;
diff --git a/drivers/gpu/drm/nouveau/nouveau_display.c b/drivers/gpu/drm/nouveau/nouveau_display.c
index 25ea82f8def3..24011596af43 100644
--- a/drivers/gpu/drm/nouveau/nouveau_display.c
+++ b/drivers/gpu/drm/nouveau/nouveau_display.c
@@ -68,20 +68,100 @@ nouveau_display_vblank_disable(struct drm_device *dev, int head)
 		nouveau_event_put(disp->vblank[head]);
 }
 
+static inline int
+calc(int blanks, int blanke, int total, int line)
+{
+	if (blanke >= blanks) {
+		if (line >= blanks)
+			line -= total;
+	} else {
+		if (line >= blanks)
+			line -= total;
+		line -= blanke + 1;
+	}
+	return line;
+}
+
+int
+nouveau_display_scanoutpos_head(struct drm_crtc *crtc, int *vpos, int *hpos,
+				ktime_t *stime, ktime_t *etime)
+{
+	const u32 mthd = NV04_DISP_SCANOUTPOS + nouveau_crtc(crtc)->index;
+	struct nouveau_display *disp = nouveau_display(crtc->dev);
+	struct nv04_display_scanoutpos args;
+	int ret, retry = 1;
+
+	do {
+		ret = nv_exec(disp->core, mthd, &args, sizeof(args));
+		if (ret != 0)
+			return 0;
+
+		if (args.vline) {
+			ret |= DRM_SCANOUTPOS_ACCURATE;
+			ret |= DRM_SCANOUTPOS_VALID;
+			break;
+		}
+
+		if (retry) ndelay(crtc->linedur_ns);
+	} while (retry--);
+
+	*hpos = calc(args.hblanks, args.hblanke, args.htotal, args.hline);
+	*vpos = calc(args.vblanks, args.vblanke, args.vtotal, args.vline);
+	if (stime) *stime = ns_to_ktime(args.time[0]);
+	if (etime) *etime = ns_to_ktime(args.time[1]);
+
+	if (*vpos < 0)
+		ret |= DRM_SCANOUTPOS_INVBL;
+	return ret;
+}
+
+int
+nouveau_display_scanoutpos(struct drm_device *dev, int head, unsigned int flags,
+			   int *vpos, int *hpos, ktime_t *stime, ktime_t *etime)
+{
+	struct drm_crtc *crtc;
+
+	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
+		if (nouveau_crtc(crtc)->index == head) {
+			return nouveau_display_scanoutpos_head(crtc, vpos, hpos,
+							       stime, etime);
+		}
+	}
+
+	return 0;
+}
+
+int
+nouveau_display_vblstamp(struct drm_device *dev, int head, int *max_error,
+			 struct timeval *time, unsigned flags)
+{
+	struct drm_crtc *crtc;
+
+	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
+		if (nouveau_crtc(crtc)->index == head) {
+			return drm_calc_vbltimestamp_from_scanoutpos(dev,
+					head, max_error, time, flags, crtc,
+					&crtc->hwmode);
+		}
+	}
+
+	return -EINVAL;
+}
+
 static void
 nouveau_display_vblank_fini(struct drm_device *dev)
 {
 	struct nouveau_display *disp = nouveau_display(dev);
 	int i;
 
+	drm_vblank_cleanup(dev);
+
 	if (disp->vblank) {
 		for (i = 0; i < dev->mode_config.num_crtc; i++)
 			nouveau_event_ref(NULL, &disp->vblank[i]);
 		kfree(disp->vblank);
 		disp->vblank = NULL;
 	}
-
-	drm_vblank_cleanup(dev);
 }
 
 static int
@@ -407,10 +487,31 @@ nouveau_display_create(struct drm_device *dev)
 	drm_kms_helper_poll_disable(dev);
 
 	if (drm->vbios.dcb.entries) {
-		if (nv_device(drm->device)->card_type < NV_50)
-			ret = nv04_display_create(dev);
-		else
-			ret = nv50_display_create(dev);
+		static const u16 oclass[] = {
+			NVF0_DISP_CLASS,
+			NVE0_DISP_CLASS,
+			NVD0_DISP_CLASS,
+			NVA3_DISP_CLASS,
+			NV94_DISP_CLASS,
+			NVA0_DISP_CLASS,
+			NV84_DISP_CLASS,
+			NV50_DISP_CLASS,
+			NV04_DISP_CLASS,
+		};
+		int i;
+
+		for (i = 0, ret = -ENODEV; ret && i < ARRAY_SIZE(oclass); i++) {
+			ret = nouveau_object_new(nv_object(drm), NVDRM_DEVICE,
+						 NVDRM_DISPLAY, oclass[i],
+						 NULL, 0, &disp->core);
+		}
+
+		if (ret == 0) {
+			if (nv_mclass(disp->core) < NV50_DISP_CLASS)
+				ret = nv04_display_create(dev);
+			else
+				ret = nv50_display_create(dev);
+		}
 	} else {
 		ret = 0;
 	}
@@ -439,6 +540,7 @@ void
 nouveau_display_destroy(struct drm_device *dev)
 {
 	struct nouveau_display *disp = nouveau_display(dev);
+	struct nouveau_drm *drm = nouveau_drm(dev);
 
 	nouveau_backlight_exit(dev);
 	nouveau_display_vblank_fini(dev);
@@ -449,6 +551,8 @@ nouveau_display_destroy(struct drm_device *dev)
 	if (disp->dtor)
 		disp->dtor(dev);
 
+	nouveau_object_del(nv_object(drm), NVDRM_DEVICE, NVDRM_DISPLAY);
+
 	nouveau_drm(dev)->display = NULL;
 	kfree(disp);
 }
@@ -603,6 +707,14 @@ nouveau_crtc_page_flip(struct drm_crtc *crtc, struct drm_framebuffer *fb,
 	if (!s)
 		return -ENOMEM;
 
+	if (new_bo != old_bo) {
+		ret = nouveau_bo_pin(new_bo, TTM_PL_FLAG_VRAM);
+		if (ret)
+			goto fail_free;
+	}
+
+	mutex_lock(&chan->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);
@@ -610,15 +722,8 @@ nouveau_crtc_page_flip(struct drm_crtc *crtc, struct drm_framebuffer *fb,
 	ret = nouveau_fence_sync(fence, chan);
 	nouveau_fence_unref(&fence);
 	if (ret)
-		goto fail_free;
-
-	if (new_bo != old_bo) {
-		ret = nouveau_bo_pin(new_bo, TTM_PL_FLAG_VRAM);
-		if (ret)
-			goto fail_free;
-	}
+		goto fail_unpin;
 
-	mutex_lock(&chan->cli->mutex);
 	ret = ttm_bo_reserve(&old_bo->bo, true, false, false, NULL);
 	if (ret)
 		goto fail_unpin;
diff --git a/drivers/gpu/drm/nouveau/nouveau_display.h b/drivers/gpu/drm/nouveau/nouveau_display.h
index 8bc8bab90e8d..a71cf77e55b2 100644
--- a/drivers/gpu/drm/nouveau/nouveau_display.h
+++ b/drivers/gpu/drm/nouveau/nouveau_display.h
@@ -36,6 +36,7 @@ struct nouveau_display {
 	int  (*init)(struct drm_device *);
 	void (*fini)(struct drm_device *);
 
+	struct nouveau_object *core;
 	struct nouveau_eventh **vblank;
 
 	struct drm_property *dithering_mode;
@@ -63,6 +64,10 @@ void nouveau_display_repin(struct drm_device *dev);
 void nouveau_display_resume(struct drm_device *dev);
 int  nouveau_display_vblank_enable(struct drm_device *, int);
 void nouveau_display_vblank_disable(struct drm_device *, int);
+int  nouveau_display_scanoutpos(struct drm_device *, int, unsigned int,
+				int *, int *, ktime_t *, ktime_t *);
+int  nouveau_display_vblstamp(struct drm_device *, int, int *,
+			      struct timeval *, unsigned);
 
 int  nouveau_crtc_page_flip(struct drm_crtc *crtc, struct drm_framebuffer *fb,
 			    struct drm_pending_vblank_event *event,
diff --git a/drivers/gpu/drm/nouveau/nouveau_dma.c b/drivers/gpu/drm/nouveau/nouveau_dma.c
index 40f91e1e5842..c177272152e2 100644
--- a/drivers/gpu/drm/nouveau/nouveau_dma.c
+++ b/drivers/gpu/drm/nouveau/nouveau_dma.c
@@ -100,7 +100,7 @@ nv50_dma_push(struct nouveau_channel *chan, struct nouveau_bo *bo,
 
 	chan->dma.ib_put = (chan->dma.ib_put + 1) & chan->dma.ib_max;
 
-	DRM_MEMORYBARRIER();
+	mb();
 	/* Flush writes. */
 	nouveau_bo_rd32(pb, 0);
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_dma.h b/drivers/gpu/drm/nouveau/nouveau_dma.h
index 984004d66a6d..dc0e0c5cadb4 100644
--- a/drivers/gpu/drm/nouveau/nouveau_dma.h
+++ b/drivers/gpu/drm/nouveau/nouveau_dma.h
@@ -155,7 +155,7 @@ BEGIN_IMC0(struct nouveau_channel *chan, int subc, int mthd, u16 data)
 }
 
 #define WRITE_PUT(val) do {                                                    \
-	DRM_MEMORYBARRIER();                                                   \
+	mb();                                                   \
 	nouveau_bo_rd32(chan->push.buffer, 0);                                 \
 	nv_wo32(chan->object, chan->user_put, ((val) << 2) + chan->push.vma.offset);  \
 } while (0)
diff --git a/drivers/gpu/drm/nouveau/nouveau_drm.c b/drivers/gpu/drm/nouveau/nouveau_drm.c
index 98a22e6e27a1..78c8e7146d56 100644
--- a/drivers/gpu/drm/nouveau/nouveau_drm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_drm.c
@@ -503,19 +503,21 @@ nouveau_do_suspend(struct drm_device *dev)
 	if (drm->cechan) {
 		ret = nouveau_channel_idle(drm->cechan);
 		if (ret)
-			return ret;
+			goto fail_display;
 	}
 
 	if (drm->channel) {
 		ret = nouveau_channel_idle(drm->channel);
 		if (ret)
-			return ret;
+			goto fail_display;
 	}
 
 	NV_INFO(drm, "suspending client object trees...\n");
 	if (drm->fence && nouveau_fence(drm)->suspend) {
-		if (!nouveau_fence(drm)->suspend(drm))
-			return -ENOMEM;
+		if (!nouveau_fence(drm)->suspend(drm)) {
+			ret = -ENOMEM;
+			goto fail_display;
+		}
 	}
 
 	list_for_each_entry(cli, &drm->clients, head) {
@@ -537,6 +539,10 @@ fail_client:
 		nouveau_client_init(&cli->base);
 	}
 
+	if (drm->fence && nouveau_fence(drm)->resume)
+		nouveau_fence(drm)->resume(drm);
+
+fail_display:
 	if (dev->mode_config.num_crtc) {
 		NV_INFO(drm, "resuming display...\n");
 		nouveau_display_resume(dev);
@@ -798,6 +804,8 @@ driver = {
 	.get_vblank_counter = drm_vblank_count,
 	.enable_vblank = nouveau_display_vblank_enable,
 	.disable_vblank = nouveau_display_vblank_disable,
+	.get_scanout_position = nouveau_display_scanoutpos,
+	.get_vblank_timestamp = nouveau_display_vblstamp,
 
 	.ioctls = nouveau_ioctls,
 	.num_ioctls = ARRAY_SIZE(nouveau_ioctls),
diff --git a/drivers/gpu/drm/nouveau/nouveau_drm.h b/drivers/gpu/drm/nouveau/nouveau_drm.h
index 4b0fb6c66be9..23ca7a517246 100644
--- a/drivers/gpu/drm/nouveau/nouveau_drm.h
+++ b/drivers/gpu/drm/nouveau/nouveau_drm.h
@@ -54,6 +54,7 @@ enum nouveau_drm_handle {
 	NVDRM_CLIENT  = 0xffffffff,
 	NVDRM_DEVICE  = 0xdddddddd,
 	NVDRM_CONTROL = 0xdddddddc,
+	NVDRM_DISPLAY = 0xd1500000,
 	NVDRM_PUSH    = 0xbbbb0000, /* |= client chid */
 	NVDRM_CHAN    = 0xcccc0000, /* |= client chid */
 	NVDRM_NVSW    = 0x55550000,
diff --git a/drivers/gpu/drm/nouveau/nouveau_fence.c b/drivers/gpu/drm/nouveau/nouveau_fence.c
index 40cf52e6d6d2..90074d620e31 100644
--- a/drivers/gpu/drm/nouveau/nouveau_fence.c
+++ b/drivers/gpu/drm/nouveau/nouveau_fence.c
@@ -143,7 +143,7 @@ nouveau_fence_emit(struct nouveau_fence *fence, struct nouveau_channel *chan)
 	int ret;
 
 	fence->channel  = chan;
-	fence->timeout  = jiffies + (15 * DRM_HZ);
+	fence->timeout  = jiffies + (15 * HZ);
 	fence->sequence = ++fctx->sequence;
 
 	ret = fctx->emit(fence);
diff --git a/drivers/gpu/drm/nouveau/nouveau_gem.c b/drivers/gpu/drm/nouveau/nouveau_gem.c
index 78a27f8ad7d9..27c3fd89e8ce 100644
--- a/drivers/gpu/drm/nouveau/nouveau_gem.c
+++ b/drivers/gpu/drm/nouveau/nouveau_gem.c
@@ -463,12 +463,6 @@ validate_list(struct nouveau_channel *chan, struct nouveau_cli *cli,
 	list_for_each_entry(nvbo, list, entry) {
 		struct drm_nouveau_gem_pushbuf_bo *b = &pbbo[nvbo->pbbo_index];
 
-		ret = validate_sync(chan, nvbo);
-		if (unlikely(ret)) {
-			NV_ERROR(cli, "fail pre-validate sync\n");
-			return ret;
-		}
-
 		ret = nouveau_gem_set_domain(&nvbo->gem, b->read_domains,
 					     b->write_domains,
 					     b->valid_domains);
@@ -506,7 +500,7 @@ validate_list(struct nouveau_channel *chan, struct nouveau_cli *cli,
 			b->presumed.valid = 0;
 			relocs++;
 
-			if (DRM_COPY_TO_USER(&upbbo[nvbo->pbbo_index].presumed,
+			if (copy_to_user(&upbbo[nvbo->pbbo_index].presumed,
 					     &b->presumed, sizeof(b->presumed)))
 				return -EFAULT;
 		}
@@ -593,7 +587,7 @@ u_memcpya(uint64_t user, unsigned nmemb, unsigned size)
 	if (!mem)
 		return ERR_PTR(-ENOMEM);
 
-	if (DRM_COPY_FROM_USER(mem, userptr, size)) {
+	if (copy_from_user(mem, userptr, size)) {
 		u_free(mem);
 		return ERR_PTR(-EFAULT);
 	}
diff --git a/drivers/gpu/drm/nouveau/nouveau_sgdma.c b/drivers/gpu/drm/nouveau/nouveau_sgdma.c
index 0843ebc910d4..a4d22e5eb176 100644
--- a/drivers/gpu/drm/nouveau/nouveau_sgdma.c
+++ b/drivers/gpu/drm/nouveau/nouveau_sgdma.c
@@ -31,16 +31,17 @@ nv04_sgdma_bind(struct ttm_tt *ttm, struct ttm_mem_reg *mem)
 {
 	struct nouveau_sgdma_be *nvbe = (struct nouveau_sgdma_be *)ttm;
 	struct nouveau_mem *node = mem->mm_node;
-	u64 size = mem->num_pages << 12;
 
 	if (ttm->sg) {
-		node->sg = ttm->sg;
-		nouveau_vm_map_sg_table(&node->vma[0], 0, size, node);
+		node->sg    = ttm->sg;
+		node->pages = NULL;
 	} else {
+		node->sg    = NULL;
 		node->pages = nvbe->ttm.dma_address;
-		nouveau_vm_map_sg(&node->vma[0], 0, size, node);
 	}
+	node->size = (mem->num_pages << PAGE_SHIFT) >> 12;
 
+	nouveau_vm_map(&node->vma[0], node);
 	nvbe->node = node;
 	return 0;
 }
@@ -67,9 +68,13 @@ nv50_sgdma_bind(struct ttm_tt *ttm, struct ttm_mem_reg *mem)
 
 	/* noop: bound in move_notify() */
 	if (ttm->sg) {
-		node->sg = ttm->sg;
-	} else
+		node->sg    = ttm->sg;
+		node->pages = NULL;
+	} else {
+		node->sg    = NULL;
 		node->pages = nvbe->ttm.dma_address;
+	}
+	node->size = (mem->num_pages << PAGE_SHIFT) >> 12;
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_ttm.c b/drivers/gpu/drm/nouveau/nouveau_ttm.c
index 19e3757291fb..d45d50da978f 100644
--- a/drivers/gpu/drm/nouveau/nouveau_ttm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_ttm.c
@@ -171,6 +171,7 @@ nouveau_gart_manager_new(struct ttm_mem_type_manager *man,
 	node = kzalloc(sizeof(*node), GFP_KERNEL);
 	if (!node)
 		return -ENOMEM;
+
 	node->page_shift = 12;
 
 	switch (nv_device(drm->device)->card_type) {
diff --git a/drivers/gpu/drm/nouveau/nv50_display.c b/drivers/gpu/drm/nouveau/nv50_display.c
index 4e384a2f99c3..2dccafc6e9db 100644
--- a/drivers/gpu/drm/nouveau/nv50_display.c
+++ b/drivers/gpu/drm/nouveau/nv50_display.c
@@ -1035,6 +1035,7 @@ static bool
 nv50_crtc_mode_fixup(struct drm_crtc *crtc, const struct drm_display_mode *mode,
 		     struct drm_display_mode *adjusted_mode)
 {
+	drm_mode_set_crtcinfo(adjusted_mode, CRTC_INTERLACE_HALVE_V);
 	return true;
 }
 
@@ -2199,16 +2200,6 @@ nv50_display_destroy(struct drm_device *dev)
 int
 nv50_display_create(struct drm_device *dev)
 {
-	static const u16 oclass[] = {
-		NVF0_DISP_CLASS,
-		NVE0_DISP_CLASS,
-		NVD0_DISP_CLASS,
-		NVA3_DISP_CLASS,
-		NV94_DISP_CLASS,
-		NVA0_DISP_CLASS,
-		NV84_DISP_CLASS,
-		NV50_DISP_CLASS,
-	};
 	struct nouveau_device *device = nouveau_dev(dev);
 	struct nouveau_drm *drm = nouveau_drm(dev);
 	struct dcb_table *dcb = &drm->vbios.dcb;
@@ -2225,6 +2216,7 @@ nv50_display_create(struct drm_device *dev)
 	nouveau_display(dev)->dtor = nv50_display_destroy;
 	nouveau_display(dev)->init = nv50_display_init;
 	nouveau_display(dev)->fini = nv50_display_fini;
+	disp->core = nouveau_display(dev)->core;
 
 	/* small shared memory area we use for notifiers and semaphores */
 	ret = nouveau_bo_new(dev, 4096, 0x1000, TTM_PL_FLAG_VRAM,
@@ -2243,17 +2235,6 @@ nv50_display_create(struct drm_device *dev)
 	if (ret)
 		goto out;
 
-	/* attempt to allocate a supported evo display class */
-	ret = -ENODEV;
-	for (i = 0; ret && i < ARRAY_SIZE(oclass); i++) {
-		ret = nouveau_object_new(nv_object(drm), NVDRM_DEVICE,
-					 0xd1500000, oclass[i], NULL, 0,
-					 &disp->core);
-	}
-
-	if (ret)
-		goto out;
-
 	/* allocate master evo channel */
 	ret = nv50_dmac_create(disp->core, NV50_DISP_MAST_CLASS, 0,
 			      &(struct nv50_display_mast_class) {
diff --git a/drivers/gpu/drm/omapdrm/omap_crtc.c b/drivers/gpu/drm/omapdrm/omap_crtc.c
index 0fd2eb139f6e..4313bb0a49a6 100644
--- a/drivers/gpu/drm/omapdrm/omap_crtc.c
+++ b/drivers/gpu/drm/omapdrm/omap_crtc.c
@@ -411,7 +411,7 @@ static void omap_crtc_error_irq(struct omap_drm_irq *irq, uint32_t irqstatus)
 	struct drm_crtc *crtc = &omap_crtc->base;
 	DRM_ERROR("%s: errors: %08x\n", omap_crtc->name, irqstatus);
 	/* avoid getting in a flood, unregister the irq until next vblank */
-	omap_irq_unregister(crtc->dev, &omap_crtc->error_irq);
+	__omap_irq_unregister(crtc->dev, &omap_crtc->error_irq);
 }
 
 static void omap_crtc_apply_irq(struct omap_drm_irq *irq, uint32_t irqstatus)
@@ -421,13 +421,13 @@ static void omap_crtc_apply_irq(struct omap_drm_irq *irq, uint32_t irqstatus)
 	struct drm_crtc *crtc = &omap_crtc->base;
 
 	if (!omap_crtc->error_irq.registered)
-		omap_irq_register(crtc->dev, &omap_crtc->error_irq);
+		__omap_irq_register(crtc->dev, &omap_crtc->error_irq);
 
 	if (!dispc_mgr_go_busy(omap_crtc->channel)) {
 		struct omap_drm_private *priv =
 				crtc->dev->dev_private;
 		DBG("%s: apply done", omap_crtc->name);
-		omap_irq_unregister(crtc->dev, &omap_crtc->apply_irq);
+		__omap_irq_unregister(crtc->dev, &omap_crtc->apply_irq);
 		queue_work(priv->wq, &omap_crtc->apply_work);
 	}
 }
@@ -623,6 +623,11 @@ void omap_crtc_pre_init(void)
 	dss_install_mgr_ops(&mgr_ops);
 }
 
+void omap_crtc_pre_uninit(void)
+{
+	dss_uninstall_mgr_ops();
+}
+
 /* initialize crtc */
 struct drm_crtc *omap_crtc_init(struct drm_device *dev,
 		struct drm_plane *plane, enum omap_channel channel, int id)
diff --git a/drivers/gpu/drm/omapdrm/omap_debugfs.c b/drivers/gpu/drm/omapdrm/omap_debugfs.c
index c27f59da7f29..d4c04d69fc4d 100644
--- a/drivers/gpu/drm/omapdrm/omap_debugfs.c
+++ b/drivers/gpu/drm/omapdrm/omap_debugfs.c
@@ -48,7 +48,7 @@ static int mm_show(struct seq_file *m, void *arg)
 {
 	struct drm_info_node *node = (struct drm_info_node *) m->private;
 	struct drm_device *dev = node->minor->dev;
-	return drm_mm_dump_table(m, dev->mm_private);
+	return drm_mm_dump_table(m, &dev->vma_offset_manager->vm_addr_space_mm);
 }
 
 static int fb_show(struct seq_file *m, void *arg)
diff --git a/drivers/gpu/drm/omapdrm/omap_dmm_tiler.c b/drivers/gpu/drm/omapdrm/omap_dmm_tiler.c
index 701c4c10e08b..f926b4caf449 100644
--- a/drivers/gpu/drm/omapdrm/omap_dmm_tiler.c
+++ b/drivers/gpu/drm/omapdrm/omap_dmm_tiler.c
@@ -969,12 +969,21 @@ static const struct dev_pm_ops omap_dmm_pm_ops = {
 };
 #endif
 
+#if defined(CONFIG_OF)
+static const struct of_device_id dmm_of_match[] = {
+	{ .compatible = "ti,omap4-dmm", },
+	{ .compatible = "ti,omap5-dmm", },
+	{},
+};
+#endif
+
 struct platform_driver omap_dmm_driver = {
 	.probe = omap_dmm_probe,
 	.remove = omap_dmm_remove,
 	.driver = {
 		.owner = THIS_MODULE,
 		.name = DMM_DRIVER_NAME,
+		.of_match_table = of_match_ptr(dmm_of_match),
 #ifdef CONFIG_PM
 		.pm = &omap_dmm_pm_ops,
 #endif
diff --git a/drivers/gpu/drm/omapdrm/omap_drv.c b/drivers/gpu/drm/omapdrm/omap_drv.c
index e7fa3cd96743..bf39fcc49e0f 100644
--- a/drivers/gpu/drm/omapdrm/omap_drv.c
+++ b/drivers/gpu/drm/omapdrm/omap_drv.c
@@ -86,6 +86,47 @@ static bool channel_used(struct drm_device *dev, enum omap_channel channel)
 
 	return false;
 }
+static void omap_disconnect_dssdevs(void)
+{
+	struct omap_dss_device *dssdev = NULL;
+
+	for_each_dss_dev(dssdev)
+		dssdev->driver->disconnect(dssdev);
+}
+
+static int omap_connect_dssdevs(void)
+{
+	int r;
+	struct omap_dss_device *dssdev = NULL;
+	bool no_displays = true;
+
+	for_each_dss_dev(dssdev) {
+		r = dssdev->driver->connect(dssdev);
+		if (r == -EPROBE_DEFER) {
+			omap_dss_put_device(dssdev);
+			goto cleanup;
+		} else if (r) {
+			dev_warn(dssdev->dev, "could not connect display: %s\n",
+				dssdev->name);
+		} else {
+			no_displays = false;
+		}
+	}
+
+	if (no_displays)
+		return -EPROBE_DEFER;
+
+	return 0;
+
+cleanup:
+	/*
+	 * if we are deferring probe, we disconnect the devices we previously
+	 * connected
+	 */
+	omap_disconnect_dssdevs();
+
+	return r;
+}
 
 static int omap_modeset_init(struct drm_device *dev)
 {
@@ -95,9 +136,6 @@ static int omap_modeset_init(struct drm_device *dev)
 	int num_mgrs = dss_feat_get_num_mgrs();
 	int num_crtcs;
 	int i, id = 0;
-	int r;
-
-	omap_crtc_pre_init();
 
 	drm_mode_config_init(dev);
 
@@ -119,26 +157,8 @@ static int omap_modeset_init(struct drm_device *dev)
 		enum omap_channel channel;
 		struct omap_overlay_manager *mgr;
 
-		if (!dssdev->driver) {
-			dev_warn(dev->dev, "%s has no driver.. skipping it\n",
-					dssdev->name);
-			continue;
-		}
-
-		if (!(dssdev->driver->get_timings ||
-					dssdev->driver->read_edid)) {
-			dev_warn(dev->dev, "%s driver does not support "
-				"get_timings or read_edid.. skipping it!\n",
-				dssdev->name);
-			continue;
-		}
-
-		r = dssdev->driver->connect(dssdev);
-		if (r) {
-			dev_err(dev->dev, "could not connect display: %s\n",
-					dssdev->name);
+		if (!omapdss_device_is_connected(dssdev))
 			continue;
-		}
 
 		encoder = omap_encoder_init(dev, dssdev);
 
@@ -497,16 +517,16 @@ static int dev_unload(struct drm_device *dev)
 	DBG("unload: dev=%p", dev);
 
 	drm_kms_helper_poll_fini(dev);
-	drm_vblank_cleanup(dev);
-	omap_drm_irq_uninstall(dev);
 
 	omap_fbdev_free(dev);
 	omap_modeset_free(dev);
 	omap_gem_deinit(dev);
 
-	flush_workqueue(priv->wq);
 	destroy_workqueue(priv->wq);
 
+	drm_vblank_cleanup(dev);
+	omap_drm_irq_uninstall(dev);
+
 	kfree(dev->dev_private);
 	dev->dev_private = NULL;
 
@@ -655,9 +675,19 @@ static void pdev_shutdown(struct platform_device *device)
 
 static int pdev_probe(struct platform_device *device)
 {
+	int r;
+
 	if (omapdss_is_initialized() == false)
 		return -EPROBE_DEFER;
 
+	omap_crtc_pre_init();
+
+	r = omap_connect_dssdevs();
+	if (r) {
+		omap_crtc_pre_uninit();
+		return r;
+	}
+
 	DBG("%s", device->name);
 	return drm_platform_init(&omap_drm_driver, device);
 }
@@ -665,9 +695,11 @@ static int pdev_probe(struct platform_device *device)
 static int pdev_remove(struct platform_device *device)
 {
 	DBG("");
-	drm_platform_exit(&omap_drm_driver, device);
 
-	platform_driver_unregister(&omap_dmm_driver);
+	omap_disconnect_dssdevs();
+	omap_crtc_pre_uninit();
+
+	drm_put_dev(platform_get_drvdata(device));
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/omapdrm/omap_drv.h b/drivers/gpu/drm/omapdrm/omap_drv.h
index 07847693cf49..428b2981fd68 100644
--- a/drivers/gpu/drm/omapdrm/omap_drv.h
+++ b/drivers/gpu/drm/omapdrm/omap_drv.h
@@ -141,10 +141,12 @@ int omap_gem_resume(struct device *dev);
 
 int omap_irq_enable_vblank(struct drm_device *dev, int crtc_id);
 void omap_irq_disable_vblank(struct drm_device *dev, int crtc_id);
-irqreturn_t omap_irq_handler(DRM_IRQ_ARGS);
+irqreturn_t omap_irq_handler(int irq, void *arg);
 void omap_irq_preinstall(struct drm_device *dev);
 int omap_irq_postinstall(struct drm_device *dev);
 void omap_irq_uninstall(struct drm_device *dev);
+void __omap_irq_register(struct drm_device *dev, struct omap_drm_irq *irq);
+void __omap_irq_unregister(struct drm_device *dev, struct omap_drm_irq *irq);
 void omap_irq_register(struct drm_device *dev, struct omap_drm_irq *irq);
 void omap_irq_unregister(struct drm_device *dev, struct omap_drm_irq *irq);
 int omap_drm_irq_uninstall(struct drm_device *dev);
@@ -158,6 +160,7 @@ enum omap_channel omap_crtc_channel(struct drm_crtc *crtc);
 int omap_crtc_apply(struct drm_crtc *crtc,
 		struct omap_drm_apply *apply);
 void omap_crtc_pre_init(void);
+void omap_crtc_pre_uninit(void);
 struct drm_crtc *omap_crtc_init(struct drm_device *dev,
 		struct drm_plane *plane, enum omap_channel channel, int id);
 
diff --git a/drivers/gpu/drm/omapdrm/omap_encoder.c b/drivers/gpu/drm/omapdrm/omap_encoder.c
index 6a12e899235b..5290a88c681d 100644
--- a/drivers/gpu/drm/omapdrm/omap_encoder.c
+++ b/drivers/gpu/drm/omapdrm/omap_encoder.c
@@ -51,6 +51,9 @@ struct omap_dss_device *omap_encoder_get_dssdev(struct drm_encoder *encoder)
 static void omap_encoder_destroy(struct drm_encoder *encoder)
 {
 	struct omap_encoder *omap_encoder = to_omap_encoder(encoder);
+
+	omap_encoder_set_enabled(encoder, false);
+
 	drm_encoder_cleanup(encoder);
 	kfree(omap_encoder);
 }
diff --git a/drivers/gpu/drm/omapdrm/omap_fb.c b/drivers/gpu/drm/omapdrm/omap_fb.c
index f2b8f0668c0c..f466c4aaee94 100644
--- a/drivers/gpu/drm/omapdrm/omap_fb.c
+++ b/drivers/gpu/drm/omapdrm/omap_fb.c
@@ -123,12 +123,16 @@ static int omap_framebuffer_dirty(struct drm_framebuffer *fb,
 {
 	int i;
 
+	drm_modeset_lock_all(fb->dev);
+
 	for (i = 0; i < num_clips; i++) {
 		omap_framebuffer_flush(fb, clips[i].x1, clips[i].y1,
 					clips[i].x2 - clips[i].x1,
 					clips[i].y2 - clips[i].y1);
 	}
 
+	drm_modeset_unlock_all(fb->dev);
+
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/omapdrm/omap_irq.c b/drivers/gpu/drm/omapdrm/omap_irq.c
index cb858600185f..f035d2bceae7 100644
--- a/drivers/gpu/drm/omapdrm/omap_irq.c
+++ b/drivers/gpu/drm/omapdrm/omap_irq.c
@@ -45,12 +45,11 @@ static void omap_irq_update(struct drm_device *dev)
 	dispc_read_irqenable();        /* flush posted write */
 }
 
-void omap_irq_register(struct drm_device *dev, struct omap_drm_irq *irq)
+void __omap_irq_register(struct drm_device *dev, struct omap_drm_irq *irq)
 {
 	struct omap_drm_private *priv = dev->dev_private;
 	unsigned long flags;
 
-	dispc_runtime_get();
 	spin_lock_irqsave(&list_lock, flags);
 
 	if (!WARN_ON(irq->registered)) {
@@ -60,14 +59,21 @@ void omap_irq_register(struct drm_device *dev, struct omap_drm_irq *irq)
 	}
 
 	spin_unlock_irqrestore(&list_lock, flags);
+}
+
+void omap_irq_register(struct drm_device *dev, struct omap_drm_irq *irq)
+{
+	dispc_runtime_get();
+
+	__omap_irq_register(dev, irq);
+
 	dispc_runtime_put();
 }
 
-void omap_irq_unregister(struct drm_device *dev, struct omap_drm_irq *irq)
+void __omap_irq_unregister(struct drm_device *dev, struct omap_drm_irq *irq)
 {
 	unsigned long flags;
 
-	dispc_runtime_get();
 	spin_lock_irqsave(&list_lock, flags);
 
 	if (!WARN_ON(!irq->registered)) {
@@ -77,6 +83,14 @@ void omap_irq_unregister(struct drm_device *dev, struct omap_drm_irq *irq)
 	}
 
 	spin_unlock_irqrestore(&list_lock, flags);
+}
+
+void omap_irq_unregister(struct drm_device *dev, struct omap_drm_irq *irq)
+{
+	dispc_runtime_get();
+
+	__omap_irq_unregister(dev, irq);
+
 	dispc_runtime_put();
 }
 
@@ -173,7 +187,7 @@ void omap_irq_disable_vblank(struct drm_device *dev, int crtc_id)
 	dispc_runtime_put();
 }
 
-irqreturn_t omap_irq_handler(DRM_IRQ_ARGS)
+irqreturn_t omap_irq_handler(int irq, void *arg)
 {
 	struct drm_device *dev = (struct drm_device *) arg;
 	struct omap_drm_private *priv = dev->dev_private;
@@ -308,7 +322,7 @@ int omap_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++) {
-			DRM_WAKEUP(&dev->vblank[i].queue);
+			wake_up(&dev->vblank[i].queue);
 			dev->vblank[i].enabled = false;
 			dev->vblank[i].last =
 				dev->driver->get_vblank_counter(dev, i);
diff --git a/drivers/gpu/drm/panel/Kconfig b/drivers/gpu/drm/panel/Kconfig
new file mode 100644
index 000000000000..3e0f13d1bc84
--- /dev/null
+++ b/drivers/gpu/drm/panel/Kconfig
@@ -0,0 +1,19 @@
+config DRM_PANEL
+	bool
+	depends on DRM
+	help
+	  Panel registration and lookup framework.
+
+menu "Display Panels"
+	depends on DRM_PANEL
+
+config DRM_PANEL_SIMPLE
+	tristate "support for simple panels"
+	depends on OF
+	help
+	  DRM panel driver for dumb panels that need at most a regulator and
+	  a GPIO to be powered up. Optionally a backlight can be attached so
+	  that it can be automatically turned off when the panel goes into a
+	  low power state.
+
+endmenu
diff --git a/drivers/gpu/drm/panel/Makefile b/drivers/gpu/drm/panel/Makefile
new file mode 100644
index 000000000000..af9dfa235b94
--- /dev/null
+++ b/drivers/gpu/drm/panel/Makefile
@@ -0,0 +1 @@
+obj-$(CONFIG_DRM_PANEL_SIMPLE) += panel-simple.o
diff --git a/drivers/gpu/drm/panel/panel-simple.c b/drivers/gpu/drm/panel/panel-simple.c
new file mode 100644
index 000000000000..59d52ca2c67f
--- /dev/null
+++ b/drivers/gpu/drm/panel/panel-simple.c
@@ -0,0 +1,548 @@
+/*
+ * Copyright (C) 2013, NVIDIA Corporation.  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 AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ */
+
+#include <linux/backlight.h>
+#include <linux/gpio.h>
+#include <linux/module.h>
+#include <linux/of_gpio.h>
+#include <linux/of_platform.h>
+#include <linux/platform_device.h>
+#include <linux/regulator/consumer.h>
+
+#include <drm/drmP.h>
+#include <drm/drm_crtc.h>
+#include <drm/drm_mipi_dsi.h>
+#include <drm/drm_panel.h>
+
+struct panel_desc {
+	const struct drm_display_mode *modes;
+	unsigned int num_modes;
+
+	struct {
+		unsigned int width;
+		unsigned int height;
+	} size;
+};
+
+/* TODO: convert to gpiod_*() API once it's been merged */
+#define GPIO_ACTIVE_LOW	(1 << 0)
+
+struct panel_simple {
+	struct drm_panel base;
+	bool enabled;
+
+	const struct panel_desc *desc;
+
+	struct backlight_device *backlight;
+	struct regulator *supply;
+	struct i2c_adapter *ddc;
+
+	unsigned long enable_gpio_flags;
+	int enable_gpio;
+};
+
+static inline struct panel_simple *to_panel_simple(struct drm_panel *panel)
+{
+	return container_of(panel, struct panel_simple, base);
+}
+
+static int panel_simple_get_fixed_modes(struct panel_simple *panel)
+{
+	struct drm_connector *connector = panel->base.connector;
+	struct drm_device *drm = panel->base.drm;
+	struct drm_display_mode *mode;
+	unsigned int i, num = 0;
+
+	if (!panel->desc)
+		return 0;
+
+	for (i = 0; i < panel->desc->num_modes; i++) {
+		const struct drm_display_mode *m = &panel->desc->modes[i];
+
+		mode = drm_mode_duplicate(drm, m);
+		if (!mode) {
+			dev_err(drm->dev, "failed to add mode %ux%u@%u\n",
+				m->hdisplay, m->vdisplay, m->vrefresh);
+			continue;
+		}
+
+		drm_mode_set_name(mode);
+
+		drm_mode_probed_add(connector, mode);
+		num++;
+	}
+
+	connector->display_info.width_mm = panel->desc->size.width;
+	connector->display_info.height_mm = panel->desc->size.height;
+
+	return num;
+}
+
+static int panel_simple_disable(struct drm_panel *panel)
+{
+	struct panel_simple *p = to_panel_simple(panel);
+
+	if (!p->enabled)
+		return 0;
+
+	if (p->backlight) {
+		p->backlight->props.power = FB_BLANK_POWERDOWN;
+		backlight_update_status(p->backlight);
+	}
+
+	if (gpio_is_valid(p->enable_gpio)) {
+		if (p->enable_gpio_flags & GPIO_ACTIVE_LOW)
+			gpio_set_value(p->enable_gpio, 1);
+		else
+			gpio_set_value(p->enable_gpio, 0);
+	}
+
+	regulator_disable(p->supply);
+	p->enabled = false;
+
+	return 0;
+}
+
+static int panel_simple_enable(struct drm_panel *panel)
+{
+	struct panel_simple *p = to_panel_simple(panel);
+	int err;
+
+	if (p->enabled)
+		return 0;
+
+	err = regulator_enable(p->supply);
+	if (err < 0) {
+		dev_err(panel->dev, "failed to enable supply: %d\n", err);
+		return err;
+	}
+
+	if (gpio_is_valid(p->enable_gpio)) {
+		if (p->enable_gpio_flags & GPIO_ACTIVE_LOW)
+			gpio_set_value(p->enable_gpio, 0);
+		else
+			gpio_set_value(p->enable_gpio, 1);
+	}
+
+	if (p->backlight) {
+		p->backlight->props.power = FB_BLANK_UNBLANK;
+		backlight_update_status(p->backlight);
+	}
+
+	p->enabled = true;
+
+	return 0;
+}
+
+static int panel_simple_get_modes(struct drm_panel *panel)
+{
+	struct panel_simple *p = to_panel_simple(panel);
+	int num = 0;
+
+	/* probe EDID if a DDC bus is available */
+	if (p->ddc) {
+		struct edid *edid = drm_get_edid(panel->connector, p->ddc);
+		drm_mode_connector_update_edid_property(panel->connector, edid);
+		if (edid) {
+			num += drm_add_edid_modes(panel->connector, edid);
+			kfree(edid);
+		}
+	}
+
+	/* add hard-coded panel modes */
+	num += panel_simple_get_fixed_modes(p);
+
+	return num;
+}
+
+static const struct drm_panel_funcs panel_simple_funcs = {
+	.disable = panel_simple_disable,
+	.enable = panel_simple_enable,
+	.get_modes = panel_simple_get_modes,
+};
+
+static int panel_simple_probe(struct device *dev, const struct panel_desc *desc)
+{
+	struct device_node *backlight, *ddc;
+	struct panel_simple *panel;
+	enum of_gpio_flags flags;
+	int err;
+
+	panel = devm_kzalloc(dev, sizeof(*panel), GFP_KERNEL);
+	if (!panel)
+		return -ENOMEM;
+
+	panel->enabled = false;
+	panel->desc = desc;
+
+	panel->supply = devm_regulator_get(dev, "power");
+	if (IS_ERR(panel->supply))
+		return PTR_ERR(panel->supply);
+
+	panel->enable_gpio = of_get_named_gpio_flags(dev->of_node,
+						     "enable-gpios", 0,
+						     &flags);
+	if (gpio_is_valid(panel->enable_gpio)) {
+		unsigned int value;
+
+		if (flags & OF_GPIO_ACTIVE_LOW)
+			panel->enable_gpio_flags |= GPIO_ACTIVE_LOW;
+
+		err = gpio_request(panel->enable_gpio, "enable");
+		if (err < 0) {
+			dev_err(dev, "failed to request GPIO#%u: %d\n",
+				panel->enable_gpio, err);
+			return err;
+		}
+
+		value = (panel->enable_gpio_flags & GPIO_ACTIVE_LOW) != 0;
+
+		err = gpio_direction_output(panel->enable_gpio, value);
+		if (err < 0) {
+			dev_err(dev, "failed to setup GPIO%u: %d\n",
+				panel->enable_gpio, err);
+			goto free_gpio;
+		}
+	}
+
+	backlight = of_parse_phandle(dev->of_node, "backlight", 0);
+	if (backlight) {
+		panel->backlight = of_find_backlight_by_node(backlight);
+		of_node_put(backlight);
+
+		if (!panel->backlight) {
+			err = -EPROBE_DEFER;
+			goto free_gpio;
+		}
+	}
+
+	ddc = of_parse_phandle(dev->of_node, "ddc-i2c-bus", 0);
+	if (ddc) {
+		panel->ddc = of_find_i2c_adapter_by_node(ddc);
+		of_node_put(ddc);
+
+		if (!panel->ddc) {
+			err = -EPROBE_DEFER;
+			goto free_backlight;
+		}
+	}
+
+	drm_panel_init(&panel->base);
+	panel->base.dev = dev;
+	panel->base.funcs = &panel_simple_funcs;
+
+	err = drm_panel_add(&panel->base);
+	if (err < 0)
+		goto free_ddc;
+
+	dev_set_drvdata(dev, panel);
+
+	return 0;
+
+free_ddc:
+	if (panel->ddc)
+		put_device(&panel->ddc->dev);
+free_backlight:
+	if (panel->backlight)
+		put_device(&panel->backlight->dev);
+free_gpio:
+	if (gpio_is_valid(panel->enable_gpio))
+		gpio_free(panel->enable_gpio);
+
+	return err;
+}
+
+static int panel_simple_remove(struct device *dev)
+{
+	struct panel_simple *panel = dev_get_drvdata(dev);
+
+	drm_panel_detach(&panel->base);
+	drm_panel_remove(&panel->base);
+
+	panel_simple_disable(&panel->base);
+
+	if (panel->ddc)
+		put_device(&panel->ddc->dev);
+
+	if (panel->backlight)
+		put_device(&panel->backlight->dev);
+
+	if (gpio_is_valid(panel->enable_gpio))
+		gpio_free(panel->enable_gpio);
+
+	regulator_disable(panel->supply);
+
+	return 0;
+}
+
+static const struct drm_display_mode auo_b101aw03_mode = {
+	.clock = 51450,
+	.hdisplay = 1024,
+	.hsync_start = 1024 + 156,
+	.hsync_end = 1024 + 156 + 8,
+	.htotal = 1024 + 156 + 8 + 156,
+	.vdisplay = 600,
+	.vsync_start = 600 + 16,
+	.vsync_end = 600 + 16 + 6,
+	.vtotal = 600 + 16 + 6 + 16,
+	.vrefresh = 60,
+};
+
+static const struct panel_desc auo_b101aw03 = {
+	.modes = &auo_b101aw03_mode,
+	.num_modes = 1,
+	.size = {
+		.width = 223,
+		.height = 125,
+	},
+};
+
+static const struct drm_display_mode chunghwa_claa101wa01a_mode = {
+	.clock = 72070,
+	.hdisplay = 1366,
+	.hsync_start = 1366 + 58,
+	.hsync_end = 1366 + 58 + 58,
+	.htotal = 1366 + 58 + 58 + 58,
+	.vdisplay = 768,
+	.vsync_start = 768 + 4,
+	.vsync_end = 768 + 4 + 4,
+	.vtotal = 768 + 4 + 4 + 4,
+	.vrefresh = 60,
+};
+
+static const struct panel_desc chunghwa_claa101wa01a = {
+	.modes = &chunghwa_claa101wa01a_mode,
+	.num_modes = 1,
+	.size = {
+		.width = 220,
+		.height = 120,
+	},
+};
+
+static const struct drm_display_mode chunghwa_claa101wb01_mode = {
+	.clock = 69300,
+	.hdisplay = 1366,
+	.hsync_start = 1366 + 48,
+	.hsync_end = 1366 + 48 + 32,
+	.htotal = 1366 + 48 + 32 + 20,
+	.vdisplay = 768,
+	.vsync_start = 768 + 16,
+	.vsync_end = 768 + 16 + 8,
+	.vtotal = 768 + 16 + 8 + 16,
+	.vrefresh = 60,
+};
+
+static const struct panel_desc chunghwa_claa101wb01 = {
+	.modes = &chunghwa_claa101wb01_mode,
+	.num_modes = 1,
+	.size = {
+		.width = 223,
+		.height = 125,
+	},
+};
+
+static const struct drm_display_mode samsung_ltn101nt05_mode = {
+	.clock = 54030,
+	.hdisplay = 1024,
+	.hsync_start = 1024 + 24,
+	.hsync_end = 1024 + 24 + 136,
+	.htotal = 1024 + 24 + 136 + 160,
+	.vdisplay = 600,
+	.vsync_start = 600 + 3,
+	.vsync_end = 600 + 3 + 6,
+	.vtotal = 600 + 3 + 6 + 61,
+	.vrefresh = 60,
+};
+
+static const struct panel_desc samsung_ltn101nt05 = {
+	.modes = &samsung_ltn101nt05_mode,
+	.num_modes = 1,
+	.size = {
+		.width = 1024,
+		.height = 600,
+	},
+};
+
+static const struct of_device_id platform_of_match[] = {
+	{
+		.compatible = "auo,b101aw03",
+		.data = &auo_b101aw03,
+	}, {
+		.compatible = "chunghwa,claa101wa01a",
+		.data = &chunghwa_claa101wa01a
+	}, {
+		.compatible = "chunghwa,claa101wb01",
+		.data = &chunghwa_claa101wb01
+	}, {
+		.compatible = "samsung,ltn101nt05",
+		.data = &samsung_ltn101nt05,
+	}, {
+		.compatible = "simple-panel",
+	}, {
+		/* sentinel */
+	}
+};
+MODULE_DEVICE_TABLE(of, platform_of_match);
+
+static int panel_simple_platform_probe(struct platform_device *pdev)
+{
+	const struct of_device_id *id;
+
+	id = of_match_node(platform_of_match, pdev->dev.of_node);
+	if (!id)
+		return -ENODEV;
+
+	return panel_simple_probe(&pdev->dev, id->data);
+}
+
+static int panel_simple_platform_remove(struct platform_device *pdev)
+{
+	return panel_simple_remove(&pdev->dev);
+}
+
+static struct platform_driver panel_simple_platform_driver = {
+	.driver = {
+		.name = "panel-simple",
+		.owner = THIS_MODULE,
+		.of_match_table = platform_of_match,
+	},
+	.probe = panel_simple_platform_probe,
+	.remove = panel_simple_platform_remove,
+};
+
+struct panel_desc_dsi {
+	struct panel_desc desc;
+
+	enum mipi_dsi_pixel_format format;
+	unsigned int lanes;
+};
+
+static const struct drm_display_mode panasonic_vvx10f004b00_mode = {
+	.clock = 157200,
+	.hdisplay = 1920,
+	.hsync_start = 1920 + 154,
+	.hsync_end = 1920 + 154 + 16,
+	.htotal = 1920 + 154 + 16 + 32,
+	.vdisplay = 1200,
+	.vsync_start = 1200 + 17,
+	.vsync_end = 1200 + 17 + 2,
+	.vtotal = 1200 + 17 + 2 + 16,
+	.vrefresh = 60,
+};
+
+static const struct panel_desc_dsi panasonic_vvx10f004b00 = {
+	.desc = {
+		.modes = &panasonic_vvx10f004b00_mode,
+		.num_modes = 1,
+		.size = {
+			.width = 217,
+			.height = 136,
+		},
+	},
+	.format = MIPI_DSI_FMT_RGB888,
+	.lanes = 4,
+};
+
+static const struct of_device_id dsi_of_match[] = {
+	{
+		.compatible = "panasonic,vvx10f004b00",
+		.data = &panasonic_vvx10f004b00
+	}, {
+		/* sentinel */
+	}
+};
+MODULE_DEVICE_TABLE(of, dsi_of_match);
+
+static int panel_simple_dsi_probe(struct mipi_dsi_device *dsi)
+{
+	const struct panel_desc_dsi *desc;
+	const struct of_device_id *id;
+	int err;
+
+	id = of_match_node(dsi_of_match, dsi->dev.of_node);
+	if (!id)
+		return -ENODEV;
+
+	desc = id->data;
+
+	err = panel_simple_probe(&dsi->dev, &desc->desc);
+	if (err < 0)
+		return err;
+
+	dsi->format = desc->format;
+	dsi->lanes = desc->lanes;
+
+	return mipi_dsi_attach(dsi);
+}
+
+static int panel_simple_dsi_remove(struct mipi_dsi_device *dsi)
+{
+	int err;
+
+	err = mipi_dsi_detach(dsi);
+	if (err < 0)
+		dev_err(&dsi->dev, "failed to detach from DSI host: %d\n", err);
+
+	return panel_simple_remove(&dsi->dev);
+}
+
+static struct mipi_dsi_driver panel_simple_dsi_driver = {
+	.driver = {
+		.name = "panel-simple-dsi",
+		.owner = THIS_MODULE,
+		.of_match_table = dsi_of_match,
+	},
+	.probe = panel_simple_dsi_probe,
+	.remove = panel_simple_dsi_remove,
+};
+
+static int __init panel_simple_init(void)
+{
+	int err;
+
+	err = platform_driver_register(&panel_simple_platform_driver);
+	if (err < 0)
+		return err;
+
+	if (IS_ENABLED(CONFIG_DRM_MIPI_DSI)) {
+		err = mipi_dsi_driver_register(&panel_simple_dsi_driver);
+		if (err < 0)
+			return err;
+	}
+
+	return 0;
+}
+module_init(panel_simple_init);
+
+static void __exit panel_simple_exit(void)
+{
+	if (IS_ENABLED(CONFIG_DRM_MIPI_DSI))
+		mipi_dsi_driver_unregister(&panel_simple_dsi_driver);
+
+	platform_driver_unregister(&panel_simple_platform_driver);
+}
+module_exit(panel_simple_exit);
+
+MODULE_AUTHOR("Thierry Reding <treding@nvidia.com>");
+MODULE_DESCRIPTION("DRM Driver for Simple Panels");
+MODULE_LICENSE("GPL and additional rights");
diff --git a/drivers/gpu/drm/qxl/qxl_display.c b/drivers/gpu/drm/qxl/qxl_display.c
index d70aafb83307..798bde2e5881 100644
--- a/drivers/gpu/drm/qxl/qxl_display.c
+++ b/drivers/gpu/drm/qxl/qxl_display.c
@@ -399,10 +399,14 @@ static int qxl_framebuffer_surface_dirty(struct drm_framebuffer *fb,
 	struct qxl_bo *qobj;
 	int inc = 1;
 
+	drm_modeset_lock_all(fb->dev);
+
 	qobj = gem_to_qxl_bo(qxl_fb->obj);
 	/* if we aren't primary surface ignore this */
-	if (!qobj->is_primary)
+	if (!qobj->is_primary) {
+		drm_modeset_unlock_all(fb->dev);
 		return 0;
+	}
 
 	if (!num_clips) {
 		num_clips = 1;
@@ -417,6 +421,9 @@ static int qxl_framebuffer_surface_dirty(struct drm_framebuffer *fb,
 
 	qxl_draw_dirty_fb(qdev, qxl_fb, qobj, flags, color,
 			  clips, num_clips, inc);
+
+	drm_modeset_unlock_all(fb->dev);
+
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/qxl/qxl_drv.h b/drivers/gpu/drm/qxl/qxl_drv.h
index 7bda32f68d3b..36ed40ba773f 100644
--- a/drivers/gpu/drm/qxl/qxl_drv.h
+++ b/drivers/gpu/drm/qxl/qxl_drv.h
@@ -534,7 +534,7 @@ void qxl_debugfs_takedown(struct drm_minor *minor);
 
 /* qxl_irq.c */
 int qxl_irq_init(struct qxl_device *qdev);
-irqreturn_t qxl_irq_handler(DRM_IRQ_ARGS);
+irqreturn_t qxl_irq_handler(int irq, void *arg);
 
 /* qxl_fb.c */
 int qxl_fb_init(struct qxl_device *qdev);
diff --git a/drivers/gpu/drm/qxl/qxl_ioctl.c b/drivers/gpu/drm/qxl/qxl_ioctl.c
index 7b95c75e9626..0bb86e6d41b4 100644
--- a/drivers/gpu/drm/qxl/qxl_ioctl.c
+++ b/drivers/gpu/drm/qxl/qxl_ioctl.c
@@ -200,7 +200,7 @@ static int qxl_process_single_command(struct qxl_device *qdev,
 	for (i = 0; i < cmd->relocs_num; ++i) {
 		struct drm_qxl_reloc reloc;
 
-		if (DRM_COPY_FROM_USER(&reloc,
+		if (copy_from_user(&reloc,
 				       &((struct drm_qxl_reloc *)(uintptr_t)cmd->relocs)[i],
 				       sizeof(reloc))) {
 			ret = -EFAULT;
@@ -297,7 +297,7 @@ static int qxl_execbuffer_ioctl(struct drm_device *dev, void *data,
 		struct drm_qxl_command *commands =
 			(struct drm_qxl_command *)(uintptr_t)execbuffer->commands;
 
-		if (DRM_COPY_FROM_USER(&user_cmd, &commands[cmd_num],
+		if (copy_from_user(&user_cmd, &commands[cmd_num],
 				       sizeof(user_cmd)))
 			return -EFAULT;
 
diff --git a/drivers/gpu/drm/qxl/qxl_irq.c b/drivers/gpu/drm/qxl/qxl_irq.c
index 21393dc4700a..28f84b4fce32 100644
--- a/drivers/gpu/drm/qxl/qxl_irq.c
+++ b/drivers/gpu/drm/qxl/qxl_irq.c
@@ -25,7 +25,7 @@
 
 #include "qxl_drv.h"
 
-irqreturn_t qxl_irq_handler(DRM_IRQ_ARGS)
+irqreturn_t qxl_irq_handler(int irq, void *arg)
 {
 	struct drm_device *dev = (struct drm_device *) arg;
 	struct qxl_device *qdev = (struct qxl_device *)dev->dev_private;
diff --git a/drivers/gpu/drm/qxl/qxl_kms.c b/drivers/gpu/drm/qxl/qxl_kms.c
index e5ca498be920..fd88eb4a3f79 100644
--- a/drivers/gpu/drm/qxl/qxl_kms.c
+++ b/drivers/gpu/drm/qxl/qxl_kms.c
@@ -115,7 +115,7 @@ static void qxl_gc_work(struct work_struct *work)
 	qxl_garbage_collect(qdev);
 }
 
-int qxl_device_init(struct qxl_device *qdev,
+static int qxl_device_init(struct qxl_device *qdev,
 		    struct drm_device *ddev,
 		    struct pci_dev *pdev,
 		    unsigned long flags)
diff --git a/drivers/gpu/drm/r128/r128_cce.c b/drivers/gpu/drm/r128/r128_cce.c
index c451257f08fb..59459fe4e8c5 100644
--- a/drivers/gpu/drm/r128/r128_cce.c
+++ b/drivers/gpu/drm/r128/r128_cce.c
@@ -892,10 +892,10 @@ static int r128_cce_get_buffers(struct drm_device *dev,
 
 		buf->file_priv = file_priv;
 
-		if (DRM_COPY_TO_USER(&d->request_indices[i], &buf->idx,
+		if (copy_to_user(&d->request_indices[i], &buf->idx,
 				     sizeof(buf->idx)))
 			return -EFAULT;
-		if (DRM_COPY_TO_USER(&d->request_sizes[i], &buf->total,
+		if (copy_to_user(&d->request_sizes[i], &buf->total,
 				     sizeof(buf->total)))
 			return -EFAULT;
 
diff --git a/drivers/gpu/drm/r128/r128_drv.h b/drivers/gpu/drm/r128/r128_drv.h
index 56eb5e3f5439..5bf3f5ff805d 100644
--- a/drivers/gpu/drm/r128/r128_drv.h
+++ b/drivers/gpu/drm/r128/r128_drv.h
@@ -154,7 +154,7 @@ extern int r128_do_cleanup_cce(struct drm_device *dev);
 extern int r128_enable_vblank(struct drm_device *dev, int crtc);
 extern void r128_disable_vblank(struct drm_device *dev, int crtc);
 extern u32 r128_get_vblank_counter(struct drm_device *dev, int crtc);
-extern irqreturn_t r128_driver_irq_handler(DRM_IRQ_ARGS);
+extern irqreturn_t r128_driver_irq_handler(int irq, void *arg);
 extern void r128_driver_irq_preinstall(struct drm_device *dev);
 extern int r128_driver_irq_postinstall(struct drm_device *dev);
 extern void r128_driver_irq_uninstall(struct drm_device *dev);
@@ -514,7 +514,7 @@ do {									\
 	if (R128_VERBOSE)						\
 		DRM_INFO("COMMIT_RING() tail=0x%06x\n",			\
 			 dev_priv->ring.tail);				\
-	DRM_MEMORYBARRIER();						\
+	mb();						\
 	R128_WRITE(R128_PM4_BUFFER_DL_WPTR, dev_priv->ring.tail);	\
 	R128_READ(R128_PM4_BUFFER_DL_WPTR);				\
 } while (0)
diff --git a/drivers/gpu/drm/r128/r128_ioc32.c b/drivers/gpu/drm/r128/r128_ioc32.c
index a954c548201e..b0d0fd3e4376 100644
--- a/drivers/gpu/drm/r128/r128_ioc32.c
+++ b/drivers/gpu/drm/r128/r128_ioc32.c
@@ -33,6 +33,7 @@
 
 #include <drm/drmP.h>
 #include <drm/r128_drm.h>
+#include "r128_drv.h"
 
 typedef struct drm_r128_init32 {
 	int func;
diff --git a/drivers/gpu/drm/r128/r128_irq.c b/drivers/gpu/drm/r128/r128_irq.c
index 2ea4f09d2691..c2ae496babb7 100644
--- a/drivers/gpu/drm/r128/r128_irq.c
+++ b/drivers/gpu/drm/r128/r128_irq.c
@@ -44,7 +44,7 @@ u32 r128_get_vblank_counter(struct drm_device *dev, int crtc)
 	return atomic_read(&dev_priv->vbl_received);
 }
 
-irqreturn_t r128_driver_irq_handler(DRM_IRQ_ARGS)
+irqreturn_t r128_driver_irq_handler(int irq, void *arg)
 {
 	struct drm_device *dev = (struct drm_device *) arg;
 	drm_r128_private_t *dev_priv = (drm_r128_private_t *) dev->dev_private;
diff --git a/drivers/gpu/drm/r128/r128_state.c b/drivers/gpu/drm/r128/r128_state.c
index 01dd9aef9f0e..e806dacd452f 100644
--- a/drivers/gpu/drm/r128/r128_state.c
+++ b/drivers/gpu/drm/r128/r128_state.c
@@ -895,31 +895,22 @@ static int r128_cce_dispatch_write_span(struct drm_device *dev,
 	if (count > 4096 || count <= 0)
 		return -EMSGSIZE;
 
-	if (DRM_COPY_FROM_USER(&x, depth->x, sizeof(x)))
+	if (copy_from_user(&x, depth->x, sizeof(x)))
 		return -EFAULT;
-	if (DRM_COPY_FROM_USER(&y, depth->y, sizeof(y)))
+	if (copy_from_user(&y, depth->y, sizeof(y)))
 		return -EFAULT;
 
 	buffer_size = depth->n * sizeof(u32);
-	buffer = kmalloc(buffer_size, GFP_KERNEL);
-	if (buffer == NULL)
-		return -ENOMEM;
-	if (DRM_COPY_FROM_USER(buffer, depth->buffer, buffer_size)) {
-		kfree(buffer);
-		return -EFAULT;
-	}
+	buffer = memdup_user(depth->buffer, buffer_size);
+	if (IS_ERR(buffer))
+		return PTR_ERR(buffer);
 
 	mask_size = depth->n * sizeof(u8);
 	if (depth->mask) {
-		mask = kmalloc(mask_size, GFP_KERNEL);
-		if (mask == NULL) {
+		mask = memdup_user(depth->mask, mask_size);
+		if (IS_ERR(mask)) {
 			kfree(buffer);
-			return -ENOMEM;
-		}
-		if (DRM_COPY_FROM_USER(mask, depth->mask, mask_size)) {
-			kfree(buffer);
-			kfree(mask);
-			return -EFAULT;
+			return PTR_ERR(mask);
 		}
 
 		for (i = 0; i < count; i++, x++) {
@@ -999,46 +990,33 @@ static int r128_cce_dispatch_write_pixels(struct drm_device *dev,
 		kfree(x);
 		return -ENOMEM;
 	}
-	if (DRM_COPY_FROM_USER(x, depth->x, xbuf_size)) {
+	if (copy_from_user(x, depth->x, xbuf_size)) {
 		kfree(x);
 		kfree(y);
 		return -EFAULT;
 	}
-	if (DRM_COPY_FROM_USER(y, depth->y, xbuf_size)) {
+	if (copy_from_user(y, depth->y, xbuf_size)) {
 		kfree(x);
 		kfree(y);
 		return -EFAULT;
 	}
 
 	buffer_size = depth->n * sizeof(u32);
-	buffer = kmalloc(buffer_size, GFP_KERNEL);
-	if (buffer == NULL) {
-		kfree(x);
-		kfree(y);
-		return -ENOMEM;
-	}
-	if (DRM_COPY_FROM_USER(buffer, depth->buffer, buffer_size)) {
+	buffer = memdup_user(depth->buffer, buffer_size);
+	if (IS_ERR(buffer)) {
 		kfree(x);
 		kfree(y);
-		kfree(buffer);
-		return -EFAULT;
+		return PTR_ERR(buffer);
 	}
 
 	if (depth->mask) {
 		mask_size = depth->n * sizeof(u8);
-		mask = kmalloc(mask_size, GFP_KERNEL);
-		if (mask == NULL) {
-			kfree(x);
-			kfree(y);
-			kfree(buffer);
-			return -ENOMEM;
-		}
-		if (DRM_COPY_FROM_USER(mask, depth->mask, mask_size)) {
+		mask = memdup_user(depth->mask, mask_size);
+		if (IS_ERR(mask)) {
 			kfree(x);
 			kfree(y);
 			kfree(buffer);
-			kfree(mask);
-			return -EFAULT;
+			return PTR_ERR(mask);
 		}
 
 		for (i = 0; i < count; i++) {
@@ -1107,9 +1085,9 @@ static int r128_cce_dispatch_read_span(struct drm_device *dev,
 	if (count > 4096 || count <= 0)
 		return -EMSGSIZE;
 
-	if (DRM_COPY_FROM_USER(&x, depth->x, sizeof(x)))
+	if (copy_from_user(&x, depth->x, sizeof(x)))
 		return -EFAULT;
-	if (DRM_COPY_FROM_USER(&y, depth->y, sizeof(y)))
+	if (copy_from_user(&y, depth->y, sizeof(y)))
 		return -EFAULT;
 
 	BEGIN_RING(7);
@@ -1162,12 +1140,12 @@ static int r128_cce_dispatch_read_pixels(struct drm_device *dev,
 		kfree(x);
 		return -ENOMEM;
 	}
-	if (DRM_COPY_FROM_USER(x, depth->x, xbuf_size)) {
+	if (copy_from_user(x, depth->x, xbuf_size)) {
 		kfree(x);
 		kfree(y);
 		return -EFAULT;
 	}
-	if (DRM_COPY_FROM_USER(y, depth->y, ybuf_size)) {
+	if (copy_from_user(y, depth->y, ybuf_size)) {
 		kfree(x);
 		kfree(y);
 		return -EFAULT;
@@ -1524,7 +1502,7 @@ static int r128_cce_stipple(struct drm_device *dev, void *data, struct drm_file
 
 	DEV_INIT_TEST_WITH_RETURN(dev_priv);
 
-	if (DRM_COPY_FROM_USER(&mask, stipple->mask, 32 * sizeof(u32)))
+	if (copy_from_user(&mask, stipple->mask, 32 * sizeof(u32)))
 		return -EFAULT;
 
 	RING_SPACE_TEST_WITH_RETURN(dev_priv);
@@ -1622,7 +1600,7 @@ static int r128_getparam(struct drm_device *dev, void *data, struct drm_file *fi
 		return -EINVAL;
 	}
 
-	if (DRM_COPY_TO_USER(param->value, &value, sizeof(int))) {
+	if (copy_to_user(param->value, &value, sizeof(int))) {
 		DRM_ERROR("copy_to_user\n");
 		return -EFAULT;
 	}
diff --git a/drivers/gpu/drm/radeon/atombios_crtc.c b/drivers/gpu/drm/radeon/atombios_crtc.c
index 0b9621c9aeea..a9338c85630f 100644
--- a/drivers/gpu/drm/radeon/atombios_crtc.c
+++ b/drivers/gpu/drm/radeon/atombios_crtc.c
@@ -209,6 +209,16 @@ static void atombios_enable_crtc_memreq(struct drm_crtc *crtc, int state)
 	atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args);
 }
 
+static const u32 vga_control_regs[6] =
+{
+	AVIVO_D1VGA_CONTROL,
+	AVIVO_D2VGA_CONTROL,
+	EVERGREEN_D3VGA_CONTROL,
+	EVERGREEN_D4VGA_CONTROL,
+	EVERGREEN_D5VGA_CONTROL,
+	EVERGREEN_D6VGA_CONTROL,
+};
+
 static void atombios_blank_crtc(struct drm_crtc *crtc, int state)
 {
 	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
@@ -216,13 +226,23 @@ static void atombios_blank_crtc(struct drm_crtc *crtc, int state)
 	struct radeon_device *rdev = dev->dev_private;
 	int index = GetIndexIntoMasterTable(COMMAND, BlankCRTC);
 	BLANK_CRTC_PS_ALLOCATION args;
+	u32 vga_control = 0;
 
 	memset(&args, 0, sizeof(args));
 
+	if (ASIC_IS_DCE8(rdev)) {
+		vga_control = RREG32(vga_control_regs[radeon_crtc->crtc_id]);
+		WREG32(vga_control_regs[radeon_crtc->crtc_id], vga_control | 1);
+	}
+
 	args.ucCRTC = radeon_crtc->crtc_id;
 	args.ucBlanking = state;
 
 	atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args);
+
+	if (ASIC_IS_DCE8(rdev)) {
+		WREG32(vga_control_regs[radeon_crtc->crtc_id], vga_control);
+	}
 }
 
 static void atombios_powergate_crtc(struct drm_crtc *crtc, int state)
@@ -423,7 +443,17 @@ static void atombios_crtc_program_ss(struct radeon_device *rdev,
 	int index = GetIndexIntoMasterTable(COMMAND, EnableSpreadSpectrumOnPPLL);
 	union atom_enable_ss args;
 
-	if (!enable) {
+	if (enable) {
+		/* Don't mess with SS if percentage is 0 or external ss.
+		 * SS is already disabled previously, and disabling it
+		 * again can cause display problems if the pll is already
+		 * programmed.
+		 */
+		if (ss->percentage == 0)
+			return;
+		if (ss->type & ATOM_EXTERNAL_SS_MASK)
+			return;
+	} else {
 		for (i = 0; i < rdev->num_crtc; i++) {
 			if (rdev->mode_info.crtcs[i] &&
 			    rdev->mode_info.crtcs[i]->enabled &&
@@ -459,8 +489,6 @@ static void atombios_crtc_program_ss(struct radeon_device *rdev,
 		args.v3.usSpreadSpectrumAmount = cpu_to_le16(ss->amount);
 		args.v3.usSpreadSpectrumStep = cpu_to_le16(ss->step);
 		args.v3.ucEnable = enable;
-		if ((ss->percentage == 0) || (ss->type & ATOM_EXTERNAL_SS_MASK) || ASIC_IS_DCE61(rdev))
-			args.v3.ucEnable = ATOM_DISABLE;
 	} else if (ASIC_IS_DCE4(rdev)) {
 		args.v2.usSpreadSpectrumPercentage = cpu_to_le16(ss->percentage);
 		args.v2.ucSpreadSpectrumType = ss->type & ATOM_SS_CENTRE_SPREAD_MODE_MASK;
@@ -480,8 +508,6 @@ static void atombios_crtc_program_ss(struct radeon_device *rdev,
 		args.v2.usSpreadSpectrumAmount = cpu_to_le16(ss->amount);
 		args.v2.usSpreadSpectrumStep = cpu_to_le16(ss->step);
 		args.v2.ucEnable = enable;
-		if ((ss->percentage == 0) || (ss->type & ATOM_EXTERNAL_SS_MASK) || ASIC_IS_DCE41(rdev))
-			args.v2.ucEnable = ATOM_DISABLE;
 	} else if (ASIC_IS_DCE3(rdev)) {
 		args.v1.usSpreadSpectrumPercentage = cpu_to_le16(ss->percentage);
 		args.v1.ucSpreadSpectrumType = ss->type & ATOM_SS_CENTRE_SPREAD_MODE_MASK;
@@ -503,8 +529,7 @@ static void atombios_crtc_program_ss(struct radeon_device *rdev,
 		args.lvds_ss_2.ucSpreadSpectrumRange = ss->range;
 		args.lvds_ss_2.ucEnable = enable;
 	} else {
-		if ((enable == ATOM_DISABLE) || (ss->percentage == 0) ||
-		    (ss->type & ATOM_EXTERNAL_SS_MASK)) {
+		if (enable == ATOM_DISABLE) {
 			atombios_disable_ss(rdev, pll_id);
 			return;
 		}
@@ -938,11 +963,14 @@ static bool atombios_crtc_prepare_pll(struct drm_crtc *crtc, struct drm_display_
 							radeon_atombios_get_ppll_ss_info(rdev,
 											 &radeon_crtc->ss,
 											 ATOM_DP_SS_ID1);
-				} else
+				} else {
 					radeon_crtc->ss_enabled =
 						radeon_atombios_get_ppll_ss_info(rdev,
 										 &radeon_crtc->ss,
 										 ATOM_DP_SS_ID1);
+				}
+				/* disable spread spectrum on DCE3 DP */
+				radeon_crtc->ss_enabled = false;
 			}
 			break;
 		case ATOM_ENCODER_MODE_LVDS:
@@ -1039,15 +1067,17 @@ static void atombios_crtc_set_pll(struct drm_crtc *crtc, struct drm_display_mode
 		/* calculate ss amount and step size */
 		if (ASIC_IS_DCE4(rdev)) {
 			u32 step_size;
-			u32 amount = (((fb_div * 10) + frac_fb_div) * radeon_crtc->ss.percentage) / 10000;
+			u32 amount = (((fb_div * 10) + frac_fb_div) *
+				      (u32)radeon_crtc->ss.percentage) /
+				(100 * (u32)radeon_crtc->ss.percentage_divider);
 			radeon_crtc->ss.amount = (amount / 10) & ATOM_PPLL_SS_AMOUNT_V2_FBDIV_MASK;
 			radeon_crtc->ss.amount |= ((amount - (amount / 10)) << ATOM_PPLL_SS_AMOUNT_V2_NFRAC_SHIFT) &
 				ATOM_PPLL_SS_AMOUNT_V2_NFRAC_MASK;
 			if (radeon_crtc->ss.type & ATOM_PPLL_SS_TYPE_V2_CENTRE_SPREAD)
-				step_size = (4 * amount * ref_div * (radeon_crtc->ss.rate * 2048)) /
+				step_size = (4 * amount * ref_div * ((u32)radeon_crtc->ss.rate * 2048)) /
 					(125 * 25 * pll->reference_freq / 100);
 			else
-				step_size = (2 * amount * ref_div * (radeon_crtc->ss.rate * 2048)) /
+				step_size = (2 * amount * ref_div * ((u32)radeon_crtc->ss.rate * 2048)) /
 					(125 * 25 * pll->reference_freq / 100);
 			radeon_crtc->ss.step = step_size;
 		}
diff --git a/drivers/gpu/drm/radeon/atombios_dp.c b/drivers/gpu/drm/radeon/atombios_dp.c
index fb3ae07a1469..4ad7643fce5f 100644
--- a/drivers/gpu/drm/radeon/atombios_dp.c
+++ b/drivers/gpu/drm/radeon/atombios_dp.c
@@ -157,21 +157,22 @@ static int radeon_dp_aux_native_write(struct radeon_connector *radeon_connector,
 
 	msg[0] = address;
 	msg[1] = address >> 8;
-	msg[2] = AUX_NATIVE_WRITE << 4;
+	msg[2] = DP_AUX_NATIVE_WRITE << 4;
 	msg[3] = (msg_bytes << 4) | (send_bytes - 1);
 	memcpy(&msg[4], send, send_bytes);
 
-	for (retry = 0; retry < 4; retry++) {
+	for (retry = 0; retry < 7; retry++) {
 		ret = radeon_process_aux_ch(dig_connector->dp_i2c_bus,
 					    msg, msg_bytes, NULL, 0, delay, &ack);
 		if (ret == -EBUSY)
 			continue;
 		else if (ret < 0)
 			return ret;
-		if ((ack & AUX_NATIVE_REPLY_MASK) == AUX_NATIVE_REPLY_ACK)
+		ack >>= 4;
+		if ((ack & DP_AUX_NATIVE_REPLY_MASK) == DP_AUX_NATIVE_REPLY_ACK)
 			return send_bytes;
-		else if ((ack & AUX_NATIVE_REPLY_MASK) == AUX_NATIVE_REPLY_DEFER)
-			udelay(400);
+		else if ((ack & DP_AUX_NATIVE_REPLY_MASK) == DP_AUX_NATIVE_REPLY_DEFER)
+			usleep_range(400, 500);
 		else
 			return -EIO;
 	}
@@ -191,20 +192,21 @@ static int radeon_dp_aux_native_read(struct radeon_connector *radeon_connector,
 
 	msg[0] = address;
 	msg[1] = address >> 8;
-	msg[2] = AUX_NATIVE_READ << 4;
+	msg[2] = DP_AUX_NATIVE_READ << 4;
 	msg[3] = (msg_bytes << 4) | (recv_bytes - 1);
 
-	for (retry = 0; retry < 4; retry++) {
+	for (retry = 0; retry < 7; retry++) {
 		ret = radeon_process_aux_ch(dig_connector->dp_i2c_bus,
 					    msg, msg_bytes, recv, recv_bytes, delay, &ack);
 		if (ret == -EBUSY)
 			continue;
 		else if (ret < 0)
 			return ret;
-		if ((ack & AUX_NATIVE_REPLY_MASK) == AUX_NATIVE_REPLY_ACK)
+		ack >>= 4;
+		if ((ack & DP_AUX_NATIVE_REPLY_MASK) == DP_AUX_NATIVE_REPLY_ACK)
 			return ret;
-		else if ((ack & AUX_NATIVE_REPLY_MASK) == AUX_NATIVE_REPLY_DEFER)
-			udelay(400);
+		else if ((ack & DP_AUX_NATIVE_REPLY_MASK) == DP_AUX_NATIVE_REPLY_DEFER)
+			usleep_range(400, 500);
 		else if (ret == 0)
 			return -EPROTO;
 		else
@@ -246,12 +248,12 @@ int radeon_dp_i2c_aux_ch(struct i2c_adapter *adapter, int mode,
 
 	/* Set up the command byte */
 	if (mode & MODE_I2C_READ)
-		msg[2] = AUX_I2C_READ << 4;
+		msg[2] = DP_AUX_I2C_READ << 4;
 	else
-		msg[2] = AUX_I2C_WRITE << 4;
+		msg[2] = DP_AUX_I2C_WRITE << 4;
 
 	if (!(mode & MODE_I2C_STOP))
-		msg[2] |= AUX_I2C_MOT << 4;
+		msg[2] |= DP_AUX_I2C_MOT << 4;
 
 	msg[0] = address;
 	msg[1] = address >> 8;
@@ -272,7 +274,7 @@ int radeon_dp_i2c_aux_ch(struct i2c_adapter *adapter, int mode,
 		break;
 	}
 
-	for (retry = 0; retry < 4; retry++) {
+	for (retry = 0; retry < 7; retry++) {
 		ret = radeon_process_aux_ch(auxch,
 					    msg, msg_bytes, reply, reply_bytes, 0, &ack);
 		if (ret == -EBUSY)
@@ -282,35 +284,35 @@ int radeon_dp_i2c_aux_ch(struct i2c_adapter *adapter, int mode,
 			return ret;
 		}
 
-		switch (ack & AUX_NATIVE_REPLY_MASK) {
-		case AUX_NATIVE_REPLY_ACK:
+		switch ((ack >> 4) & DP_AUX_NATIVE_REPLY_MASK) {
+		case DP_AUX_NATIVE_REPLY_ACK:
 			/* I2C-over-AUX Reply field is only valid
 			 * when paired with AUX ACK.
 			 */
 			break;
-		case AUX_NATIVE_REPLY_NACK:
+		case DP_AUX_NATIVE_REPLY_NACK:
 			DRM_DEBUG_KMS("aux_ch native nack\n");
 			return -EREMOTEIO;
-		case AUX_NATIVE_REPLY_DEFER:
+		case DP_AUX_NATIVE_REPLY_DEFER:
 			DRM_DEBUG_KMS("aux_ch native defer\n");
-			udelay(400);
+			usleep_range(500, 600);
 			continue;
 		default:
 			DRM_ERROR("aux_ch invalid native reply 0x%02x\n", ack);
 			return -EREMOTEIO;
 		}
 
-		switch (ack & AUX_I2C_REPLY_MASK) {
-		case AUX_I2C_REPLY_ACK:
+		switch ((ack >> 4) & DP_AUX_I2C_REPLY_MASK) {
+		case DP_AUX_I2C_REPLY_ACK:
 			if (mode == MODE_I2C_READ)
 				*read_byte = reply[0];
 			return ret;
-		case AUX_I2C_REPLY_NACK:
+		case DP_AUX_I2C_REPLY_NACK:
 			DRM_DEBUG_KMS("aux_i2c nack\n");
 			return -EREMOTEIO;
-		case AUX_I2C_REPLY_DEFER:
+		case DP_AUX_I2C_REPLY_DEFER:
 			DRM_DEBUG_KMS("aux_i2c defer\n");
-			udelay(400);
+			usleep_range(400, 500);
 			break;
 		default:
 			DRM_ERROR("aux_i2c invalid reply 0x%02x\n", ack);
@@ -671,9 +673,11 @@ static int radeon_dp_link_train_init(struct radeon_dp_link_train_info *dp_info)
 	u8 tmp;
 
 	/* power up the sink */
-	if (dp_info->dpcd[0] >= 0x11)
+	if (dp_info->dpcd[0] >= 0x11) {
 		radeon_write_dpcd_reg(dp_info->radeon_connector,
 				      DP_SET_POWER, DP_SET_POWER_D0);
+		usleep_range(1000, 2000);
+	}
 
 	/* possibly enable downspread on the sink */
 	if (dp_info->dpcd[3] & 0x1)
diff --git a/drivers/gpu/drm/radeon/atombios_i2c.c b/drivers/gpu/drm/radeon/atombios_i2c.c
index f685035dbe39..b5162c3b6111 100644
--- a/drivers/gpu/drm/radeon/atombios_i2c.c
+++ b/drivers/gpu/drm/radeon/atombios_i2c.c
@@ -27,8 +27,6 @@
 #include "radeon.h"
 #include "atom.h"
 
-extern void radeon_atom_copy_swap(u8 *dst, u8 *src, u8 num_bytes, bool to_le);
-
 #define TARGET_HW_I2C_CLOCK 50
 
 /* these are a limitation of ProcessI2cChannelTransaction not the hw */
diff --git a/drivers/gpu/drm/radeon/btc_dpm.c b/drivers/gpu/drm/radeon/btc_dpm.c
index 9b6950d9b3c0..0fbd36f3d4e9 100644
--- a/drivers/gpu/drm/radeon/btc_dpm.c
+++ b/drivers/gpu/drm/radeon/btc_dpm.c
@@ -49,6 +49,7 @@ struct rv7xx_ps *rv770_get_ps(struct radeon_ps *rps);
 struct rv7xx_power_info *rv770_get_pi(struct radeon_device *rdev);
 struct evergreen_power_info *evergreen_get_pi(struct radeon_device *rdev);
 
+extern int ni_mc_load_microcode(struct radeon_device *rdev);
 
 //********* BARTS **************//
 static const u32 barts_cgcg_cgls_default[] =
@@ -2510,21 +2511,6 @@ int btc_dpm_enable(struct radeon_device *rdev)
 	if (eg_pi->ls_clock_gating)
 		btc_ls_clock_gating_enable(rdev, true);
 
-	if (rdev->irq.installed &&
-	    r600_is_internal_thermal_sensor(rdev->pm.int_thermal_type)) {
-		PPSMC_Result result;
-
-		ret = rv770_set_thermal_temperature_range(rdev, R600_TEMP_RANGE_MIN, R600_TEMP_RANGE_MAX);
-		if (ret)
-			return ret;
-		rdev->irq.dpm_thermal = true;
-		radeon_irq_set(rdev);
-		result = rv770_send_msg_to_smc(rdev, PPSMC_MSG_EnableThermalInterrupt);
-
-		if (result != PPSMC_Result_OK)
-			DRM_DEBUG_KMS("Could not enable thermal interrupts.\n");
-	}
-
 	rv770_enable_auto_throttle_source(rdev, RADEON_DPM_AUTO_THROTTLE_SRC_THERMAL, true);
 
 	btc_init_stutter_mode(rdev);
@@ -2576,7 +2562,11 @@ void btc_dpm_disable(struct radeon_device *rdev)
 void btc_dpm_setup_asic(struct radeon_device *rdev)
 {
 	struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev);
+	int r;
 
+	r = ni_mc_load_microcode(rdev);
+	if (r)
+		DRM_ERROR("Failed to load MC firmware!\n");
 	rv770_get_memory_type(rdev);
 	rv740_read_clock_registers(rdev);
 	btc_read_arb_registers(rdev);
diff --git a/drivers/gpu/drm/radeon/ci_dpm.c b/drivers/gpu/drm/radeon/ci_dpm.c
index 1ed479976358..8d49104ca6c2 100644
--- a/drivers/gpu/drm/radeon/ci_dpm.c
+++ b/drivers/gpu/drm/radeon/ci_dpm.c
@@ -171,8 +171,7 @@ extern void si_trim_voltage_table_to_fit_state_table(struct radeon_device *rdev,
 						     struct atom_voltage_table *voltage_table);
 extern void cik_enter_rlc_safe_mode(struct radeon_device *rdev);
 extern void cik_exit_rlc_safe_mode(struct radeon_device *rdev);
-extern void cik_update_cg(struct radeon_device *rdev,
-			  u32 block, bool enable);
+extern int ci_mc_load_microcode(struct radeon_device *rdev);
 
 static int ci_get_std_voltage_value_sidd(struct radeon_device *rdev,
 					 struct atom_voltage_table_entry *voltage_table,
@@ -4503,8 +4502,8 @@ static void ci_get_memory_type(struct radeon_device *rdev)
 
 }
 
-void ci_update_current_ps(struct radeon_device *rdev,
-			  struct radeon_ps *rps)
+static void ci_update_current_ps(struct radeon_device *rdev,
+				 struct radeon_ps *rps)
 {
 	struct ci_ps *new_ps = ci_get_ps(rps);
 	struct ci_power_info *pi = ci_get_pi(rdev);
@@ -4514,8 +4513,8 @@ void ci_update_current_ps(struct radeon_device *rdev,
 	pi->current_rps.ps_priv = &pi->current_ps;
 }
 
-void ci_update_requested_ps(struct radeon_device *rdev,
-			    struct radeon_ps *rps)
+static void ci_update_requested_ps(struct radeon_device *rdev,
+				   struct radeon_ps *rps)
 {
 	struct ci_ps *new_ps = ci_get_ps(rps);
 	struct ci_power_info *pi = ci_get_pi(rdev);
@@ -4549,6 +4548,11 @@ void ci_dpm_post_set_power_state(struct radeon_device *rdev)
 
 void ci_dpm_setup_asic(struct radeon_device *rdev)
 {
+	int r;
+
+	r = ci_mc_load_microcode(rdev);
+	if (r)
+		DRM_ERROR("Failed to load MC firmware!\n");
 	ci_read_clock_registers(rdev);
 	ci_get_memory_type(rdev);
 	ci_enable_acpi_power_management(rdev);
@@ -4561,13 +4565,6 @@ int ci_dpm_enable(struct radeon_device *rdev)
 	struct radeon_ps *boot_ps = rdev->pm.dpm.boot_ps;
 	int ret;
 
-	cik_update_cg(rdev, (RADEON_CG_BLOCK_GFX |
-			     RADEON_CG_BLOCK_MC |
-			     RADEON_CG_BLOCK_SDMA |
-			     RADEON_CG_BLOCK_BIF |
-			     RADEON_CG_BLOCK_UVD |
-			     RADEON_CG_BLOCK_HDP), false);
-
 	if (ci_is_smc_running(rdev))
 		return -EINVAL;
 	if (pi->voltage_control != CISLANDS_VOLTAGE_CONTROL_NONE) {
@@ -4665,6 +4662,18 @@ int ci_dpm_enable(struct radeon_device *rdev)
 		DRM_ERROR("ci_enable_power_containment failed\n");
 		return ret;
 	}
+
+	ci_enable_auto_throttle_source(rdev, RADEON_DPM_AUTO_THROTTLE_SRC_THERMAL, true);
+
+	ci_update_current_ps(rdev, boot_ps);
+
+	return 0;
+}
+
+int ci_dpm_late_enable(struct radeon_device *rdev)
+{
+	int ret;
+
 	if (rdev->irq.installed &&
 	    r600_is_internal_thermal_sensor(rdev->pm.int_thermal_type)) {
 #if 0
@@ -4685,19 +4694,8 @@ int ci_dpm_enable(struct radeon_device *rdev)
 #endif
 	}
 
-	ci_enable_auto_throttle_source(rdev, RADEON_DPM_AUTO_THROTTLE_SRC_THERMAL, true);
-
 	ci_dpm_powergate_uvd(rdev, true);
 
-	cik_update_cg(rdev, (RADEON_CG_BLOCK_GFX |
-			     RADEON_CG_BLOCK_MC |
-			     RADEON_CG_BLOCK_SDMA |
-			     RADEON_CG_BLOCK_BIF |
-			     RADEON_CG_BLOCK_UVD |
-			     RADEON_CG_BLOCK_HDP), true);
-
-	ci_update_current_ps(rdev, boot_ps);
-
 	return 0;
 }
 
@@ -4706,12 +4704,6 @@ void ci_dpm_disable(struct radeon_device *rdev)
 	struct ci_power_info *pi = ci_get_pi(rdev);
 	struct radeon_ps *boot_ps = rdev->pm.dpm.boot_ps;
 
-	cik_update_cg(rdev, (RADEON_CG_BLOCK_GFX |
-			     RADEON_CG_BLOCK_MC |
-			     RADEON_CG_BLOCK_SDMA |
-			     RADEON_CG_BLOCK_UVD |
-			     RADEON_CG_BLOCK_HDP), false);
-
 	ci_dpm_powergate_uvd(rdev, false);
 
 	if (!ci_is_smc_running(rdev))
@@ -4742,13 +4734,6 @@ int ci_dpm_set_power_state(struct radeon_device *rdev)
 	struct radeon_ps *old_ps = &pi->current_rps;
 	int ret;
 
-	cik_update_cg(rdev, (RADEON_CG_BLOCK_GFX |
-			     RADEON_CG_BLOCK_MC |
-			     RADEON_CG_BLOCK_SDMA |
-			     RADEON_CG_BLOCK_BIF |
-			     RADEON_CG_BLOCK_UVD |
-			     RADEON_CG_BLOCK_HDP), false);
-
 	ci_find_dpm_states_clocks_in_dpm_table(rdev, new_ps);
 	if (pi->pcie_performance_request)
 		ci_request_link_speed_change_before_state_change(rdev, new_ps, old_ps);
@@ -4804,13 +4789,6 @@ int ci_dpm_set_power_state(struct radeon_device *rdev)
 	if (pi->pcie_performance_request)
 		ci_notify_link_speed_change_after_state_change(rdev, new_ps, old_ps);
 
-	cik_update_cg(rdev, (RADEON_CG_BLOCK_GFX |
-			     RADEON_CG_BLOCK_MC |
-			     RADEON_CG_BLOCK_SDMA |
-			     RADEON_CG_BLOCK_BIF |
-			     RADEON_CG_BLOCK_UVD |
-			     RADEON_CG_BLOCK_HDP), true);
-
 	return 0;
 }
 
@@ -5023,8 +5001,8 @@ static int ci_parse_power_table(struct radeon_device *rdev)
 	return 0;
 }
 
-int ci_get_vbios_boot_values(struct radeon_device *rdev,
-			     struct ci_vbios_boot_state *boot_state)
+static int ci_get_vbios_boot_values(struct radeon_device *rdev,
+				    struct ci_vbios_boot_state *boot_state)
 {
 	struct radeon_mode_info *mode_info = &rdev->mode_info;
 	int index = GetIndexIntoMasterTable(DATA, FirmwareInfo);
diff --git a/drivers/gpu/drm/radeon/ci_smc.c b/drivers/gpu/drm/radeon/ci_smc.c
index 9c745dd22438..8debc9d47362 100644
--- a/drivers/gpu/drm/radeon/ci_smc.c
+++ b/drivers/gpu/drm/radeon/ci_smc.c
@@ -28,6 +28,7 @@
 #include "cikd.h"
 #include "ppsmc.h"
 #include "radeon_ucode.h"
+#include "ci_dpm.h"
 
 static int ci_set_smc_sram_address(struct radeon_device *rdev,
 				   u32 smc_address, u32 limit)
diff --git a/drivers/gpu/drm/radeon/cik.c b/drivers/gpu/drm/radeon/cik.c
index e950fabd7f5e..e6419ca7cd37 100644
--- a/drivers/gpu/drm/radeon/cik.c
+++ b/drivers/gpu/drm/radeon/cik.c
@@ -1697,7 +1697,7 @@ static void cik_srbm_select(struct radeon_device *rdev,
  * Load the GDDR MC ucode into the hw (CIK).
  * Returns 0 on success, error on failure.
  */
-static int ci_mc_load_microcode(struct radeon_device *rdev)
+int ci_mc_load_microcode(struct radeon_device *rdev)
 {
 	const __be32 *fw_data;
 	u32 running, blackout = 0;
@@ -3487,6 +3487,51 @@ int cik_ring_test(struct radeon_device *rdev, struct radeon_ring *ring)
 }
 
 /**
+ * cik_hdp_flush_cp_ring_emit - emit an hdp flush on the cp
+ *
+ * @rdev: radeon_device pointer
+ * @ridx: radeon ring index
+ *
+ * Emits an hdp flush on the cp.
+ */
+static void cik_hdp_flush_cp_ring_emit(struct radeon_device *rdev,
+				       int ridx)
+{
+	struct radeon_ring *ring = &rdev->ring[ridx];
+	u32 ref_and_mask;
+
+	switch (ring->idx) {
+	case CAYMAN_RING_TYPE_CP1_INDEX:
+	case CAYMAN_RING_TYPE_CP2_INDEX:
+	default:
+		switch (ring->me) {
+		case 0:
+			ref_and_mask = CP2 << ring->pipe;
+			break;
+		case 1:
+			ref_and_mask = CP6 << ring->pipe;
+			break;
+		default:
+			return;
+		}
+		break;
+	case RADEON_RING_TYPE_GFX_INDEX:
+		ref_and_mask = CP0;
+		break;
+	}
+
+	radeon_ring_write(ring, PACKET3(PACKET3_WAIT_REG_MEM, 5));
+	radeon_ring_write(ring, (WAIT_REG_MEM_OPERATION(1) | /* write, wait, write */
+				 WAIT_REG_MEM_FUNCTION(3) |  /* == */
+				 WAIT_REG_MEM_ENGINE(1)));   /* pfp */
+	radeon_ring_write(ring, GPU_HDP_FLUSH_REQ >> 2);
+	radeon_ring_write(ring, GPU_HDP_FLUSH_DONE >> 2);
+	radeon_ring_write(ring, ref_and_mask);
+	radeon_ring_write(ring, ref_and_mask);
+	radeon_ring_write(ring, 0x20); /* poll interval */
+}
+
+/**
  * cik_fence_gfx_ring_emit - emit a fence on the gfx ring
  *
  * @rdev: radeon_device pointer
@@ -3512,15 +3557,7 @@ void cik_fence_gfx_ring_emit(struct radeon_device *rdev,
 	radeon_ring_write(ring, fence->seq);
 	radeon_ring_write(ring, 0);
 	/* HDP flush */
-	/* We should be using the new WAIT_REG_MEM special op packet here
-	 * but it causes the CP to hang
-	 */
-	radeon_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3));
-	radeon_ring_write(ring, (WRITE_DATA_ENGINE_SEL(0) |
-				 WRITE_DATA_DST_SEL(0)));
-	radeon_ring_write(ring, HDP_MEM_COHERENCY_FLUSH_CNTL >> 2);
-	radeon_ring_write(ring, 0);
-	radeon_ring_write(ring, 0);
+	cik_hdp_flush_cp_ring_emit(rdev, fence->ring);
 }
 
 /**
@@ -3550,15 +3587,7 @@ void cik_fence_compute_ring_emit(struct radeon_device *rdev,
 	radeon_ring_write(ring, fence->seq);
 	radeon_ring_write(ring, 0);
 	/* HDP flush */
-	/* We should be using the new WAIT_REG_MEM special op packet here
-	 * but it causes the CP to hang
-	 */
-	radeon_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3));
-	radeon_ring_write(ring, (WRITE_DATA_ENGINE_SEL(0) |
-				 WRITE_DATA_DST_SEL(0)));
-	radeon_ring_write(ring, HDP_MEM_COHERENCY_FLUSH_CNTL >> 2);
-	radeon_ring_write(ring, 0);
-	radeon_ring_write(ring, 0);
+	cik_hdp_flush_cp_ring_emit(rdev, fence->ring);
 }
 
 bool cik_semaphore_ring_emit(struct radeon_device *rdev,
@@ -3566,8 +3595,6 @@ bool cik_semaphore_ring_emit(struct radeon_device *rdev,
 			     struct radeon_semaphore *semaphore,
 			     bool emit_wait)
 {
-/* TODO: figure out why semaphore cause lockups */
-#if 0
 	uint64_t addr = semaphore->gpu_addr;
 	unsigned sel = emit_wait ? PACKET3_SEM_SEL_WAIT : PACKET3_SEM_SEL_SIGNAL;
 
@@ -3576,9 +3603,6 @@ bool cik_semaphore_ring_emit(struct radeon_device *rdev,
 	radeon_ring_write(ring, (upper_32_bits(addr) & 0xffff) | sel);
 
 	return true;
-#else
-	return false;
-#endif
 }
 
 /**
@@ -3816,6 +3840,8 @@ static void cik_cp_gfx_enable(struct radeon_device *rdev, bool enable)
 	if (enable)
 		WREG32(CP_ME_CNTL, 0);
 	else {
+		if (rdev->asic->copy.copy_ring_index == RADEON_RING_TYPE_GFX_INDEX)
+			radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
 		WREG32(CP_ME_CNTL, (CP_ME_HALT | CP_PFP_HALT | CP_CE_HALT));
 		rdev->ring[RADEON_RING_TYPE_GFX_INDEX].ready = false;
 	}
@@ -4014,18 +4040,50 @@ static int cik_cp_gfx_resume(struct radeon_device *rdev)
 		rdev->ring[RADEON_RING_TYPE_GFX_INDEX].ready = false;
 		return r;
 	}
+
+	if (rdev->asic->copy.copy_ring_index == RADEON_RING_TYPE_GFX_INDEX)
+		radeon_ttm_set_active_vram_size(rdev, rdev->mc.real_vram_size);
+
 	return 0;
 }
 
-u32 cik_compute_ring_get_rptr(struct radeon_device *rdev,
-			      struct radeon_ring *ring)
+u32 cik_gfx_get_rptr(struct radeon_device *rdev,
+		     struct radeon_ring *ring)
 {
 	u32 rptr;
 
+	if (rdev->wb.enabled)
+		rptr = rdev->wb.wb[ring->rptr_offs/4];
+	else
+		rptr = RREG32(CP_RB0_RPTR);
+
+	return rptr;
+}
+
+u32 cik_gfx_get_wptr(struct radeon_device *rdev,
+		     struct radeon_ring *ring)
+{
+	u32 wptr;
+
+	wptr = RREG32(CP_RB0_WPTR);
+
+	return wptr;
+}
 
+void cik_gfx_set_wptr(struct radeon_device *rdev,
+		      struct radeon_ring *ring)
+{
+	WREG32(CP_RB0_WPTR, ring->wptr);
+	(void)RREG32(CP_RB0_WPTR);
+}
+
+u32 cik_compute_get_rptr(struct radeon_device *rdev,
+			 struct radeon_ring *ring)
+{
+	u32 rptr;
 
 	if (rdev->wb.enabled) {
-		rptr = le32_to_cpu(rdev->wb.wb[ring->rptr_offs/4]);
+		rptr = rdev->wb.wb[ring->rptr_offs/4];
 	} else {
 		mutex_lock(&rdev->srbm_mutex);
 		cik_srbm_select(rdev, ring->me, ring->pipe, ring->queue, 0);
@@ -4037,13 +4095,14 @@ u32 cik_compute_ring_get_rptr(struct radeon_device *rdev,
 	return rptr;
 }
 
-u32 cik_compute_ring_get_wptr(struct radeon_device *rdev,
-			      struct radeon_ring *ring)
+u32 cik_compute_get_wptr(struct radeon_device *rdev,
+			 struct radeon_ring *ring)
 {
 	u32 wptr;
 
 	if (rdev->wb.enabled) {
-		wptr = le32_to_cpu(rdev->wb.wb[ring->wptr_offs/4]);
+		/* XXX check if swapping is necessary on BE */
+		wptr = rdev->wb.wb[ring->wptr_offs/4];
 	} else {
 		mutex_lock(&rdev->srbm_mutex);
 		cik_srbm_select(rdev, ring->me, ring->pipe, ring->queue, 0);
@@ -4055,10 +4114,11 @@ u32 cik_compute_ring_get_wptr(struct radeon_device *rdev,
 	return wptr;
 }
 
-void cik_compute_ring_set_wptr(struct radeon_device *rdev,
-			       struct radeon_ring *ring)
+void cik_compute_set_wptr(struct radeon_device *rdev,
+			  struct radeon_ring *ring)
 {
-	rdev->wb.wb[ring->wptr_offs/4] = cpu_to_le32(ring->wptr);
+	/* XXX check if swapping is necessary on BE */
+	rdev->wb.wb[ring->wptr_offs/4] = ring->wptr;
 	WDOORBELL32(ring->doorbell_index, ring->wptr);
 }
 
@@ -4852,6 +4912,160 @@ static void cik_gpu_soft_reset(struct radeon_device *rdev, u32 reset_mask)
 	cik_print_gpu_status_regs(rdev);
 }
 
+struct kv_reset_save_regs {
+	u32 gmcon_reng_execute;
+	u32 gmcon_misc;
+	u32 gmcon_misc3;
+};
+
+static void kv_save_regs_for_reset(struct radeon_device *rdev,
+				   struct kv_reset_save_regs *save)
+{
+	save->gmcon_reng_execute = RREG32(GMCON_RENG_EXECUTE);
+	save->gmcon_misc = RREG32(GMCON_MISC);
+	save->gmcon_misc3 = RREG32(GMCON_MISC3);
+
+	WREG32(GMCON_RENG_EXECUTE, save->gmcon_reng_execute & ~RENG_EXECUTE_ON_PWR_UP);
+	WREG32(GMCON_MISC, save->gmcon_misc & ~(RENG_EXECUTE_ON_REG_UPDATE |
+						STCTRL_STUTTER_EN));
+}
+
+static void kv_restore_regs_for_reset(struct radeon_device *rdev,
+				      struct kv_reset_save_regs *save)
+{
+	int i;
+
+	WREG32(GMCON_PGFSM_WRITE, 0);
+	WREG32(GMCON_PGFSM_CONFIG, 0x200010ff);
+
+	for (i = 0; i < 5; i++)
+		WREG32(GMCON_PGFSM_WRITE, 0);
+
+	WREG32(GMCON_PGFSM_WRITE, 0);
+	WREG32(GMCON_PGFSM_CONFIG, 0x300010ff);
+
+	for (i = 0; i < 5; i++)
+		WREG32(GMCON_PGFSM_WRITE, 0);
+
+	WREG32(GMCON_PGFSM_WRITE, 0x210000);
+	WREG32(GMCON_PGFSM_CONFIG, 0xa00010ff);
+
+	for (i = 0; i < 5; i++)
+		WREG32(GMCON_PGFSM_WRITE, 0);
+
+	WREG32(GMCON_PGFSM_WRITE, 0x21003);
+	WREG32(GMCON_PGFSM_CONFIG, 0xb00010ff);
+
+	for (i = 0; i < 5; i++)
+		WREG32(GMCON_PGFSM_WRITE, 0);
+
+	WREG32(GMCON_PGFSM_WRITE, 0x2b00);
+	WREG32(GMCON_PGFSM_CONFIG, 0xc00010ff);
+
+	for (i = 0; i < 5; i++)
+		WREG32(GMCON_PGFSM_WRITE, 0);
+
+	WREG32(GMCON_PGFSM_WRITE, 0);
+	WREG32(GMCON_PGFSM_CONFIG, 0xd00010ff);
+
+	for (i = 0; i < 5; i++)
+		WREG32(GMCON_PGFSM_WRITE, 0);
+
+	WREG32(GMCON_PGFSM_WRITE, 0x420000);
+	WREG32(GMCON_PGFSM_CONFIG, 0x100010ff);
+
+	for (i = 0; i < 5; i++)
+		WREG32(GMCON_PGFSM_WRITE, 0);
+
+	WREG32(GMCON_PGFSM_WRITE, 0x120202);
+	WREG32(GMCON_PGFSM_CONFIG, 0x500010ff);
+
+	for (i = 0; i < 5; i++)
+		WREG32(GMCON_PGFSM_WRITE, 0);
+
+	WREG32(GMCON_PGFSM_WRITE, 0x3e3e36);
+	WREG32(GMCON_PGFSM_CONFIG, 0x600010ff);
+
+	for (i = 0; i < 5; i++)
+		WREG32(GMCON_PGFSM_WRITE, 0);
+
+	WREG32(GMCON_PGFSM_WRITE, 0x373f3e);
+	WREG32(GMCON_PGFSM_CONFIG, 0x700010ff);
+
+	for (i = 0; i < 5; i++)
+		WREG32(GMCON_PGFSM_WRITE, 0);
+
+	WREG32(GMCON_PGFSM_WRITE, 0x3e1332);
+	WREG32(GMCON_PGFSM_CONFIG, 0xe00010ff);
+
+	WREG32(GMCON_MISC3, save->gmcon_misc3);
+	WREG32(GMCON_MISC, save->gmcon_misc);
+	WREG32(GMCON_RENG_EXECUTE, save->gmcon_reng_execute);
+}
+
+static void cik_gpu_pci_config_reset(struct radeon_device *rdev)
+{
+	struct evergreen_mc_save save;
+	struct kv_reset_save_regs kv_save = { 0 };
+	u32 tmp, i;
+
+	dev_info(rdev->dev, "GPU pci config reset\n");
+
+	/* disable dpm? */
+
+	/* disable cg/pg */
+	cik_fini_pg(rdev);
+	cik_fini_cg(rdev);
+
+	/* Disable GFX parsing/prefetching */
+	WREG32(CP_ME_CNTL, CP_ME_HALT | CP_PFP_HALT | CP_CE_HALT);
+
+	/* Disable MEC parsing/prefetching */
+	WREG32(CP_MEC_CNTL, MEC_ME1_HALT | MEC_ME2_HALT);
+
+	/* sdma0 */
+	tmp = RREG32(SDMA0_ME_CNTL + SDMA0_REGISTER_OFFSET);
+	tmp |= SDMA_HALT;
+	WREG32(SDMA0_ME_CNTL + SDMA0_REGISTER_OFFSET, tmp);
+	/* sdma1 */
+	tmp = RREG32(SDMA0_ME_CNTL + SDMA1_REGISTER_OFFSET);
+	tmp |= SDMA_HALT;
+	WREG32(SDMA0_ME_CNTL + SDMA1_REGISTER_OFFSET, tmp);
+	/* XXX other engines? */
+
+	/* halt the rlc, disable cp internal ints */
+	cik_rlc_stop(rdev);
+
+	udelay(50);
+
+	/* disable mem access */
+	evergreen_mc_stop(rdev, &save);
+	if (evergreen_mc_wait_for_idle(rdev)) {
+		dev_warn(rdev->dev, "Wait for MC idle timed out !\n");
+	}
+
+	if (rdev->flags & RADEON_IS_IGP)
+		kv_save_regs_for_reset(rdev, &kv_save);
+
+	/* disable BM */
+	pci_clear_master(rdev->pdev);
+	/* reset */
+	radeon_pci_config_reset(rdev);
+
+	udelay(100);
+
+	/* wait for asic to come out of reset */
+	for (i = 0; i < rdev->usec_timeout; i++) {
+		if (RREG32(CONFIG_MEMSIZE) != 0xffffffff)
+			break;
+		udelay(1);
+	}
+
+	/* does asic init need to be run first??? */
+	if (rdev->flags & RADEON_IS_IGP)
+		kv_restore_regs_for_reset(rdev, &kv_save);
+}
+
 /**
  * cik_asic_reset - soft reset GPU
  *
@@ -4870,10 +5084,17 @@ int cik_asic_reset(struct radeon_device *rdev)
 	if (reset_mask)
 		r600_set_bios_scratch_engine_hung(rdev, true);
 
+	/* try soft reset */
 	cik_gpu_soft_reset(rdev, reset_mask);
 
 	reset_mask = cik_gpu_check_soft_reset(rdev);
 
+	/* try pci config reset */
+	if (reset_mask && radeon_hard_reset)
+		cik_gpu_pci_config_reset(rdev);
+
+	reset_mask = cik_gpu_check_soft_reset(rdev);
+
 	if (!reset_mask)
 		r600_set_bios_scratch_engine_hung(rdev, false);
 
@@ -5138,20 +5359,6 @@ static int cik_pcie_gart_enable(struct radeon_device *rdev)
 				WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT |
 				WRITE_PROTECTION_FAULT_ENABLE_DEFAULT);
 
-	/* TC cache setup ??? */
-	WREG32(TC_CFG_L1_LOAD_POLICY0, 0);
-	WREG32(TC_CFG_L1_LOAD_POLICY1, 0);
-	WREG32(TC_CFG_L1_STORE_POLICY, 0);
-
-	WREG32(TC_CFG_L2_LOAD_POLICY0, 0);
-	WREG32(TC_CFG_L2_LOAD_POLICY1, 0);
-	WREG32(TC_CFG_L2_STORE_POLICY0, 0);
-	WREG32(TC_CFG_L2_STORE_POLICY1, 0);
-	WREG32(TC_CFG_L2_ATOMIC_POLICY, 0);
-
-	WREG32(TC_CFG_L1_VOLATILE, 0);
-	WREG32(TC_CFG_L2_VOLATILE, 0);
-
 	if (rdev->family == CHIP_KAVERI) {
 		u32 tmp = RREG32(CHUB_CONTROL);
 		tmp &= ~BYPASS_VM;
@@ -5367,16 +5574,7 @@ void cik_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm)
 	radeon_ring_write(ring, VMID(0));
 
 	/* HDP flush */
-	/* We should be using the WAIT_REG_MEM packet here like in
-	 * cik_fence_ring_emit(), but it causes the CP to hang in this
-	 * context...
-	 */
-	radeon_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3));
-	radeon_ring_write(ring, (WRITE_DATA_ENGINE_SEL(0) |
-				 WRITE_DATA_DST_SEL(0)));
-	radeon_ring_write(ring, HDP_MEM_COHERENCY_FLUSH_CNTL >> 2);
-	radeon_ring_write(ring, 0);
-	radeon_ring_write(ring, 0);
+	cik_hdp_flush_cp_ring_emit(rdev, ridx);
 
 	/* bits 0-15 are the VM contexts0-15 */
 	radeon_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3));
@@ -7503,26 +7701,7 @@ static int cik_startup(struct radeon_device *rdev)
 
 	cik_mc_program(rdev);
 
-	if (rdev->flags & RADEON_IS_IGP) {
-		if (!rdev->me_fw || !rdev->pfp_fw || !rdev->ce_fw ||
-		    !rdev->mec_fw || !rdev->sdma_fw || !rdev->rlc_fw) {
-			r = cik_init_microcode(rdev);
-			if (r) {
-				DRM_ERROR("Failed to load firmware!\n");
-				return r;
-			}
-		}
-	} else {
-		if (!rdev->me_fw || !rdev->pfp_fw || !rdev->ce_fw ||
-		    !rdev->mec_fw || !rdev->sdma_fw || !rdev->rlc_fw ||
-		    !rdev->mc_fw) {
-			r = cik_init_microcode(rdev);
-			if (r) {
-				DRM_ERROR("Failed to load firmware!\n");
-				return r;
-			}
-		}
-
+	if (!(rdev->flags & RADEON_IS_IGP) && !rdev->pm.dpm_enabled) {
 		r = ci_mc_load_microcode(rdev);
 		if (r) {
 			DRM_ERROR("Failed to load MC firmware!\n");
@@ -7627,7 +7806,6 @@ static int cik_startup(struct radeon_device *rdev)
 
 	ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, RADEON_WB_CP_RPTR_OFFSET,
-			     CP_RB0_RPTR, CP_RB0_WPTR,
 			     PACKET3(PACKET3_NOP, 0x3FFF));
 	if (r)
 		return r;
@@ -7636,7 +7814,6 @@ static int cik_startup(struct radeon_device *rdev)
 	/* type-2 packets are deprecated on MEC, use type-3 instead */
 	ring = &rdev->ring[CAYMAN_RING_TYPE_CP1_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, RADEON_WB_CP1_RPTR_OFFSET,
-			     CP_HQD_PQ_RPTR, CP_HQD_PQ_WPTR,
 			     PACKET3(PACKET3_NOP, 0x3FFF));
 	if (r)
 		return r;
@@ -7648,7 +7825,6 @@ static int cik_startup(struct radeon_device *rdev)
 	/* type-2 packets are deprecated on MEC, use type-3 instead */
 	ring = &rdev->ring[CAYMAN_RING_TYPE_CP2_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, RADEON_WB_CP2_RPTR_OFFSET,
-			     CP_HQD_PQ_RPTR, CP_HQD_PQ_WPTR,
 			     PACKET3(PACKET3_NOP, 0x3FFF));
 	if (r)
 		return r;
@@ -7660,16 +7836,12 @@ static int cik_startup(struct radeon_device *rdev)
 
 	ring = &rdev->ring[R600_RING_TYPE_DMA_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, R600_WB_DMA_RPTR_OFFSET,
-			     SDMA0_GFX_RB_RPTR + SDMA0_REGISTER_OFFSET,
-			     SDMA0_GFX_RB_WPTR + SDMA0_REGISTER_OFFSET,
 			     SDMA_PACKET(SDMA_OPCODE_NOP, 0, 0));
 	if (r)
 		return r;
 
 	ring = &rdev->ring[CAYMAN_RING_TYPE_DMA1_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, CAYMAN_WB_DMA1_RPTR_OFFSET,
-			     SDMA0_GFX_RB_RPTR + SDMA1_REGISTER_OFFSET,
-			     SDMA0_GFX_RB_WPTR + SDMA1_REGISTER_OFFSET,
 			     SDMA_PACKET(SDMA_OPCODE_NOP, 0, 0));
 	if (r)
 		return r;
@@ -7685,7 +7857,6 @@ static int cik_startup(struct radeon_device *rdev)
 	ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
 	if (ring->ring_size) {
 		r = radeon_ring_init(rdev, ring, ring->ring_size, 0,
-				     UVD_RBC_RB_RPTR, UVD_RBC_RB_WPTR,
 				     RADEON_CP_PACKET2);
 		if (!r)
 			r = uvd_v1_0_init(rdev);
@@ -7731,6 +7902,8 @@ int cik_resume(struct radeon_device *rdev)
 	/* init golden registers */
 	cik_init_golden_registers(rdev);
 
+	radeon_pm_resume(rdev);
+
 	rdev->accel_working = true;
 	r = cik_startup(rdev);
 	if (r) {
@@ -7754,6 +7927,7 @@ int cik_resume(struct radeon_device *rdev)
  */
 int cik_suspend(struct radeon_device *rdev)
 {
+	radeon_pm_suspend(rdev);
 	dce6_audio_fini(rdev);
 	radeon_vm_manager_fini(rdev);
 	cik_cp_enable(rdev, false);
@@ -7835,6 +8009,30 @@ int cik_init(struct radeon_device *rdev)
 	if (r)
 		return r;
 
+	if (rdev->flags & RADEON_IS_IGP) {
+		if (!rdev->me_fw || !rdev->pfp_fw || !rdev->ce_fw ||
+		    !rdev->mec_fw || !rdev->sdma_fw || !rdev->rlc_fw) {
+			r = cik_init_microcode(rdev);
+			if (r) {
+				DRM_ERROR("Failed to load firmware!\n");
+				return r;
+			}
+		}
+	} else {
+		if (!rdev->me_fw || !rdev->pfp_fw || !rdev->ce_fw ||
+		    !rdev->mec_fw || !rdev->sdma_fw || !rdev->rlc_fw ||
+		    !rdev->mc_fw) {
+			r = cik_init_microcode(rdev);
+			if (r) {
+				DRM_ERROR("Failed to load firmware!\n");
+				return r;
+			}
+		}
+	}
+
+	/* Initialize power management */
+	radeon_pm_init(rdev);
+
 	ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX];
 	ring->ring_obj = NULL;
 	r600_ring_init(rdev, ring, 1024 * 1024);
@@ -7915,6 +8113,7 @@ int cik_init(struct radeon_device *rdev)
  */
 void cik_fini(struct radeon_device *rdev)
 {
+	radeon_pm_fini(rdev);
 	cik_cp_fini(rdev);
 	cik_sdma_fini(rdev);
 	cik_fini_pg(rdev);
diff --git a/drivers/gpu/drm/radeon/cik_sdma.c b/drivers/gpu/drm/radeon/cik_sdma.c
index d08b83c6267b..1ecb3f1070e3 100644
--- a/drivers/gpu/drm/radeon/cik_sdma.c
+++ b/drivers/gpu/drm/radeon/cik_sdma.c
@@ -52,6 +52,75 @@ u32 cik_gpu_check_soft_reset(struct radeon_device *rdev);
  */
 
 /**
+ * cik_sdma_get_rptr - get the current read pointer
+ *
+ * @rdev: radeon_device pointer
+ * @ring: radeon ring pointer
+ *
+ * Get the current rptr from the hardware (CIK+).
+ */
+uint32_t cik_sdma_get_rptr(struct radeon_device *rdev,
+			   struct radeon_ring *ring)
+{
+	u32 rptr, reg;
+
+	if (rdev->wb.enabled) {
+		rptr = rdev->wb.wb[ring->rptr_offs/4];
+	} else {
+		if (ring->idx == R600_RING_TYPE_DMA_INDEX)
+			reg = SDMA0_GFX_RB_RPTR + SDMA0_REGISTER_OFFSET;
+		else
+			reg = SDMA0_GFX_RB_RPTR + SDMA1_REGISTER_OFFSET;
+
+		rptr = RREG32(reg);
+	}
+
+	return (rptr & 0x3fffc) >> 2;
+}
+
+/**
+ * cik_sdma_get_wptr - get the current write pointer
+ *
+ * @rdev: radeon_device pointer
+ * @ring: radeon ring pointer
+ *
+ * Get the current wptr from the hardware (CIK+).
+ */
+uint32_t cik_sdma_get_wptr(struct radeon_device *rdev,
+			   struct radeon_ring *ring)
+{
+	u32 reg;
+
+	if (ring->idx == R600_RING_TYPE_DMA_INDEX)
+		reg = SDMA0_GFX_RB_WPTR + SDMA0_REGISTER_OFFSET;
+	else
+		reg = SDMA0_GFX_RB_WPTR + SDMA1_REGISTER_OFFSET;
+
+	return (RREG32(reg) & 0x3fffc) >> 2;
+}
+
+/**
+ * cik_sdma_set_wptr - commit the write pointer
+ *
+ * @rdev: radeon_device pointer
+ * @ring: radeon ring pointer
+ *
+ * Write the wptr back to the hardware (CIK+).
+ */
+void cik_sdma_set_wptr(struct radeon_device *rdev,
+		       struct radeon_ring *ring)
+{
+	u32 reg;
+
+	if (ring->idx == R600_RING_TYPE_DMA_INDEX)
+		reg = SDMA0_GFX_RB_WPTR + SDMA0_REGISTER_OFFSET;
+	else
+		reg = SDMA0_GFX_RB_WPTR + SDMA1_REGISTER_OFFSET;
+
+	WREG32(reg, (ring->wptr << 2) & 0x3fffc);
+}
+
+/**
  * cik_sdma_ring_ib_execute - Schedule an IB on the DMA engine
  *
  * @rdev: radeon_device pointer
@@ -88,6 +157,35 @@ void cik_sdma_ring_ib_execute(struct radeon_device *rdev,
 }
 
 /**
+ * cik_sdma_hdp_flush_ring_emit - emit an hdp flush on the DMA ring
+ *
+ * @rdev: radeon_device pointer
+ * @ridx: radeon ring index
+ *
+ * Emit an hdp flush packet on the requested DMA ring.
+ */
+static void cik_sdma_hdp_flush_ring_emit(struct radeon_device *rdev,
+					 int ridx)
+{
+	struct radeon_ring *ring = &rdev->ring[ridx];
+	u32 extra_bits = (SDMA_POLL_REG_MEM_EXTRA_OP(1) |
+			  SDMA_POLL_REG_MEM_EXTRA_FUNC(3)); /* == */
+	u32 ref_and_mask;
+
+	if (ridx == R600_RING_TYPE_DMA_INDEX)
+		ref_and_mask = SDMA0;
+	else
+		ref_and_mask = SDMA1;
+
+	radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_POLL_REG_MEM, 0, extra_bits));
+	radeon_ring_write(ring, GPU_HDP_FLUSH_DONE);
+	radeon_ring_write(ring, GPU_HDP_FLUSH_REQ);
+	radeon_ring_write(ring, ref_and_mask); /* reference */
+	radeon_ring_write(ring, ref_and_mask); /* mask */
+	radeon_ring_write(ring, (0xfff << 16) | 10); /* retry count, poll interval */
+}
+
+/**
  * cik_sdma_fence_ring_emit - emit a fence on the DMA ring
  *
  * @rdev: radeon_device pointer
@@ -111,12 +209,7 @@ void cik_sdma_fence_ring_emit(struct radeon_device *rdev,
 	/* generate an interrupt */
 	radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_TRAP, 0, 0));
 	/* flush HDP */
-	/* We should be using the new POLL_REG_MEM special op packet here
-	 * but it causes sDMA to hang sometimes
-	 */
-	radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_SRBM_WRITE, 0, 0xf000));
-	radeon_ring_write(ring, HDP_MEM_COHERENCY_FLUSH_CNTL >> 2);
-	radeon_ring_write(ring, 0);
+	cik_sdma_hdp_flush_ring_emit(rdev, fence->ring);
 }
 
 /**
@@ -157,7 +250,9 @@ static void cik_sdma_gfx_stop(struct radeon_device *rdev)
 	u32 rb_cntl, reg_offset;
 	int i;
 
-	radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
+	if ((rdev->asic->copy.copy_ring_index == R600_RING_TYPE_DMA_INDEX) ||
+	    (rdev->asic->copy.copy_ring_index == CAYMAN_RING_TYPE_DMA1_INDEX))
+		radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
 
 	for (i = 0; i < 2; i++) {
 		if (i == 0)
@@ -288,7 +383,9 @@ static int cik_sdma_gfx_resume(struct radeon_device *rdev)
 		}
 	}
 
-	radeon_ttm_set_active_vram_size(rdev, rdev->mc.real_vram_size);
+	if ((rdev->asic->copy.copy_ring_index == R600_RING_TYPE_DMA_INDEX) ||
+	    (rdev->asic->copy.copy_ring_index == CAYMAN_RING_TYPE_DMA1_INDEX))
+		radeon_ttm_set_active_vram_size(rdev, rdev->mc.real_vram_size);
 
 	return 0;
 }
@@ -747,12 +844,7 @@ void cik_dma_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm
 	radeon_ring_write(ring, VMID(0));
 
 	/* flush HDP */
-	/* We should be using the new POLL_REG_MEM special op packet here
-	 * but it causes sDMA to hang sometimes
-	 */
-	radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_SRBM_WRITE, 0, 0xf000));
-	radeon_ring_write(ring, HDP_MEM_COHERENCY_FLUSH_CNTL >> 2);
-	radeon_ring_write(ring, 0);
+	cik_sdma_hdp_flush_ring_emit(rdev, ridx);
 
 	/* flush TLB */
 	radeon_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_SRBM_WRITE, 0, 0xf000));
diff --git a/drivers/gpu/drm/radeon/cikd.h b/drivers/gpu/drm/radeon/cikd.h
index 5964af5e5b2d..98bae9d7b74d 100644
--- a/drivers/gpu/drm/radeon/cikd.h
+++ b/drivers/gpu/drm/radeon/cikd.h
@@ -724,6 +724,17 @@
 
 #define ATC_MISC_CG           				0x3350
 
+#define GMCON_RENG_EXECUTE				0x3508
+#define 	RENG_EXECUTE_ON_PWR_UP			(1 << 0)
+#define GMCON_MISC					0x350c
+#define 	RENG_EXECUTE_ON_REG_UPDATE		(1 << 11)
+#define 	STCTRL_STUTTER_EN			(1 << 16)
+
+#define GMCON_PGFSM_CONFIG				0x3538
+#define GMCON_PGFSM_WRITE				0x353c
+#define GMCON_PGFSM_READ				0x3540
+#define GMCON_MISC3					0x3544
+
 #define MC_SEQ_CNTL_3                                     0x3600
 #       define CAC_EN                                     (1 << 31)
 #define MC_SEQ_G5PDX_CTRL                                 0x3604
diff --git a/drivers/gpu/drm/radeon/cypress_dpm.c b/drivers/gpu/drm/radeon/cypress_dpm.c
index 920e1e4a52c5..cf783fc0ef21 100644
--- a/drivers/gpu/drm/radeon/cypress_dpm.c
+++ b/drivers/gpu/drm/radeon/cypress_dpm.c
@@ -1905,21 +1905,6 @@ int cypress_dpm_enable(struct radeon_device *rdev)
 	if (pi->mg_clock_gating)
 		cypress_mg_clock_gating_enable(rdev, true);
 
-	if (rdev->irq.installed &&
-	    r600_is_internal_thermal_sensor(rdev->pm.int_thermal_type)) {
-		PPSMC_Result result;
-
-		ret = rv770_set_thermal_temperature_range(rdev, R600_TEMP_RANGE_MIN, R600_TEMP_RANGE_MAX);
-		if (ret)
-			return ret;
-		rdev->irq.dpm_thermal = true;
-		radeon_irq_set(rdev);
-		result = rv770_send_msg_to_smc(rdev, PPSMC_MSG_EnableThermalInterrupt);
-
-		if (result != PPSMC_Result_OK)
-			DRM_DEBUG_KMS("Could not enable thermal interrupts.\n");
-	}
-
 	rv770_enable_auto_throttle_source(rdev, RADEON_DPM_AUTO_THROTTLE_SRC_THERMAL, true);
 
 	return 0;
diff --git a/drivers/gpu/drm/radeon/evergreen.c b/drivers/gpu/drm/radeon/evergreen.c
index 9702e55e924e..f2b9e21ce4da 100644
--- a/drivers/gpu/drm/radeon/evergreen.c
+++ b/drivers/gpu/drm/radeon/evergreen.c
@@ -146,6 +146,7 @@ extern u32 si_get_csb_size(struct radeon_device *rdev);
 extern void si_get_csb_buffer(struct radeon_device *rdev, volatile u32 *buffer);
 extern u32 cik_get_csb_size(struct radeon_device *rdev);
 extern void cik_get_csb_buffer(struct radeon_device *rdev, volatile u32 *buffer);
+extern void rv770_set_clk_bypass_mode(struct radeon_device *rdev);
 
 static const u32 evergreen_golden_registers[] =
 {
@@ -3867,6 +3868,48 @@ static void evergreen_gpu_soft_reset(struct radeon_device *rdev, u32 reset_mask)
 	evergreen_print_gpu_status_regs(rdev);
 }
 
+void evergreen_gpu_pci_config_reset(struct radeon_device *rdev)
+{
+	struct evergreen_mc_save save;
+	u32 tmp, i;
+
+	dev_info(rdev->dev, "GPU pci config reset\n");
+
+	/* disable dpm? */
+
+	/* Disable CP parsing/prefetching */
+	WREG32(CP_ME_CNTL, CP_ME_HALT | CP_PFP_HALT);
+	udelay(50);
+	/* Disable DMA */
+	tmp = RREG32(DMA_RB_CNTL);
+	tmp &= ~DMA_RB_ENABLE;
+	WREG32(DMA_RB_CNTL, tmp);
+	/* XXX other engines? */
+
+	/* halt the rlc */
+	r600_rlc_stop(rdev);
+
+	udelay(50);
+
+	/* set mclk/sclk to bypass */
+	rv770_set_clk_bypass_mode(rdev);
+	/* disable BM */
+	pci_clear_master(rdev->pdev);
+	/* disable mem access */
+	evergreen_mc_stop(rdev, &save);
+	if (evergreen_mc_wait_for_idle(rdev)) {
+		dev_warn(rdev->dev, "Wait for MC idle timed out !\n");
+	}
+	/* reset */
+	radeon_pci_config_reset(rdev);
+	/* wait for asic to come out of reset */
+	for (i = 0; i < rdev->usec_timeout; i++) {
+		if (RREG32(CONFIG_MEMSIZE) != 0xffffffff)
+			break;
+		udelay(1);
+	}
+}
+
 int evergreen_asic_reset(struct radeon_device *rdev)
 {
 	u32 reset_mask;
@@ -3876,10 +3919,17 @@ int evergreen_asic_reset(struct radeon_device *rdev)
 	if (reset_mask)
 		r600_set_bios_scratch_engine_hung(rdev, true);
 
+	/* try soft reset */
 	evergreen_gpu_soft_reset(rdev, reset_mask);
 
 	reset_mask = evergreen_gpu_check_soft_reset(rdev);
 
+	/* try pci config reset */
+	if (reset_mask && radeon_hard_reset)
+		evergreen_gpu_pci_config_reset(rdev);
+
+	reset_mask = evergreen_gpu_check_soft_reset(rdev);
+
 	if (!reset_mask)
 		r600_set_bios_scratch_engine_hung(rdev, false);
 
@@ -4298,8 +4348,8 @@ void evergreen_disable_interrupt_state(struct radeon_device *rdev)
 		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC5_REGISTER_OFFSET, 0);
 	}
 
-	/* only one DAC on DCE6 */
-	if (!ASIC_IS_DCE6(rdev))
+	/* only one DAC on DCE5 */
+	if (!ASIC_IS_DCE5(rdev))
 		WREG32(DACA_AUTODETECT_INT_CONTROL, 0);
 	WREG32(DACB_AUTODETECT_INT_CONTROL, 0);
 
@@ -5109,27 +5159,12 @@ static int evergreen_startup(struct radeon_device *rdev)
 
 	evergreen_mc_program(rdev);
 
-	if (ASIC_IS_DCE5(rdev)) {
-		if (!rdev->me_fw || !rdev->pfp_fw || !rdev->rlc_fw || !rdev->mc_fw) {
-			r = ni_init_microcode(rdev);
-			if (r) {
-				DRM_ERROR("Failed to load firmware!\n");
-				return r;
-			}
-		}
+	if (ASIC_IS_DCE5(rdev) && !rdev->pm.dpm_enabled) {
 		r = ni_mc_load_microcode(rdev);
 		if (r) {
 			DRM_ERROR("Failed to load MC firmware!\n");
 			return r;
 		}
-	} else {
-		if (!rdev->me_fw || !rdev->pfp_fw || !rdev->rlc_fw) {
-			r = r600_init_microcode(rdev);
-			if (r) {
-				DRM_ERROR("Failed to load firmware!\n");
-				return r;
-			}
-		}
 	}
 
 	if (rdev->flags & RADEON_IS_AGP) {
@@ -5199,14 +5234,12 @@ static int evergreen_startup(struct radeon_device *rdev)
 
 	ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, RADEON_WB_CP_RPTR_OFFSET,
-			     R600_CP_RB_RPTR, R600_CP_RB_WPTR,
 			     RADEON_CP_PACKET2);
 	if (r)
 		return r;
 
 	ring = &rdev->ring[R600_RING_TYPE_DMA_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, R600_WB_DMA_RPTR_OFFSET,
-			     DMA_RB_RPTR, DMA_RB_WPTR,
 			     DMA_PACKET(DMA_PACKET_NOP, 0, 0));
 	if (r)
 		return r;
@@ -5224,7 +5257,6 @@ static int evergreen_startup(struct radeon_device *rdev)
 	ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
 	if (ring->ring_size) {
 		r = radeon_ring_init(rdev, ring, ring->ring_size, 0,
-				     UVD_RBC_RB_RPTR, UVD_RBC_RB_WPTR,
 				     RADEON_CP_PACKET2);
 		if (!r)
 			r = uvd_v1_0_init(rdev);
@@ -5267,6 +5299,8 @@ int evergreen_resume(struct radeon_device *rdev)
 	/* init golden registers */
 	evergreen_init_golden_registers(rdev);
 
+	radeon_pm_resume(rdev);
+
 	rdev->accel_working = true;
 	r = evergreen_startup(rdev);
 	if (r) {
@@ -5281,6 +5315,7 @@ int evergreen_resume(struct radeon_device *rdev)
 
 int evergreen_suspend(struct radeon_device *rdev)
 {
+	radeon_pm_suspend(rdev);
 	r600_audio_fini(rdev);
 	uvd_v1_0_fini(rdev);
 	radeon_uvd_suspend(rdev);
@@ -5357,6 +5392,27 @@ int evergreen_init(struct radeon_device *rdev)
 	if (r)
 		return r;
 
+	if (ASIC_IS_DCE5(rdev)) {
+		if (!rdev->me_fw || !rdev->pfp_fw || !rdev->rlc_fw || !rdev->mc_fw) {
+			r = ni_init_microcode(rdev);
+			if (r) {
+				DRM_ERROR("Failed to load firmware!\n");
+				return r;
+			}
+		}
+	} else {
+		if (!rdev->me_fw || !rdev->pfp_fw || !rdev->rlc_fw) {
+			r = r600_init_microcode(rdev);
+			if (r) {
+				DRM_ERROR("Failed to load firmware!\n");
+				return r;
+			}
+		}
+	}
+
+	/* Initialize power management */
+	radeon_pm_init(rdev);
+
 	rdev->ring[RADEON_RING_TYPE_GFX_INDEX].ring_obj = NULL;
 	r600_ring_init(rdev, &rdev->ring[RADEON_RING_TYPE_GFX_INDEX], 1024 * 1024);
 
@@ -5409,6 +5465,7 @@ int evergreen_init(struct radeon_device *rdev)
 
 void evergreen_fini(struct radeon_device *rdev)
 {
+	radeon_pm_fini(rdev);
 	r600_audio_fini(rdev);
 	r700_cp_fini(rdev);
 	r600_dma_fini(rdev);
diff --git a/drivers/gpu/drm/radeon/evergreen_cs.c b/drivers/gpu/drm/radeon/evergreen_cs.c
index eb8ac315f92f..c7cac07f139b 100644
--- a/drivers/gpu/drm/radeon/evergreen_cs.c
+++ b/drivers/gpu/drm/radeon/evergreen_cs.c
@@ -967,7 +967,10 @@ static int evergreen_cs_track_check(struct radeon_cs_parser *p)
 	if (track->cb_dirty) {
 		tmp = track->cb_target_mask;
 		for (i = 0; i < 8; i++) {
-			if ((tmp >> (i * 4)) & 0xF) {
+			u32 format = G_028C70_FORMAT(track->cb_color_info[i]);
+
+			if (format != V_028C70_COLOR_INVALID &&
+			    (tmp >> (i * 4)) & 0xF) {
 				/* at least one component is enabled */
 				if (track->cb_color_bo[i] == NULL) {
 					dev_warn(p->dev, "%s:%d mask 0x%08X | 0x%08X no cb for %d\n",
diff --git a/drivers/gpu/drm/radeon/evergreen_reg.h b/drivers/gpu/drm/radeon/evergreen_reg.h
index 8a4e641f0e3c..a0f63ff5a5e9 100644
--- a/drivers/gpu/drm/radeon/evergreen_reg.h
+++ b/drivers/gpu/drm/radeon/evergreen_reg.h
@@ -33,6 +33,7 @@
 #define EVERGREEN_PIF_PHY0_DATA                         0xc
 #define EVERGREEN_PIF_PHY1_INDEX                        0x10
 #define EVERGREEN_PIF_PHY1_DATA                         0x14
+#define EVERGREEN_MM_INDEX_HI                           0x18
 
 #define EVERGREEN_VGA_MEMORY_BASE_ADDRESS               0x310
 #define EVERGREEN_VGA_MEMORY_BASE_ADDRESS_HIGH          0x324
diff --git a/drivers/gpu/drm/radeon/evergreend.h b/drivers/gpu/drm/radeon/evergreend.h
index 17f990798992..f9c7963b3ee6 100644
--- a/drivers/gpu/drm/radeon/evergreend.h
+++ b/drivers/gpu/drm/radeon/evergreend.h
@@ -82,12 +82,16 @@
 #define	CG_SPLL_FUNC_CNTL_2				0x604
 #define		SCLK_MUX_SEL(x)				((x) << 0)
 #define		SCLK_MUX_SEL_MASK			(0x1ff << 0)
+#define		SCLK_MUX_UPDATE				(1 << 26)
 #define	CG_SPLL_FUNC_CNTL_3				0x608
 #define		SPLL_FB_DIV(x)				((x) << 0)
 #define		SPLL_FB_DIV_MASK			(0x3ffffff << 0)
 #define		SPLL_DITHEN				(1 << 28)
+#define	CG_SPLL_STATUS					0x60c
+#define		SPLL_CHG_STATUS				(1 << 1)
 
 #define MPLL_CNTL_MODE                                  0x61c
+#       define MPLL_MCLK_SEL                            (1 << 11)
 #       define SS_SSEN                                  (1 << 24)
 #       define SS_DSMODE_EN                             (1 << 25)
 
diff --git a/drivers/gpu/drm/radeon/kv_dpm.c b/drivers/gpu/drm/radeon/kv_dpm.c
index b41905573cd2..b6e01d5d2cce 100644
--- a/drivers/gpu/drm/radeon/kv_dpm.c
+++ b/drivers/gpu/drm/radeon/kv_dpm.c
@@ -1126,11 +1126,6 @@ int kv_dpm_enable(struct radeon_device *rdev)
 	struct kv_power_info *pi = kv_get_pi(rdev);
 	int ret;
 
-	cik_update_cg(rdev, (RADEON_CG_BLOCK_GFX |
-			     RADEON_CG_BLOCK_SDMA |
-			     RADEON_CG_BLOCK_BIF |
-			     RADEON_CG_BLOCK_HDP), false);
-
 	ret = kv_process_firmware_header(rdev);
 	if (ret) {
 		DRM_ERROR("kv_process_firmware_header failed\n");
@@ -1215,6 +1210,21 @@ int kv_dpm_enable(struct radeon_device *rdev)
 
 	kv_reset_acp_boot_level(rdev);
 
+	ret = kv_smc_bapm_enable(rdev, false);
+	if (ret) {
+		DRM_ERROR("kv_smc_bapm_enable failed\n");
+		return ret;
+	}
+
+	kv_update_current_ps(rdev, rdev->pm.dpm.boot_ps);
+
+	return ret;
+}
+
+int kv_dpm_late_enable(struct radeon_device *rdev)
+{
+	int ret;
+
 	if (rdev->irq.installed &&
 	    r600_is_internal_thermal_sensor(rdev->pm.int_thermal_type)) {
 		ret = kv_set_thermal_temperature_range(rdev, R600_TEMP_RANGE_MIN, R600_TEMP_RANGE_MAX);
@@ -1226,35 +1236,17 @@ int kv_dpm_enable(struct radeon_device *rdev)
 		radeon_irq_set(rdev);
 	}
 
-	ret = kv_smc_bapm_enable(rdev, false);
-	if (ret) {
-		DRM_ERROR("kv_smc_bapm_enable failed\n");
-		return ret;
-	}
-
 	/* powerdown unused blocks for now */
 	kv_dpm_powergate_acp(rdev, true);
 	kv_dpm_powergate_samu(rdev, true);
 	kv_dpm_powergate_vce(rdev, true);
 	kv_dpm_powergate_uvd(rdev, true);
 
-	cik_update_cg(rdev, (RADEON_CG_BLOCK_GFX |
-			     RADEON_CG_BLOCK_SDMA |
-			     RADEON_CG_BLOCK_BIF |
-			     RADEON_CG_BLOCK_HDP), true);
-
-	kv_update_current_ps(rdev, rdev->pm.dpm.boot_ps);
-
 	return ret;
 }
 
 void kv_dpm_disable(struct radeon_device *rdev)
 {
-	cik_update_cg(rdev, (RADEON_CG_BLOCK_GFX |
-			     RADEON_CG_BLOCK_SDMA |
-			     RADEON_CG_BLOCK_BIF |
-			     RADEON_CG_BLOCK_HDP), false);
-
 	kv_smc_bapm_enable(rdev, false);
 
 	/* powerup blocks */
@@ -1779,11 +1771,6 @@ int kv_dpm_set_power_state(struct radeon_device *rdev)
 	/*struct radeon_ps *old_ps = &pi->current_rps;*/
 	int ret;
 
-	cik_update_cg(rdev, (RADEON_CG_BLOCK_GFX |
-			     RADEON_CG_BLOCK_SDMA |
-			     RADEON_CG_BLOCK_BIF |
-			     RADEON_CG_BLOCK_HDP), false);
-
 	if (pi->bapm_enable) {
 		ret = kv_smc_bapm_enable(rdev, rdev->pm.dpm.ac_power);
 		if (ret) {
@@ -1849,11 +1836,6 @@ int kv_dpm_set_power_state(struct radeon_device *rdev)
 		}
 	}
 
-	cik_update_cg(rdev, (RADEON_CG_BLOCK_GFX |
-			     RADEON_CG_BLOCK_SDMA |
-			     RADEON_CG_BLOCK_BIF |
-			     RADEON_CG_BLOCK_HDP), true);
-
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/radeon/ni.c b/drivers/gpu/drm/radeon/ni.c
index f59a9e9fccf8..ea932ac66fc6 100644
--- a/drivers/gpu/drm/radeon/ni.c
+++ b/drivers/gpu/drm/radeon/ni.c
@@ -174,6 +174,7 @@ extern void evergreen_pcie_gen2_enable(struct radeon_device *rdev);
 extern void evergreen_program_aspm(struct radeon_device *rdev);
 extern void sumo_rlc_fini(struct radeon_device *rdev);
 extern int sumo_rlc_init(struct radeon_device *rdev);
+extern void evergreen_gpu_pci_config_reset(struct radeon_device *rdev);
 
 /* Firmware Names */
 MODULE_FIRMWARE("radeon/BARTS_pfp.bin");
@@ -1330,13 +1331,12 @@ void cayman_fence_ring_emit(struct radeon_device *rdev,
 {
 	struct radeon_ring *ring = &rdev->ring[fence->ring];
 	u64 addr = rdev->fence_drv[fence->ring].gpu_addr;
+	u32 cp_coher_cntl = PACKET3_FULL_CACHE_ENA | PACKET3_TC_ACTION_ENA |
+		PACKET3_SH_ACTION_ENA;
 
 	/* flush read cache over gart for this vmid */
-	radeon_ring_write(ring, PACKET3(PACKET3_SET_CONFIG_REG, 1));
-	radeon_ring_write(ring, (CP_COHER_CNTL2 - PACKET3_SET_CONFIG_REG_START) >> 2);
-	radeon_ring_write(ring, 0);
 	radeon_ring_write(ring, PACKET3(PACKET3_SURFACE_SYNC, 3));
-	radeon_ring_write(ring, PACKET3_TC_ACTION_ENA | PACKET3_SH_ACTION_ENA);
+	radeon_ring_write(ring, PACKET3_ENGINE_ME | cp_coher_cntl);
 	radeon_ring_write(ring, 0xFFFFFFFF);
 	radeon_ring_write(ring, 0);
 	radeon_ring_write(ring, 10); /* poll interval */
@@ -1352,6 +1352,8 @@ void cayman_fence_ring_emit(struct radeon_device *rdev,
 void cayman_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib)
 {
 	struct radeon_ring *ring = &rdev->ring[ib->ring];
+	u32 cp_coher_cntl = PACKET3_FULL_CACHE_ENA | PACKET3_TC_ACTION_ENA |
+		PACKET3_SH_ACTION_ENA;
 
 	/* set to DX10/11 mode */
 	radeon_ring_write(ring, PACKET3(PACKET3_MODE_CONTROL, 0));
@@ -1376,14 +1378,11 @@ void cayman_ring_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib)
 			  (ib->vm ? (ib->vm->id << 24) : 0));
 
 	/* flush read cache over gart for this vmid */
-	radeon_ring_write(ring, PACKET3(PACKET3_SET_CONFIG_REG, 1));
-	radeon_ring_write(ring, (CP_COHER_CNTL2 - PACKET3_SET_CONFIG_REG_START) >> 2);
-	radeon_ring_write(ring, ib->vm ? ib->vm->id : 0);
 	radeon_ring_write(ring, PACKET3(PACKET3_SURFACE_SYNC, 3));
-	radeon_ring_write(ring, PACKET3_TC_ACTION_ENA | PACKET3_SH_ACTION_ENA);
+	radeon_ring_write(ring, PACKET3_ENGINE_ME | cp_coher_cntl);
 	radeon_ring_write(ring, 0xFFFFFFFF);
 	radeon_ring_write(ring, 0);
-	radeon_ring_write(ring, 10); /* poll interval */
+	radeon_ring_write(ring, ((ib->vm ? ib->vm->id : 0) << 24) | 10); /* poll interval */
 }
 
 static void cayman_cp_enable(struct radeon_device *rdev, bool enable)
@@ -1391,13 +1390,63 @@ static void cayman_cp_enable(struct radeon_device *rdev, bool enable)
 	if (enable)
 		WREG32(CP_ME_CNTL, 0);
 	else {
-		radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
+		if (rdev->asic->copy.copy_ring_index == RADEON_RING_TYPE_GFX_INDEX)
+			radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
 		WREG32(CP_ME_CNTL, (CP_ME_HALT | CP_PFP_HALT));
 		WREG32(SCRATCH_UMSK, 0);
 		rdev->ring[RADEON_RING_TYPE_GFX_INDEX].ready = false;
 	}
 }
 
+u32 cayman_gfx_get_rptr(struct radeon_device *rdev,
+			struct radeon_ring *ring)
+{
+	u32 rptr;
+
+	if (rdev->wb.enabled)
+		rptr = rdev->wb.wb[ring->rptr_offs/4];
+	else {
+		if (ring->idx == RADEON_RING_TYPE_GFX_INDEX)
+			rptr = RREG32(CP_RB0_RPTR);
+		else if (ring->idx == CAYMAN_RING_TYPE_CP1_INDEX)
+			rptr = RREG32(CP_RB1_RPTR);
+		else
+			rptr = RREG32(CP_RB2_RPTR);
+	}
+
+	return rptr;
+}
+
+u32 cayman_gfx_get_wptr(struct radeon_device *rdev,
+			struct radeon_ring *ring)
+{
+	u32 wptr;
+
+	if (ring->idx == RADEON_RING_TYPE_GFX_INDEX)
+		wptr = RREG32(CP_RB0_WPTR);
+	else if (ring->idx == CAYMAN_RING_TYPE_CP1_INDEX)
+		wptr = RREG32(CP_RB1_WPTR);
+	else
+		wptr = RREG32(CP_RB2_WPTR);
+
+	return wptr;
+}
+
+void cayman_gfx_set_wptr(struct radeon_device *rdev,
+			 struct radeon_ring *ring)
+{
+	if (ring->idx == RADEON_RING_TYPE_GFX_INDEX) {
+		WREG32(CP_RB0_WPTR, ring->wptr);
+		(void)RREG32(CP_RB0_WPTR);
+	} else if (ring->idx == CAYMAN_RING_TYPE_CP1_INDEX) {
+		WREG32(CP_RB1_WPTR, ring->wptr);
+		(void)RREG32(CP_RB1_WPTR);
+	} else {
+		WREG32(CP_RB2_WPTR, ring->wptr);
+		(void)RREG32(CP_RB2_WPTR);
+	}
+}
+
 static int cayman_cp_load_microcode(struct radeon_device *rdev)
 {
 	const __be32 *fw_data;
@@ -1526,6 +1575,16 @@ static int cayman_cp_resume(struct radeon_device *rdev)
 		CP_RB1_BASE,
 		CP_RB2_BASE
 	};
+	static const unsigned cp_rb_rptr[] = {
+		CP_RB0_RPTR,
+		CP_RB1_RPTR,
+		CP_RB2_RPTR
+	};
+	static const unsigned cp_rb_wptr[] = {
+		CP_RB0_WPTR,
+		CP_RB1_WPTR,
+		CP_RB2_WPTR
+	};
 	struct radeon_ring *ring;
 	int i, r;
 
@@ -1584,8 +1643,8 @@ static int cayman_cp_resume(struct radeon_device *rdev)
 		WREG32_P(cp_rb_cntl[i], RB_RPTR_WR_ENA, ~RB_RPTR_WR_ENA);
 
 		ring->rptr = ring->wptr = 0;
-		WREG32(ring->rptr_reg, ring->rptr);
-		WREG32(ring->wptr_reg, ring->wptr);
+		WREG32(cp_rb_rptr[i], ring->rptr);
+		WREG32(cp_rb_wptr[i], ring->wptr);
 
 		mdelay(1);
 		WREG32_P(cp_rb_cntl[i], 0, ~RB_RPTR_WR_ENA);
@@ -1605,6 +1664,9 @@ static int cayman_cp_resume(struct radeon_device *rdev)
 		return r;
 	}
 
+	if (rdev->asic->copy.copy_ring_index == RADEON_RING_TYPE_GFX_INDEX)
+		radeon_ttm_set_active_vram_size(rdev, rdev->mc.real_vram_size);
+
 	return 0;
 }
 
@@ -1831,8 +1893,10 @@ int cayman_asic_reset(struct radeon_device *rdev)
 
 	reset_mask = cayman_gpu_check_soft_reset(rdev);
 
-	if (!reset_mask)
-		r600_set_bios_scratch_engine_hung(rdev, false);
+	if (reset_mask)
+		evergreen_gpu_pci_config_reset(rdev);
+
+	r600_set_bios_scratch_engine_hung(rdev, false);
 
 	return 0;
 }
@@ -1878,23 +1942,7 @@ static int cayman_startup(struct radeon_device *rdev)
 
 	evergreen_mc_program(rdev);
 
-	if (rdev->flags & RADEON_IS_IGP) {
-		if (!rdev->me_fw || !rdev->pfp_fw || !rdev->rlc_fw) {
-			r = ni_init_microcode(rdev);
-			if (r) {
-				DRM_ERROR("Failed to load firmware!\n");
-				return r;
-			}
-		}
-	} else {
-		if (!rdev->me_fw || !rdev->pfp_fw || !rdev->rlc_fw || !rdev->mc_fw) {
-			r = ni_init_microcode(rdev);
-			if (r) {
-				DRM_ERROR("Failed to load firmware!\n");
-				return r;
-			}
-		}
-
+	if (!(rdev->flags & RADEON_IS_IGP) && !rdev->pm.dpm_enabled) {
 		r = ni_mc_load_microcode(rdev);
 		if (r) {
 			DRM_ERROR("Failed to load MC firmware!\n");
@@ -1981,23 +2029,18 @@ static int cayman_startup(struct radeon_device *rdev)
 	evergreen_irq_set(rdev);
 
 	r = radeon_ring_init(rdev, ring, ring->ring_size, RADEON_WB_CP_RPTR_OFFSET,
-			     CP_RB0_RPTR, CP_RB0_WPTR,
 			     RADEON_CP_PACKET2);
 	if (r)
 		return r;
 
 	ring = &rdev->ring[R600_RING_TYPE_DMA_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, R600_WB_DMA_RPTR_OFFSET,
-			     DMA_RB_RPTR + DMA0_REGISTER_OFFSET,
-			     DMA_RB_WPTR + DMA0_REGISTER_OFFSET,
 			     DMA_PACKET(DMA_PACKET_NOP, 0, 0, 0));
 	if (r)
 		return r;
 
 	ring = &rdev->ring[CAYMAN_RING_TYPE_DMA1_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, CAYMAN_WB_DMA1_RPTR_OFFSET,
-			     DMA_RB_RPTR + DMA1_REGISTER_OFFSET,
-			     DMA_RB_WPTR + DMA1_REGISTER_OFFSET,
 			     DMA_PACKET(DMA_PACKET_NOP, 0, 0, 0));
 	if (r)
 		return r;
@@ -2016,7 +2059,6 @@ static int cayman_startup(struct radeon_device *rdev)
 	ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
 	if (ring->ring_size) {
 		r = radeon_ring_init(rdev, ring, ring->ring_size, 0,
-				     UVD_RBC_RB_RPTR, UVD_RBC_RB_WPTR,
 				     RADEON_CP_PACKET2);
 		if (!r)
 			r = uvd_v1_0_init(rdev);
@@ -2063,6 +2105,8 @@ int cayman_resume(struct radeon_device *rdev)
 	/* init golden registers */
 	ni_init_golden_registers(rdev);
 
+	radeon_pm_resume(rdev);
+
 	rdev->accel_working = true;
 	r = cayman_startup(rdev);
 	if (r) {
@@ -2075,6 +2119,7 @@ int cayman_resume(struct radeon_device *rdev)
 
 int cayman_suspend(struct radeon_device *rdev)
 {
+	radeon_pm_suspend(rdev);
 	if (ASIC_IS_DCE6(rdev))
 		dce6_audio_fini(rdev);
 	else
@@ -2145,6 +2190,27 @@ int cayman_init(struct radeon_device *rdev)
 	if (r)
 		return r;
 
+	if (rdev->flags & RADEON_IS_IGP) {
+		if (!rdev->me_fw || !rdev->pfp_fw || !rdev->rlc_fw) {
+			r = ni_init_microcode(rdev);
+			if (r) {
+				DRM_ERROR("Failed to load firmware!\n");
+				return r;
+			}
+		}
+	} else {
+		if (!rdev->me_fw || !rdev->pfp_fw || !rdev->rlc_fw || !rdev->mc_fw) {
+			r = ni_init_microcode(rdev);
+			if (r) {
+				DRM_ERROR("Failed to load firmware!\n");
+				return r;
+			}
+		}
+	}
+
+	/* Initialize power management */
+	radeon_pm_init(rdev);
+
 	ring->ring_obj = NULL;
 	r600_ring_init(rdev, ring, 1024 * 1024);
 
@@ -2204,6 +2270,7 @@ int cayman_init(struct radeon_device *rdev)
 
 void cayman_fini(struct radeon_device *rdev)
 {
+	radeon_pm_fini(rdev);
 	cayman_cp_fini(rdev);
 	cayman_dma_fini(rdev);
 	r600_irq_fini(rdev);
diff --git a/drivers/gpu/drm/radeon/ni_dma.c b/drivers/gpu/drm/radeon/ni_dma.c
index bdeb65ed3658..7cf96b15377f 100644
--- a/drivers/gpu/drm/radeon/ni_dma.c
+++ b/drivers/gpu/drm/radeon/ni_dma.c
@@ -43,6 +43,75 @@ u32 cayman_gpu_check_soft_reset(struct radeon_device *rdev);
  */
 
 /**
+ * cayman_dma_get_rptr - get the current read pointer
+ *
+ * @rdev: radeon_device pointer
+ * @ring: radeon ring pointer
+ *
+ * Get the current rptr from the hardware (cayman+).
+ */
+uint32_t cayman_dma_get_rptr(struct radeon_device *rdev,
+			     struct radeon_ring *ring)
+{
+	u32 rptr, reg;
+
+	if (rdev->wb.enabled) {
+		rptr = rdev->wb.wb[ring->rptr_offs/4];
+	} else {
+		if (ring->idx == R600_RING_TYPE_DMA_INDEX)
+			reg = DMA_RB_RPTR + DMA0_REGISTER_OFFSET;
+		else
+			reg = DMA_RB_RPTR + DMA1_REGISTER_OFFSET;
+
+		rptr = RREG32(reg);
+	}
+
+	return (rptr & 0x3fffc) >> 2;
+}
+
+/**
+ * cayman_dma_get_wptr - get the current write pointer
+ *
+ * @rdev: radeon_device pointer
+ * @ring: radeon ring pointer
+ *
+ * Get the current wptr from the hardware (cayman+).
+ */
+uint32_t cayman_dma_get_wptr(struct radeon_device *rdev,
+			   struct radeon_ring *ring)
+{
+	u32 reg;
+
+	if (ring->idx == R600_RING_TYPE_DMA_INDEX)
+		reg = DMA_RB_WPTR + DMA0_REGISTER_OFFSET;
+	else
+		reg = DMA_RB_WPTR + DMA1_REGISTER_OFFSET;
+
+	return (RREG32(reg) & 0x3fffc) >> 2;
+}
+
+/**
+ * cayman_dma_set_wptr - commit the write pointer
+ *
+ * @rdev: radeon_device pointer
+ * @ring: radeon ring pointer
+ *
+ * Write the wptr back to the hardware (cayman+).
+ */
+void cayman_dma_set_wptr(struct radeon_device *rdev,
+			 struct radeon_ring *ring)
+{
+	u32 reg;
+
+	if (ring->idx == R600_RING_TYPE_DMA_INDEX)
+		reg = DMA_RB_WPTR + DMA0_REGISTER_OFFSET;
+	else
+		reg = DMA_RB_WPTR + DMA1_REGISTER_OFFSET;
+
+	WREG32(reg, (ring->wptr << 2) & 0x3fffc);
+}
+
+/**
  * cayman_dma_ring_ib_execute - Schedule an IB on the DMA engine
  *
  * @rdev: radeon_device pointer
@@ -88,7 +157,9 @@ void cayman_dma_stop(struct radeon_device *rdev)
 {
 	u32 rb_cntl;
 
-	radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
+	if ((rdev->asic->copy.copy_ring_index == R600_RING_TYPE_DMA_INDEX) ||
+	    (rdev->asic->copy.copy_ring_index == CAYMAN_RING_TYPE_DMA1_INDEX))
+		radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
 
 	/* dma0 */
 	rb_cntl = RREG32(DMA_RB_CNTL + DMA0_REGISTER_OFFSET);
@@ -190,7 +261,9 @@ int cayman_dma_resume(struct radeon_device *rdev)
 		}
 	}
 
-	radeon_ttm_set_active_vram_size(rdev, rdev->mc.real_vram_size);
+	if ((rdev->asic->copy.copy_ring_index == R600_RING_TYPE_DMA_INDEX) ||
+	    (rdev->asic->copy.copy_ring_index == CAYMAN_RING_TYPE_DMA1_INDEX))
+		radeon_ttm_set_active_vram_size(rdev, rdev->mc.real_vram_size);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/radeon/ni_dpm.c b/drivers/gpu/drm/radeon/ni_dpm.c
index 49c4d48f54d6..c351226ecb31 100644
--- a/drivers/gpu/drm/radeon/ni_dpm.c
+++ b/drivers/gpu/drm/radeon/ni_dpm.c
@@ -720,6 +720,8 @@ static const u32 cayman_sysls_enable[] =
 struct rv7xx_power_info *rv770_get_pi(struct radeon_device *rdev);
 struct evergreen_power_info *evergreen_get_pi(struct radeon_device *rdev);
 
+extern int ni_mc_load_microcode(struct radeon_device *rdev);
+
 struct ni_power_info *ni_get_pi(struct radeon_device *rdev)
 {
         struct ni_power_info *pi = rdev->pm.dpm.priv;
@@ -3565,7 +3567,11 @@ void ni_set_uvd_clock_after_set_eng_clock(struct radeon_device *rdev,
 void ni_dpm_setup_asic(struct radeon_device *rdev)
 {
 	struct evergreen_power_info *eg_pi = evergreen_get_pi(rdev);
+	int r;
 
+	r = ni_mc_load_microcode(rdev);
+	if (r)
+		DRM_ERROR("Failed to load MC firmware!\n");
 	ni_read_clock_registers(rdev);
 	btc_read_arb_registers(rdev);
 	rv770_get_memory_type(rdev);
@@ -3710,21 +3716,6 @@ int ni_dpm_enable(struct radeon_device *rdev)
 	if (eg_pi->ls_clock_gating)
 		ni_ls_clockgating_enable(rdev, true);
 
-	if (rdev->irq.installed &&
-	    r600_is_internal_thermal_sensor(rdev->pm.int_thermal_type)) {
-		PPSMC_Result result;
-
-		ret = rv770_set_thermal_temperature_range(rdev, R600_TEMP_RANGE_MIN, 0xff * 1000);
-		if (ret)
-			return ret;
-		rdev->irq.dpm_thermal = true;
-		radeon_irq_set(rdev);
-		result = rv770_send_msg_to_smc(rdev, PPSMC_MSG_EnableThermalInterrupt);
-
-		if (result != PPSMC_Result_OK)
-			DRM_DEBUG_KMS("Could not enable thermal interrupts.\n");
-	}
-
 	rv770_enable_auto_throttle_source(rdev, RADEON_DPM_AUTO_THROTTLE_SRC_THERMAL, true);
 
 	ni_update_current_ps(rdev, boot_ps);
diff --git a/drivers/gpu/drm/radeon/nid.h b/drivers/gpu/drm/radeon/nid.h
index 22421bc80c0d..d996033c243e 100644
--- a/drivers/gpu/drm/radeon/nid.h
+++ b/drivers/gpu/drm/radeon/nid.h
@@ -1154,6 +1154,7 @@
 #              define PACKET3_DB_ACTION_ENA        (1 << 26)
 #              define PACKET3_SH_ACTION_ENA        (1 << 27)
 #              define PACKET3_SX_ACTION_ENA        (1 << 28)
+#              define PACKET3_ENGINE_ME            (1 << 31)
 #define	PACKET3_ME_INITIALIZE				0x44
 #define		PACKET3_ME_INITIALIZE_DEVICE_ID(x) ((x) << 16)
 #define	PACKET3_COND_WRITE				0x45
diff --git a/drivers/gpu/drm/radeon/pptable.h b/drivers/gpu/drm/radeon/pptable.h
index da43ab328833..2d532996c697 100644
--- a/drivers/gpu/drm/radeon/pptable.h
+++ b/drivers/gpu/drm/radeon/pptable.h
@@ -23,7 +23,7 @@
 #ifndef _PPTABLE_H
 #define _PPTABLE_H
 
-#pragma pack(push, 1)
+#pragma pack(1)
 
 typedef struct _ATOM_PPLIB_THERMALCONTROLLER
 
@@ -677,6 +677,6 @@ typedef struct _ATOM_PPLIB_PPM_Table
       ULONG  ulTjmax;
 } ATOM_PPLIB_PPM_Table;
 
-#pragma pack(pop)
+#pragma pack()
 
 #endif
diff --git a/drivers/gpu/drm/radeon/r100.c b/drivers/gpu/drm/radeon/r100.c
index 10abc4d5a6cc..ef024ce3f7cc 100644
--- a/drivers/gpu/drm/radeon/r100.c
+++ b/drivers/gpu/drm/radeon/r100.c
@@ -1050,6 +1050,36 @@ static int r100_cp_init_microcode(struct radeon_device *rdev)
 	return err;
 }
 
+u32 r100_gfx_get_rptr(struct radeon_device *rdev,
+		      struct radeon_ring *ring)
+{
+	u32 rptr;
+
+	if (rdev->wb.enabled)
+		rptr = le32_to_cpu(rdev->wb.wb[ring->rptr_offs/4]);
+	else
+		rptr = RREG32(RADEON_CP_RB_RPTR);
+
+	return rptr;
+}
+
+u32 r100_gfx_get_wptr(struct radeon_device *rdev,
+		      struct radeon_ring *ring)
+{
+	u32 wptr;
+
+	wptr = RREG32(RADEON_CP_RB_WPTR);
+
+	return wptr;
+}
+
+void r100_gfx_set_wptr(struct radeon_device *rdev,
+		       struct radeon_ring *ring)
+{
+	WREG32(RADEON_CP_RB_WPTR, ring->wptr);
+	(void)RREG32(RADEON_CP_RB_WPTR);
+}
+
 static void r100_cp_load_microcode(struct radeon_device *rdev)
 {
 	const __be32 *fw_data;
@@ -1102,7 +1132,6 @@ int r100_cp_init(struct radeon_device *rdev, unsigned ring_size)
 	ring_size = (1 << (rb_bufsz + 1)) * 4;
 	r100_cp_load_microcode(rdev);
 	r = radeon_ring_init(rdev, ring, ring_size, RADEON_WB_CP_RPTR_OFFSET,
-			     RADEON_CP_RB_RPTR, RADEON_CP_RB_WPTR,
 			     RADEON_CP_PACKET2);
 	if (r) {
 		return r;
@@ -3913,6 +3942,8 @@ int r100_resume(struct radeon_device *rdev)
 	/* Initialize surface registers */
 	radeon_surface_init(rdev);
 
+	radeon_pm_resume(rdev);
+
 	rdev->accel_working = true;
 	r = r100_startup(rdev);
 	if (r) {
@@ -3923,6 +3954,7 @@ int r100_resume(struct radeon_device *rdev)
 
 int r100_suspend(struct radeon_device *rdev)
 {
+	radeon_pm_suspend(rdev);
 	r100_cp_disable(rdev);
 	radeon_wb_disable(rdev);
 	r100_irq_disable(rdev);
@@ -3933,6 +3965,7 @@ int r100_suspend(struct radeon_device *rdev)
 
 void r100_fini(struct radeon_device *rdev)
 {
+	radeon_pm_fini(rdev);
 	r100_cp_fini(rdev);
 	radeon_wb_fini(rdev);
 	radeon_ib_pool_fini(rdev);
@@ -4039,6 +4072,9 @@ int r100_init(struct radeon_device *rdev)
 	}
 	r100_set_safe_registers(rdev);
 
+	/* Initialize power management */
+	radeon_pm_init(rdev);
+
 	rdev->accel_working = true;
 	r = r100_startup(rdev);
 	if (r) {
diff --git a/drivers/gpu/drm/radeon/r300.c b/drivers/gpu/drm/radeon/r300.c
index d8dd269b9159..7c63ef840e86 100644
--- a/drivers/gpu/drm/radeon/r300.c
+++ b/drivers/gpu/drm/radeon/r300.c
@@ -1430,6 +1430,8 @@ int r300_resume(struct radeon_device *rdev)
 	/* Initialize surface registers */
 	radeon_surface_init(rdev);
 
+	radeon_pm_resume(rdev);
+
 	rdev->accel_working = true;
 	r = r300_startup(rdev);
 	if (r) {
@@ -1440,6 +1442,7 @@ int r300_resume(struct radeon_device *rdev)
 
 int r300_suspend(struct radeon_device *rdev)
 {
+	radeon_pm_suspend(rdev);
 	r100_cp_disable(rdev);
 	radeon_wb_disable(rdev);
 	r100_irq_disable(rdev);
@@ -1452,6 +1455,7 @@ int r300_suspend(struct radeon_device *rdev)
 
 void r300_fini(struct radeon_device *rdev)
 {
+	radeon_pm_fini(rdev);
 	r100_cp_fini(rdev);
 	radeon_wb_fini(rdev);
 	radeon_ib_pool_fini(rdev);
@@ -1538,6 +1542,9 @@ int r300_init(struct radeon_device *rdev)
 	}
 	r300_set_reg_safe(rdev);
 
+	/* Initialize power management */
+	radeon_pm_init(rdev);
+
 	rdev->accel_working = true;
 	r = r300_startup(rdev);
 	if (r) {
diff --git a/drivers/gpu/drm/radeon/r300_cmdbuf.c b/drivers/gpu/drm/radeon/r300_cmdbuf.c
index 60170ea5e3a2..84b1d5367a11 100644
--- a/drivers/gpu/drm/radeon/r300_cmdbuf.c
+++ b/drivers/gpu/drm/radeon/r300_cmdbuf.c
@@ -75,7 +75,7 @@ static int r300_emit_cliprects(drm_radeon_private_t *dev_priv,
 		OUT_RING(CP_PACKET0(R300_RE_CLIPRECT_TL_0, nr * 2 - 1));
 
 		for (i = 0; i < nr; ++i) {
-			if (DRM_COPY_FROM_USER
+			if (copy_from_user
 			    (&box, &cmdbuf->boxes[n + i], sizeof(box))) {
 				DRM_ERROR("copy cliprect faulted\n");
 				return -EFAULT;
@@ -928,12 +928,12 @@ static int r300_scratch(drm_radeon_private_t *dev_priv,
 		buf_idx = drm_buffer_pointer_to_dword(cmdbuf->buffer, 0);
 		*buf_idx *= 2; /* 8 bytes per buf */
 
-		if (DRM_COPY_TO_USER(ref_age_base + *buf_idx,
+		if (copy_to_user(ref_age_base + *buf_idx,
 				&dev_priv->scratch_ages[header.scratch.reg],
 				sizeof(u32)))
 			return -EINVAL;
 
-		if (DRM_COPY_FROM_USER(&h_pending,
+		if (copy_from_user(&h_pending,
 				ref_age_base + *buf_idx + 1,
 				sizeof(u32)))
 			return -EINVAL;
@@ -943,7 +943,7 @@ static int r300_scratch(drm_radeon_private_t *dev_priv,
 
 		h_pending--;
 
-		if (DRM_COPY_TO_USER(ref_age_base + *buf_idx + 1,
+		if (copy_to_user(ref_age_base + *buf_idx + 1,
 					&h_pending,
 					sizeof(u32)))
 			return -EINVAL;
diff --git a/drivers/gpu/drm/radeon/r420.c b/drivers/gpu/drm/radeon/r420.c
index 6edf2b3a52b4..3768aab2710b 100644
--- a/drivers/gpu/drm/radeon/r420.c
+++ b/drivers/gpu/drm/radeon/r420.c
@@ -325,6 +325,8 @@ int r420_resume(struct radeon_device *rdev)
 	/* Initialize surface registers */
 	radeon_surface_init(rdev);
 
+	radeon_pm_resume(rdev);
+
 	rdev->accel_working = true;
 	r = r420_startup(rdev);
 	if (r) {
@@ -335,6 +337,7 @@ int r420_resume(struct radeon_device *rdev)
 
 int r420_suspend(struct radeon_device *rdev)
 {
+	radeon_pm_suspend(rdev);
 	r420_cp_errata_fini(rdev);
 	r100_cp_disable(rdev);
 	radeon_wb_disable(rdev);
@@ -348,6 +351,7 @@ int r420_suspend(struct radeon_device *rdev)
 
 void r420_fini(struct radeon_device *rdev)
 {
+	radeon_pm_fini(rdev);
 	r100_cp_fini(rdev);
 	radeon_wb_fini(rdev);
 	radeon_ib_pool_fini(rdev);
@@ -444,6 +448,9 @@ int r420_init(struct radeon_device *rdev)
 	}
 	r420_set_reg_safe(rdev);
 
+	/* Initialize power management */
+	radeon_pm_init(rdev);
+
 	rdev->accel_working = true;
 	r = r420_startup(rdev);
 	if (r) {
diff --git a/drivers/gpu/drm/radeon/r520.c b/drivers/gpu/drm/radeon/r520.c
index e1aece73b370..e209eb75024f 100644
--- a/drivers/gpu/drm/radeon/r520.c
+++ b/drivers/gpu/drm/radeon/r520.c
@@ -240,6 +240,8 @@ int r520_resume(struct radeon_device *rdev)
 	/* Initialize surface registers */
 	radeon_surface_init(rdev);
 
+	radeon_pm_resume(rdev);
+
 	rdev->accel_working = true;
 	r = r520_startup(rdev);
 	if (r) {
@@ -312,6 +314,9 @@ int r520_init(struct radeon_device *rdev)
 		return r;
 	rv515_set_safe_registers(rdev);
 
+	/* Initialize power management */
+	radeon_pm_init(rdev);
+
 	rdev->accel_working = true;
 	r = r520_startup(rdev);
 	if (r) {
diff --git a/drivers/gpu/drm/radeon/r600.c b/drivers/gpu/drm/radeon/r600.c
index 9ad06732a78b..56140b4e5bb2 100644
--- a/drivers/gpu/drm/radeon/r600.c
+++ b/drivers/gpu/drm/radeon/r600.c
@@ -105,6 +105,7 @@ void r600_fini(struct radeon_device *rdev);
 void r600_irq_disable(struct radeon_device *rdev);
 static void r600_pcie_gen2_enable(struct radeon_device *rdev);
 extern int evergreen_rlc_resume(struct radeon_device *rdev);
+extern void rv770_set_clk_bypass_mode(struct radeon_device *rdev);
 
 /**
  * r600_get_xclk - get the xclk
@@ -1644,6 +1645,67 @@ static void r600_gpu_soft_reset(struct radeon_device *rdev, u32 reset_mask)
 	r600_print_gpu_status_regs(rdev);
 }
 
+static void r600_gpu_pci_config_reset(struct radeon_device *rdev)
+{
+	struct rv515_mc_save save;
+	u32 tmp, i;
+
+	dev_info(rdev->dev, "GPU pci config reset\n");
+
+	/* disable dpm? */
+
+	/* Disable CP parsing/prefetching */
+	if (rdev->family >= CHIP_RV770)
+		WREG32(R_0086D8_CP_ME_CNTL, S_0086D8_CP_ME_HALT(1) | S_0086D8_CP_PFP_HALT(1));
+	else
+		WREG32(R_0086D8_CP_ME_CNTL, S_0086D8_CP_ME_HALT(1));
+
+	/* disable the RLC */
+	WREG32(RLC_CNTL, 0);
+
+	/* Disable DMA */
+	tmp = RREG32(DMA_RB_CNTL);
+	tmp &= ~DMA_RB_ENABLE;
+	WREG32(DMA_RB_CNTL, tmp);
+
+	mdelay(50);
+
+	/* set mclk/sclk to bypass */
+	if (rdev->family >= CHIP_RV770)
+		rv770_set_clk_bypass_mode(rdev);
+	/* disable BM */
+	pci_clear_master(rdev->pdev);
+	/* disable mem access */
+	rv515_mc_stop(rdev, &save);
+	if (r600_mc_wait_for_idle(rdev)) {
+		dev_warn(rdev->dev, "Wait for MC idle timedout !\n");
+	}
+
+	/* BIF reset workaround.  Not sure if this is needed on 6xx */
+	tmp = RREG32(BUS_CNTL);
+	tmp |= VGA_COHE_SPEC_TIMER_DIS;
+	WREG32(BUS_CNTL, tmp);
+
+	tmp = RREG32(BIF_SCRATCH0);
+
+	/* reset */
+	radeon_pci_config_reset(rdev);
+	mdelay(1);
+
+	/* BIF reset workaround.  Not sure if this is needed on 6xx */
+	tmp = SOFT_RESET_BIF;
+	WREG32(SRBM_SOFT_RESET, tmp);
+	mdelay(1);
+	WREG32(SRBM_SOFT_RESET, 0);
+
+	/* wait for asic to come out of reset */
+	for (i = 0; i < rdev->usec_timeout; i++) {
+		if (RREG32(CONFIG_MEMSIZE) != 0xffffffff)
+			break;
+		udelay(1);
+	}
+}
+
 int r600_asic_reset(struct radeon_device *rdev)
 {
 	u32 reset_mask;
@@ -1653,10 +1715,17 @@ int r600_asic_reset(struct radeon_device *rdev)
 	if (reset_mask)
 		r600_set_bios_scratch_engine_hung(rdev, true);
 
+	/* try soft reset */
 	r600_gpu_soft_reset(rdev, reset_mask);
 
 	reset_mask = r600_gpu_check_soft_reset(rdev);
 
+	/* try pci config reset */
+	if (reset_mask && radeon_hard_reset)
+		r600_gpu_pci_config_reset(rdev);
+
+	reset_mask = r600_gpu_check_soft_reset(rdev);
+
 	if (!reset_mask)
 		r600_set_bios_scratch_engine_hung(rdev, false);
 
@@ -2185,7 +2254,8 @@ void r600_pciep_wreg(struct radeon_device *rdev, u32 reg, u32 v)
  */
 void r600_cp_stop(struct radeon_device *rdev)
 {
-	radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
+	if (rdev->asic->copy.copy_ring_index == RADEON_RING_TYPE_GFX_INDEX)
+		radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
 	WREG32(R_0086D8_CP_ME_CNTL, S_0086D8_CP_ME_HALT(1));
 	WREG32(SCRATCH_UMSK, 0);
 	rdev->ring[RADEON_RING_TYPE_GFX_INDEX].ready = false;
@@ -2382,6 +2452,36 @@ out:
 	return err;
 }
 
+u32 r600_gfx_get_rptr(struct radeon_device *rdev,
+		      struct radeon_ring *ring)
+{
+	u32 rptr;
+
+	if (rdev->wb.enabled)
+		rptr = rdev->wb.wb[ring->rptr_offs/4];
+	else
+		rptr = RREG32(R600_CP_RB_RPTR);
+
+	return rptr;
+}
+
+u32 r600_gfx_get_wptr(struct radeon_device *rdev,
+		      struct radeon_ring *ring)
+{
+	u32 wptr;
+
+	wptr = RREG32(R600_CP_RB_WPTR);
+
+	return wptr;
+}
+
+void r600_gfx_set_wptr(struct radeon_device *rdev,
+		       struct radeon_ring *ring)
+{
+	WREG32(R600_CP_RB_WPTR, ring->wptr);
+	(void)RREG32(R600_CP_RB_WPTR);
+}
+
 static int r600_cp_load_microcode(struct radeon_device *rdev)
 {
 	const __be32 *fw_data;
@@ -2513,6 +2613,10 @@ int r600_cp_resume(struct radeon_device *rdev)
 		ring->ready = false;
 		return r;
 	}
+
+	if (rdev->asic->copy.copy_ring_index == RADEON_RING_TYPE_GFX_INDEX)
+		radeon_ttm_set_active_vram_size(rdev, rdev->mc.real_vram_size);
+
 	return 0;
 }
 
@@ -2607,14 +2711,17 @@ void r600_fence_ring_emit(struct radeon_device *rdev,
 			  struct radeon_fence *fence)
 {
 	struct radeon_ring *ring = &rdev->ring[fence->ring];
+	u32 cp_coher_cntl = PACKET3_TC_ACTION_ENA | PACKET3_VC_ACTION_ENA |
+		PACKET3_SH_ACTION_ENA;
+
+	if (rdev->family >= CHIP_RV770)
+		cp_coher_cntl |= PACKET3_FULL_CACHE_ENA;
 
 	if (rdev->wb.use_event) {
 		u64 addr = rdev->fence_drv[fence->ring].gpu_addr;
 		/* flush read cache over gart */
 		radeon_ring_write(ring, PACKET3(PACKET3_SURFACE_SYNC, 3));
-		radeon_ring_write(ring, PACKET3_TC_ACTION_ENA |
-					PACKET3_VC_ACTION_ENA |
-					PACKET3_SH_ACTION_ENA);
+		radeon_ring_write(ring, cp_coher_cntl);
 		radeon_ring_write(ring, 0xFFFFFFFF);
 		radeon_ring_write(ring, 0);
 		radeon_ring_write(ring, 10); /* poll interval */
@@ -2628,9 +2735,7 @@ void r600_fence_ring_emit(struct radeon_device *rdev,
 	} else {
 		/* flush read cache over gart */
 		radeon_ring_write(ring, PACKET3(PACKET3_SURFACE_SYNC, 3));
-		radeon_ring_write(ring, PACKET3_TC_ACTION_ENA |
-					PACKET3_VC_ACTION_ENA |
-					PACKET3_SH_ACTION_ENA);
+		radeon_ring_write(ring, cp_coher_cntl);
 		radeon_ring_write(ring, 0xFFFFFFFF);
 		radeon_ring_write(ring, 0);
 		radeon_ring_write(ring, 10); /* poll interval */
@@ -2775,14 +2880,6 @@ static int r600_startup(struct radeon_device *rdev)
 
 	r600_mc_program(rdev);
 
-	if (!rdev->me_fw || !rdev->pfp_fw || !rdev->rlc_fw) {
-		r = r600_init_microcode(rdev);
-		if (r) {
-			DRM_ERROR("Failed to load firmware!\n");
-			return r;
-		}
-	}
-
 	if (rdev->flags & RADEON_IS_AGP) {
 		r600_agp_enable(rdev);
 	} else {
@@ -2803,12 +2900,6 @@ static int r600_startup(struct radeon_device *rdev)
 		return r;
 	}
 
-	r = radeon_fence_driver_start_ring(rdev, R600_RING_TYPE_DMA_INDEX);
-	if (r) {
-		dev_err(rdev->dev, "failed initializing DMA fences (%d).\n", r);
-		return r;
-	}
-
 	/* Enable IRQ */
 	if (!rdev->irq.installed) {
 		r = radeon_irq_kms_init(rdev);
@@ -2826,18 +2917,10 @@ static int r600_startup(struct radeon_device *rdev)
 
 	ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, RADEON_WB_CP_RPTR_OFFSET,
-			     R600_CP_RB_RPTR, R600_CP_RB_WPTR,
 			     RADEON_CP_PACKET2);
 	if (r)
 		return r;
 
-	ring = &rdev->ring[R600_RING_TYPE_DMA_INDEX];
-	r = radeon_ring_init(rdev, ring, ring->ring_size, R600_WB_DMA_RPTR_OFFSET,
-			     DMA_RB_RPTR, DMA_RB_WPTR,
-			     DMA_PACKET(DMA_PACKET_NOP, 0, 0, 0));
-	if (r)
-		return r;
-
 	r = r600_cp_load_microcode(rdev);
 	if (r)
 		return r;
@@ -2845,10 +2928,6 @@ static int r600_startup(struct radeon_device *rdev)
 	if (r)
 		return r;
 
-	r = r600_dma_resume(rdev);
-	if (r)
-		return r;
-
 	r = radeon_ib_pool_init(rdev);
 	if (r) {
 		dev_err(rdev->dev, "IB initialization failed (%d).\n", r);
@@ -2889,6 +2968,8 @@ int r600_resume(struct radeon_device *rdev)
 	/* post card */
 	atom_asic_init(rdev->mode_info.atom_context);
 
+	radeon_pm_resume(rdev);
+
 	rdev->accel_working = true;
 	r = r600_startup(rdev);
 	if (r) {
@@ -2902,9 +2983,9 @@ int r600_resume(struct radeon_device *rdev)
 
 int r600_suspend(struct radeon_device *rdev)
 {
+	radeon_pm_suspend(rdev);
 	r600_audio_fini(rdev);
 	r600_cp_stop(rdev);
-	r600_dma_stop(rdev);
 	r600_irq_suspend(rdev);
 	radeon_wb_disable(rdev);
 	r600_pcie_gart_disable(rdev);
@@ -2970,12 +3051,20 @@ int r600_init(struct radeon_device *rdev)
 	if (r)
 		return r;
 
+	if (!rdev->me_fw || !rdev->pfp_fw || !rdev->rlc_fw) {
+		r = r600_init_microcode(rdev);
+		if (r) {
+			DRM_ERROR("Failed to load firmware!\n");
+			return r;
+		}
+	}
+
+	/* Initialize power management */
+	radeon_pm_init(rdev);
+
 	rdev->ring[RADEON_RING_TYPE_GFX_INDEX].ring_obj = NULL;
 	r600_ring_init(rdev, &rdev->ring[RADEON_RING_TYPE_GFX_INDEX], 1024 * 1024);
 
-	rdev->ring[R600_RING_TYPE_DMA_INDEX].ring_obj = NULL;
-	r600_ring_init(rdev, &rdev->ring[R600_RING_TYPE_DMA_INDEX], 64 * 1024);
-
 	rdev->ih.ring_obj = NULL;
 	r600_ih_ring_init(rdev, 64 * 1024);
 
@@ -2988,7 +3077,6 @@ int r600_init(struct radeon_device *rdev)
 	if (r) {
 		dev_err(rdev->dev, "disabling GPU acceleration\n");
 		r600_cp_fini(rdev);
-		r600_dma_fini(rdev);
 		r600_irq_fini(rdev);
 		radeon_wb_fini(rdev);
 		radeon_ib_pool_fini(rdev);
@@ -3002,9 +3090,9 @@ int r600_init(struct radeon_device *rdev)
 
 void r600_fini(struct radeon_device *rdev)
 {
+	radeon_pm_fini(rdev);
 	r600_audio_fini(rdev);
 	r600_cp_fini(rdev);
-	r600_dma_fini(rdev);
 	r600_irq_fini(rdev);
 	radeon_wb_fini(rdev);
 	radeon_ib_pool_fini(rdev);
diff --git a/drivers/gpu/drm/radeon/r600_cp.c b/drivers/gpu/drm/radeon/r600_cp.c
index d8eb48bff0ed..8c9b7e26533c 100644
--- a/drivers/gpu/drm/radeon/r600_cp.c
+++ b/drivers/gpu/drm/radeon/r600_cp.c
@@ -2515,7 +2515,7 @@ int r600_cp_dispatch_texture(struct drm_device *dev,
 		buf = radeon_freelist_get(dev);
 		if (!buf) {
 			DRM_DEBUG("EAGAIN\n");
-			if (DRM_COPY_TO_USER(tex->image, image, sizeof(*image)))
+			if (copy_to_user(tex->image, image, sizeof(*image)))
 				return -EFAULT;
 			return -EAGAIN;
 		}
@@ -2528,7 +2528,7 @@ int r600_cp_dispatch_texture(struct drm_device *dev,
 		buffer =
 		    (u32 *) ((char *)dev->agp_buffer_map->handle + buf->offset);
 
-		if (DRM_COPY_FROM_USER(buffer, data, pass_size)) {
+		if (copy_from_user(buffer, data, pass_size)) {
 			DRM_ERROR("EFAULT on pad, %d bytes\n", pass_size);
 			return -EFAULT;
 		}
diff --git a/drivers/gpu/drm/radeon/r600_cs.c b/drivers/gpu/drm/radeon/r600_cs.c
index 5dceea6f71ae..7b399dc5fd54 100644
--- a/drivers/gpu/drm/radeon/r600_cs.c
+++ b/drivers/gpu/drm/radeon/r600_cs.c
@@ -749,7 +749,10 @@ static int r600_cs_track_check(struct radeon_cs_parser *p)
 		}
 
 		for (i = 0; i < 8; i++) {
-			if ((tmp >> (i * 4)) & 0xF) {
+			u32 format = G_0280A0_FORMAT(track->cb_color_info[i]);
+
+			if (format != V_0280A0_COLOR_INVALID &&
+			    (tmp >> (i * 4)) & 0xF) {
 				/* at least one component is enabled */
 				if (track->cb_color_bo[i] == NULL) {
 					dev_warn(p->dev, "%s:%d mask 0x%08X | 0x%08X no cb for %d\n",
@@ -2386,7 +2389,7 @@ int r600_cs_legacy(struct drm_device *dev, void *data, struct drm_file *filp,
 	ib_chunk = &parser.chunks[parser.chunk_ib_idx];
 	parser.ib.length_dw = ib_chunk->length_dw;
 	*l = parser.ib.length_dw;
-	if (DRM_COPY_FROM_USER(ib, ib_chunk->user_ptr, ib_chunk->length_dw * 4)) {
+	if (copy_from_user(ib, ib_chunk->user_ptr, ib_chunk->length_dw * 4)) {
 		r = -EFAULT;
 		r600_cs_parser_fini(&parser, r);
 		return r;
diff --git a/drivers/gpu/drm/radeon/r600_dma.c b/drivers/gpu/drm/radeon/r600_dma.c
index 7844d15c139f..b2d4c91e6272 100644
--- a/drivers/gpu/drm/radeon/r600_dma.c
+++ b/drivers/gpu/drm/radeon/r600_dma.c
@@ -51,7 +51,14 @@ u32 r600_gpu_check_soft_reset(struct radeon_device *rdev);
 uint32_t r600_dma_get_rptr(struct radeon_device *rdev,
 			   struct radeon_ring *ring)
 {
-	return (radeon_ring_generic_get_rptr(rdev, ring) & 0x3fffc) >> 2;
+	u32 rptr;
+
+	if (rdev->wb.enabled)
+		rptr = rdev->wb.wb[ring->rptr_offs/4];
+	else
+		rptr = RREG32(DMA_RB_RPTR);
+
+	return (rptr & 0x3fffc) >> 2;
 }
 
 /**
@@ -65,7 +72,7 @@ uint32_t r600_dma_get_rptr(struct radeon_device *rdev,
 uint32_t r600_dma_get_wptr(struct radeon_device *rdev,
 			   struct radeon_ring *ring)
 {
-	return (RREG32(ring->wptr_reg) & 0x3fffc) >> 2;
+	return (RREG32(DMA_RB_WPTR) & 0x3fffc) >> 2;
 }
 
 /**
@@ -79,7 +86,7 @@ uint32_t r600_dma_get_wptr(struct radeon_device *rdev,
 void r600_dma_set_wptr(struct radeon_device *rdev,
 		       struct radeon_ring *ring)
 {
-	WREG32(ring->wptr_reg, (ring->wptr << 2) & 0x3fffc);
+	WREG32(DMA_RB_WPTR, (ring->wptr << 2) & 0x3fffc);
 }
 
 /**
@@ -93,7 +100,8 @@ void r600_dma_stop(struct radeon_device *rdev)
 {
 	u32 rb_cntl = RREG32(DMA_RB_CNTL);
 
-	radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
+	if (rdev->asic->copy.copy_ring_index == R600_RING_TYPE_DMA_INDEX)
+		radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
 
 	rb_cntl &= ~DMA_RB_ENABLE;
 	WREG32(DMA_RB_CNTL, rb_cntl);
@@ -180,7 +188,8 @@ int r600_dma_resume(struct radeon_device *rdev)
 		return r;
 	}
 
-	radeon_ttm_set_active_vram_size(rdev, rdev->mc.real_vram_size);
+	if (rdev->asic->copy.copy_ring_index == R600_RING_TYPE_DMA_INDEX)
+		radeon_ttm_set_active_vram_size(rdev, rdev->mc.real_vram_size);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/radeon/r600_dpm.c b/drivers/gpu/drm/radeon/r600_dpm.c
index 5513d8f06252..e4cc9b314ce9 100644
--- a/drivers/gpu/drm/radeon/r600_dpm.c
+++ b/drivers/gpu/drm/radeon/r600_dpm.c
@@ -729,8 +729,8 @@ bool r600_is_uvd_state(u32 class, u32 class2)
 	return false;
 }
 
-int r600_set_thermal_temperature_range(struct radeon_device *rdev,
-				       int min_temp, int max_temp)
+static int r600_set_thermal_temperature_range(struct radeon_device *rdev,
+					      int min_temp, int max_temp)
 {
 	int low_temp = 0 * 1000;
 	int high_temp = 255 * 1000;
@@ -777,6 +777,22 @@ bool r600_is_internal_thermal_sensor(enum radeon_int_thermal_type sensor)
 	}
 }
 
+int r600_dpm_late_enable(struct radeon_device *rdev)
+{
+	int ret;
+
+	if (rdev->irq.installed &&
+	    r600_is_internal_thermal_sensor(rdev->pm.int_thermal_type)) {
+		ret = r600_set_thermal_temperature_range(rdev, R600_TEMP_RANGE_MIN, R600_TEMP_RANGE_MAX);
+		if (ret)
+			return ret;
+		rdev->irq.dpm_thermal = true;
+		radeon_irq_set(rdev);
+	}
+
+	return 0;
+}
+
 union power_info {
 	struct _ATOM_POWERPLAY_INFO info;
 	struct _ATOM_POWERPLAY_INFO_V2 info_2;
diff --git a/drivers/gpu/drm/radeon/r600_dpm.h b/drivers/gpu/drm/radeon/r600_dpm.h
index 1000bf9719f2..07eab2b04e81 100644
--- a/drivers/gpu/drm/radeon/r600_dpm.h
+++ b/drivers/gpu/drm/radeon/r600_dpm.h
@@ -213,8 +213,6 @@ void r600_wait_for_power_level(struct radeon_device *rdev,
 void r600_start_dpm(struct radeon_device *rdev);
 void r600_stop_dpm(struct radeon_device *rdev);
 
-int r600_set_thermal_temperature_range(struct radeon_device *rdev,
-				       int min_temp, int max_temp);
 bool r600_is_internal_thermal_sensor(enum radeon_int_thermal_type sensor);
 
 int r600_parse_extended_power_table(struct radeon_device *rdev);
diff --git a/drivers/gpu/drm/radeon/r600_hdmi.c b/drivers/gpu/drm/radeon/r600_hdmi.c
index b7d3ecba43e3..3016fc14f502 100644
--- a/drivers/gpu/drm/radeon/r600_hdmi.c
+++ b/drivers/gpu/drm/radeon/r600_hdmi.c
@@ -250,7 +250,7 @@ static void r600_hdmi_audio_workaround(struct drm_encoder *encoder)
 		 value, ~HDMI0_AUDIO_TEST_EN);
 }
 
-void r600_audio_set_dto(struct drm_encoder *encoder, u32 clock)
+static void r600_audio_set_dto(struct drm_encoder *encoder, u32 clock)
 {
 	struct drm_device *dev = encoder->dev;
 	struct radeon_device *rdev = dev->dev_private;
diff --git a/drivers/gpu/drm/radeon/r600d.h b/drivers/gpu/drm/radeon/r600d.h
index ebe38724a976..37455f65107f 100644
--- a/drivers/gpu/drm/radeon/r600d.h
+++ b/drivers/gpu/drm/radeon/r600d.h
@@ -701,11 +701,18 @@
 #define RLC_UCODE_DATA                                    0x3f30
 
 #define SRBM_SOFT_RESET                                   0xe60
+#       define SOFT_RESET_BIF                             (1 << 1)
 #       define SOFT_RESET_DMA                             (1 << 12)
 #       define SOFT_RESET_RLC                             (1 << 13)
 #       define SOFT_RESET_UVD                             (1 << 18)
 #       define RV770_SOFT_RESET_DMA                       (1 << 20)
 
+#define BIF_SCRATCH0                                      0x5438
+
+#define BUS_CNTL                                          0x5420
+#       define BIOS_ROM_DIS                               (1 << 1)
+#       define VGA_COHE_SPEC_TIMER_DIS                    (1 << 9)
+
 #define CP_INT_CNTL                                       0xc124
 #       define CNTX_BUSY_INT_ENABLE                       (1 << 19)
 #       define CNTX_EMPTY_INT_ENABLE                      (1 << 20)
@@ -1575,6 +1582,7 @@
 #              define PACKET3_CP_DMA_CMD_DAIC      (1 << 29)
 #define	PACKET3_SURFACE_SYNC				0x43
 #              define PACKET3_CB0_DEST_BASE_ENA    (1 << 6)
+#              define PACKET3_FULL_CACHE_ENA       (1 << 20) /* r7xx+ only */
 #              define PACKET3_TC_ACTION_ENA        (1 << 23)
 #              define PACKET3_VC_ACTION_ENA        (1 << 24)
 #              define PACKET3_CB_ACTION_ENA        (1 << 25)
diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h
index 45e1f447bc79..4a8ac1cd6b4c 100644
--- a/drivers/gpu/drm/radeon/radeon.h
+++ b/drivers/gpu/drm/radeon/radeon.h
@@ -99,6 +99,7 @@ extern int radeon_fastfb;
 extern int radeon_dpm;
 extern int radeon_aspm;
 extern int radeon_runtime_pm;
+extern int radeon_hard_reset;
 
 /*
  * Copy from radeon_drv.h so we don't have to include both and have conflicting
@@ -139,6 +140,9 @@ extern int radeon_runtime_pm;
 #define RADEON_VA_RESERVED_SIZE			(8 << 20)
 #define RADEON_IB_VM_MAX_SIZE			(64 << 10)
 
+/* hard reset data */
+#define RADEON_ASIC_RESET_DATA                  0x39d5e86b
+
 /* reset flags */
 #define RADEON_RESET_GFX			(1 << 0)
 #define RADEON_RESET_COMPUTE			(1 << 1)
@@ -252,6 +256,7 @@ struct radeon_clock {
  * Power management
  */
 int radeon_pm_init(struct radeon_device *rdev);
+int radeon_pm_late_init(struct radeon_device *rdev);
 void radeon_pm_fini(struct radeon_device *rdev);
 void radeon_pm_compute_clocks(struct radeon_device *rdev);
 void radeon_pm_suspend(struct radeon_device *rdev);
@@ -413,6 +418,11 @@ struct radeon_mman {
 	struct ttm_bo_device		bdev;
 	bool				mem_global_referenced;
 	bool				initialized;
+
+#if defined(CONFIG_DEBUG_FS)
+	struct dentry			*vram;
+	struct dentry			*gtt;
+#endif
 };
 
 /* bo virtual address in a specific vm */
@@ -779,13 +789,11 @@ struct radeon_ring {
 	volatile uint32_t	*ring;
 	unsigned		rptr;
 	unsigned		rptr_offs;
-	unsigned		rptr_reg;
 	unsigned		rptr_save_reg;
 	u64			next_rptr_gpu_addr;
 	volatile u32		*next_rptr_cpu_addr;
 	unsigned		wptr;
 	unsigned		wptr_old;
-	unsigned		wptr_reg;
 	unsigned		ring_size;
 	unsigned		ring_free_dw;
 	int			count_dw;
@@ -859,6 +867,8 @@ struct radeon_vm {
 	struct radeon_fence		*fence;
 	/* last flush or NULL if we still need to flush */
 	struct radeon_fence		*last_flush;
+	/* last use of vmid */
+	struct radeon_fence		*last_id_use;
 };
 
 struct radeon_vm_manager {
@@ -949,7 +959,7 @@ unsigned radeon_ring_backup(struct radeon_device *rdev, struct radeon_ring *ring
 int radeon_ring_restore(struct radeon_device *rdev, struct radeon_ring *ring,
 			unsigned size, uint32_t *data);
 int radeon_ring_init(struct radeon_device *rdev, struct radeon_ring *cp, unsigned ring_size,
-		     unsigned rptr_offs, unsigned rptr_reg, unsigned wptr_reg, u32 nop);
+		     unsigned rptr_offs, u32 nop);
 void radeon_ring_fini(struct radeon_device *rdev, struct radeon_ring *cp);
 
 
@@ -1775,6 +1785,7 @@ struct radeon_asic {
 		int (*init)(struct radeon_device *rdev);
 		void (*setup_asic)(struct radeon_device *rdev);
 		int (*enable)(struct radeon_device *rdev);
+		int (*late_enable)(struct radeon_device *rdev);
 		void (*disable)(struct radeon_device *rdev);
 		int (*pre_set_power_state)(struct radeon_device *rdev);
 		int (*set_power_state)(struct radeon_device *rdev);
@@ -2650,6 +2661,7 @@ void radeon_ring_write(struct radeon_ring *ring, uint32_t v);
 #define radeon_dpm_init(rdev) rdev->asic->dpm.init((rdev))
 #define radeon_dpm_setup_asic(rdev) rdev->asic->dpm.setup_asic((rdev))
 #define radeon_dpm_enable(rdev) rdev->asic->dpm.enable((rdev))
+#define radeon_dpm_late_enable(rdev) rdev->asic->dpm.late_enable((rdev))
 #define radeon_dpm_disable(rdev) rdev->asic->dpm.disable((rdev))
 #define radeon_dpm_pre_set_power_state(rdev) rdev->asic->dpm.pre_set_power_state((rdev))
 #define radeon_dpm_set_power_state(rdev) rdev->asic->dpm.set_power_state((rdev))
@@ -2668,6 +2680,7 @@ void radeon_ring_write(struct radeon_ring *ring, uint32_t v);
 /* Common functions */
 /* AGP */
 extern int radeon_gpu_reset(struct radeon_device *rdev);
+extern void radeon_pci_config_reset(struct radeon_device *rdev);
 extern void r600_set_bios_scratch_engine_hung(struct radeon_device *rdev, bool hung);
 extern void radeon_agp_disable(struct radeon_device *rdev);
 extern int radeon_modeset_init(struct radeon_device *rdev);
diff --git a/drivers/gpu/drm/radeon/radeon_asic.c b/drivers/gpu/drm/radeon/radeon_asic.c
index c0425bb6223a..f74db43346fd 100644
--- a/drivers/gpu/drm/radeon/radeon_asic.c
+++ b/drivers/gpu/drm/radeon/radeon_asic.c
@@ -182,9 +182,9 @@ static struct radeon_asic_ring r100_gfx_ring = {
 	.ring_test = &r100_ring_test,
 	.ib_test = &r100_ib_test,
 	.is_lockup = &r100_gpu_is_lockup,
-	.get_rptr = &radeon_ring_generic_get_rptr,
-	.get_wptr = &radeon_ring_generic_get_wptr,
-	.set_wptr = &radeon_ring_generic_set_wptr,
+	.get_rptr = &r100_gfx_get_rptr,
+	.get_wptr = &r100_gfx_get_wptr,
+	.set_wptr = &r100_gfx_set_wptr,
 };
 
 static struct radeon_asic r100_asic = {
@@ -330,9 +330,9 @@ static struct radeon_asic_ring r300_gfx_ring = {
 	.ring_test = &r100_ring_test,
 	.ib_test = &r100_ib_test,
 	.is_lockup = &r100_gpu_is_lockup,
-	.get_rptr = &radeon_ring_generic_get_rptr,
-	.get_wptr = &radeon_ring_generic_get_wptr,
-	.set_wptr = &radeon_ring_generic_set_wptr,
+	.get_rptr = &r100_gfx_get_rptr,
+	.get_wptr = &r100_gfx_get_wptr,
+	.set_wptr = &r100_gfx_set_wptr,
 };
 
 static struct radeon_asic r300_asic = {
@@ -883,9 +883,9 @@ static struct radeon_asic_ring r600_gfx_ring = {
 	.ring_test = &r600_ring_test,
 	.ib_test = &r600_ib_test,
 	.is_lockup = &r600_gfx_is_lockup,
-	.get_rptr = &radeon_ring_generic_get_rptr,
-	.get_wptr = &radeon_ring_generic_get_wptr,
-	.set_wptr = &radeon_ring_generic_set_wptr,
+	.get_rptr = &r600_gfx_get_rptr,
+	.get_wptr = &r600_gfx_get_wptr,
+	.set_wptr = &r600_gfx_set_wptr,
 };
 
 static struct radeon_asic_ring r600_dma_ring = {
@@ -1045,6 +1045,7 @@ static struct radeon_asic rv6xx_asic = {
 		.init = &rv6xx_dpm_init,
 		.setup_asic = &rv6xx_setup_asic,
 		.enable = &rv6xx_dpm_enable,
+		.late_enable = &r600_dpm_late_enable,
 		.disable = &rv6xx_dpm_disable,
 		.pre_set_power_state = &r600_dpm_pre_set_power_state,
 		.set_power_state = &rv6xx_dpm_set_power_state,
@@ -1135,6 +1136,7 @@ static struct radeon_asic rs780_asic = {
 		.init = &rs780_dpm_init,
 		.setup_asic = &rs780_dpm_setup_asic,
 		.enable = &rs780_dpm_enable,
+		.late_enable = &r600_dpm_late_enable,
 		.disable = &rs780_dpm_disable,
 		.pre_set_power_state = &r600_dpm_pre_set_power_state,
 		.set_power_state = &rs780_dpm_set_power_state,
@@ -1239,6 +1241,7 @@ static struct radeon_asic rv770_asic = {
 		.init = &rv770_dpm_init,
 		.setup_asic = &rv770_dpm_setup_asic,
 		.enable = &rv770_dpm_enable,
+		.late_enable = &rv770_dpm_late_enable,
 		.disable = &rv770_dpm_disable,
 		.pre_set_power_state = &r600_dpm_pre_set_power_state,
 		.set_power_state = &rv770_dpm_set_power_state,
@@ -1267,9 +1270,9 @@ static struct radeon_asic_ring evergreen_gfx_ring = {
 	.ring_test = &r600_ring_test,
 	.ib_test = &r600_ib_test,
 	.is_lockup = &evergreen_gfx_is_lockup,
-	.get_rptr = &radeon_ring_generic_get_rptr,
-	.get_wptr = &radeon_ring_generic_get_wptr,
-	.set_wptr = &radeon_ring_generic_set_wptr,
+	.get_rptr = &r600_gfx_get_rptr,
+	.get_wptr = &r600_gfx_get_wptr,
+	.set_wptr = &r600_gfx_set_wptr,
 };
 
 static struct radeon_asic_ring evergreen_dma_ring = {
@@ -1357,6 +1360,7 @@ static struct radeon_asic evergreen_asic = {
 		.init = &cypress_dpm_init,
 		.setup_asic = &cypress_dpm_setup_asic,
 		.enable = &cypress_dpm_enable,
+		.late_enable = &rv770_dpm_late_enable,
 		.disable = &cypress_dpm_disable,
 		.pre_set_power_state = &r600_dpm_pre_set_power_state,
 		.set_power_state = &cypress_dpm_set_power_state,
@@ -1449,6 +1453,7 @@ static struct radeon_asic sumo_asic = {
 		.init = &sumo_dpm_init,
 		.setup_asic = &sumo_dpm_setup_asic,
 		.enable = &sumo_dpm_enable,
+		.late_enable = &sumo_dpm_late_enable,
 		.disable = &sumo_dpm_disable,
 		.pre_set_power_state = &sumo_dpm_pre_set_power_state,
 		.set_power_state = &sumo_dpm_set_power_state,
@@ -1540,6 +1545,7 @@ static struct radeon_asic btc_asic = {
 		.init = &btc_dpm_init,
 		.setup_asic = &btc_dpm_setup_asic,
 		.enable = &btc_dpm_enable,
+		.late_enable = &rv770_dpm_late_enable,
 		.disable = &btc_dpm_disable,
 		.pre_set_power_state = &btc_dpm_pre_set_power_state,
 		.set_power_state = &btc_dpm_set_power_state,
@@ -1570,9 +1576,9 @@ static struct radeon_asic_ring cayman_gfx_ring = {
 	.ib_test = &r600_ib_test,
 	.is_lockup = &cayman_gfx_is_lockup,
 	.vm_flush = &cayman_vm_flush,
-	.get_rptr = &radeon_ring_generic_get_rptr,
-	.get_wptr = &radeon_ring_generic_get_wptr,
-	.set_wptr = &radeon_ring_generic_set_wptr,
+	.get_rptr = &cayman_gfx_get_rptr,
+	.get_wptr = &cayman_gfx_get_wptr,
+	.set_wptr = &cayman_gfx_set_wptr,
 };
 
 static struct radeon_asic_ring cayman_dma_ring = {
@@ -1585,9 +1591,9 @@ static struct radeon_asic_ring cayman_dma_ring = {
 	.ib_test = &r600_dma_ib_test,
 	.is_lockup = &cayman_dma_is_lockup,
 	.vm_flush = &cayman_dma_vm_flush,
-	.get_rptr = &r600_dma_get_rptr,
-	.get_wptr = &r600_dma_get_wptr,
-	.set_wptr = &r600_dma_set_wptr
+	.get_rptr = &cayman_dma_get_rptr,
+	.get_wptr = &cayman_dma_get_wptr,
+	.set_wptr = &cayman_dma_set_wptr
 };
 
 static struct radeon_asic_ring cayman_uvd_ring = {
@@ -1683,6 +1689,7 @@ static struct radeon_asic cayman_asic = {
 		.init = &ni_dpm_init,
 		.setup_asic = &ni_dpm_setup_asic,
 		.enable = &ni_dpm_enable,
+		.late_enable = &rv770_dpm_late_enable,
 		.disable = &ni_dpm_disable,
 		.pre_set_power_state = &ni_dpm_pre_set_power_state,
 		.set_power_state = &ni_dpm_set_power_state,
@@ -1783,6 +1790,7 @@ static struct radeon_asic trinity_asic = {
 		.init = &trinity_dpm_init,
 		.setup_asic = &trinity_dpm_setup_asic,
 		.enable = &trinity_dpm_enable,
+		.late_enable = &trinity_dpm_late_enable,
 		.disable = &trinity_dpm_disable,
 		.pre_set_power_state = &trinity_dpm_pre_set_power_state,
 		.set_power_state = &trinity_dpm_set_power_state,
@@ -1813,9 +1821,9 @@ static struct radeon_asic_ring si_gfx_ring = {
 	.ib_test = &r600_ib_test,
 	.is_lockup = &si_gfx_is_lockup,
 	.vm_flush = &si_vm_flush,
-	.get_rptr = &radeon_ring_generic_get_rptr,
-	.get_wptr = &radeon_ring_generic_get_wptr,
-	.set_wptr = &radeon_ring_generic_set_wptr,
+	.get_rptr = &cayman_gfx_get_rptr,
+	.get_wptr = &cayman_gfx_get_wptr,
+	.set_wptr = &cayman_gfx_set_wptr,
 };
 
 static struct radeon_asic_ring si_dma_ring = {
@@ -1828,9 +1836,9 @@ static struct radeon_asic_ring si_dma_ring = {
 	.ib_test = &r600_dma_ib_test,
 	.is_lockup = &si_dma_is_lockup,
 	.vm_flush = &si_dma_vm_flush,
-	.get_rptr = &r600_dma_get_rptr,
-	.get_wptr = &r600_dma_get_wptr,
-	.set_wptr = &r600_dma_set_wptr,
+	.get_rptr = &cayman_dma_get_rptr,
+	.get_wptr = &cayman_dma_get_wptr,
+	.set_wptr = &cayman_dma_set_wptr,
 };
 
 static struct radeon_asic si_asic = {
@@ -1913,6 +1921,7 @@ static struct radeon_asic si_asic = {
 		.init = &si_dpm_init,
 		.setup_asic = &si_dpm_setup_asic,
 		.enable = &si_dpm_enable,
+		.late_enable = &si_dpm_late_enable,
 		.disable = &si_dpm_disable,
 		.pre_set_power_state = &si_dpm_pre_set_power_state,
 		.set_power_state = &si_dpm_set_power_state,
@@ -1943,9 +1952,9 @@ static struct radeon_asic_ring ci_gfx_ring = {
 	.ib_test = &cik_ib_test,
 	.is_lockup = &cik_gfx_is_lockup,
 	.vm_flush = &cik_vm_flush,
-	.get_rptr = &radeon_ring_generic_get_rptr,
-	.get_wptr = &radeon_ring_generic_get_wptr,
-	.set_wptr = &radeon_ring_generic_set_wptr,
+	.get_rptr = &cik_gfx_get_rptr,
+	.get_wptr = &cik_gfx_get_wptr,
+	.set_wptr = &cik_gfx_set_wptr,
 };
 
 static struct radeon_asic_ring ci_cp_ring = {
@@ -1958,9 +1967,9 @@ static struct radeon_asic_ring ci_cp_ring = {
 	.ib_test = &cik_ib_test,
 	.is_lockup = &cik_gfx_is_lockup,
 	.vm_flush = &cik_vm_flush,
-	.get_rptr = &cik_compute_ring_get_rptr,
-	.get_wptr = &cik_compute_ring_get_wptr,
-	.set_wptr = &cik_compute_ring_set_wptr,
+	.get_rptr = &cik_compute_get_rptr,
+	.get_wptr = &cik_compute_get_wptr,
+	.set_wptr = &cik_compute_set_wptr,
 };
 
 static struct radeon_asic_ring ci_dma_ring = {
@@ -1973,9 +1982,9 @@ static struct radeon_asic_ring ci_dma_ring = {
 	.ib_test = &cik_sdma_ib_test,
 	.is_lockup = &cik_sdma_is_lockup,
 	.vm_flush = &cik_dma_vm_flush,
-	.get_rptr = &r600_dma_get_rptr,
-	.get_wptr = &r600_dma_get_wptr,
-	.set_wptr = &r600_dma_set_wptr,
+	.get_rptr = &cik_sdma_get_rptr,
+	.get_wptr = &cik_sdma_get_wptr,
+	.set_wptr = &cik_sdma_set_wptr,
 };
 
 static struct radeon_asic ci_asic = {
@@ -2058,6 +2067,7 @@ static struct radeon_asic ci_asic = {
 		.init = &ci_dpm_init,
 		.setup_asic = &ci_dpm_setup_asic,
 		.enable = &ci_dpm_enable,
+		.late_enable = &ci_dpm_late_enable,
 		.disable = &ci_dpm_disable,
 		.pre_set_power_state = &ci_dpm_pre_set_power_state,
 		.set_power_state = &ci_dpm_set_power_state,
@@ -2159,6 +2169,7 @@ static struct radeon_asic kv_asic = {
 		.init = &kv_dpm_init,
 		.setup_asic = &kv_dpm_setup_asic,
 		.enable = &kv_dpm_enable,
+		.late_enable = &kv_dpm_late_enable,
 		.disable = &kv_dpm_disable,
 		.pre_set_power_state = &kv_dpm_pre_set_power_state,
 		.set_power_state = &kv_dpm_set_power_state,
@@ -2449,7 +2460,7 @@ int radeon_asic_init(struct radeon_device *rdev)
 			rdev->cg_flags =
 				RADEON_CG_SUPPORT_GFX_MGCG |
 				RADEON_CG_SUPPORT_GFX_MGLS |
-				/*RADEON_CG_SUPPORT_GFX_CGCG |*/
+				RADEON_CG_SUPPORT_GFX_CGCG |
 				RADEON_CG_SUPPORT_GFX_CGLS |
 				RADEON_CG_SUPPORT_GFX_CGTS |
 				RADEON_CG_SUPPORT_GFX_CGTS_LS |
@@ -2468,7 +2479,7 @@ int radeon_asic_init(struct radeon_device *rdev)
 			rdev->cg_flags =
 				RADEON_CG_SUPPORT_GFX_MGCG |
 				RADEON_CG_SUPPORT_GFX_MGLS |
-				/*RADEON_CG_SUPPORT_GFX_CGCG |*/
+				RADEON_CG_SUPPORT_GFX_CGCG |
 				RADEON_CG_SUPPORT_GFX_CGLS |
 				RADEON_CG_SUPPORT_GFX_CGTS |
 				RADEON_CG_SUPPORT_GFX_CP_LS |
@@ -2493,7 +2504,7 @@ int radeon_asic_init(struct radeon_device *rdev)
 			rdev->cg_flags =
 				RADEON_CG_SUPPORT_GFX_MGCG |
 				RADEON_CG_SUPPORT_GFX_MGLS |
-				/*RADEON_CG_SUPPORT_GFX_CGCG |*/
+				RADEON_CG_SUPPORT_GFX_CGCG |
 				RADEON_CG_SUPPORT_GFX_CGLS |
 				RADEON_CG_SUPPORT_GFX_CGTS |
 				RADEON_CG_SUPPORT_GFX_CGTS_LS |
@@ -2521,7 +2532,7 @@ int radeon_asic_init(struct radeon_device *rdev)
 			rdev->cg_flags =
 				RADEON_CG_SUPPORT_GFX_MGCG |
 				RADEON_CG_SUPPORT_GFX_MGLS |
-				/*RADEON_CG_SUPPORT_GFX_CGCG |*/
+				RADEON_CG_SUPPORT_GFX_CGCG |
 				RADEON_CG_SUPPORT_GFX_CGLS |
 				RADEON_CG_SUPPORT_GFX_CGTS |
 				RADEON_CG_SUPPORT_GFX_CGTS_LS |
diff --git a/drivers/gpu/drm/radeon/radeon_asic.h b/drivers/gpu/drm/radeon/radeon_asic.h
index c9fd97b58076..b3bc433eed4c 100644
--- a/drivers/gpu/drm/radeon/radeon_asic.h
+++ b/drivers/gpu/drm/radeon/radeon_asic.h
@@ -47,13 +47,6 @@ u8 atombios_get_backlight_level(struct radeon_encoder *radeon_encoder);
 void radeon_legacy_set_backlight_level(struct radeon_encoder *radeon_encoder, u8 level);
 u8 radeon_legacy_get_backlight_level(struct radeon_encoder *radeon_encoder);
 
-u32 radeon_ring_generic_get_rptr(struct radeon_device *rdev,
-				 struct radeon_ring *ring);
-u32 radeon_ring_generic_get_wptr(struct radeon_device *rdev,
-				 struct radeon_ring *ring);
-void radeon_ring_generic_set_wptr(struct radeon_device *rdev,
-				  struct radeon_ring *ring);
-
 /*
  * r100,rv100,rs100,rv200,rs200
  */
@@ -148,6 +141,13 @@ extern void r100_post_page_flip(struct radeon_device *rdev, int crtc);
 extern void r100_wait_for_vblank(struct radeon_device *rdev, int crtc);
 extern int r100_mc_wait_for_idle(struct radeon_device *rdev);
 
+u32 r100_gfx_get_rptr(struct radeon_device *rdev,
+		      struct radeon_ring *ring);
+u32 r100_gfx_get_wptr(struct radeon_device *rdev,
+		      struct radeon_ring *ring);
+void r100_gfx_set_wptr(struct radeon_device *rdev,
+		       struct radeon_ring *ring);
+
 /*
  * r200,rv250,rs300,rv280
  */
@@ -368,6 +368,12 @@ int r600_mc_wait_for_idle(struct radeon_device *rdev);
 int r600_pcie_gart_init(struct radeon_device *rdev);
 void r600_scratch_init(struct radeon_device *rdev);
 int r600_init_microcode(struct radeon_device *rdev);
+u32 r600_gfx_get_rptr(struct radeon_device *rdev,
+		      struct radeon_ring *ring);
+u32 r600_gfx_get_wptr(struct radeon_device *rdev,
+		      struct radeon_ring *ring);
+void r600_gfx_set_wptr(struct radeon_device *rdev,
+		       struct radeon_ring *ring);
 /* r600 irq */
 int r600_irq_process(struct radeon_device *rdev);
 int r600_irq_init(struct radeon_device *rdev);
@@ -392,6 +398,7 @@ int rv6xx_get_temp(struct radeon_device *rdev);
 int r600_set_uvd_clocks(struct radeon_device *rdev, u32 vclk, u32 dclk);
 int r600_dpm_pre_set_power_state(struct radeon_device *rdev);
 void r600_dpm_post_set_power_state(struct radeon_device *rdev);
+int r600_dpm_late_enable(struct radeon_device *rdev);
 /* r600 dma */
 uint32_t r600_dma_get_rptr(struct radeon_device *rdev,
 			   struct radeon_ring *ring);
@@ -454,6 +461,7 @@ int rv770_get_temp(struct radeon_device *rdev);
 /* rv7xx pm */
 int rv770_dpm_init(struct radeon_device *rdev);
 int rv770_dpm_enable(struct radeon_device *rdev);
+int rv770_dpm_late_enable(struct radeon_device *rdev);
 void rv770_dpm_disable(struct radeon_device *rdev);
 int rv770_dpm_set_power_state(struct radeon_device *rdev);
 void rv770_dpm_setup_asic(struct radeon_device *rdev);
@@ -545,6 +553,7 @@ u32 btc_dpm_get_mclk(struct radeon_device *rdev, bool low);
 bool btc_dpm_vblank_too_short(struct radeon_device *rdev);
 int sumo_dpm_init(struct radeon_device *rdev);
 int sumo_dpm_enable(struct radeon_device *rdev);
+int sumo_dpm_late_enable(struct radeon_device *rdev);
 void sumo_dpm_disable(struct radeon_device *rdev);
 int sumo_dpm_pre_set_power_state(struct radeon_device *rdev);
 int sumo_dpm_set_power_state(struct radeon_device *rdev);
@@ -591,6 +600,19 @@ void cayman_dma_vm_set_page(struct radeon_device *rdev,
 
 void cayman_dma_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm);
 
+u32 cayman_gfx_get_rptr(struct radeon_device *rdev,
+			struct radeon_ring *ring);
+u32 cayman_gfx_get_wptr(struct radeon_device *rdev,
+			struct radeon_ring *ring);
+void cayman_gfx_set_wptr(struct radeon_device *rdev,
+			 struct radeon_ring *ring);
+uint32_t cayman_dma_get_rptr(struct radeon_device *rdev,
+			     struct radeon_ring *ring);
+uint32_t cayman_dma_get_wptr(struct radeon_device *rdev,
+			     struct radeon_ring *ring);
+void cayman_dma_set_wptr(struct radeon_device *rdev,
+			 struct radeon_ring *ring);
+
 int ni_dpm_init(struct radeon_device *rdev);
 void ni_dpm_setup_asic(struct radeon_device *rdev);
 int ni_dpm_enable(struct radeon_device *rdev);
@@ -610,6 +632,7 @@ int ni_dpm_force_performance_level(struct radeon_device *rdev,
 bool ni_dpm_vblank_too_short(struct radeon_device *rdev);
 int trinity_dpm_init(struct radeon_device *rdev);
 int trinity_dpm_enable(struct radeon_device *rdev);
+int trinity_dpm_late_enable(struct radeon_device *rdev);
 void trinity_dpm_disable(struct radeon_device *rdev);
 int trinity_dpm_pre_set_power_state(struct radeon_device *rdev);
 int trinity_dpm_set_power_state(struct radeon_device *rdev);
@@ -669,6 +692,7 @@ int si_get_temp(struct radeon_device *rdev);
 int si_dpm_init(struct radeon_device *rdev);
 void si_dpm_setup_asic(struct radeon_device *rdev);
 int si_dpm_enable(struct radeon_device *rdev);
+int si_dpm_late_enable(struct radeon_device *rdev);
 void si_dpm_disable(struct radeon_device *rdev);
 int si_dpm_pre_set_power_state(struct radeon_device *rdev);
 int si_dpm_set_power_state(struct radeon_device *rdev);
@@ -739,17 +763,30 @@ void cik_sdma_vm_set_page(struct radeon_device *rdev,
 			  uint32_t incr, uint32_t flags);
 void cik_dma_vm_flush(struct radeon_device *rdev, int ridx, struct radeon_vm *vm);
 int cik_ib_parse(struct radeon_device *rdev, struct radeon_ib *ib);
-u32 cik_compute_ring_get_rptr(struct radeon_device *rdev,
-			      struct radeon_ring *ring);
-u32 cik_compute_ring_get_wptr(struct radeon_device *rdev,
-			      struct radeon_ring *ring);
-void cik_compute_ring_set_wptr(struct radeon_device *rdev,
-			       struct radeon_ring *ring);
+u32 cik_gfx_get_rptr(struct radeon_device *rdev,
+		     struct radeon_ring *ring);
+u32 cik_gfx_get_wptr(struct radeon_device *rdev,
+		     struct radeon_ring *ring);
+void cik_gfx_set_wptr(struct radeon_device *rdev,
+		      struct radeon_ring *ring);
+u32 cik_compute_get_rptr(struct radeon_device *rdev,
+			 struct radeon_ring *ring);
+u32 cik_compute_get_wptr(struct radeon_device *rdev,
+			 struct radeon_ring *ring);
+void cik_compute_set_wptr(struct radeon_device *rdev,
+			  struct radeon_ring *ring);
+u32 cik_sdma_get_rptr(struct radeon_device *rdev,
+		      struct radeon_ring *ring);
+u32 cik_sdma_get_wptr(struct radeon_device *rdev,
+		      struct radeon_ring *ring);
+void cik_sdma_set_wptr(struct radeon_device *rdev,
+		       struct radeon_ring *ring);
 int ci_get_temp(struct radeon_device *rdev);
 int kv_get_temp(struct radeon_device *rdev);
 
 int ci_dpm_init(struct radeon_device *rdev);
 int ci_dpm_enable(struct radeon_device *rdev);
+int ci_dpm_late_enable(struct radeon_device *rdev);
 void ci_dpm_disable(struct radeon_device *rdev);
 int ci_dpm_pre_set_power_state(struct radeon_device *rdev);
 int ci_dpm_set_power_state(struct radeon_device *rdev);
@@ -770,6 +807,7 @@ void ci_dpm_powergate_uvd(struct radeon_device *rdev, bool gate);
 
 int kv_dpm_init(struct radeon_device *rdev);
 int kv_dpm_enable(struct radeon_device *rdev);
+int kv_dpm_late_enable(struct radeon_device *rdev);
 void kv_dpm_disable(struct radeon_device *rdev);
 int kv_dpm_pre_set_power_state(struct radeon_device *rdev);
 int kv_dpm_set_power_state(struct radeon_device *rdev);
diff --git a/drivers/gpu/drm/radeon/radeon_atombios.c b/drivers/gpu/drm/radeon/radeon_atombios.c
index 5c39bf7c3d88..30844814c25a 100644
--- a/drivers/gpu/drm/radeon/radeon_atombios.c
+++ b/drivers/gpu/drm/radeon/radeon_atombios.c
@@ -30,27 +30,10 @@
 #include "atom.h"
 #include "atom-bits.h"
 
-/* from radeon_encoder.c */
-extern uint32_t
-radeon_get_encoder_enum(struct drm_device *dev, uint32_t supported_device,
-			uint8_t dac);
-extern void radeon_link_encoder_connector(struct drm_device *dev);
 extern void
 radeon_add_atom_encoder(struct drm_device *dev, uint32_t encoder_enum,
 			uint32_t supported_device, u16 caps);
 
-/* from radeon_connector.c */
-extern void
-radeon_add_atom_connector(struct drm_device *dev,
-			  uint32_t connector_id,
-			  uint32_t supported_device,
-			  int connector_type,
-			  struct radeon_i2c_bus_rec *i2c_bus,
-			  uint32_t igp_lane_info,
-			  uint16_t connector_object_id,
-			  struct radeon_hpd *hpd,
-			  struct radeon_router *router);
-
 /* from radeon_legacy_encoder.c */
 extern void
 radeon_add_legacy_encoder(struct drm_device *dev, uint32_t encoder_enum,
@@ -1528,6 +1511,7 @@ bool radeon_atombios_get_asic_ss_info(struct radeon_device *rdev,
 						le16_to_cpu(ss_assign->v1.usSpreadSpectrumPercentage);
 					ss->type = ss_assign->v1.ucSpreadSpectrumMode;
 					ss->rate = le16_to_cpu(ss_assign->v1.usSpreadRateInKhz);
+					ss->percentage_divider = 100;
 					return true;
 				}
 				ss_assign = (union asic_ss_assignment *)
@@ -1545,6 +1529,7 @@ bool radeon_atombios_get_asic_ss_info(struct radeon_device *rdev,
 						le16_to_cpu(ss_assign->v2.usSpreadSpectrumPercentage);
 					ss->type = ss_assign->v2.ucSpreadSpectrumMode;
 					ss->rate = le16_to_cpu(ss_assign->v2.usSpreadRateIn10Hz);
+					ss->percentage_divider = 100;
 					if ((crev == 2) &&
 					    ((id == ASIC_INTERNAL_ENGINE_SS) ||
 					     (id == ASIC_INTERNAL_MEMORY_SS)))
@@ -1566,6 +1551,11 @@ bool radeon_atombios_get_asic_ss_info(struct radeon_device *rdev,
 						le16_to_cpu(ss_assign->v3.usSpreadSpectrumPercentage);
 					ss->type = ss_assign->v3.ucSpreadSpectrumMode;
 					ss->rate = le16_to_cpu(ss_assign->v3.usSpreadRateIn10Hz);
+					if (ss_assign->v3.ucSpreadSpectrumMode &
+					    SS_MODE_V3_PERCENTAGE_DIV_BY_1000_MASK)
+						ss->percentage_divider = 1000;
+					else
+						ss->percentage_divider = 100;
 					if ((id == ASIC_INTERNAL_ENGINE_SS) ||
 					    (id == ASIC_INTERNAL_MEMORY_SS))
 						ss->rate /= 100;
@@ -1809,7 +1799,8 @@ bool radeon_atom_get_tv_timings(struct radeon_device *rdev, int index,
 		if (misc & ATOM_DOUBLE_CLOCK_MODE)
 			mode->flags |= DRM_MODE_FLAG_DBLSCAN;
 
-		mode->clock = le16_to_cpu(tv_info->aModeTimings[index].usPixelClock) * 10;
+		mode->crtc_clock = mode->clock =
+			le16_to_cpu(tv_info->aModeTimings[index].usPixelClock) * 10;
 
 		if (index == 1) {
 			/* PAL timings appear to have wrong values for totals */
@@ -1852,7 +1843,8 @@ bool radeon_atom_get_tv_timings(struct radeon_device *rdev, int index,
 		if (misc & ATOM_DOUBLE_CLOCK_MODE)
 			mode->flags |= DRM_MODE_FLAG_DBLSCAN;
 
-		mode->clock = le16_to_cpu(dtd_timings->usPixClk) * 10;
+		mode->crtc_clock = mode->clock =
+			le16_to_cpu(dtd_timings->usPixClk) * 10;
 		break;
 	}
 	return true;
@@ -3884,16 +3876,18 @@ int radeon_atom_init_mc_reg_table(struct radeon_device *rdev,
 							((u8 *)format + sizeof(ATOM_INIT_REG_INDEX_FORMAT));
 					}
 					reg_table->last = i;
-					while ((*(u32 *)reg_data != END_OF_REG_DATA_BLOCK) &&
+					while ((le32_to_cpu(*(u32 *)reg_data) != END_OF_REG_DATA_BLOCK) &&
 					       (num_ranges < VBIOS_MAX_AC_TIMING_ENTRIES)) {
-						t_mem_id = (u8)((*(u32 *)reg_data & MEM_ID_MASK) >> MEM_ID_SHIFT);
+						t_mem_id = (u8)((le32_to_cpu(*(u32 *)reg_data) & MEM_ID_MASK)
+								>> MEM_ID_SHIFT);
 						if (module_index == t_mem_id) {
 							reg_table->mc_reg_table_entry[num_ranges].mclk_max =
-								(u32)((*(u32 *)reg_data & CLOCK_RANGE_MASK) >> CLOCK_RANGE_SHIFT);
+								(u32)((le32_to_cpu(*(u32 *)reg_data) & CLOCK_RANGE_MASK)
+								      >> CLOCK_RANGE_SHIFT);
 							for (i = 0, j = 1; i < reg_table->last; i++) {
 								if ((reg_table->mc_reg_address[i].pre_reg_data & LOW_NIBBLE_MASK) == DATA_FROM_TABLE) {
 									reg_table->mc_reg_table_entry[num_ranges].mc_data[i] =
-										(u32)*((u32 *)reg_data + j);
+										(u32)le32_to_cpu(*((u32 *)reg_data + j));
 									j++;
 								} else if ((reg_table->mc_reg_address[i].pre_reg_data & LOW_NIBBLE_MASK) == DATA_EQU_PREV) {
 									reg_table->mc_reg_table_entry[num_ranges].mc_data[i] =
@@ -3905,7 +3899,7 @@ int radeon_atom_init_mc_reg_table(struct radeon_device *rdev,
 						reg_data = (ATOM_MEMORY_SETTING_DATA_BLOCK *)
 							((u8 *)reg_data + le16_to_cpu(reg_block->usRegDataBlkSize));
 					}
-					if (*(u32 *)reg_data != END_OF_REG_DATA_BLOCK)
+					if (le32_to_cpu(*(u32 *)reg_data) != END_OF_REG_DATA_BLOCK)
 						return -EINVAL;
 					reg_table->num_entries = num_ranges;
 				} else
@@ -3944,6 +3938,10 @@ void radeon_atom_initialize_bios_scratch_regs(struct drm_device *dev)
 	/* tell the bios not to handle mode switching */
 	bios_6_scratch |= ATOM_S6_ACC_BLOCK_DISPLAY_SWITCH;
 
+	/* clear the vbios dpms state */
+	if (ASIC_IS_DCE4(rdev))
+		bios_2_scratch &= ~ATOM_S2_DEVICE_DPMS_STATE;
+
 	if (rdev->family >= CHIP_R600) {
 		WREG32(R600_BIOS_2_SCRATCH, bios_2_scratch);
 		WREG32(R600_BIOS_6_SCRATCH, bios_6_scratch);
diff --git a/drivers/gpu/drm/radeon/radeon_combios.c b/drivers/gpu/drm/radeon/radeon_combios.c
index 68ce36056019..6651177110f0 100644
--- a/drivers/gpu/drm/radeon/radeon_combios.c
+++ b/drivers/gpu/drm/radeon/radeon_combios.c
@@ -37,22 +37,6 @@
 #include <asm/pci-bridge.h>
 #endif /* CONFIG_PPC_PMAC */
 
-/* from radeon_encoder.c */
-extern uint32_t
-radeon_get_encoder_enum(struct drm_device *dev, uint32_t supported_device,
-			uint8_t dac);
-extern void radeon_link_encoder_connector(struct drm_device *dev);
-
-/* from radeon_connector.c */
-extern void
-radeon_add_legacy_connector(struct drm_device *dev,
-			    uint32_t connector_id,
-			    uint32_t supported_device,
-			    int connector_type,
-			    struct radeon_i2c_bus_rec *i2c_bus,
-			    uint16_t connector_object_id,
-			    struct radeon_hpd *hpd);
-
 /* from radeon_legacy_encoder.c */
 extern void
 radeon_add_legacy_encoder(struct drm_device *dev, uint32_t encoder_enum,
diff --git a/drivers/gpu/drm/radeon/radeon_connectors.c b/drivers/gpu/drm/radeon/radeon_connectors.c
index 20a768ac89a8..82d4f865546e 100644
--- a/drivers/gpu/drm/radeon/radeon_connectors.c
+++ b/drivers/gpu/drm/radeon/radeon_connectors.c
@@ -33,15 +33,6 @@
 
 #include <linux/pm_runtime.h>
 
-extern void
-radeon_combios_connected_scratch_regs(struct drm_connector *connector,
-				      struct drm_encoder *encoder,
-				      bool connected);
-extern void
-radeon_atombios_connected_scratch_regs(struct drm_connector *connector,
-				       struct drm_encoder *encoder,
-				       bool connected);
-
 void radeon_connector_hotplug(struct drm_connector *connector)
 {
 	struct drm_device *dev = connector->dev;
diff --git a/drivers/gpu/drm/radeon/radeon_cp.c b/drivers/gpu/drm/radeon/radeon_cp.c
index 3cae2bbc1854..bb0d5c3a8311 100644
--- a/drivers/gpu/drm/radeon/radeon_cp.c
+++ b/drivers/gpu/drm/radeon/radeon_cp.c
@@ -2020,10 +2020,10 @@ static int radeon_cp_get_buffers(struct drm_device *dev,
 
 		buf->file_priv = file_priv;
 
-		if (DRM_COPY_TO_USER(&d->request_indices[i], &buf->idx,
+		if (copy_to_user(&d->request_indices[i], &buf->idx,
 				     sizeof(buf->idx)))
 			return -EFAULT;
-		if (DRM_COPY_TO_USER(&d->request_sizes[i], &buf->total,
+		if (copy_to_user(&d->request_sizes[i], &buf->total,
 				     sizeof(buf->total)))
 			return -EFAULT;
 
@@ -2228,7 +2228,7 @@ void radeon_commit_ring(drm_radeon_private_t *dev_priv)
 
 	dev_priv->ring.tail &= dev_priv->ring.tail_mask;
 
-	DRM_MEMORYBARRIER();
+	mb();
 	GET_RING_HEAD( dev_priv );
 
 	if ((dev_priv->flags & RADEON_FAMILY_MASK) >= CHIP_R600) {
diff --git a/drivers/gpu/drm/radeon/radeon_cs.c b/drivers/gpu/drm/radeon/radeon_cs.c
index 0b366169d64d..dfb5a1db87d4 100644
--- a/drivers/gpu/drm/radeon/radeon_cs.c
+++ b/drivers/gpu/drm/radeon/radeon_cs.c
@@ -138,7 +138,7 @@ static int radeon_cs_get_ring(struct radeon_cs_parser *p, u32 ring, s32 priority
 				p->ring = R600_RING_TYPE_DMA_INDEX;
 			else
 				p->ring = CAYMAN_RING_TYPE_DMA1_INDEX;
-		} else if (p->rdev->family >= CHIP_R600) {
+		} else if (p->rdev->family >= CHIP_RV770) {
 			p->ring = R600_RING_TYPE_DMA_INDEX;
 		} else {
 			return -EINVAL;
@@ -192,7 +192,7 @@ int radeon_cs_parser_init(struct radeon_cs_parser *p, void *data)
 		return -ENOMEM;
 	}
 	chunk_array_ptr = (uint64_t *)(unsigned long)(cs->chunks);
-	if (DRM_COPY_FROM_USER(p->chunks_array, chunk_array_ptr,
+	if (copy_from_user(p->chunks_array, chunk_array_ptr,
 			       sizeof(uint64_t)*cs->num_chunks)) {
 		return -EFAULT;
 	}
@@ -208,7 +208,7 @@ int radeon_cs_parser_init(struct radeon_cs_parser *p, void *data)
 		uint32_t __user *cdata;
 
 		chunk_ptr = (void __user*)(unsigned long)p->chunks_array[i];
-		if (DRM_COPY_FROM_USER(&user_chunk, chunk_ptr,
+		if (copy_from_user(&user_chunk, chunk_ptr,
 				       sizeof(struct drm_radeon_cs_chunk))) {
 			return -EFAULT;
 		}
@@ -252,7 +252,7 @@ int radeon_cs_parser_init(struct radeon_cs_parser *p, void *data)
 		if (p->chunks[i].kdata == NULL) {
 			return -ENOMEM;
 		}
-		if (DRM_COPY_FROM_USER(p->chunks[i].kdata, cdata, size)) {
+		if (copy_from_user(p->chunks[i].kdata, cdata, size)) {
 			return -EFAULT;
 		}
 		if (p->chunks[i].chunk_id == RADEON_CHUNK_ID_FLAGS) {
@@ -472,7 +472,7 @@ static int radeon_cs_ib_fill(struct radeon_device *rdev, struct radeon_cs_parser
 			}
 			parser->const_ib.is_const_ib = true;
 			parser->const_ib.length_dw = ib_chunk->length_dw;
-			if (DRM_COPY_FROM_USER(parser->const_ib.ptr,
+			if (copy_from_user(parser->const_ib.ptr,
 					       ib_chunk->user_ptr,
 					       ib_chunk->length_dw * 4))
 				return -EFAULT;
@@ -495,7 +495,7 @@ static int radeon_cs_ib_fill(struct radeon_device *rdev, struct radeon_cs_parser
 	parser->ib.length_dw = ib_chunk->length_dw;
 	if (ib_chunk->kdata)
 		memcpy(parser->ib.ptr, ib_chunk->kdata, ib_chunk->length_dw * 4);
-	else if (DRM_COPY_FROM_USER(parser->ib.ptr, ib_chunk->user_ptr, ib_chunk->length_dw * 4))
+	else if (copy_from_user(parser->ib.ptr, ib_chunk->user_ptr, ib_chunk->length_dw * 4))
 		return -EFAULT;
 	return 0;
 }
diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c
index 39b033b441d2..b012cbbc3ed5 100644
--- a/drivers/gpu/drm/radeon/radeon_device.c
+++ b/drivers/gpu/drm/radeon/radeon_device.c
@@ -144,6 +144,11 @@ void radeon_program_register_sequence(struct radeon_device *rdev,
 	}
 }
 
+void radeon_pci_config_reset(struct radeon_device *rdev)
+{
+	pci_write_config_dword(rdev->pdev, 0x7c, RADEON_ASIC_RESET_DATA);
+}
+
 /**
  * radeon_surface_init - Clear GPU surface registers.
  *
@@ -249,7 +254,7 @@ void radeon_scratch_free(struct radeon_device *rdev, uint32_t reg)
  * Init doorbell driver information (CIK)
  * Returns 0 on success, error on failure.
  */
-int radeon_doorbell_init(struct radeon_device *rdev)
+static int radeon_doorbell_init(struct radeon_device *rdev)
 {
 	/* doorbell bar mapping */
 	rdev->doorbell.base = pci_resource_start(rdev->pdev, 2);
@@ -278,7 +283,7 @@ int radeon_doorbell_init(struct radeon_device *rdev)
  *
  * Tear down doorbell driver information (CIK)
  */
-void radeon_doorbell_fini(struct radeon_device *rdev)
+static void radeon_doorbell_fini(struct radeon_device *rdev)
 {
 	iounmap(rdev->doorbell.ptr);
 	rdev->doorbell.ptr = NULL;
@@ -1330,6 +1335,7 @@ int radeon_device_init(struct radeon_device *rdev,
 		if (r)
 			return r;
 	}
+
 	if ((radeon_testing & 1)) {
 		if (rdev->accel_working)
 			radeon_test_moves(rdev);
@@ -1455,7 +1461,6 @@ int radeon_suspend_kms(struct drm_device *dev, bool suspend, bool fbcon)
 
 	radeon_save_bios_scratch_regs(rdev);
 
-	radeon_pm_suspend(rdev);
 	radeon_suspend(rdev);
 	radeon_hpd_fini(rdev);
 	/* evict remaining vram memory */
@@ -1516,14 +1521,22 @@ int radeon_resume_kms(struct drm_device *dev, bool resume, bool fbcon)
 	if (r)
 		DRM_ERROR("ib ring test failed (%d).\n", r);
 
-	radeon_pm_resume(rdev);
+	if (rdev->pm.dpm_enabled) {
+		/* do dpm late init */
+		r = radeon_pm_late_init(rdev);
+		if (r) {
+			rdev->pm.dpm_enabled = false;
+			DRM_ERROR("radeon_pm_late_init failed, disabling dpm\n");
+		}
+	}
+
 	radeon_restore_bios_scratch_regs(rdev);
 
 	if (fbcon) {
 		radeon_fbdev_set_suspend(rdev, 0);
 		console_unlock();
 	}
-       
+
 	/* init dig PHYs, disp eng pll */
 	if (rdev->is_atom_bios) {
 		radeon_atom_encoder_init(rdev);
diff --git a/drivers/gpu/drm/radeon/radeon_display.c b/drivers/gpu/drm/radeon/radeon_display.c
index 7b253815a323..d680608f6f5b 100644
--- a/drivers/gpu/drm/radeon/radeon_display.c
+++ b/drivers/gpu/drm/radeon/radeon_display.c
@@ -306,7 +306,7 @@ void radeon_crtc_handle_flip(struct radeon_device *rdev, int crtc_id)
 	 * to complete in this vblank?
 	 */
 	if (update_pending &&
-	    (DRM_SCANOUTPOS_VALID & radeon_get_crtc_scanoutpos(rdev->ddev, crtc_id,
+	    (DRM_SCANOUTPOS_VALID & radeon_get_crtc_scanoutpos(rdev->ddev, crtc_id, 0,
 							       &vpos, &hpos, NULL, NULL)) &&
 	    ((vpos >= (99 * rdev->mode_info.crtcs[crtc_id]->base.hwmode.crtc_vdisplay)/100) ||
 	     (vpos < 0 && !ASIC_IS_AVIVO(rdev)))) {
@@ -1464,12 +1464,22 @@ int radeon_modeset_init(struct radeon_device *rdev)
 	/* setup afmt */
 	radeon_afmt_init(rdev);
 
-	/* Initialize power management */
-	radeon_pm_init(rdev);
-
 	radeon_fbdev_init(rdev);
 	drm_kms_helper_poll_init(rdev->ddev);
 
+	if (rdev->pm.dpm_enabled) {
+		/* do dpm late init */
+		ret = radeon_pm_late_init(rdev);
+		if (ret) {
+			rdev->pm.dpm_enabled = false;
+			DRM_ERROR("radeon_pm_late_init failed, disabling dpm\n");
+		}
+		/* set the dpm state for PX since there won't be
+		 * a modeset to call this.
+		 */
+		radeon_pm_compute_clocks(rdev);
+	}
+
 	return 0;
 }
 
@@ -1477,7 +1487,6 @@ void radeon_modeset_fini(struct radeon_device *rdev)
 {
 	radeon_fbdev_fini(rdev);
 	kfree(rdev->mode_info.bios_hardcoded_edid);
-	radeon_pm_fini(rdev);
 
 	if (rdev->mode_info.mode_config_initialized) {
 		radeon_afmt_fini(rdev);
@@ -1601,6 +1610,7 @@ bool radeon_crtc_scaling_mode_fixup(struct drm_crtc *crtc,
  *
  * \param dev Device to query.
  * \param crtc Crtc to query.
+ * \param flags Flags from caller (DRM_CALLED_FROM_VBLIRQ or 0).
  * \param *vpos Location where vertical scanout position should be stored.
  * \param *hpos Location where horizontal scanout position should go.
  * \param *stime Target location for timestamp taken immediately before
@@ -1622,8 +1632,8 @@ bool radeon_crtc_scaling_mode_fixup(struct drm_crtc *crtc,
  * unknown small number of scanlines wrt. real scanout position.
  *
  */
-int radeon_get_crtc_scanoutpos(struct drm_device *dev, int crtc, int *vpos, int *hpos,
-			       ktime_t *stime, ktime_t *etime)
+int radeon_get_crtc_scanoutpos(struct drm_device *dev, int crtc, unsigned int flags,
+			       int *vpos, int *hpos, ktime_t *stime, ktime_t *etime)
 {
 	u32 stat_crtc = 0, vbl = 0, position = 0;
 	int vbl_start, vbl_end, vtotal, ret = 0;
@@ -1765,5 +1775,27 @@ int radeon_get_crtc_scanoutpos(struct drm_device *dev, int crtc, int *vpos, int
 	if (in_vbl)
 		ret |= DRM_SCANOUTPOS_INVBL;
 
+	/* Is vpos outside nominal vblank area, but less than
+	 * 1/100 of a frame height away from start of vblank?
+	 * If so, assume this isn't a massively delayed vblank
+	 * interrupt, but a vblank interrupt that fired a few
+	 * microseconds before true start of vblank. Compensate
+	 * by adding a full frame duration to the final timestamp.
+	 * Happens, e.g., on ATI R500, R600.
+	 *
+	 * We only do this if DRM_CALLED_FROM_VBLIRQ.
+	 */
+	if ((flags & DRM_CALLED_FROM_VBLIRQ) && !in_vbl) {
+		vbl_start = rdev->mode_info.crtcs[crtc]->base.hwmode.crtc_vdisplay;
+		vtotal = rdev->mode_info.crtcs[crtc]->base.hwmode.crtc_vtotal;
+
+		if (vbl_start - *vpos < vtotal / 100) {
+			*vpos -= vtotal;
+
+			/* Signal this correction as "applied". */
+			ret |= 0x8;
+		}
+	}
+
 	return ret;
 }
diff --git a/drivers/gpu/drm/radeon/radeon_drv.c b/drivers/gpu/drm/radeon/radeon_drv.c
index db39ea36bf22..ec8c388eec17 100644
--- a/drivers/gpu/drm/radeon/radeon_drv.c
+++ b/drivers/gpu/drm/radeon/radeon_drv.c
@@ -102,13 +102,14 @@ int radeon_get_vblank_timestamp_kms(struct drm_device *dev, int crtc,
 void radeon_driver_irq_preinstall_kms(struct drm_device *dev);
 int radeon_driver_irq_postinstall_kms(struct drm_device *dev);
 void radeon_driver_irq_uninstall_kms(struct drm_device *dev);
-irqreturn_t radeon_driver_irq_handler_kms(DRM_IRQ_ARGS);
+irqreturn_t radeon_driver_irq_handler_kms(int irq, void *arg);
 void radeon_gem_object_free(struct drm_gem_object *obj);
 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);
 extern int radeon_get_crtc_scanoutpos(struct drm_device *dev, int crtc,
+				      unsigned int flags,
 				      int *vpos, int *hpos, ktime_t *stime,
 				      ktime_t *etime);
 extern const struct drm_ioctl_desc radeon_ioctls_kms[];
@@ -168,6 +169,7 @@ int radeon_fastfb = 0;
 int radeon_dpm = -1;
 int radeon_aspm = -1;
 int radeon_runtime_pm = -1;
+int radeon_hard_reset = 0;
 
 MODULE_PARM_DESC(no_wb, "Disable AGP writeback for scratch registers");
 module_param_named(no_wb, radeon_no_wb, int, 0444);
@@ -232,6 +234,9 @@ module_param_named(aspm, radeon_aspm, int, 0444);
 MODULE_PARM_DESC(runpm, "PX runtime pm (1 = force enable, 0 = disable, -1 = PX only default)");
 module_param_named(runpm, radeon_runtime_pm, int, 0444);
 
+MODULE_PARM_DESC(hard_reset, "PCI config reset (1 = force enable, 0 = disable (default))");
+module_param_named(hard_reset, radeon_hard_reset, int, 0444);
+
 static struct pci_device_id pciidlist[] = {
 	radeon_PCI_IDS
 };
@@ -400,6 +405,9 @@ static int radeon_pmops_runtime_suspend(struct device *dev)
 	if (radeon_runtime_pm == 0)
 		return -EINVAL;
 
+	if (radeon_runtime_pm == -1 && !radeon_is_px())
+		return -EINVAL;
+
 	drm_dev->switch_power_state = DRM_SWITCH_POWER_CHANGING;
 	drm_kms_helper_poll_disable(drm_dev);
 	vga_switcheroo_set_dynamic_switch(pdev, VGA_SWITCHEROO_OFF);
@@ -422,6 +430,9 @@ static int radeon_pmops_runtime_resume(struct device *dev)
 	if (radeon_runtime_pm == 0)
 		return -EINVAL;
 
+	if (radeon_runtime_pm == -1 && !radeon_is_px())
+		return -EINVAL;
+
 	drm_dev->switch_power_state = DRM_SWITCH_POWER_CHANGING;
 
 	pci_set_power_state(pdev, PCI_D0);
diff --git a/drivers/gpu/drm/radeon/radeon_drv.h b/drivers/gpu/drm/radeon/radeon_drv.h
index 00e0d449021c..dafd812e4571 100644
--- a/drivers/gpu/drm/radeon/radeon_drv.h
+++ b/drivers/gpu/drm/radeon/radeon_drv.h
@@ -405,7 +405,7 @@ extern void radeon_do_release(struct drm_device * dev);
 extern u32 radeon_get_vblank_counter(struct drm_device *dev, int crtc);
 extern int radeon_enable_vblank(struct drm_device *dev, int crtc);
 extern void radeon_disable_vblank(struct drm_device *dev, int crtc);
-extern irqreturn_t radeon_driver_irq_handler(DRM_IRQ_ARGS);
+extern irqreturn_t radeon_driver_irq_handler(int irq, void *arg);
 extern void radeon_driver_irq_preinstall(struct drm_device * dev);
 extern int radeon_driver_irq_postinstall(struct drm_device *dev);
 extern void radeon_driver_irq_uninstall(struct drm_device * dev);
diff --git a/drivers/gpu/drm/radeon/radeon_fence.c b/drivers/gpu/drm/radeon/radeon_fence.c
index d3a86e43c012..c37cb79a9489 100644
--- a/drivers/gpu/drm/radeon/radeon_fence.c
+++ b/drivers/gpu/drm/radeon/radeon_fence.c
@@ -121,7 +121,7 @@ int radeon_fence_emit(struct radeon_device *rdev,
 	(*fence)->seq = ++rdev->fence_drv[ring].sync_seq[ring];
 	(*fence)->ring = ring;
 	radeon_fence_ring_emit(rdev, ring, *fence);
-	trace_radeon_fence_emit(rdev->ddev, (*fence)->seq);
+	trace_radeon_fence_emit(rdev->ddev, ring, (*fence)->seq);
 	return 0;
 }
 
@@ -313,7 +313,7 @@ static int radeon_fence_wait_seq(struct radeon_device *rdev, u64 *target_seq,
 				continue;
 
 			last_seq[i] = atomic64_read(&rdev->fence_drv[i].last_seq);
-			trace_radeon_fence_wait_begin(rdev->ddev, target_seq[i]);
+			trace_radeon_fence_wait_begin(rdev->ddev, i, target_seq[i]);
 			radeon_irq_kms_sw_irq_get(rdev, i);
 		}
 
@@ -332,7 +332,7 @@ static int radeon_fence_wait_seq(struct radeon_device *rdev, u64 *target_seq,
 				continue;
 
 			radeon_irq_kms_sw_irq_put(rdev, i);
-			trace_radeon_fence_wait_end(rdev->ddev, target_seq[i]);
+			trace_radeon_fence_wait_end(rdev->ddev, i, target_seq[i]);
 		}
 
 		if (unlikely(r < 0))
@@ -841,6 +841,8 @@ static int radeon_debugfs_fence_info(struct seq_file *m, void *data)
 		if (!rdev->fence_drv[i].initialized)
 			continue;
 
+		radeon_fence_process(rdev, i);
+
 		seq_printf(m, "--- ring %d ---\n", i);
 		seq_printf(m, "Last signaled fence 0x%016llx\n",
 			   (unsigned long long)atomic64_read(&rdev->fence_drv[i].last_seq));
diff --git a/drivers/gpu/drm/radeon/radeon_gart.c b/drivers/gpu/drm/radeon/radeon_gart.c
index 96e440061bdb..a8f9b463bf2a 100644
--- a/drivers/gpu/drm/radeon/radeon_gart.c
+++ b/drivers/gpu/drm/radeon/radeon_gart.c
@@ -713,7 +713,7 @@ struct radeon_fence *radeon_vm_grab_id(struct radeon_device *rdev,
 	unsigned i;
 
 	/* check if the id is still valid */
-	if (vm->fence && vm->fence == rdev->vm_manager.active[vm->id])
+	if (vm->last_id_use && vm->last_id_use == rdev->vm_manager.active[vm->id])
 		return NULL;
 
 	/* we definately need to flush */
@@ -726,6 +726,7 @@ struct radeon_fence *radeon_vm_grab_id(struct radeon_device *rdev,
 		if (fence == NULL) {
 			/* found a free one */
 			vm->id = i;
+			trace_radeon_vm_grab_id(vm->id, ring);
 			return NULL;
 		}
 
@@ -769,6 +770,9 @@ void radeon_vm_fence(struct radeon_device *rdev,
 
 	radeon_fence_unref(&vm->fence);
 	vm->fence = radeon_fence_ref(fence);
+
+	radeon_fence_unref(&vm->last_id_use);
+	vm->last_id_use = radeon_fence_ref(fence);
 }
 
 /**
@@ -1303,6 +1307,8 @@ void radeon_vm_init(struct radeon_device *rdev, struct radeon_vm *vm)
 {
 	vm->id = 0;
 	vm->fence = NULL;
+	vm->last_flush = NULL;
+	vm->last_id_use = NULL;
 	mutex_init(&vm->mutex);
 	INIT_LIST_HEAD(&vm->list);
 	INIT_LIST_HEAD(&vm->va);
@@ -1341,5 +1347,6 @@ void radeon_vm_fini(struct radeon_device *rdev, struct radeon_vm *vm)
 	}
 	radeon_fence_unref(&vm->fence);
 	radeon_fence_unref(&vm->last_flush);
+	radeon_fence_unref(&vm->last_id_use);
 	mutex_unlock(&vm->mutex);
 }
diff --git a/drivers/gpu/drm/radeon/radeon_gem.c b/drivers/gpu/drm/radeon/radeon_gem.c
index 805c5e566b9a..b96c819024b3 100644
--- a/drivers/gpu/drm/radeon/radeon_gem.c
+++ b/drivers/gpu/drm/radeon/radeon_gem.c
@@ -86,7 +86,7 @@ retry:
 	return 0;
 }
 
-int radeon_gem_set_domain(struct drm_gem_object *gobj,
+static int radeon_gem_set_domain(struct drm_gem_object *gobj,
 			  uint32_t rdomain, uint32_t wdomain)
 {
 	struct radeon_bo *robj;
diff --git a/drivers/gpu/drm/radeon/radeon_i2c.c b/drivers/gpu/drm/radeon/radeon_i2c.c
index fc60b74ee304..e24ca6ab96de 100644
--- a/drivers/gpu/drm/radeon/radeon_i2c.c
+++ b/drivers/gpu/drm/radeon/radeon_i2c.c
@@ -1020,6 +1020,9 @@ void radeon_i2c_destroy(struct radeon_i2c_chan *i2c)
 /* Add the default buses */
 void radeon_i2c_init(struct radeon_device *rdev)
 {
+	if (radeon_hw_i2c)
+		DRM_INFO("hw_i2c forced on, you may experience display detection problems!\n");
+
 	if (rdev->is_atom_bios)
 		radeon_atombios_i2c_init(rdev);
 	else
diff --git a/drivers/gpu/drm/radeon/radeon_irq.c b/drivers/gpu/drm/radeon/radeon_irq.c
index 8d68e972789a..244b19bab2e7 100644
--- a/drivers/gpu/drm/radeon/radeon_irq.c
+++ b/drivers/gpu/drm/radeon/radeon_irq.c
@@ -181,7 +181,7 @@ static u32 radeon_acknowledge_irqs(drm_radeon_private_t *dev_priv, u32 *r500_dis
  * tied to dma at all, this is just a hangover from dri prehistory.
  */
 
-irqreturn_t radeon_driver_irq_handler(DRM_IRQ_ARGS)
+irqreturn_t radeon_driver_irq_handler(int irq, void *arg)
 {
 	struct drm_device *dev = (struct drm_device *) arg;
 	drm_radeon_private_t *dev_priv =
@@ -203,7 +203,7 @@ irqreturn_t radeon_driver_irq_handler(DRM_IRQ_ARGS)
 
 	/* SW interrupt */
 	if (stat & RADEON_SW_INT_TEST)
-		DRM_WAKEUP(&dev_priv->swi_queue);
+		wake_up(&dev_priv->swi_queue);
 
 	/* VBLANK interrupt */
 	if ((dev_priv->flags & RADEON_FAMILY_MASK) >= CHIP_RS600) {
@@ -249,7 +249,7 @@ static int radeon_wait_irq(struct drm_device * dev, int swi_nr)
 
 	dev_priv->stats.boxes |= RADEON_BOX_WAIT_IDLE;
 
-	DRM_WAIT_ON(ret, dev_priv->swi_queue, 3 * DRM_HZ,
+	DRM_WAIT_ON(ret, dev_priv->swi_queue, 3 * HZ,
 		    RADEON_READ(RADEON_LAST_SWI_REG) >= swi_nr);
 
 	return ret;
@@ -302,7 +302,7 @@ int radeon_irq_emit(struct drm_device *dev, void *data, struct drm_file *file_pr
 
 	result = radeon_emit_irq(dev);
 
-	if (DRM_COPY_TO_USER(emit->irq_seq, &result, sizeof(int))) {
+	if (copy_to_user(emit->irq_seq, &result, sizeof(int))) {
 		DRM_ERROR("copy_to_user\n");
 		return -EFAULT;
 	}
@@ -354,7 +354,7 @@ int radeon_driver_irq_postinstall(struct drm_device *dev)
 	    (drm_radeon_private_t *) dev->dev_private;
 
 	atomic_set(&dev_priv->swi_emitted, 0);
-	DRM_INIT_WAITQUEUE(&dev_priv->swi_queue);
+	init_waitqueue_head(&dev_priv->swi_queue);
 
 	dev->max_vblank_count = 0x001fffff;
 
diff --git a/drivers/gpu/drm/radeon/radeon_irq_kms.c b/drivers/gpu/drm/radeon/radeon_irq_kms.c
index ec6240b00469..089c9ffb0aa9 100644
--- a/drivers/gpu/drm/radeon/radeon_irq_kms.c
+++ b/drivers/gpu/drm/radeon/radeon_irq_kms.c
@@ -39,13 +39,13 @@
 /**
  * radeon_driver_irq_handler_kms - irq handler for KMS
  *
- * @DRM_IRQ_ARGS: args
+ * @int irq, void *arg: args
  *
  * This is the irq handler for the radeon KMS driver (all asics).
  * radeon_irq_process is a macro that points to the per-asic
  * irq handler callback.
  */
-irqreturn_t radeon_driver_irq_handler_kms(DRM_IRQ_ARGS)
+irqreturn_t radeon_driver_irq_handler_kms(int irq, void *arg)
 {
 	struct drm_device *dev = (struct drm_device *) arg;
 	struct radeon_device *rdev = dev->dev_private;
diff --git a/drivers/gpu/drm/radeon/radeon_kms.c b/drivers/gpu/drm/radeon/radeon_kms.c
index 21d593c0ecaf..114d1672d616 100644
--- a/drivers/gpu/drm/radeon/radeon_kms.c
+++ b/drivers/gpu/drm/radeon/radeon_kms.c
@@ -191,7 +191,7 @@ static void radeon_set_filp_rights(struct drm_device *dev,
  * etc. (all asics).
  * Returns 0 on success, -EINVAL on failure.
  */
-int radeon_info_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
+static int radeon_info_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
 {
 	struct radeon_device *rdev = dev->dev_private;
 	struct drm_radeon_info *info = data;
@@ -223,7 +223,7 @@ int radeon_info_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
 			*value = rdev->accel_working;
 		break;
 	case RADEON_INFO_CRTC_FROM_ID:
-		if (DRM_COPY_FROM_USER(value, value_ptr, sizeof(uint32_t))) {
+		if (copy_from_user(value, value_ptr, sizeof(uint32_t))) {
 			DRM_ERROR("copy_from_user %s:%u\n", __func__, __LINE__);
 			return -EFAULT;
 		}
@@ -269,7 +269,7 @@ int radeon_info_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
 		 *
 		 * When returning, the value is 1 if filp owns hyper-z access,
 		 * 0 otherwise. */
-		if (DRM_COPY_FROM_USER(value, value_ptr, sizeof(uint32_t))) {
+		if (copy_from_user(value, value_ptr, sizeof(uint32_t))) {
 			DRM_ERROR("copy_from_user %s:%u\n", __func__, __LINE__);
 			return -EFAULT;
 		}
@@ -281,7 +281,7 @@ int radeon_info_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
 		break;
 	case RADEON_INFO_WANT_CMASK:
 		/* The same logic as Hyper-Z. */
-		if (DRM_COPY_FROM_USER(value, value_ptr, sizeof(uint32_t))) {
+		if (copy_from_user(value, value_ptr, sizeof(uint32_t))) {
 			DRM_ERROR("copy_from_user %s:%u\n", __func__, __LINE__);
 			return -EFAULT;
 		}
@@ -417,7 +417,7 @@ int radeon_info_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
 		*value = rdev->fastfb_working;
 		break;
 	case RADEON_INFO_RING_WORKING:
-		if (DRM_COPY_FROM_USER(value, value_ptr, sizeof(uint32_t))) {
+		if (copy_from_user(value, value_ptr, sizeof(uint32_t))) {
 			DRM_ERROR("copy_from_user %s:%u\n", __func__, __LINE__);
 			return -EFAULT;
 		}
@@ -470,11 +470,18 @@ int radeon_info_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
 			DRM_DEBUG_KMS("BACKEND_ENABLED_MASK is si+ only!\n");
 		}
 		break;
+	case RADEON_INFO_MAX_SCLK:
+		if ((rdev->pm.pm_method == PM_METHOD_DPM) &&
+		    rdev->pm.dpm_enabled)
+			*value = rdev->pm.dpm.dyn_state.max_clock_voltage_on_ac.sclk * 10;
+		else
+			*value = rdev->pm.default_sclk * 10;
+		break;
 	default:
 		DRM_DEBUG_KMS("Invalid request %d\n", info->request);
 		return -EINVAL;
 	}
-	if (DRM_COPY_TO_USER(value_ptr, (char*)value, value_size)) {
+	if (copy_to_user(value_ptr, (char*)value, value_size)) {
 		DRM_ERROR("copy_to_user %s:%u\n", __func__, __LINE__);
 		return -EFAULT;
 	}
@@ -712,11 +719,12 @@ int radeon_get_vblank_timestamp_kms(struct drm_device *dev, int crtc,
 	/* Helper routine in DRM core does all the work: */
 	return drm_calc_vbltimestamp_from_scanoutpos(dev, crtc, max_error,
 						     vblank_time, flags,
-						     drmcrtc);
+						     drmcrtc, &drmcrtc->hwmode);
 }
 
 #define KMS_INVALID_IOCTL(name)						\
-int name(struct drm_device *dev, void *data, struct drm_file *file_priv)\
+static int name(struct drm_device *dev, void *data, struct drm_file	\
+		*file_priv)						\
 {									\
 	DRM_ERROR("invalid ioctl with kms %s\n", __func__);		\
 	return -EINVAL;							\
diff --git a/drivers/gpu/drm/radeon/radeon_mem.c b/drivers/gpu/drm/radeon/radeon_mem.c
index d54d2d7c9031..146d253f1131 100644
--- a/drivers/gpu/drm/radeon/radeon_mem.c
+++ b/drivers/gpu/drm/radeon/radeon_mem.c
@@ -243,7 +243,7 @@ int radeon_mem_alloc(struct drm_device *dev, void *data, struct drm_file *file_p
 	if (!block)
 		return -ENOMEM;
 
-	if (DRM_COPY_TO_USER(alloc->region_offset, &block->start,
+	if (copy_to_user(alloc->region_offset, &block->start,
 			     sizeof(int))) {
 		DRM_ERROR("copy_to_user\n");
 		return -EFAULT;
diff --git a/drivers/gpu/drm/radeon/radeon_mode.h b/drivers/gpu/drm/radeon/radeon_mode.h
index 3f0dd664af90..402dbe32c234 100644
--- a/drivers/gpu/drm/radeon/radeon_mode.h
+++ b/drivers/gpu/drm/radeon/radeon_mode.h
@@ -291,6 +291,7 @@ struct radeon_tv_regs {
 
 struct radeon_atom_ss {
 	uint16_t percentage;
+	uint16_t percentage_divider;
 	uint8_t type;
 	uint16_t step;
 	uint8_t delay;
@@ -624,6 +625,30 @@ struct atom_voltage_table
 	struct atom_voltage_table_entry entries[MAX_VOLTAGE_ENTRIES];
 };
 
+
+extern void
+radeon_add_atom_connector(struct drm_device *dev,
+			  uint32_t connector_id,
+			  uint32_t supported_device,
+			  int connector_type,
+			  struct radeon_i2c_bus_rec *i2c_bus,
+			  uint32_t igp_lane_info,
+			  uint16_t connector_object_id,
+			  struct radeon_hpd *hpd,
+			  struct radeon_router *router);
+extern void
+radeon_add_legacy_connector(struct drm_device *dev,
+			    uint32_t connector_id,
+			    uint32_t supported_device,
+			    int connector_type,
+			    struct radeon_i2c_bus_rec *i2c_bus,
+			    uint16_t connector_object_id,
+			    struct radeon_hpd *hpd);
+extern uint32_t
+radeon_get_encoder_enum(struct drm_device *dev, uint32_t supported_device,
+			uint8_t dac);
+extern void radeon_link_encoder_connector(struct drm_device *dev);
+
 extern enum radeon_tv_std
 radeon_combios_get_tv_info(struct radeon_device *rdev);
 extern enum radeon_tv_std
@@ -631,6 +656,15 @@ radeon_atombios_get_tv_info(struct radeon_device *rdev);
 extern void radeon_atombios_get_default_voltages(struct radeon_device *rdev,
 						 u16 *vddc, u16 *vddci, u16 *mvdd);
 
+extern void
+radeon_combios_connected_scratch_regs(struct drm_connector *connector,
+				      struct drm_encoder *encoder,
+				      bool connected);
+extern void
+radeon_atombios_connected_scratch_regs(struct drm_connector *connector,
+				       struct drm_encoder *encoder,
+				       bool connected);
+
 extern struct drm_connector *
 radeon_get_connector_for_encoder(struct drm_encoder *encoder);
 extern struct drm_connector *
@@ -666,6 +700,7 @@ extern void radeon_atom_ext_encoder_setup_ddc(struct drm_encoder *encoder);
 extern struct drm_encoder *radeon_get_external_encoder(struct drm_encoder *encoder);
 extern int radeon_dp_i2c_aux_ch(struct i2c_adapter *adapter, int mode,
 				u8 write_byte, u8 *read_byte);
+void radeon_atom_copy_swap(u8 *dst, u8 *src, u8 num_bytes, bool to_le);
 
 extern void radeon_i2c_init(struct radeon_device *rdev);
 extern void radeon_i2c_fini(struct radeon_device *rdev);
@@ -766,6 +801,7 @@ extern int radeon_crtc_cursor_move(struct drm_crtc *crtc,
 				   int x, int y);
 
 extern int radeon_get_crtc_scanoutpos(struct drm_device *dev, int crtc,
+				      unsigned int flags,
 				      int *vpos, int *hpos, ktime_t *stime,
 				      ktime_t *etime);
 
diff --git a/drivers/gpu/drm/radeon/radeon_object.c b/drivers/gpu/drm/radeon/radeon_object.c
index c0fa4aa9ceea..08595cf90b01 100644
--- a/drivers/gpu/drm/radeon/radeon_object.c
+++ b/drivers/gpu/drm/radeon/radeon_object.c
@@ -46,7 +46,7 @@ static void radeon_bo_clear_surface_reg(struct radeon_bo *bo);
  * function are calling it.
  */
 
-void radeon_bo_clear_va(struct radeon_bo *bo)
+static void radeon_bo_clear_va(struct radeon_bo *bo)
 {
 	struct radeon_bo_va *bo_va, *tmp;
 
diff --git a/drivers/gpu/drm/radeon/radeon_pm.c b/drivers/gpu/drm/radeon/radeon_pm.c
index 984097b907ef..8e8153e471c2 100644
--- a/drivers/gpu/drm/radeon/radeon_pm.c
+++ b/drivers/gpu/drm/radeon/radeon_pm.c
@@ -924,6 +924,10 @@ void radeon_dpm_enable_uvd(struct radeon_device *rdev, bool enable)
 
 	if (rdev->asic->dpm.powergate_uvd) {
 		mutex_lock(&rdev->pm.mutex);
+		/* don't powergate anything if we
+		   have active but pause streams */
+		enable |= rdev->pm.dpm.sd > 0;
+		enable |= rdev->pm.dpm.hd > 0;
 		/* enable/disable UVD */
 		radeon_dpm_powergate_uvd(rdev, !enable);
 		mutex_unlock(&rdev->pm.mutex);
@@ -1010,8 +1014,10 @@ static void radeon_pm_resume_old(struct radeon_device *rdev)
 	rdev->pm.current_clock_mode_index = 0;
 	rdev->pm.current_sclk = rdev->pm.default_sclk;
 	rdev->pm.current_mclk = rdev->pm.default_mclk;
-	rdev->pm.current_vddc = rdev->pm.power_state[rdev->pm.default_power_state_index].clock_info[0].voltage.voltage;
-	rdev->pm.current_vddci = rdev->pm.power_state[rdev->pm.default_power_state_index].clock_info[0].voltage.vddci;
+	if (rdev->pm.power_state) {
+		rdev->pm.current_vddc = rdev->pm.power_state[rdev->pm.default_power_state_index].clock_info[0].voltage.voltage;
+		rdev->pm.current_vddci = rdev->pm.power_state[rdev->pm.default_power_state_index].clock_info[0].voltage.vddci;
+	}
 	if (rdev->pm.pm_method == PM_METHOD_DYNPM
 	    && rdev->pm.dynpm_state == DYNPM_STATE_SUSPENDED) {
 		rdev->pm.dynpm_state = DYNPM_STATE_ACTIVE;
@@ -1032,25 +1038,27 @@ static void radeon_pm_resume_dpm(struct radeon_device *rdev)
 	radeon_dpm_setup_asic(rdev);
 	ret = radeon_dpm_enable(rdev);
 	mutex_unlock(&rdev->pm.mutex);
-	if (ret) {
-		DRM_ERROR("radeon: dpm resume failed\n");
-		if ((rdev->family >= CHIP_BARTS) &&
-		    (rdev->family <= CHIP_CAYMAN) &&
-		    rdev->mc_fw) {
-			if (rdev->pm.default_vddc)
-				radeon_atom_set_voltage(rdev, rdev->pm.default_vddc,
-							SET_VOLTAGE_TYPE_ASIC_VDDC);
-			if (rdev->pm.default_vddci)
-				radeon_atom_set_voltage(rdev, rdev->pm.default_vddci,
-							SET_VOLTAGE_TYPE_ASIC_VDDCI);
-			if (rdev->pm.default_sclk)
-				radeon_set_engine_clock(rdev, rdev->pm.default_sclk);
-			if (rdev->pm.default_mclk)
-				radeon_set_memory_clock(rdev, rdev->pm.default_mclk);
-		}
-	} else {
-		rdev->pm.dpm_enabled = true;
-		radeon_pm_compute_clocks(rdev);
+	if (ret)
+		goto dpm_resume_fail;
+	rdev->pm.dpm_enabled = true;
+	radeon_pm_compute_clocks(rdev);
+	return;
+
+dpm_resume_fail:
+	DRM_ERROR("radeon: dpm resume failed\n");
+	if ((rdev->family >= CHIP_BARTS) &&
+	    (rdev->family <= CHIP_CAYMAN) &&
+	    rdev->mc_fw) {
+		if (rdev->pm.default_vddc)
+			radeon_atom_set_voltage(rdev, rdev->pm.default_vddc,
+						SET_VOLTAGE_TYPE_ASIC_VDDC);
+		if (rdev->pm.default_vddci)
+			radeon_atom_set_voltage(rdev, rdev->pm.default_vddci,
+						SET_VOLTAGE_TYPE_ASIC_VDDCI);
+		if (rdev->pm.default_sclk)
+			radeon_set_engine_clock(rdev, rdev->pm.default_sclk);
+		if (rdev->pm.default_mclk)
+			radeon_set_memory_clock(rdev, rdev->pm.default_mclk);
 	}
 }
 
@@ -1170,51 +1178,50 @@ static int radeon_pm_init_dpm(struct radeon_device *rdev)
 	radeon_dpm_setup_asic(rdev);
 	ret = radeon_dpm_enable(rdev);
 	mutex_unlock(&rdev->pm.mutex);
-	if (ret) {
-		rdev->pm.dpm_enabled = false;
-		if ((rdev->family >= CHIP_BARTS) &&
-		    (rdev->family <= CHIP_CAYMAN) &&
-		    rdev->mc_fw) {
-			if (rdev->pm.default_vddc)
-				radeon_atom_set_voltage(rdev, rdev->pm.default_vddc,
-							SET_VOLTAGE_TYPE_ASIC_VDDC);
-			if (rdev->pm.default_vddci)
-				radeon_atom_set_voltage(rdev, rdev->pm.default_vddci,
-							SET_VOLTAGE_TYPE_ASIC_VDDCI);
-			if (rdev->pm.default_sclk)
-				radeon_set_engine_clock(rdev, rdev->pm.default_sclk);
-			if (rdev->pm.default_mclk)
-				radeon_set_memory_clock(rdev, rdev->pm.default_mclk);
-		}
-		DRM_ERROR("radeon: dpm initialization failed\n");
-		return ret;
-	}
+	if (ret)
+		goto dpm_failed;
 	rdev->pm.dpm_enabled = true;
-	radeon_pm_compute_clocks(rdev);
 
-	if (rdev->pm.num_power_states > 1) {
-		ret = device_create_file(rdev->dev, &dev_attr_power_dpm_state);
-		if (ret)
-			DRM_ERROR("failed to create device file for dpm state\n");
-		ret = device_create_file(rdev->dev, &dev_attr_power_dpm_force_performance_level);
-		if (ret)
-			DRM_ERROR("failed to create device file for dpm state\n");
-		/* XXX: these are noops for dpm but are here for backwards compat */
-		ret = device_create_file(rdev->dev, &dev_attr_power_profile);
-		if (ret)
-			DRM_ERROR("failed to create device file for power profile\n");
-		ret = device_create_file(rdev->dev, &dev_attr_power_method);
-		if (ret)
-			DRM_ERROR("failed to create device file for power method\n");
-
-		if (radeon_debugfs_pm_init(rdev)) {
-			DRM_ERROR("Failed to register debugfs file for dpm!\n");
-		}
+	ret = device_create_file(rdev->dev, &dev_attr_power_dpm_state);
+	if (ret)
+		DRM_ERROR("failed to create device file for dpm state\n");
+	ret = device_create_file(rdev->dev, &dev_attr_power_dpm_force_performance_level);
+	if (ret)
+		DRM_ERROR("failed to create device file for dpm state\n");
+	/* XXX: these are noops for dpm but are here for backwards compat */
+	ret = device_create_file(rdev->dev, &dev_attr_power_profile);
+	if (ret)
+		DRM_ERROR("failed to create device file for power profile\n");
+	ret = device_create_file(rdev->dev, &dev_attr_power_method);
+	if (ret)
+		DRM_ERROR("failed to create device file for power method\n");
 
-		DRM_INFO("radeon: dpm initialized\n");
+	if (radeon_debugfs_pm_init(rdev)) {
+		DRM_ERROR("Failed to register debugfs file for dpm!\n");
 	}
 
+	DRM_INFO("radeon: dpm initialized\n");
+
 	return 0;
+
+dpm_failed:
+	rdev->pm.dpm_enabled = false;
+	if ((rdev->family >= CHIP_BARTS) &&
+	    (rdev->family <= CHIP_CAYMAN) &&
+	    rdev->mc_fw) {
+		if (rdev->pm.default_vddc)
+			radeon_atom_set_voltage(rdev, rdev->pm.default_vddc,
+						SET_VOLTAGE_TYPE_ASIC_VDDC);
+		if (rdev->pm.default_vddci)
+			radeon_atom_set_voltage(rdev, rdev->pm.default_vddci,
+						SET_VOLTAGE_TYPE_ASIC_VDDCI);
+		if (rdev->pm.default_sclk)
+			radeon_set_engine_clock(rdev, rdev->pm.default_sclk);
+		if (rdev->pm.default_mclk)
+			radeon_set_memory_clock(rdev, rdev->pm.default_mclk);
+	}
+	DRM_ERROR("radeon: dpm initialization failed\n");
+	return ret;
 }
 
 int radeon_pm_init(struct radeon_device *rdev)
@@ -1228,11 +1235,10 @@ int radeon_pm_init(struct radeon_device *rdev)
 	case CHIP_RV670:
 	case CHIP_RS780:
 	case CHIP_RS880:
+	case CHIP_BARTS:
+	case CHIP_TURKS:
+	case CHIP_CAICOS:
 	case CHIP_CAYMAN:
-	case CHIP_BONAIRE:
-	case CHIP_KABINI:
-	case CHIP_KAVERI:
-	case CHIP_HAWAII:
 		/* DPM requires the RLC, RV770+ dGPU requires SMC */
 		if (!rdev->rlc_fw)
 			rdev->pm.pm_method = PM_METHOD_PROFILE;
@@ -1257,15 +1263,16 @@ int radeon_pm_init(struct radeon_device *rdev)
 	case CHIP_PALM:
 	case CHIP_SUMO:
 	case CHIP_SUMO2:
-	case CHIP_BARTS:
-	case CHIP_TURKS:
-	case CHIP_CAICOS:
 	case CHIP_ARUBA:
 	case CHIP_TAHITI:
 	case CHIP_PITCAIRN:
 	case CHIP_VERDE:
 	case CHIP_OLAND:
 	case CHIP_HAINAN:
+	case CHIP_BONAIRE:
+	case CHIP_KABINI:
+	case CHIP_KAVERI:
+	case CHIP_HAWAII:
 		/* DPM requires the RLC, RV770+ dGPU requires SMC */
 		if (!rdev->rlc_fw)
 			rdev->pm.pm_method = PM_METHOD_PROFILE;
@@ -1290,6 +1297,18 @@ int radeon_pm_init(struct radeon_device *rdev)
 		return radeon_pm_init_old(rdev);
 }
 
+int radeon_pm_late_init(struct radeon_device *rdev)
+{
+	int ret = 0;
+
+	if (rdev->pm.pm_method == PM_METHOD_DPM) {
+		mutex_lock(&rdev->pm.mutex);
+		ret = radeon_dpm_late_enable(rdev);
+		mutex_unlock(&rdev->pm.mutex);
+	}
+	return ret;
+}
+
 static void radeon_pm_fini_old(struct radeon_device *rdev)
 {
 	if (rdev->pm.num_power_states > 1) {
@@ -1420,6 +1439,9 @@ static void radeon_pm_compute_clocks_dpm(struct radeon_device *rdev)
 	struct drm_crtc *crtc;
 	struct radeon_crtc *radeon_crtc;
 
+	if (!rdev->pm.dpm_enabled)
+		return;
+
 	mutex_lock(&rdev->pm.mutex);
 
 	/* update active crtc counts */
@@ -1464,7 +1486,7 @@ static bool radeon_pm_in_vbl(struct radeon_device *rdev)
 	 */
 	for (crtc = 0; (crtc < rdev->num_crtc) && in_vbl; crtc++) {
 		if (rdev->pm.active_crtcs & (1 << crtc)) {
-			vbl_status = radeon_get_crtc_scanoutpos(rdev->ddev, crtc, &vpos, &hpos, NULL, NULL);
+			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))
 				in_vbl = false;
diff --git a/drivers/gpu/drm/radeon/radeon_ring.c b/drivers/gpu/drm/radeon/radeon_ring.c
index 9214403ae173..1b783f0e6d3a 100644
--- a/drivers/gpu/drm/radeon/radeon_ring.c
+++ b/drivers/gpu/drm/radeon/radeon_ring.c
@@ -332,36 +332,6 @@ bool radeon_ring_supports_scratch_reg(struct radeon_device *rdev,
 	}
 }
 
-u32 radeon_ring_generic_get_rptr(struct radeon_device *rdev,
-				 struct radeon_ring *ring)
-{
-	u32 rptr;
-
-	if (rdev->wb.enabled)
-		rptr = le32_to_cpu(rdev->wb.wb[ring->rptr_offs/4]);
-	else
-		rptr = RREG32(ring->rptr_reg);
-
-	return rptr;
-}
-
-u32 radeon_ring_generic_get_wptr(struct radeon_device *rdev,
-				 struct radeon_ring *ring)
-{
-	u32 wptr;
-
-	wptr = RREG32(ring->wptr_reg);
-
-	return wptr;
-}
-
-void radeon_ring_generic_set_wptr(struct radeon_device *rdev,
-				  struct radeon_ring *ring)
-{
-	WREG32(ring->wptr_reg, ring->wptr);
-	(void)RREG32(ring->wptr_reg);
-}
-
 /**
  * radeon_ring_free_size - update the free size
  *
@@ -463,7 +433,7 @@ void radeon_ring_commit(struct radeon_device *rdev, struct radeon_ring *ring)
 	while (ring->wptr & ring->align_mask) {
 		radeon_ring_write(ring, ring->nop);
 	}
-	DRM_MEMORYBARRIER();
+	mb();
 	radeon_ring_set_wptr(rdev, ring);
 }
 
@@ -689,22 +659,18 @@ int radeon_ring_restore(struct radeon_device *rdev, struct radeon_ring *ring,
  * @ring: radeon_ring structure holding ring information
  * @ring_size: size of the ring
  * @rptr_offs: offset of the rptr writeback location in the WB buffer
- * @rptr_reg: MMIO offset of the rptr register
- * @wptr_reg: MMIO offset of the wptr register
  * @nop: nop packet for this ring
  *
  * Initialize the driver information for the selected ring (all asics).
  * Returns 0 on success, error on failure.
  */
 int radeon_ring_init(struct radeon_device *rdev, struct radeon_ring *ring, unsigned ring_size,
-		     unsigned rptr_offs, unsigned rptr_reg, unsigned wptr_reg, u32 nop)
+		     unsigned rptr_offs, u32 nop)
 {
 	int r;
 
 	ring->ring_size = ring_size;
 	ring->rptr_offs = rptr_offs;
-	ring->rptr_reg = rptr_reg;
-	ring->wptr_reg = wptr_reg;
 	ring->nop = nop;
 	/* Allocate ring buffer */
 	if (ring->ring_obj == NULL) {
@@ -790,34 +756,54 @@ static int radeon_debugfs_ring_info(struct seq_file *m, void *data)
 	struct radeon_device *rdev = dev->dev_private;
 	int ridx = *(int*)node->info_ent->data;
 	struct radeon_ring *ring = &rdev->ring[ridx];
+
+	uint32_t rptr, wptr, rptr_next;
 	unsigned count, i, j;
-	u32 tmp;
 
 	radeon_ring_free_size(rdev, ring);
 	count = (ring->ring_size / 4) - ring->ring_free_dw;
-	tmp = radeon_ring_get_wptr(rdev, ring);
-	seq_printf(m, "wptr(0x%04x): 0x%08x [%5d]\n", ring->wptr_reg, tmp, tmp);
-	tmp = radeon_ring_get_rptr(rdev, ring);
-	seq_printf(m, "rptr(0x%04x): 0x%08x [%5d]\n", ring->rptr_reg, tmp, tmp);
+
+	wptr = radeon_ring_get_wptr(rdev, ring);
+	seq_printf(m, "wptr: 0x%08x [%5d]\n",
+		   wptr, wptr);
+
+	rptr = radeon_ring_get_rptr(rdev, ring);
+	seq_printf(m, "rptr: 0x%08x [%5d]\n",
+		   rptr, rptr);
+
 	if (ring->rptr_save_reg) {
-		seq_printf(m, "rptr next(0x%04x): 0x%08x\n", ring->rptr_save_reg,
-			   RREG32(ring->rptr_save_reg));
-	}
-	seq_printf(m, "driver's copy of the wptr: 0x%08x [%5d]\n", ring->wptr, ring->wptr);
-	seq_printf(m, "driver's copy of the rptr: 0x%08x [%5d]\n", ring->rptr, ring->rptr);
-	seq_printf(m, "last semaphore signal addr : 0x%016llx\n", ring->last_semaphore_signal_addr);
-	seq_printf(m, "last semaphore wait addr   : 0x%016llx\n", ring->last_semaphore_wait_addr);
+		rptr_next = RREG32(ring->rptr_save_reg);
+		seq_printf(m, "rptr next(0x%04x): 0x%08x [%5d]\n",
+			   ring->rptr_save_reg, rptr_next, rptr_next);
+	} else
+		rptr_next = ~0;
+
+	seq_printf(m, "driver's copy of the wptr: 0x%08x [%5d]\n",
+		   ring->wptr, ring->wptr);
+	seq_printf(m, "driver's copy of the rptr: 0x%08x [%5d]\n",
+		   ring->rptr, ring->rptr);
+	seq_printf(m, "last semaphore signal addr : 0x%016llx\n",
+		   ring->last_semaphore_signal_addr);
+	seq_printf(m, "last semaphore wait addr   : 0x%016llx\n",
+		   ring->last_semaphore_wait_addr);
 	seq_printf(m, "%u free dwords in ring\n", ring->ring_free_dw);
 	seq_printf(m, "%u dwords in ring\n", count);
+
+	if (!ring->ready)
+		return 0;
+
 	/* print 8 dw before current rptr as often it's the last executed
 	 * packet that is the root issue
 	 */
-	i = (ring->rptr + ring->ptr_mask + 1 - 32) & ring->ptr_mask;
-	if (ring->ready) {
-		for (j = 0; j <= (count + 32); j++) {
-			seq_printf(m, "r[%5d]=0x%08x\n", i, ring->ring[i]);
-			i = (i + 1) & ring->ptr_mask;
-		}
+	i = (rptr + ring->ptr_mask + 1 - 32) & ring->ptr_mask;
+	for (j = 0; j <= (count + 32); j++) {
+		seq_printf(m, "r[%5d]=0x%08x", i, ring->ring[i]);
+		if (rptr == i)
+			seq_puts(m, " *");
+		if (rptr_next == i)
+			seq_puts(m, " #");
+		seq_puts(m, "\n");
+		i = (i + 1) & ring->ptr_mask;
 	}
 	return 0;
 }
diff --git a/drivers/gpu/drm/radeon/radeon_sa.c b/drivers/gpu/drm/radeon/radeon_sa.c
index f0bac68254b7..c0625805cdd7 100644
--- a/drivers/gpu/drm/radeon/radeon_sa.c
+++ b/drivers/gpu/drm/radeon/radeon_sa.c
@@ -402,13 +402,15 @@ void radeon_sa_bo_dump_debug_info(struct radeon_sa_manager *sa_manager,
 
 	spin_lock(&sa_manager->wq.lock);
 	list_for_each_entry(i, &sa_manager->olist, olist) {
+		uint64_t soffset = i->soffset + sa_manager->gpu_addr;
+		uint64_t eoffset = i->eoffset + sa_manager->gpu_addr;
 		if (&i->olist == sa_manager->hole) {
 			seq_printf(m, ">");
 		} else {
 			seq_printf(m, " ");
 		}
-		seq_printf(m, "[0x%08x 0x%08x] size %8d",
-			   i->soffset, i->eoffset, i->eoffset - i->soffset);
+		seq_printf(m, "[0x%010llx 0x%010llx] size %8lld",
+			   soffset, eoffset, eoffset - soffset);
 		if (i->fence) {
 			seq_printf(m, " protected by 0x%016llx on ring %d",
 				   i->fence->seq, i->fence->ring);
diff --git a/drivers/gpu/drm/radeon/radeon_state.c b/drivers/gpu/drm/radeon/radeon_state.c
index 4d20910899d4..956ab7f14e16 100644
--- a/drivers/gpu/drm/radeon/radeon_state.c
+++ b/drivers/gpu/drm/radeon/radeon_state.c
@@ -1810,7 +1810,7 @@ static int radeon_cp_dispatch_texture(struct drm_device * dev,
 		}
 		if (!buf) {
 			DRM_DEBUG("EAGAIN\n");
-			if (DRM_COPY_TO_USER(tex->image, image, sizeof(*image)))
+			if (copy_to_user(tex->image, image, sizeof(*image)))
 				return -EFAULT;
 			return -EAGAIN;
 		}
@@ -1823,7 +1823,7 @@ static int radeon_cp_dispatch_texture(struct drm_device * dev,
 
 #define RADEON_COPY_MT(_buf, _data, _width) \
 	do { \
-		if (DRM_COPY_FROM_USER(_buf, _data, (_width))) {\
+		if (copy_from_user(_buf, _data, (_width))) {\
 			DRM_ERROR("EFAULT on pad, %d bytes\n", (_width)); \
 			return -EFAULT; \
 		} \
@@ -2168,7 +2168,7 @@ static int radeon_cp_clear(struct drm_device *dev, void *data, struct drm_file *
 	if (sarea_priv->nbox > RADEON_NR_SAREA_CLIPRECTS)
 		sarea_priv->nbox = RADEON_NR_SAREA_CLIPRECTS;
 
-	if (DRM_COPY_FROM_USER(&depth_boxes, clear->depth_boxes,
+	if (copy_from_user(&depth_boxes, clear->depth_boxes,
 			       sarea_priv->nbox * sizeof(depth_boxes[0])))
 		return -EFAULT;
 
@@ -2436,7 +2436,7 @@ static int radeon_cp_texture(struct drm_device *dev, void *data, struct drm_file
 		return -EINVAL;
 	}
 
-	if (DRM_COPY_FROM_USER(&image,
+	if (copy_from_user(&image,
 			       (drm_radeon_tex_image_t __user *) tex->image,
 			       sizeof(image)))
 		return -EFAULT;
@@ -2460,7 +2460,7 @@ static int radeon_cp_stipple(struct drm_device *dev, void *data, struct drm_file
 
 	LOCK_TEST_WITH_RETURN(dev, file_priv);
 
-	if (DRM_COPY_FROM_USER(&mask, stipple->mask, 32 * sizeof(u32)))
+	if (copy_from_user(&mask, stipple->mask, 32 * sizeof(u32)))
 		return -EFAULT;
 
 	RING_SPACE_TEST_WITH_RETURN(dev_priv);
@@ -2585,13 +2585,13 @@ static int radeon_cp_vertex2(struct drm_device *dev, void *data, struct drm_file
 		drm_radeon_prim_t prim;
 		drm_radeon_tcl_prim_t tclprim;
 
-		if (DRM_COPY_FROM_USER(&prim, &vertex->prim[i], sizeof(prim)))
+		if (copy_from_user(&prim, &vertex->prim[i], sizeof(prim)))
 			return -EFAULT;
 
 		if (prim.stateidx != laststate) {
 			drm_radeon_state_t state;
 
-			if (DRM_COPY_FROM_USER(&state,
+			if (copy_from_user(&state,
 					       &vertex->state[prim.stateidx],
 					       sizeof(state)))
 				return -EFAULT;
@@ -2799,7 +2799,7 @@ static int radeon_emit_packet3_cliprect(struct drm_device *dev,
 
 	do {
 		if (i < cmdbuf->nbox) {
-			if (DRM_COPY_FROM_USER(&box, &boxes[i], sizeof(box)))
+			if (copy_from_user(&box, &boxes[i], sizeof(box)))
 				return -EFAULT;
 			/* FIXME The second and subsequent times round
 			 * this loop, send a WAIT_UNTIL_3D_IDLE before
@@ -3116,7 +3116,7 @@ static int radeon_cp_getparam(struct drm_device *dev, void *data, struct drm_fil
 		return -EINVAL;
 	}
 
-	if (DRM_COPY_TO_USER(param->value, &value, sizeof(int))) {
+	if (copy_to_user(param->value, &value, sizeof(int))) {
 		DRM_ERROR("copy_to_user\n");
 		return -EFAULT;
 	}
diff --git a/drivers/gpu/drm/radeon/radeon_trace.h b/drivers/gpu/drm/radeon/radeon_trace.h
index 0473257d4078..f749f2c3bbdb 100644
--- a/drivers/gpu/drm/radeon/radeon_trace.h
+++ b/drivers/gpu/drm/radeon/radeon_trace.h
@@ -106,42 +106,45 @@ TRACE_EVENT(radeon_vm_set_page,
 
 DECLARE_EVENT_CLASS(radeon_fence_request,
 
-	    TP_PROTO(struct drm_device *dev, u32 seqno),
+	    TP_PROTO(struct drm_device *dev, int ring, u32 seqno),
 
-	    TP_ARGS(dev, seqno),
+	    TP_ARGS(dev, ring, seqno),
 
 	    TP_STRUCT__entry(
 			     __field(u32, dev)
+			     __field(int, ring)
 			     __field(u32, seqno)
 			     ),
 
 	    TP_fast_assign(
 			   __entry->dev = dev->primary->index;
+			   __entry->ring = ring;
 			   __entry->seqno = seqno;
 			   ),
 
-	    TP_printk("dev=%u, seqno=%u", __entry->dev, __entry->seqno)
+	    TP_printk("dev=%u, ring=%d, seqno=%u",
+		      __entry->dev, __entry->ring, __entry->seqno)
 );
 
 DEFINE_EVENT(radeon_fence_request, radeon_fence_emit,
 
-	    TP_PROTO(struct drm_device *dev, u32 seqno),
+	    TP_PROTO(struct drm_device *dev, int ring, u32 seqno),
 
-	    TP_ARGS(dev, seqno)
+	    TP_ARGS(dev, ring, seqno)
 );
 
 DEFINE_EVENT(radeon_fence_request, radeon_fence_wait_begin,
 
-	    TP_PROTO(struct drm_device *dev, u32 seqno),
+	    TP_PROTO(struct drm_device *dev, int ring, u32 seqno),
 
-	    TP_ARGS(dev, seqno)
+	    TP_ARGS(dev, ring, seqno)
 );
 
 DEFINE_EVENT(radeon_fence_request, radeon_fence_wait_end,
 
-	    TP_PROTO(struct drm_device *dev, u32 seqno),
+	    TP_PROTO(struct drm_device *dev, int ring, u32 seqno),
 
-	    TP_ARGS(dev, seqno)
+	    TP_ARGS(dev, ring, seqno)
 );
 
 DECLARE_EVENT_CLASS(radeon_semaphore_request,
diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c b/drivers/gpu/drm/radeon/radeon_ttm.c
index 71245d6f34a2..77f5b0c3edb8 100644
--- a/drivers/gpu/drm/radeon/radeon_ttm.c
+++ b/drivers/gpu/drm/radeon/radeon_ttm.c
@@ -39,12 +39,14 @@
 #include <linux/seq_file.h>
 #include <linux/slab.h>
 #include <linux/swiotlb.h>
+#include <linux/debugfs.h>
 #include "radeon_reg.h"
 #include "radeon.h"
 
 #define DRM_FILE_PAGE_OFFSET (0x100000000ULL >> PAGE_SHIFT)
 
 static int radeon_ttm_debugfs_init(struct radeon_device *rdev);
+static void radeon_ttm_debugfs_fini(struct radeon_device *rdev);
 
 static struct radeon_device *radeon_get_rdev(struct ttm_bo_device *bdev)
 {
@@ -142,7 +144,7 @@ static int radeon_init_mem_type(struct ttm_bo_device *bdev, uint32_t type,
 		man->flags = TTM_MEMTYPE_FLAG_MAPPABLE | TTM_MEMTYPE_FLAG_CMA;
 #if __OS_HAS_AGP
 		if (rdev->flags & RADEON_IS_AGP) {
-			if (!(drm_core_has_AGP(rdev->ddev) && rdev->ddev->agp)) {
+			if (!rdev->ddev->agp) {
 				DRM_ERROR("AGP is not enabled for memory type %u\n",
 					  (unsigned)type);
 				return -EINVAL;
@@ -753,6 +755,7 @@ void radeon_ttm_fini(struct radeon_device *rdev)
 
 	if (!rdev->mman.initialized)
 		return;
+	radeon_ttm_debugfs_fini(rdev);
 	if (rdev->stollen_vga_memory) {
 		r = radeon_bo_reserve(rdev->stollen_vga_memory, false);
 		if (r == 0) {
@@ -832,16 +835,15 @@ int radeon_mmap(struct file *filp, struct vm_area_struct *vma)
 	return 0;
 }
 
-
-#define RADEON_DEBUGFS_MEM_TYPES 2
-
 #if defined(CONFIG_DEBUG_FS)
+
 static int radeon_mm_dump_table(struct seq_file *m, void *data)
 {
 	struct drm_info_node *node = (struct drm_info_node *)m->private;
-	struct drm_mm *mm = (struct drm_mm *)node->info_ent->data;
+	unsigned ttm_pl = *(int *)node->info_ent->data;
 	struct drm_device *dev = node->minor->dev;
 	struct radeon_device *rdev = dev->dev_private;
+	struct drm_mm *mm = (struct drm_mm *)rdev->mman.bdev.man[ttm_pl].priv;
 	int ret;
 	struct ttm_bo_global *glob = rdev->mman.bdev.glob;
 
@@ -850,46 +852,169 @@ static int radeon_mm_dump_table(struct seq_file *m, void *data)
 	spin_unlock(&glob->lru_lock);
 	return ret;
 }
+
+static int ttm_pl_vram = TTM_PL_VRAM;
+static int ttm_pl_tt = TTM_PL_TT;
+
+static struct drm_info_list radeon_ttm_debugfs_list[] = {
+	{"radeon_vram_mm", radeon_mm_dump_table, 0, &ttm_pl_vram},
+	{"radeon_gtt_mm", radeon_mm_dump_table, 0, &ttm_pl_tt},
+	{"ttm_page_pool", ttm_page_alloc_debugfs, 0, NULL},
+#ifdef CONFIG_SWIOTLB
+	{"ttm_dma_page_pool", ttm_dma_page_alloc_debugfs, 0, NULL}
 #endif
+};
 
-static int radeon_ttm_debugfs_init(struct radeon_device *rdev)
+static int radeon_ttm_vram_open(struct inode *inode, struct file *filep)
 {
-#if defined(CONFIG_DEBUG_FS)
-	static struct drm_info_list radeon_mem_types_list[RADEON_DEBUGFS_MEM_TYPES+2];
-	static char radeon_mem_types_names[RADEON_DEBUGFS_MEM_TYPES+2][32];
-	unsigned i;
+	struct radeon_device *rdev = inode->i_private;
+	i_size_write(inode, rdev->mc.mc_vram_size);
+	filep->private_data = inode->i_private;
+	return 0;
+}
 
-	for (i = 0; i < RADEON_DEBUGFS_MEM_TYPES; i++) {
-		if (i == 0)
-			sprintf(radeon_mem_types_names[i], "radeon_vram_mm");
-		else
-			sprintf(radeon_mem_types_names[i], "radeon_gtt_mm");
-		radeon_mem_types_list[i].name = radeon_mem_types_names[i];
-		radeon_mem_types_list[i].show = &radeon_mm_dump_table;
-		radeon_mem_types_list[i].driver_features = 0;
-		if (i == 0)
-			radeon_mem_types_list[i].data = rdev->mman.bdev.man[TTM_PL_VRAM].priv;
-		else
-			radeon_mem_types_list[i].data = rdev->mman.bdev.man[TTM_PL_TT].priv;
+static ssize_t radeon_ttm_vram_read(struct file *f, char __user *buf,
+				    size_t size, loff_t *pos)
+{
+	struct radeon_device *rdev = f->private_data;
+	ssize_t result = 0;
+	int r;
 
+	if (size & 0x3 || *pos & 0x3)
+		return -EINVAL;
+
+	while (size) {
+		unsigned long flags;
+		uint32_t value;
+
+		if (*pos >= rdev->mc.mc_vram_size)
+			return result;
+
+		spin_lock_irqsave(&rdev->mmio_idx_lock, flags);
+		WREG32(RADEON_MM_INDEX, ((uint32_t)*pos) | 0x80000000);
+		if (rdev->family >= CHIP_CEDAR)
+			WREG32(EVERGREEN_MM_INDEX_HI, *pos >> 31);
+		value = RREG32(RADEON_MM_DATA);
+		spin_unlock_irqrestore(&rdev->mmio_idx_lock, flags);
+
+		r = put_user(value, (uint32_t *)buf);
+		if (r)
+			return r;
+
+		result += 4;
+		buf += 4;
+		*pos += 4;
+		size -= 4;
 	}
-	/* Add ttm page pool to debugfs */
-	sprintf(radeon_mem_types_names[i], "ttm_page_pool");
-	radeon_mem_types_list[i].name = radeon_mem_types_names[i];
-	radeon_mem_types_list[i].show = &ttm_page_alloc_debugfs;
-	radeon_mem_types_list[i].driver_features = 0;
-	radeon_mem_types_list[i++].data = NULL;
-#ifdef CONFIG_SWIOTLB
-	if (swiotlb_nr_tbl()) {
-		sprintf(radeon_mem_types_names[i], "ttm_dma_page_pool");
-		radeon_mem_types_list[i].name = radeon_mem_types_names[i];
-		radeon_mem_types_list[i].show = &ttm_dma_page_alloc_debugfs;
-		radeon_mem_types_list[i].driver_features = 0;
-		radeon_mem_types_list[i++].data = NULL;
+
+	return result;
+}
+
+static const struct file_operations radeon_ttm_vram_fops = {
+	.owner = THIS_MODULE,
+	.open = radeon_ttm_vram_open,
+	.read = radeon_ttm_vram_read,
+	.llseek = default_llseek
+};
+
+static int radeon_ttm_gtt_open(struct inode *inode, struct file *filep)
+{
+	struct radeon_device *rdev = inode->i_private;
+	i_size_write(inode, rdev->mc.gtt_size);
+	filep->private_data = inode->i_private;
+	return 0;
+}
+
+static ssize_t radeon_ttm_gtt_read(struct file *f, char __user *buf,
+				   size_t size, loff_t *pos)
+{
+	struct radeon_device *rdev = f->private_data;
+	ssize_t result = 0;
+	int r;
+
+	while (size) {
+		loff_t p = *pos / PAGE_SIZE;
+		unsigned off = *pos & ~PAGE_MASK;
+		ssize_t cur_size = min(size, PAGE_SIZE - off);
+		struct page *page;
+		void *ptr;
+
+		if (p >= rdev->gart.num_cpu_pages)
+			return result;
+
+		page = rdev->gart.pages[p];
+		if (page) {
+			ptr = kmap(page);
+			ptr += off;
+
+			r = copy_to_user(buf, ptr, cur_size);
+			kunmap(rdev->gart.pages[p]);
+		} else
+			r = clear_user(buf, cur_size);
+
+		if (r)
+			return -EFAULT;
+
+		result += cur_size;
+		buf += cur_size;
+		*pos += cur_size;
+		size -= cur_size;
 	}
+
+	return result;
+}
+
+static const struct file_operations radeon_ttm_gtt_fops = {
+	.owner = THIS_MODULE,
+	.open = radeon_ttm_gtt_open,
+	.read = radeon_ttm_gtt_read,
+	.llseek = default_llseek
+};
+
 #endif
-	return radeon_debugfs_add_files(rdev, radeon_mem_types_list, i);
 
+static int radeon_ttm_debugfs_init(struct radeon_device *rdev)
+{
+#if defined(CONFIG_DEBUG_FS)
+	unsigned count;
+
+	struct drm_minor *minor = rdev->ddev->primary;
+	struct dentry *ent, *root = minor->debugfs_root;
+
+	ent = debugfs_create_file("radeon_vram", S_IFREG | S_IRUGO, root,
+				  rdev, &radeon_ttm_vram_fops);
+	if (IS_ERR(ent))
+		return PTR_ERR(ent);
+	rdev->mman.vram = ent;
+
+	ent = debugfs_create_file("radeon_gtt", S_IFREG | S_IRUGO, root,
+				  rdev, &radeon_ttm_gtt_fops);
+	if (IS_ERR(ent))
+		return PTR_ERR(ent);
+	rdev->mman.gtt = ent;
+
+	count = ARRAY_SIZE(radeon_ttm_debugfs_list);
+
+#ifdef CONFIG_SWIOTLB
+	if (!swiotlb_nr_tbl())
+		--count;
 #endif
+
+	return radeon_debugfs_add_files(rdev, radeon_ttm_debugfs_list, count);
+#else
+
 	return 0;
+#endif
+}
+
+static void radeon_ttm_debugfs_fini(struct radeon_device *rdev)
+{
+#if defined(CONFIG_DEBUG_FS)
+
+	debugfs_remove(rdev->mman.vram);
+	rdev->mman.vram = NULL;
+
+	debugfs_remove(rdev->mman.gtt);
+	rdev->mman.gtt = NULL;
+#endif
 }
diff --git a/drivers/gpu/drm/radeon/radeon_uvd.c b/drivers/gpu/drm/radeon/radeon_uvd.c
index b9c0529b4a2e..6781fee1eaad 100644
--- a/drivers/gpu/drm/radeon/radeon_uvd.c
+++ b/drivers/gpu/drm/radeon/radeon_uvd.c
@@ -91,6 +91,7 @@ int radeon_uvd_init(struct radeon_device *rdev)
 	case CHIP_VERDE:
 	case CHIP_PITCAIRN:
 	case CHIP_ARUBA:
+	case CHIP_OLAND:
 		fw_name = FIRMWARE_TAHITI;
 		break;
 
@@ -778,6 +779,8 @@ static void radeon_uvd_idle_work_handler(struct work_struct *work)
 
 	if (radeon_fence_count_emitted(rdev, R600_RING_TYPE_UVD_INDEX) == 0) {
 		if ((rdev->pm.pm_method == PM_METHOD_DPM) && rdev->pm.dpm_enabled) {
+			radeon_uvd_count_handles(rdev, &rdev->pm.dpm.sd,
+						 &rdev->pm.dpm.hd);
 			radeon_dpm_enable_uvd(rdev, false);
 		} else {
 			radeon_set_uvd_clocks(rdev, 0, 0);
diff --git a/drivers/gpu/drm/radeon/rs400.c b/drivers/gpu/drm/radeon/rs400.c
index 9566b5940a5a..b5c2369cda2f 100644
--- a/drivers/gpu/drm/radeon/rs400.c
+++ b/drivers/gpu/drm/radeon/rs400.c
@@ -474,6 +474,8 @@ int rs400_resume(struct radeon_device *rdev)
 	/* Initialize surface registers */
 	radeon_surface_init(rdev);
 
+	radeon_pm_resume(rdev);
+
 	rdev->accel_working = true;
 	r = rs400_startup(rdev);
 	if (r) {
@@ -484,6 +486,7 @@ int rs400_resume(struct radeon_device *rdev)
 
 int rs400_suspend(struct radeon_device *rdev)
 {
+	radeon_pm_suspend(rdev);
 	r100_cp_disable(rdev);
 	radeon_wb_disable(rdev);
 	r100_irq_disable(rdev);
@@ -493,6 +496,7 @@ int rs400_suspend(struct radeon_device *rdev)
 
 void rs400_fini(struct radeon_device *rdev)
 {
+	radeon_pm_fini(rdev);
 	r100_cp_fini(rdev);
 	radeon_wb_fini(rdev);
 	radeon_ib_pool_fini(rdev);
@@ -560,6 +564,9 @@ int rs400_init(struct radeon_device *rdev)
 		return r;
 	r300_set_reg_safe(rdev);
 
+	/* Initialize power management */
+	radeon_pm_init(rdev);
+
 	rdev->accel_working = true;
 	r = rs400_startup(rdev);
 	if (r) {
diff --git a/drivers/gpu/drm/radeon/rs600.c b/drivers/gpu/drm/radeon/rs600.c
index 76cc8d3aafec..fdcde7693032 100644
--- a/drivers/gpu/drm/radeon/rs600.c
+++ b/drivers/gpu/drm/radeon/rs600.c
@@ -1048,6 +1048,8 @@ int rs600_resume(struct radeon_device *rdev)
 	/* Initialize surface registers */
 	radeon_surface_init(rdev);
 
+	radeon_pm_resume(rdev);
+
 	rdev->accel_working = true;
 	r = rs600_startup(rdev);
 	if (r) {
@@ -1058,6 +1060,7 @@ int rs600_resume(struct radeon_device *rdev)
 
 int rs600_suspend(struct radeon_device *rdev)
 {
+	radeon_pm_suspend(rdev);
 	r600_audio_fini(rdev);
 	r100_cp_disable(rdev);
 	radeon_wb_disable(rdev);
@@ -1068,6 +1071,7 @@ int rs600_suspend(struct radeon_device *rdev)
 
 void rs600_fini(struct radeon_device *rdev)
 {
+	radeon_pm_fini(rdev);
 	r600_audio_fini(rdev);
 	r100_cp_fini(rdev);
 	radeon_wb_fini(rdev);
@@ -1136,6 +1140,9 @@ int rs600_init(struct radeon_device *rdev)
 		return r;
 	rs600_set_safe_registers(rdev);
 
+	/* Initialize power management */
+	radeon_pm_init(rdev);
+
 	rdev->accel_working = true;
 	r = rs600_startup(rdev);
 	if (r) {
diff --git a/drivers/gpu/drm/radeon/rs690.c b/drivers/gpu/drm/radeon/rs690.c
index e7dab069cccf..35950738bd5e 100644
--- a/drivers/gpu/drm/radeon/rs690.c
+++ b/drivers/gpu/drm/radeon/rs690.c
@@ -756,6 +756,8 @@ int rs690_resume(struct radeon_device *rdev)
 	/* Initialize surface registers */
 	radeon_surface_init(rdev);
 
+	radeon_pm_resume(rdev);
+
 	rdev->accel_working = true;
 	r = rs690_startup(rdev);
 	if (r) {
@@ -766,6 +768,7 @@ int rs690_resume(struct radeon_device *rdev)
 
 int rs690_suspend(struct radeon_device *rdev)
 {
+	radeon_pm_suspend(rdev);
 	r600_audio_fini(rdev);
 	r100_cp_disable(rdev);
 	radeon_wb_disable(rdev);
@@ -776,6 +779,7 @@ int rs690_suspend(struct radeon_device *rdev)
 
 void rs690_fini(struct radeon_device *rdev)
 {
+	radeon_pm_fini(rdev);
 	r600_audio_fini(rdev);
 	r100_cp_fini(rdev);
 	radeon_wb_fini(rdev);
@@ -845,6 +849,9 @@ int rs690_init(struct radeon_device *rdev)
 		return r;
 	rs600_set_safe_registers(rdev);
 
+	/* Initialize power management */
+	radeon_pm_init(rdev);
+
 	rdev->accel_working = true;
 	r = rs690_startup(rdev);
 	if (r) {
diff --git a/drivers/gpu/drm/radeon/rs780_dpm.c b/drivers/gpu/drm/radeon/rs780_dpm.c
index 6af8505cf4d2..8512085b0aef 100644
--- a/drivers/gpu/drm/radeon/rs780_dpm.c
+++ b/drivers/gpu/drm/radeon/rs780_dpm.c
@@ -623,14 +623,6 @@ int rs780_dpm_enable(struct radeon_device *rdev)
 	if (pi->gfx_clock_gating)
 		r600_gfx_clockgating_enable(rdev, true);
 
-	if (rdev->irq.installed && (rdev->pm.int_thermal_type == THERMAL_TYPE_RV6XX)) {
-		ret = r600_set_thermal_temperature_range(rdev, R600_TEMP_RANGE_MIN, R600_TEMP_RANGE_MAX);
-		if (ret)
-			return ret;
-		rdev->irq.dpm_thermal = true;
-		radeon_irq_set(rdev);
-	}
-
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/radeon/rv515.c b/drivers/gpu/drm/radeon/rv515.c
index 5d1c316115ef..98e8138ff779 100644
--- a/drivers/gpu/drm/radeon/rv515.c
+++ b/drivers/gpu/drm/radeon/rv515.c
@@ -586,6 +586,8 @@ int rv515_resume(struct radeon_device *rdev)
 	/* Initialize surface registers */
 	radeon_surface_init(rdev);
 
+	radeon_pm_resume(rdev);
+
 	rdev->accel_working = true;
 	r =  rv515_startup(rdev);
 	if (r) {
@@ -596,6 +598,7 @@ int rv515_resume(struct radeon_device *rdev)
 
 int rv515_suspend(struct radeon_device *rdev)
 {
+	radeon_pm_suspend(rdev);
 	r100_cp_disable(rdev);
 	radeon_wb_disable(rdev);
 	rs600_irq_disable(rdev);
@@ -612,6 +615,7 @@ void rv515_set_safe_registers(struct radeon_device *rdev)
 
 void rv515_fini(struct radeon_device *rdev)
 {
+	radeon_pm_fini(rdev);
 	r100_cp_fini(rdev);
 	radeon_wb_fini(rdev);
 	radeon_ib_pool_fini(rdev);
@@ -685,6 +689,9 @@ int rv515_init(struct radeon_device *rdev)
 		return r;
 	rv515_set_safe_registers(rdev);
 
+	/* Initialize power management */
+	radeon_pm_init(rdev);
+
 	rdev->accel_working = true;
 	r = rv515_startup(rdev);
 	if (r) {
diff --git a/drivers/gpu/drm/radeon/rv6xx_dpm.c b/drivers/gpu/drm/radeon/rv6xx_dpm.c
index 26633a025252..bebf31c4d841 100644
--- a/drivers/gpu/drm/radeon/rv6xx_dpm.c
+++ b/drivers/gpu/drm/radeon/rv6xx_dpm.c
@@ -1546,7 +1546,6 @@ int rv6xx_dpm_enable(struct radeon_device *rdev)
 {
 	struct rv6xx_power_info *pi = rv6xx_get_pi(rdev);
 	struct radeon_ps *boot_ps = rdev->pm.dpm.boot_ps;
-	int ret;
 
 	if (r600_dynamicpm_enabled(rdev))
 		return -EINVAL;
@@ -1594,15 +1593,6 @@ int rv6xx_dpm_enable(struct radeon_device *rdev)
 	r600_power_level_enable(rdev, R600_POWER_LEVEL_MEDIUM, true);
 	r600_power_level_enable(rdev, R600_POWER_LEVEL_HIGH, true);
 
-	if (rdev->irq.installed &&
-	    r600_is_internal_thermal_sensor(rdev->pm.int_thermal_type)) {
-		ret = r600_set_thermal_temperature_range(rdev, R600_TEMP_RANGE_MIN, R600_TEMP_RANGE_MAX);
-		if (ret)
-			return ret;
-		rdev->irq.dpm_thermal = true;
-		radeon_irq_set(rdev);
-	}
-
 	rv6xx_enable_auto_throttle_source(rdev, RADEON_DPM_AUTO_THROTTLE_SRC_THERMAL, true);
 
 	r600_start_dpm(rdev);
diff --git a/drivers/gpu/drm/radeon/rv770.c b/drivers/gpu/drm/radeon/rv770.c
index 9f5846743c9e..6c772e58c784 100644
--- a/drivers/gpu/drm/radeon/rv770.c
+++ b/drivers/gpu/drm/radeon/rv770.c
@@ -1071,7 +1071,8 @@ static void rv770_mc_program(struct radeon_device *rdev)
  */
 void r700_cp_stop(struct radeon_device *rdev)
 {
-	radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
+	if (rdev->asic->copy.copy_ring_index == RADEON_RING_TYPE_GFX_INDEX)
+		radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
 	WREG32(CP_ME_CNTL, (CP_ME_HALT | CP_PFP_HALT));
 	WREG32(SCRATCH_UMSK, 0);
 	rdev->ring[RADEON_RING_TYPE_GFX_INDEX].ready = false;
@@ -1123,6 +1124,35 @@ void r700_cp_fini(struct radeon_device *rdev)
 	radeon_scratch_free(rdev, ring->rptr_save_reg);
 }
 
+void rv770_set_clk_bypass_mode(struct radeon_device *rdev)
+{
+	u32 tmp, i;
+
+	if (rdev->flags & RADEON_IS_IGP)
+		return;
+
+	tmp = RREG32(CG_SPLL_FUNC_CNTL_2);
+	tmp &= SCLK_MUX_SEL_MASK;
+	tmp |= SCLK_MUX_SEL(1) | SCLK_MUX_UPDATE;
+	WREG32(CG_SPLL_FUNC_CNTL_2, tmp);
+
+	for (i = 0; i < rdev->usec_timeout; i++) {
+		if (RREG32(CG_SPLL_STATUS) & SPLL_CHG_STATUS)
+			break;
+		udelay(1);
+	}
+
+	tmp &= ~SCLK_MUX_UPDATE;
+	WREG32(CG_SPLL_FUNC_CNTL_2, tmp);
+
+	tmp = RREG32(MPLL_CNTL_MODE);
+	if ((rdev->family == CHIP_RV710) || (rdev->family == CHIP_RV730))
+		tmp &= ~RV730_MPLL_MCLK_SEL;
+	else
+		tmp &= ~MPLL_MCLK_SEL;
+	WREG32(MPLL_CNTL_MODE, tmp);
+}
+
 /*
  * Core functions
  */
@@ -1665,14 +1695,6 @@ static int rv770_startup(struct radeon_device *rdev)
 
 	rv770_mc_program(rdev);
 
-	if (!rdev->me_fw || !rdev->pfp_fw || !rdev->rlc_fw) {
-		r = r600_init_microcode(rdev);
-		if (r) {
-			DRM_ERROR("Failed to load firmware!\n");
-			return r;
-		}
-	}
-
 	if (rdev->flags & RADEON_IS_AGP) {
 		rv770_agp_enable(rdev);
 	} else {
@@ -1728,14 +1750,12 @@ static int rv770_startup(struct radeon_device *rdev)
 
 	ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, RADEON_WB_CP_RPTR_OFFSET,
-			     R600_CP_RB_RPTR, R600_CP_RB_WPTR,
 			     RADEON_CP_PACKET2);
 	if (r)
 		return r;
 
 	ring = &rdev->ring[R600_RING_TYPE_DMA_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, R600_WB_DMA_RPTR_OFFSET,
-			     DMA_RB_RPTR, DMA_RB_WPTR,
 			     DMA_PACKET(DMA_PACKET_NOP, 0, 0, 0));
 	if (r)
 		return r;
@@ -1754,7 +1774,6 @@ static int rv770_startup(struct radeon_device *rdev)
 	ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
 	if (ring->ring_size) {
 		r = radeon_ring_init(rdev, ring, ring->ring_size, 0,
-				     UVD_RBC_RB_RPTR, UVD_RBC_RB_WPTR,
 				     RADEON_CP_PACKET2);
 		if (!r)
 			r = uvd_v1_0_init(rdev);
@@ -1792,6 +1811,8 @@ int rv770_resume(struct radeon_device *rdev)
 	/* init golden registers */
 	rv770_init_golden_registers(rdev);
 
+	radeon_pm_resume(rdev);
+
 	rdev->accel_working = true;
 	r = rv770_startup(rdev);
 	if (r) {
@@ -1806,6 +1827,7 @@ int rv770_resume(struct radeon_device *rdev)
 
 int rv770_suspend(struct radeon_device *rdev)
 {
+	radeon_pm_suspend(rdev);
 	r600_audio_fini(rdev);
 	uvd_v1_0_fini(rdev);
 	radeon_uvd_suspend(rdev);
@@ -1876,6 +1898,17 @@ int rv770_init(struct radeon_device *rdev)
 	if (r)
 		return r;
 
+	if (!rdev->me_fw || !rdev->pfp_fw || !rdev->rlc_fw) {
+		r = r600_init_microcode(rdev);
+		if (r) {
+			DRM_ERROR("Failed to load firmware!\n");
+			return r;
+		}
+	}
+
+	/* Initialize power management */
+	radeon_pm_init(rdev);
+
 	rdev->ring[RADEON_RING_TYPE_GFX_INDEX].ring_obj = NULL;
 	r600_ring_init(rdev, &rdev->ring[RADEON_RING_TYPE_GFX_INDEX], 1024 * 1024);
 
@@ -1915,6 +1948,7 @@ int rv770_init(struct radeon_device *rdev)
 
 void rv770_fini(struct radeon_device *rdev)
 {
+	radeon_pm_fini(rdev);
 	r700_cp_fini(rdev);
 	r600_dma_fini(rdev);
 	r600_irq_fini(rdev);
diff --git a/drivers/gpu/drm/radeon/rv770_dpm.c b/drivers/gpu/drm/radeon/rv770_dpm.c
index 374499db20c7..80c595aba359 100644
--- a/drivers/gpu/drm/radeon/rv770_dpm.c
+++ b/drivers/gpu/drm/radeon/rv770_dpm.c
@@ -1863,8 +1863,8 @@ void rv770_enable_auto_throttle_source(struct radeon_device *rdev,
 	}
 }
 
-int rv770_set_thermal_temperature_range(struct radeon_device *rdev,
-					int min_temp, int max_temp)
+static int rv770_set_thermal_temperature_range(struct radeon_device *rdev,
+					       int min_temp, int max_temp)
 {
 	int low_temp = 0 * 1000;
 	int high_temp = 255 * 1000;
@@ -1966,6 +1966,15 @@ int rv770_dpm_enable(struct radeon_device *rdev)
 	if (pi->mg_clock_gating)
 		rv770_mg_clock_gating_enable(rdev, true);
 
+	rv770_enable_auto_throttle_source(rdev, RADEON_DPM_AUTO_THROTTLE_SRC_THERMAL, true);
+
+	return 0;
+}
+
+int rv770_dpm_late_enable(struct radeon_device *rdev)
+{
+	int ret;
+
 	if (rdev->irq.installed &&
 	    r600_is_internal_thermal_sensor(rdev->pm.int_thermal_type)) {
 		PPSMC_Result result;
@@ -1981,8 +1990,6 @@ int rv770_dpm_enable(struct radeon_device *rdev)
 			DRM_DEBUG_KMS("Could not enable thermal interrupts.\n");
 	}
 
-	rv770_enable_auto_throttle_source(rdev, RADEON_DPM_AUTO_THROTTLE_SRC_THERMAL, true);
-
 	return 0;
 }
 
@@ -2244,14 +2251,12 @@ static void rv7xx_parse_pplib_clock_info(struct radeon_device *rdev,
 		pl->vddci = vddci;
 	}
 
-	if (rdev->family >= CHIP_BARTS) {
-		if ((rps->class & ATOM_PPLIB_CLASSIFICATION_UI_MASK) ==
-		    ATOM_PPLIB_CLASSIFICATION_UI_PERFORMANCE) {
-			rdev->pm.dpm.dyn_state.max_clock_voltage_on_ac.sclk = pl->sclk;
-			rdev->pm.dpm.dyn_state.max_clock_voltage_on_ac.mclk = pl->mclk;
-			rdev->pm.dpm.dyn_state.max_clock_voltage_on_ac.vddc = pl->vddc;
-			rdev->pm.dpm.dyn_state.max_clock_voltage_on_ac.vddci = pl->vddci;
-		}
+	if ((rps->class & ATOM_PPLIB_CLASSIFICATION_UI_MASK) ==
+	    ATOM_PPLIB_CLASSIFICATION_UI_PERFORMANCE) {
+		rdev->pm.dpm.dyn_state.max_clock_voltage_on_ac.sclk = pl->sclk;
+		rdev->pm.dpm.dyn_state.max_clock_voltage_on_ac.mclk = pl->mclk;
+		rdev->pm.dpm.dyn_state.max_clock_voltage_on_ac.vddc = pl->vddc;
+		rdev->pm.dpm.dyn_state.max_clock_voltage_on_ac.vddci = pl->vddci;
 	}
 }
 
@@ -2531,6 +2536,12 @@ bool rv770_dpm_vblank_too_short(struct radeon_device *rdev)
 	    (rdev->pdev->subsystem_device == 0x1c42))
 		switch_limit = 200;
 
+	/* RV770 */
+	/* mclk switching doesn't seem to work reliably on desktop RV770s */
+	if ((rdev->family == CHIP_RV770) &&
+	    !(rdev->flags & RADEON_IS_MOBILITY))
+		switch_limit = 0xffffffff; /* disable mclk switching */
+
 	if (vblank_time < switch_limit)
 		return true;
 	else
diff --git a/drivers/gpu/drm/radeon/rv770_dpm.h b/drivers/gpu/drm/radeon/rv770_dpm.h
index 9244effc6b59..f776634840c9 100644
--- a/drivers/gpu/drm/radeon/rv770_dpm.h
+++ b/drivers/gpu/drm/radeon/rv770_dpm.h
@@ -283,8 +283,4 @@ int rv770_read_smc_soft_register(struct radeon_device *rdev,
 int rv770_write_smc_soft_register(struct radeon_device *rdev,
 				  u16 reg_offset, u32 value);
 
-/* thermal */
-int rv770_set_thermal_temperature_range(struct radeon_device *rdev,
-					int min_temp, int max_temp);
-
 #endif
diff --git a/drivers/gpu/drm/radeon/rv770d.h b/drivers/gpu/drm/radeon/rv770d.h
index 1ae277152cc7..3cf1e2921545 100644
--- a/drivers/gpu/drm/radeon/rv770d.h
+++ b/drivers/gpu/drm/radeon/rv770d.h
@@ -100,14 +100,21 @@
 #define	CG_SPLL_FUNC_CNTL_2				0x604
 #define		SCLK_MUX_SEL(x)				((x) << 0)
 #define		SCLK_MUX_SEL_MASK			(0x1ff << 0)
+#define		SCLK_MUX_UPDATE				(1 << 26)
 #define	CG_SPLL_FUNC_CNTL_3				0x608
 #define		SPLL_FB_DIV(x)				((x) << 0)
 #define		SPLL_FB_DIV_MASK			(0x3ffffff << 0)
 #define		SPLL_DITHEN				(1 << 28)
+#define	CG_SPLL_STATUS					0x60c
+#define		SPLL_CHG_STATUS				(1 << 1)
 
 #define	SPLL_CNTL_MODE					0x610
 #define		SPLL_DIV_SYNC				(1 << 5)
 
+#define MPLL_CNTL_MODE                                  0x61c
+#       define MPLL_MCLK_SEL                            (1 << 11)
+#       define RV730_MPLL_MCLK_SEL                      (1 << 25)
+
 #define	MPLL_AD_FUNC_CNTL				0x624
 #define		CLKF(x)					((x) << 0)
 #define		CLKF_MASK				(0x7f << 0)
diff --git a/drivers/gpu/drm/radeon/si.c b/drivers/gpu/drm/radeon/si.c
index 85e1edfaa3be..09ec4f6c53bb 100644
--- a/drivers/gpu/drm/radeon/si.c
+++ b/drivers/gpu/drm/radeon/si.c
@@ -80,6 +80,8 @@ extern void evergreen_print_gpu_status_regs(struct radeon_device *rdev);
 extern bool evergreen_is_display_hung(struct radeon_device *rdev);
 static void si_enable_gui_idle_interrupt(struct radeon_device *rdev,
 					 bool enable);
+static void si_init_pg(struct radeon_device *rdev);
+static void si_init_cg(struct radeon_device *rdev);
 static void si_fini_pg(struct radeon_device *rdev);
 static void si_fini_cg(struct radeon_device *rdev);
 static void si_rlc_stop(struct radeon_device *rdev);
@@ -1460,7 +1462,7 @@ static const u32 hainan_io_mc_regs[TAHITI_IO_MC_REGS_SIZE][2] = {
 };
 
 /* ucode loading */
-static int si_mc_load_microcode(struct radeon_device *rdev)
+int si_mc_load_microcode(struct radeon_device *rdev)
 {
 	const __be32 *fw_data;
 	u32 running, blackout = 0;
@@ -3247,7 +3249,8 @@ static void si_cp_enable(struct radeon_device *rdev, bool enable)
 	if (enable)
 		WREG32(CP_ME_CNTL, 0);
 	else {
-		radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
+		if (rdev->asic->copy.copy_ring_index == RADEON_RING_TYPE_GFX_INDEX)
+			radeon_ttm_set_active_vram_size(rdev, rdev->mc.visible_vram_size);
 		WREG32(CP_ME_CNTL, (CP_ME_HALT | CP_PFP_HALT | CP_CE_HALT));
 		WREG32(SCRATCH_UMSK, 0);
 		rdev->ring[RADEON_RING_TYPE_GFX_INDEX].ready = false;
@@ -3508,6 +3511,9 @@ static int si_cp_resume(struct radeon_device *rdev)
 
 	si_enable_gui_idle_interrupt(rdev, true);
 
+	if (rdev->asic->copy.copy_ring_index == RADEON_RING_TYPE_GFX_INDEX)
+		radeon_ttm_set_active_vram_size(rdev, rdev->mc.real_vram_size);
+
 	return 0;
 }
 
@@ -3724,6 +3730,106 @@ static void si_gpu_soft_reset(struct radeon_device *rdev, u32 reset_mask)
 	evergreen_print_gpu_status_regs(rdev);
 }
 
+static void si_set_clk_bypass_mode(struct radeon_device *rdev)
+{
+	u32 tmp, i;
+
+	tmp = RREG32(CG_SPLL_FUNC_CNTL);
+	tmp |= SPLL_BYPASS_EN;
+	WREG32(CG_SPLL_FUNC_CNTL, tmp);
+
+	tmp = RREG32(CG_SPLL_FUNC_CNTL_2);
+	tmp |= SPLL_CTLREQ_CHG;
+	WREG32(CG_SPLL_FUNC_CNTL_2, tmp);
+
+	for (i = 0; i < rdev->usec_timeout; i++) {
+		if (RREG32(SPLL_STATUS) & SPLL_CHG_STATUS)
+			break;
+		udelay(1);
+	}
+
+	tmp = RREG32(CG_SPLL_FUNC_CNTL_2);
+	tmp &= ~(SPLL_CTLREQ_CHG | SCLK_MUX_UPDATE);
+	WREG32(CG_SPLL_FUNC_CNTL_2, tmp);
+
+	tmp = RREG32(MPLL_CNTL_MODE);
+	tmp &= ~MPLL_MCLK_SEL;
+	WREG32(MPLL_CNTL_MODE, tmp);
+}
+
+static void si_spll_powerdown(struct radeon_device *rdev)
+{
+	u32 tmp;
+
+	tmp = RREG32(SPLL_CNTL_MODE);
+	tmp |= SPLL_SW_DIR_CONTROL;
+	WREG32(SPLL_CNTL_MODE, tmp);
+
+	tmp = RREG32(CG_SPLL_FUNC_CNTL);
+	tmp |= SPLL_RESET;
+	WREG32(CG_SPLL_FUNC_CNTL, tmp);
+
+	tmp = RREG32(CG_SPLL_FUNC_CNTL);
+	tmp |= SPLL_SLEEP;
+	WREG32(CG_SPLL_FUNC_CNTL, tmp);
+
+	tmp = RREG32(SPLL_CNTL_MODE);
+	tmp &= ~SPLL_SW_DIR_CONTROL;
+	WREG32(SPLL_CNTL_MODE, tmp);
+}
+
+static void si_gpu_pci_config_reset(struct radeon_device *rdev)
+{
+	struct evergreen_mc_save save;
+	u32 tmp, i;
+
+	dev_info(rdev->dev, "GPU pci config reset\n");
+
+	/* disable dpm? */
+
+	/* disable cg/pg */
+	si_fini_pg(rdev);
+	si_fini_cg(rdev);
+
+	/* Disable CP parsing/prefetching */
+	WREG32(CP_ME_CNTL, CP_ME_HALT | CP_PFP_HALT | CP_CE_HALT);
+	/* dma0 */
+	tmp = RREG32(DMA_RB_CNTL + DMA0_REGISTER_OFFSET);
+	tmp &= ~DMA_RB_ENABLE;
+	WREG32(DMA_RB_CNTL + DMA0_REGISTER_OFFSET, tmp);
+	/* dma1 */
+	tmp = RREG32(DMA_RB_CNTL + DMA1_REGISTER_OFFSET);
+	tmp &= ~DMA_RB_ENABLE;
+	WREG32(DMA_RB_CNTL + DMA1_REGISTER_OFFSET, tmp);
+	/* XXX other engines? */
+
+	/* halt the rlc, disable cp internal ints */
+	si_rlc_stop(rdev);
+
+	udelay(50);
+
+	/* disable mem access */
+	evergreen_mc_stop(rdev, &save);
+	if (evergreen_mc_wait_for_idle(rdev)) {
+		dev_warn(rdev->dev, "Wait for MC idle timed out !\n");
+	}
+
+	/* set mclk/sclk to bypass */
+	si_set_clk_bypass_mode(rdev);
+	/* powerdown spll */
+	si_spll_powerdown(rdev);
+	/* disable BM */
+	pci_clear_master(rdev->pdev);
+	/* reset */
+	radeon_pci_config_reset(rdev);
+	/* wait for asic to come out of reset */
+	for (i = 0; i < rdev->usec_timeout; i++) {
+		if (RREG32(CONFIG_MEMSIZE) != 0xffffffff)
+			break;
+		udelay(1);
+	}
+}
+
 int si_asic_reset(struct radeon_device *rdev)
 {
 	u32 reset_mask;
@@ -3733,10 +3839,17 @@ int si_asic_reset(struct radeon_device *rdev)
 	if (reset_mask)
 		r600_set_bios_scratch_engine_hung(rdev, true);
 
+	/* try soft reset */
 	si_gpu_soft_reset(rdev, reset_mask);
 
 	reset_mask = si_gpu_check_soft_reset(rdev);
 
+	/* try pci config reset */
+	if (reset_mask && radeon_hard_reset)
+		si_gpu_pci_config_reset(rdev);
+
+	reset_mask = si_gpu_check_soft_reset(rdev);
+
 	if (!reset_mask)
 		r600_set_bios_scratch_engine_hung(rdev, false);
 
@@ -5212,8 +5325,8 @@ static void si_enable_hdp_ls(struct radeon_device *rdev,
 		WREG32(HDP_MEM_POWER_LS, data);
 }
 
-void si_update_cg(struct radeon_device *rdev,
-		  u32 block, bool enable)
+static void si_update_cg(struct radeon_device *rdev,
+			 u32 block, bool enable)
 {
 	if (block & RADEON_CG_BLOCK_GFX) {
 		si_enable_gui_idle_interrupt(rdev, false);
@@ -5379,6 +5492,9 @@ static void si_init_pg(struct radeon_device *rdev)
 		si_init_ao_cu_mask(rdev);
 		if (rdev->pg_flags & RADEON_PG_SUPPORT_GFX_PG) {
 			si_init_gfx_cgpg(rdev);
+		} else {
+			WREG32(RLC_SAVE_AND_RESTORE_BASE, rdev->rlc.save_restore_gpu_addr >> 8);
+			WREG32(RLC_CLEAR_STATE_RESTORE_BASE, rdev->rlc.clear_state_gpu_addr >> 8);
 		}
 		si_enable_dma_pg(rdev, true);
 		si_enable_gfx_cgpg(rdev, true);
@@ -5566,7 +5682,7 @@ static void si_disable_interrupt_state(struct radeon_device *rdev)
 	}
 
 	if (!ASIC_IS_NODCE(rdev)) {
-		WREG32(DACA_AUTODETECT_INT_CONTROL, 0);
+		WREG32(DAC_AUTODETECT_INT_CONTROL, 0);
 
 		tmp = RREG32(DC_HPD1_INT_CONTROL) & DC_HPDx_INT_POLARITY;
 		WREG32(DC_HPD1_INT_CONTROL, tmp);
@@ -6324,21 +6440,14 @@ static int si_startup(struct radeon_device *rdev)
 
 	si_mc_program(rdev);
 
-	if (!rdev->me_fw || !rdev->pfp_fw || !rdev->ce_fw ||
-	    !rdev->rlc_fw || !rdev->mc_fw) {
-		r = si_init_microcode(rdev);
+	if (!rdev->pm.dpm_enabled) {
+		r = si_mc_load_microcode(rdev);
 		if (r) {
-			DRM_ERROR("Failed to load firmware!\n");
+			DRM_ERROR("Failed to load MC firmware!\n");
 			return r;
 		}
 	}
 
-	r = si_mc_load_microcode(rdev);
-	if (r) {
-		DRM_ERROR("Failed to load MC firmware!\n");
-		return r;
-	}
-
 	r = si_pcie_gart_enable(rdev);
 	if (r)
 		return r;
@@ -6421,37 +6530,30 @@ static int si_startup(struct radeon_device *rdev)
 
 	ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, RADEON_WB_CP_RPTR_OFFSET,
-			     CP_RB0_RPTR, CP_RB0_WPTR,
 			     RADEON_CP_PACKET2);
 	if (r)
 		return r;
 
 	ring = &rdev->ring[CAYMAN_RING_TYPE_CP1_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, RADEON_WB_CP1_RPTR_OFFSET,
-			     CP_RB1_RPTR, CP_RB1_WPTR,
 			     RADEON_CP_PACKET2);
 	if (r)
 		return r;
 
 	ring = &rdev->ring[CAYMAN_RING_TYPE_CP2_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, RADEON_WB_CP2_RPTR_OFFSET,
-			     CP_RB2_RPTR, CP_RB2_WPTR,
 			     RADEON_CP_PACKET2);
 	if (r)
 		return r;
 
 	ring = &rdev->ring[R600_RING_TYPE_DMA_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, R600_WB_DMA_RPTR_OFFSET,
-			     DMA_RB_RPTR + DMA0_REGISTER_OFFSET,
-			     DMA_RB_WPTR + DMA0_REGISTER_OFFSET,
 			     DMA_PACKET(DMA_PACKET_NOP, 0, 0, 0, 0));
 	if (r)
 		return r;
 
 	ring = &rdev->ring[CAYMAN_RING_TYPE_DMA1_INDEX];
 	r = radeon_ring_init(rdev, ring, ring->ring_size, CAYMAN_WB_DMA1_RPTR_OFFSET,
-			     DMA_RB_RPTR + DMA1_REGISTER_OFFSET,
-			     DMA_RB_WPTR + DMA1_REGISTER_OFFSET,
 			     DMA_PACKET(DMA_PACKET_NOP, 0, 0, 0, 0));
 	if (r)
 		return r;
@@ -6471,7 +6573,6 @@ static int si_startup(struct radeon_device *rdev)
 		ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
 		if (ring->ring_size) {
 			r = radeon_ring_init(rdev, ring, ring->ring_size, 0,
-					     UVD_RBC_RB_RPTR, UVD_RBC_RB_WPTR,
 					     RADEON_CP_PACKET2);
 			if (!r)
 				r = uvd_v1_0_init(rdev);
@@ -6513,6 +6614,8 @@ int si_resume(struct radeon_device *rdev)
 	/* init golden registers */
 	si_init_golden_registers(rdev);
 
+	radeon_pm_resume(rdev);
+
 	rdev->accel_working = true;
 	r = si_startup(rdev);
 	if (r) {
@@ -6527,6 +6630,7 @@ int si_resume(struct radeon_device *rdev)
 
 int si_suspend(struct radeon_device *rdev)
 {
+	radeon_pm_suspend(rdev);
 	dce6_audio_fini(rdev);
 	radeon_vm_manager_fini(rdev);
 	si_cp_enable(rdev, false);
@@ -6600,6 +6704,18 @@ int si_init(struct radeon_device *rdev)
 	if (r)
 		return r;
 
+	if (!rdev->me_fw || !rdev->pfp_fw || !rdev->ce_fw ||
+	    !rdev->rlc_fw || !rdev->mc_fw) {
+		r = si_init_microcode(rdev);
+		if (r) {
+			DRM_ERROR("Failed to load firmware!\n");
+			return r;
+		}
+	}
+
+	/* Initialize power management */
+	radeon_pm_init(rdev);
+
 	ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX];
 	ring->ring_obj = NULL;
 	r600_ring_init(rdev, ring, 1024 * 1024);
@@ -6666,6 +6782,7 @@ int si_init(struct radeon_device *rdev)
 
 void si_fini(struct radeon_device *rdev)
 {
+	radeon_pm_fini(rdev);
 	si_cp_fini(rdev);
 	cayman_dma_fini(rdev);
 	si_fini_pg(rdev);
diff --git a/drivers/gpu/drm/radeon/si_dpm.c b/drivers/gpu/drm/radeon/si_dpm.c
index 0b00c790fb77..0471501338fb 100644
--- a/drivers/gpu/drm/radeon/si_dpm.c
+++ b/drivers/gpu/drm/radeon/si_dpm.c
@@ -1738,6 +1738,8 @@ struct evergreen_power_info *evergreen_get_pi(struct radeon_device *rdev);
 struct ni_power_info *ni_get_pi(struct radeon_device *rdev);
 struct ni_ps *ni_get_ps(struct radeon_ps *rps);
 
+extern int si_mc_load_microcode(struct radeon_device *rdev);
+
 static int si_populate_voltage_value(struct radeon_device *rdev,
 				     const struct atom_voltage_table *table,
 				     u16 value, SISLANDS_SMC_VOLTAGE_VALUE *voltage);
@@ -1753,9 +1755,6 @@ static int si_calculate_sclk_params(struct radeon_device *rdev,
 				    u32 engine_clock,
 				    SISLANDS_SMC_SCLK_VALUE *sclk);
 
-extern void si_update_cg(struct radeon_device *rdev,
-			 u32 block, bool enable);
-
 static struct si_power_info *si_get_pi(struct radeon_device *rdev)
 {
         struct si_power_info *pi = rdev->pm.dpm.priv;
@@ -2396,7 +2395,7 @@ static int si_populate_sq_ramping_values(struct radeon_device *rdev,
 	if (SISLANDS_DPM2_SQ_RAMP_STI_SIZE > (STI_SIZE_MASK >> STI_SIZE_SHIFT))
 		enable_sq_ramping = false;
 
-	if (NISLANDS_DPM2_SQ_RAMP_LTI_RATIO <= (LTI_RATIO_MASK >> LTI_RATIO_SHIFT))
+	if (SISLANDS_DPM2_SQ_RAMP_LTI_RATIO <= (LTI_RATIO_MASK >> LTI_RATIO_SHIFT))
 		enable_sq_ramping = false;
 
 	for (i = 0; i < state->performance_level_count; i++) {
@@ -3591,10 +3590,9 @@ static void si_program_display_gap(struct radeon_device *rdev)
 
 	/* Setting this to false forces the performance state to low if the crtcs are disabled.
 	 * This can be a problem on PowerXpress systems or if you want to use the card
-	 * for offscreen rendering or compute if there are no crtcs enabled.  Set it to
-	 * true for now so that performance scales even if the displays are off.
+	 * for offscreen rendering or compute if there are no crtcs enabled.
 	 */
-	si_notify_smc_display_change(rdev, true /*rdev->pm.dpm.new_active_crtc_count > 0*/);
+	si_notify_smc_display_change(rdev, rdev->pm.dpm.new_active_crtc_count > 0);
 }
 
 static void si_enable_spread_spectrum(struct radeon_device *rdev, bool enable)
@@ -5414,7 +5412,7 @@ static void si_populate_mc_reg_addresses(struct radeon_device *rdev,
 
 	for (i = 0, j = 0; j < si_pi->mc_reg_table.last; j++) {
 		if (si_pi->mc_reg_table.valid_flag & (1 << j)) {
-			if (i >= SMC_NISLANDS_MC_REGISTER_ARRAY_SIZE)
+			if (i >= SMC_SISLANDS_MC_REGISTER_ARRAY_SIZE)
 				break;
 			mc_reg_table->address[i].s0 =
 				cpu_to_be16(si_pi->mc_reg_table.mc_reg_address[j].s0);
@@ -5754,6 +5752,11 @@ static void si_set_pcie_lane_width_in_smc(struct radeon_device *rdev,
 
 void si_dpm_setup_asic(struct radeon_device *rdev)
 {
+	int r;
+
+	r = si_mc_load_microcode(rdev);
+	if (r)
+		DRM_ERROR("Failed to load MC firmware!\n");
 	rv770_get_memory_type(rdev);
 	si_read_clock_registers(rdev);
 	si_enable_acpi_power_management(rdev);
@@ -5791,13 +5794,6 @@ int si_dpm_enable(struct radeon_device *rdev)
 	struct radeon_ps *boot_ps = rdev->pm.dpm.boot_ps;
 	int ret;
 
-	si_update_cg(rdev, (RADEON_CG_BLOCK_GFX |
-			    RADEON_CG_BLOCK_MC |
-			    RADEON_CG_BLOCK_SDMA |
-			    RADEON_CG_BLOCK_BIF |
-			    RADEON_CG_BLOCK_UVD |
-			    RADEON_CG_BLOCK_HDP), false);
-
 	if (si_is_smc_running(rdev))
 		return -EINVAL;
 	if (pi->voltage_control)
@@ -5900,6 +5896,17 @@ int si_dpm_enable(struct radeon_device *rdev)
 	si_enable_sclk_control(rdev, true);
 	si_start_dpm(rdev);
 
+	si_enable_auto_throttle_source(rdev, RADEON_DPM_AUTO_THROTTLE_SRC_THERMAL, true);
+
+	ni_update_current_ps(rdev, boot_ps);
+
+	return 0;
+}
+
+int si_dpm_late_enable(struct radeon_device *rdev)
+{
+	int ret;
+
 	if (rdev->irq.installed &&
 	    r600_is_internal_thermal_sensor(rdev->pm.int_thermal_type)) {
 		PPSMC_Result result;
@@ -5915,17 +5922,6 @@ int si_dpm_enable(struct radeon_device *rdev)
 			DRM_DEBUG_KMS("Could not enable thermal interrupts.\n");
 	}
 
-	si_enable_auto_throttle_source(rdev, RADEON_DPM_AUTO_THROTTLE_SRC_THERMAL, true);
-
-	si_update_cg(rdev, (RADEON_CG_BLOCK_GFX |
-			    RADEON_CG_BLOCK_MC |
-			    RADEON_CG_BLOCK_SDMA |
-			    RADEON_CG_BLOCK_BIF |
-			    RADEON_CG_BLOCK_UVD |
-			    RADEON_CG_BLOCK_HDP), true);
-
-	ni_update_current_ps(rdev, boot_ps);
-
 	return 0;
 }
 
@@ -5934,13 +5930,6 @@ void si_dpm_disable(struct radeon_device *rdev)
 	struct rv7xx_power_info *pi = rv770_get_pi(rdev);
 	struct radeon_ps *boot_ps = rdev->pm.dpm.boot_ps;
 
-	si_update_cg(rdev, (RADEON_CG_BLOCK_GFX |
-			    RADEON_CG_BLOCK_MC |
-			    RADEON_CG_BLOCK_SDMA |
-			    RADEON_CG_BLOCK_BIF |
-			    RADEON_CG_BLOCK_UVD |
-			    RADEON_CG_BLOCK_HDP), false);
-
 	if (!si_is_smc_running(rdev))
 		return;
 	si_disable_ulv(rdev);
@@ -6005,13 +5994,6 @@ int si_dpm_set_power_state(struct radeon_device *rdev)
 	struct radeon_ps *old_ps = &eg_pi->current_rps;
 	int ret;
 
-	si_update_cg(rdev, (RADEON_CG_BLOCK_GFX |
-			    RADEON_CG_BLOCK_MC |
-			    RADEON_CG_BLOCK_SDMA |
-			    RADEON_CG_BLOCK_BIF |
-			    RADEON_CG_BLOCK_UVD |
-			    RADEON_CG_BLOCK_HDP), false);
-
 	ret = si_disable_ulv(rdev);
 	if (ret) {
 		DRM_ERROR("si_disable_ulv failed\n");
@@ -6104,13 +6086,6 @@ int si_dpm_set_power_state(struct radeon_device *rdev)
 		return ret;
 	}
 
-	si_update_cg(rdev, (RADEON_CG_BLOCK_GFX |
-			    RADEON_CG_BLOCK_MC |
-			    RADEON_CG_BLOCK_SDMA |
-			    RADEON_CG_BLOCK_BIF |
-			    RADEON_CG_BLOCK_UVD |
-			    RADEON_CG_BLOCK_HDP), true);
-
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/radeon/si_smc.c b/drivers/gpu/drm/radeon/si_smc.c
index d422a1cbf727..e80efcf0c230 100644
--- a/drivers/gpu/drm/radeon/si_smc.c
+++ b/drivers/gpu/drm/radeon/si_smc.c
@@ -28,6 +28,7 @@
 #include "sid.h"
 #include "ppsmc.h"
 #include "radeon_ucode.h"
+#include "sislands_smc.h"
 
 static int si_set_smc_sram_address(struct radeon_device *rdev,
 				   u32 smc_address, u32 limit)
diff --git a/drivers/gpu/drm/radeon/sid.h b/drivers/gpu/drm/radeon/sid.h
index b322acc48097..9239a6d29128 100644
--- a/drivers/gpu/drm/radeon/sid.h
+++ b/drivers/gpu/drm/radeon/sid.h
@@ -94,6 +94,8 @@
 #define	CG_SPLL_FUNC_CNTL_2				0x604
 #define		SCLK_MUX_SEL(x)				((x) << 0)
 #define		SCLK_MUX_SEL_MASK			(0x1ff << 0)
+#define		SPLL_CTLREQ_CHG				(1 << 23)
+#define		SCLK_MUX_UPDATE				(1 << 26)
 #define	CG_SPLL_FUNC_CNTL_3				0x608
 #define		SPLL_FB_DIV(x)				((x) << 0)
 #define		SPLL_FB_DIV_MASK			(0x3ffffff << 0)
@@ -101,7 +103,10 @@
 #define		SPLL_DITHEN				(1 << 28)
 #define	CG_SPLL_FUNC_CNTL_4				0x60c
 
+#define	SPLL_STATUS					0x614
+#define		SPLL_CHG_STATUS				(1 << 1)
 #define	SPLL_CNTL_MODE					0x618
+#define		SPLL_SW_DIR_CONTROL			(1 << 0)
 #	define SPLL_REFCLK_SEL(x)			((x) << 8)
 #	define SPLL_REFCLK_SEL_MASK			0xFF00
 
@@ -559,6 +564,8 @@
 #       define MRDCK0_BYPASS                            (1 << 24)
 #       define MRDCK1_BYPASS                            (1 << 25)
 
+#define	MPLL_CNTL_MODE					0x2bb0
+#       define MPLL_MCLK_SEL                            (1 << 11)
 #define	MPLL_FUNC_CNTL					0x2bb4
 #define		BWCTRL(x)				((x) << 20)
 #define		BWCTRL_MASK				(0xff << 20)
@@ -815,7 +822,7 @@
 #       define GRPH_PFLIP_INT_MASK                      (1 << 0)
 #       define GRPH_PFLIP_INT_TYPE                      (1 << 8)
 
-#define	DACA_AUTODETECT_INT_CONTROL			0x66c8
+#define	DAC_AUTODETECT_INT_CONTROL			0x67c8
 
 #define DC_HPD1_INT_STATUS                              0x601c
 #define DC_HPD2_INT_STATUS                              0x6028
diff --git a/drivers/gpu/drm/radeon/sislands_smc.h b/drivers/gpu/drm/radeon/sislands_smc.h
index 5578e9837026..10e945a49479 100644
--- a/drivers/gpu/drm/radeon/sislands_smc.h
+++ b/drivers/gpu/drm/radeon/sislands_smc.h
@@ -374,8 +374,6 @@ typedef struct Smc_SIslands_DTE_Configuration Smc_SIslands_DTE_Configuration;
 
 #pragma pack(pop)
 
-int si_set_smc_sram_address(struct radeon_device *rdev,
-			    u32 smc_address, u32 limit);
 int si_copy_bytes_to_smc(struct radeon_device *rdev,
 			 u32 smc_start_address,
 			 const u8 *src, u32 byte_count, u32 limit);
diff --git a/drivers/gpu/drm/radeon/sumo_dpm.c b/drivers/gpu/drm/radeon/sumo_dpm.c
index 96ea6db8bf57..f121efe12dc5 100644
--- a/drivers/gpu/drm/radeon/sumo_dpm.c
+++ b/drivers/gpu/drm/radeon/sumo_dpm.c
@@ -71,7 +71,7 @@ static const u32 sumo_dtc[SUMO_PM_NUMBER_OF_TC] =
 	SUMO_DTC_DFLT_14,
 };
 
-struct sumo_ps *sumo_get_ps(struct radeon_ps *rps)
+static struct sumo_ps *sumo_get_ps(struct radeon_ps *rps)
 {
 	struct sumo_ps *ps = rps->ps_priv;
 
@@ -1202,14 +1202,10 @@ static void sumo_update_requested_ps(struct radeon_device *rdev,
 int sumo_dpm_enable(struct radeon_device *rdev)
 {
 	struct sumo_power_info *pi = sumo_get_pi(rdev);
-	int ret;
 
 	if (sumo_dpm_enabled(rdev))
 		return -EINVAL;
 
-	ret = sumo_enable_clock_power_gating(rdev);
-	if (ret)
-		return ret;
 	sumo_program_bootup_state(rdev);
 	sumo_init_bsp(rdev);
 	sumo_reset_am(rdev);
@@ -1233,6 +1229,19 @@ int sumo_dpm_enable(struct radeon_device *rdev)
 	if (pi->enable_boost)
 		sumo_enable_boost_timer(rdev);
 
+	sumo_update_current_ps(rdev, rdev->pm.dpm.boot_ps);
+
+	return 0;
+}
+
+int sumo_dpm_late_enable(struct radeon_device *rdev)
+{
+	int ret;
+
+	ret = sumo_enable_clock_power_gating(rdev);
+	if (ret)
+		return ret;
+
 	if (rdev->irq.installed &&
 	    r600_is_internal_thermal_sensor(rdev->pm.int_thermal_type)) {
 		ret = sumo_set_thermal_temperature_range(rdev, R600_TEMP_RANGE_MIN, R600_TEMP_RANGE_MAX);
@@ -1242,8 +1251,6 @@ int sumo_dpm_enable(struct radeon_device *rdev)
 		radeon_irq_set(rdev);
 	}
 
-	sumo_update_current_ps(rdev, rdev->pm.dpm.boot_ps);
-
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/radeon/sumo_smc.c b/drivers/gpu/drm/radeon/sumo_smc.c
index 18abba5b5810..fb081d2ae374 100644
--- a/drivers/gpu/drm/radeon/sumo_smc.c
+++ b/drivers/gpu/drm/radeon/sumo_smc.c
@@ -31,7 +31,6 @@
 #define SUMO_SMU_SERVICE_ROUTINE_ALTVDDNB_NOTIFY  27
 #define SUMO_SMU_SERVICE_ROUTINE_GFX_SRV_ID_20  20
 
-struct sumo_ps *sumo_get_ps(struct radeon_ps *rps);
 struct sumo_power_info *sumo_get_pi(struct radeon_device *rdev);
 
 static void sumo_send_msg_to_smu(struct radeon_device *rdev, u32 id)
diff --git a/drivers/gpu/drm/radeon/trinity_dpm.c b/drivers/gpu/drm/radeon/trinity_dpm.c
index d700698a1f22..2d447192d6f7 100644
--- a/drivers/gpu/drm/radeon/trinity_dpm.c
+++ b/drivers/gpu/drm/radeon/trinity_dpm.c
@@ -342,14 +342,14 @@ static void trinity_apply_state_adjust_rules(struct radeon_device *rdev,
 					     struct radeon_ps *new_rps,
 					     struct radeon_ps *old_rps);
 
-struct trinity_ps *trinity_get_ps(struct radeon_ps *rps)
+static struct trinity_ps *trinity_get_ps(struct radeon_ps *rps)
 {
 	struct trinity_ps *ps = rps->ps_priv;
 
 	return ps;
 }
 
-struct trinity_power_info *trinity_get_pi(struct radeon_device *rdev)
+static struct trinity_power_info *trinity_get_pi(struct radeon_device *rdev)
 {
 	struct trinity_power_info *pi = rdev->pm.dpm.priv;
 
@@ -1082,7 +1082,6 @@ void trinity_dpm_enable_bapm(struct radeon_device *rdev, bool enable)
 int trinity_dpm_enable(struct radeon_device *rdev)
 {
 	struct trinity_power_info *pi = trinity_get_pi(rdev);
-	int ret;
 
 	trinity_acquire_mutex(rdev);
 
@@ -1091,7 +1090,6 @@ int trinity_dpm_enable(struct radeon_device *rdev)
 		return -EINVAL;
 	}
 
-	trinity_enable_clock_power_gating(rdev);
 	trinity_program_bootup_state(rdev);
 	sumo_program_vc(rdev, 0x00C00033);
 	trinity_start_am(rdev);
@@ -1105,6 +1103,18 @@ int trinity_dpm_enable(struct radeon_device *rdev)
 	trinity_dpm_bapm_enable(rdev, false);
 	trinity_release_mutex(rdev);
 
+	trinity_update_current_ps(rdev, rdev->pm.dpm.boot_ps);
+
+	return 0;
+}
+
+int trinity_dpm_late_enable(struct radeon_device *rdev)
+{
+	int ret;
+
+	trinity_acquire_mutex(rdev);
+	trinity_enable_clock_power_gating(rdev);
+
 	if (rdev->irq.installed &&
 	    r600_is_internal_thermal_sensor(rdev->pm.int_thermal_type)) {
 		ret = trinity_set_thermal_temperature_range(rdev, R600_TEMP_RANGE_MIN, R600_TEMP_RANGE_MAX);
@@ -1115,8 +1125,7 @@ int trinity_dpm_enable(struct radeon_device *rdev)
 		rdev->irq.dpm_thermal = true;
 		radeon_irq_set(rdev);
 	}
-
-	trinity_update_current_ps(rdev, rdev->pm.dpm.boot_ps);
+	trinity_release_mutex(rdev);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/radeon/trinity_smc.c b/drivers/gpu/drm/radeon/trinity_smc.c
index 9672bcbc7312..99dd0455334d 100644
--- a/drivers/gpu/drm/radeon/trinity_smc.c
+++ b/drivers/gpu/drm/radeon/trinity_smc.c
@@ -27,9 +27,6 @@
 #include "trinity_dpm.h"
 #include "ppsmc.h"
 
-struct trinity_ps *trinity_get_ps(struct radeon_ps *rps);
-struct trinity_power_info *trinity_get_pi(struct radeon_device *rdev);
-
 static int trinity_notify_message_to_smu(struct radeon_device *rdev, u32 id)
 {
 	int i;
diff --git a/drivers/gpu/drm/radeon/uvd_v2_2.c b/drivers/gpu/drm/radeon/uvd_v2_2.c
index b19ef4951085..824550db3fed 100644
--- a/drivers/gpu/drm/radeon/uvd_v2_2.c
+++ b/drivers/gpu/drm/radeon/uvd_v2_2.c
@@ -153,6 +153,7 @@ int uvd_v2_2_resume(struct radeon_device *rdev)
 		chip_id = 0x01000015;
 		break;
 	case CHIP_PITCAIRN:
+	case CHIP_OLAND:
 		chip_id = 0x01000016;
 		break;
 	case CHIP_ARUBA:
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_crtc.c b/drivers/gpu/drm/rcar-du/rcar_du_crtc.c
index a9d24e4bf792..fbf4be316d0b 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_crtc.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_crtc.c
@@ -371,7 +371,6 @@ static int rcar_du_crtc_mode_set(struct drm_crtc *crtc,
 		goto error;
 
 	rcrtc->plane->format = format;
-	rcrtc->plane->pitch = crtc->fb->pitches[0];
 
 	rcrtc->plane->src_x = x;
 	rcrtc->plane->src_y = y;
@@ -413,7 +412,7 @@ static int rcar_du_crtc_mode_set_base(struct drm_crtc *crtc, int x, int y,
 	rcrtc->plane->src_x = x;
 	rcrtc->plane->src_y = y;
 
-	rcar_du_crtc_update_base(to_rcar_crtc(crtc));
+	rcar_du_crtc_update_base(rcrtc);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_drv.c b/drivers/gpu/drm/rcar-du/rcar_du_drv.c
index 0023f9719cf1..792fd1d20e86 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_drv.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_drv.c
@@ -224,7 +224,9 @@ static int rcar_du_probe(struct platform_device *pdev)
 
 static int rcar_du_remove(struct platform_device *pdev)
 {
-	drm_platform_exit(&rcar_du_driver, pdev);
+	struct rcar_du_device *rcdu = platform_get_drvdata(pdev);
+
+	drm_put_dev(rcdu->ddev);
 
 	return 0;
 }
@@ -249,8 +251,8 @@ static const struct rcar_du_device_info rcar_du_r8a7779_info = {
 };
 
 static const struct rcar_du_device_info rcar_du_r8a7790_info = {
-	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK | RCAR_DU_FEATURE_ALIGN_128B
-		  | RCAR_DU_FEATURE_DEFR8,
+	.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
@@ -272,9 +274,29 @@ static const struct rcar_du_device_info rcar_du_r8a7790_info = {
 	.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 },
 	{ }
 };
 
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_drv.h b/drivers/gpu/drm/rcar-du/rcar_du_drv.h
index 65d2d636b002..e31b735d3f25 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_drv.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_drv.h
@@ -28,8 +28,10 @@ struct rcar_du_device;
 struct rcar_du_lvdsenc;
 
 #define RCAR_DU_FEATURE_CRTC_IRQ_CLOCK	(1 << 0)	/* Per-CRTC IRQ and clock */
-#define RCAR_DU_FEATURE_ALIGN_128B	(1 << 1)	/* Align pitches to 128 bytes */
-#define RCAR_DU_FEATURE_DEFR8		(1 << 2)	/* Has DEFR8 register */
+#define RCAR_DU_FEATURE_DEFR8		(1 << 1)	/* Has DEFR8 register */
+
+#define RCAR_DU_QUIRK_ALIGN_128B	(1 << 0)	/* Align pitches to 128 bytes */
+#define RCAR_DU_QUIRK_LVDS_LANES	(1 << 1)	/* LVDS lanes 1 and 3 inverted */
 
 /*
  * struct rcar_du_output_routing - Output routing specification
@@ -48,12 +50,14 @@ struct rcar_du_output_routing {
 /*
  * struct rcar_du_device_info - DU model-specific information
  * @features: device features (RCAR_DU_FEATURE_*)
+ * @quirks: device quirks (RCAR_DU_QUIRK_*)
  * @num_crtcs: total number of CRTCs
  * @routes: array of CRTC to output routes, indexed by output (RCAR_DU_OUTPUT_*)
  * @num_lvds: number of internal LVDS encoders
  */
 struct rcar_du_device_info {
 	unsigned int features;
+	unsigned int quirks;
 	unsigned int num_crtcs;
 	struct rcar_du_output_routing routes[RCAR_DU_OUTPUT_MAX];
 	unsigned int num_lvds;
@@ -84,6 +88,12 @@ static inline bool rcar_du_has(struct rcar_du_device *rcdu,
 	return rcdu->info->features & feature;
 }
 
+static inline bool rcar_du_needs(struct rcar_du_device *rcdu,
+				 unsigned int quirk)
+{
+	return rcdu->info->quirks & quirk;
+}
+
 static inline u32 rcar_du_read(struct rcar_du_device *rcdu, u32 reg)
 {
 	return ioread32(rcdu->mmio + reg);
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_kms.c b/drivers/gpu/drm/rcar-du/rcar_du_kms.c
index b31ac080c4a7..fbeabd9a281f 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_kms.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_kms.c
@@ -119,7 +119,7 @@ int rcar_du_dumb_create(struct drm_file *file, struct drm_device *dev,
 	/* The R8A7779 DU requires a 16 pixels pitch alignment as documented,
 	 * but the R8A7790 DU seems to require a 128 bytes pitch alignment.
 	 */
-	if (rcar_du_has(rcdu, RCAR_DU_FEATURE_ALIGN_128B))
+	if (rcar_du_needs(rcdu, RCAR_DU_QUIRK_ALIGN_128B))
 		align = 128;
 	else
 		align = 16 * args->bpp / 8;
@@ -144,7 +144,7 @@ rcar_du_fb_create(struct drm_device *dev, struct drm_file *file_priv,
 		return ERR_PTR(-EINVAL);
 	}
 
-	if (rcar_du_has(rcdu, RCAR_DU_FEATURE_ALIGN_128B))
+	if (rcar_du_needs(rcdu, RCAR_DU_QUIRK_ALIGN_128B))
 		align = 128;
 	else
 		align = 16 * format->bpp / 8;
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.c b/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.c
index a0f6a1781925..df30a075d793 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_lvdsenc.c
@@ -44,6 +44,7 @@ static int rcar_du_lvdsenc_start(struct rcar_du_lvdsenc *lvds,
 	const struct drm_display_mode *mode = &rcrtc->crtc.mode;
 	unsigned int freq = mode->clock;
 	u32 lvdcr0;
+	u32 lvdhcr;
 	u32 pllcr;
 	int ret;
 
@@ -72,15 +73,19 @@ static int rcar_du_lvdsenc_start(struct rcar_du_lvdsenc *lvds,
 	 * VSYNC -> CTRL1
 	 * DISP  -> CTRL2
 	 * 0     -> CTRL3
-	 *
-	 * Channels 1 and 3 are switched on ES1.
 	 */
 	rcar_lvds_write(lvds, LVDCTRCR, LVDCTRCR_CTR3SEL_ZERO |
 			LVDCTRCR_CTR2SEL_DISP | LVDCTRCR_CTR1SEL_VSYNC |
 			LVDCTRCR_CTR0SEL_HSYNC);
-	rcar_lvds_write(lvds, LVDCHCR,
-			LVDCHCR_CHSEL_CH(0, 0) | LVDCHCR_CHSEL_CH(1, 3) |
-			LVDCHCR_CHSEL_CH(2, 2) | LVDCHCR_CHSEL_CH(3, 1));
+
+	if (rcar_du_needs(lvds->dev, RCAR_DU_QUIRK_LVDS_LANES))
+		lvdhcr = LVDCHCR_CHSEL_CH(0, 0) | LVDCHCR_CHSEL_CH(1, 3)
+		       | LVDCHCR_CHSEL_CH(2, 2) | LVDCHCR_CHSEL_CH(3, 1);
+	else
+		lvdhcr = LVDCHCR_CHSEL_CH(0, 0) | LVDCHCR_CHSEL_CH(1, 1)
+		       | LVDCHCR_CHSEL_CH(2, 2) | LVDCHCR_CHSEL_CH(3, 3);
+
+	rcar_lvds_write(lvds, LVDCHCR, lvdhcr);
 
 	/* Select the input, hardcode mode 0, enable LVDS operation and turn
 	 * bias circuitry on.
@@ -144,18 +149,9 @@ static int rcar_du_lvdsenc_get_resources(struct rcar_du_lvdsenc *lvds,
 	sprintf(name, "lvds.%u", lvds->index);
 
 	mem = platform_get_resource_byname(pdev, IORESOURCE_MEM, name);
-	if (mem == NULL) {
-		dev_err(&pdev->dev, "failed to get memory resource for %s\n",
-			name);
-		return -EINVAL;
-	}
-
 	lvds->mmio = devm_ioremap_resource(&pdev->dev, mem);
-	if (lvds->mmio == NULL) {
-		dev_err(&pdev->dev, "failed to remap memory resource for %s\n",
-			name);
-		return -ENOMEM;
-	}
+	if (IS_ERR(lvds->mmio))
+		return PTR_ERR(lvds->mmio);
 
 	lvds->clock = devm_clk_get(&pdev->dev, name);
 	if (IS_ERR(lvds->clock)) {
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_plane.c b/drivers/gpu/drm/rcar-du/rcar_du_plane.c
index 53000644733f..3fb69d9ae61b 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_plane.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_plane.c
@@ -104,6 +104,15 @@ void rcar_du_plane_update_base(struct rcar_du_plane *plane)
 {
 	struct rcar_du_group *rgrp = plane->group;
 	unsigned int index = plane->hwindex;
+	u32 mwr;
+
+	/* Memory pitch (expressed in pixels) */
+	if (plane->format->planes == 2)
+		mwr = plane->pitch;
+	else
+		mwr = plane->pitch * 8 / plane->format->bpp;
+
+	rcar_du_plane_write(rgrp, index, PnMWR, mwr);
 
 	/* The Y position is expressed in raster line units and must be doubled
 	 * for 32bpp formats, according to the R8A7790 datasheet. No mention of
@@ -133,6 +142,8 @@ void rcar_du_plane_compute_base(struct rcar_du_plane *plane,
 {
 	struct drm_gem_cma_object *gem;
 
+	plane->pitch = fb->pitches[0];
+
 	gem = drm_fb_cma_get_gem_obj(fb, 0);
 	plane->dma[0] = gem->paddr + fb->offsets[0];
 
@@ -209,7 +220,6 @@ static void __rcar_du_plane_setup(struct rcar_du_plane *plane,
 	struct rcar_du_group *rgrp = plane->group;
 	u32 ddcr2 = PnDDCR2_CODE;
 	u32 ddcr4;
-	u32 mwr;
 
 	/* Data format
 	 *
@@ -240,14 +250,6 @@ static void __rcar_du_plane_setup(struct rcar_du_plane *plane,
 	rcar_du_plane_write(rgrp, index, PnDDCR2, ddcr2);
 	rcar_du_plane_write(rgrp, index, PnDDCR4, ddcr4);
 
-	/* Memory pitch (expressed in pixels) */
-	if (plane->format->planes == 2)
-		mwr = plane->pitch;
-	else
-		mwr = plane->pitch * 8 / plane->format->bpp;
-
-	rcar_du_plane_write(rgrp, index, PnMWR, mwr);
-
 	/* Destination position and size */
 	rcar_du_plane_write(rgrp, index, PnDSXR, plane->width);
 	rcar_du_plane_write(rgrp, index, PnDSYR, plane->height);
@@ -309,7 +311,6 @@ rcar_du_plane_update(struct drm_plane *plane, struct drm_crtc *crtc,
 
 	rplane->crtc = crtc;
 	rplane->format = format;
-	rplane->pitch = fb->pitches[0];
 
 	rplane->src_x = src_x >> 16;
 	rplane->src_y = src_y >> 16;
diff --git a/drivers/gpu/drm/savage/savage_bci.c b/drivers/gpu/drm/savage/savage_bci.c
index b17d0710871a..d2b2df9e26f3 100644
--- a/drivers/gpu/drm/savage/savage_bci.c
+++ b/drivers/gpu/drm/savage/savage_bci.c
@@ -49,7 +49,7 @@ savage_bci_wait_fifo_shadow(drm_savage_private_t * dev_priv, unsigned int n)
 #endif
 
 	for (i = 0; i < SAVAGE_DEFAULT_USEC_TIMEOUT; i++) {
-		DRM_MEMORYBARRIER();
+		mb();
 		status = dev_priv->status_ptr[0];
 		if ((status & mask) < threshold)
 			return 0;
@@ -123,7 +123,7 @@ savage_bci_wait_event_shadow(drm_savage_private_t * dev_priv, uint16_t e)
 	int i;
 
 	for (i = 0; i < SAVAGE_EVENT_USEC_TIMEOUT; i++) {
-		DRM_MEMORYBARRIER();
+		mb();
 		status = dev_priv->status_ptr[1];
 		if ((((status & 0xffff) - e) & 0xffff) <= 0x7fff ||
 		    (status & 0xffff) == 0)
@@ -449,7 +449,7 @@ static void savage_dma_flush(drm_savage_private_t * dev_priv)
 		}
 	}
 
-	DRM_MEMORYBARRIER();
+	mb();
 
 	/* do flush ... */
 	phys_addr = dev_priv->cmd_dma->offset +
@@ -990,10 +990,10 @@ static int savage_bci_get_buffers(struct drm_device *dev,
 
 		buf->file_priv = file_priv;
 
-		if (DRM_COPY_TO_USER(&d->request_indices[i],
+		if (copy_to_user(&d->request_indices[i],
 				     &buf->idx, sizeof(buf->idx)))
 			return -EFAULT;
-		if (DRM_COPY_TO_USER(&d->request_sizes[i],
+		if (copy_to_user(&d->request_sizes[i],
 				     &buf->total, sizeof(buf->total)))
 			return -EFAULT;
 
diff --git a/drivers/gpu/drm/savage/savage_state.c b/drivers/gpu/drm/savage/savage_state.c
index b35e75ed890c..c01ad0aeaa58 100644
--- a/drivers/gpu/drm/savage/savage_state.c
+++ b/drivers/gpu/drm/savage/savage_state.c
@@ -992,7 +992,7 @@ int savage_bci_cmdbuf(struct drm_device *dev, void *data, struct drm_file *file_
 		if (kcmd_addr == NULL)
 			return -ENOMEM;
 
-		if (DRM_COPY_FROM_USER(kcmd_addr, cmdbuf->cmd_addr,
+		if (copy_from_user(kcmd_addr, cmdbuf->cmd_addr,
 				       cmdbuf->size * 8))
 		{
 			kfree(kcmd_addr);
@@ -1007,7 +1007,7 @@ int savage_bci_cmdbuf(struct drm_device *dev, void *data, struct drm_file *file_
 			goto done;
 		}
 
-		if (DRM_COPY_FROM_USER(kvb_addr, cmdbuf->vb_addr,
+		if (copy_from_user(kvb_addr, cmdbuf->vb_addr,
 				       cmdbuf->vb_size)) {
 			ret = -EFAULT;
 			goto done;
@@ -1022,7 +1022,7 @@ int savage_bci_cmdbuf(struct drm_device *dev, void *data, struct drm_file *file_
 			goto done;
 		}
 
-		if (DRM_COPY_FROM_USER(kbox_addr, cmdbuf->box_addr,
+		if (copy_from_user(kbox_addr, cmdbuf->box_addr,
 				       cmdbuf->nbox * sizeof(struct drm_clip_rect))) {
 			ret = -EFAULT;
 			goto done;
@@ -1032,7 +1032,7 @@ int savage_bci_cmdbuf(struct drm_device *dev, void *data, struct drm_file *file_
 
 	/* Make sure writes to DMA buffers are finished before sending
 	 * DMA commands to the graphics hardware. */
-	DRM_MEMORYBARRIER();
+	mb();
 
 	/* Coming from user space. Don't know if the Xserver has
 	 * emitted wait commands. Assuming the worst. */
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_crtc.c b/drivers/gpu/drm/shmobile/shmob_drm_crtc.c
index 562f9a401cf6..0428076f1ce8 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_crtc.c
+++ b/drivers/gpu/drm/shmobile/shmob_drm_crtc.c
@@ -37,14 +37,21 @@
  * Clock management
  */
 
-static void shmob_drm_clk_on(struct shmob_drm_device *sdev)
+static int shmob_drm_clk_on(struct shmob_drm_device *sdev)
 {
-	if (sdev->clock)
-		clk_prepare_enable(sdev->clock);
+	int ret;
+
+	if (sdev->clock) {
+		ret = clk_prepare_enable(sdev->clock);
+		if (ret < 0)
+			return ret;
+	}
 #if 0
 	if (sdev->meram_dev && sdev->meram_dev->pdev)
 		pm_runtime_get_sync(&sdev->meram_dev->pdev->dev);
 #endif
+
+	return 0;
 }
 
 static void shmob_drm_clk_off(struct shmob_drm_device *sdev)
@@ -161,6 +168,7 @@ static void shmob_drm_crtc_start(struct shmob_drm_crtc *scrtc)
 	struct drm_device *dev = sdev->ddev;
 	struct drm_plane *plane;
 	u32 value;
+	int ret;
 
 	if (scrtc->started)
 		return;
@@ -170,7 +178,9 @@ static void shmob_drm_crtc_start(struct shmob_drm_crtc *scrtc)
 		return;
 
 	/* Enable clocks before accessing the hardware. */
-	shmob_drm_clk_on(sdev);
+	ret = shmob_drm_clk_on(sdev);
+	if (ret < 0)
+		return;
 
 	/* Reset and enable the LCDC. */
 	lcdc_write(sdev, LDCNT2R, lcdc_read(sdev, LDCNT2R) | LDCNT2R_BR);
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_drv.c b/drivers/gpu/drm/shmobile/shmob_drm_drv.c
index 015551866b4a..c839c9c89efb 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_drv.c
+++ b/drivers/gpu/drm/shmobile/shmob_drm_drv.c
@@ -336,7 +336,9 @@ static int shmob_drm_probe(struct platform_device *pdev)
 
 static int shmob_drm_remove(struct platform_device *pdev)
 {
-	drm_platform_exit(&shmob_drm_driver, pdev);
+	struct shmob_drm_device *sdev = platform_get_drvdata(pdev);
+
+	drm_put_dev(sdev->ddev);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/sis/sis_drv.c b/drivers/gpu/drm/sis/sis_drv.c
index 4383b74a3aa4..756f787b7143 100644
--- a/drivers/gpu/drm/sis/sis_drv.c
+++ b/drivers/gpu/drm/sis/sis_drv.c
@@ -94,7 +94,7 @@ static int sis_driver_open(struct drm_device *dev, struct drm_file *file)
 	return 0;
 }
 
-void sis_driver_postclose(struct drm_device *dev, struct drm_file *file)
+static void sis_driver_postclose(struct drm_device *dev, struct drm_file *file)
 {
 	struct sis_file_private *file_priv = file->driver_priv;
 
diff --git a/drivers/gpu/drm/sis/sis_mm.c b/drivers/gpu/drm/sis/sis_mm.c
index 01857d836350..0573be0d2933 100644
--- a/drivers/gpu/drm/sis/sis_mm.c
+++ b/drivers/gpu/drm/sis/sis_mm.c
@@ -266,7 +266,7 @@ int sis_idle(struct drm_device *dev)
 	 * because its polling frequency is too low.
 	 */
 
-	end = jiffies + (DRM_HZ * 3);
+	end = jiffies + (HZ * 3);
 
 	for (i = 0; i < 4; ++i) {
 		do {
diff --git a/drivers/gpu/drm/tegra/Kconfig b/drivers/gpu/drm/tegra/Kconfig
index 8db9b3bce001..354ddb29231f 100644
--- a/drivers/gpu/drm/tegra/Kconfig
+++ b/drivers/gpu/drm/tegra/Kconfig
@@ -1,14 +1,12 @@
 config DRM_TEGRA
-	bool "NVIDIA Tegra DRM"
-	depends on ARCH_TEGRA || ARCH_MULTIPLATFORM
+	tristate "NVIDIA Tegra DRM"
+	depends on ARCH_TEGRA || (ARM && COMPILE_TEST)
 	depends on DRM
 	depends on RESET_CONTROLLER
-	select TEGRA_HOST1X
 	select DRM_KMS_HELPER
-	select DRM_KMS_FB_HELPER
-	select FB_SYS_FILLRECT
-	select FB_SYS_COPYAREA
-	select FB_SYS_IMAGEBLIT
+	select DRM_MIPI_DSI
+	select DRM_PANEL
+	select TEGRA_HOST1X
 	help
 	  Choose this option if you have an NVIDIA Tegra SoC.
 
@@ -17,6 +15,18 @@ config DRM_TEGRA
 
 if DRM_TEGRA
 
+config DRM_TEGRA_FBDEV
+	bool "Enable legacy fbdev support"
+	select DRM_KMS_FB_HELPER
+	select FB_SYS_FILLRECT
+	select FB_SYS_COPYAREA
+	select FB_SYS_IMAGEBLIT
+	default y
+	help
+	  Choose this option if you have a need for the legacy fbdev support.
+	  Note that this support also provides the Linux console on top of
+	  the Tegra modesetting driver.
+
 config DRM_TEGRA_DEBUG
 	bool "NVIDIA Tegra DRM debug support"
 	help
diff --git a/drivers/gpu/drm/tegra/Makefile b/drivers/gpu/drm/tegra/Makefile
index edc76abd58bb..8d220afbd85f 100644
--- a/drivers/gpu/drm/tegra/Makefile
+++ b/drivers/gpu/drm/tegra/Makefile
@@ -9,6 +9,8 @@ tegra-drm-y := \
 	output.o \
 	rgb.o \
 	hdmi.o \
+	mipi-phy.o \
+	dsi.o \
 	gr2d.o \
 	gr3d.o
 
diff --git a/drivers/gpu/drm/tegra/bus.c b/drivers/gpu/drm/tegra/bus.c
index 565f8f7b9a47..e38e5967d77b 100644
--- a/drivers/gpu/drm/tegra/bus.c
+++ b/drivers/gpu/drm/tegra/bus.c
@@ -46,7 +46,6 @@ int drm_host1x_init(struct drm_driver *driver, struct host1x_device *device)
 	struct drm_device *drm;
 	int ret;
 
-	INIT_LIST_HEAD(&driver->device_list);
 	driver->bus = &drm_host1x_bus;
 
 	drm = drm_dev_alloc(driver, &device->dev);
diff --git a/drivers/gpu/drm/tegra/dc.c b/drivers/gpu/drm/tegra/dc.c
index cd7f1e499616..9336006b475d 100644
--- a/drivers/gpu/drm/tegra/dc.c
+++ b/drivers/gpu/drm/tegra/dc.c
@@ -15,6 +15,10 @@
 #include "drm.h"
 #include "gem.h"
 
+struct tegra_dc_soc_info {
+	bool supports_interlacing;
+};
+
 struct tegra_plane {
 	struct drm_plane base;
 	unsigned int index;
@@ -658,19 +662,12 @@ static int tegra_crtc_mode_set(struct drm_crtc *crtc,
 	/* program display mode */
 	tegra_dc_set_timings(dc, mode);
 
-	value = DE_SELECT_ACTIVE | DE_CONTROL_NORMAL;
-	tegra_dc_writel(dc, value, DC_DISP_DATA_ENABLE_OPTIONS);
-
-	value = tegra_dc_readl(dc, DC_COM_PIN_OUTPUT_POLARITY(1));
-	value &= ~LVS_OUTPUT_POLARITY_LOW;
-	value &= ~LHS_OUTPUT_POLARITY_LOW;
-	tegra_dc_writel(dc, value, DC_COM_PIN_OUTPUT_POLARITY(1));
-
-	value = DISP_DATA_FORMAT_DF1P1C | DISP_ALIGNMENT_MSB |
-		DISP_ORDER_RED_BLUE;
-	tegra_dc_writel(dc, value, DC_DISP_DISP_INTERFACE_CONTROL);
-
-	tegra_dc_writel(dc, 0x00010001, DC_DISP_SHIFT_CLOCK_OPTIONS);
+	/* interlacing isn't supported yet, so disable it */
+	if (dc->soc->supports_interlacing) {
+		value = tegra_dc_readl(dc, DC_DISP_INTERLACE_CONTROL);
+		value &= ~INTERLACE_ENABLE;
+		tegra_dc_writel(dc, value, DC_DISP_INTERLACE_CONTROL);
+	}
 
 	value = SHIFT_CLK_DIVIDER(div) | PIXEL_CLK_DIVIDER_PCD1;
 	tegra_dc_writel(dc, value, DC_DISP_DISP_CLOCK_CONTROL);
@@ -735,10 +732,6 @@ static void tegra_crtc_prepare(struct drm_crtc *crtc)
 		PW4_ENABLE | PM0_ENABLE | PM1_ENABLE;
 	tegra_dc_writel(dc, value, DC_CMD_DISPLAY_POWER_CONTROL);
 
-	value = tegra_dc_readl(dc, DC_CMD_DISPLAY_COMMAND);
-	value |= DISP_CTRL_MODE_C_DISPLAY;
-	tegra_dc_writel(dc, value, DC_CMD_DISPLAY_COMMAND);
-
 	/* initialize timer */
 	value = CURSOR_THRESHOLD(0) | WINDOW_A_THRESHOLD(0x20) |
 		WINDOW_B_THRESHOLD(0x20) | WINDOW_C_THRESHOLD(0x20);
@@ -1107,8 +1100,6 @@ static int tegra_dc_init(struct host1x_client *client)
 	struct tegra_dc *dc = host1x_client_to_dc(client);
 	int err;
 
-	dc->pipe = tegra->drm->mode_config.num_crtc;
-
 	drm_crtc_init(tegra->drm, &dc->base, &tegra_crtc_funcs);
 	drm_mode_crtc_set_gamma_size(&dc->base, 256);
 	drm_crtc_helper_add(&dc->base, &tegra_crtc_helper_funcs);
@@ -1167,8 +1158,71 @@ static const struct host1x_client_ops dc_client_ops = {
 	.exit = tegra_dc_exit,
 };
 
+static const struct tegra_dc_soc_info tegra20_dc_soc_info = {
+	.supports_interlacing = false,
+};
+
+static const struct tegra_dc_soc_info tegra30_dc_soc_info = {
+	.supports_interlacing = false,
+};
+
+static const struct tegra_dc_soc_info tegra124_dc_soc_info = {
+	.supports_interlacing = true,
+};
+
+static const struct of_device_id tegra_dc_of_match[] = {
+	{
+		.compatible = "nvidia,tegra124-dc",
+		.data = &tegra124_dc_soc_info,
+	}, {
+		.compatible = "nvidia,tegra30-dc",
+		.data = &tegra30_dc_soc_info,
+	}, {
+		.compatible = "nvidia,tegra20-dc",
+		.data = &tegra20_dc_soc_info,
+	}, {
+		/* sentinel */
+	}
+};
+
+static int tegra_dc_parse_dt(struct tegra_dc *dc)
+{
+	struct device_node *np;
+	u32 value = 0;
+	int err;
+
+	err = of_property_read_u32(dc->dev->of_node, "nvidia,head", &value);
+	if (err < 0) {
+		dev_err(dc->dev, "missing \"nvidia,head\" property\n");
+
+		/*
+		 * If the nvidia,head property isn't present, try to find the
+		 * correct head number by looking up the position of this
+		 * display controller's node within the device tree. Assuming
+		 * that the nodes are ordered properly in the DTS file and
+		 * that the translation into a flattened device tree blob
+		 * preserves that ordering this will actually yield the right
+		 * head number.
+		 *
+		 * If those assumptions don't hold, this will still work for
+		 * cases where only a single display controller is used.
+		 */
+		for_each_matching_node(np, tegra_dc_of_match) {
+			if (np == dc->dev->of_node)
+				break;
+
+			value++;
+		}
+	}
+
+	dc->pipe = value;
+
+	return 0;
+}
+
 static int tegra_dc_probe(struct platform_device *pdev)
 {
+	const struct of_device_id *id;
 	struct resource *regs;
 	struct tegra_dc *dc;
 	int err;
@@ -1177,9 +1231,18 @@ static int tegra_dc_probe(struct platform_device *pdev)
 	if (!dc)
 		return -ENOMEM;
 
+	id = of_match_node(tegra_dc_of_match, pdev->dev.of_node);
+	if (!id)
+		return -ENODEV;
+
 	spin_lock_init(&dc->lock);
 	INIT_LIST_HEAD(&dc->list);
 	dc->dev = &pdev->dev;
+	dc->soc = id->data;
+
+	err = tegra_dc_parse_dt(dc);
+	if (err < 0)
+		return err;
 
 	dc->clk = devm_clk_get(&pdev->dev, NULL);
 	if (IS_ERR(dc->clk)) {
@@ -1253,12 +1316,6 @@ static int tegra_dc_remove(struct platform_device *pdev)
 	return 0;
 }
 
-static struct of_device_id tegra_dc_of_match[] = {
-	{ .compatible = "nvidia,tegra30-dc", },
-	{ .compatible = "nvidia,tegra20-dc", },
-	{ },
-};
-
 struct platform_driver tegra_dc_driver = {
 	.driver = {
 		.name = "tegra-dc",
diff --git a/drivers/gpu/drm/tegra/dc.h b/drivers/gpu/drm/tegra/dc.h
index 91bbda291470..3c2c0ea1cd87 100644
--- a/drivers/gpu/drm/tegra/dc.h
+++ b/drivers/gpu/drm/tegra/dc.h
@@ -28,6 +28,7 @@
 #define DISP_CTRL_MODE_STOP (0 << 5)
 #define DISP_CTRL_MODE_C_DISPLAY (1 << 5)
 #define DISP_CTRL_MODE_NC_DISPLAY (2 << 5)
+#define DISP_CTRL_MODE_MASK (3 << 5)
 #define DC_CMD_SIGNAL_RAISE			0x033
 #define DC_CMD_DISPLAY_POWER_CONTROL		0x036
 #define PW0_ENABLE (1 <<  0)
@@ -116,6 +117,7 @@
 
 #define DC_DISP_DISP_WIN_OPTIONS		0x402
 #define HDMI_ENABLE (1 << 30)
+#define DSI_ENABLE  (1 << 29)
 
 #define DC_DISP_DISP_MEM_HIGH_PRIORITY		0x403
 #define CURSOR_THRESHOLD(x)   (((x) & 0x03) << 24)
@@ -238,6 +240,8 @@
 #define DITHER_CONTROL_ERRDIFF (3 << 8)
 
 #define DC_DISP_SHIFT_CLOCK_OPTIONS		0x431
+#define  SC1_H_QUALIFIER_NONE	(1 << 16)
+#define  SC0_H_QUALIFIER_NONE	(1 <<  0)
 
 #define DC_DISP_DATA_ENABLE_OPTIONS		0x432
 #define DE_SELECT_ACTIVE_BLANK  (0 << 0)
@@ -292,6 +296,11 @@
 #define DC_DISP_SD_HW_K_VALUES			0x4dd
 #define DC_DISP_SD_MAN_K_VALUES			0x4de
 
+#define DC_DISP_INTERLACE_CONTROL		0x4e5
+#define  INTERLACE_STATUS (1 << 2)
+#define  INTERLACE_START  (1 << 1)
+#define  INTERLACE_ENABLE (1 << 0)
+
 #define DC_WIN_CSC_YOF				0x611
 #define DC_WIN_CSC_KYRGB			0x612
 #define DC_WIN_CSC_KUR				0x613
diff --git a/drivers/gpu/drm/tegra/drm.c b/drivers/gpu/drm/tegra/drm.c
index 07eba596d458..88a529008ce0 100644
--- a/drivers/gpu/drm/tegra/drm.c
+++ b/drivers/gpu/drm/tegra/drm.c
@@ -104,9 +104,11 @@ static void tegra_drm_context_free(struct tegra_drm_context *context)
 
 static void tegra_drm_lastclose(struct drm_device *drm)
 {
+#ifdef CONFIG_TEGRA_DRM_FBDEV
 	struct tegra_drm *tegra = drm->dev_private;
 
 	tegra_fbdev_restore_mode(tegra->fbdev);
+#endif
 }
 
 static struct host1x_bo *
@@ -578,7 +580,7 @@ static void tegra_debugfs_cleanup(struct drm_minor *minor)
 #endif
 
 static struct drm_driver tegra_drm_driver = {
-	.driver_features = DRIVER_MODESET | DRIVER_GEM,
+	.driver_features = DRIVER_MODESET | DRIVER_GEM | DRIVER_PRIME,
 	.load = tegra_drm_load,
 	.unload = tegra_drm_unload,
 	.open = tegra_drm_open,
@@ -596,6 +598,12 @@ static struct drm_driver tegra_drm_driver = {
 
 	.gem_free_object = tegra_bo_free_object,
 	.gem_vm_ops = &tegra_bo_vm_ops,
+
+	.prime_handle_to_fd = drm_gem_prime_handle_to_fd,
+	.prime_fd_to_handle = drm_gem_prime_fd_to_handle,
+	.gem_prime_export = tegra_gem_prime_export,
+	.gem_prime_import = tegra_gem_prime_import,
+
 	.dumb_create = tegra_bo_dumb_create,
 	.dumb_map_offset = tegra_bo_dumb_map_offset,
 	.dumb_destroy = drm_gem_dumb_destroy,
@@ -653,8 +661,10 @@ static const struct of_device_id host1x_drm_subdevs[] = {
 	{ .compatible = "nvidia,tegra30-hdmi", },
 	{ .compatible = "nvidia,tegra30-gr2d", },
 	{ .compatible = "nvidia,tegra30-gr3d", },
+	{ .compatible = "nvidia,tegra114-dsi", },
 	{ .compatible = "nvidia,tegra114-hdmi", },
 	{ .compatible = "nvidia,tegra114-gr3d", },
+	{ .compatible = "nvidia,tegra124-dc", },
 	{ /* sentinel */ }
 };
 
@@ -677,10 +687,14 @@ static int __init host1x_drm_init(void)
 	if (err < 0)
 		goto unregister_host1x;
 
-	err = platform_driver_register(&tegra_hdmi_driver);
+	err = platform_driver_register(&tegra_dsi_driver);
 	if (err < 0)
 		goto unregister_dc;
 
+	err = platform_driver_register(&tegra_hdmi_driver);
+	if (err < 0)
+		goto unregister_dsi;
+
 	err = platform_driver_register(&tegra_gr2d_driver);
 	if (err < 0)
 		goto unregister_hdmi;
@@ -695,6 +709,8 @@ unregister_gr2d:
 	platform_driver_unregister(&tegra_gr2d_driver);
 unregister_hdmi:
 	platform_driver_unregister(&tegra_hdmi_driver);
+unregister_dsi:
+	platform_driver_unregister(&tegra_dsi_driver);
 unregister_dc:
 	platform_driver_unregister(&tegra_dc_driver);
 unregister_host1x:
@@ -708,6 +724,7 @@ static void __exit host1x_drm_exit(void)
 	platform_driver_unregister(&tegra_gr3d_driver);
 	platform_driver_unregister(&tegra_gr2d_driver);
 	platform_driver_unregister(&tegra_hdmi_driver);
+	platform_driver_unregister(&tegra_dsi_driver);
 	platform_driver_unregister(&tegra_dc_driver);
 	host1x_driver_unregister(&host1x_drm_driver);
 }
diff --git a/drivers/gpu/drm/tegra/drm.h b/drivers/gpu/drm/tegra/drm.h
index 266aae08a3bd..bf1cac7658f8 100644
--- a/drivers/gpu/drm/tegra/drm.h
+++ b/drivers/gpu/drm/tegra/drm.h
@@ -27,10 +27,12 @@ struct tegra_fb {
 	unsigned int num_planes;
 };
 
+#ifdef CONFIG_DRM_TEGRA_FBDEV
 struct tegra_fbdev {
 	struct drm_fb_helper base;
 	struct tegra_fb *fb;
 };
+#endif
 
 struct tegra_drm {
 	struct drm_device *drm;
@@ -38,7 +40,9 @@ struct tegra_drm {
 	struct mutex clients_lock;
 	struct list_head clients;
 
+#ifdef CONFIG_DRM_TEGRA_FBDEV
 	struct tegra_fbdev *fbdev;
+#endif
 };
 
 struct tegra_drm_client;
@@ -84,6 +88,7 @@ extern int tegra_drm_unregister_client(struct tegra_drm *tegra,
 extern int tegra_drm_init(struct tegra_drm *tegra, struct drm_device *drm);
 extern int tegra_drm_exit(struct tegra_drm *tegra);
 
+struct tegra_dc_soc_info;
 struct tegra_output;
 
 struct tegra_dc {
@@ -109,6 +114,8 @@ struct tegra_dc {
 
 	/* page-flip handling */
 	struct drm_pending_vblank_event *event;
+
+	const struct tegra_dc_soc_info *soc;
 };
 
 static inline struct tegra_dc *
@@ -177,6 +184,7 @@ struct tegra_output_ops {
 enum tegra_output_type {
 	TEGRA_OUTPUT_RGB,
 	TEGRA_OUTPUT_HDMI,
+	TEGRA_OUTPUT_DSI,
 };
 
 struct tegra_output {
@@ -186,6 +194,7 @@ struct tegra_output {
 	const struct tegra_output_ops *ops;
 	enum tegra_output_type type;
 
+	struct drm_panel *panel;
 	struct i2c_adapter *ddc;
 	const struct edid *edid;
 	unsigned int hpd_irq;
@@ -263,9 +272,12 @@ bool tegra_fb_is_bottom_up(struct drm_framebuffer *framebuffer);
 bool tegra_fb_is_tiled(struct drm_framebuffer *framebuffer);
 extern int tegra_drm_fb_init(struct drm_device *drm);
 extern void tegra_drm_fb_exit(struct drm_device *drm);
+#ifdef CONFIG_DRM_TEGRA_FBDEV
 extern void tegra_fbdev_restore_mode(struct tegra_fbdev *fbdev);
+#endif
 
 extern struct platform_driver tegra_dc_driver;
+extern struct platform_driver tegra_dsi_driver;
 extern struct platform_driver tegra_hdmi_driver;
 extern struct platform_driver tegra_gr2d_driver;
 extern struct platform_driver tegra_gr3d_driver;
diff --git a/drivers/gpu/drm/tegra/dsi.c b/drivers/gpu/drm/tegra/dsi.c
new file mode 100644
index 000000000000..d452faab0235
--- /dev/null
+++ b/drivers/gpu/drm/tegra/dsi.c
@@ -0,0 +1,971 @@
+/*
+ * Copyright (C) 2013 NVIDIA Corporation
+ *
+ * Permission to use, copy, modify, distribute, and sell this software and its
+ * documentation for any purpose is hereby granted without fee, provided that
+ * the above copyright notice appear in all copies and that both that copyright
+ * notice and this permission notice appear in supporting documentation, and
+ * that the name of the copyright holders not be used in advertising or
+ * publicity pertaining to distribution of the software without specific,
+ * written prior permission.  The copyright holders make no representations
+ * about the suitability of this software for any purpose.  It is provided "as
+ * is" without express or implied warranty.
+ *
+ * THE COPYRIGHT HOLDERS DISCLAIM ALL WARRANTIES WITH REGARD TO THIS SOFTWARE,
+ * INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN NO
+ * EVENT SHALL THE COPYRIGHT HOLDERS BE LIABLE FOR ANY SPECIAL, INDIRECT OR
+ * CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE,
+ * DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER
+ * TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE
+ * OF THIS SOFTWARE.
+ */
+
+#include <linux/clk.h>
+#include <linux/debugfs.h>
+#include <linux/host1x.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/platform_device.h>
+#include <linux/reset.h>
+
+#include <drm/drm_mipi_dsi.h>
+#include <drm/drm_panel.h>
+
+#include <video/mipi_display.h>
+
+#include "dc.h"
+#include "drm.h"
+#include "dsi.h"
+#include "mipi-phy.h"
+
+#define DSI_VIDEO_FIFO_DEPTH (1920 / 4)
+#define DSI_HOST_FIFO_DEPTH 64
+
+struct tegra_dsi {
+	struct host1x_client client;
+	struct tegra_output output;
+	struct device *dev;
+
+	void __iomem *regs;
+
+	struct reset_control *rst;
+	struct clk *clk_parent;
+	struct clk *clk_lp;
+	struct clk *clk;
+
+	struct drm_info_list *debugfs_files;
+	struct drm_minor *minor;
+	struct dentry *debugfs;
+
+	enum mipi_dsi_pixel_format format;
+	unsigned int lanes;
+
+	struct tegra_mipi_device *mipi;
+	struct mipi_dsi_host host;
+};
+
+static inline struct tegra_dsi *
+host1x_client_to_dsi(struct host1x_client *client)
+{
+	return container_of(client, struct tegra_dsi, client);
+}
+
+static inline struct tegra_dsi *host_to_tegra(struct mipi_dsi_host *host)
+{
+	return container_of(host, struct tegra_dsi, host);
+}
+
+static inline struct tegra_dsi *to_dsi(struct tegra_output *output)
+{
+	return container_of(output, struct tegra_dsi, output);
+}
+
+static inline unsigned long tegra_dsi_readl(struct tegra_dsi *dsi,
+					    unsigned long reg)
+{
+	return readl(dsi->regs + (reg << 2));
+}
+
+static inline void tegra_dsi_writel(struct tegra_dsi *dsi, unsigned long value,
+				    unsigned long reg)
+{
+	writel(value, dsi->regs + (reg << 2));
+}
+
+static int tegra_dsi_show_regs(struct seq_file *s, void *data)
+{
+	struct drm_info_node *node = s->private;
+	struct tegra_dsi *dsi = node->info_ent->data;
+
+#define DUMP_REG(name)						\
+	seq_printf(s, "%-32s %#05x %08lx\n", #name, name,	\
+		   tegra_dsi_readl(dsi, name))
+
+	DUMP_REG(DSI_INCR_SYNCPT);
+	DUMP_REG(DSI_INCR_SYNCPT_CONTROL);
+	DUMP_REG(DSI_INCR_SYNCPT_ERROR);
+	DUMP_REG(DSI_CTXSW);
+	DUMP_REG(DSI_RD_DATA);
+	DUMP_REG(DSI_WR_DATA);
+	DUMP_REG(DSI_POWER_CONTROL);
+	DUMP_REG(DSI_INT_ENABLE);
+	DUMP_REG(DSI_INT_STATUS);
+	DUMP_REG(DSI_INT_MASK);
+	DUMP_REG(DSI_HOST_CONTROL);
+	DUMP_REG(DSI_CONTROL);
+	DUMP_REG(DSI_SOL_DELAY);
+	DUMP_REG(DSI_MAX_THRESHOLD);
+	DUMP_REG(DSI_TRIGGER);
+	DUMP_REG(DSI_TX_CRC);
+	DUMP_REG(DSI_STATUS);
+
+	DUMP_REG(DSI_INIT_SEQ_CONTROL);
+	DUMP_REG(DSI_INIT_SEQ_DATA_0);
+	DUMP_REG(DSI_INIT_SEQ_DATA_1);
+	DUMP_REG(DSI_INIT_SEQ_DATA_2);
+	DUMP_REG(DSI_INIT_SEQ_DATA_3);
+	DUMP_REG(DSI_INIT_SEQ_DATA_4);
+	DUMP_REG(DSI_INIT_SEQ_DATA_5);
+	DUMP_REG(DSI_INIT_SEQ_DATA_6);
+	DUMP_REG(DSI_INIT_SEQ_DATA_7);
+
+	DUMP_REG(DSI_PKT_SEQ_0_LO);
+	DUMP_REG(DSI_PKT_SEQ_0_HI);
+	DUMP_REG(DSI_PKT_SEQ_1_LO);
+	DUMP_REG(DSI_PKT_SEQ_1_HI);
+	DUMP_REG(DSI_PKT_SEQ_2_LO);
+	DUMP_REG(DSI_PKT_SEQ_2_HI);
+	DUMP_REG(DSI_PKT_SEQ_3_LO);
+	DUMP_REG(DSI_PKT_SEQ_3_HI);
+	DUMP_REG(DSI_PKT_SEQ_4_LO);
+	DUMP_REG(DSI_PKT_SEQ_4_HI);
+	DUMP_REG(DSI_PKT_SEQ_5_LO);
+	DUMP_REG(DSI_PKT_SEQ_5_HI);
+
+	DUMP_REG(DSI_DCS_CMDS);
+
+	DUMP_REG(DSI_PKT_LEN_0_1);
+	DUMP_REG(DSI_PKT_LEN_2_3);
+	DUMP_REG(DSI_PKT_LEN_4_5);
+	DUMP_REG(DSI_PKT_LEN_6_7);
+
+	DUMP_REG(DSI_PHY_TIMING_0);
+	DUMP_REG(DSI_PHY_TIMING_1);
+	DUMP_REG(DSI_PHY_TIMING_2);
+	DUMP_REG(DSI_BTA_TIMING);
+
+	DUMP_REG(DSI_TIMEOUT_0);
+	DUMP_REG(DSI_TIMEOUT_1);
+	DUMP_REG(DSI_TO_TALLY);
+
+	DUMP_REG(DSI_PAD_CONTROL_0);
+	DUMP_REG(DSI_PAD_CONTROL_CD);
+	DUMP_REG(DSI_PAD_CD_STATUS);
+	DUMP_REG(DSI_VIDEO_MODE_CONTROL);
+	DUMP_REG(DSI_PAD_CONTROL_1);
+	DUMP_REG(DSI_PAD_CONTROL_2);
+	DUMP_REG(DSI_PAD_CONTROL_3);
+	DUMP_REG(DSI_PAD_CONTROL_4);
+
+	DUMP_REG(DSI_GANGED_MODE_CONTROL);
+	DUMP_REG(DSI_GANGED_MODE_START);
+	DUMP_REG(DSI_GANGED_MODE_SIZE);
+
+	DUMP_REG(DSI_RAW_DATA_BYTE_COUNT);
+	DUMP_REG(DSI_ULTRA_LOW_POWER_CONTROL);
+
+	DUMP_REG(DSI_INIT_SEQ_DATA_8);
+	DUMP_REG(DSI_INIT_SEQ_DATA_9);
+	DUMP_REG(DSI_INIT_SEQ_DATA_10);
+	DUMP_REG(DSI_INIT_SEQ_DATA_11);
+	DUMP_REG(DSI_INIT_SEQ_DATA_12);
+	DUMP_REG(DSI_INIT_SEQ_DATA_13);
+	DUMP_REG(DSI_INIT_SEQ_DATA_14);
+	DUMP_REG(DSI_INIT_SEQ_DATA_15);
+
+#undef DUMP_REG
+
+	return 0;
+}
+
+static struct drm_info_list debugfs_files[] = {
+	{ "regs", tegra_dsi_show_regs, 0, NULL },
+};
+
+static int tegra_dsi_debugfs_init(struct tegra_dsi *dsi,
+				  struct drm_minor *minor)
+{
+	const char *name = dev_name(dsi->dev);
+	unsigned int i;
+	int err;
+
+	dsi->debugfs = debugfs_create_dir(name, minor->debugfs_root);
+	if (!dsi->debugfs)
+		return -ENOMEM;
+
+	dsi->debugfs_files = kmemdup(debugfs_files, sizeof(debugfs_files),
+				     GFP_KERNEL);
+	if (!dsi->debugfs_files) {
+		err = -ENOMEM;
+		goto remove;
+	}
+
+	for (i = 0; i < ARRAY_SIZE(debugfs_files); i++)
+		dsi->debugfs_files[i].data = dsi;
+
+	err = drm_debugfs_create_files(dsi->debugfs_files,
+				       ARRAY_SIZE(debugfs_files),
+				       dsi->debugfs, minor);
+	if (err < 0)
+		goto free;
+
+	dsi->minor = minor;
+
+	return 0;
+
+free:
+	kfree(dsi->debugfs_files);
+	dsi->debugfs_files = NULL;
+remove:
+	debugfs_remove(dsi->debugfs);
+	dsi->debugfs = NULL;
+
+	return err;
+}
+
+static int tegra_dsi_debugfs_exit(struct tegra_dsi *dsi)
+{
+	drm_debugfs_remove_files(dsi->debugfs_files, ARRAY_SIZE(debugfs_files),
+				 dsi->minor);
+	dsi->minor = NULL;
+
+	kfree(dsi->debugfs_files);
+	dsi->debugfs_files = NULL;
+
+	debugfs_remove(dsi->debugfs);
+	dsi->debugfs = NULL;
+
+	return 0;
+}
+
+#define PKT_ID0(id)	((((id) & 0x3f) <<  3) | (1 <<  9))
+#define PKT_LEN0(len)	(((len) & 0x07) <<  0)
+#define PKT_ID1(id)	((((id) & 0x3f) << 13) | (1 << 19))
+#define PKT_LEN1(len)	(((len) & 0x07) << 10)
+#define PKT_ID2(id)	((((id) & 0x3f) << 23) | (1 << 29))
+#define PKT_LEN2(len)	(((len) & 0x07) << 20)
+
+#define PKT_LP		(1 << 30)
+#define NUM_PKT_SEQ	12
+
+/* non-burst mode with sync-end */
+static const u32 pkt_seq_vnb_syne[NUM_PKT_SEQ] = {
+	[ 0] = PKT_ID0(MIPI_DSI_V_SYNC_START) | PKT_LEN0(0) |
+	       PKT_ID1(MIPI_DSI_BLANKING_PACKET) | PKT_LEN1(1) |
+	       PKT_ID2(MIPI_DSI_H_SYNC_END) | PKT_LEN2(0) |
+	       PKT_LP,
+	[ 1] = 0,
+	[ 2] = PKT_ID0(MIPI_DSI_V_SYNC_END) | PKT_LEN0(0) |
+	       PKT_ID1(MIPI_DSI_BLANKING_PACKET) | PKT_LEN1(1) |
+	       PKT_ID2(MIPI_DSI_H_SYNC_END) | PKT_LEN2(0) |
+	       PKT_LP,
+	[ 3] = 0,
+	[ 4] = PKT_ID0(MIPI_DSI_H_SYNC_START) | PKT_LEN0(0) |
+	       PKT_ID1(MIPI_DSI_BLANKING_PACKET) | PKT_LEN1(1) |
+	       PKT_ID2(MIPI_DSI_H_SYNC_END) | PKT_LEN2(0) |
+	       PKT_LP,
+	[ 5] = 0,
+	[ 6] = PKT_ID0(MIPI_DSI_H_SYNC_START) | PKT_LEN0(0) |
+	       PKT_ID1(MIPI_DSI_BLANKING_PACKET) | PKT_LEN1(1) |
+	       PKT_ID2(MIPI_DSI_H_SYNC_END) | PKT_LEN2(0),
+	[ 7] = PKT_ID0(MIPI_DSI_BLANKING_PACKET) | PKT_LEN0(2) |
+	       PKT_ID1(MIPI_DSI_PACKED_PIXEL_STREAM_24) | PKT_LEN1(3) |
+	       PKT_ID2(MIPI_DSI_BLANKING_PACKET) | PKT_LEN2(4),
+	[ 8] = PKT_ID0(MIPI_DSI_H_SYNC_START) | PKT_LEN0(0) |
+	       PKT_ID1(MIPI_DSI_BLANKING_PACKET) | PKT_LEN1(1) |
+	       PKT_ID2(MIPI_DSI_H_SYNC_END) | PKT_LEN2(0) |
+	       PKT_LP,
+	[ 9] = 0,
+	[10] = PKT_ID0(MIPI_DSI_H_SYNC_START) | PKT_LEN0(0) |
+	       PKT_ID1(MIPI_DSI_BLANKING_PACKET) | PKT_LEN1(1) |
+	       PKT_ID2(MIPI_DSI_H_SYNC_END) | PKT_LEN2(0),
+	[11] = PKT_ID0(MIPI_DSI_BLANKING_PACKET) | PKT_LEN0(2) |
+	       PKT_ID1(MIPI_DSI_PACKED_PIXEL_STREAM_24) | PKT_LEN1(3) |
+	       PKT_ID2(MIPI_DSI_BLANKING_PACKET) | PKT_LEN2(4),
+};
+
+static int tegra_dsi_set_phy_timing(struct tegra_dsi *dsi)
+{
+	struct mipi_dphy_timing timing;
+	unsigned long value, period;
+	long rate;
+	int err;
+
+	rate = clk_get_rate(dsi->clk);
+	if (rate < 0)
+		return rate;
+
+	period = DIV_ROUND_CLOSEST(1000000000UL, rate * 2);
+
+	err = mipi_dphy_timing_get_default(&timing, period);
+	if (err < 0)
+		return err;
+
+	err = mipi_dphy_timing_validate(&timing, period);
+	if (err < 0) {
+		dev_err(dsi->dev, "failed to validate D-PHY timing: %d\n", err);
+		return err;
+	}
+
+	/*
+	 * The D-PHY timing fields below are expressed in byte-clock cycles,
+	 * so multiply the period by 8.
+	 */
+	period *= 8;
+
+	value = DSI_TIMING_FIELD(timing.hsexit, period, 1) << 24 |
+		DSI_TIMING_FIELD(timing.hstrail, period, 0) << 16 |
+		DSI_TIMING_FIELD(timing.hszero, period, 3) << 8 |
+		DSI_TIMING_FIELD(timing.hsprepare, period, 1);
+	tegra_dsi_writel(dsi, value, DSI_PHY_TIMING_0);
+
+	value = DSI_TIMING_FIELD(timing.clktrail, period, 1) << 24 |
+		DSI_TIMING_FIELD(timing.clkpost, period, 1) << 16 |
+		DSI_TIMING_FIELD(timing.clkzero, period, 1) << 8 |
+		DSI_TIMING_FIELD(timing.lpx, period, 1);
+	tegra_dsi_writel(dsi, value, DSI_PHY_TIMING_1);
+
+	value = DSI_TIMING_FIELD(timing.clkprepare, period, 1) << 16 |
+		DSI_TIMING_FIELD(timing.clkpre, period, 1) << 8 |
+		DSI_TIMING_FIELD(0xff * period, period, 0) << 0;
+	tegra_dsi_writel(dsi, value, DSI_PHY_TIMING_2);
+
+	value = DSI_TIMING_FIELD(timing.taget, period, 1) << 16 |
+		DSI_TIMING_FIELD(timing.tasure, period, 1) << 8 |
+		DSI_TIMING_FIELD(timing.tago, period, 1);
+	tegra_dsi_writel(dsi, value, DSI_BTA_TIMING);
+
+	return 0;
+}
+
+static int tegra_dsi_get_muldiv(enum mipi_dsi_pixel_format format,
+				unsigned int *mulp, unsigned int *divp)
+{
+	switch (format) {
+	case MIPI_DSI_FMT_RGB666_PACKED:
+	case MIPI_DSI_FMT_RGB888:
+		*mulp = 3;
+		*divp = 1;
+		break;
+
+	case MIPI_DSI_FMT_RGB565:
+		*mulp = 2;
+		*divp = 1;
+		break;
+
+	case MIPI_DSI_FMT_RGB666:
+		*mulp = 9;
+		*divp = 4;
+		break;
+
+	default:
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int tegra_output_dsi_enable(struct tegra_output *output)
+{
+	struct tegra_dc *dc = to_tegra_dc(output->encoder.crtc);
+	struct drm_display_mode *mode = &dc->base.mode;
+	unsigned int hact, hsw, hbp, hfp, i, mul, div;
+	struct tegra_dsi *dsi = to_dsi(output);
+	/* FIXME: don't hardcode this */
+	const u32 *pkt_seq = pkt_seq_vnb_syne;
+	unsigned long value;
+	int err;
+
+	err = tegra_dsi_get_muldiv(dsi->format, &mul, &div);
+	if (err < 0)
+		return err;
+
+	err = clk_enable(dsi->clk);
+	if (err < 0)
+		return err;
+
+	reset_control_deassert(dsi->rst);
+
+	value = DSI_CONTROL_CHANNEL(0) | DSI_CONTROL_FORMAT(dsi->format) |
+		DSI_CONTROL_LANES(dsi->lanes - 1) |
+		DSI_CONTROL_SOURCE(dc->pipe);
+	tegra_dsi_writel(dsi, value, DSI_CONTROL);
+
+	tegra_dsi_writel(dsi, DSI_VIDEO_FIFO_DEPTH, DSI_MAX_THRESHOLD);
+
+	value = DSI_HOST_CONTROL_HS | DSI_HOST_CONTROL_CS |
+		DSI_HOST_CONTROL_ECC;
+	tegra_dsi_writel(dsi, value, DSI_HOST_CONTROL);
+
+	value = tegra_dsi_readl(dsi, DSI_CONTROL);
+	value |= DSI_CONTROL_HS_CLK_CTRL;
+	value &= ~DSI_CONTROL_TX_TRIG(3);
+	value &= ~DSI_CONTROL_DCS_ENABLE;
+	value |= DSI_CONTROL_VIDEO_ENABLE;
+	value &= ~DSI_CONTROL_HOST_ENABLE;
+	tegra_dsi_writel(dsi, value, DSI_CONTROL);
+
+	err = tegra_dsi_set_phy_timing(dsi);
+	if (err < 0)
+		return err;
+
+	for (i = 0; i < NUM_PKT_SEQ; i++)
+		tegra_dsi_writel(dsi, pkt_seq[i], DSI_PKT_SEQ_0_LO + i);
+
+	/* horizontal active pixels */
+	hact = mode->hdisplay * mul / div;
+
+	/* horizontal sync width */
+	hsw = (mode->hsync_end - mode->hsync_start) * mul / div;
+	hsw -= 10;
+
+	/* horizontal back porch */
+	hbp = (mode->htotal - mode->hsync_end) * mul / div;
+	hbp -= 14;
+
+	/* horizontal front porch */
+	hfp = (mode->hsync_start  - mode->hdisplay) * mul / div;
+	hfp -= 8;
+
+	tegra_dsi_writel(dsi, hsw << 16 | 0, DSI_PKT_LEN_0_1);
+	tegra_dsi_writel(dsi, hact << 16 | hbp, DSI_PKT_LEN_2_3);
+	tegra_dsi_writel(dsi, hfp, DSI_PKT_LEN_4_5);
+	tegra_dsi_writel(dsi, 0x0f0f << 16, DSI_PKT_LEN_6_7);
+
+	/* set SOL delay */
+	tegra_dsi_writel(dsi, 8 * mul / div, DSI_SOL_DELAY);
+
+	/* enable display controller */
+	value = tegra_dc_readl(dc, DC_DISP_DISP_WIN_OPTIONS);
+	value |= DSI_ENABLE;
+	tegra_dc_writel(dc, value, DC_DISP_DISP_WIN_OPTIONS);
+
+	value = tegra_dc_readl(dc, DC_CMD_DISPLAY_COMMAND);
+	value &= ~DISP_CTRL_MODE_MASK;
+	value |= DISP_CTRL_MODE_C_DISPLAY;
+	tegra_dc_writel(dc, value, DC_CMD_DISPLAY_COMMAND);
+
+	value = tegra_dc_readl(dc, DC_CMD_DISPLAY_POWER_CONTROL);
+	value |= PW0_ENABLE | PW1_ENABLE | PW2_ENABLE | PW3_ENABLE |
+		 PW4_ENABLE | PM0_ENABLE | PM1_ENABLE;
+	tegra_dc_writel(dc, value, DC_CMD_DISPLAY_POWER_CONTROL);
+
+	tegra_dc_writel(dc, GENERAL_ACT_REQ << 8, DC_CMD_STATE_CONTROL);
+	tegra_dc_writel(dc, GENERAL_ACT_REQ, DC_CMD_STATE_CONTROL);
+
+	/* enable DSI controller */
+	value = tegra_dsi_readl(dsi, DSI_POWER_CONTROL);
+	value |= DSI_POWER_CONTROL_ENABLE;
+	tegra_dsi_writel(dsi, value, DSI_POWER_CONTROL);
+
+	return 0;
+}
+
+static int tegra_output_dsi_disable(struct tegra_output *output)
+{
+	struct tegra_dc *dc = to_tegra_dc(output->encoder.crtc);
+	struct tegra_dsi *dsi = to_dsi(output);
+	unsigned long value;
+
+	/* disable DSI controller */
+	value = tegra_dsi_readl(dsi, DSI_POWER_CONTROL);
+	value &= DSI_POWER_CONTROL_ENABLE;
+	tegra_dsi_writel(dsi, value, DSI_POWER_CONTROL);
+
+	/*
+	 * The following accesses registers of the display controller, so make
+	 * sure it's only executed when the output is attached to one.
+	 */
+	if (dc) {
+		value = tegra_dc_readl(dc, DC_CMD_DISPLAY_POWER_CONTROL);
+		value &= ~(PW0_ENABLE | PW1_ENABLE | PW2_ENABLE | PW3_ENABLE |
+			   PW4_ENABLE | PM0_ENABLE | PM1_ENABLE);
+		tegra_dc_writel(dc, value, DC_CMD_DISPLAY_POWER_CONTROL);
+
+		value = tegra_dc_readl(dc, DC_CMD_DISPLAY_COMMAND);
+		value &= ~DISP_CTRL_MODE_MASK;
+		tegra_dc_writel(dc, value, DC_CMD_DISPLAY_COMMAND);
+
+		value = tegra_dc_readl(dc, DC_DISP_DISP_WIN_OPTIONS);
+		value &= ~DSI_ENABLE;
+		tegra_dc_writel(dc, value, DC_DISP_DISP_WIN_OPTIONS);
+
+		tegra_dc_writel(dc, GENERAL_ACT_REQ << 8, DC_CMD_STATE_CONTROL);
+		tegra_dc_writel(dc, GENERAL_ACT_REQ, DC_CMD_STATE_CONTROL);
+	}
+
+	clk_disable(dsi->clk);
+
+	return 0;
+}
+
+static int tegra_output_dsi_setup_clock(struct tegra_output *output,
+					struct clk *clk, unsigned long pclk)
+{
+	struct tegra_dc *dc = to_tegra_dc(output->encoder.crtc);
+	struct drm_display_mode *mode = &dc->base.mode;
+	unsigned int timeout, mul, div, vrefresh;
+	struct tegra_dsi *dsi = to_dsi(output);
+	unsigned long bclk, plld, value;
+	struct clk *base;
+	int err;
+
+	err = tegra_dsi_get_muldiv(dsi->format, &mul, &div);
+	if (err < 0)
+		return err;
+
+	vrefresh = drm_mode_vrefresh(mode);
+
+	pclk = mode->htotal * mode->vtotal * vrefresh;
+	bclk = (pclk * mul) / (div * dsi->lanes);
+	plld = DIV_ROUND_UP(bclk * 8, 1000000);
+	pclk = (plld * 1000000) / 2;
+
+	err = clk_set_parent(clk, dsi->clk_parent);
+	if (err < 0) {
+		dev_err(dsi->dev, "failed to set parent clock: %d\n", err);
+		return err;
+	}
+
+	base = clk_get_parent(dsi->clk_parent);
+
+	/*
+	 * This assumes that the parent clock is pll_d_out0 or pll_d2_out
+	 * respectively, each of which divides the base pll_d by 2.
+	 */
+	err = clk_set_rate(base, pclk * 2);
+	if (err < 0) {
+		dev_err(dsi->dev, "failed to set base clock rate to %lu Hz\n",
+			pclk * 2);
+		return err;
+	}
+
+	/*
+	 * XXX: Move the below somewhere else so that we don't need to have
+	 * access to the vrefresh in this function?
+	 */
+
+	/* one frame high-speed transmission timeout */
+	timeout = (bclk / vrefresh) / 512;
+	value = DSI_TIMEOUT_LRX(0x2000) | DSI_TIMEOUT_HTX(timeout);
+	tegra_dsi_writel(dsi, value, DSI_TIMEOUT_0);
+
+	/* 2 ms peripheral timeout for panel */
+	timeout = 2 * bclk / 512 * 1000;
+	value = DSI_TIMEOUT_PR(timeout) | DSI_TIMEOUT_TA(0x2000);
+	tegra_dsi_writel(dsi, value, DSI_TIMEOUT_1);
+
+	value = DSI_TALLY_TA(0) | DSI_TALLY_LRX(0) | DSI_TALLY_HTX(0);
+	tegra_dsi_writel(dsi, value, DSI_TO_TALLY);
+
+	return 0;
+}
+
+static int tegra_output_dsi_check_mode(struct tegra_output *output,
+				       struct drm_display_mode *mode,
+				       enum drm_mode_status *status)
+{
+	/*
+	 * FIXME: For now, always assume that the mode is okay.
+	 */
+
+	*status = MODE_OK;
+
+	return 0;
+}
+
+static const struct tegra_output_ops dsi_ops = {
+	.enable = tegra_output_dsi_enable,
+	.disable = tegra_output_dsi_disable,
+	.setup_clock = tegra_output_dsi_setup_clock,
+	.check_mode = tegra_output_dsi_check_mode,
+};
+
+static int tegra_dsi_pad_enable(struct tegra_dsi *dsi)
+{
+	unsigned long value;
+
+	value = DSI_PAD_CONTROL_VS1_PULLDN(0) | DSI_PAD_CONTROL_VS1_PDIO(0);
+	tegra_dsi_writel(dsi, value, DSI_PAD_CONTROL_0);
+
+	return 0;
+}
+
+static int tegra_dsi_pad_calibrate(struct tegra_dsi *dsi)
+{
+	unsigned long value;
+
+	tegra_dsi_writel(dsi, 0, DSI_PAD_CONTROL_0);
+	tegra_dsi_writel(dsi, 0, DSI_PAD_CONTROL_1);
+	tegra_dsi_writel(dsi, 0, DSI_PAD_CONTROL_2);
+	tegra_dsi_writel(dsi, 0, DSI_PAD_CONTROL_3);
+	tegra_dsi_writel(dsi, 0, DSI_PAD_CONTROL_4);
+
+	/* start calibration */
+	tegra_dsi_pad_enable(dsi);
+
+	value = DSI_PAD_SLEW_UP(0x7) | DSI_PAD_SLEW_DN(0x7) |
+		DSI_PAD_LP_UP(0x1) | DSI_PAD_LP_DN(0x1) |
+		DSI_PAD_OUT_CLK(0x0);
+	tegra_dsi_writel(dsi, value, DSI_PAD_CONTROL_2);
+
+	return tegra_mipi_calibrate(dsi->mipi);
+}
+
+static int tegra_dsi_init(struct host1x_client *client)
+{
+	struct tegra_drm *tegra = dev_get_drvdata(client->parent);
+	struct tegra_dsi *dsi = host1x_client_to_dsi(client);
+	unsigned long value, i;
+	int err;
+
+	dsi->output.type = TEGRA_OUTPUT_DSI;
+	dsi->output.dev = client->dev;
+	dsi->output.ops = &dsi_ops;
+
+	err = tegra_output_init(tegra->drm, &dsi->output);
+	if (err < 0) {
+		dev_err(client->dev, "output setup failed: %d\n", err);
+		return err;
+	}
+
+	if (IS_ENABLED(CONFIG_DEBUG_FS)) {
+		err = tegra_dsi_debugfs_init(dsi, tegra->drm->primary);
+		if (err < 0)
+			dev_err(dsi->dev, "debugfs setup failed: %d\n", err);
+	}
+
+	/*
+	 * enable high-speed mode, checksum generation, ECC generation and
+	 * disable raw mode
+	 */
+	value = tegra_dsi_readl(dsi, DSI_HOST_CONTROL);
+	value |= DSI_HOST_CONTROL_ECC | DSI_HOST_CONTROL_CS |
+		 DSI_HOST_CONTROL_HS;
+	value &= ~DSI_HOST_CONTROL_RAW;
+	tegra_dsi_writel(dsi, value, DSI_HOST_CONTROL);
+
+	tegra_dsi_writel(dsi, 0, DSI_SOL_DELAY);
+	tegra_dsi_writel(dsi, 0, DSI_MAX_THRESHOLD);
+
+	tegra_dsi_writel(dsi, 0, DSI_INIT_SEQ_CONTROL);
+
+	for (i = 0; i < 8; i++) {
+		tegra_dsi_writel(dsi, 0, DSI_INIT_SEQ_DATA_0 + i);
+		tegra_dsi_writel(dsi, 0, DSI_INIT_SEQ_DATA_8 + i);
+	}
+
+	for (i = 0; i < 12; i++)
+		tegra_dsi_writel(dsi, 0, DSI_PKT_SEQ_0_LO + i);
+
+	tegra_dsi_writel(dsi, 0, DSI_DCS_CMDS);
+
+	err = tegra_dsi_pad_calibrate(dsi);
+	if (err < 0) {
+		dev_err(dsi->dev, "MIPI calibration failed: %d\n", err);
+		return err;
+	}
+
+	tegra_dsi_writel(dsi, DSI_POWER_CONTROL_ENABLE, DSI_POWER_CONTROL);
+	usleep_range(300, 1000);
+
+	return 0;
+}
+
+static int tegra_dsi_exit(struct host1x_client *client)
+{
+	struct tegra_dsi *dsi = host1x_client_to_dsi(client);
+	int err;
+
+	if (IS_ENABLED(CONFIG_DEBUG_FS)) {
+		err = tegra_dsi_debugfs_exit(dsi);
+		if (err < 0)
+			dev_err(dsi->dev, "debugfs cleanup failed: %d\n", err);
+	}
+
+	err = tegra_output_disable(&dsi->output);
+	if (err < 0) {
+		dev_err(client->dev, "output failed to disable: %d\n", err);
+		return err;
+	}
+
+	err = tegra_output_exit(&dsi->output);
+	if (err < 0) {
+		dev_err(client->dev, "output cleanup failed: %d\n", err);
+		return err;
+	}
+
+	return 0;
+}
+
+static const struct host1x_client_ops dsi_client_ops = {
+	.init = tegra_dsi_init,
+	.exit = tegra_dsi_exit,
+};
+
+static int tegra_dsi_setup_clocks(struct tegra_dsi *dsi)
+{
+	struct clk *parent;
+	int err;
+
+	parent = clk_get_parent(dsi->clk);
+	if (!parent)
+		return -EINVAL;
+
+	err = clk_set_parent(parent, dsi->clk_parent);
+	if (err < 0)
+		return err;
+
+	return 0;
+}
+
+static void tegra_dsi_initialize(struct tegra_dsi *dsi)
+{
+	unsigned int i;
+
+	tegra_dsi_writel(dsi, 0, DSI_POWER_CONTROL);
+
+	tegra_dsi_writel(dsi, 0, DSI_INT_ENABLE);
+	tegra_dsi_writel(dsi, 0, DSI_INT_STATUS);
+	tegra_dsi_writel(dsi, 0, DSI_INT_MASK);
+
+	tegra_dsi_writel(dsi, 0, DSI_HOST_CONTROL);
+	tegra_dsi_writel(dsi, 0, DSI_CONTROL);
+
+	tegra_dsi_writel(dsi, 0, DSI_SOL_DELAY);
+	tegra_dsi_writel(dsi, 0, DSI_MAX_THRESHOLD);
+
+	tegra_dsi_writel(dsi, 0, DSI_INIT_SEQ_CONTROL);
+
+	for (i = 0; i < 8; i++) {
+		tegra_dsi_writel(dsi, 0, DSI_INIT_SEQ_DATA_0 + i);
+		tegra_dsi_writel(dsi, 0, DSI_INIT_SEQ_DATA_8 + i);
+	}
+
+	for (i = 0; i < 12; i++)
+		tegra_dsi_writel(dsi, 0, DSI_PKT_SEQ_0_LO + i);
+
+	tegra_dsi_writel(dsi, 0, DSI_DCS_CMDS);
+
+	for (i = 0; i < 4; i++)
+		tegra_dsi_writel(dsi, 0, DSI_PKT_LEN_0_1 + i);
+
+	tegra_dsi_writel(dsi, 0x00000000, DSI_PHY_TIMING_0);
+	tegra_dsi_writel(dsi, 0x00000000, DSI_PHY_TIMING_1);
+	tegra_dsi_writel(dsi, 0x000000ff, DSI_PHY_TIMING_2);
+	tegra_dsi_writel(dsi, 0x00000000, DSI_BTA_TIMING);
+
+	tegra_dsi_writel(dsi, 0, DSI_TIMEOUT_0);
+	tegra_dsi_writel(dsi, 0, DSI_TIMEOUT_1);
+	tegra_dsi_writel(dsi, 0, DSI_TO_TALLY);
+
+	tegra_dsi_writel(dsi, 0, DSI_PAD_CONTROL_0);
+	tegra_dsi_writel(dsi, 0, DSI_PAD_CONTROL_CD);
+	tegra_dsi_writel(dsi, 0, DSI_PAD_CD_STATUS);
+	tegra_dsi_writel(dsi, 0, DSI_VIDEO_MODE_CONTROL);
+	tegra_dsi_writel(dsi, 0, DSI_PAD_CONTROL_1);
+	tegra_dsi_writel(dsi, 0, DSI_PAD_CONTROL_2);
+	tegra_dsi_writel(dsi, 0, DSI_PAD_CONTROL_3);
+	tegra_dsi_writel(dsi, 0, DSI_PAD_CONTROL_4);
+
+	tegra_dsi_writel(dsi, 0, DSI_GANGED_MODE_CONTROL);
+	tegra_dsi_writel(dsi, 0, DSI_GANGED_MODE_START);
+	tegra_dsi_writel(dsi, 0, DSI_GANGED_MODE_SIZE);
+}
+
+static int tegra_dsi_host_attach(struct mipi_dsi_host *host,
+				 struct mipi_dsi_device *device)
+{
+	struct tegra_dsi *dsi = host_to_tegra(host);
+	struct tegra_output *output = &dsi->output;
+
+	dsi->format = device->format;
+	dsi->lanes = device->lanes;
+
+	output->panel = of_drm_find_panel(device->dev.of_node);
+	if (output->panel) {
+		if (output->connector.dev)
+			drm_helper_hpd_irq_event(output->connector.dev);
+	}
+
+	return 0;
+}
+
+static int tegra_dsi_host_detach(struct mipi_dsi_host *host,
+				 struct mipi_dsi_device *device)
+{
+	struct tegra_dsi *dsi = host_to_tegra(host);
+	struct tegra_output *output = &dsi->output;
+
+	if (output->panel && &device->dev == output->panel->dev) {
+		if (output->connector.dev)
+			drm_helper_hpd_irq_event(output->connector.dev);
+
+		output->panel = NULL;
+	}
+
+	return 0;
+}
+
+static const struct mipi_dsi_host_ops tegra_dsi_host_ops = {
+	.attach = tegra_dsi_host_attach,
+	.detach = tegra_dsi_host_detach,
+};
+
+static int tegra_dsi_probe(struct platform_device *pdev)
+{
+	struct tegra_dsi *dsi;
+	struct resource *regs;
+	int err;
+
+	dsi = devm_kzalloc(&pdev->dev, sizeof(*dsi), GFP_KERNEL);
+	if (!dsi)
+		return -ENOMEM;
+
+	dsi->output.dev = dsi->dev = &pdev->dev;
+
+	err = tegra_output_probe(&dsi->output);
+	if (err < 0)
+		return err;
+
+	/*
+	 * Assume these values by default. When a DSI peripheral driver
+	 * attaches to the DSI host, the parameters will be taken from
+	 * the attached device.
+	 */
+	dsi->format = MIPI_DSI_FMT_RGB888;
+	dsi->lanes = 4;
+
+	dsi->rst = devm_reset_control_get(&pdev->dev, "dsi");
+	if (IS_ERR(dsi->rst))
+		return PTR_ERR(dsi->rst);
+
+	dsi->clk = devm_clk_get(&pdev->dev, NULL);
+	if (IS_ERR(dsi->clk)) {
+		dev_err(&pdev->dev, "cannot get DSI clock\n");
+		return PTR_ERR(dsi->clk);
+	}
+
+	err = clk_prepare_enable(dsi->clk);
+	if (err < 0) {
+		dev_err(&pdev->dev, "cannot enable DSI clock\n");
+		return err;
+	}
+
+	dsi->clk_lp = devm_clk_get(&pdev->dev, "lp");
+	if (IS_ERR(dsi->clk_lp)) {
+		dev_err(&pdev->dev, "cannot get low-power clock\n");
+		return PTR_ERR(dsi->clk_lp);
+	}
+
+	err = clk_prepare_enable(dsi->clk_lp);
+	if (err < 0) {
+		dev_err(&pdev->dev, "cannot enable low-power clock\n");
+		return err;
+	}
+
+	dsi->clk_parent = devm_clk_get(&pdev->dev, "parent");
+	if (IS_ERR(dsi->clk_parent)) {
+		dev_err(&pdev->dev, "cannot get parent clock\n");
+		return PTR_ERR(dsi->clk_parent);
+	}
+
+	err = clk_prepare_enable(dsi->clk_parent);
+	if (err < 0) {
+		dev_err(&pdev->dev, "cannot enable parent clock\n");
+		return err;
+	}
+
+	err = tegra_dsi_setup_clocks(dsi);
+	if (err < 0) {
+		dev_err(&pdev->dev, "cannot setup clocks\n");
+		return err;
+	}
+
+	regs = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	dsi->regs = devm_ioremap_resource(&pdev->dev, regs);
+	if (IS_ERR(dsi->regs))
+		return PTR_ERR(dsi->regs);
+
+	tegra_dsi_initialize(dsi);
+
+	dsi->mipi = tegra_mipi_request(&pdev->dev);
+	if (IS_ERR(dsi->mipi))
+		return PTR_ERR(dsi->mipi);
+
+	dsi->host.ops = &tegra_dsi_host_ops;
+	dsi->host.dev = &pdev->dev;
+
+	err = mipi_dsi_host_register(&dsi->host);
+	if (err < 0) {
+		dev_err(&pdev->dev, "failed to register DSI host: %d\n", err);
+		return err;
+	}
+
+	INIT_LIST_HEAD(&dsi->client.list);
+	dsi->client.ops = &dsi_client_ops;
+	dsi->client.dev = &pdev->dev;
+
+	err = host1x_client_register(&dsi->client);
+	if (err < 0) {
+		dev_err(&pdev->dev, "failed to register host1x client: %d\n",
+			err);
+		return err;
+	}
+
+	platform_set_drvdata(pdev, dsi);
+
+	return 0;
+}
+
+static int tegra_dsi_remove(struct platform_device *pdev)
+{
+	struct tegra_dsi *dsi = platform_get_drvdata(pdev);
+	int err;
+
+	err = host1x_client_unregister(&dsi->client);
+	if (err < 0) {
+		dev_err(&pdev->dev, "failed to unregister host1x client: %d\n",
+			err);
+		return err;
+	}
+
+	mipi_dsi_host_unregister(&dsi->host);
+	tegra_mipi_free(dsi->mipi);
+
+	clk_disable_unprepare(dsi->clk_parent);
+	clk_disable_unprepare(dsi->clk_lp);
+	clk_disable_unprepare(dsi->clk);
+
+	err = tegra_output_remove(&dsi->output);
+	if (err < 0) {
+		dev_err(&pdev->dev, "failed to remove output: %d\n", err);
+		return err;
+	}
+
+	return 0;
+}
+
+static const struct of_device_id tegra_dsi_of_match[] = {
+	{ .compatible = "nvidia,tegra114-dsi", },
+	{ },
+};
+
+struct platform_driver tegra_dsi_driver = {
+	.driver = {
+		.name = "tegra-dsi",
+		.of_match_table = tegra_dsi_of_match,
+	},
+	.probe = tegra_dsi_probe,
+	.remove = tegra_dsi_remove,
+};
diff --git a/drivers/gpu/drm/tegra/dsi.h b/drivers/gpu/drm/tegra/dsi.h
new file mode 100644
index 000000000000..00e79c1f448c
--- /dev/null
+++ b/drivers/gpu/drm/tegra/dsi.h
@@ -0,0 +1,134 @@
+/*
+ * Copyright (C) 2013 NVIDIA Corporation
+ *
+ * Permission to use, copy, modify, distribute, and sell this software and its
+ * documentation for any purpose is hereby granted without fee, provided that
+ * the above copyright notice appear in all copies and that both that copyright
+ * notice and this permission notice appear in supporting documentation, and
+ * that the name of the copyright holders not be used in advertising or
+ * publicity pertaining to distribution of the software without specific,
+ * written prior permission.  The copyright holders make no representations
+ * about the suitability of this software for any purpose.  It is provided "as
+ * is" without express or implied warranty.
+ *
+ * THE COPYRIGHT HOLDERS DISCLAIM ALL WARRANTIES WITH REGARD TO THIS SOFTWARE,
+ * INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN NO
+ * EVENT SHALL THE COPYRIGHT HOLDERS BE LIABLE FOR ANY SPECIAL, INDIRECT OR
+ * CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE,
+ * DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER
+ * TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE
+ * OF THIS SOFTWARE.
+ */
+
+#ifndef DRM_TEGRA_DSI_H
+#define DRM_TEGRA_DSI_H
+
+#define DSI_INCR_SYNCPT			0x00
+#define DSI_INCR_SYNCPT_CONTROL		0x01
+#define DSI_INCR_SYNCPT_ERROR		0x02
+#define DSI_CTXSW			0x08
+#define DSI_RD_DATA			0x09
+#define DSI_WR_DATA			0x0a
+#define DSI_POWER_CONTROL		0x0b
+#define DSI_POWER_CONTROL_ENABLE	(1 << 0)
+#define DSI_INT_ENABLE			0x0c
+#define DSI_INT_STATUS			0x0d
+#define DSI_INT_MASK			0x0e
+#define DSI_HOST_CONTROL		0x0f
+#define DSI_HOST_CONTROL_RAW		(1 << 6)
+#define DSI_HOST_CONTROL_HS		(1 << 5)
+#define DSI_HOST_CONTROL_BTA		(1 << 2)
+#define DSI_HOST_CONTROL_CS		(1 << 1)
+#define DSI_HOST_CONTROL_ECC		(1 << 0)
+#define DSI_CONTROL			0x10
+#define DSI_CONTROL_HS_CLK_CTRL		(1 << 20)
+#define DSI_CONTROL_CHANNEL(c)		(((c) & 0x3) << 16)
+#define DSI_CONTROL_FORMAT(f)		(((f) & 0x3) << 12)
+#define DSI_CONTROL_TX_TRIG(x)		(((x) & 0x3) <<  8)
+#define DSI_CONTROL_LANES(n)		(((n) & 0x3) <<  4)
+#define DSI_CONTROL_DCS_ENABLE		(1 << 3)
+#define DSI_CONTROL_SOURCE(s)		(((s) & 0x1) <<  2)
+#define DSI_CONTROL_VIDEO_ENABLE	(1 << 1)
+#define DSI_CONTROL_HOST_ENABLE		(1 << 0)
+#define DSI_SOL_DELAY			0x11
+#define DSI_MAX_THRESHOLD		0x12
+#define DSI_TRIGGER			0x13
+#define DSI_TX_CRC			0x14
+#define DSI_STATUS			0x15
+#define DSI_STATUS_IDLE			(1 << 10)
+#define DSI_INIT_SEQ_CONTROL		0x1a
+#define DSI_INIT_SEQ_DATA_0		0x1b
+#define DSI_INIT_SEQ_DATA_1		0x1c
+#define DSI_INIT_SEQ_DATA_2		0x1d
+#define DSI_INIT_SEQ_DATA_3		0x1e
+#define DSI_INIT_SEQ_DATA_4		0x1f
+#define DSI_INIT_SEQ_DATA_5		0x20
+#define DSI_INIT_SEQ_DATA_6		0x21
+#define DSI_INIT_SEQ_DATA_7		0x22
+#define DSI_PKT_SEQ_0_LO		0x23
+#define DSI_PKT_SEQ_0_HI		0x24
+#define DSI_PKT_SEQ_1_LO		0x25
+#define DSI_PKT_SEQ_1_HI		0x26
+#define DSI_PKT_SEQ_2_LO		0x27
+#define DSI_PKT_SEQ_2_HI		0x28
+#define DSI_PKT_SEQ_3_LO		0x29
+#define DSI_PKT_SEQ_3_HI		0x2a
+#define DSI_PKT_SEQ_4_LO		0x2b
+#define DSI_PKT_SEQ_4_HI		0x2c
+#define DSI_PKT_SEQ_5_LO		0x2d
+#define DSI_PKT_SEQ_5_HI		0x2e
+#define DSI_DCS_CMDS			0x33
+#define DSI_PKT_LEN_0_1			0x34
+#define DSI_PKT_LEN_2_3			0x35
+#define DSI_PKT_LEN_4_5			0x36
+#define DSI_PKT_LEN_6_7			0x37
+#define DSI_PHY_TIMING_0		0x3c
+#define DSI_PHY_TIMING_1		0x3d
+#define DSI_PHY_TIMING_2		0x3e
+#define DSI_BTA_TIMING			0x3f
+
+#define DSI_TIMING_FIELD(value, period, hwinc) \
+	((DIV_ROUND_CLOSEST(value, period) - (hwinc)) & 0xff)
+
+#define DSI_TIMEOUT_0			0x44
+#define DSI_TIMEOUT_LRX(x)		(((x) & 0xffff) << 16)
+#define DSI_TIMEOUT_HTX(x)		(((x) & 0xffff) <<  0)
+#define DSI_TIMEOUT_1			0x45
+#define DSI_TIMEOUT_PR(x)		(((x) & 0xffff) << 16)
+#define DSI_TIMEOUT_TA(x)		(((x) & 0xffff) <<  0)
+#define DSI_TO_TALLY			0x46
+#define DSI_TALLY_TA(x)			(((x) & 0xff) << 16)
+#define DSI_TALLY_LRX(x)		(((x) & 0xff) <<  8)
+#define DSI_TALLY_HTX(x)		(((x) & 0xff) <<  0)
+#define DSI_PAD_CONTROL_0		0x4b
+#define DSI_PAD_CONTROL_VS1_PDIO(x)	(((x) & 0xf) <<  0)
+#define DSI_PAD_CONTROL_VS1_PDIO_CLK	(1 <<  8)
+#define DSI_PAD_CONTROL_VS1_PULLDN(x)	(((x) & 0xf) << 16)
+#define DSI_PAD_CONTROL_VS1_PULLDN_CLK	(1 << 24)
+#define DSI_PAD_CONTROL_CD		0x4c
+#define DSI_PAD_CD_STATUS		0x4d
+#define DSI_VIDEO_MODE_CONTROL		0x4e
+#define DSI_PAD_CONTROL_1		0x4f
+#define DSI_PAD_CONTROL_2		0x50
+#define DSI_PAD_OUT_CLK(x)		(((x) & 0x7) <<  0)
+#define DSI_PAD_LP_DN(x)		(((x) & 0x7) <<  4)
+#define DSI_PAD_LP_UP(x)		(((x) & 0x7) <<  8)
+#define DSI_PAD_SLEW_DN(x)		(((x) & 0x7) << 12)
+#define DSI_PAD_SLEW_UP(x)		(((x) & 0x7) << 16)
+#define DSI_PAD_CONTROL_3		0x51
+#define DSI_PAD_CONTROL_4		0x52
+#define DSI_GANGED_MODE_CONTROL		0x53
+#define DSI_GANGED_MODE_START		0x54
+#define DSI_GANGED_MODE_SIZE		0x55
+#define DSI_RAW_DATA_BYTE_COUNT		0x56
+#define DSI_ULTRA_LOW_POWER_CONTROL	0x57
+#define DSI_INIT_SEQ_DATA_8		0x58
+#define DSI_INIT_SEQ_DATA_9		0x59
+#define DSI_INIT_SEQ_DATA_10		0x5a
+#define DSI_INIT_SEQ_DATA_11		0x5b
+#define DSI_INIT_SEQ_DATA_12		0x5c
+#define DSI_INIT_SEQ_DATA_13		0x5d
+#define DSI_INIT_SEQ_DATA_14		0x5e
+#define DSI_INIT_SEQ_DATA_15		0x5f
+
+#endif
diff --git a/drivers/gpu/drm/tegra/fb.c b/drivers/gpu/drm/tegra/fb.c
index a3835e7de184..f7fca09d4921 100644
--- a/drivers/gpu/drm/tegra/fb.c
+++ b/drivers/gpu/drm/tegra/fb.c
@@ -18,10 +18,12 @@ static inline struct tegra_fb *to_tegra_fb(struct drm_framebuffer *fb)
 	return container_of(fb, struct tegra_fb, base);
 }
 
+#ifdef CONFIG_DRM_TEGRA_FBDEV
 static inline struct tegra_fbdev *to_tegra_fbdev(struct drm_fb_helper *helper)
 {
 	return container_of(helper, struct tegra_fbdev, base);
 }
+#endif
 
 struct tegra_bo *tegra_fb_get_plane(struct drm_framebuffer *framebuffer,
 				    unsigned int index)
@@ -98,8 +100,10 @@ static struct tegra_fb *tegra_fb_alloc(struct drm_device *drm,
 		return ERR_PTR(-ENOMEM);
 
 	fb->planes = kzalloc(num_planes * sizeof(*planes), GFP_KERNEL);
-	if (!fb->planes)
+	if (!fb->planes) {
+		kfree(fb);
 		return ERR_PTR(-ENOMEM);
+	}
 
 	fb->num_planes = num_planes;
 
@@ -172,6 +176,7 @@ unreference:
 	return ERR_PTR(err);
 }
 
+#ifdef CONFIG_DRM_TEGRA_FBDEV
 static struct fb_ops tegra_fb_ops = {
 	.owner = THIS_MODULE,
 	.fb_fillrect = sys_fillrect,
@@ -339,6 +344,15 @@ static void tegra_fbdev_free(struct tegra_fbdev *fbdev)
 	kfree(fbdev);
 }
 
+void tegra_fbdev_restore_mode(struct tegra_fbdev *fbdev)
+{
+	if (fbdev) {
+		drm_modeset_lock_all(fbdev->base.dev);
+		drm_fb_helper_restore_fbdev_mode(&fbdev->base);
+		drm_modeset_unlock_all(fbdev->base.dev);
+	}
+}
+
 static void tegra_fb_output_poll_changed(struct drm_device *drm)
 {
 	struct tegra_drm *tegra = drm->dev_private;
@@ -346,16 +360,20 @@ static void tegra_fb_output_poll_changed(struct drm_device *drm)
 	if (tegra->fbdev)
 		drm_fb_helper_hotplug_event(&tegra->fbdev->base);
 }
+#endif
 
 static const struct drm_mode_config_funcs tegra_drm_mode_funcs = {
 	.fb_create = tegra_fb_create,
+#ifdef CONFIG_DRM_TEGRA_FBDEV
 	.output_poll_changed = tegra_fb_output_poll_changed,
+#endif
 };
 
 int tegra_drm_fb_init(struct drm_device *drm)
 {
+#ifdef CONFIG_DRM_TEGRA_FBDEV
 	struct tegra_drm *tegra = drm->dev_private;
-	struct tegra_fbdev *fbdev;
+#endif
 
 	drm->mode_config.min_width = 0;
 	drm->mode_config.min_height = 0;
@@ -365,28 +383,21 @@ int tegra_drm_fb_init(struct drm_device *drm)
 
 	drm->mode_config.funcs = &tegra_drm_mode_funcs;
 
-	fbdev = tegra_fbdev_create(drm, 32, drm->mode_config.num_crtc,
-				   drm->mode_config.num_connector);
-	if (IS_ERR(fbdev))
-		return PTR_ERR(fbdev);
-
-	tegra->fbdev = fbdev;
+#ifdef CONFIG_DRM_TEGRA_FBDEV
+	tegra->fbdev = tegra_fbdev_create(drm, 32, drm->mode_config.num_crtc,
+					  drm->mode_config.num_connector);
+	if (IS_ERR(tegra->fbdev))
+		return PTR_ERR(tegra->fbdev);
+#endif
 
 	return 0;
 }
 
 void tegra_drm_fb_exit(struct drm_device *drm)
 {
+#ifdef CONFIG_DRM_TEGRA_FBDEV
 	struct tegra_drm *tegra = drm->dev_private;
 
 	tegra_fbdev_free(tegra->fbdev);
-}
-
-void tegra_fbdev_restore_mode(struct tegra_fbdev *fbdev)
-{
-	if (fbdev) {
-		drm_modeset_lock_all(fbdev->base.dev);
-		drm_fb_helper_restore_fbdev_mode(&fbdev->base);
-		drm_modeset_unlock_all(fbdev->base.dev);
-	}
+#endif
 }
diff --git a/drivers/gpu/drm/tegra/gem.c b/drivers/gpu/drm/tegra/gem.c
index 28a9cbc07ab9..ef853e558036 100644
--- a/drivers/gpu/drm/tegra/gem.c
+++ b/drivers/gpu/drm/tegra/gem.c
@@ -18,6 +18,7 @@
  * GNU General Public License for more details.
  */
 
+#include <linux/dma-buf.h>
 #include <drm/tegra_drm.h>
 
 #include "gem.h"
@@ -83,7 +84,7 @@ static struct host1x_bo *tegra_bo_get(struct host1x_bo *bo)
 	return bo;
 }
 
-const struct host1x_bo_ops tegra_bo_ops = {
+static const struct host1x_bo_ops tegra_bo_ops = {
 	.get = tegra_bo_get,
 	.put = tegra_bo_put,
 	.pin = tegra_bo_pin,
@@ -145,7 +146,6 @@ err_dma:
 	kfree(bo);
 
 	return ERR_PTR(err);
-
 }
 
 struct tegra_bo *tegra_bo_create_with_handle(struct drm_file *file,
@@ -174,13 +174,87 @@ err:
 	return ERR_PTR(ret);
 }
 
+struct tegra_bo *tegra_bo_import(struct drm_device *drm, struct dma_buf *buf)
+{
+	struct dma_buf_attachment *attach;
+	struct tegra_bo *bo;
+	ssize_t size;
+	int err;
+
+	bo = kzalloc(sizeof(*bo), GFP_KERNEL);
+	if (!bo)
+		return ERR_PTR(-ENOMEM);
+
+	host1x_bo_init(&bo->base, &tegra_bo_ops);
+	size = round_up(buf->size, PAGE_SIZE);
+
+	err = drm_gem_object_init(drm, &bo->gem, size);
+	if (err < 0)
+		goto free;
+
+	err = drm_gem_create_mmap_offset(&bo->gem);
+	if (err < 0)
+		goto release;
+
+	attach = dma_buf_attach(buf, drm->dev);
+	if (IS_ERR(attach)) {
+		err = PTR_ERR(attach);
+		goto free_mmap;
+	}
+
+	get_dma_buf(buf);
+
+	bo->sgt = dma_buf_map_attachment(attach, DMA_TO_DEVICE);
+	if (!bo->sgt) {
+		err = -ENOMEM;
+		goto detach;
+	}
+
+	if (IS_ERR(bo->sgt)) {
+		err = PTR_ERR(bo->sgt);
+		goto detach;
+	}
+
+	if (bo->sgt->nents > 1) {
+		err = -EINVAL;
+		goto detach;
+	}
+
+	bo->paddr = sg_dma_address(bo->sgt->sgl);
+	bo->gem.import_attach = attach;
+
+	return bo;
+
+detach:
+	if (!IS_ERR_OR_NULL(bo->sgt))
+		dma_buf_unmap_attachment(attach, bo->sgt, DMA_TO_DEVICE);
+
+	dma_buf_detach(buf, attach);
+	dma_buf_put(buf);
+free_mmap:
+	drm_gem_free_mmap_offset(&bo->gem);
+release:
+	drm_gem_object_release(&bo->gem);
+free:
+	kfree(bo);
+
+	return ERR_PTR(err);
+}
+
 void tegra_bo_free_object(struct drm_gem_object *gem)
 {
 	struct tegra_bo *bo = to_tegra_bo(gem);
 
+	if (gem->import_attach) {
+		dma_buf_unmap_attachment(gem->import_attach, bo->sgt,
+					 DMA_TO_DEVICE);
+		drm_prime_gem_destroy(gem, NULL);
+	} else {
+		tegra_bo_destroy(gem->dev, bo);
+	}
+
 	drm_gem_free_mmap_offset(gem);
 	drm_gem_object_release(gem);
-	tegra_bo_destroy(gem->dev, bo);
 
 	kfree(bo);
 }
@@ -256,3 +330,106 @@ int tegra_drm_mmap(struct file *file, struct vm_area_struct *vma)
 
 	return ret;
 }
+
+static struct sg_table *
+tegra_gem_prime_map_dma_buf(struct dma_buf_attachment *attach,
+			    enum dma_data_direction dir)
+{
+	struct drm_gem_object *gem = attach->dmabuf->priv;
+	struct tegra_bo *bo = to_tegra_bo(gem);
+	struct sg_table *sgt;
+
+	sgt = kmalloc(sizeof(*sgt), GFP_KERNEL);
+	if (!sgt)
+		return NULL;
+
+	if (sg_alloc_table(sgt, 1, GFP_KERNEL)) {
+		kfree(sgt);
+		return NULL;
+	}
+
+	sg_dma_address(sgt->sgl) = bo->paddr;
+	sg_dma_len(sgt->sgl) = gem->size;
+
+	return sgt;
+}
+
+static void tegra_gem_prime_unmap_dma_buf(struct dma_buf_attachment *attach,
+					  struct sg_table *sgt,
+					  enum dma_data_direction dir)
+{
+	sg_free_table(sgt);
+	kfree(sgt);
+}
+
+static void tegra_gem_prime_release(struct dma_buf *buf)
+{
+	drm_gem_dmabuf_release(buf);
+}
+
+static void *tegra_gem_prime_kmap_atomic(struct dma_buf *buf,
+					 unsigned long page)
+{
+	return NULL;
+}
+
+static void tegra_gem_prime_kunmap_atomic(struct dma_buf *buf,
+					  unsigned long page,
+					  void *addr)
+{
+}
+
+static void *tegra_gem_prime_kmap(struct dma_buf *buf, unsigned long page)
+{
+	return NULL;
+}
+
+static void tegra_gem_prime_kunmap(struct dma_buf *buf, unsigned long page,
+				   void *addr)
+{
+}
+
+static int tegra_gem_prime_mmap(struct dma_buf *buf, struct vm_area_struct *vma)
+{
+	return -EINVAL;
+}
+
+static const struct dma_buf_ops tegra_gem_prime_dmabuf_ops = {
+	.map_dma_buf = tegra_gem_prime_map_dma_buf,
+	.unmap_dma_buf = tegra_gem_prime_unmap_dma_buf,
+	.release = tegra_gem_prime_release,
+	.kmap_atomic = tegra_gem_prime_kmap_atomic,
+	.kunmap_atomic = tegra_gem_prime_kunmap_atomic,
+	.kmap = tegra_gem_prime_kmap,
+	.kunmap = tegra_gem_prime_kunmap,
+	.mmap = tegra_gem_prime_mmap,
+};
+
+struct dma_buf *tegra_gem_prime_export(struct drm_device *drm,
+				       struct drm_gem_object *gem,
+				       int flags)
+{
+	return dma_buf_export(gem, &tegra_gem_prime_dmabuf_ops, gem->size,
+			      flags);
+}
+
+struct drm_gem_object *tegra_gem_prime_import(struct drm_device *drm,
+					      struct dma_buf *buf)
+{
+	struct tegra_bo *bo;
+
+	if (buf->ops == &tegra_gem_prime_dmabuf_ops) {
+		struct drm_gem_object *gem = buf->priv;
+
+		if (gem->dev == drm) {
+			drm_gem_object_reference(gem);
+			return gem;
+		}
+	}
+
+	bo = tegra_bo_import(drm, buf);
+	if (IS_ERR(bo))
+		return ERR_CAST(bo);
+
+	return &bo->gem;
+}
diff --git a/drivers/gpu/drm/tegra/gem.h b/drivers/gpu/drm/tegra/gem.h
index 7674000bf47d..ffd4f792b410 100644
--- a/drivers/gpu/drm/tegra/gem.h
+++ b/drivers/gpu/drm/tegra/gem.h
@@ -31,6 +31,7 @@ struct tegra_bo {
 	struct drm_gem_object gem;
 	struct host1x_bo base;
 	unsigned long flags;
+	struct sg_table *sgt;
 	dma_addr_t paddr;
 	void *vaddr;
 };
@@ -40,8 +41,6 @@ static inline struct tegra_bo *to_tegra_bo(struct drm_gem_object *gem)
 	return container_of(gem, struct tegra_bo, gem);
 }
 
-extern const struct host1x_bo_ops tegra_bo_ops;
-
 struct tegra_bo *tegra_bo_create(struct drm_device *drm, unsigned int size,
 				 unsigned long flags);
 struct tegra_bo *tegra_bo_create_with_handle(struct drm_file *file,
@@ -59,4 +58,10 @@ int tegra_drm_mmap(struct file *file, struct vm_area_struct *vma);
 
 extern const struct vm_operations_struct tegra_bo_vm_ops;
 
+struct dma_buf *tegra_gem_prime_export(struct drm_device *drm,
+				       struct drm_gem_object *gem,
+				       int flags);
+struct drm_gem_object *tegra_gem_prime_import(struct drm_device *drm,
+					      struct dma_buf *buf);
+
 #endif
diff --git a/drivers/gpu/drm/tegra/hdmi.c b/drivers/gpu/drm/tegra/hdmi.c
index 7f6253ea5cb5..6928015d11a4 100644
--- a/drivers/gpu/drm/tegra/hdmi.c
+++ b/drivers/gpu/drm/tegra/hdmi.c
@@ -40,6 +40,7 @@ struct tegra_hdmi {
 	struct host1x_client client;
 	struct tegra_output output;
 	struct device *dev;
+	bool enabled;
 
 	struct regulator *vdd;
 	struct regulator *pll;
@@ -379,7 +380,7 @@ static void tegra_hdmi_setup_audio_fs_tables(struct tegra_hdmi *hdmi)
 
 		if (f > 96000)
 			delta = 2;
-		else if (f > 480000)
+		else if (f > 48000)
 			delta = 6;
 		else
 			delta = 9;
@@ -699,6 +700,9 @@ static int tegra_output_hdmi_enable(struct tegra_output *output)
 	int retries = 1000;
 	int err;
 
+	if (hdmi->enabled)
+		return 0;
+
 	hdmi->dvi = !tegra_output_is_hdmi(output);
 
 	pclk = mode->clock * 1000;
@@ -839,10 +843,6 @@ static int tegra_output_hdmi_enable(struct tegra_output *output)
 	value |= SOR_CSTM_ROTCLK(2);
 	tegra_hdmi_writel(hdmi, value, HDMI_NV_PDISP_SOR_CSTM);
 
-	tegra_dc_writel(dc, DISP_CTRL_MODE_STOP, DC_CMD_DISPLAY_COMMAND);
-	tegra_dc_writel(dc, GENERAL_ACT_REQ << 8, DC_CMD_STATE_CONTROL);
-	tegra_dc_writel(dc, GENERAL_ACT_REQ, DC_CMD_STATE_CONTROL);
-
 	/* start SOR */
 	tegra_hdmi_writel(hdmi,
 			  SOR_PWR_NORMAL_STATE_PU |
@@ -892,31 +892,67 @@ static int tegra_output_hdmi_enable(struct tegra_output *output)
 			  HDMI_NV_PDISP_SOR_STATE1);
 	tegra_hdmi_writel(hdmi, 0, HDMI_NV_PDISP_SOR_STATE0);
 
-	tegra_dc_writel(dc, HDMI_ENABLE, DC_DISP_DISP_WIN_OPTIONS);
-
-	value = PW0_ENABLE | PW1_ENABLE | PW2_ENABLE | PW3_ENABLE |
-		PW4_ENABLE | PM0_ENABLE | PM1_ENABLE;
-	tegra_dc_writel(dc, value, DC_CMD_DISPLAY_POWER_CONTROL);
+	value = tegra_dc_readl(dc, DC_DISP_DISP_WIN_OPTIONS);
+	value |= HDMI_ENABLE;
+	tegra_dc_writel(dc, value, DC_DISP_DISP_WIN_OPTIONS);
 
-	value = DISP_CTRL_MODE_C_DISPLAY;
+	value = tegra_dc_readl(dc, DC_CMD_DISPLAY_COMMAND);
+	value &= ~DISP_CTRL_MODE_MASK;
+	value |= DISP_CTRL_MODE_C_DISPLAY;
 	tegra_dc_writel(dc, value, DC_CMD_DISPLAY_COMMAND);
 
+	value = tegra_dc_readl(dc, DC_CMD_DISPLAY_POWER_CONTROL);
+	value |= PW0_ENABLE | PW1_ENABLE | PW2_ENABLE | PW3_ENABLE |
+		 PW4_ENABLE | PM0_ENABLE | PM1_ENABLE;
+	tegra_dc_writel(dc, value, DC_CMD_DISPLAY_POWER_CONTROL);
+
 	tegra_dc_writel(dc, GENERAL_ACT_REQ << 8, DC_CMD_STATE_CONTROL);
 	tegra_dc_writel(dc, GENERAL_ACT_REQ, DC_CMD_STATE_CONTROL);
 
 	/* TODO: add HDCP support */
 
+	hdmi->enabled = true;
+
 	return 0;
 }
 
 static int tegra_output_hdmi_disable(struct tegra_output *output)
 {
+	struct tegra_dc *dc = to_tegra_dc(output->encoder.crtc);
 	struct tegra_hdmi *hdmi = to_hdmi(output);
+	unsigned long value;
+
+	if (!hdmi->enabled)
+		return 0;
+
+	/*
+	 * The following accesses registers of the display controller, so make
+	 * sure it's only executed when the output is attached to one.
+	 */
+	if (dc) {
+		value = tegra_dc_readl(dc, DC_CMD_DISPLAY_POWER_CONTROL);
+		value &= ~(PW0_ENABLE | PW1_ENABLE | PW2_ENABLE | PW3_ENABLE |
+			   PW4_ENABLE | PM0_ENABLE | PM1_ENABLE);
+		tegra_dc_writel(dc, value, DC_CMD_DISPLAY_POWER_CONTROL);
+
+		value = tegra_dc_readl(dc, DC_CMD_DISPLAY_COMMAND);
+		value &= ~DISP_CTRL_MODE_MASK;
+		tegra_dc_writel(dc, value, DC_CMD_DISPLAY_COMMAND);
+
+		value = tegra_dc_readl(dc, DC_DISP_DISP_WIN_OPTIONS);
+		value &= ~HDMI_ENABLE;
+		tegra_dc_writel(dc, value, DC_DISP_DISP_WIN_OPTIONS);
+
+		tegra_dc_writel(dc, GENERAL_ACT_REQ << 8, DC_CMD_STATE_CONTROL);
+		tegra_dc_writel(dc, GENERAL_ACT_REQ, DC_CMD_STATE_CONTROL);
+	}
 
 	reset_control_assert(hdmi->rst);
 	clk_disable(hdmi->clk);
 	regulator_disable(hdmi->pll);
 
+	hdmi->enabled = false;
+
 	return 0;
 }
 
@@ -960,7 +996,7 @@ static int tegra_output_hdmi_check_mode(struct tegra_output *output,
 	parent = clk_get_parent(hdmi->clk_parent);
 
 	err = clk_round_rate(parent, pclk * 4);
-	if (err < 0)
+	if (err <= 0)
 		*status = MODE_NOCLOCK;
 	else
 		*status = MODE_OK;
@@ -1382,9 +1418,6 @@ static int tegra_hdmi_probe(struct platform_device *pdev)
 		return err;
 
 	regs = platform_get_resource(pdev, IORESOURCE_MEM, 0);
-	if (!regs)
-		return -ENXIO;
-
 	hdmi->regs = devm_ioremap_resource(&pdev->dev, regs);
 	if (IS_ERR(hdmi->regs))
 		return PTR_ERR(hdmi->regs);
diff --git a/drivers/gpu/drm/tegra/mipi-phy.c b/drivers/gpu/drm/tegra/mipi-phy.c
new file mode 100644
index 000000000000..e2c4aedaee78
--- /dev/null
+++ b/drivers/gpu/drm/tegra/mipi-phy.c
@@ -0,0 +1,138 @@
+/*
+ * Copyright (C) 2013 NVIDIA Corporation
+ *
+ * Permission to use, copy, modify, distribute, and sell this software and its
+ * documentation for any purpose is hereby granted without fee, provided that
+ * the above copyright notice appear in all copies and that both that copyright
+ * notice and this permission notice appear in supporting documentation, and
+ * that the name of the copyright holders not be used in advertising or
+ * publicity pertaining to distribution of the software without specific,
+ * written prior permission.  The copyright holders make no representations
+ * about the suitability of this software for any purpose.  It is provided "as
+ * is" without express or implied warranty.
+ *
+ * THE COPYRIGHT HOLDERS DISCLAIM ALL WARRANTIES WITH REGARD TO THIS SOFTWARE,
+ * INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN NO
+ * EVENT SHALL THE COPYRIGHT HOLDERS BE LIABLE FOR ANY SPECIAL, INDIRECT OR
+ * CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE,
+ * DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER
+ * TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE
+ * OF THIS SOFTWARE.
+ */
+
+#include <linux/errno.h>
+#include <linux/kernel.h>
+
+#include "mipi-phy.h"
+
+/*
+ * Default D-PHY timings based on MIPI D-PHY specification. Derived from
+ * the valid ranges specified in Section 5.9 of the D-PHY specification
+ * with minor adjustments.
+ */
+int mipi_dphy_timing_get_default(struct mipi_dphy_timing *timing,
+				 unsigned long period)
+{
+	timing->clkmiss = 0;
+	timing->clkpost = 70 + 52 * period;
+	timing->clkpre = 8;
+	timing->clkprepare = 65;
+	timing->clksettle = 95;
+	timing->clktermen = 0;
+	timing->clktrail = 80;
+	timing->clkzero = 260;
+	timing->dtermen = 0;
+	timing->eot = 0;
+	timing->hsexit = 120;
+	timing->hsprepare = 65 + 5 * period;
+	timing->hszero = 145 + 5 * period;
+	timing->hssettle = 85 + 6 * period;
+	timing->hsskip = 40;
+	timing->hstrail = max(8 * period, 60 + 4 * period);
+	timing->init = 100000;
+	timing->lpx = 60;
+	timing->taget = 5 * timing->lpx;
+	timing->tago = 4 * timing->lpx;
+	timing->tasure = 2 * timing->lpx;
+	timing->wakeup = 1000000;
+
+	return 0;
+}
+
+/*
+ * Validate D-PHY timing according to MIPI Alliance Specification for D-PHY,
+ * Section 5.9 "Global Operation Timing Parameters".
+ */
+int mipi_dphy_timing_validate(struct mipi_dphy_timing *timing,
+			      unsigned long period)
+{
+	if (timing->clkmiss > 60)
+		return -EINVAL;
+
+	if (timing->clkpost < (60 + 52 * period))
+		return -EINVAL;
+
+	if (timing->clkpre < 8)
+		return -EINVAL;
+
+	if (timing->clkprepare < 38 || timing->clkprepare > 95)
+		return -EINVAL;
+
+	if (timing->clksettle < 95 || timing->clksettle > 300)
+		return -EINVAL;
+
+	if (timing->clktermen > 38)
+		return -EINVAL;
+
+	if (timing->clktrail < 60)
+		return -EINVAL;
+
+	if (timing->clkprepare + timing->clkzero < 300)
+		return -EINVAL;
+
+	if (timing->dtermen > 35 + 4 * period)
+		return -EINVAL;
+
+	if (timing->eot > 105 + 12 * period)
+		return -EINVAL;
+
+	if (timing->hsexit < 100)
+		return -EINVAL;
+
+	if (timing->hsprepare < 40 + 4 * period ||
+	    timing->hsprepare > 85 + 6 * period)
+		return -EINVAL;
+
+	if (timing->hsprepare + timing->hszero < 145 + 10 * period)
+		return -EINVAL;
+
+	if ((timing->hssettle < 85 + 6 * period) ||
+	    (timing->hssettle > 145 + 10 * period))
+		return -EINVAL;
+
+	if (timing->hsskip < 40 || timing->hsskip > 55 + 4 * period)
+		return -EINVAL;
+
+	if (timing->hstrail < max(8 * period, 60 + 4 * period))
+		return -EINVAL;
+
+	if (timing->init < 100000)
+		return -EINVAL;
+
+	if (timing->lpx < 50)
+		return -EINVAL;
+
+	if (timing->taget != 5 * timing->lpx)
+		return -EINVAL;
+
+	if (timing->tago != 4 * timing->lpx)
+		return -EINVAL;
+
+	if (timing->tasure < timing->lpx || timing->tasure > 2 * timing->lpx)
+		return -EINVAL;
+
+	if (timing->wakeup < 1000000)
+		return -EINVAL;
+
+	return 0;
+}
diff --git a/drivers/gpu/drm/tegra/mipi-phy.h b/drivers/gpu/drm/tegra/mipi-phy.h
new file mode 100644
index 000000000000..d3591694432d
--- /dev/null
+++ b/drivers/gpu/drm/tegra/mipi-phy.h
@@ -0,0 +1,65 @@
+/*
+ * Copyright (C) 2013 NVIDIA Corporation
+ *
+ * Permission to use, copy, modify, distribute, and sell this software and its
+ * documentation for any purpose is hereby granted without fee, provided that
+ * the above copyright notice appear in all copies and that both that copyright
+ * notice and this permission notice appear in supporting documentation, and
+ * that the name of the copyright holders not be used in advertising or
+ * publicity pertaining to distribution of the software without specific,
+ * written prior permission.  The copyright holders make no representations
+ * about the suitability of this software for any purpose.  It is provided "as
+ * is" without express or implied warranty.
+ *
+ * THE COPYRIGHT HOLDERS DISCLAIM ALL WARRANTIES WITH REGARD TO THIS SOFTWARE,
+ * INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN NO
+ * EVENT SHALL THE COPYRIGHT HOLDERS BE LIABLE FOR ANY SPECIAL, INDIRECT OR
+ * CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE,
+ * DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER
+ * TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE
+ * OF THIS SOFTWARE.
+ */
+
+#ifndef DRM_TEGRA_MIPI_PHY_H
+#define DRM_TEGRA_MIPI_PHY_H
+
+/*
+ * D-PHY timing parameters
+ *
+ * A detailed description of these parameters can be found in the  MIPI
+ * Alliance Specification for D-PHY, Section 5.9 "Global Operation Timing
+ * Parameters".
+ *
+ * All parameters are specified in nanoseconds.
+ */
+struct mipi_dphy_timing {
+	unsigned int clkmiss;
+	unsigned int clkpost;
+	unsigned int clkpre;
+	unsigned int clkprepare;
+	unsigned int clksettle;
+	unsigned int clktermen;
+	unsigned int clktrail;
+	unsigned int clkzero;
+	unsigned int dtermen;
+	unsigned int eot;
+	unsigned int hsexit;
+	unsigned int hsprepare;
+	unsigned int hszero;
+	unsigned int hssettle;
+	unsigned int hsskip;
+	unsigned int hstrail;
+	unsigned int init;
+	unsigned int lpx;
+	unsigned int taget;
+	unsigned int tago;
+	unsigned int tasure;
+	unsigned int wakeup;
+};
+
+int mipi_dphy_timing_get_default(struct mipi_dphy_timing *timing,
+				 unsigned long period);
+int mipi_dphy_timing_validate(struct mipi_dphy_timing *timing,
+			      unsigned long period);
+
+#endif
diff --git a/drivers/gpu/drm/tegra/output.c b/drivers/gpu/drm/tegra/output.c
index 2cb0065e0578..57cecbd18ca8 100644
--- a/drivers/gpu/drm/tegra/output.c
+++ b/drivers/gpu/drm/tegra/output.c
@@ -9,6 +9,7 @@
 
 #include <linux/of_gpio.h>
 
+#include <drm/drm_panel.h>
 #include "drm.h"
 
 static int tegra_connector_get_modes(struct drm_connector *connector)
@@ -17,6 +18,16 @@ static int tegra_connector_get_modes(struct drm_connector *connector)
 	struct edid *edid = NULL;
 	int err = 0;
 
+	/*
+	 * If the panel provides one or more modes, use them exclusively and
+	 * ignore any other means of obtaining a mode.
+	 */
+	if (output->panel) {
+		err = output->panel->funcs->get_modes(output->panel);
+		if (err > 0)
+			return err;
+	}
+
 	if (output->edid)
 		edid = kmemdup(output->edid, sizeof(*edid), GFP_KERNEL);
 	else if (output->ddc)
@@ -72,6 +83,11 @@ tegra_connector_detect(struct drm_connector *connector, bool force)
 		else
 			status = connector_status_connected;
 	} else {
+		if (!output->panel)
+			status = connector_status_disconnected;
+		else
+			status = connector_status_connected;
+
 		if (connector->connector_type == DRM_MODE_CONNECTOR_LVDS)
 			status = connector_status_connected;
 	}
@@ -115,6 +131,16 @@ static const struct drm_encoder_funcs encoder_funcs = {
 
 static void tegra_encoder_dpms(struct drm_encoder *encoder, int mode)
 {
+	struct tegra_output *output = encoder_to_output(encoder);
+	struct drm_panel *panel = output->panel;
+
+	if (mode != DRM_MODE_DPMS_ON) {
+		drm_panel_disable(panel);
+		tegra_output_disable(output);
+	} else {
+		tegra_output_enable(output);
+		drm_panel_enable(panel);
+	}
 }
 
 static bool tegra_encoder_mode_fixup(struct drm_encoder *encoder,
@@ -163,14 +189,22 @@ static irqreturn_t hpd_irq(int irq, void *data)
 
 int tegra_output_probe(struct tegra_output *output)
 {
+	struct device_node *ddc, *panel;
 	enum of_gpio_flags flags;
-	struct device_node *ddc;
-	size_t size;
-	int err;
+	int err, size;
 
 	if (!output->of_node)
 		output->of_node = output->dev->of_node;
 
+	panel = of_parse_phandle(output->of_node, "nvidia,panel", 0);
+	if (panel) {
+		output->panel = of_drm_find_panel(panel);
+		if (!output->panel)
+			return -EPROBE_DEFER;
+
+		of_node_put(panel);
+	}
+
 	output->edid = of_get_property(output->of_node, "nvidia,edid", &size);
 
 	ddc = of_parse_phandle(output->of_node, "nvidia,ddc-i2c-bus", 0);
@@ -185,9 +219,6 @@ int tegra_output_probe(struct tegra_output *output)
 		of_node_put(ddc);
 	}
 
-	if (!output->edid && !output->ddc)
-		return -ENODEV;
-
 	output->hpd_gpio = of_get_named_gpio_flags(output->of_node,
 						   "nvidia,hpd-gpio", 0,
 						   &flags);
@@ -256,6 +287,11 @@ int tegra_output_init(struct drm_device *drm, struct tegra_output *output)
 		encoder = DRM_MODE_ENCODER_TMDS;
 		break;
 
+	case TEGRA_OUTPUT_DSI:
+		connector = DRM_MODE_CONNECTOR_DSI;
+		encoder = DRM_MODE_ENCODER_DSI;
+		break;
+
 	default:
 		connector = DRM_MODE_CONNECTOR_Unknown;
 		encoder = DRM_MODE_ENCODER_NONE;
@@ -267,6 +303,9 @@ int tegra_output_init(struct drm_device *drm, struct tegra_output *output)
 	drm_connector_helper_add(&output->connector, &connector_helper_funcs);
 	output->connector.dpms = DRM_MODE_DPMS_OFF;
 
+	if (output->panel)
+		drm_panel_attach(output->panel, &output->connector);
+
 	drm_encoder_init(drm, &output->encoder, &encoder_funcs, encoder);
 	drm_encoder_helper_add(&output->encoder, &encoder_helper_funcs);
 
diff --git a/drivers/gpu/drm/tegra/rgb.c b/drivers/gpu/drm/tegra/rgb.c
index 3b29018913a5..338f7f6561d7 100644
--- a/drivers/gpu/drm/tegra/rgb.c
+++ b/drivers/gpu/drm/tegra/rgb.c
@@ -87,15 +87,60 @@ static void tegra_dc_write_regs(struct tegra_dc *dc,
 static int tegra_output_rgb_enable(struct tegra_output *output)
 {
 	struct tegra_rgb *rgb = to_rgb(output);
+	unsigned long value;
 
 	tegra_dc_write_regs(rgb->dc, rgb_enable, ARRAY_SIZE(rgb_enable));
 
+	value = DE_SELECT_ACTIVE | DE_CONTROL_NORMAL;
+	tegra_dc_writel(rgb->dc, value, DC_DISP_DATA_ENABLE_OPTIONS);
+
+	/* XXX: parameterize? */
+	value = tegra_dc_readl(rgb->dc, DC_COM_PIN_OUTPUT_POLARITY(1));
+	value &= ~LVS_OUTPUT_POLARITY_LOW;
+	value &= ~LHS_OUTPUT_POLARITY_LOW;
+	tegra_dc_writel(rgb->dc, value, DC_COM_PIN_OUTPUT_POLARITY(1));
+
+	/* XXX: parameterize? */
+	value = DISP_DATA_FORMAT_DF1P1C | DISP_ALIGNMENT_MSB |
+		DISP_ORDER_RED_BLUE;
+	tegra_dc_writel(rgb->dc, value, DC_DISP_DISP_INTERFACE_CONTROL);
+
+	/* XXX: parameterize? */
+	value = SC0_H_QUALIFIER_NONE | SC1_H_QUALIFIER_NONE;
+	tegra_dc_writel(rgb->dc, value, DC_DISP_SHIFT_CLOCK_OPTIONS);
+
+	value = tegra_dc_readl(rgb->dc, DC_CMD_DISPLAY_COMMAND);
+	value &= ~DISP_CTRL_MODE_MASK;
+	value |= DISP_CTRL_MODE_C_DISPLAY;
+	tegra_dc_writel(rgb->dc, value, DC_CMD_DISPLAY_COMMAND);
+
+	value = tegra_dc_readl(rgb->dc, DC_CMD_DISPLAY_POWER_CONTROL);
+	value |= PW0_ENABLE | PW1_ENABLE | PW2_ENABLE | PW3_ENABLE |
+		 PW4_ENABLE | PM0_ENABLE | PM1_ENABLE;
+	tegra_dc_writel(rgb->dc, value, DC_CMD_DISPLAY_POWER_CONTROL);
+
+	tegra_dc_writel(rgb->dc, GENERAL_ACT_REQ << 8, DC_CMD_STATE_CONTROL);
+	tegra_dc_writel(rgb->dc, GENERAL_ACT_REQ, DC_CMD_STATE_CONTROL);
+
 	return 0;
 }
 
 static int tegra_output_rgb_disable(struct tegra_output *output)
 {
 	struct tegra_rgb *rgb = to_rgb(output);
+	unsigned long value;
+
+	value = tegra_dc_readl(rgb->dc, DC_CMD_DISPLAY_POWER_CONTROL);
+	value &= ~(PW0_ENABLE | PW1_ENABLE | PW2_ENABLE | PW3_ENABLE |
+		   PW4_ENABLE | PM0_ENABLE | PM1_ENABLE);
+	tegra_dc_writel(rgb->dc, value, DC_CMD_DISPLAY_POWER_CONTROL);
+
+	value = tegra_dc_readl(rgb->dc, DC_CMD_DISPLAY_COMMAND);
+	value &= ~DISP_CTRL_MODE_MASK;
+	tegra_dc_writel(rgb->dc, value, DC_CMD_DISPLAY_COMMAND);
+
+	tegra_dc_writel(rgb->dc, GENERAL_ACT_REQ << 8, DC_CMD_STATE_CONTROL);
+	tegra_dc_writel(rgb->dc, GENERAL_ACT_REQ, DC_CMD_STATE_CONTROL);
 
 	tegra_dc_write_regs(rgb->dc, rgb_disable, ARRAY_SIZE(rgb_disable));
 
@@ -213,7 +258,7 @@ int tegra_dc_rgb_init(struct drm_device *drm, struct tegra_dc *dc)
 	 * RGB outputs are an exception, so we make sure they can be attached
 	 * to only their parent display controller.
 	 */
-	rgb->output.encoder.possible_crtcs = 1 << dc->pipe;
+	rgb->output.encoder.possible_crtcs = drm_crtc_mask(&dc->base);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/tilcdc/tilcdc_drv.c b/drivers/gpu/drm/tilcdc/tilcdc_drv.c
index 116da199b942..171a8203892c 100644
--- a/drivers/gpu/drm/tilcdc/tilcdc_drv.c
+++ b/drivers/gpu/drm/tilcdc/tilcdc_drv.c
@@ -311,7 +311,7 @@ static void tilcdc_lastclose(struct drm_device *dev)
 	drm_fbdev_cma_restore_mode(priv->fbdev);
 }
 
-static irqreturn_t tilcdc_irq(DRM_IRQ_ARGS)
+static irqreturn_t tilcdc_irq(int irq, void *arg)
 {
 	struct drm_device *dev = arg;
 	struct tilcdc_drm_private *priv = dev->dev_private;
@@ -444,7 +444,7 @@ static int tilcdc_mm_show(struct seq_file *m, void *arg)
 {
 	struct drm_info_node *node = (struct drm_info_node *) m->private;
 	struct drm_device *dev = node->minor->dev;
-	return drm_mm_dump_table(m, dev->mm_private);
+	return drm_mm_dump_table(m, &dev->vma_offset_manager->vm_addr_space_mm);
 }
 
 static struct drm_info_list tilcdc_debugfs_list[] = {
@@ -594,7 +594,7 @@ static int tilcdc_pdev_probe(struct platform_device *pdev)
 
 static int tilcdc_pdev_remove(struct platform_device *pdev)
 {
-	drm_platform_exit(&tilcdc_driver, pdev);
+	drm_put_dev(platform_get_drvdata(pdev));
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c
index 07e02c4bf5a8..a06651309388 100644
--- a/drivers/gpu/drm/ttm/ttm_bo.c
+++ b/drivers/gpu/drm/ttm/ttm_bo.c
@@ -957,7 +957,7 @@ int ttm_bo_mem_space(struct ttm_buffer_object *bo,
 }
 EXPORT_SYMBOL(ttm_bo_mem_space);
 
-int ttm_bo_move_buffer(struct ttm_buffer_object *bo,
+static int ttm_bo_move_buffer(struct ttm_buffer_object *bo,
 			struct ttm_placement *placement,
 			bool interruptible,
 			bool no_wait_gpu)
diff --git a/drivers/gpu/drm/ttm/ttm_bo_util.c b/drivers/gpu/drm/ttm/ttm_bo_util.c
index 406152152315..1df856f78568 100644
--- a/drivers/gpu/drm/ttm/ttm_bo_util.c
+++ b/drivers/gpu/drm/ttm/ttm_bo_util.c
@@ -187,7 +187,7 @@ void ttm_mem_io_free_vm(struct ttm_buffer_object *bo)
 	}
 }
 
-int ttm_mem_reg_ioremap(struct ttm_bo_device *bdev, struct ttm_mem_reg *mem,
+static int ttm_mem_reg_ioremap(struct ttm_bo_device *bdev, struct ttm_mem_reg *mem,
 			void **virtual)
 {
 	struct ttm_mem_type_manager *man = &bdev->man[mem->mem_type];
@@ -219,7 +219,7 @@ int ttm_mem_reg_ioremap(struct ttm_bo_device *bdev, struct ttm_mem_reg *mem,
 	return 0;
 }
 
-void ttm_mem_reg_iounmap(struct ttm_bo_device *bdev, struct ttm_mem_reg *mem,
+static void ttm_mem_reg_iounmap(struct ttm_bo_device *bdev, struct ttm_mem_reg *mem,
 			 void *virtual)
 {
 	struct ttm_mem_type_manager *man;
@@ -594,7 +594,7 @@ int ttm_bo_kmap(struct ttm_buffer_object *bo,
 	if (start_page > bo->num_pages)
 		return -EINVAL;
 #if 0
-	if (num_pages > 1 && !DRM_SUSER(DRM_CURPROC))
+	if (num_pages > 1 && !capable(CAP_SYS_ADMIN))
 		return -EPERM;
 #endif
 	(void) ttm_mem_io_lock(man, false);
diff --git a/drivers/gpu/drm/ttm/ttm_bo_vm.c b/drivers/gpu/drm/ttm/ttm_bo_vm.c
index 6440eeac22d2..801231c9ae48 100644
--- a/drivers/gpu/drm/ttm/ttm_bo_vm.c
+++ b/drivers/gpu/drm/ttm/ttm_bo_vm.c
@@ -132,6 +132,15 @@ static int ttm_bo_vm_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 		return VM_FAULT_NOPAGE;
 	}
 
+	/*
+	 * Refuse to fault imported pages. This should be handled
+	 * (if at all) by redirecting mmap to the exporter.
+	 */
+	if (bo->ttm && (bo->ttm->page_flags & TTM_PAGE_FLAG_SG)) {
+		retval = VM_FAULT_SIGBUS;
+		goto out_unlock;
+	}
+
 	if (bdev->driver->fault_reserve_notify) {
 		ret = bdev->driver->fault_reserve_notify(bo);
 		switch (ret) {
@@ -217,10 +226,17 @@ static int ttm_bo_vm_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 			} else if (unlikely(!page)) {
 				break;
 			}
+			page->mapping = vma->vm_file->f_mapping;
+			page->index = drm_vma_node_start(&bo->vma_node) +
+				page_offset;
 			pfn = page_to_pfn(page);
 		}
 
-		ret = vm_insert_mixed(&cvma, address, pfn);
+		if (vma->vm_flags & VM_MIXEDMAP)
+			ret = vm_insert_mixed(&cvma, address, pfn);
+		else
+			ret = vm_insert_pfn(&cvma, address, pfn);
+
 		/*
 		 * Somebody beat us to this PTE or prefaulting to
 		 * an already populated PTE, or prefaulting error.
@@ -250,6 +266,8 @@ static void ttm_bo_vm_open(struct vm_area_struct *vma)
 	struct ttm_buffer_object *bo =
 	    (struct ttm_buffer_object *)vma->vm_private_data;
 
+	WARN_ON(bo->bdev->dev_mapping != vma->vm_file->f_mapping);
+
 	(void)ttm_bo_reference(bo);
 }
 
@@ -319,7 +337,14 @@ int ttm_bo_mmap(struct file *filp, struct vm_area_struct *vma,
 	 */
 
 	vma->vm_private_data = bo;
-	vma->vm_flags |= VM_IO | VM_MIXEDMAP | VM_DONTEXPAND | VM_DONTDUMP;
+
+	/*
+	 * PFNMAP is faster than MIXEDMAP due to reduced page
+	 * administration. So use MIXEDMAP only if private VMA, where
+	 * we need to support COW.
+	 */
+	vma->vm_flags |= (vma->vm_flags & VM_SHARED) ? VM_PFNMAP : VM_MIXEDMAP;
+	vma->vm_flags |= VM_IO | VM_DONTEXPAND | VM_DONTDUMP;
 	return 0;
 out_unref:
 	ttm_bo_unref(&bo);
@@ -334,7 +359,8 @@ int ttm_fbdev_mmap(struct vm_area_struct *vma, struct ttm_buffer_object *bo)
 
 	vma->vm_ops = &ttm_bo_vm_ops;
 	vma->vm_private_data = ttm_bo_reference(bo);
-	vma->vm_flags |= VM_IO | VM_MIXEDMAP | VM_DONTEXPAND;
+	vma->vm_flags |= (vma->vm_flags & VM_SHARED) ? VM_PFNMAP : VM_MIXEDMAP;
+	vma->vm_flags |= VM_IO | VM_DONTEXPAND;
 	return 0;
 }
 EXPORT_SYMBOL(ttm_fbdev_mmap);
diff --git a/drivers/gpu/drm/ttm/ttm_lock.c b/drivers/gpu/drm/ttm/ttm_lock.c
index 3daa9a3930b8..6a954544727f 100644
--- a/drivers/gpu/drm/ttm/ttm_lock.c
+++ b/drivers/gpu/drm/ttm/ttm_lock.c
@@ -186,14 +186,6 @@ int ttm_write_lock(struct ttm_lock *lock, bool interruptible)
 }
 EXPORT_SYMBOL(ttm_write_lock);
 
-void ttm_write_lock_downgrade(struct ttm_lock *lock)
-{
-	spin_lock(&lock->lock);
-	lock->rw = 1;
-	wake_up_all(&lock->queue);
-	spin_unlock(&lock->lock);
-}
-
 static int __ttm_vt_unlock(struct ttm_lock *lock)
 {
 	int ret = 0;
diff --git a/drivers/gpu/drm/ttm/ttm_object.c b/drivers/gpu/drm/ttm/ttm_object.c
index 6fe7b92a82d1..37079859afc8 100644
--- a/drivers/gpu/drm/ttm/ttm_object.c
+++ b/drivers/gpu/drm/ttm/ttm_object.c
@@ -68,7 +68,7 @@
 
 struct ttm_object_file {
 	struct ttm_object_device *tdev;
-	rwlock_t lock;
+	spinlock_t lock;
 	struct list_head ref_list;
 	struct drm_open_hash ref_hash[TTM_REF_NUM];
 	struct kref refcount;
@@ -118,6 +118,7 @@ struct ttm_object_device {
  */
 
 struct ttm_ref_object {
+	struct rcu_head rcu_head;
 	struct drm_hash_item hash;
 	struct list_head head;
 	struct kref kref;
@@ -210,10 +211,9 @@ static void ttm_release_base(struct kref *kref)
 	 * call_rcu() or ttm_base_object_kfree().
 	 */
 
-	if (base->refcount_release) {
-		ttm_object_file_unref(&base->tfile);
+	ttm_object_file_unref(&base->tfile);
+	if (base->refcount_release)
 		base->refcount_release(&base);
-	}
 }
 
 void ttm_base_object_unref(struct ttm_base_object **p_base)
@@ -229,32 +229,46 @@ EXPORT_SYMBOL(ttm_base_object_unref);
 struct ttm_base_object *ttm_base_object_lookup(struct ttm_object_file *tfile,
 					       uint32_t key)
 {
-	struct ttm_object_device *tdev = tfile->tdev;
-	struct ttm_base_object *uninitialized_var(base);
+	struct ttm_base_object *base = NULL;
 	struct drm_hash_item *hash;
+	struct drm_open_hash *ht = &tfile->ref_hash[TTM_REF_USAGE];
 	int ret;
 
 	rcu_read_lock();
-	ret = drm_ht_find_item_rcu(&tdev->object_hash, key, &hash);
+	ret = drm_ht_find_item_rcu(ht, key, &hash);
 
 	if (likely(ret == 0)) {
-		base = drm_hash_entry(hash, struct ttm_base_object, hash);
-		ret = kref_get_unless_zero(&base->refcount) ? 0 : -EINVAL;
+		base = drm_hash_entry(hash, struct ttm_ref_object, hash)->obj;
+		if (!kref_get_unless_zero(&base->refcount))
+			base = NULL;
 	}
 	rcu_read_unlock();
 
-	if (unlikely(ret != 0))
-		return NULL;
+	return base;
+}
+EXPORT_SYMBOL(ttm_base_object_lookup);
 
-	if (tfile != base->tfile && !base->shareable) {
-		pr_err("Attempted access of non-shareable object\n");
-		ttm_base_object_unref(&base);
-		return NULL;
+struct ttm_base_object *
+ttm_base_object_lookup_for_ref(struct ttm_object_device *tdev, uint32_t key)
+{
+	struct ttm_base_object *base = NULL;
+	struct drm_hash_item *hash;
+	struct drm_open_hash *ht = &tdev->object_hash;
+	int ret;
+
+	rcu_read_lock();
+	ret = drm_ht_find_item_rcu(ht, key, &hash);
+
+	if (likely(ret == 0)) {
+		base = drm_hash_entry(hash, struct ttm_base_object, hash);
+		if (!kref_get_unless_zero(&base->refcount))
+			base = NULL;
 	}
+	rcu_read_unlock();
 
 	return base;
 }
-EXPORT_SYMBOL(ttm_base_object_lookup);
+EXPORT_SYMBOL(ttm_base_object_lookup_for_ref);
 
 int ttm_ref_object_add(struct ttm_object_file *tfile,
 		       struct ttm_base_object *base,
@@ -266,21 +280,25 @@ int ttm_ref_object_add(struct ttm_object_file *tfile,
 	struct ttm_mem_global *mem_glob = tfile->tdev->mem_glob;
 	int ret = -EINVAL;
 
+	if (base->tfile != tfile && !base->shareable)
+		return -EPERM;
+
 	if (existed != NULL)
 		*existed = true;
 
 	while (ret == -EINVAL) {
-		read_lock(&tfile->lock);
-		ret = drm_ht_find_item(ht, base->hash.key, &hash);
+		rcu_read_lock();
+		ret = drm_ht_find_item_rcu(ht, base->hash.key, &hash);
 
 		if (ret == 0) {
 			ref = drm_hash_entry(hash, struct ttm_ref_object, hash);
-			kref_get(&ref->kref);
-			read_unlock(&tfile->lock);
-			break;
+			if (!kref_get_unless_zero(&ref->kref)) {
+				rcu_read_unlock();
+				break;
+			}
 		}
 
-		read_unlock(&tfile->lock);
+		rcu_read_unlock();
 		ret = ttm_mem_global_alloc(mem_glob, sizeof(*ref),
 					   false, false);
 		if (unlikely(ret != 0))
@@ -297,19 +315,19 @@ int ttm_ref_object_add(struct ttm_object_file *tfile,
 		ref->ref_type = ref_type;
 		kref_init(&ref->kref);
 
-		write_lock(&tfile->lock);
-		ret = drm_ht_insert_item(ht, &ref->hash);
+		spin_lock(&tfile->lock);
+		ret = drm_ht_insert_item_rcu(ht, &ref->hash);
 
 		if (likely(ret == 0)) {
 			list_add_tail(&ref->head, &tfile->ref_list);
 			kref_get(&base->refcount);
-			write_unlock(&tfile->lock);
+			spin_unlock(&tfile->lock);
 			if (existed != NULL)
 				*existed = false;
 			break;
 		}
 
-		write_unlock(&tfile->lock);
+		spin_unlock(&tfile->lock);
 		BUG_ON(ret != -EINVAL);
 
 		ttm_mem_global_free(mem_glob, sizeof(*ref));
@@ -330,17 +348,17 @@ static void ttm_ref_object_release(struct kref *kref)
 	struct ttm_mem_global *mem_glob = tfile->tdev->mem_glob;
 
 	ht = &tfile->ref_hash[ref->ref_type];
-	(void)drm_ht_remove_item(ht, &ref->hash);
+	(void)drm_ht_remove_item_rcu(ht, &ref->hash);
 	list_del(&ref->head);
-	write_unlock(&tfile->lock);
+	spin_unlock(&tfile->lock);
 
 	if (ref->ref_type != TTM_REF_USAGE && base->ref_obj_release)
 		base->ref_obj_release(base, ref->ref_type);
 
 	ttm_base_object_unref(&ref->obj);
 	ttm_mem_global_free(mem_glob, sizeof(*ref));
-	kfree(ref);
-	write_lock(&tfile->lock);
+	kfree_rcu(ref, rcu_head);
+	spin_lock(&tfile->lock);
 }
 
 int ttm_ref_object_base_unref(struct ttm_object_file *tfile,
@@ -351,15 +369,15 @@ int ttm_ref_object_base_unref(struct ttm_object_file *tfile,
 	struct drm_hash_item *hash;
 	int ret;
 
-	write_lock(&tfile->lock);
+	spin_lock(&tfile->lock);
 	ret = drm_ht_find_item(ht, key, &hash);
 	if (unlikely(ret != 0)) {
-		write_unlock(&tfile->lock);
+		spin_unlock(&tfile->lock);
 		return -EINVAL;
 	}
 	ref = drm_hash_entry(hash, struct ttm_ref_object, hash);
 	kref_put(&ref->kref, ttm_ref_object_release);
-	write_unlock(&tfile->lock);
+	spin_unlock(&tfile->lock);
 	return 0;
 }
 EXPORT_SYMBOL(ttm_ref_object_base_unref);
@@ -372,7 +390,7 @@ void ttm_object_file_release(struct ttm_object_file **p_tfile)
 	struct ttm_object_file *tfile = *p_tfile;
 
 	*p_tfile = NULL;
-	write_lock(&tfile->lock);
+	spin_lock(&tfile->lock);
 
 	/*
 	 * Since we release the lock within the loop, we have to
@@ -388,7 +406,7 @@ void ttm_object_file_release(struct ttm_object_file **p_tfile)
 	for (i = 0; i < TTM_REF_NUM; ++i)
 		drm_ht_remove(&tfile->ref_hash[i]);
 
-	write_unlock(&tfile->lock);
+	spin_unlock(&tfile->lock);
 	ttm_object_file_unref(&tfile);
 }
 EXPORT_SYMBOL(ttm_object_file_release);
@@ -404,7 +422,7 @@ struct ttm_object_file *ttm_object_file_init(struct ttm_object_device *tdev,
 	if (unlikely(tfile == NULL))
 		return NULL;
 
-	rwlock_init(&tfile->lock);
+	spin_lock_init(&tfile->lock);
 	tfile->tdev = tdev;
 	kref_init(&tfile->refcount);
 	INIT_LIST_HEAD(&tfile->ref_list);
diff --git a/drivers/gpu/drm/ttm/ttm_tt.c b/drivers/gpu/drm/ttm/ttm_tt.c
index 210d50365162..9af99084b344 100644
--- a/drivers/gpu/drm/ttm/ttm_tt.c
+++ b/drivers/gpu/drm/ttm/ttm_tt.c
@@ -170,9 +170,8 @@ void ttm_tt_destroy(struct ttm_tt *ttm)
 		ttm_tt_unbind(ttm);
 	}
 
-	if (ttm->state == tt_unbound) {
-		ttm->bdev->driver->ttm_tt_unpopulate(ttm);
-	}
+	if (ttm->state == tt_unbound)
+		ttm_tt_unpopulate(ttm);
 
 	if (!(ttm->page_flags & TTM_PAGE_FLAG_PERSISTENT_SWAP) &&
 	    ttm->swap_storage)
@@ -362,7 +361,7 @@ int ttm_tt_swapout(struct ttm_tt *ttm, struct file *persistent_swap_storage)
 		page_cache_release(to_page);
 	}
 
-	ttm->bdev->driver->ttm_tt_unpopulate(ttm);
+	ttm_tt_unpopulate(ttm);
 	ttm->swap_storage = swap_storage;
 	ttm->page_flags |= TTM_PAGE_FLAG_SWAPPED;
 	if (persistent_swap_storage)
@@ -375,3 +374,23 @@ out_err:
 
 	return ret;
 }
+
+static void ttm_tt_clear_mapping(struct ttm_tt *ttm)
+{
+	pgoff_t i;
+	struct page **page = ttm->pages;
+
+	for (i = 0; i < ttm->num_pages; ++i) {
+		(*page)->mapping = NULL;
+		(*page++)->index = 0;
+	}
+}
+
+void ttm_tt_unpopulate(struct ttm_tt *ttm)
+{
+	if (ttm->state == tt_unpopulated)
+		return;
+
+	ttm_tt_clear_mapping(ttm);
+	ttm->bdev->driver->ttm_tt_unpopulate(ttm);
+}
diff --git a/drivers/gpu/drm/udl/udl_fb.c b/drivers/gpu/drm/udl/udl_fb.c
index 97e9d614700f..dbadd49e4c4a 100644
--- a/drivers/gpu/drm/udl/udl_fb.c
+++ b/drivers/gpu/drm/udl/udl_fb.c
@@ -403,15 +403,17 @@ static int udl_user_framebuffer_dirty(struct drm_framebuffer *fb,
 	int i;
 	int ret = 0;
 
+	drm_modeset_lock_all(fb->dev);
+
 	if (!ufb->active_16)
-		return 0;
+		goto unlock;
 
 	if (ufb->obj->base.import_attach) {
 		ret = dma_buf_begin_cpu_access(ufb->obj->base.import_attach->dmabuf,
 					       0, ufb->obj->base.size,
 					       DMA_FROM_DEVICE);
 		if (ret)
-			return ret;
+			goto unlock;
 	}
 
 	for (i = 0; i < num_clips; i++) {
@@ -419,7 +421,7 @@ static int udl_user_framebuffer_dirty(struct drm_framebuffer *fb,
 				  clips[i].x2 - clips[i].x1,
 				  clips[i].y2 - clips[i].y1);
 		if (ret)
-			break;
+			goto unlock;
 	}
 
 	if (ufb->obj->base.import_attach) {
@@ -427,6 +429,10 @@ static int udl_user_framebuffer_dirty(struct drm_framebuffer *fb,
 				       0, ufb->obj->base.size,
 				       DMA_FROM_DEVICE);
 	}
+
+ unlock:
+	drm_modeset_unlock_all(fb->dev);
+
 	return ret;
 }
 
diff --git a/drivers/gpu/drm/via/via_dma.c b/drivers/gpu/drm/via/via_dma.c
index 652f9b43ec9d..a18479c6b6da 100644
--- a/drivers/gpu/drm/via/via_dma.c
+++ b/drivers/gpu/drm/via/via_dma.c
@@ -60,7 +60,7 @@
 	dev_priv->dma_low += 8;					\
 }
 
-#define via_flush_write_combine() DRM_MEMORYBARRIER()
+#define via_flush_write_combine() mb()
 
 #define VIA_OUT_RING_QW(w1, w2)	do {		\
 	*vb++ = (w1);				\
@@ -234,13 +234,13 @@ static int via_dma_init(struct drm_device *dev, void *data, struct drm_file *fil
 
 	switch (init->func) {
 	case VIA_INIT_DMA:
-		if (!DRM_SUSER(DRM_CURPROC))
+		if (!capable(CAP_SYS_ADMIN))
 			retcode = -EPERM;
 		else
 			retcode = via_initialize(dev, dev_priv, init);
 		break;
 	case VIA_CLEANUP_DMA:
-		if (!DRM_SUSER(DRM_CURPROC))
+		if (!capable(CAP_SYS_ADMIN))
 			retcode = -EPERM;
 		else
 			retcode = via_dma_cleanup(dev);
@@ -273,7 +273,7 @@ static int via_dispatch_cmdbuffer(struct drm_device *dev, drm_via_cmdbuffer_t *c
 	if (cmd->size > VIA_PCI_BUF_SIZE)
 		return -ENOMEM;
 
-	if (DRM_COPY_FROM_USER(dev_priv->pci_buf, cmd->buf, cmd->size))
+	if (copy_from_user(dev_priv->pci_buf, cmd->buf, cmd->size))
 		return -EFAULT;
 
 	/*
@@ -346,7 +346,7 @@ static int via_dispatch_pci_cmdbuffer(struct drm_device *dev,
 
 	if (cmd->size > VIA_PCI_BUF_SIZE)
 		return -ENOMEM;
-	if (DRM_COPY_FROM_USER(dev_priv->pci_buf, cmd->buf, cmd->size))
+	if (copy_from_user(dev_priv->pci_buf, cmd->buf, cmd->size))
 		return -EFAULT;
 
 	if ((ret =
@@ -543,7 +543,7 @@ static void via_cmdbuf_start(drm_via_private_t *dev_priv)
 
 	VIA_WRITE(VIA_REG_TRANSPACE, pause_addr_hi);
 	VIA_WRITE(VIA_REG_TRANSPACE, pause_addr_lo);
-	DRM_WRITEMEMORYBARRIER();
+	wmb();
 	VIA_WRITE(VIA_REG_TRANSPACE, command | HC_HAGPCMNT_MASK);
 	VIA_READ(VIA_REG_TRANSPACE);
 
diff --git a/drivers/gpu/drm/via/via_dmablit.c b/drivers/gpu/drm/via/via_dmablit.c
index 8b0f25904e6d..ba33cf679180 100644
--- a/drivers/gpu/drm/via/via_dmablit.c
+++ b/drivers/gpu/drm/via/via_dmablit.c
@@ -217,7 +217,7 @@ via_fire_dmablit(struct drm_device *dev, drm_via_sg_info_t *vsg, int engine)
 	VIA_WRITE(VIA_PCI_DMA_MR0  + engine*0x04, VIA_DMA_MR_CM | VIA_DMA_MR_TDIE);
 	VIA_WRITE(VIA_PCI_DMA_BCR0 + engine*0x10, 0);
 	VIA_WRITE(VIA_PCI_DMA_DPR0 + engine*0x10, vsg->chain_start);
-	DRM_WRITEMEMORYBARRIER();
+	wmb();
 	VIA_WRITE(VIA_PCI_DMA_CSR0 + engine*0x04, VIA_DMA_CSR_DE | VIA_DMA_CSR_TS);
 	VIA_READ(VIA_PCI_DMA_CSR0 + engine*0x04);
 }
@@ -338,7 +338,7 @@ via_dmablit_handler(struct drm_device *dev, int engine, int from_irq)
 
 		blitq->blits[cur]->aborted = blitq->aborting;
 		blitq->done_blit_handle++;
-		DRM_WAKEUP(blitq->blit_queue + cur);
+		wake_up(blitq->blit_queue + cur);
 
 		cur++;
 		if (cur >= VIA_NUM_BLIT_SLOTS)
@@ -363,7 +363,7 @@ via_dmablit_handler(struct drm_device *dev, int engine, int from_irq)
 
 		via_abort_dmablit(dev, engine);
 		blitq->aborting = 1;
-		blitq->end = jiffies + DRM_HZ;
+		blitq->end = jiffies + HZ;
 	}
 
 	if (!blitq->is_active) {
@@ -372,7 +372,7 @@ via_dmablit_handler(struct drm_device *dev, int engine, int from_irq)
 			blitq->is_active = 1;
 			blitq->cur = cur;
 			blitq->num_outstanding--;
-			blitq->end = jiffies + DRM_HZ;
+			blitq->end = jiffies + HZ;
 			if (!timer_pending(&blitq->poll_timer))
 				mod_timer(&blitq->poll_timer, jiffies + 1);
 		} else {
@@ -436,7 +436,7 @@ via_dmablit_sync(struct drm_device *dev, uint32_t handle, int engine)
 	int ret = 0;
 
 	if (via_dmablit_active(blitq, engine, handle, &queue)) {
-		DRM_WAIT_ON(ret, *queue, 3 * DRM_HZ,
+		DRM_WAIT_ON(ret, *queue, 3 * HZ,
 			    !via_dmablit_active(blitq, engine, handle, NULL));
 	}
 	DRM_DEBUG("DMA blit sync handle 0x%x engine %d returned %d\n",
@@ -521,7 +521,7 @@ via_dmablit_workqueue(struct work_struct *work)
 
 		spin_unlock_irqrestore(&blitq->blit_lock, irqsave);
 
-		DRM_WAKEUP(&blitq->busy_queue);
+		wake_up(&blitq->busy_queue);
 
 		via_free_sg_info(dev->pdev, cur_sg);
 		kfree(cur_sg);
@@ -561,8 +561,8 @@ via_init_dmablit(struct drm_device *dev)
 		blitq->aborting = 0;
 		spin_lock_init(&blitq->blit_lock);
 		for (j = 0; j < VIA_NUM_BLIT_SLOTS; ++j)
-			DRM_INIT_WAITQUEUE(blitq->blit_queue + j);
-		DRM_INIT_WAITQUEUE(&blitq->busy_queue);
+			init_waitqueue_head(blitq->blit_queue + j);
+		init_waitqueue_head(&blitq->busy_queue);
 		INIT_WORK(&blitq->wq, via_dmablit_workqueue);
 		setup_timer(&blitq->poll_timer, via_dmablit_timer,
 				(unsigned long)blitq);
@@ -688,7 +688,7 @@ via_dmablit_grab_slot(drm_via_blitq_t *blitq, int engine)
 	while (blitq->num_free == 0) {
 		spin_unlock_irqrestore(&blitq->blit_lock, irqsave);
 
-		DRM_WAIT_ON(ret, blitq->busy_queue, DRM_HZ, blitq->num_free > 0);
+		DRM_WAIT_ON(ret, blitq->busy_queue, HZ, blitq->num_free > 0);
 		if (ret)
 			return (-EINTR == ret) ? -EAGAIN : ret;
 
@@ -713,7 +713,7 @@ via_dmablit_release_slot(drm_via_blitq_t *blitq)
 	spin_lock_irqsave(&blitq->blit_lock, irqsave);
 	blitq->num_free++;
 	spin_unlock_irqrestore(&blitq->blit_lock, irqsave);
-	DRM_WAKEUP(&blitq->busy_queue);
+	wake_up(&blitq->busy_queue);
 }
 
 /*
diff --git a/drivers/gpu/drm/via/via_drv.c b/drivers/gpu/drm/via/via_drv.c
index 92684a9b7e34..50abc2adfaee 100644
--- a/drivers/gpu/drm/via/via_drv.c
+++ b/drivers/gpu/drm/via/via_drv.c
@@ -46,7 +46,7 @@ static int via_driver_open(struct drm_device *dev, struct drm_file *file)
 	return 0;
 }
 
-void via_driver_postclose(struct drm_device *dev, struct drm_file *file)
+static void via_driver_postclose(struct drm_device *dev, struct drm_file *file)
 {
 	struct via_file_private *file_priv = file->driver_priv;
 
diff --git a/drivers/gpu/drm/via/via_drv.h b/drivers/gpu/drm/via/via_drv.h
index a811ef2b505f..ad0273256beb 100644
--- a/drivers/gpu/drm/via/via_drv.h
+++ b/drivers/gpu/drm/via/via_drv.h
@@ -138,7 +138,7 @@ extern u32 via_get_vblank_counter(struct drm_device *dev, int crtc);
 extern int via_enable_vblank(struct drm_device *dev, int crtc);
 extern void via_disable_vblank(struct drm_device *dev, int crtc);
 
-extern irqreturn_t via_driver_irq_handler(DRM_IRQ_ARGS);
+extern irqreturn_t via_driver_irq_handler(int irq, void *arg);
 extern void via_driver_irq_preinstall(struct drm_device *dev);
 extern int via_driver_irq_postinstall(struct drm_device *dev);
 extern void via_driver_irq_uninstall(struct drm_device *dev);
diff --git a/drivers/gpu/drm/via/via_irq.c b/drivers/gpu/drm/via/via_irq.c
index ac98964297cf..1319433816d3 100644
--- a/drivers/gpu/drm/via/via_irq.c
+++ b/drivers/gpu/drm/via/via_irq.c
@@ -104,7 +104,7 @@ u32 via_get_vblank_counter(struct drm_device *dev, int crtc)
 	return atomic_read(&dev_priv->vbl_received);
 }
 
-irqreturn_t via_driver_irq_handler(DRM_IRQ_ARGS)
+irqreturn_t via_driver_irq_handler(int irq, void *arg)
 {
 	struct drm_device *dev = (struct drm_device *) arg;
 	drm_via_private_t *dev_priv = (drm_via_private_t *) dev->dev_private;
@@ -138,7 +138,7 @@ irqreturn_t via_driver_irq_handler(DRM_IRQ_ARGS)
 	for (i = 0; i < dev_priv->num_irqs; ++i) {
 		if (status & cur_irq->pending_mask) {
 			atomic_inc(&cur_irq->irq_received);
-			DRM_WAKEUP(&cur_irq->irq_queue);
+			wake_up(&cur_irq->irq_queue);
 			handled = 1;
 			if (dev_priv->irq_map[drm_via_irq_dma0_td] == i)
 				via_dmablit_handler(dev, 0, 1);
@@ -239,12 +239,12 @@ via_driver_irq_wait(struct drm_device *dev, unsigned int irq, int force_sequence
 	cur_irq = dev_priv->via_irqs + real_irq;
 
 	if (masks[real_irq][2] && !force_sequence) {
-		DRM_WAIT_ON(ret, cur_irq->irq_queue, 3 * DRM_HZ,
+		DRM_WAIT_ON(ret, cur_irq->irq_queue, 3 * HZ,
 			    ((VIA_READ(masks[irq][2]) & masks[irq][3]) ==
 			     masks[irq][4]));
 		cur_irq_sequence = atomic_read(&cur_irq->irq_received);
 	} else {
-		DRM_WAIT_ON(ret, cur_irq->irq_queue, 3 * DRM_HZ,
+		DRM_WAIT_ON(ret, cur_irq->irq_queue, 3 * HZ,
 			    (((cur_irq_sequence =
 			       atomic_read(&cur_irq->irq_received)) -
 			      *sequence) <= (1 << 23)));
@@ -287,7 +287,7 @@ void via_driver_irq_preinstall(struct drm_device *dev)
 			atomic_set(&cur_irq->irq_received, 0);
 			cur_irq->enable_mask = dev_priv->irq_masks[i][0];
 			cur_irq->pending_mask = dev_priv->irq_masks[i][1];
-			DRM_INIT_WAITQUEUE(&cur_irq->irq_queue);
+			init_waitqueue_head(&cur_irq->irq_queue);
 			dev_priv->irq_enable_mask |= cur_irq->enable_mask;
 			dev_priv->irq_pending_mask |= cur_irq->pending_mask;
 			cur_irq++;
diff --git a/drivers/gpu/drm/via/via_video.c b/drivers/gpu/drm/via/via_video.c
index 6569efa2ff6e..a9ffbad1cfdd 100644
--- a/drivers/gpu/drm/via/via_video.c
+++ b/drivers/gpu/drm/via/via_video.c
@@ -36,7 +36,7 @@ void via_init_futex(drm_via_private_t *dev_priv)
 	DRM_DEBUG("\n");
 
 	for (i = 0; i < VIA_NR_XVMC_LOCKS; ++i) {
-		DRM_INIT_WAITQUEUE(&(dev_priv->decoder_queue[i]));
+		init_waitqueue_head(&(dev_priv->decoder_queue[i]));
 		XVMCLOCKPTR(dev_priv->sarea_priv, i)->lock = 0;
 	}
 }
@@ -58,7 +58,7 @@ void via_release_futex(drm_via_private_t *dev_priv, int context)
 		if ((_DRM_LOCKING_CONTEXT(*lock) == context)) {
 			if (_DRM_LOCK_IS_HELD(*lock)
 			    && (*lock & _DRM_LOCK_CONT)) {
-				DRM_WAKEUP(&(dev_priv->decoder_queue[i]));
+				wake_up(&(dev_priv->decoder_queue[i]));
 			}
 			*lock = 0;
 		}
@@ -83,10 +83,10 @@ int via_decoder_futex(struct drm_device *dev, void *data, struct drm_file *file_
 	switch (fx->func) {
 	case VIA_FUTEX_WAIT:
 		DRM_WAIT_ON(ret, dev_priv->decoder_queue[fx->lock],
-			    (fx->ms / 10) * (DRM_HZ / 100), *lock != fx->val);
+			    (fx->ms / 10) * (HZ / 100), *lock != fx->val);
 		return ret;
 	case VIA_FUTEX_WAKE:
-		DRM_WAKEUP(&(dev_priv->decoder_queue[fx->lock]));
+		wake_up(&(dev_priv->decoder_queue[fx->lock]));
 		return 0;
 	}
 	return 0;
diff --git a/drivers/gpu/drm/vmwgfx/Makefile b/drivers/gpu/drm/vmwgfx/Makefile
index 9f8b690bcf52..458cdf6d81e8 100644
--- a/drivers/gpu/drm/vmwgfx/Makefile
+++ b/drivers/gpu/drm/vmwgfx/Makefile
@@ -6,6 +6,6 @@ vmwgfx-y := vmwgfx_execbuf.o vmwgfx_gmr.o vmwgfx_kms.o vmwgfx_drv.o \
 	    vmwgfx_fifo.o vmwgfx_irq.o vmwgfx_ldu.o vmwgfx_ttm_glue.o \
 	    vmwgfx_overlay.o vmwgfx_marker.o vmwgfx_gmrid_manager.o \
 	    vmwgfx_fence.o vmwgfx_dmabuf.o vmwgfx_scrn.o vmwgfx_context.o \
-	    vmwgfx_surface.o vmwgfx_prime.o
+	    vmwgfx_surface.o vmwgfx_prime.o vmwgfx_mob.o vmwgfx_shader.o
 
 obj-$(CONFIG_DRM_VMWGFX) := vmwgfx.o
diff --git a/drivers/gpu/drm/vmwgfx/svga3d_reg.h b/drivers/gpu/drm/vmwgfx/svga3d_reg.h
index d0e085ee8249..d95335cb90bd 100644
--- a/drivers/gpu/drm/vmwgfx/svga3d_reg.h
+++ b/drivers/gpu/drm/vmwgfx/svga3d_reg.h
@@ -34,6 +34,8 @@
 
 #include "svga_reg.h"
 
+typedef uint32 PPN;
+typedef __le64 PPN64;
 
 /*
  * 3D Hardware Version
@@ -71,6 +73,9 @@ typedef uint32 SVGA3dBool; /* 32-bit Bool definition */
 #define SVGA3D_MAX_CONTEXT_IDS                  256
 #define SVGA3D_MAX_SURFACE_IDS                  (32 * 1024)
 
+#define SVGA3D_NUM_TEXTURE_UNITS                32
+#define SVGA3D_NUM_LIGHTS                       8
+
 /*
  * Surface formats.
  *
@@ -81,6 +86,7 @@ typedef uint32 SVGA3dBool; /* 32-bit Bool definition */
  */
 
 typedef enum SVGA3dSurfaceFormat {
+   SVGA3D_FORMAT_MIN                   = 0,
    SVGA3D_FORMAT_INVALID               = 0,
 
    SVGA3D_X8R8G8B8                     = 1,
@@ -134,12 +140,6 @@ typedef enum SVGA3dSurfaceFormat {
    SVGA3D_RG_S10E5                     = 35,
    SVGA3D_RG_S23E8                     = 36,
 
-   /*
-    * Any surface can be used as a buffer object, but SVGA3D_BUFFER is
-    * the most efficient format to use when creating new surfaces
-    * expressly for index or vertex data.
-    */
-
    SVGA3D_BUFFER                       = 37,
 
    SVGA3D_Z_D24X8                      = 38,
@@ -159,15 +159,114 @@ typedef enum SVGA3dSurfaceFormat {
    /* Video format with alpha */
    SVGA3D_AYUV                         = 45,
 
+   SVGA3D_R32G32B32A32_TYPELESS        = 46,
+   SVGA3D_R32G32B32A32_FLOAT           = 25,
+   SVGA3D_R32G32B32A32_UINT            = 47,
+   SVGA3D_R32G32B32A32_SINT            = 48,
+   SVGA3D_R32G32B32_TYPELESS           = 49,
+   SVGA3D_R32G32B32_FLOAT              = 50,
+   SVGA3D_R32G32B32_UINT               = 51,
+   SVGA3D_R32G32B32_SINT               = 52,
+   SVGA3D_R16G16B16A16_TYPELESS        = 53,
+   SVGA3D_R16G16B16A16_FLOAT           = 24,
+   SVGA3D_R16G16B16A16_UNORM           = 41,
+   SVGA3D_R16G16B16A16_UINT            = 54,
+   SVGA3D_R16G16B16A16_SNORM           = 55,
+   SVGA3D_R16G16B16A16_SINT            = 56,
+   SVGA3D_R32G32_TYPELESS              = 57,
+   SVGA3D_R32G32_FLOAT                 = 36,
+   SVGA3D_R32G32_UINT                  = 58,
+   SVGA3D_R32G32_SINT                  = 59,
+   SVGA3D_R32G8X24_TYPELESS            = 60,
+   SVGA3D_D32_FLOAT_S8X24_UINT         = 61,
+   SVGA3D_R32_FLOAT_X8X24_TYPELESS     = 62,
+   SVGA3D_X32_TYPELESS_G8X24_UINT      = 63,
+   SVGA3D_R10G10B10A2_TYPELESS         = 64,
+   SVGA3D_R10G10B10A2_UNORM            = 26,
+   SVGA3D_R10G10B10A2_UINT             = 65,
+   SVGA3D_R11G11B10_FLOAT              = 66,
+   SVGA3D_R8G8B8A8_TYPELESS            = 67,
+   SVGA3D_R8G8B8A8_UNORM               = 68,
+   SVGA3D_R8G8B8A8_UNORM_SRGB          = 69,
+   SVGA3D_R8G8B8A8_UINT                = 70,
+   SVGA3D_R8G8B8A8_SNORM               = 28,
+   SVGA3D_R8G8B8A8_SINT                = 71,
+   SVGA3D_R16G16_TYPELESS              = 72,
+   SVGA3D_R16G16_FLOAT                 = 35,
+   SVGA3D_R16G16_UNORM                 = 40,
+   SVGA3D_R16G16_UINT                  = 73,
+   SVGA3D_R16G16_SNORM                 = 39,
+   SVGA3D_R16G16_SINT                  = 74,
+   SVGA3D_R32_TYPELESS                 = 75,
+   SVGA3D_D32_FLOAT                    = 76,
+   SVGA3D_R32_FLOAT                    = 34,
+   SVGA3D_R32_UINT                     = 77,
+   SVGA3D_R32_SINT                     = 78,
+   SVGA3D_R24G8_TYPELESS               = 79,
+   SVGA3D_D24_UNORM_S8_UINT            = 80,
+   SVGA3D_R24_UNORM_X8_TYPELESS        = 81,
+   SVGA3D_X24_TYPELESS_G8_UINT         = 82,
+   SVGA3D_R8G8_TYPELESS                = 83,
+   SVGA3D_R8G8_UNORM                   = 84,
+   SVGA3D_R8G8_UINT                    = 85,
+   SVGA3D_R8G8_SNORM                   = 27,
+   SVGA3D_R8G8_SINT                    = 86,
+   SVGA3D_R16_TYPELESS                 = 87,
+   SVGA3D_R16_FLOAT                    = 33,
+   SVGA3D_D16_UNORM                    = 8,
+   SVGA3D_R16_UNORM                    = 88,
+   SVGA3D_R16_UINT                     = 89,
+   SVGA3D_R16_SNORM                    = 90,
+   SVGA3D_R16_SINT                     = 91,
+   SVGA3D_R8_TYPELESS                  = 92,
+   SVGA3D_R8_UNORM                     = 93,
+   SVGA3D_R8_UINT                      = 94,
+   SVGA3D_R8_SNORM                     = 95,
+   SVGA3D_R8_SINT                      = 96,
+   SVGA3D_A8_UNORM                     = 32,
+   SVGA3D_R1_UNORM                     = 97,
+   SVGA3D_R9G9B9E5_SHAREDEXP           = 98,
+   SVGA3D_R8G8_B8G8_UNORM              = 99,
+   SVGA3D_G8R8_G8B8_UNORM              = 100,
+   SVGA3D_BC1_TYPELESS                 = 101,
+   SVGA3D_BC1_UNORM                    = 15,
+   SVGA3D_BC1_UNORM_SRGB               = 102,
+   SVGA3D_BC2_TYPELESS                 = 103,
+   SVGA3D_BC2_UNORM                    = 17,
+   SVGA3D_BC2_UNORM_SRGB               = 104,
+   SVGA3D_BC3_TYPELESS                 = 105,
+   SVGA3D_BC3_UNORM                    = 19,
+   SVGA3D_BC3_UNORM_SRGB               = 106,
+   SVGA3D_BC4_TYPELESS                 = 107,
    SVGA3D_BC4_UNORM                    = 108,
+   SVGA3D_BC4_SNORM                    = 109,
+   SVGA3D_BC5_TYPELESS                 = 110,
    SVGA3D_BC5_UNORM                    = 111,
+   SVGA3D_BC5_SNORM                    = 112,
+   SVGA3D_B5G6R5_UNORM                 = 3,
+   SVGA3D_B5G5R5A1_UNORM               = 5,
+   SVGA3D_B8G8R8A8_UNORM               = 2,
+   SVGA3D_B8G8R8X8_UNORM               = 1,
+   SVGA3D_R10G10B10_XR_BIAS_A2_UNORM   = 113,
+   SVGA3D_B8G8R8A8_TYPELESS            = 114,
+   SVGA3D_B8G8R8A8_UNORM_SRGB          = 115,
+   SVGA3D_B8G8R8X8_TYPELESS            = 116,
+   SVGA3D_B8G8R8X8_UNORM_SRGB          = 117,
 
    /* Advanced D3D9 depth formats. */
    SVGA3D_Z_DF16                       = 118,
    SVGA3D_Z_DF24                       = 119,
    SVGA3D_Z_D24S8_INT                  = 120,
 
-   SVGA3D_FORMAT_MAX
+   /* Planar video formats. */
+   SVGA3D_YV12                         = 121,
+
+   /* Shader constant formats. */
+   SVGA3D_SURFACE_SHADERCONST_FLOAT    = 122,
+   SVGA3D_SURFACE_SHADERCONST_INT      = 123,
+   SVGA3D_SURFACE_SHADERCONST_BOOL     = 124,
+
+   SVGA3D_FORMAT_MAX                   = 125,
 } SVGA3dSurfaceFormat;
 
 typedef uint32 SVGA3dColor; /* a, r, g, b */
@@ -957,15 +1056,21 @@ typedef enum {
 } SVGA3dCubeFace;
 
 typedef enum {
+   SVGA3D_SHADERTYPE_INVALID                    = 0,
+   SVGA3D_SHADERTYPE_MIN                        = 1,
    SVGA3D_SHADERTYPE_VS                         = 1,
    SVGA3D_SHADERTYPE_PS                         = 2,
-   SVGA3D_SHADERTYPE_MAX
+   SVGA3D_SHADERTYPE_MAX                        = 3,
+   SVGA3D_SHADERTYPE_GS                         = 3,
 } SVGA3dShaderType;
 
+#define SVGA3D_NUM_SHADERTYPE (SVGA3D_SHADERTYPE_MAX - SVGA3D_SHADERTYPE_MIN)
+
 typedef enum {
    SVGA3D_CONST_TYPE_FLOAT                      = 0,
    SVGA3D_CONST_TYPE_INT                        = 1,
    SVGA3D_CONST_TYPE_BOOL                       = 2,
+   SVGA3D_CONST_TYPE_MAX
 } SVGA3dShaderConstType;
 
 #define SVGA3D_MAX_SURFACE_FACES                6
@@ -1056,9 +1161,74 @@ typedef enum {
 #define SVGA_3D_CMD_GENERATE_MIPMAPS       SVGA_3D_CMD_BASE + 31
 #define SVGA_3D_CMD_ACTIVATE_SURFACE       SVGA_3D_CMD_BASE + 40
 #define SVGA_3D_CMD_DEACTIVATE_SURFACE     SVGA_3D_CMD_BASE + 41
-#define SVGA_3D_CMD_MAX                    SVGA_3D_CMD_BASE + 42
-
-#define SVGA_3D_CMD_FUTURE_MAX             2000
+#define SVGA_3D_CMD_SCREEN_DMA               1082
+#define SVGA_3D_CMD_SET_UNITY_SURFACE_COOKIE 1083
+#define SVGA_3D_CMD_OPEN_CONTEXT_SURFACE     1084
+
+#define SVGA_3D_CMD_LOGICOPS_BITBLT          1085
+#define SVGA_3D_CMD_LOGICOPS_TRANSBLT        1086
+#define SVGA_3D_CMD_LOGICOPS_STRETCHBLT      1087
+#define SVGA_3D_CMD_LOGICOPS_COLORFILL       1088
+#define SVGA_3D_CMD_LOGICOPS_ALPHABLEND      1089
+#define SVGA_3D_CMD_LOGICOPS_CLEARTYPEBLEND  1090
+
+#define SVGA_3D_CMD_SET_OTABLE_BASE          1091
+#define SVGA_3D_CMD_READBACK_OTABLE          1092
+
+#define SVGA_3D_CMD_DEFINE_GB_MOB            1093
+#define SVGA_3D_CMD_DESTROY_GB_MOB           1094
+#define SVGA_3D_CMD_REDEFINE_GB_MOB          1095
+#define SVGA_3D_CMD_UPDATE_GB_MOB_MAPPING    1096
+
+#define SVGA_3D_CMD_DEFINE_GB_SURFACE        1097
+#define SVGA_3D_CMD_DESTROY_GB_SURFACE       1098
+#define SVGA_3D_CMD_BIND_GB_SURFACE          1099
+#define SVGA_3D_CMD_COND_BIND_GB_SURFACE     1100
+#define SVGA_3D_CMD_UPDATE_GB_IMAGE          1101
+#define SVGA_3D_CMD_UPDATE_GB_SURFACE        1102
+#define SVGA_3D_CMD_READBACK_GB_IMAGE        1103
+#define SVGA_3D_CMD_READBACK_GB_SURFACE      1104
+#define SVGA_3D_CMD_INVALIDATE_GB_IMAGE      1105
+#define SVGA_3D_CMD_INVALIDATE_GB_SURFACE    1106
+
+#define SVGA_3D_CMD_DEFINE_GB_CONTEXT        1107
+#define SVGA_3D_CMD_DESTROY_GB_CONTEXT       1108
+#define SVGA_3D_CMD_BIND_GB_CONTEXT          1109
+#define SVGA_3D_CMD_READBACK_GB_CONTEXT      1110
+#define SVGA_3D_CMD_INVALIDATE_GB_CONTEXT    1111
+
+#define SVGA_3D_CMD_DEFINE_GB_SHADER         1112
+#define SVGA_3D_CMD_DESTROY_GB_SHADER        1113
+#define SVGA_3D_CMD_BIND_GB_SHADER           1114
+
+#define SVGA_3D_CMD_SET_OTABLE_BASE64        1115
+
+#define SVGA_3D_CMD_BEGIN_GB_QUERY           1116
+#define SVGA_3D_CMD_END_GB_QUERY             1117
+#define SVGA_3D_CMD_WAIT_FOR_GB_QUERY        1118
+
+#define SVGA_3D_CMD_NOP                      1119
+
+#define SVGA_3D_CMD_ENABLE_GART              1120
+#define SVGA_3D_CMD_DISABLE_GART             1121
+#define SVGA_3D_CMD_MAP_MOB_INTO_GART        1122
+#define SVGA_3D_CMD_UNMAP_GART_RANGE         1123
+
+#define SVGA_3D_CMD_DEFINE_GB_SCREENTARGET   1124
+#define SVGA_3D_CMD_DESTROY_GB_SCREENTARGET  1125
+#define SVGA_3D_CMD_BIND_GB_SCREENTARGET     1126
+#define SVGA_3D_CMD_UPDATE_GB_SCREENTARGET   1127
+
+#define SVGA_3D_CMD_READBACK_GB_IMAGE_PARTIAL   1128
+#define SVGA_3D_CMD_INVALIDATE_GB_IMAGE_PARTIAL 1129
+
+#define SVGA_3D_CMD_SET_GB_SHADERCONSTS_INLINE  1130
+
+#define SVGA_3D_CMD_DEFINE_GB_MOB64          1135
+#define SVGA_3D_CMD_REDEFINE_GB_MOB64        1136
+
+#define SVGA_3D_CMD_MAX                      1142
+#define SVGA_3D_CMD_FUTURE_MAX               3000
 
 /*
  * Common substructures used in multiple FIFO commands:
@@ -1750,6 +1920,495 @@ struct {
 
 
 /*
+ * Guest-backed surface definitions.
+ */
+
+typedef uint32 SVGAMobId;
+
+typedef enum SVGAMobFormat {
+   SVGA3D_MOBFMT_INVALID = SVGA3D_INVALID_ID,
+   SVGA3D_MOBFMT_PTDEPTH_0 = 0,
+   SVGA3D_MOBFMT_PTDEPTH_1 = 1,
+   SVGA3D_MOBFMT_PTDEPTH_2 = 2,
+   SVGA3D_MOBFMT_RANGE     = 3,
+   SVGA3D_MOBFMT_PTDEPTH64_0 = 4,
+   SVGA3D_MOBFMT_PTDEPTH64_1 = 5,
+   SVGA3D_MOBFMT_PTDEPTH64_2 = 6,
+   SVGA3D_MOBFMT_MAX,
+} SVGAMobFormat;
+
+/*
+ * Sizes of opaque types.
+ */
+
+#define SVGA3D_OTABLE_MOB_ENTRY_SIZE 16
+#define SVGA3D_OTABLE_CONTEXT_ENTRY_SIZE 8
+#define SVGA3D_OTABLE_SURFACE_ENTRY_SIZE 64
+#define SVGA3D_OTABLE_SHADER_ENTRY_SIZE 16
+#define SVGA3D_OTABLE_SCREEN_TARGET_ENTRY_SIZE 64
+#define SVGA3D_CONTEXT_DATA_SIZE 16384
+
+/*
+ * SVGA3dCmdSetOTableBase --
+ *
+ * This command allows the guest to specify the base PPN of the
+ * specified object table.
+ */
+
+typedef enum {
+   SVGA_OTABLE_MOB           = 0,
+   SVGA_OTABLE_MIN           = 0,
+   SVGA_OTABLE_SURFACE       = 1,
+   SVGA_OTABLE_CONTEXT       = 2,
+   SVGA_OTABLE_SHADER        = 3,
+   SVGA_OTABLE_SCREEN_TARGET = 4,
+   SVGA_OTABLE_DX9_MAX       = 5,
+   SVGA_OTABLE_MAX           = 8
+} SVGAOTableType;
+
+typedef
+struct {
+   SVGAOTableType type;
+   PPN baseAddress;
+   uint32 sizeInBytes;
+   uint32 validSizeInBytes;
+   SVGAMobFormat ptDepth;
+}
+__attribute__((__packed__))
+SVGA3dCmdSetOTableBase;  /* SVGA_3D_CMD_SET_OTABLE_BASE */
+
+typedef
+struct {
+   SVGAOTableType type;
+   PPN64 baseAddress;
+   uint32 sizeInBytes;
+   uint32 validSizeInBytes;
+   SVGAMobFormat ptDepth;
+}
+__attribute__((__packed__))
+SVGA3dCmdSetOTableBase64;  /* SVGA_3D_CMD_SET_OTABLE_BASE64 */
+
+typedef
+struct {
+   SVGAOTableType type;
+}
+__attribute__((__packed__))
+SVGA3dCmdReadbackOTable;  /* SVGA_3D_CMD_READBACK_OTABLE */
+
+/*
+ * Define a memory object (Mob) in the OTable.
+ */
+
+typedef
+struct SVGA3dCmdDefineGBMob {
+   SVGAMobId mobid;
+   SVGAMobFormat ptDepth;
+   PPN base;
+   uint32 sizeInBytes;
+}
+__attribute__((__packed__))
+SVGA3dCmdDefineGBMob;   /* SVGA_3D_CMD_DEFINE_GB_MOB */
+
+
+/*
+ * Destroys an object in the OTable.
+ */
+
+typedef
+struct SVGA3dCmdDestroyGBMob {
+   SVGAMobId mobid;
+}
+__attribute__((__packed__))
+SVGA3dCmdDestroyGBMob;   /* SVGA_3D_CMD_DESTROY_GB_MOB */
+
+/*
+ * Redefine an object in the OTable.
+ */
+
+typedef
+struct SVGA3dCmdRedefineGBMob {
+   SVGAMobId mobid;
+   SVGAMobFormat ptDepth;
+   PPN base;
+   uint32 sizeInBytes;
+}
+__attribute__((__packed__))
+SVGA3dCmdRedefineGBMob;   /* SVGA_3D_CMD_REDEFINE_GB_MOB */
+
+/*
+ * Define a memory object (Mob) in the OTable with a PPN64 base.
+ */
+
+typedef
+struct SVGA3dCmdDefineGBMob64 {
+   SVGAMobId mobid;
+   SVGAMobFormat ptDepth;
+   PPN64 base;
+   uint32 sizeInBytes;
+}
+__attribute__((__packed__))
+SVGA3dCmdDefineGBMob64;   /* SVGA_3D_CMD_DEFINE_GB_MOB64 */
+
+/*
+ * Redefine an object in the OTable with PPN64 base.
+ */
+
+typedef
+struct SVGA3dCmdRedefineGBMob64 {
+   SVGAMobId mobid;
+   SVGAMobFormat ptDepth;
+   PPN64 base;
+   uint32 sizeInBytes;
+}
+__attribute__((__packed__))
+SVGA3dCmdRedefineGBMob64;   /* SVGA_3D_CMD_REDEFINE_GB_MOB64 */
+
+/*
+ * Notification that the page tables have been modified.
+ */
+
+typedef
+struct SVGA3dCmdUpdateGBMobMapping {
+   SVGAMobId mobid;
+}
+__attribute__((__packed__))
+SVGA3dCmdUpdateGBMobMapping;   /* SVGA_3D_CMD_UPDATE_GB_MOB_MAPPING */
+
+/*
+ * Define a guest-backed surface.
+ */
+
+typedef
+struct SVGA3dCmdDefineGBSurface {
+   uint32 sid;
+   SVGA3dSurfaceFlags surfaceFlags;
+   SVGA3dSurfaceFormat format;
+   uint32 numMipLevels;
+   uint32 multisampleCount;
+   SVGA3dTextureFilter autogenFilter;
+   SVGA3dSize size;
+} SVGA3dCmdDefineGBSurface;   /* SVGA_3D_CMD_DEFINE_GB_SURFACE */
+
+/*
+ * Destroy a guest-backed surface.
+ */
+
+typedef
+struct SVGA3dCmdDestroyGBSurface {
+   uint32 sid;
+} SVGA3dCmdDestroyGBSurface;   /* SVGA_3D_CMD_DESTROY_GB_SURFACE */
+
+/*
+ * Bind a guest-backed surface to an object.
+ */
+
+typedef
+struct SVGA3dCmdBindGBSurface {
+   uint32 sid;
+   SVGAMobId mobid;
+} SVGA3dCmdBindGBSurface;   /* SVGA_3D_CMD_BIND_GB_SURFACE */
+
+/*
+ * Conditionally bind a mob to a guest backed surface if testMobid
+ * matches the currently bound mob.  Optionally issue a readback on
+ * the surface while it is still bound to the old mobid if the mobid
+ * is changed by this command.
+ */
+
+#define SVGA3D_COND_BIND_GB_SURFACE_FLAG_READBACK (1 << 0)
+
+typedef
+struct{
+   uint32 sid;
+   SVGAMobId testMobid;
+   SVGAMobId mobid;
+   uint32 flags;
+}
+SVGA3dCmdCondBindGBSurface;          /* SVGA_3D_CMD_COND_BIND_GB_SURFACE */
+
+/*
+ * Update an image in a guest-backed surface.
+ * (Inform the device that the guest-contents have been updated.)
+ */
+
+typedef
+struct SVGA3dCmdUpdateGBImage {
+   SVGA3dSurfaceImageId image;
+   SVGA3dBox box;
+} SVGA3dCmdUpdateGBImage;   /* SVGA_3D_CMD_UPDATE_GB_IMAGE */
+
+/*
+ * Update an entire guest-backed surface.
+ * (Inform the device that the guest-contents have been updated.)
+ */
+
+typedef
+struct SVGA3dCmdUpdateGBSurface {
+   uint32 sid;
+} SVGA3dCmdUpdateGBSurface;   /* SVGA_3D_CMD_UPDATE_GB_SURFACE */
+
+/*
+ * Readback an image in a guest-backed surface.
+ * (Request the device to flush the dirty contents into the guest.)
+ */
+
+typedef
+struct SVGA3dCmdReadbackGBImage {
+   SVGA3dSurfaceImageId image;
+} SVGA3dCmdReadbackGBImage;   /* SVGA_3D_CMD_READBACK_GB_IMAGE*/
+
+/*
+ * Readback an entire guest-backed surface.
+ * (Request the device to flush the dirty contents into the guest.)
+ */
+
+typedef
+struct SVGA3dCmdReadbackGBSurface {
+   uint32 sid;
+} SVGA3dCmdReadbackGBSurface;   /* SVGA_3D_CMD_READBACK_GB_SURFACE */
+
+/*
+ * Readback a sub rect of an image in a guest-backed surface.  After
+ * issuing this command the driver is required to issue an update call
+ * of the same region before issuing any other commands that reference
+ * this surface or rendering is not guaranteed.
+ */
+
+typedef
+struct SVGA3dCmdReadbackGBImagePartial {
+   SVGA3dSurfaceImageId image;
+   SVGA3dBox box;
+   uint32 invertBox;
+}
+SVGA3dCmdReadbackGBImagePartial; /* SVGA_3D_CMD_READBACK_GB_IMAGE_PARTIAL */
+
+/*
+ * Invalidate an image in a guest-backed surface.
+ * (Notify the device that the contents can be lost.)
+ */
+
+typedef
+struct SVGA3dCmdInvalidateGBImage {
+   SVGA3dSurfaceImageId image;
+} SVGA3dCmdInvalidateGBImage;   /* SVGA_3D_CMD_INVALIDATE_GB_IMAGE */
+
+/*
+ * Invalidate an entire guest-backed surface.
+ * (Notify the device that the contents if all images can be lost.)
+ */
+
+typedef
+struct SVGA3dCmdInvalidateGBSurface {
+   uint32 sid;
+} SVGA3dCmdInvalidateGBSurface; /* SVGA_3D_CMD_INVALIDATE_GB_SURFACE */
+
+/*
+ * Invalidate a sub rect of an image in a guest-backed surface.  After
+ * issuing this command the driver is required to issue an update call
+ * of the same region before issuing any other commands that reference
+ * this surface or rendering is not guaranteed.
+ */
+
+typedef
+struct SVGA3dCmdInvalidateGBImagePartial {
+   SVGA3dSurfaceImageId image;
+   SVGA3dBox box;
+   uint32 invertBox;
+}
+SVGA3dCmdInvalidateGBImagePartial; /* SVGA_3D_CMD_INVALIDATE_GB_IMAGE_PARTIAL */
+
+/*
+ * Define a guest-backed context.
+ */
+
+typedef
+struct SVGA3dCmdDefineGBContext {
+   uint32 cid;
+} SVGA3dCmdDefineGBContext;   /* SVGA_3D_CMD_DEFINE_GB_CONTEXT */
+
+/*
+ * Destroy a guest-backed context.
+ */
+
+typedef
+struct SVGA3dCmdDestroyGBContext {
+   uint32 cid;
+} SVGA3dCmdDestroyGBContext;   /* SVGA_3D_CMD_DESTROY_GB_CONTEXT */
+
+/*
+ * Bind a guest-backed context.
+ *
+ * validContents should be set to 0 for new contexts,
+ * and 1 if this is an old context which is getting paged
+ * back on to the device.
+ *
+ * For new contexts, it is recommended that the driver
+ * issue commands to initialize all interesting state
+ * prior to rendering.
+ */
+
+typedef
+struct SVGA3dCmdBindGBContext {
+   uint32 cid;
+   SVGAMobId mobid;
+   uint32 validContents;
+} SVGA3dCmdBindGBContext;   /* SVGA_3D_CMD_BIND_GB_CONTEXT */
+
+/*
+ * Readback a guest-backed context.
+ * (Request that the device flush the contents back into guest memory.)
+ */
+
+typedef
+struct SVGA3dCmdReadbackGBContext {
+   uint32 cid;
+} SVGA3dCmdReadbackGBContext;   /* SVGA_3D_CMD_READBACK_GB_CONTEXT */
+
+/*
+ * Invalidate a guest-backed context.
+ */
+typedef
+struct SVGA3dCmdInvalidateGBContext {
+   uint32 cid;
+} SVGA3dCmdInvalidateGBContext;   /* SVGA_3D_CMD_INVALIDATE_GB_CONTEXT */
+
+/*
+ * Define a guest-backed shader.
+ */
+
+typedef
+struct SVGA3dCmdDefineGBShader {
+   uint32 shid;
+   SVGA3dShaderType type;
+   uint32 sizeInBytes;
+} SVGA3dCmdDefineGBShader;   /* SVGA_3D_CMD_DEFINE_GB_SHADER */
+
+/*
+ * Bind a guest-backed shader.
+ */
+
+typedef struct SVGA3dCmdBindGBShader {
+   uint32 shid;
+   SVGAMobId mobid;
+   uint32 offsetInBytes;
+} SVGA3dCmdBindGBShader;   /* SVGA_3D_CMD_BIND_GB_SHADER */
+
+/*
+ * Destroy a guest-backed shader.
+ */
+
+typedef struct SVGA3dCmdDestroyGBShader {
+   uint32 shid;
+} SVGA3dCmdDestroyGBShader;   /* SVGA_3D_CMD_DESTROY_GB_SHADER */
+
+typedef
+struct {
+   uint32                  cid;
+   uint32                  regStart;
+   SVGA3dShaderType        shaderType;
+   SVGA3dShaderConstType   constType;
+
+   /*
+    * Followed by a variable number of shader constants.
+    *
+    * Note that FLOAT and INT constants are 4-dwords in length, while
+    * BOOL constants are 1-dword in length.
+    */
+} SVGA3dCmdSetGBShaderConstInline;
+/* SVGA_3D_CMD_SET_GB_SHADERCONSTS_INLINE */
+
+typedef
+struct {
+   uint32               cid;
+   SVGA3dQueryType      type;
+} SVGA3dCmdBeginGBQuery;           /* SVGA_3D_CMD_BEGIN_GB_QUERY */
+
+typedef
+struct {
+   uint32               cid;
+   SVGA3dQueryType      type;
+   SVGAMobId mobid;
+   uint32 offset;
+} SVGA3dCmdEndGBQuery;                  /* SVGA_3D_CMD_END_GB_QUERY */
+
+
+/*
+ * SVGA_3D_CMD_WAIT_FOR_GB_QUERY --
+ *
+ *    The semantics of this command are identical to the
+ *    SVGA_3D_CMD_WAIT_FOR_QUERY except that the results are written
+ *    to a Mob instead of a GMR.
+ */
+
+typedef
+struct {
+   uint32               cid;
+   SVGA3dQueryType      type;
+   SVGAMobId mobid;
+   uint32 offset;
+} SVGA3dCmdWaitForGBQuery;          /* SVGA_3D_CMD_WAIT_FOR_GB_QUERY */
+
+typedef
+struct {
+   SVGAMobId mobid;
+   uint32 fbOffset;
+   uint32 initalized;
+}
+SVGA3dCmdEnableGart;              /* SVGA_3D_CMD_ENABLE_GART */
+
+typedef
+struct {
+   SVGAMobId mobid;
+   uint32 gartOffset;
+}
+SVGA3dCmdMapMobIntoGart;          /* SVGA_3D_CMD_MAP_MOB_INTO_GART */
+
+
+typedef
+struct {
+   uint32 gartOffset;
+   uint32 numPages;
+}
+SVGA3dCmdUnmapGartRange;          /* SVGA_3D_CMD_UNMAP_GART_RANGE */
+
+
+/*
+ * Screen Targets
+ */
+#define SVGA_STFLAG_PRIMARY (1 << 0)
+
+typedef
+struct {
+   uint32 stid;
+   uint32 width;
+   uint32 height;
+   int32 xRoot;
+   int32 yRoot;
+   uint32 flags;
+}
+SVGA3dCmdDefineGBScreenTarget;    /* SVGA_3D_CMD_DEFINE_GB_SCREENTARGET */
+
+typedef
+struct {
+   uint32 stid;
+}
+SVGA3dCmdDestroyGBScreenTarget;  /* SVGA_3D_CMD_DESTROY_GB_SCREENTARGET */
+
+typedef
+struct {
+   uint32 stid;
+   SVGA3dSurfaceImageId image;
+}
+SVGA3dCmdBindGBScreenTarget;  /* SVGA_3D_CMD_BIND_GB_SCREENTARGET */
+
+typedef
+struct {
+   uint32 stid;
+   SVGA3dBox box;
+}
+SVGA3dCmdUpdateGBScreenTarget;  /* SVGA_3D_CMD_UPDATE_GB_SCREENTARGET */
+
+/*
  * Capability query index.
  *
  * Notes:
@@ -1879,10 +2538,41 @@ typedef enum {
    SVGA3D_DEVCAP_SURFACEFMT_BC5_UNORM              = 83,
 
    /*
-    * Don't add new caps into the previous section; the values in this
-    * enumeration must not change. You can put new values right before
-    * SVGA3D_DEVCAP_MAX.
+    * Deprecated.
     */
+   SVGA3D_DEVCAP_VGPU10                            = 84,
+
+   /*
+    * This contains several SVGA_3D_CAPS_VIDEO_DECODE elements
+    * ored together, one for every type of video decoding supported.
+    */
+   SVGA3D_DEVCAP_VIDEO_DECODE                      = 85,
+
+   /*
+    * This contains several SVGA_3D_CAPS_VIDEO_PROCESS elements
+    * ored together, one for every type of video processing supported.
+    */
+   SVGA3D_DEVCAP_VIDEO_PROCESS                     = 86,
+
+   SVGA3D_DEVCAP_LINE_AA                           = 87,  /* boolean */
+   SVGA3D_DEVCAP_LINE_STIPPLE                      = 88,  /* boolean */
+   SVGA3D_DEVCAP_MAX_LINE_WIDTH                    = 89,  /* float */
+   SVGA3D_DEVCAP_MAX_AA_LINE_WIDTH                 = 90,  /* float */
+
+   SVGA3D_DEVCAP_SURFACEFMT_YV12                   = 91,
+
+   /*
+    * Does the host support the SVGA logic ops commands?
+    */
+   SVGA3D_DEVCAP_LOGICOPS                          = 92,
+
+   /*
+    * What support does the host have for screen targets?
+    *
+    * See the SVGA3D_SCREENTARGET_CAP bits below.
+    */
+   SVGA3D_DEVCAP_SCREENTARGETS                     = 93,
+
    SVGA3D_DEVCAP_MAX                                  /* This must be the last index. */
 } SVGA3dDevCapIndex;
 
diff --git a/drivers/gpu/drm/vmwgfx/svga_reg.h b/drivers/gpu/drm/vmwgfx/svga_reg.h
index 01f63cb49678..71defa4d2d75 100644
--- a/drivers/gpu/drm/vmwgfx/svga_reg.h
+++ b/drivers/gpu/drm/vmwgfx/svga_reg.h
@@ -169,7 +169,10 @@ enum {
    SVGA_REG_TRACES = 45,            /* Enable trace-based updates even when FIFO is on */
    SVGA_REG_GMRS_MAX_PAGES = 46,    /* Maximum number of 4KB pages for all GMRs */
    SVGA_REG_MEMORY_SIZE = 47,       /* Total dedicated device memory excluding FIFO */
-   SVGA_REG_TOP = 48,               /* Must be 1 more than the last register */
+   SVGA_REG_MAX_PRIMARY_BOUNDING_BOX_MEM = 50,   /* Max primary memory */
+   SVGA_REG_SUGGESTED_GBOBJECT_MEM_SIZE_KB = 51, /* Suggested limit on mob mem */
+   SVGA_REG_DEV_CAP = 52,           /* Write dev cap index, read value */
+   SVGA_REG_TOP = 53,               /* Must be 1 more than the last register */
 
    SVGA_PALETTE_BASE = 1024,        /* Base of SVGA color map */
    /* Next 768 (== 256*3) registers exist for colormap */
@@ -431,7 +434,10 @@ struct SVGASignedPoint {
 #define SVGA_CAP_TRACES             0x00200000
 #define SVGA_CAP_GMR2               0x00400000
 #define SVGA_CAP_SCREEN_OBJECT_2    0x00800000
-
+#define SVGA_CAP_COMMAND_BUFFERS    0x01000000
+#define SVGA_CAP_DEAD1              0x02000000
+#define SVGA_CAP_CMD_BUFFERS_2      0x04000000
+#define SVGA_CAP_GBOBJECTS          0x08000000
 
 /*
  * FIFO register indices.
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_buffer.c b/drivers/gpu/drm/vmwgfx/vmwgfx_buffer.c
index 0489c6152482..6327cfc36805 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_buffer.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_buffer.c
@@ -40,6 +40,10 @@ static uint32_t vram_ne_placement_flags = TTM_PL_FLAG_VRAM |
 static uint32_t sys_placement_flags = TTM_PL_FLAG_SYSTEM |
 	TTM_PL_FLAG_CACHED;
 
+static uint32_t sys_ne_placement_flags = TTM_PL_FLAG_SYSTEM |
+	TTM_PL_FLAG_CACHED |
+	TTM_PL_FLAG_NO_EVICT;
+
 static uint32_t gmr_placement_flags = VMW_PL_FLAG_GMR |
 	TTM_PL_FLAG_CACHED;
 
@@ -47,6 +51,9 @@ static uint32_t gmr_ne_placement_flags = VMW_PL_FLAG_GMR |
 	TTM_PL_FLAG_CACHED |
 	TTM_PL_FLAG_NO_EVICT;
 
+static uint32_t mob_placement_flags = VMW_PL_FLAG_MOB |
+	TTM_PL_FLAG_CACHED;
+
 struct ttm_placement vmw_vram_placement = {
 	.fpfn = 0,
 	.lpfn = 0,
@@ -116,16 +123,26 @@ struct ttm_placement vmw_sys_placement = {
 	.busy_placement = &sys_placement_flags
 };
 
+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_GMR | TTM_PL_FLAG_CACHED,
+	VMW_PL_FLAG_MOB | TTM_PL_FLAG_CACHED
 };
 
 struct ttm_placement vmw_evictable_placement = {
 	.fpfn = 0,
 	.lpfn = 0,
-	.num_placement = 3,
+	.num_placement = 4,
 	.placement = evictable_placement_flags,
 	.num_busy_placement = 1,
 	.busy_placement = &sys_placement_flags
@@ -140,10 +157,21 @@ struct ttm_placement vmw_srf_placement = {
 	.busy_placement = gmr_vram_placement_flags
 };
 
+struct ttm_placement vmw_mob_placement = {
+	.fpfn = 0,
+	.lpfn = 0,
+	.num_placement = 1,
+	.num_busy_placement = 1,
+	.placement = &mob_placement_flags,
+	.busy_placement = &mob_placement_flags
+};
+
 struct vmw_ttm_tt {
 	struct ttm_dma_tt dma_ttm;
 	struct vmw_private *dev_priv;
 	int gmr_id;
+	struct vmw_mob *mob;
+	int mem_type;
 	struct sg_table sgt;
 	struct vmw_sg_table vsgt;
 	uint64_t sg_alloc_size;
@@ -244,6 +272,7 @@ void vmw_piter_start(struct vmw_piter *viter, const struct vmw_sg_table *vsgt,
 		viter->dma_address = &__vmw_piter_dma_addr;
 		viter->page = &__vmw_piter_non_sg_page;
 		viter->addrs = vsgt->addrs;
+		viter->pages = vsgt->pages;
 		break;
 	case vmw_dma_map_populate:
 	case vmw_dma_map_bind:
@@ -424,6 +453,63 @@ static void vmw_ttm_unmap_dma(struct vmw_ttm_tt *vmw_tt)
 	vmw_tt->mapped = false;
 }
 
+
+/**
+ * vmw_bo_map_dma - Make sure buffer object pages are visible to the device
+ *
+ * @bo: Pointer to a struct ttm_buffer_object
+ *
+ * Wrapper around vmw_ttm_map_dma, that takes a TTM buffer object pointer
+ * instead of a pointer to a struct vmw_ttm_backend as argument.
+ * Note that the buffer object must be either pinned or reserved before
+ * calling this function.
+ */
+int vmw_bo_map_dma(struct ttm_buffer_object *bo)
+{
+	struct vmw_ttm_tt *vmw_tt =
+		container_of(bo->ttm, struct vmw_ttm_tt, dma_ttm.ttm);
+
+	return vmw_ttm_map_dma(vmw_tt);
+}
+
+
+/**
+ * vmw_bo_unmap_dma - Make sure buffer object pages are visible to the device
+ *
+ * @bo: Pointer to a struct ttm_buffer_object
+ *
+ * Wrapper around vmw_ttm_unmap_dma, that takes a TTM buffer object pointer
+ * instead of a pointer to a struct vmw_ttm_backend as argument.
+ */
+void vmw_bo_unmap_dma(struct ttm_buffer_object *bo)
+{
+	struct vmw_ttm_tt *vmw_tt =
+		container_of(bo->ttm, struct vmw_ttm_tt, dma_ttm.ttm);
+
+	vmw_ttm_unmap_dma(vmw_tt);
+}
+
+
+/**
+ * vmw_bo_sg_table - Return a struct vmw_sg_table object for a
+ * TTM buffer object
+ *
+ * @bo: Pointer to a struct ttm_buffer_object
+ *
+ * Returns a pointer to a struct vmw_sg_table object. The object should
+ * not be freed after use.
+ * Note that for the device addresses to be valid, the buffer object must
+ * either be reserved or pinned.
+ */
+const struct vmw_sg_table *vmw_bo_sg_table(struct ttm_buffer_object *bo)
+{
+	struct vmw_ttm_tt *vmw_tt =
+		container_of(bo->ttm, struct vmw_ttm_tt, dma_ttm.ttm);
+
+	return &vmw_tt->vsgt;
+}
+
+
 static int vmw_ttm_bind(struct ttm_tt *ttm, struct ttm_mem_reg *bo_mem)
 {
 	struct vmw_ttm_tt *vmw_be =
@@ -435,9 +521,27 @@ static int vmw_ttm_bind(struct ttm_tt *ttm, struct ttm_mem_reg *bo_mem)
 		return ret;
 
 	vmw_be->gmr_id = bo_mem->start;
+	vmw_be->mem_type = bo_mem->mem_type;
+
+	switch (bo_mem->mem_type) {
+	case VMW_PL_GMR:
+		return vmw_gmr_bind(vmw_be->dev_priv, &vmw_be->vsgt,
+				    ttm->num_pages, vmw_be->gmr_id);
+	case VMW_PL_MOB:
+		if (unlikely(vmw_be->mob == NULL)) {
+			vmw_be->mob =
+				vmw_mob_create(ttm->num_pages);
+			if (unlikely(vmw_be->mob == NULL))
+				return -ENOMEM;
+		}
 
-	return vmw_gmr_bind(vmw_be->dev_priv, &vmw_be->vsgt,
-			    ttm->num_pages, vmw_be->gmr_id);
+		return vmw_mob_bind(vmw_be->dev_priv, vmw_be->mob,
+				    &vmw_be->vsgt, ttm->num_pages,
+				    vmw_be->gmr_id);
+	default:
+		BUG();
+	}
+	return 0;
 }
 
 static int vmw_ttm_unbind(struct ttm_tt *ttm)
@@ -445,7 +549,16 @@ static int vmw_ttm_unbind(struct ttm_tt *ttm)
 	struct vmw_ttm_tt *vmw_be =
 		container_of(ttm, struct vmw_ttm_tt, dma_ttm.ttm);
 
-	vmw_gmr_unbind(vmw_be->dev_priv, vmw_be->gmr_id);
+	switch (vmw_be->mem_type) {
+	case VMW_PL_GMR:
+		vmw_gmr_unbind(vmw_be->dev_priv, vmw_be->gmr_id);
+		break;
+	case VMW_PL_MOB:
+		vmw_mob_unbind(vmw_be->dev_priv, vmw_be->mob);
+		break;
+	default:
+		BUG();
+	}
 
 	if (vmw_be->dev_priv->map_mode == vmw_dma_map_bind)
 		vmw_ttm_unmap_dma(vmw_be);
@@ -453,6 +566,7 @@ static int vmw_ttm_unbind(struct ttm_tt *ttm)
 	return 0;
 }
 
+
 static void vmw_ttm_destroy(struct ttm_tt *ttm)
 {
 	struct vmw_ttm_tt *vmw_be =
@@ -463,9 +577,14 @@ static void vmw_ttm_destroy(struct ttm_tt *ttm)
 		ttm_dma_tt_fini(&vmw_be->dma_ttm);
 	else
 		ttm_tt_fini(ttm);
+
+	if (vmw_be->mob)
+		vmw_mob_destroy(vmw_be->mob);
+
 	kfree(vmw_be);
 }
 
+
 static int vmw_ttm_populate(struct ttm_tt *ttm)
 {
 	struct vmw_ttm_tt *vmw_tt =
@@ -500,6 +619,12 @@ static void vmw_ttm_unpopulate(struct ttm_tt *ttm)
 	struct vmw_private *dev_priv = vmw_tt->dev_priv;
 	struct ttm_mem_global *glob = vmw_mem_glob(dev_priv);
 
+
+	if (vmw_tt->mob) {
+		vmw_mob_destroy(vmw_tt->mob);
+		vmw_tt->mob = NULL;
+	}
+
 	vmw_ttm_unmap_dma(vmw_tt);
 	if (dev_priv->map_mode == vmw_dma_alloc_coherent) {
 		size_t size =
@@ -517,7 +642,7 @@ static struct ttm_backend_func vmw_ttm_func = {
 	.destroy = vmw_ttm_destroy,
 };
 
-struct ttm_tt *vmw_ttm_tt_create(struct ttm_bo_device *bdev,
+static struct ttm_tt *vmw_ttm_tt_create(struct ttm_bo_device *bdev,
 				 unsigned long size, uint32_t page_flags,
 				 struct page *dummy_read_page)
 {
@@ -530,6 +655,7 @@ struct ttm_tt *vmw_ttm_tt_create(struct ttm_bo_device *bdev,
 
 	vmw_be->dma_ttm.ttm.func = &vmw_ttm_func;
 	vmw_be->dev_priv = container_of(bdev, struct vmw_private, bdev);
+	vmw_be->mob = NULL;
 
 	if (vmw_be->dev_priv->map_mode == vmw_dma_alloc_coherent)
 		ret = ttm_dma_tt_init(&vmw_be->dma_ttm, bdev, size, page_flags,
@@ -546,12 +672,12 @@ out_no_init:
 	return NULL;
 }
 
-int vmw_invalidate_caches(struct ttm_bo_device *bdev, uint32_t flags)
+static int vmw_invalidate_caches(struct ttm_bo_device *bdev, uint32_t flags)
 {
 	return 0;
 }
 
-int vmw_init_mem_type(struct ttm_bo_device *bdev, uint32_t type,
+static int vmw_init_mem_type(struct ttm_bo_device *bdev, uint32_t type,
 		      struct ttm_mem_type_manager *man)
 {
 	switch (type) {
@@ -571,6 +697,7 @@ int vmw_init_mem_type(struct ttm_bo_device *bdev, uint32_t type,
 		man->default_caching = TTM_PL_FLAG_CACHED;
 		break;
 	case VMW_PL_GMR:
+	case VMW_PL_MOB:
 		/*
 		 * "Guest Memory Regions" is an aperture like feature with
 		 *  one slot per bo. There is an upper limit of the number of
@@ -589,7 +716,7 @@ int vmw_init_mem_type(struct ttm_bo_device *bdev, uint32_t type,
 	return 0;
 }
 
-void vmw_evict_flags(struct ttm_buffer_object *bo,
+static void vmw_evict_flags(struct ttm_buffer_object *bo,
 		     struct ttm_placement *placement)
 {
 	*placement = vmw_sys_placement;
@@ -618,6 +745,7 @@ static int vmw_ttm_io_mem_reserve(struct ttm_bo_device *bdev, struct ttm_mem_reg
 	switch (mem->mem_type) {
 	case TTM_PL_SYSTEM:
 	case VMW_PL_GMR:
+	case VMW_PL_MOB:
 		return 0;
 	case TTM_PL_VRAM:
 		mem->bus.offset = mem->start << PAGE_SHIFT;
@@ -677,6 +805,38 @@ static int vmw_sync_obj_wait(void *sync_obj, bool lazy, bool interruptible)
 				  VMW_FENCE_WAIT_TIMEOUT);
 }
 
+/**
+ * vmw_move_notify - TTM move_notify_callback
+ *
+ * @bo:             The TTM buffer object about to move.
+ * @mem:            The truct ttm_mem_reg indicating to what memory
+ *                  region the move is taking place.
+ *
+ * Calls move_notify for all subsystems needing it.
+ * (currently only resources).
+ */
+static void vmw_move_notify(struct ttm_buffer_object *bo,
+			    struct ttm_mem_reg *mem)
+{
+	vmw_resource_move_notify(bo, mem);
+}
+
+
+/**
+ * vmw_swap_notify - TTM move_notify_callback
+ *
+ * @bo:             The TTM buffer object about to be swapped out.
+ */
+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);
+}
+
+
 struct ttm_bo_driver vmw_bo_driver = {
 	.ttm_tt_create = &vmw_ttm_tt_create,
 	.ttm_tt_populate = &vmw_ttm_populate,
@@ -691,8 +851,8 @@ struct ttm_bo_driver vmw_bo_driver = {
 	.sync_obj_flush = vmw_sync_obj_flush,
 	.sync_obj_unref = vmw_sync_obj_unref,
 	.sync_obj_ref = vmw_sync_obj_ref,
-	.move_notify = NULL,
-	.swap_notify = NULL,
+	.move_notify = vmw_move_notify,
+	.swap_notify = vmw_swap_notify,
 	.fault_reserve_notify = &vmw_ttm_fault_reserve_notify,
 	.io_mem_reserve = &vmw_ttm_io_mem_reserve,
 	.io_mem_free = &vmw_ttm_io_mem_free,
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_context.c b/drivers/gpu/drm/vmwgfx/vmwgfx_context.c
index 00ae0925aca8..82c41daebc0e 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_context.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_context.c
@@ -32,12 +32,28 @@
 struct vmw_user_context {
 	struct ttm_base_object base;
 	struct vmw_resource res;
+	struct vmw_ctx_binding_state cbs;
 };
 
+
+
+typedef int (*vmw_scrub_func)(struct vmw_ctx_bindinfo *);
+
 static void vmw_user_context_free(struct vmw_resource *res);
 static struct vmw_resource *
 vmw_user_context_base_to_res(struct ttm_base_object *base);
 
+static int vmw_gb_context_create(struct vmw_resource *res);
+static int vmw_gb_context_bind(struct vmw_resource *res,
+			       struct ttm_validate_buffer *val_buf);
+static int vmw_gb_context_unbind(struct vmw_resource *res,
+				 bool readback,
+				 struct ttm_validate_buffer *val_buf);
+static int vmw_gb_context_destroy(struct vmw_resource *res);
+static int vmw_context_scrub_shader(struct vmw_ctx_bindinfo *bi);
+static int vmw_context_scrub_render_target(struct vmw_ctx_bindinfo *bi);
+static int vmw_context_scrub_texture(struct vmw_ctx_bindinfo *bi);
+static void vmw_context_binding_state_kill(struct vmw_ctx_binding_state *cbs);
 static uint64_t vmw_user_context_size;
 
 static const struct vmw_user_resource_conv user_context_conv = {
@@ -62,6 +78,23 @@ static const struct vmw_res_func vmw_legacy_context_func = {
 	.unbind = NULL
 };
 
+static const struct vmw_res_func vmw_gb_context_func = {
+	.res_type = vmw_res_context,
+	.needs_backup = true,
+	.may_evict = true,
+	.type_name = "guest backed contexts",
+	.backup_placement = &vmw_mob_placement,
+	.create = vmw_gb_context_create,
+	.destroy = vmw_gb_context_destroy,
+	.bind = vmw_gb_context_bind,
+	.unbind = vmw_gb_context_unbind
+};
+
+static const vmw_scrub_func vmw_scrub_funcs[vmw_ctx_binding_max] = {
+	[vmw_ctx_binding_shader] = vmw_context_scrub_shader,
+	[vmw_ctx_binding_rt] = vmw_context_scrub_render_target,
+	[vmw_ctx_binding_tex] = vmw_context_scrub_texture };
+
 /**
  * Context management:
  */
@@ -76,6 +109,16 @@ static void vmw_hw_context_destroy(struct vmw_resource *res)
 	} *cmd;
 
 
+	if (res->func->destroy == vmw_gb_context_destroy) {
+		mutex_lock(&dev_priv->cmdbuf_mutex);
+		(void) vmw_gb_context_destroy(res);
+		if (dev_priv->pinned_bo != NULL &&
+		    !dev_priv->query_cid_valid)
+			__vmw_execbuf_release_pinned_bo(dev_priv, NULL);
+		mutex_unlock(&dev_priv->cmdbuf_mutex);
+		return;
+	}
+
 	vmw_execbuf_release_pinned_bo(dev_priv);
 	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
 	if (unlikely(cmd == NULL)) {
@@ -92,6 +135,33 @@ static void vmw_hw_context_destroy(struct vmw_resource *res)
 	vmw_3d_resource_dec(dev_priv, false);
 }
 
+static int vmw_gb_context_init(struct vmw_private *dev_priv,
+			       struct vmw_resource *res,
+			       void (*res_free) (struct vmw_resource *res))
+{
+	int ret;
+	struct vmw_user_context *uctx =
+		container_of(res, struct vmw_user_context, res);
+
+	ret = vmw_resource_init(dev_priv, res, true,
+				res_free, &vmw_gb_context_func);
+	res->backup_size = SVGA3D_CONTEXT_DATA_SIZE;
+
+	if (unlikely(ret != 0)) {
+		if (res_free)
+			res_free(res);
+		else
+			kfree(res);
+		return ret;
+	}
+
+	memset(&uctx->cbs, 0, sizeof(uctx->cbs));
+	INIT_LIST_HEAD(&uctx->cbs.list);
+
+	vmw_resource_activate(res, vmw_hw_context_destroy);
+	return 0;
+}
+
 static int vmw_context_init(struct vmw_private *dev_priv,
 			    struct vmw_resource *res,
 			    void (*res_free) (struct vmw_resource *res))
@@ -103,6 +173,9 @@ static int vmw_context_init(struct vmw_private *dev_priv,
 		SVGA3dCmdDefineContext body;
 	} *cmd;
 
+	if (dev_priv->has_mob)
+		return vmw_gb_context_init(dev_priv, res, res_free);
+
 	ret = vmw_resource_init(dev_priv, res, false,
 				res_free, &vmw_legacy_context_func);
 
@@ -154,6 +227,184 @@ struct vmw_resource *vmw_context_alloc(struct vmw_private *dev_priv)
 	return (ret == 0) ? res : NULL;
 }
 
+
+static int vmw_gb_context_create(struct vmw_resource *res)
+{
+	struct vmw_private *dev_priv = res->dev_priv;
+	int ret;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdDefineGBContext body;
+	} *cmd;
+
+	if (likely(res->id != -1))
+		return 0;
+
+	ret = vmw_resource_alloc_id(res);
+	if (unlikely(ret != 0)) {
+		DRM_ERROR("Failed to allocate a context id.\n");
+		goto out_no_id;
+	}
+
+	if (unlikely(res->id >= VMWGFX_NUM_GB_CONTEXT)) {
+		ret = -EBUSY;
+		goto out_no_fifo;
+	}
+
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for context "
+			  "creation.\n");
+		ret = -ENOMEM;
+		goto out_no_fifo;
+	}
+
+	cmd->header.id = SVGA_3D_CMD_DEFINE_GB_CONTEXT;
+	cmd->header.size = sizeof(cmd->body);
+	cmd->body.cid = res->id;
+	vmw_fifo_commit(dev_priv, sizeof(*cmd));
+	(void) vmw_3d_resource_inc(dev_priv, false);
+
+	return 0;
+
+out_no_fifo:
+	vmw_resource_release_id(res);
+out_no_id:
+	return ret;
+}
+
+static int vmw_gb_context_bind(struct vmw_resource *res,
+			       struct ttm_validate_buffer *val_buf)
+{
+	struct vmw_private *dev_priv = res->dev_priv;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdBindGBContext body;
+	} *cmd;
+	struct ttm_buffer_object *bo = val_buf->bo;
+
+	BUG_ON(bo->mem.mem_type != VMW_PL_MOB);
+
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for context "
+			  "binding.\n");
+		return -ENOMEM;
+	}
+
+	cmd->header.id = SVGA_3D_CMD_BIND_GB_CONTEXT;
+	cmd->header.size = sizeof(cmd->body);
+	cmd->body.cid = res->id;
+	cmd->body.mobid = bo->mem.start;
+	cmd->body.validContents = res->backup_dirty;
+	res->backup_dirty = false;
+	vmw_fifo_commit(dev_priv, sizeof(*cmd));
+
+	return 0;
+}
+
+static int vmw_gb_context_unbind(struct vmw_resource *res,
+				 bool readback,
+				 struct ttm_validate_buffer *val_buf)
+{
+	struct vmw_private *dev_priv = res->dev_priv;
+	struct ttm_buffer_object *bo = val_buf->bo;
+	struct vmw_fence_obj *fence;
+	struct vmw_user_context *uctx =
+		container_of(res, struct vmw_user_context, res);
+
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdReadbackGBContext body;
+	} *cmd1;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdBindGBContext body;
+	} *cmd2;
+	uint32_t submit_size;
+	uint8_t *cmd;
+
+
+	BUG_ON(bo->mem.mem_type != VMW_PL_MOB);
+
+	mutex_lock(&dev_priv->binding_mutex);
+	vmw_context_binding_state_kill(&uctx->cbs);
+
+	submit_size = sizeof(*cmd2) + (readback ? sizeof(*cmd1) : 0);
+
+	cmd = vmw_fifo_reserve(dev_priv, submit_size);
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for context "
+			  "unbinding.\n");
+		mutex_unlock(&dev_priv->binding_mutex);
+		return -ENOMEM;
+	}
+
+	cmd2 = (void *) cmd;
+	if (readback) {
+		cmd1 = (void *) cmd;
+		cmd1->header.id = SVGA_3D_CMD_READBACK_GB_CONTEXT;
+		cmd1->header.size = sizeof(cmd1->body);
+		cmd1->body.cid = res->id;
+		cmd2 = (void *) (&cmd1[1]);
+	}
+	cmd2->header.id = SVGA_3D_CMD_BIND_GB_CONTEXT;
+	cmd2->header.size = sizeof(cmd2->body);
+	cmd2->body.cid = res->id;
+	cmd2->body.mobid = SVGA3D_INVALID_ID;
+
+	vmw_fifo_commit(dev_priv, submit_size);
+	mutex_unlock(&dev_priv->binding_mutex);
+
+	/*
+	 * Create a fence object and fence the backup buffer.
+	 */
+
+	(void) vmw_execbuf_fence_commands(NULL, dev_priv,
+					  &fence, NULL);
+
+	vmw_fence_single_bo(bo, fence);
+
+	if (likely(fence != NULL))
+		vmw_fence_obj_unreference(&fence);
+
+	return 0;
+}
+
+static int vmw_gb_context_destroy(struct vmw_resource *res)
+{
+	struct vmw_private *dev_priv = res->dev_priv;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdDestroyGBContext body;
+	} *cmd;
+	struct vmw_user_context *uctx =
+		container_of(res, struct vmw_user_context, res);
+
+	BUG_ON(!list_empty(&uctx->cbs.list));
+
+	if (likely(res->id == -1))
+		return 0;
+
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for context "
+			  "destruction.\n");
+		return -ENOMEM;
+	}
+
+	cmd->header.id = SVGA_3D_CMD_DESTROY_GB_CONTEXT;
+	cmd->header.size = sizeof(cmd->body);
+	cmd->body.cid = res->id;
+	vmw_fifo_commit(dev_priv, sizeof(*cmd));
+	if (dev_priv->query_cid == res->id)
+		dev_priv->query_cid_valid = false;
+	vmw_resource_release_id(res);
+	vmw_3d_resource_dec(dev_priv, false);
+
+	return 0;
+}
+
 /**
  * User-space context management:
  */
@@ -272,3 +523,283 @@ out_unlock:
 	return ret;
 
 }
+
+/**
+ * vmw_context_scrub_shader - scrub a shader binding from a context.
+ *
+ * @bi: single binding information.
+ */
+static int vmw_context_scrub_shader(struct vmw_ctx_bindinfo *bi)
+{
+	struct vmw_private *dev_priv = bi->ctx->dev_priv;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdSetShader body;
+	} *cmd;
+
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for shader "
+			  "unbinding.\n");
+		return -ENOMEM;
+	}
+
+	cmd->header.id = SVGA_3D_CMD_SET_SHADER;
+	cmd->header.size = sizeof(cmd->body);
+	cmd->body.cid = bi->ctx->id;
+	cmd->body.type = bi->i1.shader_type;
+	cmd->body.shid = SVGA3D_INVALID_ID;
+	vmw_fifo_commit(dev_priv, sizeof(*cmd));
+
+	return 0;
+}
+
+/**
+ * vmw_context_scrub_render_target - scrub a render target binding
+ * from a context.
+ *
+ * @bi: single binding information.
+ */
+static int vmw_context_scrub_render_target(struct vmw_ctx_bindinfo *bi)
+{
+	struct vmw_private *dev_priv = bi->ctx->dev_priv;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdSetRenderTarget body;
+	} *cmd;
+
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for render target "
+			  "unbinding.\n");
+		return -ENOMEM;
+	}
+
+	cmd->header.id = SVGA_3D_CMD_SETRENDERTARGET;
+	cmd->header.size = sizeof(cmd->body);
+	cmd->body.cid = bi->ctx->id;
+	cmd->body.type = bi->i1.rt_type;
+	cmd->body.target.sid = SVGA3D_INVALID_ID;
+	cmd->body.target.face = 0;
+	cmd->body.target.mipmap = 0;
+	vmw_fifo_commit(dev_priv, sizeof(*cmd));
+
+	return 0;
+}
+
+/**
+ * vmw_context_scrub_texture - scrub a texture binding from a context.
+ *
+ * @bi: single binding information.
+ *
+ * TODO: Possibly complement this function with a function that takes
+ * a list of texture bindings and combines them to a single command.
+ */
+static int vmw_context_scrub_texture(struct vmw_ctx_bindinfo *bi)
+{
+	struct vmw_private *dev_priv = bi->ctx->dev_priv;
+	struct {
+		SVGA3dCmdHeader header;
+		struct {
+			SVGA3dCmdSetTextureState c;
+			SVGA3dTextureState s1;
+		} body;
+	} *cmd;
+
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for texture "
+			  "unbinding.\n");
+		return -ENOMEM;
+	}
+
+
+	cmd->header.id = SVGA_3D_CMD_SETTEXTURESTATE;
+	cmd->header.size = sizeof(cmd->body);
+	cmd->body.c.cid = bi->ctx->id;
+	cmd->body.s1.stage = bi->i1.texture_stage;
+	cmd->body.s1.name = SVGA3D_TS_BIND_TEXTURE;
+	cmd->body.s1.value = (uint32) SVGA3D_INVALID_ID;
+	vmw_fifo_commit(dev_priv, sizeof(*cmd));
+
+	return 0;
+}
+
+/**
+ * vmw_context_binding_drop: Stop tracking a context binding
+ *
+ * @cb: Pointer to binding tracker storage.
+ *
+ * Stops tracking a context binding, and re-initializes its storage.
+ * Typically used when the context binding is replaced with a binding to
+ * another (or the same, for that matter) resource.
+ */
+static void vmw_context_binding_drop(struct vmw_ctx_binding *cb)
+{
+	list_del(&cb->ctx_list);
+	if (!list_empty(&cb->res_list))
+		list_del(&cb->res_list);
+	cb->bi.ctx = NULL;
+}
+
+/**
+ * vmw_context_binding_add: Start tracking a context binding
+ *
+ * @cbs: Pointer to the context binding state tracker.
+ * @bi: Information about the binding to track.
+ *
+ * Performs basic checks on the binding to make sure arguments are within
+ * bounds and then starts tracking the binding in the context binding
+ * state structure @cbs.
+ */
+int vmw_context_binding_add(struct vmw_ctx_binding_state *cbs,
+			    const struct vmw_ctx_bindinfo *bi)
+{
+	struct vmw_ctx_binding *loc;
+
+	switch (bi->bt) {
+	case vmw_ctx_binding_rt:
+		if (unlikely((unsigned)bi->i1.rt_type >= SVGA3D_RT_MAX)) {
+			DRM_ERROR("Illegal render target type %u.\n",
+				  (unsigned) bi->i1.rt_type);
+			return -EINVAL;
+		}
+		loc = &cbs->render_targets[bi->i1.rt_type];
+		break;
+	case vmw_ctx_binding_tex:
+		if (unlikely((unsigned)bi->i1.texture_stage >=
+			     SVGA3D_NUM_TEXTURE_UNITS)) {
+			DRM_ERROR("Illegal texture/sampler unit %u.\n",
+				  (unsigned) bi->i1.texture_stage);
+			return -EINVAL;
+		}
+		loc = &cbs->texture_units[bi->i1.texture_stage];
+		break;
+	case vmw_ctx_binding_shader:
+		if (unlikely((unsigned)bi->i1.shader_type >=
+			     SVGA3D_SHADERTYPE_MAX)) {
+			DRM_ERROR("Illegal shader type %u.\n",
+				  (unsigned) bi->i1.shader_type);
+			return -EINVAL;
+		}
+		loc = &cbs->shaders[bi->i1.shader_type];
+		break;
+	default:
+		BUG();
+	}
+
+	if (loc->bi.ctx != NULL)
+		vmw_context_binding_drop(loc);
+
+	loc->bi = *bi;
+	list_add_tail(&loc->ctx_list, &cbs->list);
+	INIT_LIST_HEAD(&loc->res_list);
+
+	return 0;
+}
+
+/**
+ * vmw_context_binding_transfer: Transfer a context binding tracking entry.
+ *
+ * @cbs: Pointer to the persistent context binding state tracker.
+ * @bi: Information about the binding to track.
+ *
+ */
+static void vmw_context_binding_transfer(struct vmw_ctx_binding_state *cbs,
+					 const struct vmw_ctx_bindinfo *bi)
+{
+	struct vmw_ctx_binding *loc;
+
+	switch (bi->bt) {
+	case vmw_ctx_binding_rt:
+		loc = &cbs->render_targets[bi->i1.rt_type];
+		break;
+	case vmw_ctx_binding_tex:
+		loc = &cbs->texture_units[bi->i1.texture_stage];
+		break;
+	case vmw_ctx_binding_shader:
+		loc = &cbs->shaders[bi->i1.shader_type];
+		break;
+	default:
+		BUG();
+	}
+
+	if (loc->bi.ctx != NULL)
+		vmw_context_binding_drop(loc);
+
+	loc->bi = *bi;
+	list_add_tail(&loc->ctx_list, &cbs->list);
+	if (bi->res != NULL)
+		list_add_tail(&loc->res_list, &bi->res->binding_head);
+	else
+		INIT_LIST_HEAD(&loc->res_list);
+}
+
+/**
+ * vmw_context_binding_kill - Kill a binding on the device
+ * and stop tracking it.
+ *
+ * @cb: Pointer to binding tracker storage.
+ *
+ * Emits FIFO commands to scrub a binding represented by @cb.
+ * Then stops tracking the binding and re-initializes its storage.
+ */
+static void vmw_context_binding_kill(struct vmw_ctx_binding *cb)
+{
+	(void) vmw_scrub_funcs[cb->bi.bt](&cb->bi);
+	vmw_context_binding_drop(cb);
+}
+
+/**
+ * vmw_context_binding_state_kill - Kill all bindings associated with a
+ * struct vmw_ctx_binding state structure, and re-initialize the structure.
+ *
+ * @cbs: Pointer to the context binding state tracker.
+ *
+ * Emits commands to scrub all bindings associated with the
+ * context binding state tracker. Then re-initializes the whole structure.
+ */
+static void vmw_context_binding_state_kill(struct vmw_ctx_binding_state *cbs)
+{
+	struct vmw_ctx_binding *entry, *next;
+
+	list_for_each_entry_safe(entry, next, &cbs->list, ctx_list)
+		vmw_context_binding_kill(entry);
+}
+
+/**
+ * vmw_context_binding_res_list_kill - Kill all bindings on a
+ * resource binding list
+ *
+ * @head: list head of resource binding list
+ *
+ * Kills all bindings associated with a specific resource. Typically
+ * called before the resource is destroyed.
+ */
+void vmw_context_binding_res_list_kill(struct list_head *head)
+{
+	struct vmw_ctx_binding *entry, *next;
+
+	list_for_each_entry_safe(entry, next, head, res_list)
+		vmw_context_binding_kill(entry);
+}
+
+/**
+ * vmw_context_binding_state_transfer - Commit staged binding info
+ *
+ * @ctx: Pointer to context to commit the staged binding info to.
+ * @from: Staged binding info built during execbuf.
+ *
+ * Transfers binding info from a temporary structure to the persistent
+ * structure in the context. This can be done once commands
+ */
+void vmw_context_binding_state_transfer(struct vmw_resource *ctx,
+					struct vmw_ctx_binding_state *from)
+{
+	struct vmw_user_context *uctx =
+		container_of(ctx, struct vmw_user_context, res);
+	struct vmw_ctx_binding *entry, *next;
+
+	list_for_each_entry_safe(entry, next, &from->list, ctx_list)
+		vmw_context_binding_transfer(&uctx->cbs, &entry->bi);
+}
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_dmabuf.c b/drivers/gpu/drm/vmwgfx/vmwgfx_dmabuf.c
index d4e54fcc0acd..a75840211b3c 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_dmabuf.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_dmabuf.c
@@ -290,8 +290,7 @@ void vmw_bo_get_guest_ptr(const struct ttm_buffer_object *bo,
 /**
  * vmw_bo_pin - Pin or unpin a buffer object without moving it.
  *
- * @bo: The buffer object. Must be reserved, and present either in VRAM
- * or GMR memory.
+ * @bo: The buffer object. Must be reserved.
  * @pin: Whether to pin or unpin.
  *
  */
@@ -303,10 +302,9 @@ void vmw_bo_pin(struct ttm_buffer_object *bo, bool pin)
 	int ret;
 
 	lockdep_assert_held(&bo->resv->lock.base);
-	BUG_ON(old_mem_type != TTM_PL_VRAM &&
-	       old_mem_type != VMW_PL_GMR);
 
-	pl_flags = TTM_PL_FLAG_VRAM | VMW_PL_FLAG_GMR | TTM_PL_FLAG_CACHED;
+	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;
 
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c
index c7a549694e59..9893328f8fdc 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c
@@ -112,6 +112,21 @@
 #define DRM_IOCTL_VMW_UPDATE_LAYOUT				\
 	DRM_IOW(DRM_COMMAND_BASE + DRM_VMW_UPDATE_LAYOUT,	\
 		 struct drm_vmw_update_layout_arg)
+#define DRM_IOCTL_VMW_CREATE_SHADER				\
+	DRM_IOWR(DRM_COMMAND_BASE + DRM_VMW_CREATE_SHADER,	\
+		 struct drm_vmw_shader_create_arg)
+#define DRM_IOCTL_VMW_UNREF_SHADER				\
+	DRM_IOW(DRM_COMMAND_BASE + DRM_VMW_UNREF_SHADER,	\
+		 struct drm_vmw_shader_arg)
+#define DRM_IOCTL_VMW_GB_SURFACE_CREATE				\
+	DRM_IOWR(DRM_COMMAND_BASE + DRM_VMW_GB_SURFACE_CREATE,	\
+		 union drm_vmw_gb_surface_create_arg)
+#define DRM_IOCTL_VMW_GB_SURFACE_REF				\
+	DRM_IOWR(DRM_COMMAND_BASE + DRM_VMW_GB_SURFACE_REF,	\
+		 union drm_vmw_gb_surface_reference_arg)
+#define DRM_IOCTL_VMW_SYNCCPU					\
+	DRM_IOW(DRM_COMMAND_BASE + DRM_VMW_SYNCCPU,		\
+		 struct drm_vmw_synccpu_arg)
 
 /**
  * The core DRM version of this macro doesn't account for
@@ -177,6 +192,21 @@ static const struct drm_ioctl_desc vmw_ioctls[] = {
 	VMW_IOCTL_DEF(VMW_UPDATE_LAYOUT,
 		      vmw_kms_update_layout_ioctl,
 		      DRM_MASTER | DRM_UNLOCKED),
+	VMW_IOCTL_DEF(VMW_CREATE_SHADER,
+		      vmw_shader_define_ioctl,
+		      DRM_AUTH | DRM_UNLOCKED),
+	VMW_IOCTL_DEF(VMW_UNREF_SHADER,
+		      vmw_shader_destroy_ioctl,
+		      DRM_AUTH | DRM_UNLOCKED),
+	VMW_IOCTL_DEF(VMW_GB_SURFACE_CREATE,
+		      vmw_gb_surface_define_ioctl,
+		      DRM_AUTH | DRM_UNLOCKED),
+	VMW_IOCTL_DEF(VMW_GB_SURFACE_REF,
+		      vmw_gb_surface_reference_ioctl,
+		      DRM_AUTH | DRM_UNLOCKED),
+	VMW_IOCTL_DEF(VMW_SYNCCPU,
+		      vmw_user_dmabuf_synccpu_ioctl,
+		      DRM_AUTH | DRM_UNLOCKED),
 };
 
 static struct pci_device_id vmw_pci_id_list[] = {
@@ -189,6 +219,7 @@ static int enable_fbdev = IS_ENABLED(CONFIG_DRM_VMWGFX_FBCON);
 static int vmw_force_iommu;
 static int vmw_restrict_iommu;
 static int vmw_force_coherent;
+static int vmw_restrict_dma_mask;
 
 static int vmw_probe(struct pci_dev *, const struct pci_device_id *);
 static void vmw_master_init(struct vmw_master *);
@@ -203,6 +234,8 @@ MODULE_PARM_DESC(restrict_iommu, "Try to limit IOMMU usage for TTM pages");
 module_param_named(restrict_iommu, vmw_restrict_iommu, int, 0600);
 MODULE_PARM_DESC(force_coherent, "Force coherent TTM pages");
 module_param_named(force_coherent, vmw_force_coherent, int, 0600);
+MODULE_PARM_DESC(restrict_dma_mask, "Restrict DMA mask to 44 bits with IOMMU");
+module_param_named(restrict_dma_mask, vmw_restrict_dma_mask, int, 0600);
 
 
 static void vmw_print_capabilities(uint32_t capabilities)
@@ -240,38 +273,52 @@ static void vmw_print_capabilities(uint32_t capabilities)
 		DRM_INFO("  GMR2.\n");
 	if (capabilities & SVGA_CAP_SCREEN_OBJECT_2)
 		DRM_INFO("  Screen Object 2.\n");
+	if (capabilities & SVGA_CAP_COMMAND_BUFFERS)
+		DRM_INFO("  Command Buffers.\n");
+	if (capabilities & SVGA_CAP_CMD_BUFFERS_2)
+		DRM_INFO("  Command Buffers 2.\n");
+	if (capabilities & SVGA_CAP_GBOBJECTS)
+		DRM_INFO("  Guest Backed Resources.\n");
 }
 
-
 /**
- * vmw_execbuf_prepare_dummy_query - Initialize a query result structure at
- * the start of a buffer object.
+ * vmw_dummy_query_bo_create - create a bo to hold a dummy query result
  *
- * @dev_priv: The device private structure.
+ * @dev_priv: A device private structure.
  *
- * This function will idle the buffer using an uninterruptible wait, then
- * map the first page and initialize a pending occlusion query result structure,
- * Finally it will unmap the buffer.
+ * This function creates a small buffer object that holds the query
+ * result for dummy queries emitted as query barriers.
+ * The function will then map the first page and initialize a pending
+ * occlusion query result structure, Finally it will unmap the buffer.
+ * No interruptible waits are done within this function.
  *
- * TODO: Since we're only mapping a single page, we should optimize the map
- * to use kmap_atomic / iomap_atomic.
+ * Returns an error if bo creation or initialization fails.
  */
-static void vmw_dummy_query_bo_prepare(struct vmw_private *dev_priv)
+static int vmw_dummy_query_bo_create(struct vmw_private *dev_priv)
 {
+	int ret;
+	struct ttm_buffer_object *bo;
 	struct ttm_bo_kmap_obj map;
 	volatile SVGA3dQueryResult *result;
 	bool dummy;
-	int ret;
-	struct ttm_bo_device *bdev = &dev_priv->bdev;
-	struct ttm_buffer_object *bo = dev_priv->dummy_query_bo;
 
-	ttm_bo_reserve(bo, false, false, false, 0);
-	spin_lock(&bdev->fence_lock);
-	ret = ttm_bo_wait(bo, false, false, false);
-	spin_unlock(&bdev->fence_lock);
+	/*
+	 * Create the bo as pinned, so that a tryreserve will
+	 * immediately succeed. This is because we're the only
+	 * user of the bo currently.
+	 */
+	ret = ttm_bo_create(&dev_priv->bdev,
+			    PAGE_SIZE,
+			    ttm_bo_type_device,
+			    &vmw_sys_ne_placement,
+			    0, false, NULL,
+			    &bo);
+
 	if (unlikely(ret != 0))
-		(void) vmw_fallback_wait(dev_priv, false, true, 0, false,
-					 10*HZ);
+		return ret;
+
+	ret = ttm_bo_reserve(bo, false, true, false, 0);
+	BUG_ON(ret != 0);
 
 	ret = ttm_bo_kmap(bo, 0, 1, &map);
 	if (likely(ret == 0)) {
@@ -280,34 +327,19 @@ static void vmw_dummy_query_bo_prepare(struct vmw_private *dev_priv)
 		result->state = SVGA3D_QUERYSTATE_PENDING;
 		result->result32 = 0xff;
 		ttm_bo_kunmap(&map);
-	} else
-		DRM_ERROR("Dummy query buffer map failed.\n");
+	}
+	vmw_bo_pin(bo, false);
 	ttm_bo_unreserve(bo);
-}
 
+	if (unlikely(ret != 0)) {
+		DRM_ERROR("Dummy query buffer map failed.\n");
+		ttm_bo_unref(&bo);
+	} else
+		dev_priv->dummy_query_bo = bo;
 
-/**
- * vmw_dummy_query_bo_create - create a bo to hold a dummy query result
- *
- * @dev_priv: A device private structure.
- *
- * This function creates a small buffer object that holds the query
- * result for dummy queries emitted as query barriers.
- * No interruptible waits are done within this function.
- *
- * Returns an error if bo creation fails.
- */
-static int vmw_dummy_query_bo_create(struct vmw_private *dev_priv)
-{
-	return ttm_bo_create(&dev_priv->bdev,
-			     PAGE_SIZE,
-			     ttm_bo_type_device,
-			     &vmw_vram_sys_placement,
-			     0, false, NULL,
-			     &dev_priv->dummy_query_bo);
+	return ret;
 }
 
-
 static int vmw_request_device(struct vmw_private *dev_priv)
 {
 	int ret;
@@ -318,14 +350,24 @@ static int vmw_request_device(struct vmw_private *dev_priv)
 		return ret;
 	}
 	vmw_fence_fifo_up(dev_priv->fman);
+	if (dev_priv->has_mob) {
+		ret = vmw_otables_setup(dev_priv);
+		if (unlikely(ret != 0)) {
+			DRM_ERROR("Unable to initialize "
+				  "guest Memory OBjects.\n");
+			goto out_no_mob;
+		}
+	}
 	ret = vmw_dummy_query_bo_create(dev_priv);
 	if (unlikely(ret != 0))
 		goto out_no_query_bo;
-	vmw_dummy_query_bo_prepare(dev_priv);
 
 	return 0;
 
 out_no_query_bo:
+	if (dev_priv->has_mob)
+		vmw_otables_takedown(dev_priv);
+out_no_mob:
 	vmw_fence_fifo_down(dev_priv->fman);
 	vmw_fifo_release(dev_priv, &dev_priv->fifo);
 	return ret;
@@ -341,10 +383,13 @@ static void vmw_release_device(struct vmw_private *dev_priv)
 	BUG_ON(dev_priv->pinned_bo != NULL);
 
 	ttm_bo_unref(&dev_priv->dummy_query_bo);
+	if (dev_priv->has_mob)
+		vmw_otables_takedown(dev_priv);
 	vmw_fence_fifo_down(dev_priv->fman);
 	vmw_fifo_release(dev_priv, &dev_priv->fifo);
 }
 
+
 /**
  * Increase the 3d resource refcount.
  * If the count was prevously zero, initialize the fifo, switching to svga
@@ -510,6 +555,33 @@ out_fixup:
 	return 0;
 }
 
+/**
+ * vmw_dma_masks - set required page- and dma masks
+ *
+ * @dev: Pointer to struct drm-device
+ *
+ * With 32-bit we can only handle 32 bit PFNs. Optionally set that
+ * restriction also for 64-bit systems.
+ */
+#ifdef CONFIG_INTEL_IOMMU
+static int vmw_dma_masks(struct vmw_private *dev_priv)
+{
+	struct drm_device *dev = dev_priv->dev;
+
+	if (intel_iommu_enabled &&
+	    (sizeof(unsigned long) == 4 || vmw_restrict_dma_mask)) {
+		DRM_INFO("Restricting DMA addresses to 44 bits.\n");
+		return dma_set_mask(dev->dev, DMA_BIT_MASK(44));
+	}
+	return 0;
+}
+#else
+static int vmw_dma_masks(struct vmw_private *dev_priv)
+{
+	return 0;
+}
+#endif
+
 static int vmw_driver_load(struct drm_device *dev, unsigned long chipset)
 {
 	struct vmw_private *dev_priv;
@@ -532,6 +604,7 @@ static int vmw_driver_load(struct drm_device *dev, unsigned long chipset)
 	mutex_init(&dev_priv->hw_mutex);
 	mutex_init(&dev_priv->cmdbuf_mutex);
 	mutex_init(&dev_priv->release_mutex);
+	mutex_init(&dev_priv->binding_mutex);
 	rwlock_init(&dev_priv->resource_lock);
 
 	for (i = vmw_res_context; i < vmw_res_max; ++i) {
@@ -578,14 +651,9 @@ static int vmw_driver_load(struct drm_device *dev, unsigned long chipset)
 
 	vmw_get_initial_size(dev_priv);
 
-	if (dev_priv->capabilities & SVGA_CAP_GMR) {
-		dev_priv->max_gmr_descriptors =
-			vmw_read(dev_priv,
-				 SVGA_REG_GMR_MAX_DESCRIPTOR_LENGTH);
+	if (dev_priv->capabilities & SVGA_CAP_GMR2) {
 		dev_priv->max_gmr_ids =
 			vmw_read(dev_priv, SVGA_REG_GMR_MAX_IDS);
-	}
-	if (dev_priv->capabilities & SVGA_CAP_GMR2) {
 		dev_priv->max_gmr_pages =
 			vmw_read(dev_priv, SVGA_REG_GMRS_MAX_PAGES);
 		dev_priv->memory_size =
@@ -598,23 +666,42 @@ static int vmw_driver_load(struct drm_device *dev, unsigned long chipset)
 		 */
 		dev_priv->memory_size = 512*1024*1024;
 	}
+	dev_priv->max_mob_pages = 0;
+	if (dev_priv->capabilities & SVGA_CAP_GBOBJECTS) {
+		uint64_t mem_size =
+			vmw_read(dev_priv,
+				 SVGA_REG_SUGGESTED_GBOBJECT_MEM_SIZE_KB);
+
+		dev_priv->max_mob_pages = mem_size * 1024 / PAGE_SIZE;
+		dev_priv->prim_bb_mem =
+			vmw_read(dev_priv,
+				 SVGA_REG_MAX_PRIMARY_BOUNDING_BOX_MEM);
+	} else
+		dev_priv->prim_bb_mem = dev_priv->vram_size;
+
+	ret = vmw_dma_masks(dev_priv);
+	if (unlikely(ret != 0)) {
+		mutex_unlock(&dev_priv->hw_mutex);
+		goto out_err0;
+	}
+
+	if (unlikely(dev_priv->prim_bb_mem < dev_priv->vram_size))
+		dev_priv->prim_bb_mem = dev_priv->vram_size;
 
 	mutex_unlock(&dev_priv->hw_mutex);
 
 	vmw_print_capabilities(dev_priv->capabilities);
 
-	if (dev_priv->capabilities & SVGA_CAP_GMR) {
+	if (dev_priv->capabilities & SVGA_CAP_GMR2) {
 		DRM_INFO("Max GMR ids is %u\n",
 			 (unsigned)dev_priv->max_gmr_ids);
-		DRM_INFO("Max GMR descriptors is %u\n",
-			 (unsigned)dev_priv->max_gmr_descriptors);
-	}
-	if (dev_priv->capabilities & SVGA_CAP_GMR2) {
 		DRM_INFO("Max number of GMR pages is %u\n",
 			 (unsigned)dev_priv->max_gmr_pages);
 		DRM_INFO("Max dedicated hypervisor surface memory is %u kiB\n",
 			 (unsigned)dev_priv->memory_size / 1024);
 	}
+	DRM_INFO("Maximum display memory size is %u kiB\n",
+		 dev_priv->prim_bb_mem / 1024);
 	DRM_INFO("VRAM at 0x%08x size is %u kiB\n",
 		 dev_priv->vram_start, dev_priv->vram_size / 1024);
 	DRM_INFO("MMIO at 0x%08x size is %u kiB\n",
@@ -649,12 +736,22 @@ static int vmw_driver_load(struct drm_device *dev, unsigned long chipset)
 	dev_priv->has_gmr = true;
 	if (((dev_priv->capabilities & (SVGA_CAP_GMR | SVGA_CAP_GMR2)) == 0) ||
 	    refuse_dma || ttm_bo_init_mm(&dev_priv->bdev, VMW_PL_GMR,
-					 dev_priv->max_gmr_ids) != 0) {
+					 VMW_PL_GMR) != 0) {
 		DRM_INFO("No GMR memory available. "
 			 "Graphics memory resources are very limited.\n");
 		dev_priv->has_gmr = false;
 	}
 
+	if (dev_priv->capabilities & SVGA_CAP_GBOBJECTS) {
+		dev_priv->has_mob = true;
+		if (ttm_bo_init_mm(&dev_priv->bdev, VMW_PL_MOB,
+				   VMW_PL_MOB) != 0) {
+			DRM_INFO("No MOB memory available. "
+				 "3D will be disabled.\n");
+			dev_priv->has_mob = false;
+		}
+	}
+
 	dev_priv->mmio_mtrr = arch_phys_wc_add(dev_priv->mmio_start,
 					       dev_priv->mmio_size);
 
@@ -757,6 +854,8 @@ out_err4:
 	iounmap(dev_priv->mmio_virt);
 out_err3:
 	arch_phys_wc_del(dev_priv->mmio_mtrr);
+	if (dev_priv->has_mob)
+		(void) ttm_bo_clean_mm(&dev_priv->bdev, VMW_PL_MOB);
 	if (dev_priv->has_gmr)
 		(void) ttm_bo_clean_mm(&dev_priv->bdev, VMW_PL_GMR);
 	(void)ttm_bo_clean_mm(&dev_priv->bdev, TTM_PL_VRAM);
@@ -801,6 +900,8 @@ static int vmw_driver_unload(struct drm_device *dev)
 	ttm_object_device_release(&dev_priv->tdev);
 	iounmap(dev_priv->mmio_virt);
 	arch_phys_wc_del(dev_priv->mmio_mtrr);
+	if (dev_priv->has_mob)
+		(void) ttm_bo_clean_mm(&dev_priv->bdev, VMW_PL_MOB);
 	if (dev_priv->has_gmr)
 		(void)ttm_bo_clean_mm(&dev_priv->bdev, VMW_PL_GMR);
 	(void)ttm_bo_clean_mm(&dev_priv->bdev, TTM_PL_VRAM);
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
index 20890ad8408b..554e7fa33082 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
@@ -40,9 +40,9 @@
 #include <drm/ttm/ttm_module.h>
 #include "vmwgfx_fence.h"
 
-#define VMWGFX_DRIVER_DATE "20120209"
+#define VMWGFX_DRIVER_DATE "20121114"
 #define VMWGFX_DRIVER_MAJOR 2
-#define VMWGFX_DRIVER_MINOR 4
+#define VMWGFX_DRIVER_MINOR 5
 #define VMWGFX_DRIVER_PATCHLEVEL 0
 #define VMWGFX_FILE_PAGE_OFFSET 0x00100000
 #define VMWGFX_FIFO_STATIC_SIZE (1024*1024)
@@ -50,14 +50,30 @@
 #define VMWGFX_MAX_VALIDATIONS 2048
 #define VMWGFX_MAX_DISPLAYS 16
 #define VMWGFX_CMD_BOUNCE_INIT_SIZE 32768
+#define VMWGFX_ENABLE_SCREEN_TARGET_OTABLE 0
+
+/*
+ * Perhaps we should have sysfs entries for these.
+ */
+#define VMWGFX_NUM_GB_CONTEXT 256
+#define VMWGFX_NUM_GB_SHADER 20000
+#define VMWGFX_NUM_GB_SURFACE 32768
+#define VMWGFX_NUM_GB_SCREEN_TARGET VMWGFX_MAX_DISPLAYS
+#define VMWGFX_NUM_MOB (VMWGFX_NUM_GB_CONTEXT +\
+			VMWGFX_NUM_GB_SHADER +\
+			VMWGFX_NUM_GB_SURFACE +\
+			VMWGFX_NUM_GB_SCREEN_TARGET)
 
 #define VMW_PL_GMR TTM_PL_PRIV0
 #define VMW_PL_FLAG_GMR TTM_PL_FLAG_PRIV0
+#define VMW_PL_MOB TTM_PL_PRIV1
+#define VMW_PL_FLAG_MOB TTM_PL_FLAG_PRIV1
 
 #define VMW_RES_CONTEXT ttm_driver_type0
 #define VMW_RES_SURFACE ttm_driver_type1
 #define VMW_RES_STREAM ttm_driver_type2
 #define VMW_RES_FENCE ttm_driver_type3
+#define VMW_RES_SHADER ttm_driver_type4
 
 struct vmw_fpriv {
 	struct drm_master *locked_master;
@@ -82,6 +98,7 @@ struct vmw_dma_buffer {
 struct vmw_validate_buffer {
 	struct ttm_validate_buffer base;
 	struct drm_hash_item hash;
+	bool validate_as_mob;
 };
 
 struct vmw_res_func;
@@ -98,6 +115,7 @@ struct vmw_resource {
 	const struct vmw_res_func *func;
 	struct list_head lru_head; /* Protected by the resource lock */
 	struct list_head mob_head; /* Protected by @backup reserved */
+	struct list_head binding_head; /* Protected by binding_mutex */
 	void (*res_free) (struct vmw_resource *res);
 	void (*hw_destroy) (struct vmw_resource *res);
 };
@@ -106,6 +124,7 @@ enum vmw_res_type {
 	vmw_res_context,
 	vmw_res_surface,
 	vmw_res_stream,
+	vmw_res_shader,
 	vmw_res_max
 };
 
@@ -154,6 +173,7 @@ struct vmw_fifo_state {
 };
 
 struct vmw_relocation {
+	SVGAMobId *mob_loc;
 	SVGAGuestPtr *location;
 	uint32_t index;
 };
@@ -229,6 +249,71 @@ struct vmw_piter {
 	struct page *(*page)(struct vmw_piter *);
 };
 
+/*
+ * enum vmw_ctx_binding_type - abstract resource to context binding types
+ */
+enum vmw_ctx_binding_type {
+	vmw_ctx_binding_shader,
+	vmw_ctx_binding_rt,
+	vmw_ctx_binding_tex,
+	vmw_ctx_binding_max
+};
+
+/**
+ * struct vmw_ctx_bindinfo - structure representing a single context binding
+ *
+ * @ctx: Pointer to the context structure. NULL means the binding is not
+ * active.
+ * @res: Non ref-counted pointer to the bound resource.
+ * @bt: The binding type.
+ * @i1: Union of information needed to unbind.
+ */
+struct vmw_ctx_bindinfo {
+	struct vmw_resource *ctx;
+	struct vmw_resource *res;
+	enum vmw_ctx_binding_type bt;
+	union {
+		SVGA3dShaderType shader_type;
+		SVGA3dRenderTargetType rt_type;
+		uint32 texture_stage;
+	} i1;
+};
+
+/**
+ * struct vmw_ctx_binding - structure representing a single context binding
+ *                        - suitable for tracking in a context
+ *
+ * @ctx_list: List head for context.
+ * @res_list: List head for bound resource.
+ * @bi: Binding info
+ */
+struct vmw_ctx_binding {
+	struct list_head ctx_list;
+	struct list_head res_list;
+	struct vmw_ctx_bindinfo bi;
+};
+
+
+/**
+ * struct vmw_ctx_binding_state - context binding state
+ *
+ * @list: linked list of individual bindings.
+ * @render_targets: Render target bindings.
+ * @texture_units: Texture units/samplers bindings.
+ * @shaders: Shader bindings.
+ *
+ * Note that this structure also provides storage space for the individual
+ * struct vmw_ctx_binding objects, so that no dynamic allocation is needed
+ * for individual bindings.
+ *
+ */
+struct vmw_ctx_binding_state {
+	struct list_head list;
+	struct vmw_ctx_binding render_targets[SVGA3D_RT_MAX];
+	struct vmw_ctx_binding texture_units[SVGA3D_NUM_TEXTURE_UNITS];
+	struct vmw_ctx_binding shaders[SVGA3D_SHADERTYPE_MAX];
+};
+
 struct vmw_sw_context{
 	struct drm_open_hash res_ht;
 	bool res_ht_initialized;
@@ -250,6 +335,7 @@ struct vmw_sw_context{
 	struct vmw_resource *last_query_ctx;
 	bool needs_post_query_barrier;
 	struct vmw_resource *error_resource;
+	struct vmw_ctx_binding_state staged_bindings;
 };
 
 struct vmw_legacy_display;
@@ -281,6 +367,7 @@ struct vmw_private {
 	unsigned int io_start;
 	uint32_t vram_start;
 	uint32_t vram_size;
+	uint32_t prim_bb_mem;
 	uint32_t mmio_start;
 	uint32_t mmio_size;
 	uint32_t fb_max_width;
@@ -290,11 +377,12 @@ struct vmw_private {
 	__le32 __iomem *mmio_virt;
 	int mmio_mtrr;
 	uint32_t capabilities;
-	uint32_t max_gmr_descriptors;
 	uint32_t max_gmr_ids;
 	uint32_t max_gmr_pages;
+	uint32_t max_mob_pages;
 	uint32_t memory_size;
 	bool has_gmr;
+	bool has_mob;
 	struct mutex hw_mutex;
 
 	/*
@@ -370,6 +458,7 @@ struct vmw_private {
 
 	struct vmw_sw_context ctx;
 	struct mutex cmdbuf_mutex;
+	struct mutex binding_mutex;
 
 	/**
 	 * Operating mode.
@@ -415,6 +504,12 @@ struct vmw_private {
 	 * DMA mapping stuff.
 	 */
 	enum vmw_dma_map_mode map_mode;
+
+	/*
+	 * Guest Backed stuff
+	 */
+	struct ttm_buffer_object *otable_bo;
+	struct vmw_otable *otables;
 };
 
 static inline struct vmw_surface *vmw_res_to_srf(struct vmw_resource *res)
@@ -471,23 +566,12 @@ extern void vmw_gmr_unbind(struct vmw_private *dev_priv, int gmr_id);
  * Resource utilities - vmwgfx_resource.c
  */
 struct vmw_user_resource_conv;
-extern const struct vmw_user_resource_conv *user_surface_converter;
-extern const struct vmw_user_resource_conv *user_context_converter;
 
-extern struct vmw_resource *vmw_context_alloc(struct vmw_private *dev_priv);
 extern void vmw_resource_unreference(struct vmw_resource **p_res);
 extern struct vmw_resource *vmw_resource_reference(struct vmw_resource *res);
 extern int vmw_resource_validate(struct vmw_resource *res);
 extern int vmw_resource_reserve(struct vmw_resource *res, bool no_backup);
 extern bool vmw_resource_needs_backup(const struct vmw_resource *res);
-extern int vmw_context_destroy_ioctl(struct drm_device *dev, void *data,
-				     struct drm_file *file_priv);
-extern int vmw_context_define_ioctl(struct drm_device *dev, void *data,
-				    struct drm_file *file_priv);
-extern int vmw_context_check(struct vmw_private *dev_priv,
-			     struct ttm_object_file *tfile,
-			     int id,
-			     struct vmw_resource **p_res);
 extern int vmw_user_lookup_handle(struct vmw_private *dev_priv,
 				  struct ttm_object_file *tfile,
 				  uint32_t handle,
@@ -499,18 +583,6 @@ extern int vmw_user_resource_lookup_handle(
 	uint32_t handle,
 	const struct vmw_user_resource_conv *converter,
 	struct vmw_resource **p_res);
-extern void vmw_surface_res_free(struct vmw_resource *res);
-extern int vmw_surface_destroy_ioctl(struct drm_device *dev, void *data,
-				     struct drm_file *file_priv);
-extern int vmw_surface_define_ioctl(struct drm_device *dev, void *data,
-				    struct drm_file *file_priv);
-extern int vmw_surface_reference_ioctl(struct drm_device *dev, void *data,
-				       struct drm_file *file_priv);
-extern int vmw_surface_check(struct vmw_private *dev_priv,
-			     struct ttm_object_file *tfile,
-			     uint32_t handle, int *id);
-extern int vmw_surface_validate(struct vmw_private *dev_priv,
-				struct vmw_surface *srf);
 extern void vmw_dmabuf_bo_free(struct ttm_buffer_object *bo);
 extern int vmw_dmabuf_init(struct vmw_private *dev_priv,
 			   struct vmw_dma_buffer *vmw_bo,
@@ -519,10 +591,21 @@ extern int vmw_dmabuf_init(struct vmw_private *dev_priv,
 			   void (*bo_free) (struct ttm_buffer_object *bo));
 extern int vmw_user_dmabuf_verify_access(struct ttm_buffer_object *bo,
 				  struct ttm_object_file *tfile);
+extern int vmw_user_dmabuf_alloc(struct vmw_private *dev_priv,
+				 struct ttm_object_file *tfile,
+				 uint32_t size,
+				 bool shareable,
+				 uint32_t *handle,
+				 struct vmw_dma_buffer **p_dma_buf);
+extern int vmw_user_dmabuf_reference(struct ttm_object_file *tfile,
+				     struct vmw_dma_buffer *dma_buf,
+				     uint32_t *handle);
 extern int vmw_dmabuf_alloc_ioctl(struct drm_device *dev, void *data,
 				  struct drm_file *file_priv);
 extern int vmw_dmabuf_unref_ioctl(struct drm_device *dev, void *data,
 				  struct drm_file *file_priv);
+extern int vmw_user_dmabuf_synccpu_ioctl(struct drm_device *dev, void *data,
+					 struct drm_file *file_priv);
 extern uint32_t vmw_dmabuf_validate_node(struct ttm_buffer_object *bo,
 					 uint32_t cur_validate_node);
 extern void vmw_dmabuf_validate_clear(struct ttm_buffer_object *bo);
@@ -622,10 +705,16 @@ extern struct ttm_placement vmw_vram_sys_placement;
 extern struct ttm_placement vmw_vram_gmr_placement;
 extern struct ttm_placement vmw_vram_gmr_ne_placement;
 extern struct ttm_placement vmw_sys_placement;
+extern struct ttm_placement vmw_sys_ne_placement;
 extern struct ttm_placement vmw_evictable_placement;
 extern struct ttm_placement vmw_srf_placement;
+extern struct ttm_placement vmw_mob_placement;
 extern struct ttm_bo_driver vmw_bo_driver;
 extern int vmw_dma_quiescent(struct drm_device *dev);
+extern int vmw_bo_map_dma(struct ttm_buffer_object *bo);
+extern void vmw_bo_unmap_dma(struct ttm_buffer_object *bo);
+extern const struct vmw_sg_table *
+vmw_bo_sg_table(struct ttm_buffer_object *bo);
 extern void vmw_piter_start(struct vmw_piter *viter,
 			    const struct vmw_sg_table *vsgt,
 			    unsigned long p_offs);
@@ -701,7 +790,7 @@ extern void vmw_execbuf_copy_fence_user(struct vmw_private *dev_priv,
  * IRQs and wating - vmwgfx_irq.c
  */
 
-extern irqreturn_t vmw_irq_handler(DRM_IRQ_ARGS);
+extern irqreturn_t vmw_irq_handler(int irq, void *arg);
 extern int vmw_wait_seqno(struct vmw_private *dev_priv, bool lazy,
 			     uint32_t seqno, bool interruptible,
 			     unsigned long timeout);
@@ -832,6 +921,76 @@ extern int vmw_prime_handle_to_fd(struct drm_device *dev,
 				  uint32_t handle, uint32_t flags,
 				  int *prime_fd);
 
+/*
+ * MemoryOBject management -  vmwgfx_mob.c
+ */
+struct vmw_mob;
+extern int vmw_mob_bind(struct vmw_private *dev_priv, struct vmw_mob *mob,
+			const struct vmw_sg_table *vsgt,
+			unsigned long num_data_pages, int32_t mob_id);
+extern void vmw_mob_unbind(struct vmw_private *dev_priv,
+			   struct vmw_mob *mob);
+extern void vmw_mob_destroy(struct vmw_mob *mob);
+extern struct vmw_mob *vmw_mob_create(unsigned long data_pages);
+extern int vmw_otables_setup(struct vmw_private *dev_priv);
+extern void vmw_otables_takedown(struct vmw_private *dev_priv);
+
+/*
+ * Context management - vmwgfx_context.c
+ */
+
+extern const struct vmw_user_resource_conv *user_context_converter;
+
+extern struct vmw_resource *vmw_context_alloc(struct vmw_private *dev_priv);
+
+extern int vmw_context_check(struct vmw_private *dev_priv,
+			     struct ttm_object_file *tfile,
+			     int id,
+			     struct vmw_resource **p_res);
+extern int vmw_context_define_ioctl(struct drm_device *dev, void *data,
+				    struct drm_file *file_priv);
+extern int vmw_context_destroy_ioctl(struct drm_device *dev, void *data,
+				     struct drm_file *file_priv);
+extern int vmw_context_binding_add(struct vmw_ctx_binding_state *cbs,
+				   const struct vmw_ctx_bindinfo *ci);
+extern void
+vmw_context_binding_state_transfer(struct vmw_resource *res,
+				   struct vmw_ctx_binding_state *cbs);
+extern void vmw_context_binding_res_list_kill(struct list_head *head);
+
+/*
+ * Surface management - vmwgfx_surface.c
+ */
+
+extern const struct vmw_user_resource_conv *user_surface_converter;
+
+extern void vmw_surface_res_free(struct vmw_resource *res);
+extern int vmw_surface_destroy_ioctl(struct drm_device *dev, void *data,
+				     struct drm_file *file_priv);
+extern int vmw_surface_define_ioctl(struct drm_device *dev, void *data,
+				    struct drm_file *file_priv);
+extern int vmw_surface_reference_ioctl(struct drm_device *dev, void *data,
+				       struct drm_file *file_priv);
+extern int vmw_gb_surface_define_ioctl(struct drm_device *dev, void *data,
+				       struct drm_file *file_priv);
+extern int vmw_gb_surface_reference_ioctl(struct drm_device *dev, void *data,
+					  struct drm_file *file_priv);
+extern int vmw_surface_check(struct vmw_private *dev_priv,
+			     struct ttm_object_file *tfile,
+			     uint32_t handle, int *id);
+extern int vmw_surface_validate(struct vmw_private *dev_priv,
+				struct vmw_surface *srf);
+
+/*
+ * Shader management - vmwgfx_shader.c
+ */
+
+extern const struct vmw_user_resource_conv *user_shader_converter;
+
+extern int vmw_shader_define_ioctl(struct drm_device *dev, void *data,
+				   struct drm_file *file_priv);
+extern int vmw_shader_destroy_ioctl(struct drm_device *dev, void *data,
+				    struct drm_file *file_priv);
 
 /**
  * Inline helper functions
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c b/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c
index 599f6469a1eb..7a5f1eb55c5a 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c
@@ -54,6 +54,8 @@ struct vmw_resource_relocation {
  * @res: Ref-counted pointer to the resource.
  * @switch_backup: Boolean whether to switch backup buffer on unreserve.
  * @new_backup: Refcounted pointer to the new backup buffer.
+ * @staged_bindings: If @res is a context, tracks bindings set up during
+ * the command batch. Otherwise NULL.
  * @new_backup_offset: New backup buffer offset if @new_backup is non-NUll.
  * @first_usage: Set to true the first time the resource is referenced in
  * the command stream.
@@ -65,12 +67,32 @@ struct vmw_resource_val_node {
 	struct drm_hash_item hash;
 	struct vmw_resource *res;
 	struct vmw_dma_buffer *new_backup;
+	struct vmw_ctx_binding_state *staged_bindings;
 	unsigned long new_backup_offset;
 	bool first_usage;
 	bool no_buffer_needed;
 };
 
 /**
+ * struct vmw_cmd_entry - Describe a command for the verifier
+ *
+ * @user_allow: Whether allowed from the execbuf ioctl.
+ * @gb_disable: Whether disabled if guest-backed objects are available.
+ * @gb_enable: Whether enabled iff guest-backed objects are available.
+ */
+struct vmw_cmd_entry {
+	int (*func) (struct vmw_private *, struct vmw_sw_context *,
+		     SVGA3dCmdHeader *);
+	bool user_allow;
+	bool gb_disable;
+	bool gb_enable;
+};
+
+#define VMW_CMD_DEF(_cmd, _func, _user_allow, _gb_disable, _gb_enable)	\
+	[(_cmd) - SVGA_3D_CMD_BASE] = {(_func), (_user_allow),\
+				       (_gb_disable), (_gb_enable)}
+
+/**
  * vmw_resource_unreserve - unreserve resources previously reserved for
  * command submission.
  *
@@ -87,6 +109,16 @@ static void vmw_resource_list_unreserve(struct list_head *list,
 		struct vmw_dma_buffer *new_backup =
 			backoff ? NULL : val->new_backup;
 
+		/*
+		 * Transfer staged context bindings to the
+		 * persistent context binding tracker.
+		 */
+		if (unlikely(val->staged_bindings)) {
+			vmw_context_binding_state_transfer
+				(val->res, val->staged_bindings);
+			kfree(val->staged_bindings);
+			val->staged_bindings = NULL;
+		}
 		vmw_resource_unreserve(res, new_backup,
 			val->new_backup_offset);
 		vmw_dmabuf_unreference(&val->new_backup);
@@ -224,6 +256,7 @@ static int vmw_cmd_ok(struct vmw_private *dev_priv,
  *
  * @sw_context: The software context used for this command submission batch.
  * @bo: The buffer object to add.
+ * @validate_as_mob: Validate this buffer as a MOB.
  * @p_val_node: If non-NULL Will be updated with the validate node number
  * on return.
  *
@@ -232,6 +265,7 @@ static int vmw_cmd_ok(struct vmw_private *dev_priv,
  */
 static int vmw_bo_to_validate_list(struct vmw_sw_context *sw_context,
 				   struct ttm_buffer_object *bo,
+				   bool validate_as_mob,
 				   uint32_t *p_val_node)
 {
 	uint32_t val_node;
@@ -244,6 +278,10 @@ static int vmw_bo_to_validate_list(struct vmw_sw_context *sw_context,
 				    &hash) == 0)) {
 		vval_buf = container_of(hash, struct vmw_validate_buffer,
 					hash);
+		if (unlikely(vval_buf->validate_as_mob != validate_as_mob)) {
+			DRM_ERROR("Inconsistent buffer usage.\n");
+			return -EINVAL;
+		}
 		val_buf = &vval_buf->base;
 		val_node = vval_buf - sw_context->val_bufs;
 	} else {
@@ -266,6 +304,7 @@ static int vmw_bo_to_validate_list(struct vmw_sw_context *sw_context,
 		val_buf->bo = ttm_bo_reference(bo);
 		val_buf->reserved = 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;
@@ -302,7 +341,8 @@ static int vmw_resources_reserve(struct vmw_sw_context *sw_context)
 			struct ttm_buffer_object *bo = &res->backup->base;
 
 			ret = vmw_bo_to_validate_list
-				(sw_context, bo, NULL);
+				(sw_context, bo,
+				 vmw_resource_needs_backup(res), NULL);
 
 			if (unlikely(ret != 0))
 				return ret;
@@ -362,8 +402,15 @@ static int vmw_cmd_res_check(struct vmw_private *dev_priv,
 	struct vmw_resource_val_node *node;
 	int ret;
 
-	if (*id == SVGA3D_INVALID_ID)
+	if (*id == SVGA3D_INVALID_ID) {
+		if (p_val)
+			*p_val = NULL;
+		if (res_type == vmw_res_context) {
+			DRM_ERROR("Illegal context invalid id.\n");
+			return -EINVAL;
+		}
 		return 0;
+	}
 
 	/*
 	 * Fastpath in case of repeated commands referencing the same
@@ -411,6 +458,18 @@ static int vmw_cmd_res_check(struct vmw_private *dev_priv,
 	rcache->node = node;
 	if (p_val)
 		*p_val = node;
+
+	if (node->first_usage && res_type == vmw_res_context) {
+		node->staged_bindings =
+			kzalloc(sizeof(*node->staged_bindings), GFP_KERNEL);
+		if (node->staged_bindings == NULL) {
+			DRM_ERROR("Failed to allocate context binding "
+				  "information.\n");
+			goto out_no_reloc;
+		}
+		INIT_LIST_HEAD(&node->staged_bindings->list);
+	}
+
 	vmw_resource_unreference(&res);
 	return 0;
 
@@ -453,17 +512,35 @@ static int vmw_cmd_set_render_target_check(struct vmw_private *dev_priv,
 		SVGA3dCmdHeader header;
 		SVGA3dCmdSetRenderTarget body;
 	} *cmd;
+	struct vmw_resource_val_node *ctx_node;
+	struct vmw_resource_val_node *res_node;
 	int ret;
 
-	ret = vmw_cmd_cid_check(dev_priv, sw_context, header);
+	cmd = container_of(header, struct vmw_sid_cmd, header);
+
+	ret = vmw_cmd_res_check(dev_priv, sw_context, vmw_res_context,
+				user_context_converter, &cmd->body.cid,
+				&ctx_node);
 	if (unlikely(ret != 0))
 		return ret;
 
-	cmd = container_of(header, struct vmw_sid_cmd, header);
 	ret = vmw_cmd_res_check(dev_priv, sw_context, vmw_res_surface,
 				user_surface_converter,
-				&cmd->body.target.sid, NULL);
-	return ret;
+				&cmd->body.target.sid, &res_node);
+	if (unlikely(ret != 0))
+		return ret;
+
+	if (dev_priv->has_mob) {
+		struct vmw_ctx_bindinfo bi;
+
+		bi.ctx = ctx_node->res;
+		bi.res = res_node ? res_node->res : NULL;
+		bi.bt = vmw_ctx_binding_rt;
+		bi.i1.rt_type = cmd->body.type;
+		return vmw_context_binding_add(ctx_node->staged_bindings, &bi);
+	}
+
+	return 0;
 }
 
 static int vmw_cmd_surface_copy_check(struct vmw_private *dev_priv,
@@ -519,11 +596,6 @@ static int vmw_cmd_blt_surf_screen_check(struct vmw_private *dev_priv,
 
 	cmd = container_of(header, struct vmw_sid_cmd, header);
 
-	if (unlikely(!sw_context->kernel)) {
-		DRM_ERROR("Kernel only SVGA3d command: %u.\n", cmd->header.id);
-		return -EPERM;
-	}
-
 	return vmw_cmd_res_check(dev_priv, sw_context, vmw_res_surface,
 				 user_surface_converter,
 				 &cmd->body.srcImage.sid, NULL);
@@ -541,11 +613,6 @@ static int vmw_cmd_present_check(struct vmw_private *dev_priv,
 
 	cmd = container_of(header, struct vmw_sid_cmd, header);
 
-	if (unlikely(!sw_context->kernel)) {
-		DRM_ERROR("Kernel only SVGA3d command: %u.\n", cmd->header.id);
-		return -EPERM;
-	}
-
 	return vmw_cmd_res_check(dev_priv, sw_context, vmw_res_surface,
 				 user_surface_converter, &cmd->body.sid,
 				 NULL);
@@ -586,7 +653,7 @@ static int vmw_query_bo_switch_prepare(struct vmw_private *dev_priv,
 			sw_context->needs_post_query_barrier = true;
 			ret = vmw_bo_to_validate_list(sw_context,
 						      sw_context->cur_query_bo,
-						      NULL);
+						      dev_priv->has_mob, NULL);
 			if (unlikely(ret != 0))
 				return ret;
 		}
@@ -594,7 +661,7 @@ static int vmw_query_bo_switch_prepare(struct vmw_private *dev_priv,
 
 		ret = vmw_bo_to_validate_list(sw_context,
 					      dev_priv->dummy_query_bo,
-					      NULL);
+					      dev_priv->has_mob, NULL);
 		if (unlikely(ret != 0))
 			return ret;
 
@@ -672,6 +739,66 @@ static void vmw_query_bo_switch_commit(struct vmw_private *dev_priv,
 }
 
 /**
+ * vmw_translate_mob_pointer - Prepare to translate a user-space buffer
+ * handle to a MOB id.
+ *
+ * @dev_priv: Pointer to a device private structure.
+ * @sw_context: The software context used for this command batch validation.
+ * @id: Pointer to the user-space handle to be translated.
+ * @vmw_bo_p: Points to a location that, on successful return will carry
+ * a reference-counted pointer to the DMA buffer identified by the
+ * user-space handle in @id.
+ *
+ * This function saves information needed to translate a user-space buffer
+ * handle to a MOB id. The translation does not take place immediately, but
+ * during a call to vmw_apply_relocations(). This function builds a relocation
+ * list and a list of buffers to validate. The former needs to be freed using
+ * either vmw_apply_relocations() or vmw_free_relocations(). The latter
+ * needs to be freed using vmw_clear_validations.
+ */
+static int vmw_translate_mob_ptr(struct vmw_private *dev_priv,
+				 struct vmw_sw_context *sw_context,
+				 SVGAMobId *id,
+				 struct vmw_dma_buffer **vmw_bo_p)
+{
+	struct vmw_dma_buffer *vmw_bo = NULL;
+	struct ttm_buffer_object *bo;
+	uint32_t handle = *id;
+	struct vmw_relocation *reloc;
+	int ret;
+
+	ret = vmw_user_dmabuf_lookup(sw_context->tfile, handle, &vmw_bo);
+	if (unlikely(ret != 0)) {
+		DRM_ERROR("Could not find or use MOB buffer.\n");
+		return -EINVAL;
+	}
+	bo = &vmw_bo->base;
+
+	if (unlikely(sw_context->cur_reloc >= VMWGFX_MAX_RELOCATIONS)) {
+		DRM_ERROR("Max number relocations per submission"
+			  " exceeded\n");
+		ret = -EINVAL;
+		goto out_no_reloc;
+	}
+
+	reloc = &sw_context->relocs[sw_context->cur_reloc++];
+	reloc->mob_loc = id;
+	reloc->location = NULL;
+
+	ret = vmw_bo_to_validate_list(sw_context, bo, true, &reloc->index);
+	if (unlikely(ret != 0))
+		goto out_no_reloc;
+
+	*vmw_bo_p = vmw_bo;
+	return 0;
+
+out_no_reloc:
+	vmw_dmabuf_unreference(&vmw_bo);
+	vmw_bo_p = NULL;
+	return ret;
+}
+
+/**
  * vmw_translate_guest_pointer - Prepare to translate a user-space buffer
  * handle to a valid SVGAGuestPtr
  *
@@ -718,7 +845,7 @@ static int vmw_translate_guest_ptr(struct vmw_private *dev_priv,
 	reloc = &sw_context->relocs[sw_context->cur_reloc++];
 	reloc->location = ptr;
 
-	ret = vmw_bo_to_validate_list(sw_context, bo, &reloc->index);
+	ret = vmw_bo_to_validate_list(sw_context, bo, false, &reloc->index);
 	if (unlikely(ret != 0))
 		goto out_no_reloc;
 
@@ -732,6 +859,30 @@ out_no_reloc:
 }
 
 /**
+ * vmw_cmd_begin_gb_query - validate a  SVGA_3D_CMD_BEGIN_GB_QUERY command.
+ *
+ * @dev_priv: Pointer to a device private struct.
+ * @sw_context: The software context used for this command submission.
+ * @header: Pointer to the command header in the command stream.
+ */
+static int vmw_cmd_begin_gb_query(struct vmw_private *dev_priv,
+				  struct vmw_sw_context *sw_context,
+				  SVGA3dCmdHeader *header)
+{
+	struct vmw_begin_gb_query_cmd {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdBeginGBQuery q;
+	} *cmd;
+
+	cmd = container_of(header, struct vmw_begin_gb_query_cmd,
+			   header);
+
+	return vmw_cmd_res_check(dev_priv, sw_context, vmw_res_context,
+				 user_context_converter, &cmd->q.cid,
+				 NULL);
+}
+
+/**
  * vmw_cmd_begin_query - validate a  SVGA_3D_CMD_BEGIN_QUERY command.
  *
  * @dev_priv: Pointer to a device private struct.
@@ -750,12 +901,64 @@ static int vmw_cmd_begin_query(struct vmw_private *dev_priv,
 	cmd = container_of(header, struct vmw_begin_query_cmd,
 			   header);
 
+	if (unlikely(dev_priv->has_mob)) {
+		struct {
+			SVGA3dCmdHeader header;
+			SVGA3dCmdBeginGBQuery q;
+		} gb_cmd;
+
+		BUG_ON(sizeof(gb_cmd) != sizeof(*cmd));
+
+		gb_cmd.header.id = SVGA_3D_CMD_BEGIN_GB_QUERY;
+		gb_cmd.header.size = cmd->header.size;
+		gb_cmd.q.cid = cmd->q.cid;
+		gb_cmd.q.type = cmd->q.type;
+
+		memcpy(cmd, &gb_cmd, sizeof(*cmd));
+		return vmw_cmd_begin_gb_query(dev_priv, sw_context, header);
+	}
+
 	return vmw_cmd_res_check(dev_priv, sw_context, vmw_res_context,
 				 user_context_converter, &cmd->q.cid,
 				 NULL);
 }
 
 /**
+ * vmw_cmd_end_gb_query - validate a  SVGA_3D_CMD_END_GB_QUERY command.
+ *
+ * @dev_priv: Pointer to a device private struct.
+ * @sw_context: The software context used for this command submission.
+ * @header: Pointer to the command header in the command stream.
+ */
+static int vmw_cmd_end_gb_query(struct vmw_private *dev_priv,
+				struct vmw_sw_context *sw_context,
+				SVGA3dCmdHeader *header)
+{
+	struct vmw_dma_buffer *vmw_bo;
+	struct vmw_query_cmd {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdEndGBQuery q;
+	} *cmd;
+	int ret;
+
+	cmd = container_of(header, struct vmw_query_cmd, header);
+	ret = vmw_cmd_cid_check(dev_priv, sw_context, header);
+	if (unlikely(ret != 0))
+		return ret;
+
+	ret = vmw_translate_mob_ptr(dev_priv, sw_context,
+				    &cmd->q.mobid,
+				    &vmw_bo);
+	if (unlikely(ret != 0))
+		return ret;
+
+	ret = vmw_query_bo_switch_prepare(dev_priv, &vmw_bo->base, sw_context);
+
+	vmw_dmabuf_unreference(&vmw_bo);
+	return ret;
+}
+
+/**
  * vmw_cmd_end_query - validate a  SVGA_3D_CMD_END_QUERY command.
  *
  * @dev_priv: Pointer to a device private struct.
@@ -774,6 +977,25 @@ static int vmw_cmd_end_query(struct vmw_private *dev_priv,
 	int ret;
 
 	cmd = container_of(header, struct vmw_query_cmd, header);
+	if (dev_priv->has_mob) {
+		struct {
+			SVGA3dCmdHeader header;
+			SVGA3dCmdEndGBQuery q;
+		} gb_cmd;
+
+		BUG_ON(sizeof(gb_cmd) != sizeof(*cmd));
+
+		gb_cmd.header.id = SVGA_3D_CMD_END_GB_QUERY;
+		gb_cmd.header.size = cmd->header.size;
+		gb_cmd.q.cid = cmd->q.cid;
+		gb_cmd.q.type = cmd->q.type;
+		gb_cmd.q.mobid = cmd->q.guestResult.gmrId;
+		gb_cmd.q.offset = cmd->q.guestResult.offset;
+
+		memcpy(cmd, &gb_cmd, sizeof(*cmd));
+		return vmw_cmd_end_gb_query(dev_priv, sw_context, header);
+	}
+
 	ret = vmw_cmd_cid_check(dev_priv, sw_context, header);
 	if (unlikely(ret != 0))
 		return ret;
@@ -790,7 +1012,40 @@ static int vmw_cmd_end_query(struct vmw_private *dev_priv,
 	return ret;
 }
 
-/*
+/**
+ * vmw_cmd_wait_gb_query - validate a  SVGA_3D_CMD_WAIT_GB_QUERY command.
+ *
+ * @dev_priv: Pointer to a device private struct.
+ * @sw_context: The software context used for this command submission.
+ * @header: Pointer to the command header in the command stream.
+ */
+static int vmw_cmd_wait_gb_query(struct vmw_private *dev_priv,
+				 struct vmw_sw_context *sw_context,
+				 SVGA3dCmdHeader *header)
+{
+	struct vmw_dma_buffer *vmw_bo;
+	struct vmw_query_cmd {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdWaitForGBQuery q;
+	} *cmd;
+	int ret;
+
+	cmd = container_of(header, struct vmw_query_cmd, header);
+	ret = vmw_cmd_cid_check(dev_priv, sw_context, header);
+	if (unlikely(ret != 0))
+		return ret;
+
+	ret = vmw_translate_mob_ptr(dev_priv, sw_context,
+				    &cmd->q.mobid,
+				    &vmw_bo);
+	if (unlikely(ret != 0))
+		return ret;
+
+	vmw_dmabuf_unreference(&vmw_bo);
+	return 0;
+}
+
+/**
  * vmw_cmd_wait_query - validate a  SVGA_3D_CMD_WAIT_QUERY command.
  *
  * @dev_priv: Pointer to a device private struct.
@@ -809,6 +1064,25 @@ static int vmw_cmd_wait_query(struct vmw_private *dev_priv,
 	int ret;
 
 	cmd = container_of(header, struct vmw_query_cmd, header);
+	if (dev_priv->has_mob) {
+		struct {
+			SVGA3dCmdHeader header;
+			SVGA3dCmdWaitForGBQuery q;
+		} gb_cmd;
+
+		BUG_ON(sizeof(gb_cmd) != sizeof(*cmd));
+
+		gb_cmd.header.id = SVGA_3D_CMD_WAIT_FOR_GB_QUERY;
+		gb_cmd.header.size = cmd->header.size;
+		gb_cmd.q.cid = cmd->q.cid;
+		gb_cmd.q.type = cmd->q.type;
+		gb_cmd.q.mobid = cmd->q.guestResult.gmrId;
+		gb_cmd.q.offset = cmd->q.guestResult.offset;
+
+		memcpy(cmd, &gb_cmd, sizeof(*cmd));
+		return vmw_cmd_wait_gb_query(dev_priv, sw_context, header);
+	}
+
 	ret = vmw_cmd_cid_check(dev_priv, sw_context, header);
 	if (unlikely(ret != 0))
 		return ret;
@@ -921,15 +1195,22 @@ static int vmw_cmd_tex_state(struct vmw_private *dev_priv,
 	struct vmw_tex_state_cmd {
 		SVGA3dCmdHeader header;
 		SVGA3dCmdSetTextureState state;
-	};
+	} *cmd;
 
 	SVGA3dTextureState *last_state = (SVGA3dTextureState *)
 	  ((unsigned long) header + header->size + sizeof(header));
 	SVGA3dTextureState *cur_state = (SVGA3dTextureState *)
 		((unsigned long) header + sizeof(struct vmw_tex_state_cmd));
+	struct vmw_resource_val_node *ctx_node;
+	struct vmw_resource_val_node *res_node;
 	int ret;
 
-	ret = vmw_cmd_cid_check(dev_priv, sw_context, header);
+	cmd = container_of(header, struct vmw_tex_state_cmd,
+			   header);
+
+	ret = vmw_cmd_res_check(dev_priv, sw_context, vmw_res_context,
+				user_context_converter, &cmd->state.cid,
+				&ctx_node);
 	if (unlikely(ret != 0))
 		return ret;
 
@@ -939,9 +1220,20 @@ static int vmw_cmd_tex_state(struct vmw_private *dev_priv,
 
 		ret = vmw_cmd_res_check(dev_priv, sw_context, vmw_res_surface,
 					user_surface_converter,
-					&cur_state->value, NULL);
+					&cur_state->value, &res_node);
 		if (unlikely(ret != 0))
 			return ret;
+
+		if (dev_priv->has_mob) {
+			struct vmw_ctx_bindinfo bi;
+
+			bi.ctx = ctx_node->res;
+			bi.res = res_node ? res_node->res : NULL;
+			bi.bt = vmw_ctx_binding_tex;
+			bi.i1.texture_stage = cur_state->stage;
+			vmw_context_binding_add(ctx_node->staged_bindings,
+						&bi);
+		}
 	}
 
 	return 0;
@@ -971,6 +1263,222 @@ static int vmw_cmd_check_define_gmrfb(struct vmw_private *dev_priv,
 }
 
 /**
+ * vmw_cmd_switch_backup - Utility function to handle backup buffer switching
+ *
+ * @dev_priv: Pointer to a device private struct.
+ * @sw_context: The software context being used for this batch.
+ * @res_type: The resource type.
+ * @converter: Information about user-space binding for this resource type.
+ * @res_id: Pointer to the user-space resource handle in the command stream.
+ * @buf_id: Pointer to the user-space backup buffer handle in the command
+ * stream.
+ * @backup_offset: Offset of backup into MOB.
+ *
+ * This function prepares for registering a switch of backup buffers
+ * in the resource metadata just prior to unreserving.
+ */
+static int vmw_cmd_switch_backup(struct vmw_private *dev_priv,
+				 struct vmw_sw_context *sw_context,
+				 enum vmw_res_type res_type,
+				 const struct vmw_user_resource_conv
+				 *converter,
+				 uint32_t *res_id,
+				 uint32_t *buf_id,
+				 unsigned long backup_offset)
+{
+	int ret;
+	struct vmw_dma_buffer *dma_buf;
+	struct vmw_resource_val_node *val_node;
+
+	ret = vmw_cmd_res_check(dev_priv, sw_context, res_type,
+				converter, res_id, &val_node);
+	if (unlikely(ret != 0))
+		return ret;
+
+	ret = vmw_translate_mob_ptr(dev_priv, sw_context, buf_id, &dma_buf);
+	if (unlikely(ret != 0))
+		return ret;
+
+	if (val_node->first_usage)
+		val_node->no_buffer_needed = true;
+
+	vmw_dmabuf_unreference(&val_node->new_backup);
+	val_node->new_backup = dma_buf;
+	val_node->new_backup_offset = backup_offset;
+
+	return 0;
+}
+
+/**
+ * vmw_cmd_bind_gb_surface - Validate an SVGA_3D_CMD_BIND_GB_SURFACE
+ * command
+ *
+ * @dev_priv: Pointer to a device private struct.
+ * @sw_context: The software context being used for this batch.
+ * @header: Pointer to the command header in the command stream.
+ */
+static int vmw_cmd_bind_gb_surface(struct vmw_private *dev_priv,
+				   struct vmw_sw_context *sw_context,
+				   SVGA3dCmdHeader *header)
+{
+	struct vmw_bind_gb_surface_cmd {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdBindGBSurface body;
+	} *cmd;
+
+	cmd = container_of(header, struct vmw_bind_gb_surface_cmd, header);
+
+	return vmw_cmd_switch_backup(dev_priv, sw_context, vmw_res_surface,
+				     user_surface_converter,
+				     &cmd->body.sid, &cmd->body.mobid,
+				     0);
+}
+
+/**
+ * vmw_cmd_update_gb_image - Validate an SVGA_3D_CMD_UPDATE_GB_IMAGE
+ * command
+ *
+ * @dev_priv: Pointer to a device private struct.
+ * @sw_context: The software context being used for this batch.
+ * @header: Pointer to the command header in the command stream.
+ */
+static int vmw_cmd_update_gb_image(struct vmw_private *dev_priv,
+				   struct vmw_sw_context *sw_context,
+				   SVGA3dCmdHeader *header)
+{
+	struct vmw_gb_surface_cmd {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdUpdateGBImage body;
+	} *cmd;
+
+	cmd = container_of(header, struct vmw_gb_surface_cmd, header);
+
+	return vmw_cmd_res_check(dev_priv, sw_context, vmw_res_surface,
+				 user_surface_converter,
+				 &cmd->body.image.sid, NULL);
+}
+
+/**
+ * vmw_cmd_update_gb_surface - Validate an SVGA_3D_CMD_UPDATE_GB_SURFACE
+ * command
+ *
+ * @dev_priv: Pointer to a device private struct.
+ * @sw_context: The software context being used for this batch.
+ * @header: Pointer to the command header in the command stream.
+ */
+static int vmw_cmd_update_gb_surface(struct vmw_private *dev_priv,
+				     struct vmw_sw_context *sw_context,
+				     SVGA3dCmdHeader *header)
+{
+	struct vmw_gb_surface_cmd {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdUpdateGBSurface body;
+	} *cmd;
+
+	cmd = container_of(header, struct vmw_gb_surface_cmd, header);
+
+	return vmw_cmd_res_check(dev_priv, sw_context, vmw_res_surface,
+				 user_surface_converter,
+				 &cmd->body.sid, NULL);
+}
+
+/**
+ * vmw_cmd_readback_gb_image - Validate an SVGA_3D_CMD_READBACK_GB_IMAGE
+ * command
+ *
+ * @dev_priv: Pointer to a device private struct.
+ * @sw_context: The software context being used for this batch.
+ * @header: Pointer to the command header in the command stream.
+ */
+static int vmw_cmd_readback_gb_image(struct vmw_private *dev_priv,
+				     struct vmw_sw_context *sw_context,
+				     SVGA3dCmdHeader *header)
+{
+	struct vmw_gb_surface_cmd {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdReadbackGBImage body;
+	} *cmd;
+
+	cmd = container_of(header, struct vmw_gb_surface_cmd, header);
+
+	return vmw_cmd_res_check(dev_priv, sw_context, vmw_res_surface,
+				 user_surface_converter,
+				 &cmd->body.image.sid, NULL);
+}
+
+/**
+ * vmw_cmd_readback_gb_surface - Validate an SVGA_3D_CMD_READBACK_GB_SURFACE
+ * command
+ *
+ * @dev_priv: Pointer to a device private struct.
+ * @sw_context: The software context being used for this batch.
+ * @header: Pointer to the command header in the command stream.
+ */
+static int vmw_cmd_readback_gb_surface(struct vmw_private *dev_priv,
+				       struct vmw_sw_context *sw_context,
+				       SVGA3dCmdHeader *header)
+{
+	struct vmw_gb_surface_cmd {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdReadbackGBSurface body;
+	} *cmd;
+
+	cmd = container_of(header, struct vmw_gb_surface_cmd, header);
+
+	return vmw_cmd_res_check(dev_priv, sw_context, vmw_res_surface,
+				 user_surface_converter,
+				 &cmd->body.sid, NULL);
+}
+
+/**
+ * vmw_cmd_invalidate_gb_image - Validate an SVGA_3D_CMD_INVALIDATE_GB_IMAGE
+ * command
+ *
+ * @dev_priv: Pointer to a device private struct.
+ * @sw_context: The software context being used for this batch.
+ * @header: Pointer to the command header in the command stream.
+ */
+static int vmw_cmd_invalidate_gb_image(struct vmw_private *dev_priv,
+				       struct vmw_sw_context *sw_context,
+				       SVGA3dCmdHeader *header)
+{
+	struct vmw_gb_surface_cmd {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdInvalidateGBImage body;
+	} *cmd;
+
+	cmd = container_of(header, struct vmw_gb_surface_cmd, header);
+
+	return vmw_cmd_res_check(dev_priv, sw_context, vmw_res_surface,
+				 user_surface_converter,
+				 &cmd->body.image.sid, NULL);
+}
+
+/**
+ * vmw_cmd_invalidate_gb_surface - Validate an
+ * SVGA_3D_CMD_INVALIDATE_GB_SURFACE command
+ *
+ * @dev_priv: Pointer to a device private struct.
+ * @sw_context: The software context being used for this batch.
+ * @header: Pointer to the command header in the command stream.
+ */
+static int vmw_cmd_invalidate_gb_surface(struct vmw_private *dev_priv,
+					 struct vmw_sw_context *sw_context,
+					 SVGA3dCmdHeader *header)
+{
+	struct vmw_gb_surface_cmd {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdInvalidateGBSurface body;
+	} *cmd;
+
+	cmd = container_of(header, struct vmw_gb_surface_cmd, header);
+
+	return vmw_cmd_res_check(dev_priv, sw_context, vmw_res_surface,
+				 user_surface_converter,
+				 &cmd->body.sid, NULL);
+}
+
+/**
  * vmw_cmd_set_shader - Validate an SVGA_3D_CMD_SET_SHADER
  * command
  *
@@ -986,18 +1494,64 @@ static int vmw_cmd_set_shader(struct vmw_private *dev_priv,
 		SVGA3dCmdHeader header;
 		SVGA3dCmdSetShader body;
 	} *cmd;
+	struct vmw_resource_val_node *ctx_node;
 	int ret;
 
 	cmd = container_of(header, struct vmw_set_shader_cmd,
 			   header);
 
-	ret = vmw_cmd_cid_check(dev_priv, sw_context, header);
+	ret = vmw_cmd_res_check(dev_priv, sw_context, vmw_res_context,
+				user_context_converter, &cmd->body.cid,
+				&ctx_node);
 	if (unlikely(ret != 0))
 		return ret;
 
+	if (dev_priv->has_mob) {
+		struct vmw_ctx_bindinfo bi;
+		struct vmw_resource_val_node *res_node;
+
+		ret = vmw_cmd_res_check(dev_priv, sw_context, vmw_res_shader,
+					user_shader_converter,
+					&cmd->body.shid, &res_node);
+		if (unlikely(ret != 0))
+			return ret;
+
+		bi.ctx = ctx_node->res;
+		bi.res = res_node ? res_node->res : NULL;
+		bi.bt = vmw_ctx_binding_shader;
+		bi.i1.shader_type = cmd->body.type;
+		return vmw_context_binding_add(ctx_node->staged_bindings, &bi);
+	}
+
 	return 0;
 }
 
+/**
+ * vmw_cmd_bind_gb_shader - Validate an SVGA_3D_CMD_BIND_GB_SHADER
+ * command
+ *
+ * @dev_priv: Pointer to a device private struct.
+ * @sw_context: The software context being used for this batch.
+ * @header: Pointer to the command header in the command stream.
+ */
+static int vmw_cmd_bind_gb_shader(struct vmw_private *dev_priv,
+				  struct vmw_sw_context *sw_context,
+				  SVGA3dCmdHeader *header)
+{
+	struct vmw_bind_gb_shader_cmd {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdBindGBShader body;
+	} *cmd;
+
+	cmd = container_of(header, struct vmw_bind_gb_shader_cmd,
+			   header);
+
+	return vmw_cmd_switch_backup(dev_priv, sw_context, vmw_res_shader,
+				     user_shader_converter,
+				     &cmd->body.shid, &cmd->body.mobid,
+				     cmd->body.offsetInBytes);
+}
+
 static int vmw_cmd_check_not_3d(struct vmw_private *dev_priv,
 				struct vmw_sw_context *sw_context,
 				void *buf, uint32_t *size)
@@ -1041,50 +1595,173 @@ static int vmw_cmd_check_not_3d(struct vmw_private *dev_priv,
 	return 0;
 }
 
-typedef int (*vmw_cmd_func) (struct vmw_private *,
-			     struct vmw_sw_context *,
-			     SVGA3dCmdHeader *);
-
-#define VMW_CMD_DEF(cmd, func) \
-	[cmd - SVGA_3D_CMD_BASE] = func
-
-static vmw_cmd_func vmw_cmd_funcs[SVGA_3D_CMD_MAX] = {
-	VMW_CMD_DEF(SVGA_3D_CMD_SURFACE_DEFINE, &vmw_cmd_invalid),
-	VMW_CMD_DEF(SVGA_3D_CMD_SURFACE_DESTROY, &vmw_cmd_invalid),
-	VMW_CMD_DEF(SVGA_3D_CMD_SURFACE_COPY, &vmw_cmd_surface_copy_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_SURFACE_STRETCHBLT, &vmw_cmd_stretch_blt_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_SURFACE_DMA, &vmw_cmd_dma),
-	VMW_CMD_DEF(SVGA_3D_CMD_CONTEXT_DEFINE, &vmw_cmd_invalid),
-	VMW_CMD_DEF(SVGA_3D_CMD_CONTEXT_DESTROY, &vmw_cmd_invalid),
-	VMW_CMD_DEF(SVGA_3D_CMD_SETTRANSFORM, &vmw_cmd_cid_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_SETZRANGE, &vmw_cmd_cid_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_SETRENDERSTATE, &vmw_cmd_cid_check),
+static const struct vmw_cmd_entry const vmw_cmd_entries[SVGA_3D_CMD_MAX] = {
+	VMW_CMD_DEF(SVGA_3D_CMD_SURFACE_DEFINE, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SURFACE_DESTROY, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SURFACE_COPY, &vmw_cmd_surface_copy_check,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SURFACE_STRETCHBLT, &vmw_cmd_stretch_blt_check,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SURFACE_DMA, &vmw_cmd_dma,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_CONTEXT_DEFINE, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_CONTEXT_DESTROY, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SETTRANSFORM, &vmw_cmd_cid_check,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SETZRANGE, &vmw_cmd_cid_check,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SETRENDERSTATE, &vmw_cmd_cid_check,
+		    true, false, false),
 	VMW_CMD_DEF(SVGA_3D_CMD_SETRENDERTARGET,
-		    &vmw_cmd_set_render_target_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_SETTEXTURESTATE, &vmw_cmd_tex_state),
-	VMW_CMD_DEF(SVGA_3D_CMD_SETMATERIAL, &vmw_cmd_cid_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_SETLIGHTDATA, &vmw_cmd_cid_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_SETLIGHTENABLED, &vmw_cmd_cid_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_SETVIEWPORT, &vmw_cmd_cid_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_SETCLIPPLANE, &vmw_cmd_cid_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_CLEAR, &vmw_cmd_cid_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_PRESENT, &vmw_cmd_present_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_SHADER_DEFINE, &vmw_cmd_cid_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_SHADER_DESTROY, &vmw_cmd_cid_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_SET_SHADER, &vmw_cmd_set_shader),
-	VMW_CMD_DEF(SVGA_3D_CMD_SET_SHADER_CONST, &vmw_cmd_cid_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_DRAW_PRIMITIVES, &vmw_cmd_draw),
-	VMW_CMD_DEF(SVGA_3D_CMD_SETSCISSORRECT, &vmw_cmd_cid_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_BEGIN_QUERY, &vmw_cmd_begin_query),
-	VMW_CMD_DEF(SVGA_3D_CMD_END_QUERY, &vmw_cmd_end_query),
-	VMW_CMD_DEF(SVGA_3D_CMD_WAIT_FOR_QUERY, &vmw_cmd_wait_query),
-	VMW_CMD_DEF(SVGA_3D_CMD_PRESENT_READBACK, &vmw_cmd_ok),
+		    &vmw_cmd_set_render_target_check, true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SETTEXTURESTATE, &vmw_cmd_tex_state,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SETMATERIAL, &vmw_cmd_cid_check,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SETLIGHTDATA, &vmw_cmd_cid_check,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SETLIGHTENABLED, &vmw_cmd_cid_check,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SETVIEWPORT, &vmw_cmd_cid_check,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SETCLIPPLANE, &vmw_cmd_cid_check,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_CLEAR, &vmw_cmd_cid_check,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_PRESENT, &vmw_cmd_present_check,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SHADER_DEFINE, &vmw_cmd_cid_check,
+		    true, true, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SHADER_DESTROY, &vmw_cmd_cid_check,
+		    true, true, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SET_SHADER, &vmw_cmd_set_shader,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SET_SHADER_CONST, &vmw_cmd_cid_check,
+		    true, true, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_DRAW_PRIMITIVES, &vmw_cmd_draw,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SETSCISSORRECT, &vmw_cmd_cid_check,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_BEGIN_QUERY, &vmw_cmd_begin_query,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_END_QUERY, &vmw_cmd_end_query,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_WAIT_FOR_QUERY, &vmw_cmd_wait_query,
+		    true, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_PRESENT_READBACK, &vmw_cmd_ok,
+		    true, false, false),
 	VMW_CMD_DEF(SVGA_3D_CMD_BLIT_SURFACE_TO_SCREEN,
-		    &vmw_cmd_blt_surf_screen_check),
-	VMW_CMD_DEF(SVGA_3D_CMD_SURFACE_DEFINE_V2, &vmw_cmd_invalid),
-	VMW_CMD_DEF(SVGA_3D_CMD_GENERATE_MIPMAPS, &vmw_cmd_invalid),
-	VMW_CMD_DEF(SVGA_3D_CMD_ACTIVATE_SURFACE, &vmw_cmd_invalid),
-	VMW_CMD_DEF(SVGA_3D_CMD_DEACTIVATE_SURFACE, &vmw_cmd_invalid),
+		    &vmw_cmd_blt_surf_screen_check, false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SURFACE_DEFINE_V2, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_GENERATE_MIPMAPS, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_ACTIVATE_SURFACE, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_DEACTIVATE_SURFACE, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SCREEN_DMA, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SET_UNITY_SURFACE_COOKIE, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_OPEN_CONTEXT_SURFACE, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_LOGICOPS_BITBLT, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_LOGICOPS_TRANSBLT, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_LOGICOPS_STRETCHBLT, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_LOGICOPS_COLORFILL, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_LOGICOPS_ALPHABLEND, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_LOGICOPS_CLEARTYPEBLEND, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_SET_OTABLE_BASE, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_READBACK_OTABLE, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_DEFINE_GB_MOB, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_DESTROY_GB_MOB, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_REDEFINE_GB_MOB, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_UPDATE_GB_MOB_MAPPING, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_DEFINE_GB_SURFACE, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_DESTROY_GB_SURFACE, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_BIND_GB_SURFACE, &vmw_cmd_bind_gb_surface,
+		    true, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_COND_BIND_GB_SURFACE, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_UPDATE_GB_IMAGE, &vmw_cmd_update_gb_image,
+		    true, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_UPDATE_GB_SURFACE,
+		    &vmw_cmd_update_gb_surface, true, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_READBACK_GB_IMAGE,
+		    &vmw_cmd_readback_gb_image, true, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_READBACK_GB_SURFACE,
+		    &vmw_cmd_readback_gb_surface, true, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_INVALIDATE_GB_IMAGE,
+		    &vmw_cmd_invalidate_gb_image, true, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_INVALIDATE_GB_SURFACE,
+		    &vmw_cmd_invalidate_gb_surface, true, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_DEFINE_GB_CONTEXT, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_DESTROY_GB_CONTEXT, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_BIND_GB_CONTEXT, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_READBACK_GB_CONTEXT, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_INVALIDATE_GB_CONTEXT, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_DEFINE_GB_SHADER, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_BIND_GB_SHADER, &vmw_cmd_bind_gb_shader,
+		    true, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_DESTROY_GB_SHADER, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_SET_OTABLE_BASE64, &vmw_cmd_invalid,
+		    false, false, false),
+	VMW_CMD_DEF(SVGA_3D_CMD_BEGIN_GB_QUERY, &vmw_cmd_begin_gb_query,
+		    true, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_END_GB_QUERY, &vmw_cmd_end_gb_query,
+		    true, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_WAIT_FOR_GB_QUERY, &vmw_cmd_wait_gb_query,
+		    true, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_NOP, &vmw_cmd_ok,
+		    true, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_ENABLE_GART, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_DISABLE_GART, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_MAP_MOB_INTO_GART, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_UNMAP_GART_RANGE, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_DEFINE_GB_SCREENTARGET, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_DESTROY_GB_SCREENTARGET, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_BIND_GB_SCREENTARGET, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_UPDATE_GB_SCREENTARGET, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_READBACK_GB_IMAGE_PARTIAL, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_INVALIDATE_GB_IMAGE_PARTIAL, &vmw_cmd_invalid,
+		    false, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_SET_GB_SHADERCONSTS_INLINE, &vmw_cmd_cid_check,
+		    true, false, true)
 };
 
 static int vmw_cmd_check(struct vmw_private *dev_priv,
@@ -1095,6 +1772,8 @@ static int vmw_cmd_check(struct vmw_private *dev_priv,
 	uint32_t size_remaining = *size;
 	SVGA3dCmdHeader *header = (SVGA3dCmdHeader *) buf;
 	int ret;
+	const struct vmw_cmd_entry *entry;
+	bool gb = dev_priv->capabilities & SVGA_CAP_GBOBJECTS;
 
 	cmd_id = le32_to_cpu(((uint32_t *)buf)[0]);
 	/* Handle any none 3D commands */
@@ -1107,18 +1786,40 @@ static int vmw_cmd_check(struct vmw_private *dev_priv,
 
 	cmd_id -= SVGA_3D_CMD_BASE;
 	if (unlikely(*size > size_remaining))
-		goto out_err;
+		goto out_invalid;
 
 	if (unlikely(cmd_id >= SVGA_3D_CMD_MAX - SVGA_3D_CMD_BASE))
-		goto out_err;
+		goto out_invalid;
+
+	entry = &vmw_cmd_entries[cmd_id];
+	if (unlikely(!entry->user_allow && !sw_context->kernel))
+		goto out_privileged;
 
-	ret = vmw_cmd_funcs[cmd_id](dev_priv, sw_context, header);
+	if (unlikely(entry->gb_disable && gb))
+		goto out_old;
+
+	if (unlikely(entry->gb_enable && !gb))
+		goto out_new;
+
+	ret = entry->func(dev_priv, sw_context, header);
 	if (unlikely(ret != 0))
-		goto out_err;
+		goto out_invalid;
 
 	return 0;
-out_err:
-	DRM_ERROR("Illegal / Invalid SVGA3D command: %d\n",
+out_invalid:
+	DRM_ERROR("Invalid SVGA3D command: %d\n",
+		  cmd_id + SVGA_3D_CMD_BASE);
+	return -EINVAL;
+out_privileged:
+	DRM_ERROR("Privileged SVGA3D command: %d\n",
+		  cmd_id + SVGA_3D_CMD_BASE);
+	return -EPERM;
+out_old:
+	DRM_ERROR("Deprecated (disallowed) SVGA3D command: %d\n",
+		  cmd_id + SVGA_3D_CMD_BASE);
+	return -EINVAL;
+out_new:
+	DRM_ERROR("SVGA3D command: %d not supported by virtual hardware.\n",
 		  cmd_id + SVGA_3D_CMD_BASE);
 	return -EINVAL;
 }
@@ -1174,6 +1875,9 @@ static void vmw_apply_relocations(struct vmw_sw_context *sw_context)
 		case VMW_PL_GMR:
 			reloc->location->gmrId = bo->mem.start;
 			break;
+		case VMW_PL_MOB:
+			*reloc->mob_loc = bo->mem.start;
+			break;
 		default:
 			BUG();
 		}
@@ -1198,6 +1902,8 @@ static void vmw_resource_list_unreference(struct list_head *list)
 	list_for_each_entry_safe(val, val_next, list, head) {
 		list_del_init(&val->head);
 		vmw_resource_unreference(&val->res);
+		if (unlikely(val->staged_bindings))
+			kfree(val->staged_bindings);
 		kfree(val);
 	}
 }
@@ -1224,7 +1930,8 @@ static void vmw_clear_validations(struct vmw_sw_context *sw_context)
 }
 
 static int vmw_validate_single_buffer(struct vmw_private *dev_priv,
-				      struct ttm_buffer_object *bo)
+				      struct ttm_buffer_object *bo,
+				      bool validate_as_mob)
 {
 	int ret;
 
@@ -1238,6 +1945,9 @@ static int vmw_validate_single_buffer(struct vmw_private *dev_priv,
 	     dev_priv->dummy_query_bo_pinned))
 		return 0;
 
+	if (validate_as_mob)
+		return ttm_bo_validate(bo, &vmw_mob_placement, true, false);
+
 	/**
 	 * Put BO in VRAM if there is space, otherwise as a GMR.
 	 * If there is no space in VRAM and GMR ids are all used up,
@@ -1259,7 +1969,6 @@ static int vmw_validate_single_buffer(struct vmw_private *dev_priv,
 	return ret;
 }
 
-
 static int vmw_validate_buffers(struct vmw_private *dev_priv,
 				struct vmw_sw_context *sw_context)
 {
@@ -1267,7 +1976,8 @@ static int vmw_validate_buffers(struct vmw_private *dev_priv,
 	int ret;
 
 	list_for_each_entry(entry, &sw_context->validate_nodes, base.head) {
-		ret = vmw_validate_single_buffer(dev_priv, entry->base.bo);
+		ret = vmw_validate_single_buffer(dev_priv, entry->base.bo,
+						 entry->validate_as_mob);
 		if (unlikely(ret != 0))
 			return ret;
 	}
@@ -1509,11 +2219,17 @@ int vmw_execbuf_process(struct drm_file *file_priv,
 			goto out_err;
 	}
 
+	ret = mutex_lock_interruptible(&dev_priv->binding_mutex);
+	if (unlikely(ret != 0)) {
+		ret = -ERESTARTSYS;
+		goto out_err;
+	}
+
 	cmd = vmw_fifo_reserve(dev_priv, command_size);
 	if (unlikely(cmd == NULL)) {
 		DRM_ERROR("Failed reserving fifo space for commands.\n");
 		ret = -ENOMEM;
-		goto out_err;
+		goto out_unlock_binding;
 	}
 
 	vmw_apply_relocations(sw_context);
@@ -1538,6 +2254,8 @@ int vmw_execbuf_process(struct drm_file *file_priv,
 		DRM_ERROR("Fence submission error. Syncing.\n");
 
 	vmw_resource_list_unreserve(&sw_context->resource_list, false);
+	mutex_unlock(&dev_priv->binding_mutex);
+
 	ttm_eu_fence_buffer_objects(&ticket, &sw_context->validate_nodes,
 				    (void *) fence);
 
@@ -1568,6 +2286,8 @@ int vmw_execbuf_process(struct drm_file *file_priv,
 
 	return 0;
 
+out_unlock_binding:
+	mutex_unlock(&dev_priv->binding_mutex);
 out_err:
 	vmw_resource_relocations_free(&sw_context->res_relocations);
 	vmw_free_relocations(sw_context);
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_fence.c b/drivers/gpu/drm/vmwgfx/vmwgfx_fence.c
index c62d20e8a6f1..436b013b4231 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_fence.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_fence.c
@@ -271,7 +271,7 @@ void vmw_fence_obj_unreference(struct vmw_fence_obj **fence_p)
 	spin_unlock_irq(&fman->lock);
 }
 
-void vmw_fences_perform_actions(struct vmw_fence_manager *fman,
+static void vmw_fences_perform_actions(struct vmw_fence_manager *fman,
 				struct list_head *list)
 {
 	struct vmw_fence_action *action, *next_action;
@@ -897,7 +897,7 @@ static void vmw_event_fence_action_cleanup(struct vmw_fence_action *action)
  * Note that the action callbacks may be executed before this function
  * returns.
  */
-void vmw_fence_obj_add_action(struct vmw_fence_obj *fence,
+static void vmw_fence_obj_add_action(struct vmw_fence_obj *fence,
 			      struct vmw_fence_action *action)
 {
 	struct vmw_fence_manager *fman = fence->fman;
@@ -993,7 +993,7 @@ struct vmw_event_fence_pending {
 	struct drm_vmw_event_fence event;
 };
 
-int vmw_event_fence_action_create(struct drm_file *file_priv,
+static int vmw_event_fence_action_create(struct drm_file *file_priv,
 				  struct vmw_fence_obj *fence,
 				  uint32_t flags,
 				  uint64_t user_data,
@@ -1080,7 +1080,8 @@ int vmw_fence_event_ioctl(struct drm_device *dev, void *data,
 	 */
 	if (arg->handle) {
 		struct ttm_base_object *base =
-			ttm_base_object_lookup(vmw_fp->tfile, arg->handle);
+			ttm_base_object_lookup_for_ref(dev_priv->tdev,
+						       arg->handle);
 
 		if (unlikely(base == NULL)) {
 			DRM_ERROR("Fence event invalid fence object handle "
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_fifo.c b/drivers/gpu/drm/vmwgfx/vmwgfx_fifo.c
index 3eb148667d63..6ccd993e26bf 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_fifo.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_fifo.c
@@ -35,6 +35,23 @@ bool vmw_fifo_have_3d(struct vmw_private *dev_priv)
 	uint32_t fifo_min, hwversion;
 	const struct vmw_fifo_state *fifo = &dev_priv->fifo;
 
+	if (!(dev_priv->capabilities & SVGA_CAP_3D))
+		return false;
+
+	if (dev_priv->capabilities & SVGA_CAP_GBOBJECTS) {
+		uint32_t result;
+
+		if (!dev_priv->has_mob)
+			return false;
+
+		mutex_lock(&dev_priv->hw_mutex);
+		vmw_write(dev_priv, SVGA_REG_DEV_CAP, SVGA3D_DEVCAP_3D);
+		result = vmw_read(dev_priv, SVGA_REG_DEV_CAP);
+		mutex_unlock(&dev_priv->hw_mutex);
+
+		return (result != 0);
+	}
+
 	if (!(dev_priv->capabilities & SVGA_CAP_EXTENDED_FIFO))
 		return false;
 
@@ -511,24 +528,16 @@ out_err:
 }
 
 /**
- * vmw_fifo_emit_dummy_query - emits a dummy query to the fifo.
+ * vmw_fifo_emit_dummy_legacy_query - emits a dummy query to the fifo using
+ * legacy query commands.
  *
  * @dev_priv: The device private structure.
  * @cid: The hardware context id used for the query.
  *
- * This function is used to emit a dummy occlusion query with
- * no primitives rendered between query begin and query end.
- * It's used to provide a query barrier, in order to know that when
- * this query is finished, all preceding queries are also finished.
- *
- * A Query results structure should have been initialized at the start
- * of the dev_priv->dummy_query_bo buffer object. And that buffer object
- * must also be either reserved or pinned when this function is called.
- *
- * Returns -ENOMEM on failure to reserve fifo space.
+ * See the vmw_fifo_emit_dummy_query documentation.
  */
-int vmw_fifo_emit_dummy_query(struct vmw_private *dev_priv,
-			      uint32_t cid)
+static int vmw_fifo_emit_dummy_legacy_query(struct vmw_private *dev_priv,
+					    uint32_t cid)
 {
 	/*
 	 * A query wait without a preceding query end will
@@ -566,3 +575,75 @@ int vmw_fifo_emit_dummy_query(struct vmw_private *dev_priv,
 
 	return 0;
 }
+
+/**
+ * vmw_fifo_emit_dummy_gb_query - emits a dummy query to the fifo using
+ * guest-backed resource query commands.
+ *
+ * @dev_priv: The device private structure.
+ * @cid: The hardware context id used for the query.
+ *
+ * See the vmw_fifo_emit_dummy_query documentation.
+ */
+static int vmw_fifo_emit_dummy_gb_query(struct vmw_private *dev_priv,
+					uint32_t cid)
+{
+	/*
+	 * A query wait without a preceding query end will
+	 * actually finish all queries for this cid
+	 * without writing to the query result structure.
+	 */
+
+	struct ttm_buffer_object *bo = dev_priv->dummy_query_bo;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdWaitForGBQuery body;
+	} *cmd;
+
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Out of fifo space for dummy query.\n");
+		return -ENOMEM;
+	}
+
+	cmd->header.id = SVGA_3D_CMD_WAIT_FOR_GB_QUERY;
+	cmd->header.size = sizeof(cmd->body);
+	cmd->body.cid = cid;
+	cmd->body.type = SVGA3D_QUERYTYPE_OCCLUSION;
+	BUG_ON(bo->mem.mem_type != VMW_PL_MOB);
+	cmd->body.mobid = bo->mem.start;
+	cmd->body.offset = 0;
+
+	vmw_fifo_commit(dev_priv, sizeof(*cmd));
+
+	return 0;
+}
+
+
+/**
+ * vmw_fifo_emit_dummy_gb_query - emits a dummy query to the fifo using
+ * appropriate resource query commands.
+ *
+ * @dev_priv: The device private structure.
+ * @cid: The hardware context id used for the query.
+ *
+ * This function is used to emit a dummy occlusion query with
+ * no primitives rendered between query begin and query end.
+ * It's used to provide a query barrier, in order to know that when
+ * this query is finished, all preceding queries are also finished.
+ *
+ * A Query results structure should have been initialized at the start
+ * of the dev_priv->dummy_query_bo buffer object. And that buffer object
+ * must also be either reserved or pinned when this function is called.
+ *
+ * Returns -ENOMEM on failure to reserve fifo space.
+ */
+int vmw_fifo_emit_dummy_query(struct vmw_private *dev_priv,
+			      uint32_t cid)
+{
+	if (dev_priv->has_mob)
+		return vmw_fifo_emit_dummy_gb_query(dev_priv, cid);
+
+	return vmw_fifo_emit_dummy_legacy_query(dev_priv, cid);
+}
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_gmr.c b/drivers/gpu/drm/vmwgfx/vmwgfx_gmr.c
index 6ef0b035becb..61d8d803199f 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_gmr.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_gmr.c
@@ -125,181 +125,27 @@ static void vmw_gmr2_unbind(struct vmw_private *dev_priv,
 }
 
 
-static void vmw_gmr_free_descriptors(struct device *dev, dma_addr_t desc_dma,
-				     struct list_head *desc_pages)
-{
-	struct page *page, *next;
-	struct svga_guest_mem_descriptor *page_virtual;
-	unsigned int desc_per_page = PAGE_SIZE /
-		sizeof(struct svga_guest_mem_descriptor) - 1;
-
-	if (list_empty(desc_pages))
-		return;
-
-	list_for_each_entry_safe(page, next, desc_pages, lru) {
-		list_del_init(&page->lru);
-
-		if (likely(desc_dma != DMA_ADDR_INVALID)) {
-			dma_unmap_page(dev, desc_dma, PAGE_SIZE,
-				       DMA_TO_DEVICE);
-		}
-
-		page_virtual = kmap_atomic(page);
-		desc_dma = (dma_addr_t)
-			le32_to_cpu(page_virtual[desc_per_page].ppn) <<
-			PAGE_SHIFT;
-		kunmap_atomic(page_virtual);
-
-		__free_page(page);
-	}
-}
-
-/**
- * FIXME: Adjust to the ttm lowmem / highmem storage to minimize
- * the number of used descriptors.
- *
- */
-
-static int vmw_gmr_build_descriptors(struct device *dev,
-				     struct list_head *desc_pages,
-				     struct vmw_piter *iter,
-				     unsigned long num_pages,
-				     dma_addr_t *first_dma)
-{
-	struct page *page;
-	struct svga_guest_mem_descriptor *page_virtual = NULL;
-	struct svga_guest_mem_descriptor *desc_virtual = NULL;
-	unsigned int desc_per_page;
-	unsigned long prev_pfn;
-	unsigned long pfn;
-	int ret;
-	dma_addr_t desc_dma;
-
-	desc_per_page = PAGE_SIZE /
-	    sizeof(struct svga_guest_mem_descriptor) - 1;
-
-	while (likely(num_pages != 0)) {
-		page = alloc_page(__GFP_HIGHMEM);
-		if (unlikely(page == NULL)) {
-			ret = -ENOMEM;
-			goto out_err;
-		}
-
-		list_add_tail(&page->lru, desc_pages);
-		page_virtual = kmap_atomic(page);
-		desc_virtual = page_virtual - 1;
-		prev_pfn = ~(0UL);
-
-		while (likely(num_pages != 0)) {
-			pfn = vmw_piter_dma_addr(iter) >> PAGE_SHIFT;
-
-			if (pfn != prev_pfn + 1) {
-
-				if (desc_virtual - page_virtual ==
-				    desc_per_page - 1)
-					break;
-
-				(++desc_virtual)->ppn = cpu_to_le32(pfn);
-				desc_virtual->num_pages = cpu_to_le32(1);
-			} else {
-				uint32_t tmp =
-				    le32_to_cpu(desc_virtual->num_pages);
-				desc_virtual->num_pages = cpu_to_le32(tmp + 1);
-			}
-			prev_pfn = pfn;
-			--num_pages;
-			vmw_piter_next(iter);
-		}
-
-		(++desc_virtual)->ppn = DMA_PAGE_INVALID;
-		desc_virtual->num_pages = cpu_to_le32(0);
-		kunmap_atomic(page_virtual);
-	}
-
-	desc_dma = 0;
-	list_for_each_entry_reverse(page, desc_pages, lru) {
-		page_virtual = kmap_atomic(page);
-		page_virtual[desc_per_page].ppn = cpu_to_le32
-			(desc_dma >> PAGE_SHIFT);
-		kunmap_atomic(page_virtual);
-		desc_dma = dma_map_page(dev, page, 0, PAGE_SIZE,
-					DMA_TO_DEVICE);
-
-		if (unlikely(dma_mapping_error(dev, desc_dma)))
-			goto out_err;
-	}
-	*first_dma = desc_dma;
-
-	return 0;
-out_err:
-	vmw_gmr_free_descriptors(dev, DMA_ADDR_INVALID, desc_pages);
-	return ret;
-}
-
-static void vmw_gmr_fire_descriptors(struct vmw_private *dev_priv,
-				     int gmr_id, dma_addr_t desc_dma)
-{
-	mutex_lock(&dev_priv->hw_mutex);
-
-	vmw_write(dev_priv, SVGA_REG_GMR_ID, gmr_id);
-	wmb();
-	vmw_write(dev_priv, SVGA_REG_GMR_DESCRIPTOR, desc_dma >> PAGE_SHIFT);
-	mb();
-
-	mutex_unlock(&dev_priv->hw_mutex);
-
-}
-
 int vmw_gmr_bind(struct vmw_private *dev_priv,
 		 const struct vmw_sg_table *vsgt,
 		 unsigned long num_pages,
 		 int gmr_id)
 {
-	struct list_head desc_pages;
-	dma_addr_t desc_dma = 0;
-	struct device *dev = dev_priv->dev->dev;
 	struct vmw_piter data_iter;
-	int ret;
 
 	vmw_piter_start(&data_iter, vsgt, 0);
 
 	if (unlikely(!vmw_piter_next(&data_iter)))
 		return 0;
 
-	if (likely(dev_priv->capabilities & SVGA_CAP_GMR2))
-		return vmw_gmr2_bind(dev_priv, &data_iter, num_pages, gmr_id);
-
-	if (unlikely(!(dev_priv->capabilities & SVGA_CAP_GMR)))
-		return -EINVAL;
-
-	if (vsgt->num_regions > dev_priv->max_gmr_descriptors)
+	if (unlikely(!(dev_priv->capabilities & SVGA_CAP_GMR2)))
 		return -EINVAL;
 
-	INIT_LIST_HEAD(&desc_pages);
-
-	ret = vmw_gmr_build_descriptors(dev, &desc_pages, &data_iter,
-					num_pages, &desc_dma);
-	if (unlikely(ret != 0))
-		return ret;
-
-	vmw_gmr_fire_descriptors(dev_priv, gmr_id, desc_dma);
-	vmw_gmr_free_descriptors(dev, desc_dma, &desc_pages);
-
-	return 0;
+	return vmw_gmr2_bind(dev_priv, &data_iter, num_pages, gmr_id);
 }
 
 
 void vmw_gmr_unbind(struct vmw_private *dev_priv, int gmr_id)
 {
-	if (likely(dev_priv->capabilities & SVGA_CAP_GMR2)) {
+	if (likely(dev_priv->capabilities & SVGA_CAP_GMR2))
 		vmw_gmr2_unbind(dev_priv, gmr_id);
-		return;
-	}
-
-	mutex_lock(&dev_priv->hw_mutex);
-	vmw_write(dev_priv, SVGA_REG_GMR_ID, gmr_id);
-	wmb();
-	vmw_write(dev_priv, SVGA_REG_GMR_DESCRIPTOR, 0);
-	mb();
-	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 c5c054ae9056..b1273e8e9a69 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_gmrid_manager.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_gmrid_manager.c
@@ -125,10 +125,21 @@ static int vmw_gmrid_man_init(struct ttm_mem_type_manager *man,
 		return -ENOMEM;
 
 	spin_lock_init(&gman->lock);
-	gman->max_gmr_pages = dev_priv->max_gmr_pages;
 	gman->used_gmr_pages = 0;
 	ida_init(&gman->gmr_ida);
-	gman->max_gmr_ids = p_size;
+
+	switch (p_size) {
+	case VMW_PL_GMR:
+		gman->max_gmr_ids = dev_priv->max_gmr_ids;
+		gman->max_gmr_pages = dev_priv->max_gmr_pages;
+		break;
+	case VMW_PL_MOB:
+		gman->max_gmr_ids = VMWGFX_NUM_MOB;
+		gman->max_gmr_pages = dev_priv->max_mob_pages;
+		break;
+	default:
+		BUG();
+	}
 	man->priv = (void *) gman;
 	return 0;
 }
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_ioctl.c b/drivers/gpu/drm/vmwgfx/vmwgfx_ioctl.c
index 45d5b5ab6ca9..116c49736763 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_ioctl.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_ioctl.c
@@ -53,7 +53,7 @@ int vmw_getparam_ioctl(struct drm_device *dev, void *data,
 		param->value = dev_priv->fifo.capabilities;
 		break;
 	case DRM_VMW_PARAM_MAX_FB_SIZE:
-		param->value = dev_priv->vram_size;
+		param->value = dev_priv->prim_bb_mem;
 		break;
 	case DRM_VMW_PARAM_FIFO_HW_VERSION:
 	{
@@ -71,6 +71,17 @@ int vmw_getparam_ioctl(struct drm_device *dev, void *data,
 	case DRM_VMW_PARAM_MAX_SURF_MEMORY:
 		param->value = dev_priv->memory_size;
 		break;
+	case DRM_VMW_PARAM_3D_CAPS_SIZE:
+		if (dev_priv->capabilities & SVGA_CAP_GBOBJECTS)
+			param->value = SVGA3D_DEVCAP_MAX;
+		else
+			param->value = (SVGA_FIFO_3D_CAPS_LAST -
+					SVGA_FIFO_3D_CAPS + 1);
+		param->value *= sizeof(uint32_t);
+		break;
+	case DRM_VMW_PARAM_MAX_MOB_MEMORY:
+		param->value = dev_priv->max_mob_pages * PAGE_SIZE;
+		break;
 	default:
 		DRM_ERROR("Illegal vmwgfx get param request: %d\n",
 			  param->param);
@@ -92,13 +103,19 @@ int vmw_get_cap_3d_ioctl(struct drm_device *dev, void *data,
 	void __user *buffer = (void __user *)((unsigned long)(arg->buffer));
 	void *bounce;
 	int ret;
+	bool gb_objects = !!(dev_priv->capabilities & SVGA_CAP_GBOBJECTS);
 
 	if (unlikely(arg->pad64 != 0)) {
 		DRM_ERROR("Illegal GET_3D_CAP argument.\n");
 		return -EINVAL;
 	}
 
-	size = (SVGA_FIFO_3D_CAPS_LAST - SVGA_FIFO_3D_CAPS + 1) << 2;
+	if (gb_objects)
+		size = SVGA3D_DEVCAP_MAX;
+	else
+		size = (SVGA_FIFO_3D_CAPS_LAST - SVGA_FIFO_3D_CAPS + 1);
+
+	size *= sizeof(uint32_t);
 
 	if (arg->max_size < size)
 		size = arg->max_size;
@@ -109,8 +126,22 @@ int vmw_get_cap_3d_ioctl(struct drm_device *dev, void *data,
 		return -ENOMEM;
 	}
 
-	fifo_mem = dev_priv->mmio_virt;
-	memcpy_fromio(bounce, &fifo_mem[SVGA_FIFO_3D_CAPS], size);
+	if (gb_objects) {
+		int i;
+		uint32_t *bounce32 = (uint32_t *) bounce;
+
+		mutex_lock(&dev_priv->hw_mutex);
+		for (i = 0; i < SVGA3D_DEVCAP_MAX; ++i) {
+			vmw_write(dev_priv, SVGA_REG_DEV_CAP, i);
+			*bounce32++ = vmw_read(dev_priv, SVGA_REG_DEV_CAP);
+		}
+		mutex_unlock(&dev_priv->hw_mutex);
+
+	} else {
+
+		fifo_mem = dev_priv->mmio_virt;
+		memcpy_fromio(bounce, &fifo_mem[SVGA_FIFO_3D_CAPS], size);
+	}
 
 	ret = copy_to_user(buffer, bounce, size);
 	if (ret)
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_irq.c b/drivers/gpu/drm/vmwgfx/vmwgfx_irq.c
index 4640adbcaf91..0c423766c441 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_irq.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_irq.c
@@ -30,7 +30,7 @@
 
 #define VMW_FENCE_WRAP (1 << 24)
 
-irqreturn_t vmw_irq_handler(DRM_IRQ_ARGS)
+irqreturn_t vmw_irq_handler(int irq, void *arg)
 {
 	struct drm_device *dev = (struct drm_device *)arg;
 	struct vmw_private *dev_priv = vmw_priv(dev);
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c b/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c
index 03f1c2038631..8a650413dea5 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c
@@ -40,7 +40,7 @@ struct vmw_clip_rect {
  * Clip @num_rects number of @rects against @clip storing the
  * results in @out_rects and the number of passed rects in @out_num.
  */
-void vmw_clip_cliprects(struct drm_clip_rect *rects,
+static void vmw_clip_cliprects(struct drm_clip_rect *rects,
 			int num_rects,
 			struct vmw_clip_rect clip,
 			SVGASignedRect *out_rects,
@@ -423,7 +423,7 @@ struct vmw_framebuffer_surface {
 	struct drm_master *master;
 };
 
-void vmw_framebuffer_surface_destroy(struct drm_framebuffer *framebuffer)
+static void vmw_framebuffer_surface_destroy(struct drm_framebuffer *framebuffer)
 {
 	struct vmw_framebuffer_surface *vfbs =
 		vmw_framebuffer_to_vfbs(framebuffer);
@@ -589,7 +589,7 @@ out_free_tmp:
 	return ret;
 }
 
-int vmw_framebuffer_surface_dirty(struct drm_framebuffer *framebuffer,
+static int vmw_framebuffer_surface_dirty(struct drm_framebuffer *framebuffer,
 				  struct drm_file *file_priv,
 				  unsigned flags, unsigned color,
 				  struct drm_clip_rect *clips,
@@ -609,9 +609,13 @@ int vmw_framebuffer_surface_dirty(struct drm_framebuffer *framebuffer,
 	if (!dev_priv->sou_priv)
 		return -EINVAL;
 
+	drm_modeset_lock_all(dev_priv->dev);
+
 	ret = ttm_read_lock(&vmaster->lock, true);
-	if (unlikely(ret != 0))
+	if (unlikely(ret != 0)) {
+		drm_modeset_unlock_all(dev_priv->dev);
 		return ret;
+	}
 
 	if (!num_clips) {
 		num_clips = 1;
@@ -629,6 +633,9 @@ int vmw_framebuffer_surface_dirty(struct drm_framebuffer *framebuffer,
 				   clips, num_clips, inc, NULL);
 
 	ttm_read_unlock(&vmaster->lock);
+
+	drm_modeset_unlock_all(dev_priv->dev);
+
 	return 0;
 }
 
@@ -665,9 +672,9 @@ static int vmw_kms_new_framebuffer_surface(struct vmw_private *dev_priv,
 
 	if (unlikely(surface->mip_levels[0] != 1 ||
 		     surface->num_sizes != 1 ||
-		     surface->sizes[0].width < mode_cmd->width ||
-		     surface->sizes[0].height < mode_cmd->height ||
-		     surface->sizes[0].depth != 1)) {
+		     surface->base_size.width < mode_cmd->width ||
+		     surface->base_size.height < mode_cmd->height ||
+		     surface->base_size.depth != 1)) {
 		DRM_ERROR("Incompatible surface dimensions "
 			  "for requested mode.\n");
 		return -EINVAL;
@@ -754,7 +761,7 @@ struct vmw_framebuffer_dmabuf {
 	struct vmw_dma_buffer *buffer;
 };
 
-void vmw_framebuffer_dmabuf_destroy(struct drm_framebuffer *framebuffer)
+static void vmw_framebuffer_dmabuf_destroy(struct drm_framebuffer *framebuffer)
 {
 	struct vmw_framebuffer_dmabuf *vfbd =
 		vmw_framebuffer_to_vfbd(framebuffer);
@@ -940,7 +947,7 @@ static int do_dmabuf_dirty_sou(struct drm_file *file_priv,
 	return ret;
 }
 
-int vmw_framebuffer_dmabuf_dirty(struct drm_framebuffer *framebuffer,
+static int vmw_framebuffer_dmabuf_dirty(struct drm_framebuffer *framebuffer,
 				 struct drm_file *file_priv,
 				 unsigned flags, unsigned color,
 				 struct drm_clip_rect *clips,
@@ -953,9 +960,13 @@ int vmw_framebuffer_dmabuf_dirty(struct drm_framebuffer *framebuffer,
 	struct drm_clip_rect norect;
 	int ret, increment = 1;
 
+	drm_modeset_lock_all(dev_priv->dev);
+
 	ret = ttm_read_lock(&vmaster->lock, true);
-	if (unlikely(ret != 0))
+	if (unlikely(ret != 0)) {
+		drm_modeset_unlock_all(dev_priv->dev);
 		return ret;
+	}
 
 	if (!num_clips) {
 		num_clips = 1;
@@ -979,6 +990,9 @@ int vmw_framebuffer_dmabuf_dirty(struct drm_framebuffer *framebuffer,
 	}
 
 	ttm_read_unlock(&vmaster->lock);
+
+	drm_modeset_unlock_all(dev_priv->dev);
+
 	return ret;
 }
 
@@ -1631,7 +1645,7 @@ bool vmw_kms_validate_mode_vram(struct vmw_private *dev_priv,
 				uint32_t pitch,
 				uint32_t height)
 {
-	return ((u64) pitch * (u64) height) < (u64) dev_priv->vram_size;
+	return ((u64) pitch * (u64) height) < (u64) dev_priv->prim_bb_mem;
 }
 
 
@@ -1663,7 +1677,7 @@ void vmw_disable_vblank(struct drm_device *dev, int crtc)
  * Small shared kms functions.
  */
 
-int vmw_du_update_layout(struct vmw_private *dev_priv, unsigned num,
+static int vmw_du_update_layout(struct vmw_private *dev_priv, unsigned num,
 			 struct drm_vmw_rect *rects)
 {
 	struct drm_device *dev = dev_priv->dev;
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_mob.c b/drivers/gpu/drm/vmwgfx/vmwgfx_mob.c
new file mode 100644
index 000000000000..4910e7b81811
--- /dev/null
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_mob.c
@@ -0,0 +1,652 @@
+/**************************************************************************
+ *
+ * Copyright © 2012 VMware, Inc., Palo Alto, CA., USA
+ * 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.
+ *
+ **************************************************************************/
+
+#include "vmwgfx_drv.h"
+
+/*
+ * If we set up the screen target otable, screen objects stop working.
+ */
+
+#define VMW_OTABLE_SETUP_SUB ((VMWGFX_ENABLE_SCREEN_TARGET_OTABLE) ? 0 : 1)
+
+#ifdef CONFIG_64BIT
+#define VMW_PPN_SIZE 8
+#define VMW_MOBFMT_PTDEPTH_0 SVGA3D_MOBFMT_PTDEPTH64_0
+#define VMW_MOBFMT_PTDEPTH_1 SVGA3D_MOBFMT_PTDEPTH64_1
+#define VMW_MOBFMT_PTDEPTH_2 SVGA3D_MOBFMT_PTDEPTH64_2
+#else
+#define VMW_PPN_SIZE 4
+#define VMW_MOBFMT_PTDEPTH_0 SVGA3D_MOBFMT_PTDEPTH_0
+#define VMW_MOBFMT_PTDEPTH_1 SVGA3D_MOBFMT_PTDEPTH_1
+#define VMW_MOBFMT_PTDEPTH_2 SVGA3D_MOBFMT_PTDEPTH_2
+#endif
+
+/*
+ * struct vmw_mob - Structure containing page table and metadata for a
+ * Guest Memory OBject.
+ *
+ * @num_pages       Number of pages that make up the page table.
+ * @pt_level        The indirection level of the page table. 0-2.
+ * @pt_root_page    DMA address of the level 0 page of the page table.
+ */
+struct vmw_mob {
+	struct ttm_buffer_object *pt_bo;
+	unsigned long num_pages;
+	unsigned pt_level;
+	dma_addr_t pt_root_page;
+	uint32_t id;
+};
+
+/*
+ * struct vmw_otable - Guest Memory OBject table metadata
+ *
+ * @size:           Size of the table (page-aligned).
+ * @page_table:     Pointer to a struct vmw_mob holding the page table.
+ */
+struct vmw_otable {
+	unsigned long size;
+	struct vmw_mob *page_table;
+};
+
+static int vmw_mob_pt_populate(struct vmw_private *dev_priv,
+			       struct vmw_mob *mob);
+static void vmw_mob_pt_setup(struct vmw_mob *mob,
+			     struct vmw_piter data_iter,
+			     unsigned long num_data_pages);
+
+/*
+ * vmw_setup_otable_base - Issue an object table base setup command to
+ * the device
+ *
+ * @dev_priv:       Pointer to a device private structure
+ * @type:           Type of object table base
+ * @offset          Start of table offset into dev_priv::otable_bo
+ * @otable          Pointer to otable metadata;
+ *
+ * This function returns -ENOMEM if it fails to reserve fifo space,
+ * and may block waiting for fifo space.
+ */
+static int vmw_setup_otable_base(struct vmw_private *dev_priv,
+				 SVGAOTableType type,
+				 unsigned long offset,
+				 struct vmw_otable *otable)
+{
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdSetOTableBase64 body;
+	} *cmd;
+	struct vmw_mob *mob;
+	const struct vmw_sg_table *vsgt;
+	struct vmw_piter iter;
+	int ret;
+
+	BUG_ON(otable->page_table != NULL);
+
+	vsgt = vmw_bo_sg_table(dev_priv->otable_bo);
+	vmw_piter_start(&iter, vsgt, offset >> PAGE_SHIFT);
+	WARN_ON(!vmw_piter_next(&iter));
+
+	mob = vmw_mob_create(otable->size >> PAGE_SHIFT);
+	if (unlikely(mob == NULL)) {
+		DRM_ERROR("Failed creating OTable page table.\n");
+		return -ENOMEM;
+	}
+
+	if (otable->size <= PAGE_SIZE) {
+		mob->pt_level = VMW_MOBFMT_PTDEPTH_0;
+		mob->pt_root_page = vmw_piter_dma_addr(&iter);
+	} else if (vsgt->num_regions == 1) {
+		mob->pt_level = SVGA3D_MOBFMT_RANGE;
+		mob->pt_root_page = vmw_piter_dma_addr(&iter);
+	} else {
+		ret = vmw_mob_pt_populate(dev_priv, mob);
+		if (unlikely(ret != 0))
+			goto out_no_populate;
+
+		vmw_mob_pt_setup(mob, iter, otable->size >> PAGE_SHIFT);
+		mob->pt_level += VMW_MOBFMT_PTDEPTH_1 - SVGA3D_MOBFMT_PTDEPTH_1;
+	}
+
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for OTable setup.\n");
+		goto out_no_fifo;
+	}
+
+	memset(cmd, 0, sizeof(*cmd));
+	cmd->header.id = SVGA_3D_CMD_SET_OTABLE_BASE64;
+	cmd->header.size = sizeof(cmd->body);
+	cmd->body.type = type;
+	cmd->body.baseAddress = cpu_to_le64(mob->pt_root_page >> PAGE_SHIFT);
+	cmd->body.sizeInBytes = otable->size;
+	cmd->body.validSizeInBytes = 0;
+	cmd->body.ptDepth = mob->pt_level;
+
+	/*
+	 * The device doesn't support this, But the otable size is
+	 * determined at compile-time, so this BUG shouldn't trigger
+	 * randomly.
+	 */
+	BUG_ON(mob->pt_level == VMW_MOBFMT_PTDEPTH_2);
+
+	vmw_fifo_commit(dev_priv, sizeof(*cmd));
+	otable->page_table = mob;
+
+	return 0;
+
+out_no_fifo:
+out_no_populate:
+	vmw_mob_destroy(mob);
+	return ret;
+}
+
+/*
+ * vmw_takedown_otable_base - Issue an object table base takedown command
+ * to the device
+ *
+ * @dev_priv:       Pointer to a device private structure
+ * @type:           Type of object table base
+ *
+ */
+static void vmw_takedown_otable_base(struct vmw_private *dev_priv,
+				     SVGAOTableType type,
+				     struct vmw_otable *otable)
+{
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdSetOTableBase body;
+	} *cmd;
+	struct ttm_buffer_object *bo;
+
+	if (otable->page_table == NULL)
+		return;
+
+	bo = otable->page_table->pt_bo;
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+	if (unlikely(cmd == NULL))
+		DRM_ERROR("Failed reserving FIFO space for OTable setup.\n");
+
+	memset(cmd, 0, sizeof(*cmd));
+	cmd->header.id = SVGA_3D_CMD_SET_OTABLE_BASE;
+	cmd->header.size = sizeof(cmd->body);
+	cmd->body.type = type;
+	cmd->body.baseAddress = 0;
+	cmd->body.sizeInBytes = 0;
+	cmd->body.validSizeInBytes = 0;
+	cmd->body.ptDepth = SVGA3D_MOBFMT_INVALID;
+	vmw_fifo_commit(dev_priv, sizeof(*cmd));
+
+	if (bo) {
+		int ret;
+
+		ret = ttm_bo_reserve(bo, false, true, false, NULL);
+		BUG_ON(ret != 0);
+
+		vmw_fence_single_bo(bo, NULL);
+		ttm_bo_unreserve(bo);
+	}
+
+	vmw_mob_destroy(otable->page_table);
+	otable->page_table = NULL;
+}
+
+/*
+ * vmw_otables_setup - Set up guest backed memory object tables
+ *
+ * @dev_priv:       Pointer to a device private structure
+ *
+ * Takes care of the device guest backed surface
+ * initialization, by setting up the guest backed memory object tables.
+ * Returns 0 on success and various error codes on failure. A succesful return
+ * means the object tables can be taken down using the vmw_otables_takedown
+ * function.
+ */
+int vmw_otables_setup(struct vmw_private *dev_priv)
+{
+	unsigned long offset;
+	unsigned long bo_size;
+	struct vmw_otable *otables;
+	SVGAOTableType i;
+	int ret;
+
+	otables = kzalloc(SVGA_OTABLE_DX9_MAX * sizeof(*otables),
+			  GFP_KERNEL);
+	if (unlikely(otables == NULL)) {
+		DRM_ERROR("Failed to allocate space for otable "
+			  "metadata.\n");
+		return -ENOMEM;
+	}
+
+	otables[SVGA_OTABLE_MOB].size =
+		VMWGFX_NUM_MOB * SVGA3D_OTABLE_MOB_ENTRY_SIZE;
+	otables[SVGA_OTABLE_SURFACE].size =
+		VMWGFX_NUM_GB_SURFACE * SVGA3D_OTABLE_SURFACE_ENTRY_SIZE;
+	otables[SVGA_OTABLE_CONTEXT].size =
+		VMWGFX_NUM_GB_CONTEXT * SVGA3D_OTABLE_CONTEXT_ENTRY_SIZE;
+	otables[SVGA_OTABLE_SHADER].size =
+		VMWGFX_NUM_GB_SHADER * SVGA3D_OTABLE_SHADER_ENTRY_SIZE;
+	otables[SVGA_OTABLE_SCREEN_TARGET].size =
+		VMWGFX_NUM_GB_SCREEN_TARGET *
+		SVGA3D_OTABLE_SCREEN_TARGET_ENTRY_SIZE;
+
+	bo_size = 0;
+	for (i = 0; i < SVGA_OTABLE_DX9_MAX; ++i) {
+		otables[i].size =
+			(otables[i].size + PAGE_SIZE - 1) & PAGE_MASK;
+		bo_size += otables[i].size;
+	}
+
+	ret = ttm_bo_create(&dev_priv->bdev, bo_size,
+			    ttm_bo_type_device,
+			    &vmw_sys_ne_placement,
+			    0, false, NULL,
+			    &dev_priv->otable_bo);
+
+	if (unlikely(ret != 0))
+		goto out_no_bo;
+
+	ret = ttm_bo_reserve(dev_priv->otable_bo, false, true, false, NULL);
+	BUG_ON(ret != 0);
+	ret = vmw_bo_driver.ttm_tt_populate(dev_priv->otable_bo->ttm);
+	if (unlikely(ret != 0))
+		goto out_unreserve;
+	ret = vmw_bo_map_dma(dev_priv->otable_bo);
+	if (unlikely(ret != 0))
+		goto out_unreserve;
+
+	ttm_bo_unreserve(dev_priv->otable_bo);
+
+	offset = 0;
+	for (i = 0; i < SVGA_OTABLE_DX9_MAX - VMW_OTABLE_SETUP_SUB; ++i) {
+		ret = vmw_setup_otable_base(dev_priv, i, offset,
+					    &otables[i]);
+		if (unlikely(ret != 0))
+			goto out_no_setup;
+		offset += otables[i].size;
+	}
+
+	dev_priv->otables = otables;
+	return 0;
+
+out_unreserve:
+	ttm_bo_unreserve(dev_priv->otable_bo);
+out_no_setup:
+	for (i = 0; i < SVGA_OTABLE_DX9_MAX - VMW_OTABLE_SETUP_SUB; ++i)
+		vmw_takedown_otable_base(dev_priv, i, &otables[i]);
+
+	ttm_bo_unref(&dev_priv->otable_bo);
+out_no_bo:
+	kfree(otables);
+	return ret;
+}
+
+
+/*
+ * vmw_otables_takedown - Take down guest backed memory object tables
+ *
+ * @dev_priv:       Pointer to a device private structure
+ *
+ * Take down the Guest Memory Object tables.
+ */
+void vmw_otables_takedown(struct vmw_private *dev_priv)
+{
+	SVGAOTableType i;
+	struct ttm_buffer_object *bo = dev_priv->otable_bo;
+	int ret;
+
+	for (i = 0; i < SVGA_OTABLE_DX9_MAX - VMW_OTABLE_SETUP_SUB; ++i)
+		vmw_takedown_otable_base(dev_priv, i,
+					 &dev_priv->otables[i]);
+
+	ret = ttm_bo_reserve(bo, false, true, false, NULL);
+	BUG_ON(ret != 0);
+
+	vmw_fence_single_bo(bo, NULL);
+	ttm_bo_unreserve(bo);
+
+	ttm_bo_unref(&dev_priv->otable_bo);
+	kfree(dev_priv->otables);
+	dev_priv->otables = NULL;
+}
+
+
+/*
+ * vmw_mob_calculate_pt_pages - Calculate the number of page table pages
+ * needed for a guest backed memory object.
+ *
+ * @data_pages:  Number of data pages in the memory object buffer.
+ */
+static unsigned long vmw_mob_calculate_pt_pages(unsigned long data_pages)
+{
+	unsigned long data_size = data_pages * PAGE_SIZE;
+	unsigned long tot_size = 0;
+
+	while (likely(data_size > PAGE_SIZE)) {
+		data_size = DIV_ROUND_UP(data_size, PAGE_SIZE);
+		data_size *= VMW_PPN_SIZE;
+		tot_size += (data_size + PAGE_SIZE - 1) & PAGE_MASK;
+	}
+
+	return tot_size >> PAGE_SHIFT;
+}
+
+/*
+ * vmw_mob_create - Create a mob, but don't populate it.
+ *
+ * @data_pages:  Number of data pages of the underlying buffer object.
+ */
+struct vmw_mob *vmw_mob_create(unsigned long data_pages)
+{
+	struct vmw_mob *mob = kzalloc(sizeof(*mob), GFP_KERNEL);
+
+	if (unlikely(mob == NULL))
+		return NULL;
+
+	mob->num_pages = vmw_mob_calculate_pt_pages(data_pages);
+
+	return mob;
+}
+
+/*
+ * vmw_mob_pt_populate - Populate the mob pagetable
+ *
+ * @mob:         Pointer to the mob the pagetable of which we want to
+ *               populate.
+ *
+ * This function allocates memory to be used for the pagetable, and
+ * adjusts TTM memory accounting accordingly. Returns ENOMEM if
+ * memory resources aren't sufficient and may cause TTM buffer objects
+ * to be swapped out by using the TTM memory accounting function.
+ */
+static int vmw_mob_pt_populate(struct vmw_private *dev_priv,
+			       struct vmw_mob *mob)
+{
+	int ret;
+	BUG_ON(mob->pt_bo != NULL);
+
+	ret = ttm_bo_create(&dev_priv->bdev, mob->num_pages * PAGE_SIZE,
+			    ttm_bo_type_device,
+			    &vmw_sys_ne_placement,
+			    0, false, NULL, &mob->pt_bo);
+	if (unlikely(ret != 0))
+		return ret;
+
+	ret = ttm_bo_reserve(mob->pt_bo, false, true, false, NULL);
+
+	BUG_ON(ret != 0);
+	ret = vmw_bo_driver.ttm_tt_populate(mob->pt_bo->ttm);
+	if (unlikely(ret != 0))
+		goto out_unreserve;
+	ret = vmw_bo_map_dma(mob->pt_bo);
+	if (unlikely(ret != 0))
+		goto out_unreserve;
+
+	ttm_bo_unreserve(mob->pt_bo);
+	
+	return 0;
+
+out_unreserve:
+	ttm_bo_unreserve(mob->pt_bo);
+	ttm_bo_unref(&mob->pt_bo);
+
+	return ret;
+}
+
+/**
+ * vmw_mob_assign_ppn - Assign a value to a page table entry
+ *
+ * @addr: Pointer to pointer to page table entry.
+ * @val: The page table entry
+ *
+ * Assigns a value to a page table entry pointed to by *@addr and increments
+ * *@addr according to the page table entry size.
+ */
+#if (VMW_PPN_SIZE == 8)
+static void vmw_mob_assign_ppn(__le32 **addr, dma_addr_t val)
+{
+	*((__le64 *) *addr) = cpu_to_le64(val >> PAGE_SHIFT);
+	*addr += 2;
+}
+#else
+static void vmw_mob_assign_ppn(__le32 **addr, dma_addr_t val)
+{
+	*(*addr)++ = cpu_to_le32(val >> PAGE_SHIFT);
+}
+#endif
+
+/*
+ * vmw_mob_build_pt - Build a pagetable
+ *
+ * @data_addr:      Array of DMA addresses to the underlying buffer
+ *                  object's data pages.
+ * @num_data_pages: Number of buffer object data pages.
+ * @pt_pages:       Array of page pointers to the page table pages.
+ *
+ * Returns the number of page table pages actually used.
+ * Uses atomic kmaps of highmem pages to avoid TLB thrashing.
+ */
+static unsigned long vmw_mob_build_pt(struct vmw_piter *data_iter,
+				      unsigned long num_data_pages,
+				      struct vmw_piter *pt_iter)
+{
+	unsigned long pt_size = num_data_pages * VMW_PPN_SIZE;
+	unsigned long num_pt_pages = DIV_ROUND_UP(pt_size, PAGE_SIZE);
+	unsigned long pt_page;
+	__le32 *addr, *save_addr;
+	unsigned long i;
+	struct page *page;
+
+	for (pt_page = 0; pt_page < num_pt_pages; ++pt_page) {
+		page = vmw_piter_page(pt_iter);
+
+		save_addr = addr = kmap_atomic(page);
+
+		for (i = 0; i < PAGE_SIZE / VMW_PPN_SIZE; ++i) {
+			vmw_mob_assign_ppn(&addr,
+					   vmw_piter_dma_addr(data_iter));
+			if (unlikely(--num_data_pages == 0))
+				break;
+			WARN_ON(!vmw_piter_next(data_iter));
+		}
+		kunmap_atomic(save_addr);
+		vmw_piter_next(pt_iter);
+	}
+
+	return num_pt_pages;
+}
+
+/*
+ * vmw_mob_build_pt - Set up a multilevel mob pagetable
+ *
+ * @mob:            Pointer to a mob whose page table needs setting up.
+ * @data_addr       Array of DMA addresses to the buffer object's data
+ *                  pages.
+ * @num_data_pages: Number of buffer object data pages.
+ *
+ * Uses tail recursion to set up a multilevel mob page table.
+ */
+static void vmw_mob_pt_setup(struct vmw_mob *mob,
+			     struct vmw_piter data_iter,
+			     unsigned long num_data_pages)
+{
+	unsigned long num_pt_pages = 0;
+	struct ttm_buffer_object *bo = mob->pt_bo;
+	struct vmw_piter save_pt_iter;
+	struct vmw_piter pt_iter;
+	const struct vmw_sg_table *vsgt;
+	int ret;
+
+	ret = ttm_bo_reserve(bo, false, true, false, NULL);
+	BUG_ON(ret != 0);
+
+	vsgt = vmw_bo_sg_table(bo);
+	vmw_piter_start(&pt_iter, vsgt, 0);
+	BUG_ON(!vmw_piter_next(&pt_iter));
+	mob->pt_level = 0;
+	while (likely(num_data_pages > 1)) {
+		++mob->pt_level;
+		BUG_ON(mob->pt_level > 2);
+		save_pt_iter = pt_iter;
+		num_pt_pages = vmw_mob_build_pt(&data_iter, num_data_pages,
+						&pt_iter);
+		data_iter = save_pt_iter;
+		num_data_pages = num_pt_pages;
+	}
+
+	mob->pt_root_page = vmw_piter_dma_addr(&save_pt_iter);
+	ttm_bo_unreserve(bo);
+}
+
+/*
+ * vmw_mob_destroy - Destroy a mob, unpopulating first if necessary.
+ *
+ * @mob:            Pointer to a mob to destroy.
+ */
+void vmw_mob_destroy(struct vmw_mob *mob)
+{
+	if (mob->pt_bo)
+		ttm_bo_unref(&mob->pt_bo);
+	kfree(mob);
+}
+
+/*
+ * vmw_mob_unbind - Hide a mob from the device.
+ *
+ * @dev_priv:       Pointer to a device private.
+ * @mob_id:         Device id of the mob to unbind.
+ */
+void vmw_mob_unbind(struct vmw_private *dev_priv,
+		    struct vmw_mob *mob)
+{
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdDestroyGBMob body;
+	} *cmd;
+	int ret;
+	struct ttm_buffer_object *bo = mob->pt_bo;
+
+	if (bo) {
+		ret = ttm_bo_reserve(bo, false, true, false, NULL);
+		/*
+		 * Noone else should be using this buffer.
+		 */
+		BUG_ON(ret != 0);
+	}
+
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for Memory "
+			  "Object unbinding.\n");
+	}
+	cmd->header.id = SVGA_3D_CMD_DESTROY_GB_MOB;
+	cmd->header.size = sizeof(cmd->body);
+	cmd->body.mobid = mob->id;
+	vmw_fifo_commit(dev_priv, sizeof(*cmd));
+	if (bo) {
+		vmw_fence_single_bo(bo, NULL);
+		ttm_bo_unreserve(bo);
+	}
+	vmw_3d_resource_dec(dev_priv, false);
+}
+
+/*
+ * vmw_mob_bind - Make a mob visible to the device after first
+ *                populating it if necessary.
+ *
+ * @dev_priv:       Pointer to a device private.
+ * @mob:            Pointer to the mob we're making visible.
+ * @data_addr:      Array of DMA addresses to the data pages of the underlying
+ *                  buffer object.
+ * @num_data_pages: Number of data pages of the underlying buffer
+ *                  object.
+ * @mob_id:         Device id of the mob to bind
+ *
+ * This function is intended to be interfaced with the ttm_tt backend
+ * code.
+ */
+int vmw_mob_bind(struct vmw_private *dev_priv,
+		 struct vmw_mob *mob,
+		 const struct vmw_sg_table *vsgt,
+		 unsigned long num_data_pages,
+		 int32_t mob_id)
+{
+	int ret;
+	bool pt_set_up = false;
+	struct vmw_piter data_iter;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdDefineGBMob64 body;
+	} *cmd;
+
+	mob->id = mob_id;
+	vmw_piter_start(&data_iter, vsgt, 0);
+	if (unlikely(!vmw_piter_next(&data_iter)))
+		return 0;
+
+	if (likely(num_data_pages == 1)) {
+		mob->pt_level = VMW_MOBFMT_PTDEPTH_0;
+		mob->pt_root_page = vmw_piter_dma_addr(&data_iter);
+	} else if (vsgt->num_regions == 1) {
+		mob->pt_level = SVGA3D_MOBFMT_RANGE;
+		mob->pt_root_page = vmw_piter_dma_addr(&data_iter);
+	} else if (unlikely(mob->pt_bo == NULL)) {
+		ret = vmw_mob_pt_populate(dev_priv, mob);
+		if (unlikely(ret != 0))
+			return ret;
+
+		vmw_mob_pt_setup(mob, data_iter, num_data_pages);
+		pt_set_up = true;
+		mob->pt_level += VMW_MOBFMT_PTDEPTH_1 - SVGA3D_MOBFMT_PTDEPTH_1;
+	}
+
+	(void) vmw_3d_resource_inc(dev_priv, false);
+
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for Memory "
+			  "Object binding.\n");
+		goto out_no_cmd_space;
+	}
+
+	cmd->header.id = SVGA_3D_CMD_DEFINE_GB_MOB64;
+	cmd->header.size = sizeof(cmd->body);
+	cmd->body.mobid = mob_id;
+	cmd->body.ptDepth = mob->pt_level;
+	cmd->body.base = cpu_to_le64(mob->pt_root_page >> PAGE_SHIFT);
+	cmd->body.sizeInBytes = num_data_pages * PAGE_SIZE;
+
+	vmw_fifo_commit(dev_priv, sizeof(*cmd));
+
+	return 0;
+
+out_no_cmd_space:
+	vmw_3d_resource_dec(dev_priv, false);
+	if (pt_set_up)
+		ttm_bo_unref(&mob->pt_bo);
+
+	return -ENOMEM;
+}
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c b/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c
index 9b5ea2ac7ddf..6fdd82d42f65 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c
@@ -215,6 +215,7 @@ int vmw_resource_init(struct vmw_private *dev_priv, struct vmw_resource *res,
 	res->func = func;
 	INIT_LIST_HEAD(&res->lru_head);
 	INIT_LIST_HEAD(&res->mob_head);
+	INIT_LIST_HEAD(&res->binding_head);
 	res->id = -1;
 	res->backup = NULL;
 	res->backup_offset = 0;
@@ -441,6 +442,21 @@ static void vmw_user_dmabuf_release(struct ttm_base_object **p_base)
 	ttm_bo_unref(&bo);
 }
 
+static void vmw_user_dmabuf_ref_obj_release(struct ttm_base_object *base,
+					    enum ttm_ref_type ref_type)
+{
+	struct vmw_user_dma_buffer *user_bo;
+	user_bo = container_of(base, struct vmw_user_dma_buffer, prime.base);
+
+	switch (ref_type) {
+	case TTM_REF_SYNCCPU_WRITE:
+		ttm_bo_synccpu_write_release(&user_bo->dma.base);
+		break;
+	default:
+		BUG();
+	}
+}
+
 /**
  * vmw_user_dmabuf_alloc - Allocate a user dma buffer
  *
@@ -471,6 +487,8 @@ int vmw_user_dmabuf_alloc(struct vmw_private *dev_priv,
 	}
 
 	ret = vmw_dmabuf_init(dev_priv, &user_bo->dma, size,
+			      (dev_priv->has_mob) ?
+			      &vmw_sys_placement :
 			      &vmw_vram_sys_placement, true,
 			      &vmw_user_dmabuf_destroy);
 	if (unlikely(ret != 0))
@@ -482,7 +500,8 @@ int vmw_user_dmabuf_alloc(struct vmw_private *dev_priv,
 				    &user_bo->prime,
 				    shareable,
 				    ttm_buffer_type,
-				    &vmw_user_dmabuf_release, NULL);
+				    &vmw_user_dmabuf_release,
+				    &vmw_user_dmabuf_ref_obj_release);
 	if (unlikely(ret != 0)) {
 		ttm_bo_unref(&tmp);
 		goto out_no_base_object;
@@ -515,6 +534,130 @@ int vmw_user_dmabuf_verify_access(struct ttm_buffer_object *bo,
 		vmw_user_bo->prime.base.shareable) ? 0 : -EPERM;
 }
 
+/**
+ * vmw_user_dmabuf_synccpu_grab - Grab a struct vmw_user_dma_buffer for cpu
+ * access, idling previous GPU operations on the buffer and optionally
+ * blocking it for further command submissions.
+ *
+ * @user_bo: Pointer to the buffer object being grabbed for CPU access
+ * @tfile: Identifying the caller.
+ * @flags: Flags indicating how the grab should be performed.
+ *
+ * A blocking grab will be automatically released when @tfile is closed.
+ */
+static int vmw_user_dmabuf_synccpu_grab(struct vmw_user_dma_buffer *user_bo,
+					struct ttm_object_file *tfile,
+					uint32_t flags)
+{
+	struct ttm_buffer_object *bo = &user_bo->dma.base;
+	bool existed;
+	int ret;
+
+	if (flags & drm_vmw_synccpu_allow_cs) {
+		struct ttm_bo_device *bdev = bo->bdev;
+
+		spin_lock(&bdev->fence_lock);
+		ret = ttm_bo_wait(bo, false, true,
+				  !!(flags & drm_vmw_synccpu_dontblock));
+		spin_unlock(&bdev->fence_lock);
+		return ret;
+	}
+
+	ret = ttm_bo_synccpu_write_grab
+		(bo, !!(flags & drm_vmw_synccpu_dontblock));
+	if (unlikely(ret != 0))
+		return ret;
+
+	ret = ttm_ref_object_add(tfile, &user_bo->prime.base,
+				 TTM_REF_SYNCCPU_WRITE, &existed);
+	if (ret != 0 || existed)
+		ttm_bo_synccpu_write_release(&user_bo->dma.base);
+
+	return ret;
+}
+
+/**
+ * vmw_user_dmabuf_synccpu_release - Release a previous grab for CPU access,
+ * and unblock command submission on the buffer if blocked.
+ *
+ * @handle: Handle identifying the buffer object.
+ * @tfile: Identifying the caller.
+ * @flags: Flags indicating the type of release.
+ */
+static int vmw_user_dmabuf_synccpu_release(uint32_t handle,
+					   struct ttm_object_file *tfile,
+					   uint32_t flags)
+{
+	if (!(flags & drm_vmw_synccpu_allow_cs))
+		return ttm_ref_object_base_unref(tfile, handle,
+						 TTM_REF_SYNCCPU_WRITE);
+
+	return 0;
+}
+
+/**
+ * vmw_user_dmabuf_synccpu_release - ioctl function implementing the synccpu
+ * functionality.
+ *
+ * @dev: Identifies the drm device.
+ * @data: Pointer to the ioctl argument.
+ * @file_priv: Identifies the caller.
+ *
+ * This function checks the ioctl arguments for validity and calls the
+ * relevant synccpu functions.
+ */
+int vmw_user_dmabuf_synccpu_ioctl(struct drm_device *dev, void *data,
+				  struct drm_file *file_priv)
+{
+	struct drm_vmw_synccpu_arg *arg =
+		(struct drm_vmw_synccpu_arg *) data;
+	struct vmw_dma_buffer *dma_buf;
+	struct vmw_user_dma_buffer *user_bo;
+	struct ttm_object_file *tfile = vmw_fpriv(file_priv)->tfile;
+	int ret;
+
+	if ((arg->flags & (drm_vmw_synccpu_read | drm_vmw_synccpu_write)) == 0
+	    || (arg->flags & ~(drm_vmw_synccpu_read | drm_vmw_synccpu_write |
+			       drm_vmw_synccpu_dontblock |
+			       drm_vmw_synccpu_allow_cs)) != 0) {
+		DRM_ERROR("Illegal synccpu flags.\n");
+		return -EINVAL;
+	}
+
+	switch (arg->op) {
+	case drm_vmw_synccpu_grab:
+		ret = vmw_user_dmabuf_lookup(tfile, arg->handle, &dma_buf);
+		if (unlikely(ret != 0))
+			return ret;
+
+		user_bo = container_of(dma_buf, struct vmw_user_dma_buffer,
+				       dma);
+		ret = vmw_user_dmabuf_synccpu_grab(user_bo, tfile, arg->flags);
+		vmw_dmabuf_unreference(&dma_buf);
+		if (unlikely(ret != 0 && ret != -ERESTARTSYS &&
+			     ret != -EBUSY)) {
+			DRM_ERROR("Failed synccpu grab on handle 0x%08x.\n",
+				  (unsigned int) arg->handle);
+			return ret;
+		}
+		break;
+	case drm_vmw_synccpu_release:
+		ret = vmw_user_dmabuf_synccpu_release(arg->handle, tfile,
+						      arg->flags);
+		if (unlikely(ret != 0)) {
+			DRM_ERROR("Failed synccpu release on handle 0x%08x.\n",
+				  (unsigned int) arg->handle);
+			return ret;
+		}
+		break;
+	default:
+		DRM_ERROR("Invalid synccpu operation.\n");
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
 int vmw_dmabuf_alloc_ioctl(struct drm_device *dev, void *data,
 			   struct drm_file *file_priv)
 {
@@ -591,7 +734,8 @@ int vmw_user_dmabuf_lookup(struct ttm_object_file *tfile,
 }
 
 int vmw_user_dmabuf_reference(struct ttm_object_file *tfile,
-			      struct vmw_dma_buffer *dma_buf)
+			      struct vmw_dma_buffer *dma_buf,
+			      uint32_t *handle)
 {
 	struct vmw_user_dma_buffer *user_bo;
 
@@ -599,6 +743,8 @@ int vmw_user_dmabuf_reference(struct ttm_object_file *tfile,
 		return -EINVAL;
 
 	user_bo = container_of(dma_buf, struct vmw_user_dma_buffer, dma);
+
+	*handle = user_bo->prime.base.hash.key;
 	return ttm_ref_object_add(tfile, &user_bo->prime.base,
 				  TTM_REF_USAGE, NULL);
 }
@@ -1291,11 +1437,54 @@ void vmw_fence_single_bo(struct ttm_buffer_object *bo,
  * @mem:            The truct ttm_mem_reg indicating to what memory
  *                  region the move is taking place.
  *
- * For now does nothing.
+ * Evicts the Guest Backed hardware resource if the backup
+ * buffer is being moved out of MOB memory.
+ * Note that this function should not race with the resource
+ * validation code as long as it accesses only members of struct
+ * resource that remain static while bo::res is !NULL and
+ * while we have @bo reserved. struct resource::backup is *not* a
+ * static member. The resource validation code will take care
+ * to set @bo::res to NULL, while having @bo reserved when the
+ * buffer is no longer bound to the resource, so @bo:res can be
+ * used to determine whether there is a need to unbind and whether
+ * it is safe to unbind.
  */
 void vmw_resource_move_notify(struct ttm_buffer_object *bo,
 			      struct ttm_mem_reg *mem)
 {
+	struct vmw_dma_buffer *dma_buf;
+
+	if (mem == NULL)
+		return;
+
+	if (bo->destroy != vmw_dmabuf_bo_free &&
+	    bo->destroy != vmw_user_dmabuf_destroy)
+		return;
+
+	dma_buf = container_of(bo, struct vmw_dma_buffer, base);
+
+	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;
+
+		list_for_each_entry_safe(res, n, &dma_buf->res_list, mob_head) {
+
+			if (unlikely(res->func->unbind == NULL))
+				continue;
+
+			(void) res->func->unbind(res, true, &val_buf);
+			res->backup_dirty = true;
+			res->res_dirty = false;
+			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/drm/vmwgfx/vmwgfx_shader.c b/drivers/gpu/drm/vmwgfx/vmwgfx_shader.c
new file mode 100644
index 000000000000..1457ec4b7125
--- /dev/null
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_shader.c
@@ -0,0 +1,441 @@
+/**************************************************************************
+ *
+ * Copyright © 2009-2012 VMware, Inc., Palo Alto, CA., USA
+ * 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.
+ *
+ **************************************************************************/
+
+#include "vmwgfx_drv.h"
+#include "vmwgfx_resource_priv.h"
+#include "ttm/ttm_placement.h"
+
+struct vmw_shader {
+	struct vmw_resource res;
+	SVGA3dShaderType type;
+	uint32_t size;
+};
+
+struct vmw_user_shader {
+	struct ttm_base_object base;
+	struct vmw_shader shader;
+};
+
+static void vmw_user_shader_free(struct vmw_resource *res);
+static struct vmw_resource *
+vmw_user_shader_base_to_res(struct ttm_base_object *base);
+
+static int vmw_gb_shader_create(struct vmw_resource *res);
+static int vmw_gb_shader_bind(struct vmw_resource *res,
+			       struct ttm_validate_buffer *val_buf);
+static int vmw_gb_shader_unbind(struct vmw_resource *res,
+				 bool readback,
+				 struct ttm_validate_buffer *val_buf);
+static int vmw_gb_shader_destroy(struct vmw_resource *res);
+
+static uint64_t vmw_user_shader_size;
+
+static const struct vmw_user_resource_conv user_shader_conv = {
+	.object_type = VMW_RES_SHADER,
+	.base_obj_to_res = vmw_user_shader_base_to_res,
+	.res_free = vmw_user_shader_free
+};
+
+const struct vmw_user_resource_conv *user_shader_converter =
+	&user_shader_conv;
+
+
+static const struct vmw_res_func vmw_gb_shader_func = {
+	.res_type = vmw_res_shader,
+	.needs_backup = true,
+	.may_evict = true,
+	.type_name = "guest backed shaders",
+	.backup_placement = &vmw_mob_placement,
+	.create = vmw_gb_shader_create,
+	.destroy = vmw_gb_shader_destroy,
+	.bind = vmw_gb_shader_bind,
+	.unbind = vmw_gb_shader_unbind
+};
+
+/**
+ * Shader management:
+ */
+
+static inline struct vmw_shader *
+vmw_res_to_shader(struct vmw_resource *res)
+{
+	return container_of(res, struct vmw_shader, res);
+}
+
+static void vmw_hw_shader_destroy(struct vmw_resource *res)
+{
+	(void) vmw_gb_shader_destroy(res);
+}
+
+static int vmw_gb_shader_init(struct vmw_private *dev_priv,
+			      struct vmw_resource *res,
+			      uint32_t size,
+			      uint64_t offset,
+			      SVGA3dShaderType type,
+			      struct vmw_dma_buffer *byte_code,
+			      void (*res_free) (struct vmw_resource *res))
+{
+	struct vmw_shader *shader = vmw_res_to_shader(res);
+	int ret;
+
+	ret = vmw_resource_init(dev_priv, res, true,
+				res_free, &vmw_gb_shader_func);
+
+
+	if (unlikely(ret != 0)) {
+		if (res_free)
+			res_free(res);
+		else
+			kfree(res);
+		return ret;
+	}
+
+	res->backup_size = size;
+	if (byte_code) {
+		res->backup = vmw_dmabuf_reference(byte_code);
+		res->backup_offset = offset;
+	}
+	shader->size = size;
+	shader->type = type;
+
+	vmw_resource_activate(res, vmw_hw_shader_destroy);
+	return 0;
+}
+
+static int vmw_gb_shader_create(struct vmw_resource *res)
+{
+	struct vmw_private *dev_priv = res->dev_priv;
+	struct vmw_shader *shader = vmw_res_to_shader(res);
+	int ret;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdDefineGBShader body;
+	} *cmd;
+
+	if (likely(res->id != -1))
+		return 0;
+
+	ret = vmw_resource_alloc_id(res);
+	if (unlikely(ret != 0)) {
+		DRM_ERROR("Failed to allocate a shader id.\n");
+		goto out_no_id;
+	}
+
+	if (unlikely(res->id >= VMWGFX_NUM_GB_SHADER)) {
+		ret = -EBUSY;
+		goto out_no_fifo;
+	}
+
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for shader "
+			  "creation.\n");
+		ret = -ENOMEM;
+		goto out_no_fifo;
+	}
+
+	cmd->header.id = SVGA_3D_CMD_DEFINE_GB_SHADER;
+	cmd->header.size = sizeof(cmd->body);
+	cmd->body.shid = res->id;
+	cmd->body.type = shader->type;
+	cmd->body.sizeInBytes = shader->size;
+	vmw_fifo_commit(dev_priv, sizeof(*cmd));
+	(void) vmw_3d_resource_inc(dev_priv, false);
+
+	return 0;
+
+out_no_fifo:
+	vmw_resource_release_id(res);
+out_no_id:
+	return ret;
+}
+
+static int vmw_gb_shader_bind(struct vmw_resource *res,
+			      struct ttm_validate_buffer *val_buf)
+{
+	struct vmw_private *dev_priv = res->dev_priv;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdBindGBShader body;
+	} *cmd;
+	struct ttm_buffer_object *bo = val_buf->bo;
+
+	BUG_ON(bo->mem.mem_type != VMW_PL_MOB);
+
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for shader "
+			  "binding.\n");
+		return -ENOMEM;
+	}
+
+	cmd->header.id = SVGA_3D_CMD_BIND_GB_SHADER;
+	cmd->header.size = sizeof(cmd->body);
+	cmd->body.shid = res->id;
+	cmd->body.mobid = bo->mem.start;
+	cmd->body.offsetInBytes = 0;
+	res->backup_dirty = false;
+	vmw_fifo_commit(dev_priv, sizeof(*cmd));
+
+	return 0;
+}
+
+static int vmw_gb_shader_unbind(struct vmw_resource *res,
+				bool readback,
+				struct ttm_validate_buffer *val_buf)
+{
+	struct vmw_private *dev_priv = res->dev_priv;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdBindGBShader body;
+	} *cmd;
+	struct vmw_fence_obj *fence;
+
+	BUG_ON(res->backup->base.mem.mem_type != VMW_PL_MOB);
+
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for shader "
+			  "unbinding.\n");
+		return -ENOMEM;
+	}
+
+	cmd->header.id = SVGA_3D_CMD_BIND_GB_SHADER;
+	cmd->header.size = sizeof(cmd->body);
+	cmd->body.shid = res->id;
+	cmd->body.mobid = SVGA3D_INVALID_ID;
+	cmd->body.offsetInBytes = 0;
+	vmw_fifo_commit(dev_priv, sizeof(*cmd));
+
+	/*
+	 * Create a fence object and fence the backup buffer.
+	 */
+
+	(void) vmw_execbuf_fence_commands(NULL, dev_priv,
+					  &fence, NULL);
+
+	vmw_fence_single_bo(val_buf->bo, fence);
+
+	if (likely(fence != NULL))
+		vmw_fence_obj_unreference(&fence);
+
+	return 0;
+}
+
+static int vmw_gb_shader_destroy(struct vmw_resource *res)
+{
+	struct vmw_private *dev_priv = res->dev_priv;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdDestroyGBShader body;
+	} *cmd;
+
+	if (likely(res->id == -1))
+		return 0;
+
+	mutex_lock(&dev_priv->binding_mutex);
+	vmw_context_binding_res_list_kill(&res->binding_head);
+
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for shader "
+			  "destruction.\n");
+		mutex_unlock(&dev_priv->binding_mutex);
+		return -ENOMEM;
+	}
+
+	cmd->header.id = SVGA_3D_CMD_DESTROY_GB_SHADER;
+	cmd->header.size = sizeof(cmd->body);
+	cmd->body.shid = res->id;
+	vmw_fifo_commit(dev_priv, sizeof(*cmd));
+	mutex_unlock(&dev_priv->binding_mutex);
+	vmw_resource_release_id(res);
+	vmw_3d_resource_dec(dev_priv, false);
+
+	return 0;
+}
+
+/**
+ * User-space shader management:
+ */
+
+static struct vmw_resource *
+vmw_user_shader_base_to_res(struct ttm_base_object *base)
+{
+	return &(container_of(base, struct vmw_user_shader, base)->
+		 shader.res);
+}
+
+static void vmw_user_shader_free(struct vmw_resource *res)
+{
+	struct vmw_user_shader *ushader =
+		container_of(res, struct vmw_user_shader, shader.res);
+	struct vmw_private *dev_priv = res->dev_priv;
+
+	ttm_base_object_kfree(ushader, base);
+	ttm_mem_global_free(vmw_mem_glob(dev_priv),
+			    vmw_user_shader_size);
+}
+
+/**
+ * This function is called when user space has no more references on the
+ * base object. It releases the base-object's reference on the resource object.
+ */
+
+static void vmw_user_shader_base_release(struct ttm_base_object **p_base)
+{
+	struct ttm_base_object *base = *p_base;
+	struct vmw_resource *res = vmw_user_shader_base_to_res(base);
+
+	*p_base = NULL;
+	vmw_resource_unreference(&res);
+}
+
+int vmw_shader_destroy_ioctl(struct drm_device *dev, void *data,
+			      struct drm_file *file_priv)
+{
+	struct drm_vmw_shader_arg *arg = (struct drm_vmw_shader_arg *)data;
+	struct ttm_object_file *tfile = vmw_fpriv(file_priv)->tfile;
+
+	return ttm_ref_object_base_unref(tfile, arg->handle,
+					 TTM_REF_USAGE);
+}
+
+int vmw_shader_define_ioctl(struct drm_device *dev, void *data,
+			     struct drm_file *file_priv)
+{
+	struct vmw_private *dev_priv = vmw_priv(dev);
+	struct vmw_user_shader *ushader;
+	struct vmw_resource *res;
+	struct vmw_resource *tmp;
+	struct drm_vmw_shader_create_arg *arg =
+		(struct drm_vmw_shader_create_arg *)data;
+	struct ttm_object_file *tfile = vmw_fpriv(file_priv)->tfile;
+	struct vmw_master *vmaster = vmw_master(file_priv->master);
+	struct vmw_dma_buffer *buffer = NULL;
+	SVGA3dShaderType shader_type;
+	int ret;
+
+	if (arg->buffer_handle != SVGA3D_INVALID_ID) {
+		ret = vmw_user_dmabuf_lookup(tfile, arg->buffer_handle,
+					     &buffer);
+		if (unlikely(ret != 0)) {
+			DRM_ERROR("Could not find buffer for shader "
+				  "creation.\n");
+			return ret;
+		}
+
+		if ((u64)buffer->base.num_pages * PAGE_SIZE <
+		    (u64)arg->size + (u64)arg->offset) {
+			DRM_ERROR("Illegal buffer- or shader size.\n");
+			ret = -EINVAL;
+			goto out_bad_arg;
+		}
+	}
+
+	switch (arg->shader_type) {
+	case drm_vmw_shader_type_vs:
+		shader_type = SVGA3D_SHADERTYPE_VS;
+		break;
+	case drm_vmw_shader_type_ps:
+		shader_type = SVGA3D_SHADERTYPE_PS;
+		break;
+	case drm_vmw_shader_type_gs:
+		shader_type = SVGA3D_SHADERTYPE_GS;
+		break;
+	default:
+		DRM_ERROR("Illegal shader type.\n");
+		ret = -EINVAL;
+		goto out_bad_arg;
+	}
+
+	/*
+	 * Approximate idr memory usage with 128 bytes. It will be limited
+	 * by maximum number_of shaders anyway.
+	 */
+
+	if (unlikely(vmw_user_shader_size == 0))
+		vmw_user_shader_size = ttm_round_pot(sizeof(*ushader))
+			+ 128;
+
+	ret = ttm_read_lock(&vmaster->lock, true);
+	if (unlikely(ret != 0))
+		return ret;
+
+	ret = ttm_mem_global_alloc(vmw_mem_glob(dev_priv),
+				   vmw_user_shader_size,
+				   false, true);
+	if (unlikely(ret != 0)) {
+		if (ret != -ERESTARTSYS)
+			DRM_ERROR("Out of graphics memory for shader"
+				  " creation.\n");
+		goto out_unlock;
+	}
+
+	ushader = kzalloc(sizeof(*ushader), GFP_KERNEL);
+	if (unlikely(ushader == NULL)) {
+		ttm_mem_global_free(vmw_mem_glob(dev_priv),
+				    vmw_user_shader_size);
+		ret = -ENOMEM;
+		goto out_unlock;
+	}
+
+	res = &ushader->shader.res;
+	ushader->base.shareable = false;
+	ushader->base.tfile = NULL;
+
+	/*
+	 * From here on, the destructor takes over resource freeing.
+	 */
+
+	ret = vmw_gb_shader_init(dev_priv, res, arg->size,
+				 arg->offset, shader_type, buffer,
+				 vmw_user_shader_free);
+	if (unlikely(ret != 0))
+		goto out_unlock;
+
+	tmp = vmw_resource_reference(res);
+	ret = ttm_base_object_init(tfile, &ushader->base, false,
+				   VMW_RES_SHADER,
+				   &vmw_user_shader_base_release, NULL);
+
+	if (unlikely(ret != 0)) {
+		vmw_resource_unreference(&tmp);
+		goto out_err;
+	}
+
+	arg->shader_handle = ushader->base.hash.key;
+out_err:
+	vmw_resource_unreference(&res);
+out_unlock:
+	ttm_read_unlock(&vmaster->lock);
+out_bad_arg:
+	vmw_dmabuf_unreference(&buffer);
+
+	return ret;
+
+}
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c b/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c
index 7de2ea8bd553..979da1c246a5 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c
@@ -41,7 +41,6 @@ struct vmw_user_surface {
 	struct ttm_prime_object prime;
 	struct vmw_surface srf;
 	uint32_t size;
-	uint32_t backup_handle;
 };
 
 /**
@@ -68,6 +67,14 @@ static int vmw_legacy_srf_unbind(struct vmw_resource *res,
 				 struct ttm_validate_buffer *val_buf);
 static int vmw_legacy_srf_create(struct vmw_resource *res);
 static int vmw_legacy_srf_destroy(struct vmw_resource *res);
+static int vmw_gb_surface_create(struct vmw_resource *res);
+static int vmw_gb_surface_bind(struct vmw_resource *res,
+			       struct ttm_validate_buffer *val_buf);
+static int vmw_gb_surface_unbind(struct vmw_resource *res,
+				 bool readback,
+				 struct ttm_validate_buffer *val_buf);
+static int vmw_gb_surface_destroy(struct vmw_resource *res);
+
 
 static const struct vmw_user_resource_conv user_surface_conv = {
 	.object_type = VMW_RES_SURFACE,
@@ -93,6 +100,18 @@ static const struct vmw_res_func vmw_legacy_surface_func = {
 	.unbind = &vmw_legacy_srf_unbind
 };
 
+static const struct vmw_res_func vmw_gb_surface_func = {
+	.res_type = vmw_res_surface,
+	.needs_backup = true,
+	.may_evict = true,
+	.type_name = "guest backed surfaces",
+	.backup_placement = &vmw_mob_placement,
+	.create = vmw_gb_surface_create,
+	.destroy = vmw_gb_surface_destroy,
+	.bind = vmw_gb_surface_bind,
+	.unbind = vmw_gb_surface_unbind
+};
+
 /**
  * struct vmw_surface_dma - SVGA3D DMA command
  */
@@ -291,6 +310,11 @@ static void vmw_hw_surface_destroy(struct vmw_resource *res)
 	struct vmw_surface *srf;
 	void *cmd;
 
+	if (res->func->destroy == vmw_gb_surface_destroy) {
+		(void) vmw_gb_surface_destroy(res);
+		return;
+	}
+
 	if (res->id != -1) {
 
 		cmd = vmw_fifo_reserve(dev_priv, vmw_surface_destroy_size());
@@ -549,12 +573,15 @@ static int vmw_surface_init(struct vmw_private *dev_priv,
 	struct vmw_resource *res = &srf->res;
 
 	BUG_ON(res_free == NULL);
-	(void) vmw_3d_resource_inc(dev_priv, false);
+	if (!dev_priv->has_mob)
+		(void) vmw_3d_resource_inc(dev_priv, false);
 	ret = vmw_resource_init(dev_priv, res, true, res_free,
+				(dev_priv->has_mob) ? &vmw_gb_surface_func :
 				&vmw_legacy_surface_func);
 
 	if (unlikely(ret != 0)) {
-		vmw_3d_resource_dec(dev_priv, false);
+		if (!dev_priv->has_mob)
+			vmw_3d_resource_dec(dev_priv, false);
 		res_free(res);
 		return ret;
 	}
@@ -750,7 +777,7 @@ int vmw_surface_define_ioctl(struct drm_device *dev, void *data,
 
 	srf->base_size = *srf->sizes;
 	srf->autogen_filter = SVGA3D_TEX_FILTER_NONE;
-	srf->multisample_count = 1;
+	srf->multisample_count = 0;
 
 	cur_bo_offset = 0;
 	cur_offset = srf->offsets;
@@ -843,6 +870,7 @@ out_unlock:
 int vmw_surface_reference_ioctl(struct drm_device *dev, void *data,
 				struct drm_file *file_priv)
 {
+	struct vmw_private *dev_priv = vmw_priv(dev);
 	union drm_vmw_surface_reference_arg *arg =
 	    (union drm_vmw_surface_reference_arg *)data;
 	struct drm_vmw_surface_arg *req = &arg->req;
@@ -854,7 +882,7 @@ int vmw_surface_reference_ioctl(struct drm_device *dev, void *data,
 	struct ttm_base_object *base;
 	int ret = -EINVAL;
 
-	base = ttm_base_object_lookup(tfile, req->sid);
+	base = ttm_base_object_lookup_for_ref(dev_priv->tdev, req->sid);
 	if (unlikely(base == NULL)) {
 		DRM_ERROR("Could not find surface to reference.\n");
 		return -EINVAL;
@@ -893,3 +921,436 @@ out_no_reference:
 
 	return ret;
 }
+
+/**
+ * vmw_surface_define_encode - Encode a surface_define command.
+ *
+ * @srf: Pointer to a struct vmw_surface object.
+ * @cmd_space: Pointer to memory area in which the commands should be encoded.
+ */
+static int vmw_gb_surface_create(struct vmw_resource *res)
+{
+	struct vmw_private *dev_priv = res->dev_priv;
+	struct vmw_surface *srf = vmw_res_to_srf(res);
+	uint32_t cmd_len, submit_len;
+	int ret;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdDefineGBSurface body;
+	} *cmd;
+
+	if (likely(res->id != -1))
+		return 0;
+
+	(void) vmw_3d_resource_inc(dev_priv, false);
+	ret = vmw_resource_alloc_id(res);
+	if (unlikely(ret != 0)) {
+		DRM_ERROR("Failed to allocate a surface id.\n");
+		goto out_no_id;
+	}
+
+	if (unlikely(res->id >= VMWGFX_NUM_GB_SURFACE)) {
+		ret = -EBUSY;
+		goto out_no_fifo;
+	}
+
+	cmd_len = sizeof(cmd->body);
+	submit_len = sizeof(*cmd);
+	cmd = vmw_fifo_reserve(dev_priv, submit_len);
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for surface "
+			  "creation.\n");
+		ret = -ENOMEM;
+		goto out_no_fifo;
+	}
+
+	cmd->header.id = SVGA_3D_CMD_DEFINE_GB_SURFACE;
+	cmd->header.size = cmd_len;
+	cmd->body.sid = srf->res.id;
+	cmd->body.surfaceFlags = srf->flags;
+	cmd->body.format = cpu_to_le32(srf->format);
+	cmd->body.numMipLevels = srf->mip_levels[0];
+	cmd->body.multisampleCount = srf->multisample_count;
+	cmd->body.autogenFilter = srf->autogen_filter;
+	cmd->body.size.width = srf->base_size.width;
+	cmd->body.size.height = srf->base_size.height;
+	cmd->body.size.depth = srf->base_size.depth;
+	vmw_fifo_commit(dev_priv, submit_len);
+
+	return 0;
+
+out_no_fifo:
+	vmw_resource_release_id(res);
+out_no_id:
+	vmw_3d_resource_dec(dev_priv, false);
+	return ret;
+}
+
+
+static int vmw_gb_surface_bind(struct vmw_resource *res,
+			       struct ttm_validate_buffer *val_buf)
+{
+	struct vmw_private *dev_priv = res->dev_priv;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdBindGBSurface body;
+	} *cmd1;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdUpdateGBSurface body;
+	} *cmd2;
+	uint32_t submit_size;
+	struct ttm_buffer_object *bo = val_buf->bo;
+
+	BUG_ON(bo->mem.mem_type != VMW_PL_MOB);
+
+	submit_size = sizeof(*cmd1) + (res->backup_dirty ? sizeof(*cmd2) : 0);
+
+	cmd1 = vmw_fifo_reserve(dev_priv, submit_size);
+	if (unlikely(cmd1 == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for surface "
+			  "binding.\n");
+		return -ENOMEM;
+	}
+
+	cmd1->header.id = SVGA_3D_CMD_BIND_GB_SURFACE;
+	cmd1->header.size = sizeof(cmd1->body);
+	cmd1->body.sid = res->id;
+	cmd1->body.mobid = bo->mem.start;
+	if (res->backup_dirty) {
+		cmd2 = (void *) &cmd1[1];
+		cmd2->header.id = SVGA_3D_CMD_UPDATE_GB_SURFACE;
+		cmd2->header.size = sizeof(cmd2->body);
+		cmd2->body.sid = res->id;
+		res->backup_dirty = false;
+	}
+	vmw_fifo_commit(dev_priv, submit_size);
+
+	return 0;
+}
+
+static int vmw_gb_surface_unbind(struct vmw_resource *res,
+				 bool readback,
+				 struct ttm_validate_buffer *val_buf)
+{
+	struct vmw_private *dev_priv = res->dev_priv;
+	struct ttm_buffer_object *bo = val_buf->bo;
+	struct vmw_fence_obj *fence;
+
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdReadbackGBSurface body;
+	} *cmd1;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdInvalidateGBSurface body;
+	} *cmd2;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdBindGBSurface body;
+	} *cmd3;
+	uint32_t submit_size;
+	uint8_t *cmd;
+
+
+	BUG_ON(bo->mem.mem_type != VMW_PL_MOB);
+
+	submit_size = sizeof(*cmd3) + (readback ? sizeof(*cmd1) : sizeof(*cmd2));
+	cmd = vmw_fifo_reserve(dev_priv, submit_size);
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for surface "
+			  "unbinding.\n");
+		return -ENOMEM;
+	}
+
+	if (readback) {
+		cmd1 = (void *) cmd;
+		cmd1->header.id = SVGA_3D_CMD_READBACK_GB_SURFACE;
+		cmd1->header.size = sizeof(cmd1->body);
+		cmd1->body.sid = res->id;
+		cmd3 = (void *) &cmd1[1];
+	} else {
+		cmd2 = (void *) cmd;
+		cmd2->header.id = SVGA_3D_CMD_INVALIDATE_GB_SURFACE;
+		cmd2->header.size = sizeof(cmd2->body);
+		cmd2->body.sid = res->id;
+		cmd3 = (void *) &cmd2[1];
+	}
+
+	cmd3->header.id = SVGA_3D_CMD_BIND_GB_SURFACE;
+	cmd3->header.size = sizeof(cmd3->body);
+	cmd3->body.sid = res->id;
+	cmd3->body.mobid = SVGA3D_INVALID_ID;
+
+	vmw_fifo_commit(dev_priv, submit_size);
+
+	/*
+	 * Create a fence object and fence the backup buffer.
+	 */
+
+	(void) vmw_execbuf_fence_commands(NULL, dev_priv,
+					  &fence, NULL);
+
+	vmw_fence_single_bo(val_buf->bo, fence);
+
+	if (likely(fence != NULL))
+		vmw_fence_obj_unreference(&fence);
+
+	return 0;
+}
+
+static int vmw_gb_surface_destroy(struct vmw_resource *res)
+{
+	struct vmw_private *dev_priv = res->dev_priv;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdDestroyGBSurface body;
+	} *cmd;
+
+	if (likely(res->id == -1))
+		return 0;
+
+	mutex_lock(&dev_priv->binding_mutex);
+	vmw_context_binding_res_list_kill(&res->binding_head);
+
+	cmd = vmw_fifo_reserve(dev_priv, sizeof(*cmd));
+	if (unlikely(cmd == NULL)) {
+		DRM_ERROR("Failed reserving FIFO space for surface "
+			  "destruction.\n");
+		mutex_unlock(&dev_priv->binding_mutex);
+		return -ENOMEM;
+	}
+
+	cmd->header.id = SVGA_3D_CMD_DESTROY_GB_SURFACE;
+	cmd->header.size = sizeof(cmd->body);
+	cmd->body.sid = res->id;
+	vmw_fifo_commit(dev_priv, sizeof(*cmd));
+	mutex_unlock(&dev_priv->binding_mutex);
+	vmw_resource_release_id(res);
+	vmw_3d_resource_dec(dev_priv, false);
+
+	return 0;
+}
+
+/**
+ * vmw_gb_surface_define_ioctl - Ioctl function implementing
+ *                               the user surface define functionality.
+ *
+ * @dev:            Pointer to a struct drm_device.
+ * @data:           Pointer to data copied from / to user-space.
+ * @file_priv:      Pointer to a drm file private structure.
+ */
+int vmw_gb_surface_define_ioctl(struct drm_device *dev, void *data,
+				struct drm_file *file_priv)
+{
+	struct vmw_private *dev_priv = vmw_priv(dev);
+	struct vmw_user_surface *user_srf;
+	struct vmw_surface *srf;
+	struct vmw_resource *res;
+	struct vmw_resource *tmp;
+	union drm_vmw_gb_surface_create_arg *arg =
+	    (union drm_vmw_gb_surface_create_arg *)data;
+	struct drm_vmw_gb_surface_create_req *req = &arg->req;
+	struct drm_vmw_gb_surface_create_rep *rep = &arg->rep;
+	struct ttm_object_file *tfile = vmw_fpriv(file_priv)->tfile;
+	int ret;
+	uint32_t size;
+	struct vmw_master *vmaster = vmw_master(file_priv->master);
+	const struct svga3d_surface_desc *desc;
+	uint32_t backup_handle;
+
+	if (unlikely(vmw_user_surface_size == 0))
+		vmw_user_surface_size = ttm_round_pot(sizeof(*user_srf)) +
+			128;
+
+	size = vmw_user_surface_size + 128;
+
+	desc = svga3dsurface_get_desc(req->format);
+	if (unlikely(desc->block_desc == SVGA3DBLOCKDESC_NONE)) {
+		DRM_ERROR("Invalid surface format for surface creation.\n");
+		return -EINVAL;
+	}
+
+	ret = ttm_read_lock(&vmaster->lock, true);
+	if (unlikely(ret != 0))
+		return ret;
+
+	ret = ttm_mem_global_alloc(vmw_mem_glob(dev_priv),
+				   size, false, true);
+	if (unlikely(ret != 0)) {
+		if (ret != -ERESTARTSYS)
+			DRM_ERROR("Out of graphics memory for surface"
+				  " creation.\n");
+		goto out_unlock;
+	}
+
+	user_srf = kzalloc(sizeof(*user_srf), GFP_KERNEL);
+	if (unlikely(user_srf == NULL)) {
+		ret = -ENOMEM;
+		goto out_no_user_srf;
+	}
+
+	srf = &user_srf->srf;
+	res = &srf->res;
+
+	srf->flags = req->svga3d_flags;
+	srf->format = req->format;
+	srf->scanout = req->drm_surface_flags & drm_vmw_surface_flag_scanout;
+	srf->mip_levels[0] = req->mip_levels;
+	srf->num_sizes = 1;
+	srf->sizes = NULL;
+	srf->offsets = NULL;
+	user_srf->size = size;
+	srf->base_size = req->base_size;
+	srf->autogen_filter = SVGA3D_TEX_FILTER_NONE;
+	srf->multisample_count = req->multisample_count;
+	res->backup_size = svga3dsurface_get_serialized_size
+	  (srf->format, srf->base_size, srf->mip_levels[0],
+	   srf->flags & SVGA3D_SURFACE_CUBEMAP);
+
+	user_srf->prime.base.shareable = false;
+	user_srf->prime.base.tfile = NULL;
+
+	/**
+	 * From this point, the generic resource management functions
+	 * destroy the object on failure.
+	 */
+
+	ret = vmw_surface_init(dev_priv, srf, vmw_user_surface_free);
+	if (unlikely(ret != 0))
+		goto out_unlock;
+
+	if (req->buffer_handle != SVGA3D_INVALID_ID) {
+		ret = vmw_user_dmabuf_lookup(tfile, req->buffer_handle,
+					     &res->backup);
+	} else if (req->drm_surface_flags &
+		   drm_vmw_surface_flag_create_buffer)
+		ret = vmw_user_dmabuf_alloc(dev_priv, tfile,
+					    res->backup_size,
+					    req->drm_surface_flags &
+					    drm_vmw_surface_flag_shareable,
+					    &backup_handle,
+					    &res->backup);
+
+	if (unlikely(ret != 0)) {
+		vmw_resource_unreference(&res);
+		goto out_unlock;
+	}
+
+	tmp = vmw_resource_reference(&srf->res);
+	ret = ttm_prime_object_init(tfile, res->backup_size, &user_srf->prime,
+				    req->drm_surface_flags &
+				    drm_vmw_surface_flag_shareable,
+				    VMW_RES_SURFACE,
+				    &vmw_user_surface_base_release, NULL);
+
+	if (unlikely(ret != 0)) {
+		vmw_resource_unreference(&tmp);
+		vmw_resource_unreference(&res);
+		goto out_unlock;
+	}
+
+	rep->handle = user_srf->prime.base.hash.key;
+	rep->backup_size = res->backup_size;
+	if (res->backup) {
+		rep->buffer_map_handle =
+			drm_vma_node_offset_addr(&res->backup->base.vma_node);
+		rep->buffer_size = res->backup->base.num_pages * PAGE_SIZE;
+		rep->buffer_handle = backup_handle;
+	} else {
+		rep->buffer_map_handle = 0;
+		rep->buffer_size = 0;
+		rep->buffer_handle = SVGA3D_INVALID_ID;
+	}
+
+	vmw_resource_unreference(&res);
+
+	ttm_read_unlock(&vmaster->lock);
+	return 0;
+out_no_user_srf:
+	ttm_mem_global_free(vmw_mem_glob(dev_priv), size);
+out_unlock:
+	ttm_read_unlock(&vmaster->lock);
+	return ret;
+}
+
+/**
+ * vmw_gb_surface_reference_ioctl - Ioctl function implementing
+ *                                  the user surface reference functionality.
+ *
+ * @dev:            Pointer to a struct drm_device.
+ * @data:           Pointer to data copied from / to user-space.
+ * @file_priv:      Pointer to a drm file private structure.
+ */
+int vmw_gb_surface_reference_ioctl(struct drm_device *dev, void *data,
+				   struct drm_file *file_priv)
+{
+	struct vmw_private *dev_priv = vmw_priv(dev);
+	union drm_vmw_gb_surface_reference_arg *arg =
+	    (union drm_vmw_gb_surface_reference_arg *)data;
+	struct drm_vmw_surface_arg *req = &arg->req;
+	struct drm_vmw_gb_surface_ref_rep *rep = &arg->rep;
+	struct ttm_object_file *tfile = vmw_fpriv(file_priv)->tfile;
+	struct vmw_surface *srf;
+	struct vmw_user_surface *user_srf;
+	struct ttm_base_object *base;
+	uint32_t backup_handle;
+	int ret = -EINVAL;
+
+	base = ttm_base_object_lookup_for_ref(dev_priv->tdev, req->sid);
+	if (unlikely(base == NULL)) {
+		DRM_ERROR("Could not find surface to reference.\n");
+		return -EINVAL;
+	}
+
+	if (unlikely(ttm_base_object_type(base) != VMW_RES_SURFACE))
+		goto out_bad_resource;
+
+	user_srf = container_of(base, struct vmw_user_surface, prime.base);
+	srf = &user_srf->srf;
+	if (srf->res.backup == NULL) {
+		DRM_ERROR("Shared GB surface is missing a backup buffer.\n");
+		goto out_bad_resource;
+	}
+
+	ret = ttm_ref_object_add(tfile, &user_srf->prime.base,
+				 TTM_REF_USAGE, NULL);
+	if (unlikely(ret != 0)) {
+		DRM_ERROR("Could not add a reference to a GB surface.\n");
+		goto out_bad_resource;
+	}
+
+	mutex_lock(&dev_priv->cmdbuf_mutex); /* Protect res->backup */
+	ret = vmw_user_dmabuf_reference(tfile, srf->res.backup,
+					&backup_handle);
+	mutex_unlock(&dev_priv->cmdbuf_mutex);
+
+	if (unlikely(ret != 0)) {
+		DRM_ERROR("Could not add a reference to a GB surface "
+			  "backup buffer.\n");
+		(void) ttm_ref_object_base_unref(vmw_fpriv(file_priv)->tfile,
+						 req->sid,
+						 TTM_REF_USAGE);
+		goto out_bad_resource;
+	}
+
+	rep->creq.svga3d_flags = srf->flags;
+	rep->creq.format = srf->format;
+	rep->creq.mip_levels = srf->mip_levels[0];
+	rep->creq.drm_surface_flags = 0;
+	rep->creq.multisample_count = srf->multisample_count;
+	rep->creq.autogen_filter = srf->autogen_filter;
+	rep->creq.buffer_handle = backup_handle;
+	rep->creq.base_size = srf->base_size;
+	rep->crep.handle = user_srf->prime.base.hash.key;
+	rep->crep.backup_size = srf->res.backup_size;
+	rep->crep.buffer_handle = backup_handle;
+	rep->crep.buffer_map_handle =
+		drm_vma_node_offset_addr(&srf->res.backup->base.vma_node);
+	rep->crep.buffer_size = srf->res.backup->base.num_pages * PAGE_SIZE;
+
+out_bad_resource:
+	ttm_base_object_unref(&base);
+
+	return ret;
+}
diff --git a/drivers/gpu/host1x/Kconfig b/drivers/gpu/host1x/Kconfig
index 7d6bed222542..b2fd029d67b3 100644
--- a/drivers/gpu/host1x/Kconfig
+++ b/drivers/gpu/host1x/Kconfig
@@ -1,6 +1,6 @@
 config TEGRA_HOST1X
 	tristate "NVIDIA Tegra host1x driver"
-	depends on ARCH_TEGRA || ARCH_MULTIPLATFORM
+	depends on ARCH_TEGRA || (ARM && COMPILE_TEST)
 	help
 	  Driver for the NVIDIA Tegra host1x hardware.
 
diff --git a/drivers/gpu/host1x/Makefile b/drivers/gpu/host1x/Makefile
index afa1e9e4e512..c1189f004441 100644
--- a/drivers/gpu/host1x/Makefile
+++ b/drivers/gpu/host1x/Makefile
@@ -7,7 +7,9 @@ host1x-y = \
 	channel.o \
 	job.o \
 	debug.o \
+	mipi.o \
 	hw/host1x01.o \
-	hw/host1x02.o
+	hw/host1x02.o \
+	hw/host1x04.o
 
 obj-$(CONFIG_TEGRA_HOST1X) += host1x.o
diff --git a/drivers/gpu/host1x/bus.c b/drivers/gpu/host1x/bus.c
index 6a929591aa73..ccdd2e6da5e3 100644
--- a/drivers/gpu/host1x/bus.c
+++ b/drivers/gpu/host1x/bus.c
@@ -188,6 +188,7 @@ int host1x_device_init(struct host1x_device *device)
 
 	return 0;
 }
+EXPORT_SYMBOL(host1x_device_init);
 
 int host1x_device_exit(struct host1x_device *device)
 {
@@ -213,6 +214,7 @@ int host1x_device_exit(struct host1x_device *device)
 
 	return 0;
 }
+EXPORT_SYMBOL(host1x_device_exit);
 
 static int host1x_register_client(struct host1x *host1x,
 				  struct host1x_client *client)
diff --git a/drivers/gpu/host1x/channel.c b/drivers/gpu/host1x/channel.c
index 83ea51b9f0fc..b4ae3affb987 100644
--- a/drivers/gpu/host1x/channel.c
+++ b/drivers/gpu/host1x/channel.c
@@ -43,6 +43,7 @@ int host1x_job_submit(struct host1x_job *job)
 
 	return host1x_hw_channel_submit(host, job);
 }
+EXPORT_SYMBOL(host1x_job_submit);
 
 struct host1x_channel *host1x_channel_get(struct host1x_channel *channel)
 {
@@ -60,6 +61,7 @@ struct host1x_channel *host1x_channel_get(struct host1x_channel *channel)
 
 	return err ? NULL : channel;
 }
+EXPORT_SYMBOL(host1x_channel_get);
 
 void host1x_channel_put(struct host1x_channel *channel)
 {
@@ -76,6 +78,7 @@ void host1x_channel_put(struct host1x_channel *channel)
 
 	mutex_unlock(&channel->reflock);
 }
+EXPORT_SYMBOL(host1x_channel_put);
 
 struct host1x_channel *host1x_channel_request(struct device *dev)
 {
@@ -115,6 +118,7 @@ fail:
 	mutex_unlock(&host->chlist_mutex);
 	return NULL;
 }
+EXPORT_SYMBOL(host1x_channel_request);
 
 void host1x_channel_free(struct host1x_channel *channel)
 {
@@ -124,3 +128,4 @@ void host1x_channel_free(struct host1x_channel *channel)
 	list_del(&channel->list);
 	kfree(channel);
 }
+EXPORT_SYMBOL(host1x_channel_free);
diff --git a/drivers/gpu/host1x/debug.c b/drivers/gpu/host1x/debug.c
index 3ec7d77de24d..ee3d12b51c50 100644
--- a/drivers/gpu/host1x/debug.c
+++ b/drivers/gpu/host1x/debug.c
@@ -96,7 +96,6 @@ static void show_all(struct host1x *m, struct output *o)
 		show_channels(ch, o, true);
 }
 
-#ifdef CONFIG_DEBUG_FS
 static void show_all_no_fifo(struct host1x *host1x, struct output *o)
 {
 	struct host1x_channel *ch;
@@ -153,7 +152,7 @@ static const struct file_operations host1x_debug_fops = {
 	.release	= single_release,
 };
 
-void host1x_debug_init(struct host1x *host1x)
+static void host1x_debugfs_init(struct host1x *host1x)
 {
 	struct dentry *de = debugfs_create_dir("tegra-host1x", NULL);
 
@@ -180,18 +179,22 @@ void host1x_debug_init(struct host1x *host1x)
 			   &host1x_debug_force_timeout_channel);
 }
 
-void host1x_debug_deinit(struct host1x *host1x)
+static void host1x_debugfs_exit(struct host1x *host1x)
 {
 	debugfs_remove_recursive(host1x->debugfs);
 }
-#else
+
 void host1x_debug_init(struct host1x *host1x)
 {
+	if (IS_ENABLED(CONFIG_DEBUG_FS))
+		host1x_debugfs_init(host1x);
 }
+
 void host1x_debug_deinit(struct host1x *host1x)
 {
+	if (IS_ENABLED(CONFIG_DEBUG_FS))
+		host1x_debugfs_exit(host1x);
 }
-#endif
 
 void host1x_debug_dump(struct host1x *host1x)
 {
diff --git a/drivers/gpu/host1x/dev.c b/drivers/gpu/host1x/dev.c
index 80da003d63de..2529908d304b 100644
--- a/drivers/gpu/host1x/dev.c
+++ b/drivers/gpu/host1x/dev.c
@@ -34,6 +34,7 @@
 #include "debug.h"
 #include "hw/host1x01.h"
 #include "hw/host1x02.h"
+#include "hw/host1x04.h"
 
 void host1x_sync_writel(struct host1x *host1x, u32 v, u32 r)
 {
@@ -77,7 +78,17 @@ static const struct host1x_info host1x02_info = {
 	.sync_offset = 0x3000,
 };
 
+static const struct host1x_info host1x04_info = {
+	.nb_channels = 12,
+	.nb_pts = 192,
+	.nb_mlocks = 16,
+	.nb_bases = 64,
+	.init = host1x04_init,
+	.sync_offset = 0x2100,
+};
+
 static struct of_device_id host1x_of_match[] = {
+	{ .compatible = "nvidia,tegra124-host1x", .data = &host1x04_info, },
 	{ .compatible = "nvidia,tegra114-host1x", .data = &host1x02_info, },
 	{ .compatible = "nvidia,tegra30-host1x", .data = &host1x01_info, },
 	{ .compatible = "nvidia,tegra20-host1x", .data = &host1x01_info, },
@@ -210,17 +221,26 @@ static int __init tegra_host1x_init(void)
 		return err;
 
 	err = platform_driver_register(&tegra_host1x_driver);
-	if (err < 0) {
-		host1x_bus_exit();
-		return err;
-	}
+	if (err < 0)
+		goto unregister_bus;
+
+	err = platform_driver_register(&tegra_mipi_driver);
+	if (err < 0)
+		goto unregister_host1x;
 
 	return 0;
+
+unregister_host1x:
+	platform_driver_unregister(&tegra_host1x_driver);
+unregister_bus:
+	host1x_bus_exit();
+	return err;
 }
 module_init(tegra_host1x_init);
 
 static void __exit tegra_host1x_exit(void)
 {
+	platform_driver_unregister(&tegra_mipi_driver);
 	platform_driver_unregister(&tegra_host1x_driver);
 	host1x_bus_exit();
 }
diff --git a/drivers/gpu/host1x/dev.h b/drivers/gpu/host1x/dev.h
index a61a976e7a42..0b6e8e9629c5 100644
--- a/drivers/gpu/host1x/dev.h
+++ b/drivers/gpu/host1x/dev.h
@@ -306,4 +306,6 @@ static inline void host1x_hw_show_mlocks(struct host1x *host, struct output *o)
 	host->debug_op->show_mlocks(host, o);
 }
 
+extern struct platform_driver tegra_mipi_driver;
+
 #endif
diff --git a/drivers/gpu/host1x/hw/host1x02.c b/drivers/gpu/host1x/hw/host1x02.c
index e98caca0ca42..928946c2144b 100644
--- a/drivers/gpu/host1x/hw/host1x02.c
+++ b/drivers/gpu/host1x/hw/host1x02.c
@@ -17,8 +17,8 @@
  */
 
 /* include hw specification */
-#include "host1x01.h"
-#include "host1x01_hardware.h"
+#include "host1x02.h"
+#include "host1x02_hardware.h"
 
 /* include code */
 #include "cdma_hw.c"
diff --git a/drivers/gpu/host1x/hw/host1x02_hardware.h b/drivers/gpu/host1x/hw/host1x02_hardware.h
new file mode 100644
index 000000000000..154901860bc6
--- /dev/null
+++ b/drivers/gpu/host1x/hw/host1x02_hardware.h
@@ -0,0 +1,142 @@
+/*
+ * Tegra host1x Register Offsets for Tegra114
+ *
+ * Copyright (c) 2010-2013 NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope 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/>.
+ */
+
+#ifndef __HOST1X_HOST1X02_HARDWARE_H
+#define __HOST1X_HOST1X02_HARDWARE_H
+
+#include <linux/types.h>
+#include <linux/bitops.h>
+
+#include "hw_host1x02_channel.h"
+#include "hw_host1x02_sync.h"
+#include "hw_host1x02_uclass.h"
+
+static inline u32 host1x_class_host_wait_syncpt(
+	unsigned indx, unsigned threshold)
+{
+	return host1x_uclass_wait_syncpt_indx_f(indx)
+		| host1x_uclass_wait_syncpt_thresh_f(threshold);
+}
+
+static inline u32 host1x_class_host_load_syncpt_base(
+	unsigned indx, unsigned threshold)
+{
+	return host1x_uclass_load_syncpt_base_base_indx_f(indx)
+		| host1x_uclass_load_syncpt_base_value_f(threshold);
+}
+
+static inline u32 host1x_class_host_wait_syncpt_base(
+	unsigned indx, unsigned base_indx, unsigned offset)
+{
+	return host1x_uclass_wait_syncpt_base_indx_f(indx)
+		| host1x_uclass_wait_syncpt_base_base_indx_f(base_indx)
+		| host1x_uclass_wait_syncpt_base_offset_f(offset);
+}
+
+static inline u32 host1x_class_host_incr_syncpt_base(
+	unsigned base_indx, unsigned offset)
+{
+	return host1x_uclass_incr_syncpt_base_base_indx_f(base_indx)
+		| host1x_uclass_incr_syncpt_base_offset_f(offset);
+}
+
+static inline u32 host1x_class_host_incr_syncpt(
+	unsigned cond, unsigned indx)
+{
+	return host1x_uclass_incr_syncpt_cond_f(cond)
+		| host1x_uclass_incr_syncpt_indx_f(indx);
+}
+
+static inline u32 host1x_class_host_indoff_reg_write(
+	unsigned mod_id, unsigned offset, bool auto_inc)
+{
+	u32 v = host1x_uclass_indoff_indbe_f(0xf)
+		| host1x_uclass_indoff_indmodid_f(mod_id)
+		| host1x_uclass_indoff_indroffset_f(offset);
+	if (auto_inc)
+		v |= host1x_uclass_indoff_autoinc_f(1);
+	return v;
+}
+
+static inline u32 host1x_class_host_indoff_reg_read(
+	unsigned mod_id, unsigned offset, bool auto_inc)
+{
+	u32 v = host1x_uclass_indoff_indmodid_f(mod_id)
+		| host1x_uclass_indoff_indroffset_f(offset)
+		| host1x_uclass_indoff_rwn_read_v();
+	if (auto_inc)
+		v |= host1x_uclass_indoff_autoinc_f(1);
+	return v;
+}
+
+/* cdma opcodes */
+static inline u32 host1x_opcode_setclass(
+	unsigned class_id, unsigned offset, unsigned mask)
+{
+	return (0 << 28) | (offset << 16) | (class_id << 6) | mask;
+}
+
+static inline u32 host1x_opcode_incr(unsigned offset, unsigned count)
+{
+	return (1 << 28) | (offset << 16) | count;
+}
+
+static inline u32 host1x_opcode_nonincr(unsigned offset, unsigned count)
+{
+	return (2 << 28) | (offset << 16) | count;
+}
+
+static inline u32 host1x_opcode_mask(unsigned offset, unsigned mask)
+{
+	return (3 << 28) | (offset << 16) | mask;
+}
+
+static inline u32 host1x_opcode_imm(unsigned offset, unsigned value)
+{
+	return (4 << 28) | (offset << 16) | value;
+}
+
+static inline u32 host1x_opcode_imm_incr_syncpt(unsigned cond, unsigned indx)
+{
+	return host1x_opcode_imm(host1x_uclass_incr_syncpt_r(),
+		host1x_class_host_incr_syncpt(cond, indx));
+}
+
+static inline u32 host1x_opcode_restart(unsigned address)
+{
+	return (5 << 28) | (address >> 4);
+}
+
+static inline u32 host1x_opcode_gather(unsigned count)
+{
+	return (6 << 28) | count;
+}
+
+static inline u32 host1x_opcode_gather_nonincr(unsigned offset,	unsigned count)
+{
+	return (6 << 28) | (offset << 16) | BIT(15) | count;
+}
+
+static inline u32 host1x_opcode_gather_incr(unsigned offset, unsigned count)
+{
+	return (6 << 28) | (offset << 16) | BIT(15) | BIT(14) | count;
+}
+
+#define HOST1X_OPCODE_NOP host1x_opcode_nonincr(0, 0)
+
+#endif
diff --git a/drivers/gpu/host1x/hw/host1x04.c b/drivers/gpu/host1x/hw/host1x04.c
new file mode 100644
index 000000000000..8007c70fa9c4
--- /dev/null
+++ b/drivers/gpu/host1x/hw/host1x04.c
@@ -0,0 +1,42 @@
+/*
+ * Host1x init for Tegra124 SoCs
+ *
+ * Copyright (c) 2013 NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope 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 hw specification */
+#include "host1x04.h"
+#include "host1x04_hardware.h"
+
+/* include code */
+#include "cdma_hw.c"
+#include "channel_hw.c"
+#include "debug_hw.c"
+#include "intr_hw.c"
+#include "syncpt_hw.c"
+
+#include "../dev.h"
+
+int host1x04_init(struct host1x *host)
+{
+	host->channel_op = &host1x_channel_ops;
+	host->cdma_op = &host1x_cdma_ops;
+	host->cdma_pb_op = &host1x_pushbuffer_ops;
+	host->syncpt_op = &host1x_syncpt_ops;
+	host->intr_op = &host1x_intr_ops;
+	host->debug_op = &host1x_debug_ops;
+
+	return 0;
+}
diff --git a/drivers/gpu/host1x/hw/host1x04.h b/drivers/gpu/host1x/hw/host1x04.h
new file mode 100644
index 000000000000..a9ab7496c06e
--- /dev/null
+++ b/drivers/gpu/host1x/hw/host1x04.h
@@ -0,0 +1,26 @@
+/*
+ * Host1x init for Tegra124 SoCs
+ *
+ * Copyright (c) 2013 NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope 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/>.
+ */
+
+#ifndef HOST1X_HOST1X04_H
+#define HOST1X_HOST1X04_H
+
+struct host1x;
+
+int host1x04_init(struct host1x *host);
+
+#endif
diff --git a/drivers/gpu/host1x/hw/host1x04_hardware.h b/drivers/gpu/host1x/hw/host1x04_hardware.h
new file mode 100644
index 000000000000..de1a38175328
--- /dev/null
+++ b/drivers/gpu/host1x/hw/host1x04_hardware.h
@@ -0,0 +1,142 @@
+/*
+ * Tegra host1x Register Offsets for Tegra124
+ *
+ * Copyright (c) 2010-2013 NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope 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/>.
+ */
+
+#ifndef __HOST1X_HOST1X04_HARDWARE_H
+#define __HOST1X_HOST1X04_HARDWARE_H
+
+#include <linux/types.h>
+#include <linux/bitops.h>
+
+#include "hw_host1x04_channel.h"
+#include "hw_host1x04_sync.h"
+#include "hw_host1x04_uclass.h"
+
+static inline u32 host1x_class_host_wait_syncpt(
+	unsigned indx, unsigned threshold)
+{
+	return host1x_uclass_wait_syncpt_indx_f(indx)
+		| host1x_uclass_wait_syncpt_thresh_f(threshold);
+}
+
+static inline u32 host1x_class_host_load_syncpt_base(
+	unsigned indx, unsigned threshold)
+{
+	return host1x_uclass_load_syncpt_base_base_indx_f(indx)
+		| host1x_uclass_load_syncpt_base_value_f(threshold);
+}
+
+static inline u32 host1x_class_host_wait_syncpt_base(
+	unsigned indx, unsigned base_indx, unsigned offset)
+{
+	return host1x_uclass_wait_syncpt_base_indx_f(indx)
+		| host1x_uclass_wait_syncpt_base_base_indx_f(base_indx)
+		| host1x_uclass_wait_syncpt_base_offset_f(offset);
+}
+
+static inline u32 host1x_class_host_incr_syncpt_base(
+	unsigned base_indx, unsigned offset)
+{
+	return host1x_uclass_incr_syncpt_base_base_indx_f(base_indx)
+		| host1x_uclass_incr_syncpt_base_offset_f(offset);
+}
+
+static inline u32 host1x_class_host_incr_syncpt(
+	unsigned cond, unsigned indx)
+{
+	return host1x_uclass_incr_syncpt_cond_f(cond)
+		| host1x_uclass_incr_syncpt_indx_f(indx);
+}
+
+static inline u32 host1x_class_host_indoff_reg_write(
+	unsigned mod_id, unsigned offset, bool auto_inc)
+{
+	u32 v = host1x_uclass_indoff_indbe_f(0xf)
+		| host1x_uclass_indoff_indmodid_f(mod_id)
+		| host1x_uclass_indoff_indroffset_f(offset);
+	if (auto_inc)
+		v |= host1x_uclass_indoff_autoinc_f(1);
+	return v;
+}
+
+static inline u32 host1x_class_host_indoff_reg_read(
+	unsigned mod_id, unsigned offset, bool auto_inc)
+{
+	u32 v = host1x_uclass_indoff_indmodid_f(mod_id)
+		| host1x_uclass_indoff_indroffset_f(offset)
+		| host1x_uclass_indoff_rwn_read_v();
+	if (auto_inc)
+		v |= host1x_uclass_indoff_autoinc_f(1);
+	return v;
+}
+
+/* cdma opcodes */
+static inline u32 host1x_opcode_setclass(
+	unsigned class_id, unsigned offset, unsigned mask)
+{
+	return (0 << 28) | (offset << 16) | (class_id << 6) | mask;
+}
+
+static inline u32 host1x_opcode_incr(unsigned offset, unsigned count)
+{
+	return (1 << 28) | (offset << 16) | count;
+}
+
+static inline u32 host1x_opcode_nonincr(unsigned offset, unsigned count)
+{
+	return (2 << 28) | (offset << 16) | count;
+}
+
+static inline u32 host1x_opcode_mask(unsigned offset, unsigned mask)
+{
+	return (3 << 28) | (offset << 16) | mask;
+}
+
+static inline u32 host1x_opcode_imm(unsigned offset, unsigned value)
+{
+	return (4 << 28) | (offset << 16) | value;
+}
+
+static inline u32 host1x_opcode_imm_incr_syncpt(unsigned cond, unsigned indx)
+{
+	return host1x_opcode_imm(host1x_uclass_incr_syncpt_r(),
+		host1x_class_host_incr_syncpt(cond, indx));
+}
+
+static inline u32 host1x_opcode_restart(unsigned address)
+{
+	return (5 << 28) | (address >> 4);
+}
+
+static inline u32 host1x_opcode_gather(unsigned count)
+{
+	return (6 << 28) | count;
+}
+
+static inline u32 host1x_opcode_gather_nonincr(unsigned offset,	unsigned count)
+{
+	return (6 << 28) | (offset << 16) | BIT(15) | count;
+}
+
+static inline u32 host1x_opcode_gather_incr(unsigned offset, unsigned count)
+{
+	return (6 << 28) | (offset << 16) | BIT(15) | BIT(14) | count;
+}
+
+#define HOST1X_OPCODE_NOP host1x_opcode_nonincr(0, 0)
+
+#endif
diff --git a/drivers/gpu/host1x/hw/hw_host1x02_uclass.h b/drivers/gpu/host1x/hw/hw_host1x02_uclass.h
index a3b3c9874413..028e49d9bac9 100644
--- a/drivers/gpu/host1x/hw/hw_host1x02_uclass.h
+++ b/drivers/gpu/host1x/hw/hw_host1x02_uclass.h
@@ -111,6 +111,12 @@ static inline u32 host1x_uclass_wait_syncpt_base_offset_f(u32 v)
 }
 #define HOST1X_UCLASS_WAIT_SYNCPT_BASE_OFFSET_F(v) \
 	host1x_uclass_wait_syncpt_base_offset_f(v)
+static inline u32 host1x_uclass_load_syncpt_base_r(void)
+{
+	return 0xb;
+}
+#define HOST1X_UCLASS_LOAD_SYNCPT_BASE \
+	host1x_uclass_load_syncpt_base_r()
 static inline u32 host1x_uclass_load_syncpt_base_base_indx_f(u32 v)
 {
 	return (v & 0xff) << 24;
diff --git a/drivers/gpu/host1x/hw/hw_host1x04_channel.h b/drivers/gpu/host1x/hw/hw_host1x04_channel.h
new file mode 100644
index 000000000000..95e6f96142b9
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x04_channel.h
@@ -0,0 +1,121 @@
+/*
+ * Copyright (c) 2013 NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope 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/>.
+ *
+ */
+
+ /*
+  * Function naming determines intended use:
+  *
+  *     <x>_r(void) : Returns the offset for register <x>.
+  *
+  *     <x>_w(void) : Returns the word offset for word (4 byte) element <x>.
+  *
+  *     <x>_<y>_s(void) : Returns size of field <y> of register <x> in bits.
+  *
+  *     <x>_<y>_f(u32 v) : Returns a value based on 'v' which has been shifted
+  *         and masked to place it at field <y> of register <x>.  This value
+  *         can be |'d with others to produce a full register value for
+  *         register <x>.
+  *
+  *     <x>_<y>_m(void) : Returns a mask for field <y> of register <x>.  This
+  *         value can be ~'d and then &'d to clear the value of field <y> for
+  *         register <x>.
+  *
+  *     <x>_<y>_<z>_f(void) : Returns the constant value <z> after being shifted
+  *         to place it at field <y> of register <x>.  This value can be |'d
+  *         with others to produce a full register value for <x>.
+  *
+  *     <x>_<y>_v(u32 r) : Returns the value of field <y> from a full register
+  *         <x> value 'r' after being shifted to place its LSB at bit 0.
+  *         This value is suitable for direct comparison with other unshifted
+  *         values appropriate for use in field <y> of register <x>.
+  *
+  *     <x>_<y>_<z>_v(void) : Returns the constant value for <z> defined for
+  *         field <y> of register <x>.  This value is suitable for direct
+  *         comparison with unshifted values appropriate for use in field <y>
+  *         of register <x>.
+  */
+
+#ifndef HOST1X_HW_HOST1X04_CHANNEL_H
+#define HOST1X_HW_HOST1X04_CHANNEL_H
+
+static inline u32 host1x_channel_fifostat_r(void)
+{
+	return 0x0;
+}
+#define HOST1X_CHANNEL_FIFOSTAT \
+	host1x_channel_fifostat_r()
+static inline u32 host1x_channel_fifostat_cfempty_v(u32 r)
+{
+	return (r >> 11) & 0x1;
+}
+#define HOST1X_CHANNEL_FIFOSTAT_CFEMPTY_V(r) \
+	host1x_channel_fifostat_cfempty_v(r)
+static inline u32 host1x_channel_dmastart_r(void)
+{
+	return 0x14;
+}
+#define HOST1X_CHANNEL_DMASTART \
+	host1x_channel_dmastart_r()
+static inline u32 host1x_channel_dmaput_r(void)
+{
+	return 0x18;
+}
+#define HOST1X_CHANNEL_DMAPUT \
+	host1x_channel_dmaput_r()
+static inline u32 host1x_channel_dmaget_r(void)
+{
+	return 0x1c;
+}
+#define HOST1X_CHANNEL_DMAGET \
+	host1x_channel_dmaget_r()
+static inline u32 host1x_channel_dmaend_r(void)
+{
+	return 0x20;
+}
+#define HOST1X_CHANNEL_DMAEND \
+	host1x_channel_dmaend_r()
+static inline u32 host1x_channel_dmactrl_r(void)
+{
+	return 0x24;
+}
+#define HOST1X_CHANNEL_DMACTRL \
+	host1x_channel_dmactrl_r()
+static inline u32 host1x_channel_dmactrl_dmastop(void)
+{
+	return 1 << 0;
+}
+#define HOST1X_CHANNEL_DMACTRL_DMASTOP \
+	host1x_channel_dmactrl_dmastop()
+static inline u32 host1x_channel_dmactrl_dmastop_v(u32 r)
+{
+	return (r >> 0) & 0x1;
+}
+#define HOST1X_CHANNEL_DMACTRL_DMASTOP_V(r) \
+	host1x_channel_dmactrl_dmastop_v(r)
+static inline u32 host1x_channel_dmactrl_dmagetrst(void)
+{
+	return 1 << 1;
+}
+#define HOST1X_CHANNEL_DMACTRL_DMAGETRST \
+	host1x_channel_dmactrl_dmagetrst()
+static inline u32 host1x_channel_dmactrl_dmainitget(void)
+{
+	return 1 << 2;
+}
+#define HOST1X_CHANNEL_DMACTRL_DMAINITGET \
+	host1x_channel_dmactrl_dmainitget()
+
+#endif
diff --git a/drivers/gpu/host1x/hw/hw_host1x04_sync.h b/drivers/gpu/host1x/hw/hw_host1x04_sync.h
new file mode 100644
index 000000000000..ef2275b5407a
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x04_sync.h
@@ -0,0 +1,243 @@
+/*
+ * Copyright (c) 2013 NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope 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/>.
+ *
+ */
+
+ /*
+  * Function naming determines intended use:
+  *
+  *     <x>_r(void) : Returns the offset for register <x>.
+  *
+  *     <x>_w(void) : Returns the word offset for word (4 byte) element <x>.
+  *
+  *     <x>_<y>_s(void) : Returns size of field <y> of register <x> in bits.
+  *
+  *     <x>_<y>_f(u32 v) : Returns a value based on 'v' which has been shifted
+  *         and masked to place it at field <y> of register <x>.  This value
+  *         can be |'d with others to produce a full register value for
+  *         register <x>.
+  *
+  *     <x>_<y>_m(void) : Returns a mask for field <y> of register <x>.  This
+  *         value can be ~'d and then &'d to clear the value of field <y> for
+  *         register <x>.
+  *
+  *     <x>_<y>_<z>_f(void) : Returns the constant value <z> after being shifted
+  *         to place it at field <y> of register <x>.  This value can be |'d
+  *         with others to produce a full register value for <x>.
+  *
+  *     <x>_<y>_v(u32 r) : Returns the value of field <y> from a full register
+  *         <x> value 'r' after being shifted to place its LSB at bit 0.
+  *         This value is suitable for direct comparison with other unshifted
+  *         values appropriate for use in field <y> of register <x>.
+  *
+  *     <x>_<y>_<z>_v(void) : Returns the constant value for <z> defined for
+  *         field <y> of register <x>.  This value is suitable for direct
+  *         comparison with unshifted values appropriate for use in field <y>
+  *         of register <x>.
+  */
+
+#ifndef HOST1X_HW_HOST1X04_SYNC_H
+#define HOST1X_HW_HOST1X04_SYNC_H
+
+#define REGISTER_STRIDE	4
+
+static inline u32 host1x_sync_syncpt_r(unsigned int id)
+{
+	return 0xf80 + id * REGISTER_STRIDE;
+}
+#define HOST1X_SYNC_SYNCPT(id) \
+	host1x_sync_syncpt_r(id)
+static inline u32 host1x_sync_syncpt_thresh_cpu0_int_status_r(unsigned int id)
+{
+	return 0xe80 + id * REGISTER_STRIDE;
+}
+#define HOST1X_SYNC_SYNCPT_THRESH_CPU0_INT_STATUS(id) \
+	host1x_sync_syncpt_thresh_cpu0_int_status_r(id)
+static inline u32 host1x_sync_syncpt_thresh_int_disable_r(unsigned int id)
+{
+	return 0xf00 + id * REGISTER_STRIDE;
+}
+#define HOST1X_SYNC_SYNCPT_THRESH_INT_DISABLE(id) \
+	host1x_sync_syncpt_thresh_int_disable_r(id)
+static inline u32 host1x_sync_syncpt_thresh_int_enable_cpu0_r(unsigned int id)
+{
+	return 0xf20 + id * REGISTER_STRIDE;
+}
+#define HOST1X_SYNC_SYNCPT_THRESH_INT_ENABLE_CPU0(id) \
+	host1x_sync_syncpt_thresh_int_enable_cpu0_r(id)
+static inline u32 host1x_sync_cf_setup_r(unsigned int channel)
+{
+	return 0xc00 + channel * REGISTER_STRIDE;
+}
+#define HOST1X_SYNC_CF_SETUP(channel) \
+	host1x_sync_cf_setup_r(channel)
+static inline u32 host1x_sync_cf_setup_base_v(u32 r)
+{
+	return (r >> 0) & 0x3ff;
+}
+#define HOST1X_SYNC_CF_SETUP_BASE_V(r) \
+	host1x_sync_cf_setup_base_v(r)
+static inline u32 host1x_sync_cf_setup_limit_v(u32 r)
+{
+	return (r >> 16) & 0x3ff;
+}
+#define HOST1X_SYNC_CF_SETUP_LIMIT_V(r) \
+	host1x_sync_cf_setup_limit_v(r)
+static inline u32 host1x_sync_cmdproc_stop_r(void)
+{
+	return 0xac;
+}
+#define HOST1X_SYNC_CMDPROC_STOP \
+	host1x_sync_cmdproc_stop_r()
+static inline u32 host1x_sync_ch_teardown_r(void)
+{
+	return 0xb0;
+}
+#define HOST1X_SYNC_CH_TEARDOWN \
+	host1x_sync_ch_teardown_r()
+static inline u32 host1x_sync_usec_clk_r(void)
+{
+	return 0x1a4;
+}
+#define HOST1X_SYNC_USEC_CLK \
+	host1x_sync_usec_clk_r()
+static inline u32 host1x_sync_ctxsw_timeout_cfg_r(void)
+{
+	return 0x1a8;
+}
+#define HOST1X_SYNC_CTXSW_TIMEOUT_CFG \
+	host1x_sync_ctxsw_timeout_cfg_r()
+static inline u32 host1x_sync_ip_busy_timeout_r(void)
+{
+	return 0x1bc;
+}
+#define HOST1X_SYNC_IP_BUSY_TIMEOUT \
+	host1x_sync_ip_busy_timeout_r()
+static inline u32 host1x_sync_mlock_owner_r(unsigned int id)
+{
+	return 0x340 + id * REGISTER_STRIDE;
+}
+#define HOST1X_SYNC_MLOCK_OWNER(id) \
+	host1x_sync_mlock_owner_r(id)
+static inline u32 host1x_sync_mlock_owner_chid_f(u32 v)
+{
+	return (v & 0xf) << 8;
+}
+#define HOST1X_SYNC_MLOCK_OWNER_CHID_F(v) \
+	host1x_sync_mlock_owner_chid_f(v)
+static inline u32 host1x_sync_mlock_owner_cpu_owns_v(u32 r)
+{
+	return (r >> 1) & 0x1;
+}
+#define HOST1X_SYNC_MLOCK_OWNER_CPU_OWNS_V(r) \
+	host1x_sync_mlock_owner_cpu_owns_v(r)
+static inline u32 host1x_sync_mlock_owner_ch_owns_v(u32 r)
+{
+	return (r >> 0) & 0x1;
+}
+#define HOST1X_SYNC_MLOCK_OWNER_CH_OWNS_V(r) \
+	host1x_sync_mlock_owner_ch_owns_v(r)
+static inline u32 host1x_sync_syncpt_int_thresh_r(unsigned int id)
+{
+	return 0x1380 + id * REGISTER_STRIDE;
+}
+#define HOST1X_SYNC_SYNCPT_INT_THRESH(id) \
+	host1x_sync_syncpt_int_thresh_r(id)
+static inline u32 host1x_sync_syncpt_base_r(unsigned int id)
+{
+	return 0x600 + id * REGISTER_STRIDE;
+}
+#define HOST1X_SYNC_SYNCPT_BASE(id) \
+	host1x_sync_syncpt_base_r(id)
+static inline u32 host1x_sync_syncpt_cpu_incr_r(unsigned int id)
+{
+	return 0xf60 + id * REGISTER_STRIDE;
+}
+#define HOST1X_SYNC_SYNCPT_CPU_INCR(id) \
+	host1x_sync_syncpt_cpu_incr_r(id)
+static inline u32 host1x_sync_cbread_r(unsigned int channel)
+{
+	return 0xc80 + channel * REGISTER_STRIDE;
+}
+#define HOST1X_SYNC_CBREAD(channel) \
+	host1x_sync_cbread_r(channel)
+static inline u32 host1x_sync_cfpeek_ctrl_r(void)
+{
+	return 0x74c;
+}
+#define HOST1X_SYNC_CFPEEK_CTRL \
+	host1x_sync_cfpeek_ctrl_r()
+static inline u32 host1x_sync_cfpeek_ctrl_addr_f(u32 v)
+{
+	return (v & 0x3ff) << 0;
+}
+#define HOST1X_SYNC_CFPEEK_CTRL_ADDR_F(v) \
+	host1x_sync_cfpeek_ctrl_addr_f(v)
+static inline u32 host1x_sync_cfpeek_ctrl_channr_f(u32 v)
+{
+	return (v & 0xf) << 16;
+}
+#define HOST1X_SYNC_CFPEEK_CTRL_CHANNR_F(v) \
+	host1x_sync_cfpeek_ctrl_channr_f(v)
+static inline u32 host1x_sync_cfpeek_ctrl_ena_f(u32 v)
+{
+	return (v & 0x1) << 31;
+}
+#define HOST1X_SYNC_CFPEEK_CTRL_ENA_F(v) \
+	host1x_sync_cfpeek_ctrl_ena_f(v)
+static inline u32 host1x_sync_cfpeek_read_r(void)
+{
+	return 0x750;
+}
+#define HOST1X_SYNC_CFPEEK_READ \
+	host1x_sync_cfpeek_read_r()
+static inline u32 host1x_sync_cfpeek_ptrs_r(void)
+{
+	return 0x754;
+}
+#define HOST1X_SYNC_CFPEEK_PTRS \
+	host1x_sync_cfpeek_ptrs_r()
+static inline u32 host1x_sync_cfpeek_ptrs_cf_rd_ptr_v(u32 r)
+{
+	return (r >> 0) & 0x3ff;
+}
+#define HOST1X_SYNC_CFPEEK_PTRS_CF_RD_PTR_V(r) \
+	host1x_sync_cfpeek_ptrs_cf_rd_ptr_v(r)
+static inline u32 host1x_sync_cfpeek_ptrs_cf_wr_ptr_v(u32 r)
+{
+	return (r >> 16) & 0x3ff;
+}
+#define HOST1X_SYNC_CFPEEK_PTRS_CF_WR_PTR_V(r) \
+	host1x_sync_cfpeek_ptrs_cf_wr_ptr_v(r)
+static inline u32 host1x_sync_cbstat_r(unsigned int channel)
+{
+	return 0xcc0 + channel * REGISTER_STRIDE;
+}
+#define HOST1X_SYNC_CBSTAT(channel) \
+	host1x_sync_cbstat_r(channel)
+static inline u32 host1x_sync_cbstat_cboffset_v(u32 r)
+{
+	return (r >> 0) & 0xffff;
+}
+#define HOST1X_SYNC_CBSTAT_CBOFFSET_V(r) \
+	host1x_sync_cbstat_cboffset_v(r)
+static inline u32 host1x_sync_cbstat_cbclass_v(u32 r)
+{
+	return (r >> 16) & 0x3ff;
+}
+#define HOST1X_SYNC_CBSTAT_CBCLASS_V(r) \
+	host1x_sync_cbstat_cbclass_v(r)
+
+#endif
diff --git a/drivers/gpu/host1x/hw/hw_host1x04_uclass.h b/drivers/gpu/host1x/hw/hw_host1x04_uclass.h
new file mode 100644
index 000000000000..d1460e971493
--- /dev/null
+++ b/drivers/gpu/host1x/hw/hw_host1x04_uclass.h
@@ -0,0 +1,181 @@
+/*
+ * Copyright (c) 2013 NVIDIA Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope 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/>.
+ *
+ */
+
+ /*
+  * Function naming determines intended use:
+  *
+  *     <x>_r(void) : Returns the offset for register <x>.
+  *
+  *     <x>_w(void) : Returns the word offset for word (4 byte) element <x>.
+  *
+  *     <x>_<y>_s(void) : Returns size of field <y> of register <x> in bits.
+  *
+  *     <x>_<y>_f(u32 v) : Returns a value based on 'v' which has been shifted
+  *         and masked to place it at field <y> of register <x>.  This value
+  *         can be |'d with others to produce a full register value for
+  *         register <x>.
+  *
+  *     <x>_<y>_m(void) : Returns a mask for field <y> of register <x>.  This
+  *         value can be ~'d and then &'d to clear the value of field <y> for
+  *         register <x>.
+  *
+  *     <x>_<y>_<z>_f(void) : Returns the constant value <z> after being shifted
+  *         to place it at field <y> of register <x>.  This value can be |'d
+  *         with others to produce a full register value for <x>.
+  *
+  *     <x>_<y>_v(u32 r) : Returns the value of field <y> from a full register
+  *         <x> value 'r' after being shifted to place its LSB at bit 0.
+  *         This value is suitable for direct comparison with other unshifted
+  *         values appropriate for use in field <y> of register <x>.
+  *
+  *     <x>_<y>_<z>_v(void) : Returns the constant value for <z> defined for
+  *         field <y> of register <x>.  This value is suitable for direct
+  *         comparison with unshifted values appropriate for use in field <y>
+  *         of register <x>.
+  */
+
+#ifndef HOST1X_HW_HOST1X04_UCLASS_H
+#define HOST1X_HW_HOST1X04_UCLASS_H
+
+static inline u32 host1x_uclass_incr_syncpt_r(void)
+{
+	return 0x0;
+}
+#define HOST1X_UCLASS_INCR_SYNCPT \
+	host1x_uclass_incr_syncpt_r()
+static inline u32 host1x_uclass_incr_syncpt_cond_f(u32 v)
+{
+	return (v & 0xff) << 8;
+}
+#define HOST1X_UCLASS_INCR_SYNCPT_COND_F(v) \
+	host1x_uclass_incr_syncpt_cond_f(v)
+static inline u32 host1x_uclass_incr_syncpt_indx_f(u32 v)
+{
+	return (v & 0xff) << 0;
+}
+#define HOST1X_UCLASS_INCR_SYNCPT_INDX_F(v) \
+	host1x_uclass_incr_syncpt_indx_f(v)
+static inline u32 host1x_uclass_wait_syncpt_r(void)
+{
+	return 0x8;
+}
+#define HOST1X_UCLASS_WAIT_SYNCPT \
+	host1x_uclass_wait_syncpt_r()
+static inline u32 host1x_uclass_wait_syncpt_indx_f(u32 v)
+{
+	return (v & 0xff) << 24;
+}
+#define HOST1X_UCLASS_WAIT_SYNCPT_INDX_F(v) \
+	host1x_uclass_wait_syncpt_indx_f(v)
+static inline u32 host1x_uclass_wait_syncpt_thresh_f(u32 v)
+{
+	return (v & 0xffffff) << 0;
+}
+#define HOST1X_UCLASS_WAIT_SYNCPT_THRESH_F(v) \
+	host1x_uclass_wait_syncpt_thresh_f(v)
+static inline u32 host1x_uclass_wait_syncpt_base_r(void)
+{
+	return 0x9;
+}
+#define HOST1X_UCLASS_WAIT_SYNCPT_BASE \
+	host1x_uclass_wait_syncpt_base_r()
+static inline u32 host1x_uclass_wait_syncpt_base_indx_f(u32 v)
+{
+	return (v & 0xff) << 24;
+}
+#define HOST1X_UCLASS_WAIT_SYNCPT_BASE_INDX_F(v) \
+	host1x_uclass_wait_syncpt_base_indx_f(v)
+static inline u32 host1x_uclass_wait_syncpt_base_base_indx_f(u32 v)
+{
+	return (v & 0xff) << 16;
+}
+#define HOST1X_UCLASS_WAIT_SYNCPT_BASE_BASE_INDX_F(v) \
+	host1x_uclass_wait_syncpt_base_base_indx_f(v)
+static inline u32 host1x_uclass_wait_syncpt_base_offset_f(u32 v)
+{
+	return (v & 0xffff) << 0;
+}
+#define HOST1X_UCLASS_WAIT_SYNCPT_BASE_OFFSET_F(v) \
+	host1x_uclass_wait_syncpt_base_offset_f(v)
+static inline u32 host1x_uclass_load_syncpt_base_r(void)
+{
+	return 0xb;
+}
+#define HOST1X_UCLASS_LOAD_SYNCPT_BASE \
+	host1x_uclass_load_syncpt_base_r()
+static inline u32 host1x_uclass_load_syncpt_base_base_indx_f(u32 v)
+{
+	return (v & 0xff) << 24;
+}
+#define HOST1X_UCLASS_LOAD_SYNCPT_BASE_BASE_INDX_F(v) \
+	host1x_uclass_load_syncpt_base_base_indx_f(v)
+static inline u32 host1x_uclass_load_syncpt_base_value_f(u32 v)
+{
+	return (v & 0xffffff) << 0;
+}
+#define HOST1X_UCLASS_LOAD_SYNCPT_BASE_VALUE_F(v) \
+	host1x_uclass_load_syncpt_base_value_f(v)
+static inline u32 host1x_uclass_incr_syncpt_base_base_indx_f(u32 v)
+{
+	return (v & 0xff) << 24;
+}
+#define HOST1X_UCLASS_INCR_SYNCPT_BASE_BASE_INDX_F(v) \
+	host1x_uclass_incr_syncpt_base_base_indx_f(v)
+static inline u32 host1x_uclass_incr_syncpt_base_offset_f(u32 v)
+{
+	return (v & 0xffffff) << 0;
+}
+#define HOST1X_UCLASS_INCR_SYNCPT_BASE_OFFSET_F(v) \
+	host1x_uclass_incr_syncpt_base_offset_f(v)
+static inline u32 host1x_uclass_indoff_r(void)
+{
+	return 0x2d;
+}
+#define HOST1X_UCLASS_INDOFF \
+	host1x_uclass_indoff_r()
+static inline u32 host1x_uclass_indoff_indbe_f(u32 v)
+{
+	return (v & 0xf) << 28;
+}
+#define HOST1X_UCLASS_INDOFF_INDBE_F(v) \
+	host1x_uclass_indoff_indbe_f(v)
+static inline u32 host1x_uclass_indoff_autoinc_f(u32 v)
+{
+	return (v & 0x1) << 27;
+}
+#define HOST1X_UCLASS_INDOFF_AUTOINC_F(v) \
+	host1x_uclass_indoff_autoinc_f(v)
+static inline u32 host1x_uclass_indoff_indmodid_f(u32 v)
+{
+	return (v & 0xff) << 18;
+}
+#define HOST1X_UCLASS_INDOFF_INDMODID_F(v) \
+	host1x_uclass_indoff_indmodid_f(v)
+static inline u32 host1x_uclass_indoff_indroffset_f(u32 v)
+{
+	return (v & 0xffff) << 2;
+}
+#define HOST1X_UCLASS_INDOFF_INDROFFSET_F(v) \
+	host1x_uclass_indoff_indroffset_f(v)
+static inline u32 host1x_uclass_indoff_rwn_read_v(void)
+{
+	return 1;
+}
+#define HOST1X_UCLASS_INDOFF_INDROFFSET_F(v) \
+	host1x_uclass_indoff_indroffset_f(v)
+
+#endif
diff --git a/drivers/gpu/host1x/hw/intr_hw.c b/drivers/gpu/host1x/hw/intr_hw.c
index b26dcc83bc1b..db9017adfe2b 100644
--- a/drivers/gpu/host1x/hw/intr_hw.c
+++ b/drivers/gpu/host1x/hw/intr_hw.c
@@ -20,7 +20,6 @@
 #include <linux/interrupt.h>
 #include <linux/irq.h>
 #include <linux/io.h>
-#include <asm/mach/irq.h>
 
 #include "../intr.h"
 #include "../dev.h"
diff --git a/drivers/gpu/host1x/job.c b/drivers/gpu/host1x/job.c
index de5ec333ce1a..1146e3bba6e1 100644
--- a/drivers/gpu/host1x/job.c
+++ b/drivers/gpu/host1x/job.c
@@ -75,12 +75,14 @@ struct host1x_job *host1x_job_alloc(struct host1x_channel *ch,
 
 	return job;
 }
+EXPORT_SYMBOL(host1x_job_alloc);
 
 struct host1x_job *host1x_job_get(struct host1x_job *job)
 {
 	kref_get(&job->ref);
 	return job;
 }
+EXPORT_SYMBOL(host1x_job_get);
 
 static void job_free(struct kref *ref)
 {
@@ -93,6 +95,7 @@ void host1x_job_put(struct host1x_job *job)
 {
 	kref_put(&job->ref, job_free);
 }
+EXPORT_SYMBOL(host1x_job_put);
 
 void host1x_job_add_gather(struct host1x_job *job, struct host1x_bo *bo,
 			   u32 words, u32 offset)
@@ -104,6 +107,7 @@ void host1x_job_add_gather(struct host1x_job *job, struct host1x_bo *bo,
 	cur_gather->offset = offset;
 	job->num_gathers++;
 }
+EXPORT_SYMBOL(host1x_job_add_gather);
 
 /*
  * NULL an already satisfied WAIT_SYNCPT host method, by patching its
@@ -560,6 +564,7 @@ out:
 
 	return err;
 }
+EXPORT_SYMBOL(host1x_job_pin);
 
 void host1x_job_unpin(struct host1x_job *job)
 {
@@ -577,6 +582,7 @@ void host1x_job_unpin(struct host1x_job *job)
 				      job->gather_copy_mapped,
 				      job->gather_copy);
 }
+EXPORT_SYMBOL(host1x_job_unpin);
 
 /*
  * Debug routine used to dump job entries
diff --git a/drivers/gpu/host1x/mipi.c b/drivers/gpu/host1x/mipi.c
new file mode 100644
index 000000000000..9882ea122024
--- /dev/null
+++ b/drivers/gpu/host1x/mipi.c
@@ -0,0 +1,275 @@
+/*
+ * Copyright (C) 2013 NVIDIA Corporation
+ *
+ * Permission to use, copy, modify, distribute, and sell this software and its
+ * documentation for any purpose is hereby granted without fee, provided that
+ * the above copyright notice appear in all copies and that both that copyright
+ * notice and this permission notice appear in supporting documentation, and
+ * that the name of the copyright holders not be used in advertising or
+ * publicity pertaining to distribution of the software without specific,
+ * written prior permission.  The copyright holders make no representations
+ * about the suitability of this software for any purpose.  It is provided "as
+ * is" without express or implied warranty.
+ *
+ * THE COPYRIGHT HOLDERS DISCLAIM ALL WARRANTIES WITH REGARD TO THIS SOFTWARE,
+ * INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN NO
+ * EVENT SHALL THE COPYRIGHT HOLDERS BE LIABLE FOR ANY SPECIAL, INDIRECT OR
+ * CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE,
+ * DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER
+ * TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE
+ * OF THIS SOFTWARE.
+ */
+
+#include <linux/clk.h>
+#include <linux/delay.h>
+#include <linux/host1x.h>
+#include <linux/io.h>
+#include <linux/of_platform.h>
+#include <linux/platform_device.h>
+#include <linux/slab.h>
+
+#include "dev.h"
+
+#define MIPI_CAL_CTRL			0x00
+#define MIPI_CAL_CTRL_START		(1 << 0)
+
+#define MIPI_CAL_AUTOCAL_CTRL		0x01
+
+#define MIPI_CAL_STATUS			0x02
+#define MIPI_CAL_STATUS_DONE		(1 << 16)
+#define MIPI_CAL_STATUS_ACTIVE		(1 <<  0)
+
+#define MIPI_CAL_CONFIG_CSIA		0x05
+#define MIPI_CAL_CONFIG_CSIB		0x06
+#define MIPI_CAL_CONFIG_CSIC		0x07
+#define MIPI_CAL_CONFIG_CSID		0x08
+#define MIPI_CAL_CONFIG_CSIE		0x09
+#define MIPI_CAL_CONFIG_DSIA		0x0e
+#define MIPI_CAL_CONFIG_DSIB		0x0f
+#define MIPI_CAL_CONFIG_DSIC		0x10
+#define MIPI_CAL_CONFIG_DSID		0x11
+
+#define MIPI_CAL_CONFIG_SELECT		(1 << 21)
+#define MIPI_CAL_CONFIG_HSPDOS(x)	(((x) & 0x1f) << 16)
+#define MIPI_CAL_CONFIG_HSPUOS(x)	(((x) & 0x1f) <<  8)
+#define MIPI_CAL_CONFIG_TERMOS(x)	(((x) & 0x1f) <<  0)
+
+#define MIPI_CAL_BIAS_PAD_CFG0		0x16
+#define MIPI_CAL_BIAS_PAD_PDVCLAMP	(1 << 1)
+#define MIPI_CAL_BIAS_PAD_E_VCLAMP_REF	(1 << 0)
+
+#define MIPI_CAL_BIAS_PAD_CFG1		0x17
+
+#define MIPI_CAL_BIAS_PAD_CFG2		0x18
+#define MIPI_CAL_BIAS_PAD_PDVREG	(1 << 1)
+
+static const struct module {
+	unsigned long reg;
+} modules[] = {
+	{ .reg = MIPI_CAL_CONFIG_CSIA },
+	{ .reg = MIPI_CAL_CONFIG_CSIB },
+	{ .reg = MIPI_CAL_CONFIG_CSIC },
+	{ .reg = MIPI_CAL_CONFIG_CSID },
+	{ .reg = MIPI_CAL_CONFIG_CSIE },
+	{ .reg = MIPI_CAL_CONFIG_DSIA },
+	{ .reg = MIPI_CAL_CONFIG_DSIB },
+	{ .reg = MIPI_CAL_CONFIG_DSIC },
+	{ .reg = MIPI_CAL_CONFIG_DSID },
+};
+
+struct tegra_mipi {
+	void __iomem *regs;
+	struct mutex lock;
+	struct clk *clk;
+};
+
+struct tegra_mipi_device {
+	struct platform_device *pdev;
+	struct tegra_mipi *mipi;
+	struct device *device;
+	unsigned long pads;
+};
+
+static inline unsigned long tegra_mipi_readl(struct tegra_mipi *mipi,
+					     unsigned long reg)
+{
+	return readl(mipi->regs + (reg << 2));
+}
+
+static inline void tegra_mipi_writel(struct tegra_mipi *mipi,
+				     unsigned long value, unsigned long reg)
+{
+	writel(value, mipi->regs + (reg << 2));
+}
+
+struct tegra_mipi_device *tegra_mipi_request(struct device *device)
+{
+	struct device_node *np = device->of_node;
+	struct tegra_mipi_device *dev;
+	struct of_phandle_args args;
+	int err;
+
+	err = of_parse_phandle_with_args(np, "nvidia,mipi-calibrate",
+					 "#nvidia,mipi-calibrate-cells", 0,
+					 &args);
+	if (err < 0)
+		return ERR_PTR(err);
+
+	dev = kzalloc(sizeof(*dev), GFP_KERNEL);
+	if (!dev) {
+		of_node_put(args.np);
+		err = -ENOMEM;
+		goto out;
+	}
+
+	dev->pdev = of_find_device_by_node(args.np);
+	if (!dev->pdev) {
+		of_node_put(args.np);
+		err = -ENODEV;
+		goto free;
+	}
+
+	of_node_put(args.np);
+
+	dev->mipi = platform_get_drvdata(dev->pdev);
+	if (!dev->mipi) {
+		err = -EPROBE_DEFER;
+		goto pdev_put;
+	}
+
+	dev->pads = args.args[0];
+	dev->device = device;
+
+	return dev;
+
+pdev_put:
+	platform_device_put(dev->pdev);
+free:
+	kfree(dev);
+out:
+	return ERR_PTR(err);
+}
+EXPORT_SYMBOL(tegra_mipi_request);
+
+void tegra_mipi_free(struct tegra_mipi_device *device)
+{
+	platform_device_put(device->pdev);
+	kfree(device);
+}
+EXPORT_SYMBOL(tegra_mipi_free);
+
+static int tegra_mipi_wait(struct tegra_mipi *mipi)
+{
+	unsigned long timeout = jiffies + msecs_to_jiffies(250);
+	unsigned long value;
+
+	while (time_before(jiffies, timeout)) {
+		value = tegra_mipi_readl(mipi, MIPI_CAL_STATUS);
+		if ((value & MIPI_CAL_STATUS_ACTIVE) == 0 &&
+		    (value & MIPI_CAL_STATUS_DONE) != 0)
+			return 0;
+
+		usleep_range(10, 50);
+	}
+
+	return -ETIMEDOUT;
+}
+
+int tegra_mipi_calibrate(struct tegra_mipi_device *device)
+{
+	unsigned long value;
+	unsigned int i;
+	int err;
+
+	err = clk_enable(device->mipi->clk);
+	if (err < 0)
+		return err;
+
+	mutex_lock(&device->mipi->lock);
+
+	value = tegra_mipi_readl(device->mipi, MIPI_CAL_BIAS_PAD_CFG0);
+	value &= ~MIPI_CAL_BIAS_PAD_PDVCLAMP;
+	value |= MIPI_CAL_BIAS_PAD_E_VCLAMP_REF;
+	tegra_mipi_writel(device->mipi, value, MIPI_CAL_BIAS_PAD_CFG0);
+
+	value = tegra_mipi_readl(device->mipi, MIPI_CAL_BIAS_PAD_CFG2);
+	value &= ~MIPI_CAL_BIAS_PAD_PDVREG;
+	tegra_mipi_writel(device->mipi, value, MIPI_CAL_BIAS_PAD_CFG2);
+
+	for (i = 0; i < ARRAY_SIZE(modules); i++) {
+		if (device->pads & BIT(i))
+			value = MIPI_CAL_CONFIG_SELECT |
+				MIPI_CAL_CONFIG_HSPDOS(0) |
+				MIPI_CAL_CONFIG_HSPUOS(4) |
+				MIPI_CAL_CONFIG_TERMOS(5);
+		else
+			value = 0;
+
+		tegra_mipi_writel(device->mipi, value, modules[i].reg);
+	}
+
+	tegra_mipi_writel(device->mipi, MIPI_CAL_CTRL_START, MIPI_CAL_CTRL);
+
+	err = tegra_mipi_wait(device->mipi);
+
+	mutex_unlock(&device->mipi->lock);
+	clk_disable(device->mipi->clk);
+
+	return err;
+}
+EXPORT_SYMBOL(tegra_mipi_calibrate);
+
+static int tegra_mipi_probe(struct platform_device *pdev)
+{
+	struct tegra_mipi *mipi;
+	struct resource *res;
+	int err;
+
+	mipi = devm_kzalloc(&pdev->dev, sizeof(*mipi), GFP_KERNEL);
+	if (!mipi)
+		return -ENOMEM;
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	mipi->regs = devm_ioremap_resource(&pdev->dev, res);
+	if (IS_ERR(mipi->regs))
+		return PTR_ERR(mipi->regs);
+
+	mutex_init(&mipi->lock);
+
+	mipi->clk = devm_clk_get(&pdev->dev, NULL);
+	if (IS_ERR(mipi->clk)) {
+		dev_err(&pdev->dev, "failed to get clock\n");
+		return PTR_ERR(mipi->clk);
+	}
+
+	err = clk_prepare(mipi->clk);
+	if (err < 0)
+		return err;
+
+	platform_set_drvdata(pdev, mipi);
+
+	return 0;
+}
+
+static int tegra_mipi_remove(struct platform_device *pdev)
+{
+	struct tegra_mipi *mipi = platform_get_drvdata(pdev);
+
+	clk_unprepare(mipi->clk);
+
+	return 0;
+}
+
+static struct of_device_id tegra_mipi_of_match[] = {
+	{ .compatible = "nvidia,tegra114-mipi", },
+	{ },
+};
+
+struct platform_driver tegra_mipi_driver = {
+	.driver = {
+		.name = "tegra-mipi",
+		.of_match_table = tegra_mipi_of_match,
+	},
+	.probe = tegra_mipi_probe,
+	.remove = tegra_mipi_remove,
+};
diff --git a/drivers/gpu/host1x/syncpt.c b/drivers/gpu/host1x/syncpt.c
index 159c479829c9..bfb09d802abd 100644
--- a/drivers/gpu/host1x/syncpt.c
+++ b/drivers/gpu/host1x/syncpt.c
@@ -93,6 +93,7 @@ u32 host1x_syncpt_id(struct host1x_syncpt *sp)
 {
 	return sp->id;
 }
+EXPORT_SYMBOL(host1x_syncpt_id);
 
 /*
  * Updates the value sent to hardware.
@@ -168,6 +169,7 @@ int host1x_syncpt_incr(struct host1x_syncpt *sp)
 {
 	return host1x_hw_syncpt_cpu_incr(sp->host, sp);
 }
+EXPORT_SYMBOL(host1x_syncpt_incr);
 
 /*
  * Updated sync point form hardware, and returns true if syncpoint is expired,
@@ -377,6 +379,7 @@ struct host1x_syncpt *host1x_syncpt_request(struct device *dev,
 	struct host1x *host = dev_get_drvdata(dev->parent);
 	return host1x_syncpt_alloc(host, dev, flags);
 }
+EXPORT_SYMBOL(host1x_syncpt_request);
 
 void host1x_syncpt_free(struct host1x_syncpt *sp)
 {
@@ -390,6 +393,7 @@ void host1x_syncpt_free(struct host1x_syncpt *sp)
 	sp->name = NULL;
 	sp->client_managed = false;
 }
+EXPORT_SYMBOL(host1x_syncpt_free);
 
 void host1x_syncpt_deinit(struct host1x *host)
 {
@@ -408,6 +412,7 @@ u32 host1x_syncpt_read_max(struct host1x_syncpt *sp)
 	smp_rmb();
 	return (u32)atomic_read(&sp->max_val);
 }
+EXPORT_SYMBOL(host1x_syncpt_read_max);
 
 /*
  * Read min, which is a shadow of the current sync point value in hardware.
@@ -417,6 +422,7 @@ u32 host1x_syncpt_read_min(struct host1x_syncpt *sp)
 	smp_rmb();
 	return (u32)atomic_read(&sp->min_val);
 }
+EXPORT_SYMBOL(host1x_syncpt_read_min);
 
 int host1x_syncpt_nb_pts(struct host1x *host)
 {
@@ -439,13 +445,16 @@ struct host1x_syncpt *host1x_syncpt_get(struct host1x *host, u32 id)
 		return NULL;
 	return host->syncpt + id;
 }
+EXPORT_SYMBOL(host1x_syncpt_get);
 
 struct host1x_syncpt_base *host1x_syncpt_get_base(struct host1x_syncpt *sp)
 {
 	return sp ? sp->base : NULL;
 }
+EXPORT_SYMBOL(host1x_syncpt_get_base);
 
 u32 host1x_syncpt_base_id(struct host1x_syncpt_base *base)
 {
 	return base->id;
 }
+EXPORT_SYMBOL(host1x_syncpt_base_id);