Merge commit '9e9a928eed8796a0a1aaed7e0b676db86ba84594' into drm-next

Merge drm-fixes into drm-next.

Both i915 and radeon need this done for later patches.

Conflicts:
	drivers/gpu/drm/drm_crtc_helper.c
	drivers/gpu/drm/i915/i915_drv.h
	drivers/gpu/drm/i915/i915_gem.c
	drivers/gpu/drm/i915/i915_gem_execbuffer.c
	drivers/gpu/drm/i915/i915_gem_gtt.c

50 files changed, 1164 insertions, 538 deletions

diff --git a/drivers/gpu/drm/drm_crtc_helper.c b/drivers/gpu/drm/drm_crtc_helper.c
index eb1c062e04b2..78b37f3febd3 100644
--- a/drivers/gpu/drm/drm_crtc_helper.c
+++ b/drivers/gpu/drm/drm_crtc_helper.c
@@ -29,6 +29,7 @@
  *      Jesse Barnes <jesse.barnes@intel.com>
  */
 
+#include <linux/kernel.h>
 #include <linux/export.h>
 #include <linux/moduleparam.h>
 
@@ -88,8 +89,15 @@ bool drm_helper_encoder_in_use(struct drm_encoder *encoder)
 	struct drm_connector *connector;
 	struct drm_device *dev = encoder->dev;
 
-	WARN_ON(!mutex_is_locked(&dev->mode_config.mutex));
-	WARN_ON(!drm_modeset_is_locked(&dev->mode_config.connection_mutex));
+	/*
+	 * We can expect this mutex to be locked if we are not panicking.
+	 * Locking is currently fubar in the panic handler.
+	 */
+	if (!oops_in_progress) {
+		WARN_ON(!mutex_is_locked(&dev->mode_config.mutex));
+		WARN_ON(!drm_modeset_is_locked(&dev->mode_config.connection_mutex));
+	}
+
 	list_for_each_entry(connector, &dev->mode_config.connector_list, head)
 		if (connector->encoder == encoder)
 			return true;
@@ -113,7 +121,13 @@ bool drm_helper_crtc_in_use(struct drm_crtc *crtc)
 	struct drm_encoder *encoder;
 	struct drm_device *dev = crtc->dev;
 
-	WARN_ON(!mutex_is_locked(&dev->mode_config.mutex));
+	/*
+	 * We can expect this mutex to be locked if we are not panicking.
+	 * Locking is currently fubar in the panic handler.
+	 */
+	if (!oops_in_progress)
+		WARN_ON(!mutex_is_locked(&dev->mode_config.mutex));
+
 	list_for_each_entry(encoder, &dev->mode_config.encoder_list, head)
 		if (encoder->crtc == crtc && drm_helper_encoder_in_use(encoder))
 			return true;
diff --git a/drivers/gpu/drm/i915/i915_dma.c b/drivers/gpu/drm/i915/i915_dma.c
index 4e70de6ed468..b9159ade5e85 100644
--- a/drivers/gpu/drm/i915/i915_dma.c
+++ b/drivers/gpu/drm/i915/i915_dma.c
@@ -1836,7 +1836,6 @@ int i915_driver_unload(struct drm_device *dev)
 		flush_workqueue(dev_priv->wq);
 
 		mutex_lock(&dev->struct_mutex);
-		i915_gem_free_all_phys_object(dev);
 		i915_gem_cleanup_ringbuffer(dev);
 		i915_gem_context_fini(dev);
 		WARN_ON(dev_priv->mm.aliasing_ppgtt);
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 8f68678f361f..8e78703e45cf 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -251,18 +251,6 @@ struct intel_ddi_plls {
 #define WATCH_LISTS	0
 #define WATCH_GTT	0
 
-#define I915_GEM_PHYS_CURSOR_0 1
-#define I915_GEM_PHYS_CURSOR_1 2
-#define I915_GEM_PHYS_OVERLAY_REGS 3
-#define I915_MAX_PHYS_OBJECT (I915_GEM_PHYS_OVERLAY_REGS)
-
-struct drm_i915_gem_phys_object {
-	int id;
-	struct page **page_list;
-	drm_dma_handle_t *handle;
-	struct drm_i915_gem_object *cur_obj;
-};
-
 struct opregion_header;
 struct opregion_acpi;
 struct opregion_swsci;
@@ -1106,9 +1094,6 @@ struct i915_gem_mm {
 	/** Bit 6 swizzling required for Y tiling */
 	uint32_t bit_6_swizzle_y;
 
-	/* storage for physical objects */
-	struct drm_i915_gem_phys_object *phys_objs[I915_MAX_PHYS_OBJECT];
-
 	/* accounting, useful for userland debugging */
 	spinlock_t object_stat_lock;
 	size_t object_memory;
@@ -1712,7 +1697,7 @@ struct drm_i915_gem_object {
 	struct drm_file *pin_filp;
 
 	/** for phy allocated objects */
-	struct drm_i915_gem_phys_object *phys_obj;
+	drm_dma_handle_t *phys_handle;
 
 	union {
 		struct i915_gem_userptr {
@@ -1916,6 +1901,9 @@ struct drm_i915_cmd_table {
 #define IS_ULT(dev)		(IS_HSW_ULT(dev) || IS_BDW_ULT(dev))
 #define IS_HSW_GT3(dev)		(IS_HASWELL(dev) && \
 				 ((dev)->pdev->device & 0x00F0) == 0x0020)
+/* ULX machines are also considered ULT. */
+#define IS_HSW_ULX(dev)		((dev)->pdev->device == 0x0A0E || \
+				 (dev)->pdev->device == 0x0A1E)
 #define IS_PRELIMINARY_HW(intel_info) ((intel_info)->is_preliminary)
 
 /*
@@ -2172,10 +2160,12 @@ void i915_gem_vma_destroy(struct i915_vma *vma);
 #define PIN_MAPPABLE 0x1
 #define PIN_NONBLOCK 0x2
 #define PIN_GLOBAL 0x4
+#define PIN_OFFSET_BIAS 0x8
+#define PIN_OFFSET_MASK (~4095)
 int __must_check i915_gem_object_pin(struct drm_i915_gem_object *obj,
 				     struct i915_address_space *vm,
 				     uint32_t alignment,
-				     unsigned flags);
+				     uint64_t flags);
 int __must_check i915_vma_unbind(struct i915_vma *vma);
 int i915_gem_object_put_pages(struct drm_i915_gem_object *obj);
 void i915_gem_release_all_mmaps(struct drm_i915_private *dev_priv);
@@ -2297,13 +2287,8 @@ i915_gem_object_pin_to_display_plane(struct drm_i915_gem_object *obj,
 				     u32 alignment,
 				     struct intel_engine_cs *pipelined);
 void i915_gem_object_unpin_from_display_plane(struct drm_i915_gem_object *obj);
-int i915_gem_attach_phys_object(struct drm_device *dev,
-				struct drm_i915_gem_object *obj,
-				int id,
+int i915_gem_object_attach_phys(struct drm_i915_gem_object *obj,
 				int align);
-void i915_gem_detach_phys_object(struct drm_device *dev,
-				 struct drm_i915_gem_object *obj);
-void i915_gem_free_all_phys_object(struct drm_device *dev);
 int i915_gem_open(struct drm_device *dev, struct drm_file *file);
 void i915_gem_release(struct drm_device *dev, struct drm_file *file);
 
@@ -2430,6 +2415,8 @@ int __must_check i915_gem_evict_something(struct drm_device *dev,
 					  int min_size,
 					  unsigned alignment,
 					  unsigned cache_level,
+					  unsigned long start,
+					  unsigned long end,
 					  unsigned flags);
 int i915_gem_evict_vm(struct i915_address_space *vm, bool do_idle);
 int i915_gem_evict_everything(struct drm_device *dev);
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 87e9b349ebef..bbcd35abf247 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -47,11 +47,6 @@ i915_gem_object_wait_rendering(struct drm_i915_gem_object *obj,
 static void
 i915_gem_object_retire(struct drm_i915_gem_object *obj);
 
-static int i915_gem_phys_pwrite(struct drm_device *dev,
-				struct drm_i915_gem_object *obj,
-				struct drm_i915_gem_pwrite *args,
-				struct drm_file *file);
-
 static void i915_gem_write_fence(struct drm_device *dev, int reg,
 				 struct drm_i915_gem_object *obj);
 static void i915_gem_object_update_fence(struct drm_i915_gem_object *obj,
@@ -214,6 +209,128 @@ i915_gem_get_aperture_ioctl(struct drm_device *dev, void *data,
 	return 0;
 }
 
+static void i915_gem_object_detach_phys(struct drm_i915_gem_object *obj)
+{
+	drm_dma_handle_t *phys = obj->phys_handle;
+
+	if (!phys)
+		return;
+
+	if (obj->madv == I915_MADV_WILLNEED) {
+		struct address_space *mapping = file_inode(obj->base.filp)->i_mapping;
+		char *vaddr = phys->vaddr;
+		int i;
+
+		for (i = 0; i < obj->base.size / PAGE_SIZE; i++) {
+			struct page *page = shmem_read_mapping_page(mapping, i);
+			if (!IS_ERR(page)) {
+				char *dst = kmap_atomic(page);
+				memcpy(dst, vaddr, PAGE_SIZE);
+				drm_clflush_virt_range(dst, PAGE_SIZE);
+				kunmap_atomic(dst);
+
+				set_page_dirty(page);
+				mark_page_accessed(page);
+				page_cache_release(page);
+			}
+			vaddr += PAGE_SIZE;
+		}
+		i915_gem_chipset_flush(obj->base.dev);
+	}
+
+#ifdef CONFIG_X86
+	set_memory_wb((unsigned long)phys->vaddr, phys->size / PAGE_SIZE);
+#endif
+	drm_pci_free(obj->base.dev, phys);
+	obj->phys_handle = NULL;
+}
+
+int
+i915_gem_object_attach_phys(struct drm_i915_gem_object *obj,
+			    int align)
+{
+	drm_dma_handle_t *phys;
+	struct address_space *mapping;
+	char *vaddr;
+	int i;
+
+	if (obj->phys_handle) {
+		if ((unsigned long)obj->phys_handle->vaddr & (align -1))
+			return -EBUSY;
+
+		return 0;
+	}
+
+	if (obj->madv != I915_MADV_WILLNEED)
+		return -EFAULT;
+
+	if (obj->base.filp == NULL)
+		return -EINVAL;
+
+	/* create a new object */
+	phys = drm_pci_alloc(obj->base.dev, obj->base.size, align);
+	if (!phys)
+		return -ENOMEM;
+
+	vaddr = phys->vaddr;
+#ifdef CONFIG_X86
+	set_memory_wc((unsigned long)vaddr, phys->size / PAGE_SIZE);
+#endif
+	mapping = file_inode(obj->base.filp)->i_mapping;
+	for (i = 0; i < obj->base.size / PAGE_SIZE; i++) {
+		struct page *page;
+		char *src;
+
+		page = shmem_read_mapping_page(mapping, i);
+		if (IS_ERR(page)) {
+#ifdef CONFIG_X86
+			set_memory_wb((unsigned long)phys->vaddr, phys->size / PAGE_SIZE);
+#endif
+			drm_pci_free(obj->base.dev, phys);
+			return PTR_ERR(page);
+		}
+
+		src = kmap_atomic(page);
+		memcpy(vaddr, src, PAGE_SIZE);
+		kunmap_atomic(src);
+
+		mark_page_accessed(page);
+		page_cache_release(page);
+
+		vaddr += PAGE_SIZE;
+	}
+
+	obj->phys_handle = phys;
+	return 0;
+}
+
+static int
+i915_gem_phys_pwrite(struct drm_i915_gem_object *obj,
+		     struct drm_i915_gem_pwrite *args,
+		     struct drm_file *file_priv)
+{
+	struct drm_device *dev = obj->base.dev;
+	void *vaddr = obj->phys_handle->vaddr + args->offset;
+	char __user *user_data = to_user_ptr(args->data_ptr);
+
+	if (__copy_from_user_inatomic_nocache(vaddr, user_data, args->size)) {
+		unsigned long unwritten;
+
+		/* The physical object once assigned is fixed for the lifetime
+		 * of the obj, so we can safely drop the lock and continue
+		 * to access vaddr.
+		 */
+		mutex_unlock(&dev->struct_mutex);
+		unwritten = copy_from_user(vaddr, user_data, args->size);
+		mutex_lock(&dev->struct_mutex);
+		if (unwritten)
+			return -EFAULT;
+	}
+
+	i915_gem_chipset_flush(dev);
+	return 0;
+}
+
 void *i915_gem_object_alloc(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -930,8 +1047,8 @@ i915_gem_pwrite_ioctl(struct drm_device *dev, void *data,
 	 * pread/pwrite currently are reading and writing from the CPU
 	 * perspective, requiring manual detiling by the client.
 	 */
-	if (obj->phys_obj) {
-		ret = i915_gem_phys_pwrite(dev, obj, args, file);
+	if (obj->phys_handle) {
+		ret = i915_gem_phys_pwrite(obj, args, file);
 		goto out;
 	}
 
@@ -3257,12 +3374,14 @@ static struct i915_vma *
 i915_gem_object_bind_to_vm(struct drm_i915_gem_object *obj,
 			   struct i915_address_space *vm,
 			   unsigned alignment,
-			   unsigned flags)
+			   uint64_t flags)
 {
 	struct drm_device *dev = obj->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	u32 size, fence_size, fence_alignment, unfenced_alignment;
-	size_t gtt_max =
+	unsigned long start =
+		flags & PIN_OFFSET_BIAS ? flags & PIN_OFFSET_MASK : 0;
+	unsigned long end =
 		flags & PIN_MAPPABLE ? dev_priv->gtt.mappable_end : vm->total;
 	struct i915_vma *vma;
 	int ret;
@@ -3291,11 +3410,11 @@ i915_gem_object_bind_to_vm(struct drm_i915_gem_object *obj,
 	/* If the object is bigger than the entire aperture, reject it early
 	 * before evicting everything in a vain attempt to find space.
 	 */
-	if (obj->base.size > gtt_max) {
-		DRM_DEBUG("Attempting to bind an object larger than the aperture: object=%zd > %s aperture=%zu\n",
+	if (obj->base.size > end) {
+		DRM_DEBUG("Attempting to bind an object larger than the aperture: object=%zd > %s aperture=%lu\n",
 			  obj->base.size,
 			  flags & PIN_MAPPABLE ? "mappable" : "total",
-			  gtt_max);
+			  end);
 		return ERR_PTR(-E2BIG);
 	}
 
@@ -3312,12 +3431,15 @@ i915_gem_object_bind_to_vm(struct drm_i915_gem_object *obj,
 search_free:
 	ret = drm_mm_insert_node_in_range_generic(&vm->mm, &vma->node,
 						  size, alignment,
-						  obj->cache_level, 0, gtt_max,
+						  obj->cache_level,
+						  start, end,
 						  DRM_MM_SEARCH_DEFAULT,
 						  DRM_MM_CREATE_DEFAULT);
 	if (ret) {
 		ret = i915_gem_evict_something(dev, vm, size, alignment,
-					       obj->cache_level, flags);
+					       obj->cache_level,
+					       start, end,
+					       flags);
 		if (ret == 0)
 			goto search_free;
 
@@ -3892,11 +4014,30 @@ i915_gem_ring_throttle(struct drm_device *dev, struct drm_file *file)
 	return ret;
 }
 
+static bool
+i915_vma_misplaced(struct i915_vma *vma, uint32_t alignment, uint64_t flags)
+{
+	struct drm_i915_gem_object *obj = vma->obj;
+
+	if (alignment &&
+	    vma->node.start & (alignment - 1))
+		return true;
+
+	if (flags & PIN_MAPPABLE && !obj->map_and_fenceable)
+		return true;
+
+	if (flags & PIN_OFFSET_BIAS &&
+	    vma->node.start < (flags & PIN_OFFSET_MASK))
+		return true;
+
+	return false;
+}
+
 int
 i915_gem_object_pin(struct drm_i915_gem_object *obj,
 		    struct i915_address_space *vm,
 		    uint32_t alignment,
-		    unsigned flags)
+		    uint64_t flags)
 {
 	struct drm_i915_private *dev_priv = obj->base.dev->dev_private;
 	struct i915_vma *vma;
@@ -3913,15 +4054,13 @@ i915_gem_object_pin(struct drm_i915_gem_object *obj,
 		if (WARN_ON(vma->pin_count == DRM_I915_GEM_OBJECT_MAX_PIN_COUNT))
 			return -EBUSY;
 
-		if ((alignment &&
-		     vma->node.start & (alignment - 1)) ||
-		    (flags & PIN_MAPPABLE && !obj->map_and_fenceable)) {
+		if (i915_vma_misplaced(vma, alignment, flags)) {
 			WARN(vma->pin_count,
 			     "bo is already pinned with incorrect alignment:"
 			     " offset=%lx, req.alignment=%x, req.map_and_fenceable=%d,"
 			     " obj->map_and_fenceable=%d\n",
 			     i915_gem_obj_offset(obj, vm), alignment,
-			     flags & PIN_MAPPABLE,
+			     !!(flags & PIN_MAPPABLE),
 			     obj->map_and_fenceable);
 			ret = i915_vma_unbind(vma);
 			if (ret)
@@ -4281,9 +4420,6 @@ void i915_gem_free_object(struct drm_gem_object *gem_obj)
 
 	trace_i915_gem_object_destroy(obj);
 
-	if (obj->phys_obj)
-		i915_gem_detach_phys_object(dev, obj);
-
 	list_for_each_entry_safe(vma, next, &obj->vma_list, vma_link) {
 		int ret;
 
@@ -4301,6 +4437,8 @@ void i915_gem_free_object(struct drm_gem_object *gem_obj)
 		}
 	}
 
+	i915_gem_object_detach_phys(obj);
+
 	/* Stolen objects don't hold a ref, but do hold pin count. Fix that up
 	 * before progressing. */
 	if (obj->stolen)
@@ -4792,190 +4930,6 @@ i915_gem_load(struct drm_device *dev)
 	register_oom_notifier(&dev_priv->mm.oom_notifier);
 }
 
-/*
- * Create a physically contiguous memory object for this object
- * e.g. for cursor + overlay regs
- */
-static int i915_gem_init_phys_object(struct drm_device *dev,
-				     int id, int size, int align)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_i915_gem_phys_object *phys_obj;
-	int ret;
-
-	if (dev_priv->mm.phys_objs[id - 1] || !size)
-		return 0;
-
-	phys_obj = kzalloc(sizeof(*phys_obj), GFP_KERNEL);
-	if (!phys_obj)
-		return -ENOMEM;
-
-	phys_obj->id = id;
-
-	phys_obj->handle = drm_pci_alloc(dev, size, align);
-	if (!phys_obj->handle) {
-		ret = -ENOMEM;
-		goto kfree_obj;
-	}
-#ifdef CONFIG_X86
-	set_memory_wc((unsigned long)phys_obj->handle->vaddr, phys_obj->handle->size / PAGE_SIZE);
-#endif
-
-	dev_priv->mm.phys_objs[id - 1] = phys_obj;
-
-	return 0;
-kfree_obj:
-	kfree(phys_obj);
-	return ret;
-}
-
-static void i915_gem_free_phys_object(struct drm_device *dev, int id)
-{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct drm_i915_gem_phys_object *phys_obj;
-
-	if (!dev_priv->mm.phys_objs[id - 1])
-		return;
-
-	phys_obj = dev_priv->mm.phys_objs[id - 1];
-	if (phys_obj->cur_obj) {
-		i915_gem_detach_phys_object(dev, phys_obj->cur_obj);
-	}
-
-#ifdef CONFIG_X86
-	set_memory_wb((unsigned long)phys_obj->handle->vaddr, phys_obj->handle->size / PAGE_SIZE);
-#endif
-	drm_pci_free(dev, phys_obj->handle);
-	kfree(phys_obj);
-	dev_priv->mm.phys_objs[id - 1] = NULL;
-}
-
-void i915_gem_free_all_phys_object(struct drm_device *dev)
-{
-	int i;
-
-	for (i = I915_GEM_PHYS_CURSOR_0; i <= I915_MAX_PHYS_OBJECT; i++)
-		i915_gem_free_phys_object(dev, i);
-}
-
-void i915_gem_detach_phys_object(struct drm_device *dev,
-				 struct drm_i915_gem_object *obj)
-{
-	struct address_space *mapping = file_inode(obj->base.filp)->i_mapping;
-	char *vaddr;
-	int i;
-	int page_count;
-
-	if (!obj->phys_obj)
-		return;
-	vaddr = obj->phys_obj->handle->vaddr;
-
-	page_count = obj->base.size / PAGE_SIZE;
-	for (i = 0; i < page_count; i++) {
-		struct page *page = shmem_read_mapping_page(mapping, i);
-		if (!IS_ERR(page)) {
-			char *dst = kmap_atomic(page);
-			memcpy(dst, vaddr + i*PAGE_SIZE, PAGE_SIZE);
-			kunmap_atomic(dst);
-
-			drm_clflush_pages(&page, 1);
-
-			set_page_dirty(page);
-			mark_page_accessed(page);
-			page_cache_release(page);
-		}
-	}
-	i915_gem_chipset_flush(dev);
-
-	obj->phys_obj->cur_obj = NULL;
-	obj->phys_obj = NULL;
-}
-
-int
-i915_gem_attach_phys_object(struct drm_device *dev,
-			    struct drm_i915_gem_object *obj,
-			    int id,
-			    int align)
-{
-	struct address_space *mapping = file_inode(obj->base.filp)->i_mapping;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	int ret = 0;
-	int page_count;
-	int i;
-
-	if (id > I915_MAX_PHYS_OBJECT)
-		return -EINVAL;
-
-	if (obj->phys_obj) {
-		if (obj->phys_obj->id == id)
-			return 0;
-		i915_gem_detach_phys_object(dev, obj);
-	}
-
-	/* create a new object */
-	if (!dev_priv->mm.phys_objs[id - 1]) {
-		ret = i915_gem_init_phys_object(dev, id,
-						obj->base.size, align);
-		if (ret) {
-			DRM_ERROR("failed to init phys object %d size: %zu\n",
-				  id, obj->base.size);
-			return ret;
-		}
-	}
-
-	/* bind to the object */
-	obj->phys_obj = dev_priv->mm.phys_objs[id - 1];
-	obj->phys_obj->cur_obj = obj;
-
-	page_count = obj->base.size / PAGE_SIZE;
-
-	for (i = 0; i < page_count; i++) {
-		struct page *page;
-		char *dst, *src;
-
-		page = shmem_read_mapping_page(mapping, i);
-		if (IS_ERR(page))
-			return PTR_ERR(page);
-
-		src = kmap_atomic(page);
-		dst = obj->phys_obj->handle->vaddr + (i * PAGE_SIZE);
-		memcpy(dst, src, PAGE_SIZE);
-		kunmap_atomic(src);
-
-		mark_page_accessed(page);
-		page_cache_release(page);
-	}
-
-	return 0;
-}
-
-static int
-i915_gem_phys_pwrite(struct drm_device *dev,
-		     struct drm_i915_gem_object *obj,
-		     struct drm_i915_gem_pwrite *args,
-		     struct drm_file *file_priv)
-{
-	void *vaddr = obj->phys_obj->handle->vaddr + args->offset;
-	char __user *user_data = to_user_ptr(args->data_ptr);
-
-	if (__copy_from_user_inatomic_nocache(vaddr, user_data, args->size)) {
-		unsigned long unwritten;
-
-		/* The physical object once assigned is fixed for the lifetime
-		 * of the obj, so we can safely drop the lock and continue
-		 * to access vaddr.
-		 */
-		mutex_unlock(&dev->struct_mutex);
-		unwritten = copy_from_user(vaddr, user_data, args->size);
-		mutex_lock(&dev->struct_mutex);
-		if (unwritten)
-			return -EFAULT;
-	}
-
-	i915_gem_chipset_flush(dev);
-	return 0;
-}
-
 void i915_gem_release(struct drm_device *dev, struct drm_file *file)
 {
 	struct drm_i915_file_private *file_priv = file->driver_priv;
diff --git a/drivers/gpu/drm/i915/i915_gem_evict.c b/drivers/gpu/drm/i915/i915_gem_evict.c
index 75fca63dc8c1..bbf4b12d842e 100644
--- a/drivers/gpu/drm/i915/i915_gem_evict.c
+++ b/drivers/gpu/drm/i915/i915_gem_evict.c
@@ -68,9 +68,9 @@ mark_free(struct i915_vma *vma, struct list_head *unwind)
 int
 i915_gem_evict_something(struct drm_device *dev, struct i915_address_space *vm,
 			 int min_size, unsigned alignment, unsigned cache_level,
+			 unsigned long start, unsigned long end,
 			 unsigned flags)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct list_head eviction_list, unwind_list;
 	struct i915_vma *vma;
 	int ret = 0;
@@ -102,11 +102,10 @@ i915_gem_evict_something(struct drm_device *dev, struct i915_address_space *vm,
 	 */
 
 	INIT_LIST_HEAD(&unwind_list);
-	if (flags & PIN_MAPPABLE) {
-		BUG_ON(!i915_is_ggtt(vm));
+	if (start != 0 || end != vm->total) {
 		drm_mm_init_scan_with_range(&vm->mm, min_size,
-					    alignment, cache_level, 0,
-					    dev_priv->gtt.mappable_end);
+					    alignment, cache_level,
+					    start, end);
 	} else
 		drm_mm_init_scan(&vm->mm, min_size, alignment, cache_level);
 
diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
index 008e208e9a3a..3a30133f93e8 100644
--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
@@ -35,6 +35,9 @@
 
 #define  __EXEC_OBJECT_HAS_PIN (1<<31)
 #define  __EXEC_OBJECT_HAS_FENCE (1<<30)
+#define  __EXEC_OBJECT_NEEDS_BIAS (1<<28)
+
+#define BATCH_OFFSET_BIAS (256*1024)
 
 struct eb_vmas {
 	struct list_head vmas;
@@ -548,7 +551,7 @@ i915_gem_execbuffer_reserve_vma(struct i915_vma *vma,
 	struct drm_i915_gem_exec_object2 *entry = vma->exec_entry;
 	bool has_fenced_gpu_access = INTEL_INFO(ring->dev)->gen < 4;
 	bool need_fence;
-	unsigned flags;
+	uint64_t flags;
 	int ret;
 
 	flags = 0;
@@ -562,6 +565,8 @@ i915_gem_execbuffer_reserve_vma(struct i915_vma *vma,
 
 	if (entry->flags & EXEC_OBJECT_NEEDS_GTT)
 		flags |= PIN_GLOBAL;
+	if (entry->flags & __EXEC_OBJECT_NEEDS_BIAS)
+		flags |= BATCH_OFFSET_BIAS | PIN_OFFSET_BIAS;
 
 	ret = i915_gem_object_pin(obj, vma->vm, entry->alignment, flags);
 	if (ret)
@@ -595,6 +600,36 @@ i915_gem_execbuffer_reserve_vma(struct i915_vma *vma,
 	return 0;
 }
 
+static bool
+eb_vma_misplaced(struct i915_vma *vma, bool has_fenced_gpu_access)
+{
+	struct drm_i915_gem_exec_object2 *entry = vma->exec_entry;
+	struct drm_i915_gem_object *obj = vma->obj;
+	bool need_fence, need_mappable;
+
+	need_fence =
+		has_fenced_gpu_access &&
+		entry->flags & EXEC_OBJECT_NEEDS_FENCE &&
+		obj->tiling_mode != I915_TILING_NONE;
+	need_mappable = need_fence || need_reloc_mappable(vma);
+
+	WARN_ON((need_mappable || need_fence) &&
+	       !i915_is_ggtt(vma->vm));
+
+	if (entry->alignment &&
+	    vma->node.start & (entry->alignment - 1))
+		return true;
+
+	if (need_mappable && !obj->map_and_fenceable)
+		return true;
+
+	if (entry->flags & __EXEC_OBJECT_NEEDS_BIAS &&
+	    vma->node.start < BATCH_OFFSET_BIAS)
+		return true;
+
+	return false;
+}
+
 static int
 i915_gem_execbuffer_reserve(struct intel_engine_cs *ring,
 			    struct list_head *vmas,
@@ -658,26 +693,10 @@ i915_gem_execbuffer_reserve(struct intel_engine_cs *ring,
 
 		/* Unbind any ill-fitting objects or pin. */
 		list_for_each_entry(vma, vmas, exec_list) {
-			struct drm_i915_gem_exec_object2 *entry = vma->exec_entry;
-			bool need_fence, need_mappable;
-
-			obj = vma->obj;
-
 			if (!drm_mm_node_allocated(&vma->node))
 				continue;
 
-			need_fence =
-				has_fenced_gpu_access &&
-				entry->flags & EXEC_OBJECT_NEEDS_FENCE &&
-				obj->tiling_mode != I915_TILING_NONE;
-			need_mappable = need_fence || need_reloc_mappable(vma);
-
-			WARN_ON((need_mappable || need_fence) &&
-			       !i915_is_ggtt(vma->vm));
-
-			if ((entry->alignment &&
-			     vma->node.start & (entry->alignment - 1)) ||
-			    (need_mappable && !obj->map_and_fenceable))
+			if (eb_vma_misplaced(vma, has_fenced_gpu_access))
 				ret = i915_vma_unbind(vma);
 			else
 				ret = i915_gem_execbuffer_reserve_vma(vma, ring, need_relocs);
@@ -778,9 +797,9 @@ i915_gem_execbuffer_relocate_slow(struct drm_device *dev,
 		 * relocations were valid.
 		 */
 		for (j = 0; j < exec[i].relocation_count; j++) {
-			if (copy_to_user(&user_relocs[j].presumed_offset,
-					 &invalid_offset,
-					 sizeof(invalid_offset))) {
+			if (__copy_to_user(&user_relocs[j].presumed_offset,
+					   &invalid_offset,
+					   sizeof(invalid_offset))) {
 				ret = -EFAULT;
 				mutex_lock(&dev->struct_mutex);
 				goto err;
@@ -1040,6 +1059,25 @@ static int gen8_dispatch_bsd_ring(struct drm_device *dev,
 	}
 }
 
+static struct drm_i915_gem_object *
+eb_get_batch(struct eb_vmas *eb)
+{
+	struct i915_vma *vma = list_entry(eb->vmas.prev, typeof(*vma), exec_list);
+
+	/*
+	 * SNA is doing fancy tricks with compressing batch buffers, which leads
+	 * to negative relocation deltas. Usually that works out ok since the
+	 * relocate address is still positive, except when the batch is placed
+	 * very low in the GTT. Ensure this doesn't happen.
+	 *
+	 * Note that actual hangs have only been observed on gen7, but for
+	 * paranoia do it everywhere.
+	 */
+	vma->exec_entry->flags |= __EXEC_OBJECT_NEEDS_BIAS;
+
+	return vma->obj;
+}
+
 static int
 i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 		       struct drm_file *file,
@@ -1220,7 +1258,7 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data,
 		goto err;
 
 	/* take note of the batch buffer before we might reorder the lists */
-	batch_obj = list_entry(eb->vmas.prev, struct i915_vma, exec_list)->obj;
+	batch_obj = eb_get_batch(eb);
 
 	/* Move the objects en-masse into the GTT, evicting if necessary. */
 	need_relocs = (args->flags & I915_EXEC_NO_RELOC) == 0;
@@ -1422,18 +1460,21 @@ i915_gem_execbuffer(struct drm_device *dev, void *data,
 
 	ret = i915_gem_do_execbuffer(dev, data, file, &exec2, exec2_list);
 	if (!ret) {
+		struct drm_i915_gem_exec_object __user *user_exec_list =
+			to_user_ptr(args->buffers_ptr);
+
 		/* Copy the new buffer offsets back to the user's exec list. */
-		for (i = 0; i < args->buffer_count; i++)
-			exec_list[i].offset = exec2_list[i].offset;
-		/* ... and back out to userspace */
-		ret = copy_to_user(to_user_ptr(args->buffers_ptr),
-				   exec_list,
-				   sizeof(*exec_list) * args->buffer_count);
-		if (ret) {
-			ret = -EFAULT;
-			DRM_DEBUG("failed to copy %d exec entries "
-				  "back to user (%d)\n",
-				  args->buffer_count, ret);
+		for (i = 0; i < args->buffer_count; i++) {
+			ret = __copy_to_user(&user_exec_list[i].offset,
+					     &exec2_list[i].offset,
+					     sizeof(user_exec_list[i].offset));
+			if (ret) {
+				ret = -EFAULT;
+				DRM_DEBUG("failed to copy %d exec entries "
+					  "back to user (%d)\n",
+					  args->buffer_count, ret);
+				break;
+			}
 		}
 	}
 
@@ -1484,14 +1525,21 @@ i915_gem_execbuffer2(struct drm_device *dev, void *data,
 	ret = i915_gem_do_execbuffer(dev, data, file, args, exec2_list);
 	if (!ret) {
 		/* Copy the new buffer offsets back to the user's exec list. */
-		ret = copy_to_user(to_user_ptr(args->buffers_ptr),
-				   exec2_list,
-				   sizeof(*exec2_list) * args->buffer_count);
-		if (ret) {
-			ret = -EFAULT;
-			DRM_DEBUG("failed to copy %d exec entries "
-				  "back to user (%d)\n",
-				  args->buffer_count, ret);
+		struct drm_i915_gem_exec_object2 *user_exec_list =
+				   to_user_ptr(args->buffers_ptr);
+		int i;
+
+		for (i = 0; i < args->buffer_count; i++) {
+			ret = __copy_to_user(&user_exec_list[i].offset,
+					     &exec2_list[i].offset,
+					     sizeof(user_exec_list[i].offset));
+			if (ret) {
+				ret = -EFAULT;
+				DRM_DEBUG("failed to copy %d exec entries "
+					  "back to user\n",
+					  args->buffer_count);
+				break;
+			}
 		}
 	}
 
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index 94916362b61c..931b906f292a 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -35,25 +35,35 @@ static void chv_setup_private_ppat(struct drm_i915_private *dev_priv);
 
 bool intel_enable_ppgtt(struct drm_device *dev, bool full)
 {
-	if (i915.enable_ppgtt == 0 || !HAS_ALIASING_PPGTT(dev))
+	if (i915.enable_ppgtt == 0)
 		return false;
 
 	if (i915.enable_ppgtt == 1 && full)
 		return false;
 
+	return true;
+}
+
+static int sanitize_enable_ppgtt(struct drm_device *dev, int enable_ppgtt)
+{
+	if (enable_ppgtt == 0 || !HAS_ALIASING_PPGTT(dev))
+		return 0;
+
+	if (enable_ppgtt == 1)
+		return 1;
+
+	if (enable_ppgtt == 2 && HAS_PPGTT(dev))
+		return 2;
+
 #ifdef CONFIG_INTEL_IOMMU
 	/* Disable ppgtt on SNB if VT-d is on. */
 	if (INTEL_INFO(dev)->gen == 6 && intel_iommu_gfx_mapped) {
 		DRM_INFO("Disabling PPGTT because VT-d is on\n");
-		return false;
+		return 0;
 	}
 #endif
 
-	/* Full ppgtt disabled by default for now due to issues. */
-	if (full)
-		return HAS_PPGTT(dev) && (i915.enable_ppgtt == 2);
-	else
-		return HAS_ALIASING_PPGTT(dev);
+	return HAS_ALIASING_PPGTT(dev) ? 1 : 0;
 }
 
 
@@ -1039,7 +1049,9 @@ alloc:
 	if (ret == -ENOSPC && !retried) {
 		ret = i915_gem_evict_something(dev, &dev_priv->gtt.base,
 					       GEN6_PD_SIZE, GEN6_PD_ALIGN,
-					       I915_CACHE_NONE, 0);
+					       I915_CACHE_NONE,
+					       0, dev_priv->gtt.base.total,
+					       0);
 		if (ret)
 			return ret;
 
@@ -2052,6 +2064,14 @@ int i915_gem_gtt_init(struct drm_device *dev)
 	if (intel_iommu_gfx_mapped)
 		DRM_INFO("VT-d active for gfx access\n");
 #endif
+	/*
+	 * i915.enable_ppgtt is read-only, so do an early pass to validate the
+	 * user's requested state against the hardware/driver capabilities.  We
+	 * do this now so that we can print out any log messages once rather
+	 * than every time we check intel_enable_ppgtt().
+	 */
+	i915.enable_ppgtt = sanitize_enable_ppgtt(dev, i915.enable_ppgtt);
+	DRM_DEBUG_DRIVER("ppgtt mode: %i\n", i915.enable_ppgtt);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/i915/intel_bios.c b/drivers/gpu/drm/i915/intel_bios.c
index 2945f57c53ee..6b6509656f16 100644
--- a/drivers/gpu/drm/i915/intel_bios.c
+++ b/drivers/gpu/drm/i915/intel_bios.c
@@ -598,47 +598,71 @@ parse_edp(struct drm_i915_private *dev_priv, struct bdb_header *bdb)
 
 	dev_priv->vbt.edp_pps = *edp_pps;
 
-	dev_priv->vbt.edp_rate = edp_link_params->rate ? DP_LINK_BW_2_7 :
-		DP_LINK_BW_1_62;
+	switch (edp_link_params->rate) {
+	case EDP_RATE_1_62:
+		dev_priv->vbt.edp_rate = DP_LINK_BW_1_62;
+		break;
+	case EDP_RATE_2_7:
+		dev_priv->vbt.edp_rate = DP_LINK_BW_2_7;
+		break;
+	default:
+		DRM_DEBUG_KMS("VBT has unknown eDP link rate value %u\n",
+			      edp_link_params->rate);
+		break;
+	}
+
 	switch (edp_link_params->lanes) {
-	case 0:
+	case EDP_LANE_1:
 		dev_priv->vbt.edp_lanes = 1;
 		break;
-	case 1:
+	case EDP_LANE_2:
 		dev_priv->vbt.edp_lanes = 2;
 		break;
-	case 3:
-	default:
+	case EDP_LANE_4:
 		dev_priv->vbt.edp_lanes = 4;
 		break;
+	default:
+		DRM_DEBUG_KMS("VBT has unknown eDP lane count value %u\n",
+			      edp_link_params->lanes);
+		break;
 	}
+
 	switch (edp_link_params->preemphasis) {
-	case 0:
+	case EDP_PREEMPHASIS_NONE:
 		dev_priv->vbt.edp_preemphasis = DP_TRAIN_PRE_EMPHASIS_0;
 		break;
-	case 1:
+	case EDP_PREEMPHASIS_3_5dB:
 		dev_priv->vbt.edp_preemphasis = DP_TRAIN_PRE_EMPHASIS_3_5;
 		break;
-	case 2:
+	case EDP_PREEMPHASIS_6dB:
 		dev_priv->vbt.edp_preemphasis = DP_TRAIN_PRE_EMPHASIS_6;
 		break;
-	case 3:
+	case EDP_PREEMPHASIS_9_5dB:
 		dev_priv->vbt.edp_preemphasis = DP_TRAIN_PRE_EMPHASIS_9_5;
 		break;
+	default:
+		DRM_DEBUG_KMS("VBT has unknown eDP pre-emphasis value %u\n",
+			      edp_link_params->preemphasis);
+		break;
 	}
+
 	switch (edp_link_params->vswing) {
-	case 0:
+	case EDP_VSWING_0_4V:
 		dev_priv->vbt.edp_vswing = DP_TRAIN_VOLTAGE_SWING_400;
 		break;
-	case 1:
+	case EDP_VSWING_0_6V:
 		dev_priv->vbt.edp_vswing = DP_TRAIN_VOLTAGE_SWING_600;
 		break;
-	case 2:
+	case EDP_VSWING_0_8V:
 		dev_priv->vbt.edp_vswing = DP_TRAIN_VOLTAGE_SWING_800;
 		break;
-	case 3:
+	case EDP_VSWING_1_2V:
 		dev_priv->vbt.edp_vswing = DP_TRAIN_VOLTAGE_SWING_1200;
 		break;
+	default:
+		DRM_DEBUG_KMS("VBT has unknown eDP voltage swing value %u\n",
+			      edp_link_params->vswing);
+		break;
 	}
 }
 
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 1ce4ad4626e4..7a4c7c98378a 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -8112,14 +8112,12 @@ static int intel_crtc_cursor_set(struct drm_crtc *crtc,
 		addr = i915_gem_obj_ggtt_offset(obj);
 	} else {
 		int align = IS_I830(dev) ? 16 * 1024 : 256;
-		ret = i915_gem_attach_phys_object(dev, obj,
-						  (intel_crtc->pipe == 0) ? I915_GEM_PHYS_CURSOR_0 : I915_GEM_PHYS_CURSOR_1,
-						  align);
+		ret = i915_gem_object_attach_phys(obj, align);
 		if (ret) {
 			DRM_DEBUG_KMS("failed to attach phys object\n");
 			goto fail_locked;
 		}
-		addr = obj->phys_obj->handle->busaddr;
+		addr = obj->phys_handle->busaddr;
 	}
 
 	if (IS_GEN2(dev))
@@ -8127,10 +8125,7 @@ static int intel_crtc_cursor_set(struct drm_crtc *crtc,
 
  finish:
 	if (intel_crtc->cursor_bo) {
-		if (INTEL_INFO(dev)->cursor_needs_physical) {
-			if (intel_crtc->cursor_bo != obj)
-				i915_gem_detach_phys_object(dev, intel_crtc->cursor_bo);
-		} else
+		if (!INTEL_INFO(dev)->cursor_needs_physical)
 			i915_gem_object_unpin_from_display_plane(intel_crtc->cursor_bo);
 		drm_gem_object_unreference(&intel_crtc->cursor_bo->base);
 	}
@@ -11808,15 +11803,6 @@ void intel_modeset_init(struct drm_device *dev)
 	}
 }
 
-static void
-intel_connector_break_all_links(struct intel_connector *connector)
-{
-	connector->base.dpms = DRM_MODE_DPMS_OFF;
-	connector->base.encoder = NULL;
-	connector->encoder->connectors_active = false;
-	connector->encoder->base.crtc = NULL;
-}
-
 static void intel_enable_pipe_a(struct drm_device *dev)
 {
 	struct intel_connector *connector;
@@ -11905,8 +11891,17 @@ static void intel_sanitize_crtc(struct intel_crtc *crtc)
 			if (connector->encoder->base.crtc != &crtc->base)
 				continue;
 
-			intel_connector_break_all_links(connector);
+			connector->base.dpms = DRM_MODE_DPMS_OFF;
+			connector->base.encoder = NULL;
 		}
+		/* multiple connectors may have the same encoder:
+		 *  handle them and break crtc link separately */
+		list_for_each_entry(connector, &dev->mode_config.connector_list,
+				    base.head)
+			if (connector->encoder->base.crtc == &crtc->base) {
+				connector->encoder->base.crtc = NULL;
+				connector->encoder->connectors_active = false;
+			}
 
 		WARN_ON(crtc->active);
 		crtc->base.enabled = false;
@@ -11997,6 +11992,8 @@ static void intel_sanitize_encoder(struct intel_encoder *encoder)
 				      encoder->base.name);
 			encoder->disable(encoder);
 		}
+		encoder->base.crtc = NULL;
+		encoder->connectors_active = false;
 
 		/* Inconsistent output/port/pipe state happens presumably due to
 		 * a bug in one of the get_hw_state functions. Or someplace else
@@ -12007,8 +12004,8 @@ static void intel_sanitize_encoder(struct intel_encoder *encoder)
 				    base.head) {
 			if (connector->encoder != encoder)
 				continue;
-
-			intel_connector_break_all_links(connector);
+			connector->base.dpms = DRM_MODE_DPMS_OFF;
+			connector->base.encoder = NULL;
 		}
 	}
 	/* Enabled encoders without active connectors will be fixed in
diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
index 2d5d9b010073..52fda950fd2a 100644
--- a/drivers/gpu/drm/i915/intel_dp.c
+++ b/drivers/gpu/drm/i915/intel_dp.c
@@ -123,7 +123,8 @@ intel_dp_max_link_bw(struct intel_dp *intel_dp)
 	case DP_LINK_BW_2_7:
 		break;
 	case DP_LINK_BW_5_4: /* 1.2 capable displays may advertise higher bw */
-		if ((IS_HASWELL(dev) || INTEL_INFO(dev)->gen >= 8) &&
+		if (((IS_HASWELL(dev) && !IS_HSW_ULX(dev)) ||
+		     INTEL_INFO(dev)->gen >= 8) &&
 		    intel_dp->dpcd[DP_DPCD_REV] >= 0x12)
 			max_link_bw = DP_LINK_BW_5_4;
 		else
@@ -138,6 +139,22 @@ intel_dp_max_link_bw(struct intel_dp *intel_dp)
 	return max_link_bw;
 }
 
+static u8 intel_dp_max_lane_count(struct intel_dp *intel_dp)
+{
+	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+	struct drm_device *dev = intel_dig_port->base.base.dev;
+	u8 source_max, sink_max;
+
+	source_max = 4;
+	if (HAS_DDI(dev) && intel_dig_port->port == PORT_A &&
+	    (intel_dig_port->saved_port_bits & DDI_A_4_LANES) == 0)
+		source_max = 2;
+
+	sink_max = drm_dp_max_lane_count(intel_dp->dpcd);
+
+	return min(source_max, sink_max);
+}
+
 /*
  * The units on the numbers in the next two are... bizarre.  Examples will
  * make it clearer; this one parallels an example in the eDP spec.
@@ -188,7 +205,7 @@ intel_dp_mode_valid(struct drm_connector *connector,
 	}
 
 	max_link_clock = drm_dp_bw_code_to_link_rate(intel_dp_max_link_bw(intel_dp));
-	max_lanes = drm_dp_max_lane_count(intel_dp->dpcd);
+	max_lanes = intel_dp_max_lane_count(intel_dp);
 
 	max_rate = intel_dp_max_data_rate(max_link_clock, max_lanes);
 	mode_rate = intel_dp_link_required(target_clock, 18);
@@ -789,8 +806,10 @@ intel_dp_compute_config(struct intel_encoder *encoder,
 	struct intel_crtc *intel_crtc = encoder->new_crtc;
 	struct intel_connector *intel_connector = intel_dp->attached_connector;
 	int lane_count, clock;
-	int max_lane_count = drm_dp_max_lane_count(intel_dp->dpcd);
+	int min_lane_count = 1;
+	int max_lane_count = intel_dp_max_lane_count(intel_dp);
 	/* Conveniently, the link BW constants become indices with a shift...*/
+	int min_clock = 0;
 	int max_clock = intel_dp_max_link_bw(intel_dp) >> 3;
 	int bpp, mode_rate;
 	static int bws[] = { DP_LINK_BW_1_62, DP_LINK_BW_2_7, DP_LINK_BW_5_4 };
@@ -824,19 +843,38 @@ intel_dp_compute_config(struct intel_encoder *encoder,
 	/* Walk through all bpp values. Luckily they're all nicely spaced with 2
 	 * bpc in between. */
 	bpp = pipe_config->pipe_bpp;
-	if (is_edp(intel_dp) && dev_priv->vbt.edp_bpp &&
-	    dev_priv->vbt.edp_bpp < bpp) {
-		DRM_DEBUG_KMS("clamping bpp for eDP panel to BIOS-provided %i\n",
-			      dev_priv->vbt.edp_bpp);
-		bpp = dev_priv->vbt.edp_bpp;
+	if (is_edp(intel_dp)) {
+		if (dev_priv->vbt.edp_bpp && dev_priv->vbt.edp_bpp < bpp) {
+			DRM_DEBUG_KMS("clamping bpp for eDP panel to BIOS-provided %i\n",
+				      dev_priv->vbt.edp_bpp);
+			bpp = dev_priv->vbt.edp_bpp;
+		}
+
+		if (IS_BROADWELL(dev)) {
+			/* Yes, it's an ugly hack. */
+			min_lane_count = max_lane_count;
+			DRM_DEBUG_KMS("forcing lane count to max (%u) on BDW\n",
+				      min_lane_count);
+		} else if (dev_priv->vbt.edp_lanes) {
+			min_lane_count = min(dev_priv->vbt.edp_lanes,
+					     max_lane_count);
+			DRM_DEBUG_KMS("using min %u lanes per VBT\n",
+				      min_lane_count);
+		}
+
+		if (dev_priv->vbt.edp_rate) {
+			min_clock = min(dev_priv->vbt.edp_rate >> 3, max_clock);
+			DRM_DEBUG_KMS("using min %02x link bw per VBT\n",
+				      bws[min_clock]);
+		}
 	}
 
 	for (; bpp >= 6*3; bpp -= 2*3) {
 		mode_rate = intel_dp_link_required(adjusted_mode->crtc_clock,
 						   bpp);
 
-		for (lane_count = 1; lane_count <= max_lane_count; lane_count <<= 1) {
-			for (clock = 0; clock <= max_clock; clock++) {
+		for (lane_count = min_lane_count; lane_count <= max_lane_count; lane_count <<= 1) {
+			for (clock = min_clock; clock <= max_clock; clock++) {
 				link_clock = drm_dp_bw_code_to_link_rate(bws[clock]);
 				link_avail = intel_dp_max_data_rate(link_clock,
 								    lane_count);
diff --git a/drivers/gpu/drm/i915/intel_fbdev.c b/drivers/gpu/drm/i915/intel_fbdev.c
index 6ea2d75464da..088fe9378a4c 100644
--- a/drivers/gpu/drm/i915/intel_fbdev.c
+++ b/drivers/gpu/drm/i915/intel_fbdev.c
@@ -387,6 +387,15 @@ static bool intel_fb_initial_config(struct drm_fb_helper *fb_helper,
 							  height);
 		}
 
+		/* No preferred mode marked by the EDID? Are there any modes? */
+		if (!modes[i] && !list_empty(&connector->modes)) {
+			DRM_DEBUG_KMS("using first mode listed on connector %s\n",
+				      connector->name);
+			modes[i] = list_first_entry(&connector->modes,
+						    struct drm_display_mode,
+						    head);
+		}
+
 		/* last resort: use current mode */
 		if (!modes[i]) {
 			/*
diff --git a/drivers/gpu/drm/i915/intel_overlay.c b/drivers/gpu/drm/i915/intel_overlay.c
index 0396d1312b5c..daa118978eec 100644
--- a/drivers/gpu/drm/i915/intel_overlay.c
+++ b/drivers/gpu/drm/i915/intel_overlay.c
@@ -193,7 +193,7 @@ intel_overlay_map_regs(struct intel_overlay *overlay)
 	struct overlay_registers __iomem *regs;
 
 	if (OVERLAY_NEEDS_PHYSICAL(overlay->dev))
-		regs = (struct overlay_registers __iomem *)overlay->reg_bo->phys_obj->handle->vaddr;
+		regs = (struct overlay_registers __iomem *)overlay->reg_bo->phys_handle->vaddr;
 	else
 		regs = io_mapping_map_wc(dev_priv->gtt.mappable,
 					 i915_gem_obj_ggtt_offset(overlay->reg_bo));
@@ -1340,14 +1340,12 @@ void intel_setup_overlay(struct drm_device *dev)
 	overlay->reg_bo = reg_bo;
 
 	if (OVERLAY_NEEDS_PHYSICAL(dev)) {
-		ret = i915_gem_attach_phys_object(dev, reg_bo,
-						  I915_GEM_PHYS_OVERLAY_REGS,
-						  PAGE_SIZE);
+		ret = i915_gem_object_attach_phys(reg_bo, PAGE_SIZE);
 		if (ret) {
 			DRM_ERROR("failed to attach phys overlay regs\n");
 			goto out_free_bo;
 		}
-		overlay->flip_addr = reg_bo->phys_obj->handle->busaddr;
+		overlay->flip_addr = reg_bo->phys_handle->busaddr;
 	} else {
 		ret = i915_gem_obj_ggtt_pin(reg_bo, PAGE_SIZE, PIN_MAPPABLE);
 		if (ret) {
@@ -1428,7 +1426,7 @@ intel_overlay_map_regs_atomic(struct intel_overlay *overlay)
 		/* Cast to make sparse happy, but it's wc memory anyway, so
 		 * equivalent to the wc io mapping on X86. */
 		regs = (struct overlay_registers __iomem *)
-			overlay->reg_bo->phys_obj->handle->vaddr;
+			overlay->reg_bo->phys_handle->vaddr;
 	else
 		regs = io_mapping_map_atomic_wc(dev_priv->gtt.mappable,
 						i915_gem_obj_ggtt_offset(overlay->reg_bo));
@@ -1462,7 +1460,7 @@ intel_overlay_capture_error_state(struct drm_device *dev)
 	error->dovsta = I915_READ(DOVSTA);
 	error->isr = I915_READ(ISR);
 	if (OVERLAY_NEEDS_PHYSICAL(overlay->dev))
-		error->base = (__force long)overlay->reg_bo->phys_obj->handle->vaddr;
+		error->base = (__force long)overlay->reg_bo->phys_handle->vaddr;
 	else
 		error->base = i915_gem_obj_ggtt_offset(overlay->reg_bo);
 
diff --git a/drivers/gpu/drm/i915/intel_panel.c b/drivers/gpu/drm/i915/intel_panel.c
index 2e1338a5d488..5e6c888b4928 100644
--- a/drivers/gpu/drm/i915/intel_panel.c
+++ b/drivers/gpu/drm/i915/intel_panel.c
@@ -567,6 +567,7 @@ void intel_panel_set_backlight(struct intel_connector *connector, u32 level,
 	enum pipe pipe = intel_get_pipe_from_connector(connector);
 	u32 freq;
 	unsigned long flags;
+	u64 n;
 
 	if (!panel->backlight.present || pipe == INVALID_PIPE)
 		return;
@@ -577,10 +578,9 @@ void intel_panel_set_backlight(struct intel_connector *connector, u32 level,
 
 	/* 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;
+	n = (u64)level * freq;
+	do_div(n, max);
+	level = n;
 
 	panel->backlight.level = level;
 	if (panel->backlight.device)
diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
index b86b58c44228..906d06f73e51 100644
--- a/drivers/gpu/drm/i915/intel_pm.c
+++ b/drivers/gpu/drm/i915/intel_pm.c
@@ -2113,6 +2113,43 @@ static void intel_print_wm_latency(struct drm_device *dev,
 	}
 }
 
+static bool ilk_increase_wm_latency(struct drm_i915_private *dev_priv,
+				    uint16_t wm[5], uint16_t min)
+{
+	int level, max_level = ilk_wm_max_level(dev_priv->dev);
+
+	if (wm[0] >= min)
+		return false;
+
+	wm[0] = max(wm[0], min);
+	for (level = 1; level <= max_level; level++)
+		wm[level] = max_t(uint16_t, wm[level], DIV_ROUND_UP(min, 5));
+
+	return true;
+}
+
+static void snb_wm_latency_quirk(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	bool changed;
+
+	/*
+	 * The BIOS provided WM memory latency values are often
+	 * inadequate for high resolution displays. Adjust them.
+	 */
+	changed = ilk_increase_wm_latency(dev_priv, dev_priv->wm.pri_latency, 12) |
+		ilk_increase_wm_latency(dev_priv, dev_priv->wm.spr_latency, 12) |
+		ilk_increase_wm_latency(dev_priv, dev_priv->wm.cur_latency, 12);
+
+	if (!changed)
+		return;
+
+	DRM_DEBUG_KMS("WM latency values increased to avoid potential underruns\n");
+	intel_print_wm_latency(dev, "Primary", dev_priv->wm.pri_latency);
+	intel_print_wm_latency(dev, "Sprite", dev_priv->wm.spr_latency);
+	intel_print_wm_latency(dev, "Cursor", dev_priv->wm.cur_latency);
+}
+
 static void ilk_setup_wm_latency(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -2130,6 +2167,9 @@ static void ilk_setup_wm_latency(struct drm_device *dev)
 	intel_print_wm_latency(dev, "Primary", dev_priv->wm.pri_latency);
 	intel_print_wm_latency(dev, "Sprite", dev_priv->wm.spr_latency);
 	intel_print_wm_latency(dev, "Cursor", dev_priv->wm.cur_latency);
+
+	if (IS_GEN6(dev))
+		snb_wm_latency_quirk(dev);
 }
 
 static void ilk_compute_wm_parameters(struct drm_crtc *crtc,
diff --git a/drivers/gpu/drm/i915/intel_uncore.c b/drivers/gpu/drm/i915/intel_uncore.c
index 9cd99d9676fd..2f5d5d3f0043 100644
--- a/drivers/gpu/drm/i915/intel_uncore.c
+++ b/drivers/gpu/drm/i915/intel_uncore.c
@@ -185,6 +185,8 @@ static void vlv_force_wake_reset(struct drm_i915_private *dev_priv)
 {
 	__raw_i915_write32(dev_priv, FORCEWAKE_VLV,
 			   _MASKED_BIT_DISABLE(0xffff));
+	__raw_i915_write32(dev_priv, FORCEWAKE_MEDIA_VLV,
+			   _MASKED_BIT_DISABLE(0xffff));
 	/* something from same cacheline, but !FORCEWAKE_VLV */
 	__raw_posting_read(dev_priv, FORCEWAKE_ACK_VLV);
 }
diff --git a/drivers/gpu/drm/nouveau/core/engine/disp/nvd0.c b/drivers/gpu/drm/nouveau/core/engine/disp/nvd0.c
index 7762665ad8fd..876de9ac3793 100644
--- a/drivers/gpu/drm/nouveau/core/engine/disp/nvd0.c
+++ b/drivers/gpu/drm/nouveau/core/engine/disp/nvd0.c
@@ -1009,7 +1009,7 @@ exec_clkcmp(struct nv50_disp_priv *priv, int head, int id,
 	}
 
 	if (outp == 8)
-		return false;
+		return conf;
 
 	data = exec_lookup(priv, head, outp, ctrl, dcb, &ver, &hdr, &cnt, &len, &info1);
 	if (data == 0x0000)
diff --git a/drivers/gpu/drm/nouveau/core/engine/graph/ctxgm107.c b/drivers/gpu/drm/nouveau/core/engine/graph/ctxgm107.c
index 1dc37b1ddbfa..b0d0fb2f4d08 100644
--- a/drivers/gpu/drm/nouveau/core/engine/graph/ctxgm107.c
+++ b/drivers/gpu/drm/nouveau/core/engine/graph/ctxgm107.c
@@ -863,7 +863,7 @@ gm107_grctx_generate_mods(struct nvc0_graph_priv *priv, struct nvc0_grctx *info)
 {
 	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_data(0x200000, 0x1000, NV_MEM_ACCESS_RW);
 
 	mmio_list(0x40800c, 0x00000000,  8, 1);
 	mmio_list(0x408010, 0x80000000,  0, 0);
@@ -877,6 +877,8 @@ gm107_grctx_generate_mods(struct nvc0_graph_priv *priv, struct nvc0_grctx *info)
 	mmio_list(0x418e24, 0x00000000,  8, 0);
 	mmio_list(0x418e28, 0x80000030,  0, 0);
 
+	mmio_list(0x4064c8, 0x018002c0,  0, 0);
+
 	mmio_list(0x418810, 0x80000000, 12, 2);
 	mmio_list(0x419848, 0x10000000, 12, 2);
 	mmio_list(0x419c2c, 0x10000000, 12, 2);
diff --git a/drivers/gpu/drm/nouveau/core/subdev/bios/base.c b/drivers/gpu/drm/nouveau/core/subdev/bios/base.c
index fb0b6b2d1427..222e8ebb669d 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/bios/base.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/bios/base.c
@@ -168,7 +168,8 @@ nouveau_bios_shadow_prom(struct nouveau_bios *bios)
 	 */
 	i = 16;
 	do {
-		if ((nv_rd32(bios, 0x300000) & 0xffff) == 0xaa55)
+		u32 data = le32_to_cpu(nv_rd32(bios, 0x300000)) & 0xffff;
+		if (data == 0xaa55)
 			break;
 	} while (i--);
 
@@ -176,14 +177,15 @@ nouveau_bios_shadow_prom(struct nouveau_bios *bios)
 		goto out;
 
 	/* read entire bios image to system memory */
-	bios->size = ((nv_rd32(bios, 0x300000) >> 16) & 0xff) * 512;
+	bios->size = (le32_to_cpu(nv_rd32(bios, 0x300000)) >> 16) & 0xff;
+	bios->size = bios->size * 512;
 	if (!bios->size)
 		goto out;
 
 	bios->data = kmalloc(bios->size, GFP_KERNEL);
 	if (bios->data) {
-		for (i = 0; i < bios->size; i+=4)
-			nv_wo32(bios, i, nv_rd32(bios, 0x300000 + i));
+		for (i = 0; i < bios->size; i += 4)
+			((u32 *)bios->data)[i/4] = nv_rd32(bios, 0x300000 + i);
 	}
 
 	/* check the PCI record header */
diff --git a/drivers/gpu/drm/nouveau/core/subdev/therm/nvd0.c b/drivers/gpu/drm/nouveau/core/subdev/therm/nvd0.c
index 43fec17ea540..bbf117be572f 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/therm/nvd0.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/therm/nvd0.c
@@ -40,6 +40,7 @@ pwm_info(struct nouveau_therm *therm, int line)
 		case 0x00: return 2;
 		case 0x19: return 1;
 		case 0x1c: return 0;
+		case 0x1e: return 2;
 		default:
 			break;
 		}
diff --git a/drivers/gpu/drm/nouveau/nouveau_acpi.c b/drivers/gpu/drm/nouveau/nouveau_acpi.c
index 83face3f608f..279206997e5c 100644
--- a/drivers/gpu/drm/nouveau/nouveau_acpi.c
+++ b/drivers/gpu/drm/nouveau/nouveau_acpi.c
@@ -389,9 +389,6 @@ bool nouveau_acpi_rom_supported(struct pci_dev *pdev)
 	acpi_status status;
 	acpi_handle dhandle, rom_handle;
 
-	if (!nouveau_dsm_priv.dsm_detected && !nouveau_dsm_priv.optimus_detected)
-		return false;
-
 	dhandle = ACPI_HANDLE(&pdev->dev);
 	if (!dhandle)
 		return false;
diff --git a/drivers/gpu/drm/nouveau/nouveau_display.c b/drivers/gpu/drm/nouveau/nouveau_display.c
index 3ff030dc1ee3..da764a4ed958 100644
--- a/drivers/gpu/drm/nouveau/nouveau_display.c
+++ b/drivers/gpu/drm/nouveau/nouveau_display.c
@@ -764,9 +764,9 @@ nouveau_crtc_page_flip(struct drm_crtc *crtc, struct drm_framebuffer *fb,
 	}
 
 	ret = nouveau_page_flip_emit(chan, old_bo, new_bo, s, &fence);
-	mutex_unlock(&chan->cli->mutex);
 	if (ret)
 		goto fail_unreserve;
+	mutex_unlock(&chan->cli->mutex);
 
 	/* Update the crtc struct and cleanup */
 	crtc->primary->fb = fb;
diff --git a/drivers/gpu/drm/radeon/atombios_crtc.c b/drivers/gpu/drm/radeon/atombios_crtc.c
index 967d193d36d0..76c30f2da3fb 100644
--- a/drivers/gpu/drm/radeon/atombios_crtc.c
+++ b/drivers/gpu/drm/radeon/atombios_crtc.c
@@ -270,8 +270,6 @@ void atombios_crtc_dpms(struct drm_crtc *crtc, int mode)
 	switch (mode) {
 	case DRM_MODE_DPMS_ON:
 		radeon_crtc->enabled = true;
-		/* adjust pm to dpms changes BEFORE enabling crtcs */
-		radeon_pm_compute_clocks(rdev);
 		atombios_enable_crtc(crtc, ATOM_ENABLE);
 		if (ASIC_IS_DCE3(rdev) && !ASIC_IS_DCE6(rdev))
 			atombios_enable_crtc_memreq(crtc, ATOM_ENABLE);
@@ -289,10 +287,10 @@ void atombios_crtc_dpms(struct drm_crtc *crtc, int mode)
 			atombios_enable_crtc_memreq(crtc, ATOM_DISABLE);
 		atombios_enable_crtc(crtc, ATOM_DISABLE);
 		radeon_crtc->enabled = false;
-		/* adjust pm to dpms changes AFTER disabling crtcs */
-		radeon_pm_compute_clocks(rdev);
 		break;
 	}
+	/* adjust pm to dpms */
+	radeon_pm_compute_clocks(rdev);
 }
 
 static void
@@ -1208,27 +1206,43 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc,
 
 		/* Set NUM_BANKS. */
 		if (rdev->family >= CHIP_TAHITI) {
-			unsigned tileb, index, num_banks, tile_split_bytes;
+			unsigned index, num_banks;
 
-			/* Calculate the macrotile mode index. */
-			tile_split_bytes = 64 << tile_split;
-			tileb = 8 * 8 * target_fb->bits_per_pixel / 8;
-			tileb = min(tile_split_bytes, tileb);
+			if (rdev->family >= CHIP_BONAIRE) {
+				unsigned tileb, tile_split_bytes;
 
-			for (index = 0; tileb > 64; index++) {
-				tileb >>= 1;
-			}
+				/* Calculate the macrotile mode index. */
+				tile_split_bytes = 64 << tile_split;
+				tileb = 8 * 8 * target_fb->bits_per_pixel / 8;
+				tileb = min(tile_split_bytes, tileb);
 
-			if (index >= 16) {
-				DRM_ERROR("Wrong screen bpp (%u) or tile split (%u)\n",
-					  target_fb->bits_per_pixel, tile_split);
-				return -EINVAL;
-			}
+				for (index = 0; tileb > 64; index++)
+					tileb >>= 1;
+
+				if (index >= 16) {
+					DRM_ERROR("Wrong screen bpp (%u) or tile split (%u)\n",
+						  target_fb->bits_per_pixel, tile_split);
+					return -EINVAL;
+				}
 
-			if (rdev->family >= CHIP_BONAIRE)
 				num_banks = (rdev->config.cik.macrotile_mode_array[index] >> 6) & 0x3;
-			else
+			} else {
+				switch (target_fb->bits_per_pixel) {
+				case 8:
+					index = 10;
+					break;
+				case 16:
+					index = SI_TILE_MODE_COLOR_2D_SCANOUT_16BPP;
+					break;
+				default:
+				case 32:
+					index = SI_TILE_MODE_COLOR_2D_SCANOUT_32BPP;
+					break;
+				}
+
 				num_banks = (rdev->config.si.tile_mode_array[index] >> 20) & 0x3;
+			}
+
 			fb_format |= EVERGREEN_GRPH_NUM_BANKS(num_banks);
 		} else {
 			/* NI and older. */
@@ -1751,8 +1765,9 @@ static int radeon_atom_pick_pll(struct drm_crtc *crtc)
 		}
 		/* otherwise, pick one of the plls */
 		if ((rdev->family == CHIP_KAVERI) ||
-		    (rdev->family == CHIP_KABINI)) {
-			/* KB/KV has PPLL1 and PPLL2 */
+		    (rdev->family == CHIP_KABINI) ||
+		    (rdev->family == CHIP_MULLINS)) {
+			/* KB/KV/ML has PPLL1 and PPLL2 */
 			pll_in_use = radeon_get_pll_use_mask(crtc);
 			if (!(pll_in_use & (1 << ATOM_PPLL2)))
 				return ATOM_PPLL2;
@@ -1916,6 +1931,9 @@ int atombios_crtc_mode_set(struct drm_crtc *crtc,
 	    (ATOM_DEVICE_TV_SUPPORT | ATOM_DEVICE_CV_SUPPORT))
 		is_tvcv = true;
 
+	if (!radeon_crtc->adjusted_clock)
+		return -EINVAL;
+
 	atombios_crtc_set_pll(crtc, adjusted_mode);
 
 	if (ASIC_IS_DCE4(rdev))
diff --git a/drivers/gpu/drm/radeon/atombios_dp.c b/drivers/gpu/drm/radeon/atombios_dp.c
index a54c44181a0f..c5b1f2da3954 100644
--- a/drivers/gpu/drm/radeon/atombios_dp.c
+++ b/drivers/gpu/drm/radeon/atombios_dp.c
@@ -390,11 +390,11 @@ static void radeon_dp_probe_oui(struct radeon_connector *radeon_connector)
 	if (!(dig_connector->dpcd[DP_DOWN_STREAM_PORT_COUNT] & DP_OUI_SUPPORT))
 		return;
 
-	if (drm_dp_dpcd_read(&radeon_connector->ddc_bus->aux, DP_SINK_OUI, buf, 3))
+	if (drm_dp_dpcd_read(&radeon_connector->ddc_bus->aux, DP_SINK_OUI, buf, 3) == 3)
 		DRM_DEBUG_KMS("Sink OUI: %02hx%02hx%02hx\n",
 			      buf[0], buf[1], buf[2]);
 
-	if (drm_dp_dpcd_read(&radeon_connector->ddc_bus->aux, DP_BRANCH_OUI, buf, 3))
+	if (drm_dp_dpcd_read(&radeon_connector->ddc_bus->aux, DP_BRANCH_OUI, buf, 3) == 3)
 		DRM_DEBUG_KMS("Branch OUI: %02hx%02hx%02hx\n",
 			      buf[0], buf[1], buf[2]);
 }
@@ -443,21 +443,23 @@ int radeon_dp_get_panel_mode(struct drm_encoder *encoder,
 
 	if (dp_bridge != ENCODER_OBJECT_ID_NONE) {
 		/* DP bridge chips */
-		drm_dp_dpcd_readb(&radeon_connector->ddc_bus->aux,
-				  DP_EDP_CONFIGURATION_CAP, &tmp);
-		if (tmp & 1)
-			panel_mode = DP_PANEL_MODE_INTERNAL_DP2_MODE;
-		else if ((dp_bridge == ENCODER_OBJECT_ID_NUTMEG) ||
-			 (dp_bridge == ENCODER_OBJECT_ID_TRAVIS))
-			panel_mode = DP_PANEL_MODE_INTERNAL_DP1_MODE;
-		else
-			panel_mode = DP_PANEL_MODE_EXTERNAL_DP_MODE;
+		if (drm_dp_dpcd_readb(&radeon_connector->ddc_bus->aux,
+				      DP_EDP_CONFIGURATION_CAP, &tmp) == 1) {
+			if (tmp & 1)
+				panel_mode = DP_PANEL_MODE_INTERNAL_DP2_MODE;
+			else if ((dp_bridge == ENCODER_OBJECT_ID_NUTMEG) ||
+				 (dp_bridge == ENCODER_OBJECT_ID_TRAVIS))
+				panel_mode = DP_PANEL_MODE_INTERNAL_DP1_MODE;
+			else
+				panel_mode = DP_PANEL_MODE_EXTERNAL_DP_MODE;
+		}
 	} else if (connector->connector_type == DRM_MODE_CONNECTOR_eDP) {
 		/* eDP */
-		drm_dp_dpcd_readb(&radeon_connector->ddc_bus->aux,
-				  DP_EDP_CONFIGURATION_CAP, &tmp);
-		if (tmp & 1)
-			panel_mode = DP_PANEL_MODE_INTERNAL_DP2_MODE;
+		if (drm_dp_dpcd_readb(&radeon_connector->ddc_bus->aux,
+				      DP_EDP_CONFIGURATION_CAP, &tmp) == 1) {
+			if (tmp & 1)
+				panel_mode = DP_PANEL_MODE_INTERNAL_DP2_MODE;
+		}
 	}
 
 	return panel_mode;
@@ -833,11 +835,15 @@ void radeon_dp_link_train(struct drm_encoder *encoder,
 	else
 		dp_info.enc_id |= ATOM_DP_CONFIG_LINK_A;
 
-	drm_dp_dpcd_readb(&radeon_connector->ddc_bus->aux, DP_MAX_LANE_COUNT, &tmp);
-	if (ASIC_IS_DCE5(rdev) && (tmp & DP_TPS3_SUPPORTED))
-		dp_info.tp3_supported = true;
-	else
+	if (drm_dp_dpcd_readb(&radeon_connector->ddc_bus->aux, DP_MAX_LANE_COUNT, &tmp)
+	    == 1) {
+		if (ASIC_IS_DCE5(rdev) && (tmp & DP_TPS3_SUPPORTED))
+			dp_info.tp3_supported = true;
+		else
+			dp_info.tp3_supported = false;
+	} else {
 		dp_info.tp3_supported = false;
+	}
 
 	memcpy(dp_info.dpcd, dig_connector->dpcd, DP_RECEIVER_CAP_SIZE);
 	dp_info.rdev = rdev;
diff --git a/drivers/gpu/drm/radeon/cik.c b/drivers/gpu/drm/radeon/cik.c
index a5181404f130..69a00d64716e 100644
--- a/drivers/gpu/drm/radeon/cik.c
+++ b/drivers/gpu/drm/radeon/cik.c
@@ -63,6 +63,12 @@ MODULE_FIRMWARE("radeon/KABINI_ce.bin");
 MODULE_FIRMWARE("radeon/KABINI_mec.bin");
 MODULE_FIRMWARE("radeon/KABINI_rlc.bin");
 MODULE_FIRMWARE("radeon/KABINI_sdma.bin");
+MODULE_FIRMWARE("radeon/MULLINS_pfp.bin");
+MODULE_FIRMWARE("radeon/MULLINS_me.bin");
+MODULE_FIRMWARE("radeon/MULLINS_ce.bin");
+MODULE_FIRMWARE("radeon/MULLINS_mec.bin");
+MODULE_FIRMWARE("radeon/MULLINS_rlc.bin");
+MODULE_FIRMWARE("radeon/MULLINS_sdma.bin");
 
 extern int r600_ih_ring_alloc(struct radeon_device *rdev);
 extern void r600_ih_ring_fini(struct radeon_device *rdev);
@@ -1473,6 +1479,43 @@ static const u32 hawaii_mgcg_cgcg_init[] =
 	0xd80c, 0xff000ff0, 0x00000100
 };
 
+static const u32 godavari_golden_registers[] =
+{
+	0x55e4, 0xff607fff, 0xfc000100,
+	0x6ed8, 0x00010101, 0x00010000,
+	0x9830, 0xffffffff, 0x00000000,
+	0x98302, 0xf00fffff, 0x00000400,
+	0x6130, 0xffffffff, 0x00010000,
+	0x5bb0, 0x000000f0, 0x00000070,
+	0x5bc0, 0xf0311fff, 0x80300000,
+	0x98f8, 0x73773777, 0x12010001,
+	0x98fc, 0xffffffff, 0x00000010,
+	0x8030, 0x00001f0f, 0x0000100a,
+	0x2f48, 0x73773777, 0x12010001,
+	0x2408, 0x000fffff, 0x000c007f,
+	0x8a14, 0xf000003f, 0x00000007,
+	0x8b24, 0xffffffff, 0x00ff0fff,
+	0x30a04, 0x0000ff0f, 0x00000000,
+	0x28a4c, 0x07ffffff, 0x06000000,
+	0x4d8, 0x00000fff, 0x00000100,
+	0xd014, 0x00010000, 0x00810001,
+	0xd814, 0x00010000, 0x00810001,
+	0x3e78, 0x00000001, 0x00000002,
+	0xc768, 0x00000008, 0x00000008,
+	0xc770, 0x00000f00, 0x00000800,
+	0xc774, 0x00000f00, 0x00000800,
+	0xc798, 0x00ffffff, 0x00ff7fbf,
+	0xc79c, 0x00ffffff, 0x00ff7faf,
+	0x8c00, 0x000000ff, 0x00000001,
+	0x214f8, 0x01ff01ff, 0x00000002,
+	0x21498, 0x007ff800, 0x00200000,
+	0x2015c, 0xffffffff, 0x00000f40,
+	0x88c4, 0x001f3ae3, 0x00000082,
+	0x88d4, 0x0000001f, 0x00000010,
+	0x30934, 0xffffffff, 0x00000000
+};
+
+
 static void cik_init_golden_registers(struct radeon_device *rdev)
 {
 	switch (rdev->family) {
@@ -1504,6 +1547,20 @@ static void cik_init_golden_registers(struct radeon_device *rdev)
 						 kalindi_golden_spm_registers,
 						 (const u32)ARRAY_SIZE(kalindi_golden_spm_registers));
 		break;
+	case CHIP_MULLINS:
+		radeon_program_register_sequence(rdev,
+						 kalindi_mgcg_cgcg_init,
+						 (const u32)ARRAY_SIZE(kalindi_mgcg_cgcg_init));
+		radeon_program_register_sequence(rdev,
+						 godavari_golden_registers,
+						 (const u32)ARRAY_SIZE(godavari_golden_registers));
+		radeon_program_register_sequence(rdev,
+						 kalindi_golden_common_registers,
+						 (const u32)ARRAY_SIZE(kalindi_golden_common_registers));
+		radeon_program_register_sequence(rdev,
+						 kalindi_golden_spm_registers,
+						 (const u32)ARRAY_SIZE(kalindi_golden_spm_registers));
+		break;
 	case CHIP_KAVERI:
 		radeon_program_register_sequence(rdev,
 						 spectre_mgcg_cgcg_init,
@@ -1834,6 +1891,15 @@ static int cik_init_microcode(struct radeon_device *rdev)
 		rlc_req_size = KB_RLC_UCODE_SIZE * 4;
 		sdma_req_size = CIK_SDMA_UCODE_SIZE * 4;
 		break;
+	case CHIP_MULLINS:
+		chip_name = "MULLINS";
+		pfp_req_size = CIK_PFP_UCODE_SIZE * 4;
+		me_req_size = CIK_ME_UCODE_SIZE * 4;
+		ce_req_size = CIK_CE_UCODE_SIZE * 4;
+		mec_req_size = CIK_MEC_UCODE_SIZE * 4;
+		rlc_req_size = ML_RLC_UCODE_SIZE * 4;
+		sdma_req_size = CIK_SDMA_UCODE_SIZE * 4;
+		break;
 	default: BUG();
 	}
 
@@ -3272,6 +3338,7 @@ static void cik_gpu_init(struct radeon_device *rdev)
 		gb_addr_config = BONAIRE_GB_ADDR_CONFIG_GOLDEN;
 		break;
 	case CHIP_KABINI:
+	case CHIP_MULLINS:
 	default:
 		rdev->config.cik.max_shader_engines = 1;
 		rdev->config.cik.max_tile_pipes = 2;
@@ -3702,6 +3769,7 @@ int cik_copy_cpdma(struct radeon_device *rdev,
 	r = radeon_fence_emit(rdev, fence, ring->idx);
 	if (r) {
 		radeon_ring_unlock_undo(rdev, ring);
+		radeon_semaphore_free(rdev, &sem, NULL);
 		return r;
 	}
 
@@ -5803,6 +5871,9 @@ static int cik_rlc_resume(struct radeon_device *rdev)
 	case CHIP_KABINI:
 		size = KB_RLC_UCODE_SIZE;
 		break;
+	case CHIP_MULLINS:
+		size = ML_RLC_UCODE_SIZE;
+		break;
 	}
 
 	cik_rlc_stop(rdev);
@@ -6551,6 +6622,7 @@ void cik_get_csb_buffer(struct radeon_device *rdev, volatile u32 *buffer)
 		buffer[count++] = cpu_to_le32(0x00000000);
 		break;
 	case CHIP_KABINI:
+	case CHIP_MULLINS:
 		buffer[count++] = cpu_to_le32(0x00000000); /* XXX */
 		buffer[count++] = cpu_to_le32(0x00000000);
 		break;
@@ -6696,6 +6768,19 @@ static void cik_disable_interrupt_state(struct radeon_device *rdev)
 		WREG32(LB_INTERRUPT_MASK + EVERGREEN_CRTC4_REGISTER_OFFSET, 0);
 		WREG32(LB_INTERRUPT_MASK + EVERGREEN_CRTC5_REGISTER_OFFSET, 0);
 	}
+	/* pflip */
+	if (rdev->num_crtc >= 2) {
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC0_REGISTER_OFFSET, 0);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC1_REGISTER_OFFSET, 0);
+	}
+	if (rdev->num_crtc >= 4) {
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC2_REGISTER_OFFSET, 0);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC3_REGISTER_OFFSET, 0);
+	}
+	if (rdev->num_crtc >= 6) {
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC4_REGISTER_OFFSET, 0);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC5_REGISTER_OFFSET, 0);
+	}
 
 	/* dac hotplug */
 	WREG32(DAC_AUTODETECT_INT_CONTROL, 0);
@@ -7052,6 +7137,25 @@ int cik_irq_set(struct radeon_device *rdev)
 		WREG32(LB_INTERRUPT_MASK + EVERGREEN_CRTC5_REGISTER_OFFSET, crtc6);
 	}
 
+	if (rdev->num_crtc >= 2) {
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC0_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC1_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
+	}
+	if (rdev->num_crtc >= 4) {
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC2_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC3_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
+	}
+	if (rdev->num_crtc >= 6) {
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC4_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC5_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
+	}
+
 	WREG32(DC_HPD1_INT_CONTROL, hpd1);
 	WREG32(DC_HPD2_INT_CONTROL, hpd2);
 	WREG32(DC_HPD3_INT_CONTROL, hpd3);
@@ -7088,6 +7192,29 @@ static inline void cik_irq_ack(struct radeon_device *rdev)
 	rdev->irq.stat_regs.cik.disp_int_cont5 = RREG32(DISP_INTERRUPT_STATUS_CONTINUE5);
 	rdev->irq.stat_regs.cik.disp_int_cont6 = RREG32(DISP_INTERRUPT_STATUS_CONTINUE6);
 
+	rdev->irq.stat_regs.cik.d1grph_int = RREG32(GRPH_INT_STATUS +
+		EVERGREEN_CRTC0_REGISTER_OFFSET);
+	rdev->irq.stat_regs.cik.d2grph_int = RREG32(GRPH_INT_STATUS +
+		EVERGREEN_CRTC1_REGISTER_OFFSET);
+	if (rdev->num_crtc >= 4) {
+		rdev->irq.stat_regs.cik.d3grph_int = RREG32(GRPH_INT_STATUS +
+			EVERGREEN_CRTC2_REGISTER_OFFSET);
+		rdev->irq.stat_regs.cik.d4grph_int = RREG32(GRPH_INT_STATUS +
+			EVERGREEN_CRTC3_REGISTER_OFFSET);
+	}
+	if (rdev->num_crtc >= 6) {
+		rdev->irq.stat_regs.cik.d5grph_int = RREG32(GRPH_INT_STATUS +
+			EVERGREEN_CRTC4_REGISTER_OFFSET);
+		rdev->irq.stat_regs.cik.d6grph_int = RREG32(GRPH_INT_STATUS +
+			EVERGREEN_CRTC5_REGISTER_OFFSET);
+	}
+
+	if (rdev->irq.stat_regs.cik.d1grph_int & GRPH_PFLIP_INT_OCCURRED)
+		WREG32(GRPH_INT_STATUS + EVERGREEN_CRTC0_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_CLEAR);
+	if (rdev->irq.stat_regs.cik.d2grph_int & GRPH_PFLIP_INT_OCCURRED)
+		WREG32(GRPH_INT_STATUS + EVERGREEN_CRTC1_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_CLEAR);
 	if (rdev->irq.stat_regs.cik.disp_int & LB_D1_VBLANK_INTERRUPT)
 		WREG32(LB_VBLANK_STATUS + EVERGREEN_CRTC0_REGISTER_OFFSET, VBLANK_ACK);
 	if (rdev->irq.stat_regs.cik.disp_int & LB_D1_VLINE_INTERRUPT)
@@ -7098,6 +7225,12 @@ static inline void cik_irq_ack(struct radeon_device *rdev)
 		WREG32(LB_VLINE_STATUS + EVERGREEN_CRTC1_REGISTER_OFFSET, VLINE_ACK);
 
 	if (rdev->num_crtc >= 4) {
+		if (rdev->irq.stat_regs.cik.d3grph_int & GRPH_PFLIP_INT_OCCURRED)
+			WREG32(GRPH_INT_STATUS + EVERGREEN_CRTC2_REGISTER_OFFSET,
+			       GRPH_PFLIP_INT_CLEAR);
+		if (rdev->irq.stat_regs.cik.d4grph_int & GRPH_PFLIP_INT_OCCURRED)
+			WREG32(GRPH_INT_STATUS + EVERGREEN_CRTC3_REGISTER_OFFSET,
+			       GRPH_PFLIP_INT_CLEAR);
 		if (rdev->irq.stat_regs.cik.disp_int_cont2 & LB_D3_VBLANK_INTERRUPT)
 			WREG32(LB_VBLANK_STATUS + EVERGREEN_CRTC2_REGISTER_OFFSET, VBLANK_ACK);
 		if (rdev->irq.stat_regs.cik.disp_int_cont2 & LB_D3_VLINE_INTERRUPT)
@@ -7109,6 +7242,12 @@ static inline void cik_irq_ack(struct radeon_device *rdev)
 	}
 
 	if (rdev->num_crtc >= 6) {
+		if (rdev->irq.stat_regs.cik.d5grph_int & GRPH_PFLIP_INT_OCCURRED)
+			WREG32(GRPH_INT_STATUS + EVERGREEN_CRTC4_REGISTER_OFFSET,
+			       GRPH_PFLIP_INT_CLEAR);
+		if (rdev->irq.stat_regs.cik.d6grph_int & GRPH_PFLIP_INT_OCCURRED)
+			WREG32(GRPH_INT_STATUS + EVERGREEN_CRTC5_REGISTER_OFFSET,
+			       GRPH_PFLIP_INT_CLEAR);
 		if (rdev->irq.stat_regs.cik.disp_int_cont4 & LB_D5_VBLANK_INTERRUPT)
 			WREG32(LB_VBLANK_STATUS + EVERGREEN_CRTC4_REGISTER_OFFSET, VBLANK_ACK);
 		if (rdev->irq.stat_regs.cik.disp_int_cont4 & LB_D5_VLINE_INTERRUPT)
@@ -7460,6 +7599,15 @@ restart_ih:
 				break;
 			}
 			break;
+		case 8: /* D1 page flip */
+		case 10: /* D2 page flip */
+		case 12: /* D3 page flip */
+		case 14: /* D4 page flip */
+		case 16: /* D5 page flip */
+		case 18: /* D6 page flip */
+			DRM_DEBUG("IH: D%d flip\n", ((src_id - 8) >> 1) + 1);
+			radeon_crtc_handle_flip(rdev, (src_id - 8) >> 1);
+			break;
 		case 42: /* HPD hotplug */
 			switch (src_data) {
 			case 0:
diff --git a/drivers/gpu/drm/radeon/cik_sdma.c b/drivers/gpu/drm/radeon/cik_sdma.c
index 3c2407bad1f0..1347162ca1a4 100644
--- a/drivers/gpu/drm/radeon/cik_sdma.c
+++ b/drivers/gpu/drm/radeon/cik_sdma.c
@@ -562,6 +562,7 @@ int cik_copy_dma(struct radeon_device *rdev,
 	r = radeon_fence_emit(rdev, fence, ring->idx);
 	if (r) {
 		radeon_ring_unlock_undo(rdev, ring);
+		radeon_semaphore_free(rdev, &sem, NULL);
 		return r;
 	}
 
diff --git a/drivers/gpu/drm/radeon/cikd.h b/drivers/gpu/drm/radeon/cikd.h
index 0b27ea08c299..ae88660f34ea 100644
--- a/drivers/gpu/drm/radeon/cikd.h
+++ b/drivers/gpu/drm/radeon/cikd.h
@@ -889,6 +889,15 @@
 #       define DC_HPD6_RX_INTERRUPT                     (1 << 18)
 #define DISP_INTERRUPT_STATUS_CONTINUE6                 0x6780
 
+/* 0x6858, 0x7458, 0x10058, 0x10c58, 0x11858, 0x12458 */
+#define GRPH_INT_STATUS                                 0x6858
+#       define GRPH_PFLIP_INT_OCCURRED                  (1 << 0)
+#       define GRPH_PFLIP_INT_CLEAR                     (1 << 8)
+/* 0x685c, 0x745c, 0x1005c, 0x10c5c, 0x1185c, 0x1245c */
+#define GRPH_INT_CONTROL                                0x685c
+#       define GRPH_PFLIP_INT_MASK                      (1 << 0)
+#       define GRPH_PFLIP_INT_TYPE                      (1 << 8)
+
 #define	DAC_AUTODETECT_INT_CONTROL			0x67c8
 
 #define DC_HPD1_INT_STATUS                              0x601c
diff --git a/drivers/gpu/drm/radeon/evergreen.c b/drivers/gpu/drm/radeon/evergreen.c
index 0318230ef274..653eff814504 100644
--- a/drivers/gpu/drm/radeon/evergreen.c
+++ b/drivers/gpu/drm/radeon/evergreen.c
@@ -4355,7 +4355,6 @@ int evergreen_irq_set(struct radeon_device *rdev)
 	u32 crtc1 = 0, crtc2 = 0, crtc3 = 0, crtc4 = 0, crtc5 = 0, crtc6 = 0;
 	u32 hpd1, hpd2, hpd3, hpd4, hpd5, hpd6;
 	u32 grbm_int_cntl = 0;
-	u32 grph1 = 0, grph2 = 0, grph3 = 0, grph4 = 0, grph5 = 0, grph6 = 0;
 	u32 afmt1 = 0, afmt2 = 0, afmt3 = 0, afmt4 = 0, afmt5 = 0, afmt6 = 0;
 	u32 dma_cntl, dma_cntl1 = 0;
 	u32 thermal_int = 0;
@@ -4538,15 +4537,21 @@ int evergreen_irq_set(struct radeon_device *rdev)
 		WREG32(INT_MASK + EVERGREEN_CRTC5_REGISTER_OFFSET, crtc6);
 	}
 
-	WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC0_REGISTER_OFFSET, grph1);
-	WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC1_REGISTER_OFFSET, grph2);
+	WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC0_REGISTER_OFFSET,
+	       GRPH_PFLIP_INT_MASK);
+	WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC1_REGISTER_OFFSET,
+	       GRPH_PFLIP_INT_MASK);
 	if (rdev->num_crtc >= 4) {
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC2_REGISTER_OFFSET, grph3);
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC3_REGISTER_OFFSET, grph4);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC2_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC3_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
 	}
 	if (rdev->num_crtc >= 6) {
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC4_REGISTER_OFFSET, grph5);
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC5_REGISTER_OFFSET, grph6);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC4_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC5_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
 	}
 
 	WREG32(DC_HPD1_INT_CONTROL, hpd1);
@@ -4935,6 +4940,15 @@ restart_ih:
 				break;
 			}
 			break;
+		case 8: /* D1 page flip */
+		case 10: /* D2 page flip */
+		case 12: /* D3 page flip */
+		case 14: /* D4 page flip */
+		case 16: /* D5 page flip */
+		case 18: /* D6 page flip */
+			DRM_DEBUG("IH: D%d flip\n", ((src_id - 8) >> 1) + 1);
+			radeon_crtc_handle_flip(rdev, (src_id - 8) >> 1);
+			break;
 		case 42: /* HPD hotplug */
 			switch (src_data) {
 			case 0:
diff --git a/drivers/gpu/drm/radeon/evergreen_dma.c b/drivers/gpu/drm/radeon/evergreen_dma.c
index 287fe966d7de..478caefe0fef 100644
--- a/drivers/gpu/drm/radeon/evergreen_dma.c
+++ b/drivers/gpu/drm/radeon/evergreen_dma.c
@@ -151,6 +151,7 @@ int evergreen_copy_dma(struct radeon_device *rdev,
 	r = radeon_fence_emit(rdev, fence, ring->idx);
 	if (r) {
 		radeon_ring_unlock_undo(rdev, ring);
+		radeon_semaphore_free(rdev, &sem, NULL);
 		return r;
 	}
 
diff --git a/drivers/gpu/drm/radeon/kv_dpm.c b/drivers/gpu/drm/radeon/kv_dpm.c
index 16ec9d56a234..3f6e817d97ee 100644
--- a/drivers/gpu/drm/radeon/kv_dpm.c
+++ b/drivers/gpu/drm/radeon/kv_dpm.c
@@ -546,6 +546,52 @@ static int kv_set_divider_value(struct radeon_device *rdev,
 	return 0;
 }
 
+static u32 kv_convert_vid2_to_vid7(struct radeon_device *rdev,
+				   struct sumo_vid_mapping_table *vid_mapping_table,
+				   u32 vid_2bit)
+{
+	struct radeon_clock_voltage_dependency_table *vddc_sclk_table =
+		&rdev->pm.dpm.dyn_state.vddc_dependency_on_sclk;
+	u32 i;
+
+	if (vddc_sclk_table && vddc_sclk_table->count) {
+		if (vid_2bit < vddc_sclk_table->count)
+			return vddc_sclk_table->entries[vid_2bit].v;
+		else
+			return vddc_sclk_table->entries[vddc_sclk_table->count - 1].v;
+	} else {
+		for (i = 0; i < vid_mapping_table->num_entries; i++) {
+			if (vid_mapping_table->entries[i].vid_2bit == vid_2bit)
+				return vid_mapping_table->entries[i].vid_7bit;
+		}
+		return vid_mapping_table->entries[vid_mapping_table->num_entries - 1].vid_7bit;
+	}
+}
+
+static u32 kv_convert_vid7_to_vid2(struct radeon_device *rdev,
+				   struct sumo_vid_mapping_table *vid_mapping_table,
+				   u32 vid_7bit)
+{
+	struct radeon_clock_voltage_dependency_table *vddc_sclk_table =
+		&rdev->pm.dpm.dyn_state.vddc_dependency_on_sclk;
+	u32 i;
+
+	if (vddc_sclk_table && vddc_sclk_table->count) {
+		for (i = 0; i < vddc_sclk_table->count; i++) {
+			if (vddc_sclk_table->entries[i].v == vid_7bit)
+				return i;
+		}
+		return vddc_sclk_table->count - 1;
+	} else {
+		for (i = 0; i < vid_mapping_table->num_entries; i++) {
+			if (vid_mapping_table->entries[i].vid_7bit == vid_7bit)
+				return vid_mapping_table->entries[i].vid_2bit;
+		}
+
+		return vid_mapping_table->entries[vid_mapping_table->num_entries - 1].vid_2bit;
+	}
+}
+
 static u16 kv_convert_8bit_index_to_voltage(struct radeon_device *rdev,
 					    u16 voltage)
 {
@@ -556,9 +602,9 @@ static u16 kv_convert_2bit_index_to_voltage(struct radeon_device *rdev,
 					    u32 vid_2bit)
 {
 	struct kv_power_info *pi = kv_get_pi(rdev);
-	u32 vid_8bit = sumo_convert_vid2_to_vid7(rdev,
-						 &pi->sys_info.vid_mapping_table,
-						 vid_2bit);
+	u32 vid_8bit = kv_convert_vid2_to_vid7(rdev,
+					       &pi->sys_info.vid_mapping_table,
+					       vid_2bit);
 
 	return kv_convert_8bit_index_to_voltage(rdev, (u16)vid_8bit);
 }
@@ -639,7 +685,7 @@ static int kv_force_lowest_valid(struct radeon_device *rdev)
 
 static int kv_unforce_levels(struct radeon_device *rdev)
 {
-	if (rdev->family == CHIP_KABINI)
+	if (rdev->family == CHIP_KABINI || rdev->family == CHIP_MULLINS)
 		return kv_notify_message_to_smu(rdev, PPSMC_MSG_NoForcedLevel);
 	else
 		return kv_set_enabled_levels(rdev);
@@ -1362,13 +1408,20 @@ static int kv_update_uvd_dpm(struct radeon_device *rdev, bool gate)
 	struct radeon_uvd_clock_voltage_dependency_table *table =
 		&rdev->pm.dpm.dyn_state.uvd_clock_voltage_dependency_table;
 	int ret;
+	u32 mask;
 
 	if (!gate) {
-		if (!pi->caps_uvd_dpm || table->count || pi->caps_stable_p_state)
+		if (table->count)
 			pi->uvd_boot_level = table->count - 1;
 		else
 			pi->uvd_boot_level = 0;
 
+		if (!pi->caps_uvd_dpm || pi->caps_stable_p_state) {
+			mask = 1 << pi->uvd_boot_level;
+		} else {
+			mask = 0x1f;
+		}
+
 		ret = kv_copy_bytes_to_smc(rdev,
 					   pi->dpm_table_start +
 					   offsetof(SMU7_Fusion_DpmTable, UvdBootLevel),
@@ -1377,11 +1430,9 @@ static int kv_update_uvd_dpm(struct radeon_device *rdev, bool gate)
 		if (ret)
 			return ret;
 
-		if (!pi->caps_uvd_dpm ||
-		    pi->caps_stable_p_state)
-			kv_send_msg_to_smc_with_parameter(rdev,
-							  PPSMC_MSG_UVDDPM_SetEnabledMask,
-							  (1 << pi->uvd_boot_level));
+		kv_send_msg_to_smc_with_parameter(rdev,
+						  PPSMC_MSG_UVDDPM_SetEnabledMask,
+						  mask);
 	}
 
 	return kv_enable_uvd_dpm(rdev, !gate);
@@ -1617,7 +1668,7 @@ static void kv_dpm_powergate_acp(struct radeon_device *rdev, bool gate)
 	if (pi->acp_power_gated == gate)
 		return;
 
-	if (rdev->family == CHIP_KABINI)
+	if (rdev->family == CHIP_KABINI || rdev->family == CHIP_MULLINS)
 		return;
 
 	pi->acp_power_gated = gate;
@@ -1786,7 +1837,7 @@ int kv_dpm_set_power_state(struct radeon_device *rdev)
 		}
 	}
 
-	if (rdev->family == CHIP_KABINI) {
+	if (rdev->family == CHIP_KABINI || rdev->family == CHIP_MULLINS) {
 		if (pi->enable_dpm) {
 			kv_set_valid_clock_range(rdev, new_ps);
 			kv_update_dfs_bypass_settings(rdev, new_ps);
@@ -1812,6 +1863,8 @@ int kv_dpm_set_power_state(struct radeon_device *rdev)
 				return ret;
 			}
 			kv_update_sclk_t(rdev);
+			if (rdev->family == CHIP_MULLINS)
+				kv_enable_nb_dpm(rdev);
 		}
 	} else {
 		if (pi->enable_dpm) {
@@ -1862,7 +1915,7 @@ void kv_dpm_reset_asic(struct radeon_device *rdev)
 {
 	struct kv_power_info *pi = kv_get_pi(rdev);
 
-	if (rdev->family == CHIP_KABINI) {
+	if (rdev->family == CHIP_KABINI || rdev->family == CHIP_MULLINS) {
 		kv_force_lowest_valid(rdev);
 		kv_init_graphics_levels(rdev);
 		kv_program_bootup_state(rdev);
@@ -1901,14 +1954,41 @@ static void kv_construct_max_power_limits_table(struct radeon_device *rdev,
 static void kv_patch_voltage_values(struct radeon_device *rdev)
 {
 	int i;
-	struct radeon_uvd_clock_voltage_dependency_table *table =
+	struct radeon_uvd_clock_voltage_dependency_table *uvd_table =
 		&rdev->pm.dpm.dyn_state.uvd_clock_voltage_dependency_table;
+	struct radeon_vce_clock_voltage_dependency_table *vce_table =
+		&rdev->pm.dpm.dyn_state.vce_clock_voltage_dependency_table;
+	struct radeon_clock_voltage_dependency_table *samu_table =
+		&rdev->pm.dpm.dyn_state.samu_clock_voltage_dependency_table;
+	struct radeon_clock_voltage_dependency_table *acp_table =
+		&rdev->pm.dpm.dyn_state.acp_clock_voltage_dependency_table;
 
-	if (table->count) {
-		for (i = 0; i < table->count; i++)
-			table->entries[i].v =
+	if (uvd_table->count) {
+		for (i = 0; i < uvd_table->count; i++)
+			uvd_table->entries[i].v =
 				kv_convert_8bit_index_to_voltage(rdev,
-								 table->entries[i].v);
+								 uvd_table->entries[i].v);
+	}
+
+	if (vce_table->count) {
+		for (i = 0; i < vce_table->count; i++)
+			vce_table->entries[i].v =
+				kv_convert_8bit_index_to_voltage(rdev,
+								 vce_table->entries[i].v);
+	}
+
+	if (samu_table->count) {
+		for (i = 0; i < samu_table->count; i++)
+			samu_table->entries[i].v =
+				kv_convert_8bit_index_to_voltage(rdev,
+								 samu_table->entries[i].v);
+	}
+
+	if (acp_table->count) {
+		for (i = 0; i < acp_table->count; i++)
+			acp_table->entries[i].v =
+				kv_convert_8bit_index_to_voltage(rdev,
+								 acp_table->entries[i].v);
 	}
 
 }
@@ -1941,7 +2021,7 @@ static int kv_force_dpm_highest(struct radeon_device *rdev)
 			break;
 	}
 
-	if (rdev->family == CHIP_KABINI)
+	if (rdev->family == CHIP_KABINI || rdev->family == CHIP_MULLINS)
 		return kv_send_msg_to_smc_with_parameter(rdev, PPSMC_MSG_DPM_ForceState, i);
 	else
 		return kv_set_enabled_level(rdev, i);
@@ -1961,7 +2041,7 @@ static int kv_force_dpm_lowest(struct radeon_device *rdev)
 			break;
 	}
 
-	if (rdev->family == CHIP_KABINI)
+	if (rdev->family == CHIP_KABINI || rdev->family == CHIP_MULLINS)
 		return kv_send_msg_to_smc_with_parameter(rdev, PPSMC_MSG_DPM_ForceState, i);
 	else
 		return kv_set_enabled_level(rdev, i);
@@ -2118,7 +2198,7 @@ static void kv_apply_state_adjust_rules(struct radeon_device *rdev,
 	else
 		pi->battery_state = false;
 
-	if (rdev->family == CHIP_KABINI) {
+	if (rdev->family == CHIP_KABINI || rdev->family == CHIP_MULLINS) {
 		ps->dpm0_pg_nb_ps_lo = 0x1;
 		ps->dpm0_pg_nb_ps_hi = 0x0;
 		ps->dpmx_nb_ps_lo = 0x1;
@@ -2179,7 +2259,7 @@ static int kv_calculate_nbps_level_settings(struct radeon_device *rdev)
 	if (pi->lowest_valid > pi->highest_valid)
 		return -EINVAL;
 
-	if (rdev->family == CHIP_KABINI) {
+	if (rdev->family == CHIP_KABINI || rdev->family == CHIP_MULLINS) {
 		for (i = pi->lowest_valid; i <= pi->highest_valid; i++) {
 			pi->graphics_level[i].GnbSlow = 1;
 			pi->graphics_level[i].ForceNbPs1 = 0;
@@ -2253,9 +2333,9 @@ static void kv_init_graphics_levels(struct radeon_device *rdev)
 				break;
 
 			kv_set_divider_value(rdev, i, table->entries[i].clk);
-			vid_2bit = sumo_convert_vid7_to_vid2(rdev,
-							     &pi->sys_info.vid_mapping_table,
-							     table->entries[i].v);
+			vid_2bit = kv_convert_vid7_to_vid2(rdev,
+							   &pi->sys_info.vid_mapping_table,
+							   table->entries[i].v);
 			kv_set_vid(rdev, i, vid_2bit);
 			kv_set_at(rdev, i, pi->at[i]);
 			kv_dpm_power_level_enabled_for_throttle(rdev, i, true);
@@ -2324,7 +2404,7 @@ static void kv_program_nbps_index_settings(struct radeon_device *rdev,
 	struct kv_power_info *pi = kv_get_pi(rdev);
 	u32 nbdpmconfig1;
 
-	if (rdev->family == CHIP_KABINI)
+	if (rdev->family == CHIP_KABINI || rdev->family == CHIP_MULLINS)
 		return;
 
 	if (pi->sys_info.nb_dpm_enable) {
@@ -2631,9 +2711,6 @@ int kv_dpm_init(struct radeon_device *rdev)
 
         pi->sram_end = SMC_RAM_END;
 
-	if (rdev->family == CHIP_KABINI)
-		pi->high_voltage_t = 4001;
-
 	pi->enable_nb_dpm = true;
 
 	pi->caps_power_containment = true;
diff --git a/drivers/gpu/drm/radeon/r600.c b/drivers/gpu/drm/radeon/r600.c
index 436e55092e9d..c75881223d18 100644
--- a/drivers/gpu/drm/radeon/r600.c
+++ b/drivers/gpu/drm/radeon/r600.c
@@ -2839,6 +2839,7 @@ int r600_copy_cpdma(struct radeon_device *rdev,
 	r = radeon_fence_emit(rdev, fence, ring->idx);
 	if (r) {
 		radeon_ring_unlock_undo(rdev, ring);
+		radeon_semaphore_free(rdev, &sem, NULL);
 		return r;
 	}
 
@@ -3505,7 +3506,6 @@ int r600_irq_set(struct radeon_device *rdev)
 	u32 hpd1, hpd2, hpd3, hpd4 = 0, hpd5 = 0, hpd6 = 0;
 	u32 grbm_int_cntl = 0;
 	u32 hdmi0, hdmi1;
-	u32 d1grph = 0, d2grph = 0;
 	u32 dma_cntl;
 	u32 thermal_int = 0;
 
@@ -3614,8 +3614,8 @@ int r600_irq_set(struct radeon_device *rdev)
 	WREG32(CP_INT_CNTL, cp_int_cntl);
 	WREG32(DMA_CNTL, dma_cntl);
 	WREG32(DxMODE_INT_MASK, mode_int);
-	WREG32(D1GRPH_INTERRUPT_CONTROL, d1grph);
-	WREG32(D2GRPH_INTERRUPT_CONTROL, d2grph);
+	WREG32(D1GRPH_INTERRUPT_CONTROL, DxGRPH_PFLIP_INT_MASK);
+	WREG32(D2GRPH_INTERRUPT_CONTROL, DxGRPH_PFLIP_INT_MASK);
 	WREG32(GRBM_INT_CNTL, grbm_int_cntl);
 	if (ASIC_IS_DCE3(rdev)) {
 		WREG32(DC_HPD1_INT_CONTROL, hpd1);
@@ -3918,6 +3918,14 @@ restart_ih:
 				break;
 			}
 			break;
+		case 9: /* D1 pflip */
+			DRM_DEBUG("IH: D1 flip\n");
+			radeon_crtc_handle_flip(rdev, 0);
+			break;
+		case 11: /* D2 pflip */
+			DRM_DEBUG("IH: D2 flip\n");
+			radeon_crtc_handle_flip(rdev, 1);
+			break;
 		case 19: /* HPD/DAC hotplug */
 			switch (src_data) {
 			case 0:
diff --git a/drivers/gpu/drm/radeon/r600_dma.c b/drivers/gpu/drm/radeon/r600_dma.c
index 53fcb28f5578..4969cef44a19 100644
--- a/drivers/gpu/drm/radeon/r600_dma.c
+++ b/drivers/gpu/drm/radeon/r600_dma.c
@@ -489,6 +489,7 @@ int r600_copy_dma(struct radeon_device *rdev,
 	r = radeon_fence_emit(rdev, fence, ring->idx);
 	if (r) {
 		radeon_ring_unlock_undo(rdev, ring);
+		radeon_semaphore_free(rdev, &sem, NULL);
 		return r;
 	}
 
diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h
index dd4da88b3ab1..7501ba318c67 100644
--- a/drivers/gpu/drm/radeon/radeon.h
+++ b/drivers/gpu/drm/radeon/radeon.h
@@ -732,6 +732,12 @@ struct cik_irq_stat_regs {
 	u32 disp_int_cont4;
 	u32 disp_int_cont5;
 	u32 disp_int_cont6;
+	u32 d1grph_int;
+	u32 d2grph_int;
+	u32 d3grph_int;
+	u32 d4grph_int;
+	u32 d5grph_int;
+	u32 d6grph_int;
 };
 
 union radeon_irq_stat_regs {
@@ -1647,6 +1653,7 @@ struct radeon_vce {
 	unsigned		fb_version;
 	atomic_t		handles[RADEON_MAX_VCE_HANDLES];
 	struct drm_file		*filp[RADEON_MAX_VCE_HANDLES];
+	unsigned		img_size[RADEON_MAX_VCE_HANDLES];
 	struct delayed_work	idle_work;
 };
 
@@ -1660,7 +1667,7 @@ int radeon_vce_get_destroy_msg(struct radeon_device *rdev, int ring,
 			       uint32_t handle, struct radeon_fence **fence);
 void radeon_vce_free_handles(struct radeon_device *rdev, struct drm_file *filp);
 void radeon_vce_note_usage(struct radeon_device *rdev);
-int radeon_vce_cs_reloc(struct radeon_cs_parser *p, int lo, int hi);
+int radeon_vce_cs_reloc(struct radeon_cs_parser *p, int lo, int hi, unsigned size);
 int radeon_vce_cs_parse(struct radeon_cs_parser *p);
 bool radeon_vce_semaphore_emit(struct radeon_device *rdev,
 			       struct radeon_ring *ring,
@@ -2644,7 +2651,8 @@ void r100_pll_errata_after_index(struct radeon_device *rdev);
 #define ASIC_IS_DCE8(rdev) ((rdev->family >= CHIP_BONAIRE))
 #define ASIC_IS_DCE81(rdev) ((rdev->family == CHIP_KAVERI))
 #define ASIC_IS_DCE82(rdev) ((rdev->family == CHIP_BONAIRE))
-#define ASIC_IS_DCE83(rdev) ((rdev->family == CHIP_KABINI))
+#define ASIC_IS_DCE83(rdev) ((rdev->family == CHIP_KABINI) || \
+			     (rdev->family == CHIP_MULLINS))
 
 #define ASIC_IS_LOMBOK(rdev) ((rdev->ddev->pdev->device == 0x6849) || \
 			      (rdev->ddev->pdev->device == 0x6850) || \
diff --git a/drivers/gpu/drm/radeon/radeon_asic.c b/drivers/gpu/drm/radeon/radeon_asic.c
index d8e1587d89cf..34ea53d980a1 100644
--- a/drivers/gpu/drm/radeon/radeon_asic.c
+++ b/drivers/gpu/drm/radeon/radeon_asic.c
@@ -2029,8 +2029,8 @@ static struct radeon_asic ci_asic = {
 		.blit_ring_index = RADEON_RING_TYPE_GFX_INDEX,
 		.dma = &cik_copy_dma,
 		.dma_ring_index = R600_RING_TYPE_DMA_INDEX,
-		.copy = &cik_copy_dma,
-		.copy_ring_index = R600_RING_TYPE_DMA_INDEX,
+		.copy = &cik_copy_cpdma,
+		.copy_ring_index = RADEON_RING_TYPE_GFX_INDEX,
 	},
 	.surface = {
 		.set_reg = r600_set_surface_reg,
@@ -2494,6 +2494,7 @@ int radeon_asic_init(struct radeon_device *rdev)
 		break;
 	case CHIP_KAVERI:
 	case CHIP_KABINI:
+	case CHIP_MULLINS:
 		rdev->asic = &kv_asic;
 		/* set num crtcs */
 		if (rdev->family == CHIP_KAVERI) {
diff --git a/drivers/gpu/drm/radeon/radeon_bios.c b/drivers/gpu/drm/radeon/radeon_bios.c
index b3633d9a5317..9ab30976287d 100644
--- a/drivers/gpu/drm/radeon/radeon_bios.c
+++ b/drivers/gpu/drm/radeon/radeon_bios.c
@@ -196,6 +196,20 @@ static bool radeon_atrm_get_bios(struct radeon_device *rdev)
 		}
 	}
 
+	if (!found) {
+		while ((pdev = pci_get_class(PCI_CLASS_DISPLAY_OTHER << 8, pdev)) != NULL) {
+			dhandle = ACPI_HANDLE(&pdev->dev);
+			if (!dhandle)
+				continue;
+
+			status = acpi_get_handle(dhandle, "ATRM", &atrm_handle);
+			if (!ACPI_FAILURE(status)) {
+				found = true;
+				break;
+			}
+		}
+	}
+
 	if (!found)
 		return false;
 
diff --git a/drivers/gpu/drm/radeon/radeon_cs.c b/drivers/gpu/drm/radeon/radeon_cs.c
index 2b6e0ebcc13a..41ecf8a60611 100644
--- a/drivers/gpu/drm/radeon/radeon_cs.c
+++ b/drivers/gpu/drm/radeon/radeon_cs.c
@@ -152,6 +152,12 @@ static int radeon_cs_parser_relocs(struct radeon_cs_parser *p)
 			uint32_t domain = r->write_domain ?
 				r->write_domain : r->read_domains;
 
+			if (domain & RADEON_GEM_DOMAIN_CPU) {
+				DRM_ERROR("RADEON_GEM_DOMAIN_CPU is not valid "
+					  "for command submission\n");
+				return -EINVAL;
+			}
+
 			p->relocs[i].domain = domain;
 			if (domain == RADEON_GEM_DOMAIN_VRAM)
 				domain |= RADEON_GEM_DOMAIN_GTT;
@@ -342,10 +348,17 @@ int radeon_cs_parser_init(struct radeon_cs_parser *p, void *data)
 			return -EINVAL;
 
 		/* we only support VM on some SI+ rings */
-		if ((p->rdev->asic->ring[p->ring]->cs_parse == NULL) &&
-		   ((p->cs_flags & RADEON_CS_USE_VM) == 0)) {
-			DRM_ERROR("Ring %d requires VM!\n", p->ring);
-			return -EINVAL;
+		if ((p->cs_flags & RADEON_CS_USE_VM) == 0) {
+			if (p->rdev->asic->ring[p->ring]->cs_parse == NULL) {
+				DRM_ERROR("Ring %d requires VM!\n", p->ring);
+				return -EINVAL;
+			}
+		} else {
+			if (p->rdev->asic->ring[p->ring]->ib_parse == NULL) {
+				DRM_ERROR("VM not supported on ring %d!\n",
+					  p->ring);
+				return -EINVAL;
+			}
 		}
 	}
 
diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c
index 9aa1afd1786e..31565de1116c 100644
--- a/drivers/gpu/drm/radeon/radeon_device.c
+++ b/drivers/gpu/drm/radeon/radeon_device.c
@@ -99,6 +99,7 @@ static const char radeon_family_name[][16] = {
 	"KAVERI",
 	"KABINI",
 	"HAWAII",
+	"MULLINS",
 	"LAST",
 };
 
@@ -1533,11 +1534,6 @@ int radeon_resume_kms(struct drm_device *dev, bool resume, bool fbcon)
 
 	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);
@@ -1562,6 +1558,16 @@ int radeon_resume_kms(struct drm_device *dev, bool resume, bool fbcon)
 	}
 
 	drm_kms_helper_poll_enable(dev);
+
+	/* set the power state here in case we are a PX system or headless */
+	if ((rdev->pm.pm_method == PM_METHOD_DPM) && rdev->pm.dpm_enabled)
+		radeon_pm_compute_clocks(rdev);
+
+	if (fbcon) {
+		radeon_fbdev_set_suspend(rdev, 0);
+		console_unlock();
+	}
+
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/radeon/radeon_display.c b/drivers/gpu/drm/radeon/radeon_display.c
index e330e762c360..a4e725c6b8c8 100644
--- a/drivers/gpu/drm/radeon/radeon_display.c
+++ b/drivers/gpu/drm/radeon/radeon_display.c
@@ -289,6 +289,10 @@ void radeon_crtc_handle_vblank(struct radeon_device *rdev, int crtc_id)
 	u32 update_pending;
 	int vpos, hpos;
 
+	/* can happen during initialization */
+	if (radeon_crtc == NULL)
+		return;
+
 	spin_lock_irqsave(&rdev->ddev->event_lock, flags);
 	work = radeon_crtc->flip_work;
 	if (work == NULL) {
@@ -872,14 +876,14 @@ static void avivo_reduce_ratio(unsigned *nom, unsigned *den,
 
 	/* make sure nominator is large enough */
         if (*nom < nom_min) {
-		tmp = (nom_min + *nom - 1) / *nom;
+		tmp = DIV_ROUND_UP(nom_min, *nom);
 		*nom *= tmp;
 		*den *= tmp;
 	}
 
 	/* make sure the denominator is large enough */
 	if (*den < den_min) {
-		tmp = (den_min + *den - 1) / *den;
+		tmp = DIV_ROUND_UP(den_min, *den);
 		*nom *= tmp;
 		*den *= tmp;
 	}
@@ -904,7 +908,7 @@ static void avivo_get_fb_ref_div(unsigned nom, unsigned den, unsigned post_div,
 				 unsigned *fb_div, unsigned *ref_div)
 {
 	/* limit reference * post divider to a maximum */
-	ref_div_max = min(210 / post_div, ref_div_max);
+	ref_div_max = max(min(100 / post_div, ref_div_max), 1u);
 
 	/* get matching reference and feedback divider */
 	*ref_div = min(max(DIV_ROUND_CLOSEST(den, post_div), 1u), ref_div_max);
@@ -1039,6 +1043,16 @@ void radeon_compute_pll_avivo(struct radeon_pll *pll,
 	/* this also makes sure that the reference divider is large enough */
 	avivo_reduce_ratio(&fb_div, &ref_div, fb_div_min, ref_div_min);
 
+	/* avoid high jitter with small fractional dividers */
+	if (pll->flags & RADEON_PLL_USE_FRAC_FB_DIV && (fb_div % 10)) {
+		fb_div_min = max(fb_div_min, (9 - (fb_div % 10)) * 20 + 50);
+		if (fb_div < fb_div_min) {
+			unsigned tmp = DIV_ROUND_UP(fb_div_min, fb_div);
+			fb_div *= tmp;
+			ref_div *= tmp;
+		}
+	}
+
 	/* and finally save the result */
 	if (pll->flags & RADEON_PLL_USE_FRAC_FB_DIV) {
 		*fb_div_p = fb_div / 10;
diff --git a/drivers/gpu/drm/radeon/radeon_family.h b/drivers/gpu/drm/radeon/radeon_family.h
index 9da5da4ffd17..4b7b87f71a63 100644
--- a/drivers/gpu/drm/radeon/radeon_family.h
+++ b/drivers/gpu/drm/radeon/radeon_family.h
@@ -97,6 +97,7 @@ enum radeon_family {
 	CHIP_KAVERI,
 	CHIP_KABINI,
 	CHIP_HAWAII,
+	CHIP_MULLINS,
 	CHIP_LAST,
 };
 
diff --git a/drivers/gpu/drm/radeon/radeon_kms.c b/drivers/gpu/drm/radeon/radeon_kms.c
index 0cc47f12d995..eaaedba04675 100644
--- a/drivers/gpu/drm/radeon/radeon_kms.c
+++ b/drivers/gpu/drm/radeon/radeon_kms.c
@@ -577,28 +577,29 @@ int radeon_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv)
 			return r;
 		}
 
-		r = radeon_bo_reserve(rdev->ring_tmp_bo.bo, false);
-		if (r) {
-			radeon_vm_fini(rdev, &fpriv->vm);
-			kfree(fpriv);
-			return r;
-		}
+		if (rdev->accel_working) {
+			r = radeon_bo_reserve(rdev->ring_tmp_bo.bo, false);
+			if (r) {
+				radeon_vm_fini(rdev, &fpriv->vm);
+				kfree(fpriv);
+				return r;
+			}
 
-		/* map the ib pool buffer read only into
-		 * virtual address space */
-		bo_va = radeon_vm_bo_add(rdev, &fpriv->vm,
-					 rdev->ring_tmp_bo.bo);
-		r = radeon_vm_bo_set_addr(rdev, bo_va, RADEON_VA_IB_OFFSET,
-					  RADEON_VM_PAGE_READABLE |
-					  RADEON_VM_PAGE_SNOOPED);
+			/* map the ib pool buffer read only into
+			 * virtual address space */
+			bo_va = radeon_vm_bo_add(rdev, &fpriv->vm,
+						 rdev->ring_tmp_bo.bo);
+			r = radeon_vm_bo_set_addr(rdev, bo_va, RADEON_VA_IB_OFFSET,
+						  RADEON_VM_PAGE_READABLE |
+						  RADEON_VM_PAGE_SNOOPED);
 
-		radeon_bo_unreserve(rdev->ring_tmp_bo.bo);
-		if (r) {
-			radeon_vm_fini(rdev, &fpriv->vm);
-			kfree(fpriv);
-			return r;
+			radeon_bo_unreserve(rdev->ring_tmp_bo.bo);
+			if (r) {
+				radeon_vm_fini(rdev, &fpriv->vm);
+				kfree(fpriv);
+				return r;
+			}
 		}
-
 		file_priv->driver_priv = fpriv;
 	}
 
@@ -626,13 +627,15 @@ void radeon_driver_postclose_kms(struct drm_device *dev,
 		struct radeon_bo_va *bo_va;
 		int r;
 
-		r = radeon_bo_reserve(rdev->ring_tmp_bo.bo, false);
-		if (!r) {
-			bo_va = radeon_vm_bo_find(&fpriv->vm,
-						  rdev->ring_tmp_bo.bo);
-			if (bo_va)
-				radeon_vm_bo_rmv(rdev, bo_va);
-			radeon_bo_unreserve(rdev->ring_tmp_bo.bo);
+		if (rdev->accel_working) {
+			r = radeon_bo_reserve(rdev->ring_tmp_bo.bo, false);
+			if (!r) {
+				bo_va = radeon_vm_bo_find(&fpriv->vm,
+							  rdev->ring_tmp_bo.bo);
+				if (bo_va)
+					radeon_vm_bo_rmv(rdev, bo_va);
+				radeon_bo_unreserve(rdev->ring_tmp_bo.bo);
+			}
 		}
 
 		radeon_vm_fini(rdev, &fpriv->vm);
diff --git a/drivers/gpu/drm/radeon/radeon_object.c b/drivers/gpu/drm/radeon/radeon_object.c
index 95197aa4de4a..2918087e572f 100644
--- a/drivers/gpu/drm/radeon/radeon_object.c
+++ b/drivers/gpu/drm/radeon/radeon_object.c
@@ -458,7 +458,7 @@ int radeon_bo_list_validate(struct radeon_device *rdev,
 			 * into account. We don't want to disallow buffer moves
 			 * completely.
 			 */
-			if (current_domain != RADEON_GEM_DOMAIN_CPU &&
+			if ((lobj->alt_domain & current_domain) != 0 &&
 			    (domain & current_domain) == 0 && /* will be moved */
 			    bytes_moved > bytes_moved_threshold) {
 				/* don't move it */
@@ -699,22 +699,30 @@ int radeon_bo_fault_reserve_notify(struct ttm_buffer_object *bo)
 	rbo = container_of(bo, struct radeon_bo, tbo);
 	radeon_bo_check_tiling(rbo, 0, 0);
 	rdev = rbo->rdev;
-	if (bo->mem.mem_type == TTM_PL_VRAM) {
-		size = bo->mem.num_pages << PAGE_SHIFT;
-		offset = bo->mem.start << PAGE_SHIFT;
-		if ((offset + size) > rdev->mc.visible_vram_size) {
-			/* hurrah the memory is not visible ! */
-			radeon_ttm_placement_from_domain(rbo, RADEON_GEM_DOMAIN_VRAM);
-			rbo->placement.lpfn = rdev->mc.visible_vram_size >> PAGE_SHIFT;
-			r = ttm_bo_validate(bo, &rbo->placement, false, false);
-			if (unlikely(r != 0))
-				return r;
-			offset = bo->mem.start << PAGE_SHIFT;
-			/* this should not happen */
-			if ((offset + size) > rdev->mc.visible_vram_size)
-				return -EINVAL;
-		}
+	if (bo->mem.mem_type != TTM_PL_VRAM)
+		return 0;
+
+	size = bo->mem.num_pages << PAGE_SHIFT;
+	offset = bo->mem.start << PAGE_SHIFT;
+	if ((offset + size) <= rdev->mc.visible_vram_size)
+		return 0;
+
+	/* hurrah the memory is not visible ! */
+	radeon_ttm_placement_from_domain(rbo, RADEON_GEM_DOMAIN_VRAM);
+	rbo->placement.lpfn = rdev->mc.visible_vram_size >> PAGE_SHIFT;
+	r = ttm_bo_validate(bo, &rbo->placement, false, false);
+	if (unlikely(r == -ENOMEM)) {
+		radeon_ttm_placement_from_domain(rbo, RADEON_GEM_DOMAIN_GTT);
+		return ttm_bo_validate(bo, &rbo->placement, false, false);
+	} else if (unlikely(r != 0)) {
+		return r;
 	}
+
+	offset = bo->mem.start << PAGE_SHIFT;
+	/* this should never happen */
+	if ((offset + size) > rdev->mc.visible_vram_size)
+		return -EINVAL;
+
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/radeon/radeon_pm.c b/drivers/gpu/drm/radeon/radeon_pm.c
index 6fac8efe8340..2bdae61c0ac0 100644
--- a/drivers/gpu/drm/radeon/radeon_pm.c
+++ b/drivers/gpu/drm/radeon/radeon_pm.c
@@ -361,6 +361,11 @@ static ssize_t radeon_set_pm_profile(struct device *dev,
 	struct drm_device *ddev = dev_get_drvdata(dev);
 	struct radeon_device *rdev = ddev->dev_private;
 
+	/* Can't set profile when the card is off */
+	if  ((rdev->flags & RADEON_IS_PX) &&
+	     (ddev->switch_power_state != DRM_SWITCH_POWER_ON))
+		return -EINVAL;
+
 	mutex_lock(&rdev->pm.mutex);
 	if (rdev->pm.pm_method == PM_METHOD_PROFILE) {
 		if (strncmp("default", buf, strlen("default")) == 0)
@@ -409,6 +414,13 @@ static ssize_t radeon_set_pm_method(struct device *dev,
 	struct drm_device *ddev = dev_get_drvdata(dev);
 	struct radeon_device *rdev = ddev->dev_private;
 
+	/* Can't set method when the card is off */
+	if  ((rdev->flags & RADEON_IS_PX) &&
+	     (ddev->switch_power_state != DRM_SWITCH_POWER_ON)) {
+		count = -EINVAL;
+		goto fail;
+	}
+
 	/* we don't support the legacy modes with dpm */
 	if (rdev->pm.pm_method == PM_METHOD_DPM) {
 		count = -EINVAL;
@@ -446,6 +458,10 @@ static ssize_t radeon_get_dpm_state(struct device *dev,
 	struct radeon_device *rdev = ddev->dev_private;
 	enum radeon_pm_state_type pm = rdev->pm.dpm.user_state;
 
+	if  ((rdev->flags & RADEON_IS_PX) &&
+	     (ddev->switch_power_state != DRM_SWITCH_POWER_ON))
+		return snprintf(buf, PAGE_SIZE, "off\n");
+
 	return snprintf(buf, PAGE_SIZE, "%s\n",
 			(pm == POWER_STATE_TYPE_BATTERY) ? "battery" :
 			(pm == POWER_STATE_TYPE_BALANCED) ? "balanced" : "performance");
@@ -459,6 +475,11 @@ static ssize_t radeon_set_dpm_state(struct device *dev,
 	struct drm_device *ddev = dev_get_drvdata(dev);
 	struct radeon_device *rdev = ddev->dev_private;
 
+	/* Can't set dpm state when the card is off */
+	if  ((rdev->flags & RADEON_IS_PX) &&
+	     (ddev->switch_power_state != DRM_SWITCH_POWER_ON))
+		return -EINVAL;
+
 	mutex_lock(&rdev->pm.mutex);
 	if (strncmp("battery", buf, strlen("battery")) == 0)
 		rdev->pm.dpm.user_state = POWER_STATE_TYPE_BATTERY;
@@ -485,6 +506,10 @@ static ssize_t radeon_get_dpm_forced_performance_level(struct device *dev,
 	struct radeon_device *rdev = ddev->dev_private;
 	enum radeon_dpm_forced_level level = rdev->pm.dpm.forced_level;
 
+	if  ((rdev->flags & RADEON_IS_PX) &&
+	     (ddev->switch_power_state != DRM_SWITCH_POWER_ON))
+		return snprintf(buf, PAGE_SIZE, "off\n");
+
 	return snprintf(buf, PAGE_SIZE, "%s\n",
 			(level == RADEON_DPM_FORCED_LEVEL_AUTO) ? "auto" :
 			(level == RADEON_DPM_FORCED_LEVEL_LOW) ? "low" : "high");
@@ -500,6 +525,11 @@ static ssize_t radeon_set_dpm_forced_performance_level(struct device *dev,
 	enum radeon_dpm_forced_level level;
 	int ret = 0;
 
+	/* Can't force performance level when the card is off */
+	if  ((rdev->flags & RADEON_IS_PX) &&
+	     (ddev->switch_power_state != DRM_SWITCH_POWER_ON))
+		return -EINVAL;
+
 	mutex_lock(&rdev->pm.mutex);
 	if (strncmp("low", buf, strlen("low")) == 0) {
 		level = RADEON_DPM_FORCED_LEVEL_LOW;
@@ -538,8 +568,14 @@ static ssize_t radeon_hwmon_show_temp(struct device *dev,
 				      char *buf)
 {
 	struct radeon_device *rdev = dev_get_drvdata(dev);
+	struct drm_device *ddev = rdev->ddev;
 	int temp;
 
+	/* Can't get temperature when the card is off */
+	if  ((rdev->flags & RADEON_IS_PX) &&
+	     (ddev->switch_power_state != DRM_SWITCH_POWER_ON))
+		return -EINVAL;
+
 	if (rdev->asic->pm.get_temperature)
 		temp = radeon_get_temperature(rdev);
 	else
@@ -1068,7 +1104,6 @@ static void radeon_pm_resume_dpm(struct radeon_device *rdev)
 	if (ret)
 		goto dpm_resume_fail;
 	rdev->pm.dpm_enabled = true;
-	radeon_pm_compute_clocks(rdev);
 	return;
 
 dpm_resume_fail:
@@ -1300,6 +1335,7 @@ int radeon_pm_init(struct radeon_device *rdev)
 	case CHIP_KABINI:
 	case CHIP_KAVERI:
 	case CHIP_HAWAII:
+	case CHIP_MULLINS:
 		/* DPM requires the RLC, RV770+ dGPU requires SMC */
 		if (!rdev->rlc_fw)
 			rdev->pm.pm_method = PM_METHOD_PROFILE;
@@ -1613,8 +1649,12 @@ static int radeon_debugfs_pm_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 radeon_device *rdev = dev->dev_private;
+	struct drm_device *ddev = rdev->ddev;
 
-	if (rdev->pm.dpm_enabled) {
+	if  ((rdev->flags & RADEON_IS_PX) &&
+	     (ddev->switch_power_state != DRM_SWITCH_POWER_ON)) {
+		seq_printf(m, "PX asic powered off\n");
+	} else if (rdev->pm.dpm_enabled) {
 		mutex_lock(&rdev->pm.mutex);
 		if (rdev->asic->dpm.debugfs_print_current_performance_level)
 			radeon_dpm_debugfs_print_current_performance_level(rdev, m);
diff --git a/drivers/gpu/drm/radeon/radeon_ucode.h b/drivers/gpu/drm/radeon/radeon_ucode.h
index 58d12938c0b8..4e7c3269b183 100644
--- a/drivers/gpu/drm/radeon/radeon_ucode.h
+++ b/drivers/gpu/drm/radeon/radeon_ucode.h
@@ -52,6 +52,7 @@
 #define BONAIRE_RLC_UCODE_SIZE       2048
 #define KB_RLC_UCODE_SIZE            2560
 #define KV_RLC_UCODE_SIZE            2560
+#define ML_RLC_UCODE_SIZE            2560
 
 /* MC */
 #define BTC_MC_UCODE_SIZE            6024
diff --git a/drivers/gpu/drm/radeon/radeon_uvd.c b/drivers/gpu/drm/radeon/radeon_uvd.c
index 5748bdaeacce..1b65ae2433cd 100644
--- a/drivers/gpu/drm/radeon/radeon_uvd.c
+++ b/drivers/gpu/drm/radeon/radeon_uvd.c
@@ -99,6 +99,7 @@ int radeon_uvd_init(struct radeon_device *rdev)
 	case CHIP_KABINI:
 	case CHIP_KAVERI:
 	case CHIP_HAWAII:
+	case CHIP_MULLINS:
 		fw_name = FIRMWARE_BONAIRE;
 		break;
 
@@ -465,6 +466,10 @@ static int radeon_uvd_cs_reloc(struct radeon_cs_parser *p,
 	cmd = radeon_get_ib_value(p, p->idx) >> 1;
 
 	if (cmd < 0x4) {
+		if (end <= start) {
+			DRM_ERROR("invalid reloc offset %X!\n", offset);
+			return -EINVAL;
+		}
 		if ((end - start) < buf_sizes[cmd]) {
 			DRM_ERROR("buffer (%d) to small (%d / %d)!\n", cmd,
 				  (unsigned)(end - start), buf_sizes[cmd]);
diff --git a/drivers/gpu/drm/radeon/radeon_vce.c b/drivers/gpu/drm/radeon/radeon_vce.c
index ced53dd03e7c..3971d968af6c 100644
--- a/drivers/gpu/drm/radeon/radeon_vce.c
+++ b/drivers/gpu/drm/radeon/radeon_vce.c
@@ -66,6 +66,7 @@ int radeon_vce_init(struct radeon_device *rdev)
 	case CHIP_BONAIRE:
 	case CHIP_KAVERI:
 	case CHIP_KABINI:
+	case CHIP_MULLINS:
 		fw_name = FIRMWARE_BONAIRE;
 		break;
 
@@ -442,13 +443,16 @@ int radeon_vce_get_destroy_msg(struct radeon_device *rdev, int ring,
  * @p: parser context
  * @lo: address of lower dword
  * @hi: address of higher dword
+ * @size: size of checker for relocation buffer
  *
  * Patch relocation inside command stream with real buffer address
  */
-int radeon_vce_cs_reloc(struct radeon_cs_parser *p, int lo, int hi)
+int radeon_vce_cs_reloc(struct radeon_cs_parser *p, int lo, int hi,
+			unsigned size)
 {
 	struct radeon_cs_chunk *relocs_chunk;
-	uint64_t offset;
+	struct radeon_cs_reloc *reloc;
+	uint64_t start, end, offset;
 	unsigned idx;
 
 	relocs_chunk = &p->chunks[p->chunk_relocs_idx];
@@ -461,15 +465,60 @@ int radeon_vce_cs_reloc(struct radeon_cs_parser *p, int lo, int hi)
 		return -EINVAL;
 	}
 
-	offset += p->relocs_ptr[(idx / 4)]->gpu_offset;
+	reloc = p->relocs_ptr[(idx / 4)];
+	start = reloc->gpu_offset;
+	end = start + radeon_bo_size(reloc->robj);
+	start += offset;
 
-        p->ib.ptr[lo] = offset & 0xFFFFFFFF;
-        p->ib.ptr[hi] = offset >> 32;
+	p->ib.ptr[lo] = start & 0xFFFFFFFF;
+	p->ib.ptr[hi] = start >> 32;
+
+	if (end <= start) {
+		DRM_ERROR("invalid reloc offset %llX!\n", offset);
+		return -EINVAL;
+	}
+	if ((end - start) < size) {
+		DRM_ERROR("buffer to small (%d / %d)!\n",
+			(unsigned)(end - start), size);
+		return -EINVAL;
+	}
 
 	return 0;
 }
 
 /**
+ * radeon_vce_validate_handle - validate stream handle
+ *
+ * @p: parser context
+ * @handle: handle to validate
+ *
+ * Validates the handle and return the found session index or -EINVAL
+ * we we don't have another free session index.
+ */
+int radeon_vce_validate_handle(struct radeon_cs_parser *p, uint32_t handle)
+{
+	unsigned i;
+
+	/* validate the handle */
+	for (i = 0; i < RADEON_MAX_VCE_HANDLES; ++i) {
+		if (atomic_read(&p->rdev->vce.handles[i]) == handle)
+			return i;
+	}
+
+	/* handle not found try to alloc a new one */
+	for (i = 0; i < RADEON_MAX_VCE_HANDLES; ++i) {
+		if (!atomic_cmpxchg(&p->rdev->vce.handles[i], 0, handle)) {
+			p->rdev->vce.filp[i] = p->filp;
+			p->rdev->vce.img_size[i] = 0;
+			return i;
+		}
+	}
+
+	DRM_ERROR("No more free VCE handles!\n");
+	return -EINVAL;
+}
+
+/**
  * radeon_vce_cs_parse - parse and validate the command stream
  *
  * @p: parser context
@@ -477,8 +526,10 @@ int radeon_vce_cs_reloc(struct radeon_cs_parser *p, int lo, int hi)
  */
 int radeon_vce_cs_parse(struct radeon_cs_parser *p)
 {
-	uint32_t handle = 0;
-	bool destroy = false;
+	int session_idx = -1;
+	bool destroyed = false;
+	uint32_t tmp, handle = 0;
+	uint32_t *size = &tmp;
 	int i, r;
 
 	while (p->idx < p->chunks[p->chunk_ib_idx].length_dw) {
@@ -490,13 +541,29 @@ int radeon_vce_cs_parse(struct radeon_cs_parser *p)
                 	return -EINVAL;
 		}
 
+		if (destroyed) {
+			DRM_ERROR("No other command allowed after destroy!\n");
+			return -EINVAL;
+		}
+
 		switch (cmd) {
 		case 0x00000001: // session
 			handle = radeon_get_ib_value(p, p->idx + 2);
+			session_idx = radeon_vce_validate_handle(p, handle);
+			if (session_idx < 0)
+				return session_idx;
+			size = &p->rdev->vce.img_size[session_idx];
 			break;
 
 		case 0x00000002: // task info
+			break;
+
 		case 0x01000001: // create
+			*size = radeon_get_ib_value(p, p->idx + 8) *
+				radeon_get_ib_value(p, p->idx + 10) *
+				8 * 3 / 2;
+			break;
+
 		case 0x04000001: // config extension
 		case 0x04000002: // pic control
 		case 0x04000005: // rate control
@@ -505,23 +572,39 @@ int radeon_vce_cs_parse(struct radeon_cs_parser *p)
 			break;
 
 		case 0x03000001: // encode
-			r = radeon_vce_cs_reloc(p, p->idx + 10, p->idx + 9);
+			r = radeon_vce_cs_reloc(p, p->idx + 10, p->idx + 9,
+						*size);
 			if (r)
 				return r;
 
-			r = radeon_vce_cs_reloc(p, p->idx + 12, p->idx + 11);
+			r = radeon_vce_cs_reloc(p, p->idx + 12, p->idx + 11,
+						*size / 3);
 			if (r)
 				return r;
 			break;
 
 		case 0x02000001: // destroy
-			destroy = true;
+			destroyed = true;
 			break;
 
 		case 0x05000001: // context buffer
+			r = radeon_vce_cs_reloc(p, p->idx + 3, p->idx + 2,
+						*size * 2);
+			if (r)
+				return r;
+			break;
+
 		case 0x05000004: // video bitstream buffer
+			tmp = radeon_get_ib_value(p, p->idx + 4);
+			r = radeon_vce_cs_reloc(p, p->idx + 3, p->idx + 2,
+						tmp);
+			if (r)
+				return r;
+			break;
+
 		case 0x05000005: // feedback buffer
-			r = radeon_vce_cs_reloc(p, p->idx + 3, p->idx + 2);
+			r = radeon_vce_cs_reloc(p, p->idx + 3, p->idx + 2,
+						4096);
 			if (r)
 				return r;
 			break;
@@ -531,33 +614,21 @@ int radeon_vce_cs_parse(struct radeon_cs_parser *p)
 			return -EINVAL;
 		}
 
+		if (session_idx == -1) {
+			DRM_ERROR("no session command at start of IB\n");
+			return -EINVAL;
+		}
+
 		p->idx += len / 4;
 	}
 
-	if (destroy) {
+	if (destroyed) {
 		/* IB contains a destroy msg, free the handle */
 		for (i = 0; i < RADEON_MAX_VCE_HANDLES; ++i)
 			atomic_cmpxchg(&p->rdev->vce.handles[i], handle, 0);
-
-		return 0;
-        }
-
-	/* create or encode, validate the handle */
-	for (i = 0; i < RADEON_MAX_VCE_HANDLES; ++i) {
-		if (atomic_read(&p->rdev->vce.handles[i]) == handle)
-			return 0;
 	}
 
-	/* handle not found try to alloc a new one */
-	for (i = 0; i < RADEON_MAX_VCE_HANDLES; ++i) {
-		if (!atomic_cmpxchg(&p->rdev->vce.handles[i], 0, handle)) {
-			p->rdev->vce.filp[i] = p->filp;
-			return 0;
-		}
-	}
-
-	DRM_ERROR("No more free VCE handles!\n");
-	return -EINVAL;
+	return 0;
 }
 
 /**
diff --git a/drivers/gpu/drm/radeon/radeon_vm.c b/drivers/gpu/drm/radeon/radeon_vm.c
index a128a4fd64b3..a72e9c81805d 100644
--- a/drivers/gpu/drm/radeon/radeon_vm.c
+++ b/drivers/gpu/drm/radeon/radeon_vm.c
@@ -130,10 +130,10 @@ struct radeon_cs_reloc *radeon_vm_get_bos(struct radeon_device *rdev,
 					  struct list_head *head)
 {
 	struct radeon_cs_reloc *list;
-	unsigned i, idx, size;
+	unsigned i, idx;
 
-	size = (radeon_vm_num_pdes(rdev) + 1) * sizeof(struct radeon_cs_reloc);
-	list = kmalloc(size, GFP_KERNEL);
+	list = kmalloc_array(vm->max_pde_used + 2,
+			     sizeof(struct radeon_cs_reloc), GFP_KERNEL);
 	if (!list)
 		return NULL;
 
@@ -585,7 +585,8 @@ int radeon_vm_update_page_directory(struct radeon_device *rdev,
 {
 	static const uint32_t incr = RADEON_VM_PTE_COUNT * 8;
 
-	uint64_t pd_addr = radeon_bo_gpu_offset(vm->page_directory);
+	struct radeon_bo *pd = vm->page_directory;
+	uint64_t pd_addr = radeon_bo_gpu_offset(pd);
 	uint64_t last_pde = ~0, last_pt = ~0;
 	unsigned count = 0, pt_idx, ndw;
 	struct radeon_ib ib;
@@ -595,7 +596,7 @@ int radeon_vm_update_page_directory(struct radeon_device *rdev,
 	ndw = 64;
 
 	/* assume the worst case */
-	ndw += vm->max_pde_used * 12;
+	ndw += vm->max_pde_used * 16;
 
 	/* update too big for an IB */
 	if (ndw > 0xfffff)
@@ -642,6 +643,7 @@ int radeon_vm_update_page_directory(struct radeon_device *rdev,
 					incr, R600_PTE_VALID);
 
 	if (ib.length_dw != 0) {
+		radeon_semaphore_sync_to(ib.semaphore, pd->tbo.sync_obj);
 		radeon_semaphore_sync_to(ib.semaphore, vm->last_id_use);
 		r = radeon_ib_schedule(rdev, &ib, NULL);
 		if (r) {
@@ -767,15 +769,18 @@ static void radeon_vm_update_ptes(struct radeon_device *rdev,
 	/* walk over the address space and update the page tables */
 	for (addr = start; addr < end; ) {
 		uint64_t pt_idx = addr >> RADEON_VM_BLOCK_SIZE;
+		struct radeon_bo *pt = vm->page_tables[pt_idx].bo;
 		unsigned nptes;
 		uint64_t pte;
 
+		radeon_semaphore_sync_to(ib->semaphore, pt->tbo.sync_obj);
+
 		if ((addr & ~mask) == (end & ~mask))
 			nptes = end - addr;
 		else
 			nptes = RADEON_VM_PTE_COUNT - (addr & mask);
 
-		pte = radeon_bo_gpu_offset(vm->page_tables[pt_idx].bo);
+		pte = radeon_bo_gpu_offset(pt);
 		pte += (addr & mask) * 8;
 
 		if ((last_pte + 8 * count) != pte) {
diff --git a/drivers/gpu/drm/radeon/rv770_dma.c b/drivers/gpu/drm/radeon/rv770_dma.c
index aca8cbe8a335..bbf2e076ee45 100644
--- a/drivers/gpu/drm/radeon/rv770_dma.c
+++ b/drivers/gpu/drm/radeon/rv770_dma.c
@@ -86,6 +86,7 @@ int rv770_copy_dma(struct radeon_device *rdev,
 	r = radeon_fence_emit(rdev, fence, ring->idx);
 	if (r) {
 		radeon_ring_unlock_undo(rdev, ring);
+		radeon_semaphore_free(rdev, &sem, NULL);
 		return r;
 	}
 
diff --git a/drivers/gpu/drm/radeon/si.c b/drivers/gpu/drm/radeon/si.c
index 5c1c0c795e98..d64ef9115b69 100644
--- a/drivers/gpu/drm/radeon/si.c
+++ b/drivers/gpu/drm/radeon/si.c
@@ -5784,7 +5784,6 @@ int si_irq_set(struct radeon_device *rdev)
 	u32 crtc1 = 0, crtc2 = 0, crtc3 = 0, crtc4 = 0, crtc5 = 0, crtc6 = 0;
 	u32 hpd1 = 0, hpd2 = 0, hpd3 = 0, hpd4 = 0, hpd5 = 0, hpd6 = 0;
 	u32 grbm_int_cntl = 0;
-	u32 grph1 = 0, grph2 = 0, grph3 = 0, grph4 = 0, grph5 = 0, grph6 = 0;
 	u32 dma_cntl, dma_cntl1;
 	u32 thermal_int = 0;
 
@@ -5923,16 +5922,22 @@ int si_irq_set(struct radeon_device *rdev)
 	}
 
 	if (rdev->num_crtc >= 2) {
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC0_REGISTER_OFFSET, grph1);
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC1_REGISTER_OFFSET, grph2);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC0_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC1_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
 	}
 	if (rdev->num_crtc >= 4) {
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC2_REGISTER_OFFSET, grph3);
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC3_REGISTER_OFFSET, grph4);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC2_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC3_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
 	}
 	if (rdev->num_crtc >= 6) {
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC4_REGISTER_OFFSET, grph5);
-		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC5_REGISTER_OFFSET, grph6);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC4_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
+		WREG32(GRPH_INT_CONTROL + EVERGREEN_CRTC5_REGISTER_OFFSET,
+		       GRPH_PFLIP_INT_MASK);
 	}
 
 	if (!ASIC_IS_NODCE(rdev)) {
@@ -6296,6 +6301,15 @@ restart_ih:
 				break;
 			}
 			break;
+		case 8: /* D1 page flip */
+		case 10: /* D2 page flip */
+		case 12: /* D3 page flip */
+		case 14: /* D4 page flip */
+		case 16: /* D5 page flip */
+		case 18: /* D6 page flip */
+			DRM_DEBUG("IH: D%d flip\n", ((src_id - 8) >> 1) + 1);
+			radeon_crtc_handle_flip(rdev, (src_id - 8) >> 1);
+			break;
 		case 42: /* HPD hotplug */
 			switch (src_data) {
 			case 0:
diff --git a/drivers/gpu/drm/radeon/si_dma.c b/drivers/gpu/drm/radeon/si_dma.c
index 952166930fb8..9a660f861d2c 100644
--- a/drivers/gpu/drm/radeon/si_dma.c
+++ b/drivers/gpu/drm/radeon/si_dma.c
@@ -231,6 +231,7 @@ int si_copy_dma(struct radeon_device *rdev,
 	r = radeon_fence_emit(rdev, fence, ring->idx);
 	if (r) {
 		radeon_ring_unlock_undo(rdev, ring);
+		radeon_semaphore_free(rdev, &sem, NULL);
 		return r;
 	}
 
diff --git a/drivers/gpu/drm/radeon/sid.h b/drivers/gpu/drm/radeon/sid.h
index da8f8674a552..fd414d34d885 100644
--- a/drivers/gpu/drm/radeon/sid.h
+++ b/drivers/gpu/drm/radeon/sid.h
@@ -107,8 +107,8 @@
 #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
+#	define SPLL_REFCLK_SEL(x)			((x) << 26)
+#	define SPLL_REFCLK_SEL_MASK			(3 << 26)
 
 #define	CG_SPLL_SPREAD_SPECTRUM				0x620
 #define		SSEN					(1 << 0)
diff --git a/drivers/gpu/drm/radeon/uvd_v1_0.c b/drivers/gpu/drm/radeon/uvd_v1_0.c
index 0a243f0e5d68..be42c8125203 100644
--- a/drivers/gpu/drm/radeon/uvd_v1_0.c
+++ b/drivers/gpu/drm/radeon/uvd_v1_0.c
@@ -83,7 +83,10 @@ int uvd_v1_0_init(struct radeon_device *rdev)
 	int r;
 
 	/* raise clocks while booting up the VCPU */
-	radeon_set_uvd_clocks(rdev, 53300, 40000);
+	if (rdev->family < CHIP_RV740)
+		radeon_set_uvd_clocks(rdev, 10000, 10000);
+	else
+		radeon_set_uvd_clocks(rdev, 53300, 40000);
 
 	r = uvd_v1_0_start(rdev);
 	if (r)
@@ -407,7 +410,10 @@ int uvd_v1_0_ib_test(struct radeon_device *rdev, struct radeon_ring *ring)
 	struct radeon_fence *fence = NULL;
 	int r;
 
-	r = radeon_set_uvd_clocks(rdev, 53300, 40000);
+	if (rdev->family < CHIP_RV740)
+		r = radeon_set_uvd_clocks(rdev, 10000, 10000);
+	else
+		r = radeon_set_uvd_clocks(rdev, 53300, 40000);
 	if (r) {
 		DRM_ERROR("radeon: failed to raise UVD clocks (%d).\n", r);
 		return r;