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

Pull drm updates from Dave Airlie:
 "This is the main drm pull request for v4.2.

  I've one other new driver from freescale on my radar, it's been posted
  and reviewed, I'd just like to get someone to give it a last look, so
  maybe I'll send it or maybe I'll leave it.

  There is no major nouveau changes in here, Ben was working on
  something big, and we agreed it was a bit late, there wasn't anything
  else he considered urgent to merge.

  There might be another msm pull for some bits that are waiting on
  arm-soc, I'll see how we time it.

  This touches some "of" stuff, acks are in place except for the fixes
  to the build in various configs,t hat I just applied.

  Summary:

  New drivers:
      - virtio-gpu:
                KMS only pieces of driver for virtio-gpu in qemu.
                This is just the first part of this driver, enough to run
                unaccelerated userspace on. As qemu merges more we'll start
                adding the 3D features for the virgl 3d work.
      - amdgpu:
                a new driver from AMD to driver their newer GPUs. (VI+)
                It contains a new cleaner userspace API, and is a clean
                break from radeon moving forward, that AMD are going to
                concentrate on. It also contains a set of register headers
                auto generated from AMD internal database.

  core:
      - atomic modesetting API completed, enabled by default now.
      - Add support for mode_id blob to atomic ioctl to complete interface.
      - bunch of Displayport MST fixes
      - lots of misc fixes.

  panel:
      - new simple panels
      - fix some long-standing build issues with bridge drivers

  radeon:
      - VCE1 support
      - add a GPU reset counter for userspace
      - lots of fixes.

  amdkfd:
      - H/W debugger support module
      - static user-mode queues
      - support killing all the waves when a process terminates
      - use standard DECLARE_BITMAP

  i915:
      - Add Broxton support
      - S3, rotation support for Skylake
      - RPS booting tuning
      - CPT modeset sequence fixes
      - ns2501 dither support
      - enable cmd parser on haswell
      - cdclk handling fixes
      - gen8 dynamic pte allocation
      - lots of atomic conversion work

  exynos:
      - Add atomic modesetting support
      - Add iommu support
      - Consolidate drm driver initialization
      - and MIC, DECON and MIPI-DSI support for exynos5433

  omapdrm:
      - atomic modesetting support (fixes lots of things in rewrite)

  tegra:
      - DP aux transaction fixes
      - iommu support fix

  msm:
      - adreno a306 support
      - various dsi bits
      - various 64-bit fixes
      - NV12MT support

  rcar-du:
      - atomic and misc fixes

  sti:
      - fix HDMI timing complaince

  tilcdc:
      - use drm component API to access tda998x driver
      - fix module unloading

  qxl:
      - stability fixes"

* 'drm-next' of git://people.freedesktop.org/~airlied/linux: (872 commits)
  drm/nouveau: Pause between setting gpu to D3hot and cutting the power
  drm/dp/mst: close deadlock in connector destruction.
  drm: Always enable atomic API
  drm/vgem: Set unique to "vgem"
  of: fix a build error to of_graph_get_endpoint_by_regs function
  drm/dp/mst: take lock around looking up the branch device on hpd irq
  drm/dp/mst: make sure mst_primary mstb is valid in work function
  of: add EXPORT_SYMBOL for of_graph_get_endpoint_by_regs
  ARM: dts: rename the clock of MIPI DSI 'pll_clk' to 'sclk_mipi'
  drm/atomic: Don't set crtc_state->enable manually
  drm/exynos: dsi: do not set TE GPIO direction by input
  drm/exynos: dsi: add support for MIC driver as a bridge
  drm/exynos: dsi: add support for Exynos5433
  drm/exynos: dsi: make use of array for clock access
  drm/exynos: dsi: make use of driver data for static values
  drm/exynos: dsi: add macros for register access
  drm/exynos: dsi: rename pll_clk to sclk_clk
  drm/exynos: mic: add MIC driver
  of: add helper for getting endpoint node of specific identifiers
  drm/exynos: add Exynos5433 decon driver
  ...

1 files changed, 588 insertions, 401 deletions

diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 2d0995e7afc3..248fd1ac7b3a 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -38,14 +38,14 @@
 #include <linux/pci.h>
 #include <linux/dma-buf.h>
 
+#define RQ_BUG_ON(expr)
+
 static void i915_gem_object_flush_gtt_write_domain(struct drm_i915_gem_object *obj);
 static void i915_gem_object_flush_cpu_write_domain(struct drm_i915_gem_object *obj);
-static __must_check int
-i915_gem_object_wait_rendering(struct drm_i915_gem_object *obj,
-			       bool readonly);
 static void
-i915_gem_object_retire(struct drm_i915_gem_object *obj);
-
+i915_gem_object_retire__write(struct drm_i915_gem_object *obj);
+static void
+i915_gem_object_retire__read(struct drm_i915_gem_object *obj, int ring);
 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,
@@ -378,13 +378,13 @@ out:
 void *i915_gem_object_alloc(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	return kmem_cache_zalloc(dev_priv->slab, GFP_KERNEL);
+	return kmem_cache_zalloc(dev_priv->objects, GFP_KERNEL);
 }
 
 void i915_gem_object_free(struct drm_i915_gem_object *obj)
 {
 	struct drm_i915_private *dev_priv = obj->base.dev->dev_private;
-	kmem_cache_free(dev_priv->slab, obj);
+	kmem_cache_free(dev_priv->objects, obj);
 }
 
 static int
@@ -518,8 +518,6 @@ int i915_gem_obj_prepare_shmem_read(struct drm_i915_gem_object *obj,
 		ret = i915_gem_object_wait_rendering(obj, true);
 		if (ret)
 			return ret;
-
-		i915_gem_object_retire(obj);
 	}
 
 	ret = i915_gem_object_get_pages(obj);
@@ -939,8 +937,6 @@ i915_gem_shmem_pwrite(struct drm_device *dev,
 		ret = i915_gem_object_wait_rendering(obj, false);
 		if (ret)
 			return ret;
-
-		i915_gem_object_retire(obj);
 	}
 	/* Same trick applies to invalidate partially written cachelines read
 	 * before writing. */
@@ -1181,12 +1177,27 @@ static bool missed_irq(struct drm_i915_private *dev_priv,
 	return test_bit(ring->id, &dev_priv->gpu_error.missed_irq_rings);
 }
 
-static bool can_wait_boost(struct drm_i915_file_private *file_priv)
+static int __i915_spin_request(struct drm_i915_gem_request *req)
 {
-	if (file_priv == NULL)
-		return true;
+	unsigned long timeout;
 
-	return !atomic_xchg(&file_priv->rps_wait_boost, true);
+	if (i915_gem_request_get_ring(req)->irq_refcount)
+		return -EBUSY;
+
+	timeout = jiffies + 1;
+	while (!need_resched()) {
+		if (i915_gem_request_completed(req, true))
+			return 0;
+
+		if (time_after_eq(jiffies, timeout))
+			break;
+
+		cpu_relax_lowlatency();
+	}
+	if (i915_gem_request_completed(req, false))
+		return 0;
+
+	return -EAGAIN;
 }
 
 /**
@@ -1210,7 +1221,7 @@ int __i915_wait_request(struct drm_i915_gem_request *req,
 			unsigned reset_counter,
 			bool interruptible,
 			s64 *timeout,
-			struct drm_i915_file_private *file_priv)
+			struct intel_rps_client *rps)
 {
 	struct intel_engine_cs *ring = i915_gem_request_get_ring(req);
 	struct drm_device *dev = ring->dev;
@@ -1224,26 +1235,32 @@ int __i915_wait_request(struct drm_i915_gem_request *req,
 
 	WARN(!intel_irqs_enabled(dev_priv), "IRQs disabled");
 
+	if (list_empty(&req->list))
+		return 0;
+
 	if (i915_gem_request_completed(req, true))
 		return 0;
 
 	timeout_expire = timeout ?
 		jiffies + nsecs_to_jiffies_timeout((u64)*timeout) : 0;
 
-	if (INTEL_INFO(dev)->gen >= 6 && ring->id == RCS && can_wait_boost(file_priv)) {
-		gen6_rps_boost(dev_priv);
-		if (file_priv)
-			mod_delayed_work(dev_priv->wq,
-					 &file_priv->mm.idle_work,
-					 msecs_to_jiffies(100));
-	}
-
-	if (!irq_test_in_progress && WARN_ON(!ring->irq_get(ring)))
-		return -ENODEV;
+	if (INTEL_INFO(dev_priv)->gen >= 6)
+		gen6_rps_boost(dev_priv, rps, req->emitted_jiffies);
 
 	/* Record current time in case interrupted by signal, or wedged */
 	trace_i915_gem_request_wait_begin(req);
 	before = ktime_get_raw_ns();
+
+	/* Optimistic spin for the next jiffie before touching IRQs */
+	ret = __i915_spin_request(req);
+	if (ret == 0)
+		goto out;
+
+	if (!irq_test_in_progress && WARN_ON(!ring->irq_get(ring))) {
+		ret = -ENODEV;
+		goto out;
+	}
+
 	for (;;) {
 		struct timer_list timer;
 
@@ -1292,14 +1309,15 @@ int __i915_wait_request(struct drm_i915_gem_request *req,
 			destroy_timer_on_stack(&timer);
 		}
 	}
-	now = ktime_get_raw_ns();
-	trace_i915_gem_request_wait_end(req);
-
 	if (!irq_test_in_progress)
 		ring->irq_put(ring);
 
 	finish_wait(&ring->irq_queue, &wait);
 
+out:
+	now = ktime_get_raw_ns();
+	trace_i915_gem_request_wait_end(req);
+
 	if (timeout) {
 		s64 tres = *timeout - (now - before);
 
@@ -1319,6 +1337,63 @@ int __i915_wait_request(struct drm_i915_gem_request *req,
 	return ret;
 }
 
+static inline void
+i915_gem_request_remove_from_client(struct drm_i915_gem_request *request)
+{
+	struct drm_i915_file_private *file_priv = request->file_priv;
+
+	if (!file_priv)
+		return;
+
+	spin_lock(&file_priv->mm.lock);
+	list_del(&request->client_list);
+	request->file_priv = NULL;
+	spin_unlock(&file_priv->mm.lock);
+}
+
+static void i915_gem_request_retire(struct drm_i915_gem_request *request)
+{
+	trace_i915_gem_request_retire(request);
+
+	/* We know the GPU must have read the request to have
+	 * sent us the seqno + interrupt, so use the position
+	 * of tail of the request to update the last known position
+	 * of the GPU head.
+	 *
+	 * Note this requires that we are always called in request
+	 * completion order.
+	 */
+	request->ringbuf->last_retired_head = request->postfix;
+
+	list_del_init(&request->list);
+	i915_gem_request_remove_from_client(request);
+
+	put_pid(request->pid);
+
+	i915_gem_request_unreference(request);
+}
+
+static void
+__i915_gem_request_retire__upto(struct drm_i915_gem_request *req)
+{
+	struct intel_engine_cs *engine = req->ring;
+	struct drm_i915_gem_request *tmp;
+
+	lockdep_assert_held(&engine->dev->struct_mutex);
+
+	if (list_empty(&req->list))
+		return;
+
+	do {
+		tmp = list_first_entry(&engine->request_list,
+				       typeof(*tmp), list);
+
+		i915_gem_request_retire(tmp);
+	} while (tmp != req);
+
+	WARN_ON(i915_verify_lists(engine->dev));
+}
+
 /**
  * Waits for a request to be signaled, and cleans up the
  * request and object lists appropriately for that event.
@@ -1329,7 +1404,6 @@ i915_wait_request(struct drm_i915_gem_request *req)
 	struct drm_device *dev;
 	struct drm_i915_private *dev_priv;
 	bool interruptible;
-	unsigned reset_counter;
 	int ret;
 
 	BUG_ON(req == NULL);
@@ -1348,29 +1422,13 @@ i915_wait_request(struct drm_i915_gem_request *req)
 	if (ret)
 		return ret;
 
-	reset_counter = atomic_read(&dev_priv->gpu_error.reset_counter);
-	i915_gem_request_reference(req);
-	ret = __i915_wait_request(req, reset_counter,
+	ret = __i915_wait_request(req,
+				  atomic_read(&dev_priv->gpu_error.reset_counter),
 				  interruptible, NULL, NULL);
-	i915_gem_request_unreference(req);
-	return ret;
-}
-
-static int
-i915_gem_object_wait_rendering__tail(struct drm_i915_gem_object *obj)
-{
-	if (!obj->active)
-		return 0;
-
-	/* Manually manage the write flush as we may have not yet
-	 * retired the buffer.
-	 *
-	 * Note that the last_write_req is always the earlier of
-	 * the two (read/write) requests, so if we haved successfully waited,
-	 * we know we have passed the last write.
-	 */
-	i915_gem_request_assign(&obj->last_write_req, NULL);
+	if (ret)
+		return ret;
 
+	__i915_gem_request_retire__upto(req);
 	return 0;
 }
 
@@ -1378,22 +1436,56 @@ i915_gem_object_wait_rendering__tail(struct drm_i915_gem_object *obj)
  * Ensures that all rendering to the object has completed and the object is
  * safe to unbind from the GTT or access from the CPU.
  */
-static __must_check int
+int
 i915_gem_object_wait_rendering(struct drm_i915_gem_object *obj,
 			       bool readonly)
 {
-	struct drm_i915_gem_request *req;
-	int ret;
+	int ret, i;
 
-	req = readonly ? obj->last_write_req : obj->last_read_req;
-	if (!req)
+	if (!obj->active)
 		return 0;
 
-	ret = i915_wait_request(req);
-	if (ret)
-		return ret;
+	if (readonly) {
+		if (obj->last_write_req != NULL) {
+			ret = i915_wait_request(obj->last_write_req);
+			if (ret)
+				return ret;
+
+			i = obj->last_write_req->ring->id;
+			if (obj->last_read_req[i] == obj->last_write_req)
+				i915_gem_object_retire__read(obj, i);
+			else
+				i915_gem_object_retire__write(obj);
+		}
+	} else {
+		for (i = 0; i < I915_NUM_RINGS; i++) {
+			if (obj->last_read_req[i] == NULL)
+				continue;
+
+			ret = i915_wait_request(obj->last_read_req[i]);
+			if (ret)
+				return ret;
+
+			i915_gem_object_retire__read(obj, i);
+		}
+		RQ_BUG_ON(obj->active);
+	}
+
+	return 0;
+}
+
+static void
+i915_gem_object_retire_request(struct drm_i915_gem_object *obj,
+			       struct drm_i915_gem_request *req)
+{
+	int ring = req->ring->id;
+
+	if (obj->last_read_req[ring] == req)
+		i915_gem_object_retire__read(obj, ring);
+	else if (obj->last_write_req == req)
+		i915_gem_object_retire__write(obj);
 
-	return i915_gem_object_wait_rendering__tail(obj);
+	__i915_gem_request_retire__upto(req);
 }
 
 /* A nonblocking variant of the above wait. This is a highly dangerous routine
@@ -1401,40 +1493,75 @@ i915_gem_object_wait_rendering(struct drm_i915_gem_object *obj,
  */
 static __must_check int
 i915_gem_object_wait_rendering__nonblocking(struct drm_i915_gem_object *obj,
-					    struct drm_i915_file_private *file_priv,
+					    struct intel_rps_client *rps,
 					    bool readonly)
 {
-	struct drm_i915_gem_request *req;
 	struct drm_device *dev = obj->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct drm_i915_gem_request *requests[I915_NUM_RINGS];
 	unsigned reset_counter;
-	int ret;
+	int ret, i, n = 0;
 
 	BUG_ON(!mutex_is_locked(&dev->struct_mutex));
 	BUG_ON(!dev_priv->mm.interruptible);
 
-	req = readonly ? obj->last_write_req : obj->last_read_req;
-	if (!req)
+	if (!obj->active)
 		return 0;
 
 	ret = i915_gem_check_wedge(&dev_priv->gpu_error, true);
 	if (ret)
 		return ret;
 
-	ret = i915_gem_check_olr(req);
-	if (ret)
-		return ret;
-
 	reset_counter = atomic_read(&dev_priv->gpu_error.reset_counter);
-	i915_gem_request_reference(req);
+
+	if (readonly) {
+		struct drm_i915_gem_request *req;
+
+		req = obj->last_write_req;
+		if (req == NULL)
+			return 0;
+
+		ret = i915_gem_check_olr(req);
+		if (ret)
+			goto err;
+
+		requests[n++] = i915_gem_request_reference(req);
+	} else {
+		for (i = 0; i < I915_NUM_RINGS; i++) {
+			struct drm_i915_gem_request *req;
+
+			req = obj->last_read_req[i];
+			if (req == NULL)
+				continue;
+
+			ret = i915_gem_check_olr(req);
+			if (ret)
+				goto err;
+
+			requests[n++] = i915_gem_request_reference(req);
+		}
+	}
+
 	mutex_unlock(&dev->struct_mutex);
-	ret = __i915_wait_request(req, reset_counter, true, NULL, file_priv);
+	for (i = 0; ret == 0 && i < n; i++)
+		ret = __i915_wait_request(requests[i], reset_counter, true,
+					  NULL, rps);
 	mutex_lock(&dev->struct_mutex);
-	i915_gem_request_unreference(req);
-	if (ret)
-		return ret;
 
-	return i915_gem_object_wait_rendering__tail(obj);
+err:
+	for (i = 0; i < n; i++) {
+		if (ret == 0)
+			i915_gem_object_retire_request(obj, requests[i]);
+		i915_gem_request_unreference(requests[i]);
+	}
+
+	return ret;
+}
+
+static struct intel_rps_client *to_rps_client(struct drm_file *file)
+{
+	struct drm_i915_file_private *fpriv = file->driver_priv;
+	return &fpriv->rps;
 }
 
 /**
@@ -1479,7 +1606,7 @@ i915_gem_set_domain_ioctl(struct drm_device *dev, void *data,
 	 * to catch cases where we are gazumped.
 	 */
 	ret = i915_gem_object_wait_rendering__nonblocking(obj,
-							  file->driver_priv,
+							  to_rps_client(file),
 							  !write_domain);
 	if (ret)
 		goto unref;
@@ -1616,6 +1743,7 @@ int i915_gem_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 	struct drm_i915_gem_object *obj = to_intel_bo(vma->vm_private_data);
 	struct drm_device *dev = obj->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct i915_ggtt_view view = i915_ggtt_view_normal;
 	pgoff_t page_offset;
 	unsigned long pfn;
 	int ret = 0;
@@ -1648,8 +1776,23 @@ int i915_gem_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 		goto unlock;
 	}
 
-	/* Now bind it into the GTT if needed */
-	ret = i915_gem_obj_ggtt_pin(obj, 0, PIN_MAPPABLE);
+	/* Use a partial view if the object is bigger than the aperture. */
+	if (obj->base.size >= dev_priv->gtt.mappable_end &&
+	    obj->tiling_mode == I915_TILING_NONE) {
+		static const unsigned int chunk_size = 256; // 1 MiB
+
+		memset(&view, 0, sizeof(view));
+		view.type = I915_GGTT_VIEW_PARTIAL;
+		view.params.partial.offset = rounddown(page_offset, chunk_size);
+		view.params.partial.size =
+			min_t(unsigned int,
+			      chunk_size,
+			      (vma->vm_end - vma->vm_start)/PAGE_SIZE -
+			      view.params.partial.offset);
+	}
+
+	/* Now pin it into the GTT if needed */
+	ret = i915_gem_object_ggtt_pin(obj, &view, 0, PIN_MAPPABLE);
 	if (ret)
 		goto unlock;
 
@@ -1662,30 +1805,50 @@ int i915_gem_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 		goto unpin;
 
 	/* Finally, remap it using the new GTT offset */
-	pfn = dev_priv->gtt.mappable_base + i915_gem_obj_ggtt_offset(obj);
+	pfn = dev_priv->gtt.mappable_base +
+		i915_gem_obj_ggtt_offset_view(obj, &view);
 	pfn >>= PAGE_SHIFT;
 
-	if (!obj->fault_mappable) {
-		unsigned long size = min_t(unsigned long,
-					   vma->vm_end - vma->vm_start,
-					   obj->base.size);
-		int i;
+	if (unlikely(view.type == I915_GGTT_VIEW_PARTIAL)) {
+		/* Overriding existing pages in partial view does not cause
+		 * us any trouble as TLBs are still valid because the fault
+		 * is due to userspace losing part of the mapping or never
+		 * having accessed it before (at this partials' range).
+		 */
+		unsigned long base = vma->vm_start +
+				     (view.params.partial.offset << PAGE_SHIFT);
+		unsigned int i;
 
-		for (i = 0; i < size >> PAGE_SHIFT; i++) {
-			ret = vm_insert_pfn(vma,
-					    (unsigned long)vma->vm_start + i * PAGE_SIZE,
-					    pfn + i);
+		for (i = 0; i < view.params.partial.size; i++) {
+			ret = vm_insert_pfn(vma, base + i * PAGE_SIZE, pfn + i);
 			if (ret)
 				break;
 		}
 
 		obj->fault_mappable = true;
-	} else
-		ret = vm_insert_pfn(vma,
-				    (unsigned long)vmf->virtual_address,
-				    pfn + page_offset);
+	} else {
+		if (!obj->fault_mappable) {
+			unsigned long size = min_t(unsigned long,
+						   vma->vm_end - vma->vm_start,
+						   obj->base.size);
+			int i;
+
+			for (i = 0; i < size >> PAGE_SHIFT; i++) {
+				ret = vm_insert_pfn(vma,
+						    (unsigned long)vma->vm_start + i * PAGE_SIZE,
+						    pfn + i);
+				if (ret)
+					break;
+			}
+
+			obj->fault_mappable = true;
+		} else
+			ret = vm_insert_pfn(vma,
+					    (unsigned long)vmf->virtual_address,
+					    pfn + page_offset);
+	}
 unpin:
-	i915_gem_object_ggtt_unpin(obj);
+	i915_gem_object_ggtt_unpin_view(obj, &view);
 unlock:
 	mutex_unlock(&dev->struct_mutex);
 out:
@@ -1864,7 +2027,6 @@ i915_gem_mmap_gtt(struct drm_file *file,
 		  uint32_t handle,
 		  uint64_t *offset)
 {
-	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_i915_gem_object *obj;
 	int ret;
 
@@ -1878,11 +2040,6 @@ i915_gem_mmap_gtt(struct drm_file *file,
 		goto unlock;
 	}
 
-	if (obj->base.size > dev_priv->gtt.mappable_end) {
-		ret = -E2BIG;
-		goto out;
-	}
-
 	if (obj->madv != I915_MADV_WILLNEED) {
 		DRM_DEBUG("Attempting to mmap a purgeable buffer\n");
 		ret = -EFAULT;
@@ -2178,81 +2335,65 @@ i915_gem_object_get_pages(struct drm_i915_gem_object *obj)
 		return ret;
 
 	list_add_tail(&obj->global_list, &dev_priv->mm.unbound_list);
+
+	obj->get_page.sg = obj->pages->sgl;
+	obj->get_page.last = 0;
+
 	return 0;
 }
 
-static void
-i915_gem_object_move_to_active(struct drm_i915_gem_object *obj,
-			       struct intel_engine_cs *ring)
+void i915_vma_move_to_active(struct i915_vma *vma,
+			     struct intel_engine_cs *ring)
 {
-	struct drm_i915_gem_request *req;
-	struct intel_engine_cs *old_ring;
-
-	BUG_ON(ring == NULL);
-
-	req = intel_ring_get_request(ring);
-	old_ring = i915_gem_request_get_ring(obj->last_read_req);
-
-	if (old_ring != ring && obj->last_write_req) {
-		/* Keep the request relative to the current ring */
-		i915_gem_request_assign(&obj->last_write_req, req);
-	}
+	struct drm_i915_gem_object *obj = vma->obj;
 
 	/* Add a reference if we're newly entering the active list. */
-	if (!obj->active) {
+	if (obj->active == 0)
 		drm_gem_object_reference(&obj->base);
-		obj->active = 1;
-	}
+	obj->active |= intel_ring_flag(ring);
 
-	list_move_tail(&obj->ring_list, &ring->active_list);
+	list_move_tail(&obj->ring_list[ring->id], &ring->active_list);
+	i915_gem_request_assign(&obj->last_read_req[ring->id],
+				intel_ring_get_request(ring));
 
-	i915_gem_request_assign(&obj->last_read_req, req);
+	list_move_tail(&vma->mm_list, &vma->vm->active_list);
 }
 
-void i915_vma_move_to_active(struct i915_vma *vma,
-			     struct intel_engine_cs *ring)
+static void
+i915_gem_object_retire__write(struct drm_i915_gem_object *obj)
 {
-	list_move_tail(&vma->mm_list, &vma->vm->active_list);
-	return i915_gem_object_move_to_active(vma->obj, ring);
+	RQ_BUG_ON(obj->last_write_req == NULL);
+	RQ_BUG_ON(!(obj->active & intel_ring_flag(obj->last_write_req->ring)));
+
+	i915_gem_request_assign(&obj->last_write_req, NULL);
+	intel_fb_obj_flush(obj, true);
 }
 
 static void
-i915_gem_object_move_to_inactive(struct drm_i915_gem_object *obj)
+i915_gem_object_retire__read(struct drm_i915_gem_object *obj, int ring)
 {
 	struct i915_vma *vma;
 
-	BUG_ON(obj->base.write_domain & ~I915_GEM_GPU_DOMAINS);
-	BUG_ON(!obj->active);
+	RQ_BUG_ON(obj->last_read_req[ring] == NULL);
+	RQ_BUG_ON(!(obj->active & (1 << ring)));
+
+	list_del_init(&obj->ring_list[ring]);
+	i915_gem_request_assign(&obj->last_read_req[ring], NULL);
+
+	if (obj->last_write_req && obj->last_write_req->ring->id == ring)
+		i915_gem_object_retire__write(obj);
+
+	obj->active &= ~(1 << ring);
+	if (obj->active)
+		return;
 
 	list_for_each_entry(vma, &obj->vma_list, vma_link) {
 		if (!list_empty(&vma->mm_list))
 			list_move_tail(&vma->mm_list, &vma->vm->inactive_list);
 	}
 
-	intel_fb_obj_flush(obj, true);
-
-	list_del_init(&obj->ring_list);
-
-	i915_gem_request_assign(&obj->last_read_req, NULL);
-	i915_gem_request_assign(&obj->last_write_req, NULL);
-	obj->base.write_domain = 0;
-
 	i915_gem_request_assign(&obj->last_fenced_req, NULL);
-
-	obj->active = 0;
 	drm_gem_object_unreference(&obj->base);
-
-	WARN_ON(i915_verify_lists(dev));
-}
-
-static void
-i915_gem_object_retire(struct drm_i915_gem_object *obj)
-{
-	if (obj->last_read_req == NULL)
-		return;
-
-	if (i915_gem_request_completed(obj->last_read_req, true))
-		i915_gem_object_move_to_inactive(obj);
 }
 
 static int
@@ -2421,7 +2562,6 @@ int __i915_add_request(struct intel_engine_cs *ring,
 
 	i915_queue_hangcheck(ring->dev);
 
-	cancel_delayed_work_sync(&dev_priv->mm.idle_work);
 	queue_delayed_work(dev_priv->wq,
 			   &dev_priv->mm.retire_work,
 			   round_jiffies_up_relative(HZ));
@@ -2430,20 +2570,6 @@ int __i915_add_request(struct intel_engine_cs *ring,
 	return 0;
 }
 
-static inline void
-i915_gem_request_remove_from_client(struct drm_i915_gem_request *request)
-{
-	struct drm_i915_file_private *file_priv = request->file_priv;
-
-	if (!file_priv)
-		return;
-
-	spin_lock(&file_priv->mm.lock);
-	list_del(&request->client_list);
-	request->file_priv = NULL;
-	spin_unlock(&file_priv->mm.lock);
-}
-
 static bool i915_context_is_banned(struct drm_i915_private *dev_priv,
 				   const struct intel_context *ctx)
 {
@@ -2489,16 +2615,6 @@ static void i915_set_reset_status(struct drm_i915_private *dev_priv,
 	}
 }
 
-static void i915_gem_free_request(struct drm_i915_gem_request *request)
-{
-	list_del(&request->list);
-	i915_gem_request_remove_from_client(request);
-
-	put_pid(request->pid);
-
-	i915_gem_request_unreference(request);
-}
-
 void i915_gem_request_free(struct kref *req_ref)
 {
 	struct drm_i915_gem_request *req = container_of(req_ref,
@@ -2516,7 +2632,45 @@ void i915_gem_request_free(struct kref *req_ref)
 		i915_gem_context_unreference(ctx);
 	}
 
-	kfree(req);
+	kmem_cache_free(req->i915->requests, req);
+}
+
+int i915_gem_request_alloc(struct intel_engine_cs *ring,
+			   struct intel_context *ctx)
+{
+	struct drm_i915_private *dev_priv = to_i915(ring->dev);
+	struct drm_i915_gem_request *req;
+	int ret;
+
+	if (ring->outstanding_lazy_request)
+		return 0;
+
+	req = kmem_cache_zalloc(dev_priv->requests, GFP_KERNEL);
+	if (req == NULL)
+		return -ENOMEM;
+
+	kref_init(&req->ref);
+	req->i915 = dev_priv;
+
+	ret = i915_gem_get_seqno(ring->dev, &req->seqno);
+	if (ret)
+		goto err;
+
+	req->ring = ring;
+
+	if (i915.enable_execlists)
+		ret = intel_logical_ring_alloc_request_extras(req, ctx);
+	else
+		ret = intel_ring_alloc_request_extras(req);
+	if (ret)
+		goto err;
+
+	ring->outstanding_lazy_request = req;
+	return 0;
+
+err:
+	kmem_cache_free(dev_priv->requests, req);
+	return ret;
 }
 
 struct drm_i915_gem_request *
@@ -2561,9 +2715,9 @@ static void i915_gem_reset_ring_cleanup(struct drm_i915_private *dev_priv,
 
 		obj = list_first_entry(&ring->active_list,
 				       struct drm_i915_gem_object,
-				       ring_list);
+				       ring_list[ring->id]);
 
-		i915_gem_object_move_to_inactive(obj);
+		i915_gem_object_retire__read(obj, ring->id);
 	}
 
 	/*
@@ -2578,7 +2732,6 @@ static void i915_gem_reset_ring_cleanup(struct drm_i915_private *dev_priv,
 				struct drm_i915_gem_request,
 				execlist_link);
 		list_del(&submit_req->execlist_link);
-		intel_runtime_pm_put(dev_priv);
 
 		if (submit_req->ctx != ring->default_context)
 			intel_lr_context_unpin(ring, submit_req->ctx);
@@ -2600,7 +2753,7 @@ static void i915_gem_reset_ring_cleanup(struct drm_i915_private *dev_priv,
 					   struct drm_i915_gem_request,
 					   list);
 
-		i915_gem_free_request(request);
+		i915_gem_request_retire(request);
 	}
 
 	/* This may not have been flushed before the reset, so clean it now */
@@ -2648,6 +2801,8 @@ void i915_gem_reset(struct drm_device *dev)
 	i915_gem_context_reset(dev);
 
 	i915_gem_restore_fences(dev);
+
+	WARN_ON(i915_verify_lists(dev));
 }
 
 /**
@@ -2656,9 +2811,6 @@ void i915_gem_reset(struct drm_device *dev)
 void
 i915_gem_retire_requests_ring(struct intel_engine_cs *ring)
 {
-	if (list_empty(&ring->request_list))
-		return;
-
 	WARN_ON(i915_verify_lists(ring->dev));
 
 	/* Retire requests first as we use it above for the early return.
@@ -2676,16 +2828,7 @@ i915_gem_retire_requests_ring(struct intel_engine_cs *ring)
 		if (!i915_gem_request_completed(request, true))
 			break;
 
-		trace_i915_gem_request_retire(request);
-
-		/* We know the GPU must have read the request to have
-		 * sent us the seqno + interrupt, so use the position
-		 * of tail of the request to update the last known position
-		 * of the GPU head.
-		 */
-		request->ringbuf->last_retired_head = request->postfix;
-
-		i915_gem_free_request(request);
+		i915_gem_request_retire(request);
 	}
 
 	/* Move any buffers on the active list that are no longer referenced
@@ -2697,12 +2840,12 @@ i915_gem_retire_requests_ring(struct intel_engine_cs *ring)
 
 		obj = list_first_entry(&ring->active_list,
 				      struct drm_i915_gem_object,
-				      ring_list);
+				      ring_list[ring->id]);
 
-		if (!i915_gem_request_completed(obj->last_read_req, true))
+		if (!list_empty(&obj->last_read_req[ring->id]->list))
 			break;
 
-		i915_gem_object_move_to_inactive(obj);
+		i915_gem_object_retire__read(obj, ring->id);
 	}
 
 	if (unlikely(ring->trace_irq_req &&
@@ -2768,8 +2911,25 @@ i915_gem_idle_work_handler(struct work_struct *work)
 {
 	struct drm_i915_private *dev_priv =
 		container_of(work, typeof(*dev_priv), mm.idle_work.work);
+	struct drm_device *dev = dev_priv->dev;
+	struct intel_engine_cs *ring;
+	int i;
+
+	for_each_ring(ring, dev_priv, i)
+		if (!list_empty(&ring->request_list))
+			return;
+
+	intel_mark_idle(dev);
+
+	if (mutex_trylock(&dev->struct_mutex)) {
+		struct intel_engine_cs *ring;
+		int i;
 
-	intel_mark_idle(dev_priv->dev);
+		for_each_ring(ring, dev_priv, i)
+			i915_gem_batch_pool_fini(&ring->batch_pool);
+
+		mutex_unlock(&dev->struct_mutex);
+	}
 }
 
 /**
@@ -2780,17 +2940,30 @@ i915_gem_idle_work_handler(struct work_struct *work)
 static int
 i915_gem_object_flush_active(struct drm_i915_gem_object *obj)
 {
-	struct intel_engine_cs *ring;
-	int ret;
+	int ret, i;
+
+	if (!obj->active)
+		return 0;
+
+	for (i = 0; i < I915_NUM_RINGS; i++) {
+		struct drm_i915_gem_request *req;
+
+		req = obj->last_read_req[i];
+		if (req == NULL)
+			continue;
 
-	if (obj->active) {
-		ring = i915_gem_request_get_ring(obj->last_read_req);
+		if (list_empty(&req->list))
+			goto retire;
 
-		ret = i915_gem_check_olr(obj->last_read_req);
+		ret = i915_gem_check_olr(req);
 		if (ret)
 			return ret;
 
-		i915_gem_retire_requests_ring(ring);
+		if (i915_gem_request_completed(req, true)) {
+			__i915_gem_request_retire__upto(req);
+retire:
+			i915_gem_object_retire__read(obj, i);
+		}
 	}
 
 	return 0;
@@ -2824,9 +2997,10 @@ i915_gem_wait_ioctl(struct drm_device *dev, void *data, struct drm_file *file)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_i915_gem_wait *args = data;
 	struct drm_i915_gem_object *obj;
-	struct drm_i915_gem_request *req;
+	struct drm_i915_gem_request *req[I915_NUM_RINGS];
 	unsigned reset_counter;
-	int ret = 0;
+	int i, n = 0;
+	int ret;
 
 	if (args->flags != 0)
 		return -EINVAL;
@@ -2846,11 +3020,9 @@ i915_gem_wait_ioctl(struct drm_device *dev, void *data, struct drm_file *file)
 	if (ret)
 		goto out;
 
-	if (!obj->active || !obj->last_read_req)
+	if (!obj->active)
 		goto out;
 
-	req = obj->last_read_req;
-
 	/* Do this after OLR check to make sure we make forward progress polling
 	 * on this IOCTL with a timeout == 0 (like busy ioctl)
 	 */
@@ -2861,15 +3033,23 @@ i915_gem_wait_ioctl(struct drm_device *dev, void *data, struct drm_file *file)
 
 	drm_gem_object_unreference(&obj->base);
 	reset_counter = atomic_read(&dev_priv->gpu_error.reset_counter);
-	i915_gem_request_reference(req);
-	mutex_unlock(&dev->struct_mutex);
 
-	ret = __i915_wait_request(req, reset_counter, true,
-				  args->timeout_ns > 0 ? &args->timeout_ns : NULL,
-				  file->driver_priv);
-	mutex_lock(&dev->struct_mutex);
-	i915_gem_request_unreference(req);
+	for (i = 0; i < I915_NUM_RINGS; i++) {
+		if (obj->last_read_req[i] == NULL)
+			continue;
+
+		req[n++] = i915_gem_request_reference(obj->last_read_req[i]);
+	}
+
 	mutex_unlock(&dev->struct_mutex);
+
+	for (i = 0; i < n; i++) {
+		if (ret == 0)
+			ret = __i915_wait_request(req[i], reset_counter, true,
+						  args->timeout_ns > 0 ? &args->timeout_ns : NULL,
+						  file->driver_priv);
+		i915_gem_request_unreference__unlocked(req[i]);
+	}
 	return ret;
 
 out:
@@ -2878,6 +3058,59 @@ out:
 	return ret;
 }
 
+static int
+__i915_gem_object_sync(struct drm_i915_gem_object *obj,
+		       struct intel_engine_cs *to,
+		       struct drm_i915_gem_request *req)
+{
+	struct intel_engine_cs *from;
+	int ret;
+
+	from = i915_gem_request_get_ring(req);
+	if (to == from)
+		return 0;
+
+	if (i915_gem_request_completed(req, true))
+		return 0;
+
+	ret = i915_gem_check_olr(req);
+	if (ret)
+		return ret;
+
+	if (!i915_semaphore_is_enabled(obj->base.dev)) {
+		struct drm_i915_private *i915 = to_i915(obj->base.dev);
+		ret = __i915_wait_request(req,
+					  atomic_read(&i915->gpu_error.reset_counter),
+					  i915->mm.interruptible,
+					  NULL,
+					  &i915->rps.semaphores);
+		if (ret)
+			return ret;
+
+		i915_gem_object_retire_request(obj, req);
+	} else {
+		int idx = intel_ring_sync_index(from, to);
+		u32 seqno = i915_gem_request_get_seqno(req);
+
+		if (seqno <= from->semaphore.sync_seqno[idx])
+			return 0;
+
+		trace_i915_gem_ring_sync_to(from, to, req);
+		ret = to->semaphore.sync_to(to, from, seqno);
+		if (ret)
+			return ret;
+
+		/* We use last_read_req because sync_to()
+		 * might have just caused seqno wrap under
+		 * the radar.
+		 */
+		from->semaphore.sync_seqno[idx] =
+			i915_gem_request_get_seqno(obj->last_read_req[from->id]);
+	}
+
+	return 0;
+}
+
 /**
  * i915_gem_object_sync - sync an object to a ring.
  *
@@ -2886,7 +3119,17 @@ out:
  *
  * This code is meant to abstract object synchronization with the GPU.
  * Calling with NULL implies synchronizing the object with the CPU
- * rather than a particular GPU ring.
+ * rather than a particular GPU ring. Conceptually we serialise writes
+ * between engines inside the GPU. We only allow on engine to write
+ * into a buffer at any time, but multiple readers. To ensure each has
+ * a coherent view of memory, we must:
+ *
+ * - If there is an outstanding write request to the object, the new
+ *   request must wait for it to complete (either CPU or in hw, requests
+ *   on the same ring will be naturally ordered).
+ *
+ * - If we are a write request (pending_write_domain is set), the new
+ *   request must wait for outstanding read requests to complete.
  *
  * Returns 0 if successful, else propagates up the lower layer error.
  */
@@ -2894,41 +3137,32 @@ int
 i915_gem_object_sync(struct drm_i915_gem_object *obj,
 		     struct intel_engine_cs *to)
 {
-	struct intel_engine_cs *from;
-	u32 seqno;
-	int ret, idx;
-
-	from = i915_gem_request_get_ring(obj->last_read_req);
-
-	if (from == NULL || to == from)
-		return 0;
-
-	if (to == NULL || !i915_semaphore_is_enabled(obj->base.dev))
-		return i915_gem_object_wait_rendering(obj, false);
+	const bool readonly = obj->base.pending_write_domain == 0;
+	struct drm_i915_gem_request *req[I915_NUM_RINGS];
+	int ret, i, n;
 
-	idx = intel_ring_sync_index(from, to);
-
-	seqno = i915_gem_request_get_seqno(obj->last_read_req);
-	/* Optimization: Avoid semaphore sync when we are sure we already
-	 * waited for an object with higher seqno */
-	if (seqno <= from->semaphore.sync_seqno[idx])
+	if (!obj->active)
 		return 0;
 
-	ret = i915_gem_check_olr(obj->last_read_req);
-	if (ret)
-		return ret;
+	if (to == NULL)
+		return i915_gem_object_wait_rendering(obj, readonly);
 
-	trace_i915_gem_ring_sync_to(from, to, obj->last_read_req);
-	ret = to->semaphore.sync_to(to, from, seqno);
-	if (!ret)
-		/* We use last_read_req because sync_to()
-		 * might have just caused seqno wrap under
-		 * the radar.
-		 */
-		from->semaphore.sync_seqno[idx] =
-				i915_gem_request_get_seqno(obj->last_read_req);
+	n = 0;
+	if (readonly) {
+		if (obj->last_write_req)
+			req[n++] = obj->last_write_req;
+	} else {
+		for (i = 0; i < I915_NUM_RINGS; i++)
+			if (obj->last_read_req[i])
+				req[n++] = obj->last_read_req[i];
+	}
+	for (i = 0; i < n; i++) {
+		ret = __i915_gem_object_sync(obj, to, req[i]);
+		if (ret)
+			return ret;
+	}
 
-	return ret;
+	return 0;
 }
 
 static void i915_gem_object_finish_gtt(struct drm_i915_gem_object *obj)
@@ -2974,7 +3208,7 @@ int i915_vma_unbind(struct i915_vma *vma)
 
 	BUG_ON(obj->pages == NULL);
 
-	ret = i915_gem_object_finish_gpu(obj);
+	ret = i915_gem_object_wait_rendering(obj, false);
 	if (ret)
 		return ret;
 	/* Continue on if we fail due to EIO, the GPU is hung so we
@@ -2994,7 +3228,8 @@ int i915_vma_unbind(struct i915_vma *vma)
 
 	trace_i915_vma_unbind(vma);
 
-	vma->unbind_vma(vma);
+	vma->vm->unbind_vma(vma);
+	vma->bound = 0;
 
 	list_del_init(&vma->mm_list);
 	if (i915_is_ggtt(vma->vm)) {
@@ -3013,10 +3248,6 @@ int i915_vma_unbind(struct i915_vma *vma)
 	/* Since the unbound list is global, only move to that list if
 	 * no more VMAs exist. */
 	if (list_empty(&obj->vma_list)) {
-		/* Throw away the active reference before
-		 * moving to the unbound list. */
-		i915_gem_object_retire(obj);
-
 		i915_gem_gtt_finish_object(obj);
 		list_move_tail(&obj->global_list, &dev_priv->mm.unbound_list);
 	}
@@ -3049,6 +3280,7 @@ int i915_gpu_idle(struct drm_device *dev)
 			return ret;
 	}
 
+	WARN_ON(i915_verify_lists(dev));
 	return 0;
 }
 
@@ -3423,7 +3655,8 @@ static bool i915_gem_valid_gtt_space(struct i915_vma *vma,
 }
 
 /**
- * Finds free space in the GTT aperture and binds the object there.
+ * Finds free space in the GTT aperture and binds the object or a view of it
+ * there.
  */
 static struct i915_vma *
 i915_gem_object_bind_to_vm(struct drm_i915_gem_object *obj,
@@ -3442,36 +3675,60 @@ i915_gem_object_bind_to_vm(struct drm_i915_gem_object *obj,
 	struct i915_vma *vma;
 	int ret;
 
-	if(WARN_ON(i915_is_ggtt(vm) != !!ggtt_view))
-		return ERR_PTR(-EINVAL);
+	if (i915_is_ggtt(vm)) {
+		u32 view_size;
+
+		if (WARN_ON(!ggtt_view))
+			return ERR_PTR(-EINVAL);
 
-	fence_size = i915_gem_get_gtt_size(dev,
-					   obj->base.size,
-					   obj->tiling_mode);
-	fence_alignment = i915_gem_get_gtt_alignment(dev,
-						     obj->base.size,
-						     obj->tiling_mode, true);
-	unfenced_alignment =
-		i915_gem_get_gtt_alignment(dev,
-					   obj->base.size,
-					   obj->tiling_mode, false);
+		view_size = i915_ggtt_view_size(obj, ggtt_view);
+
+		fence_size = i915_gem_get_gtt_size(dev,
+						   view_size,
+						   obj->tiling_mode);
+		fence_alignment = i915_gem_get_gtt_alignment(dev,
+							     view_size,
+							     obj->tiling_mode,
+							     true);
+		unfenced_alignment = i915_gem_get_gtt_alignment(dev,
+								view_size,
+								obj->tiling_mode,
+								false);
+		size = flags & PIN_MAPPABLE ? fence_size : view_size;
+	} else {
+		fence_size = i915_gem_get_gtt_size(dev,
+						   obj->base.size,
+						   obj->tiling_mode);
+		fence_alignment = i915_gem_get_gtt_alignment(dev,
+							     obj->base.size,
+							     obj->tiling_mode,
+							     true);
+		unfenced_alignment =
+			i915_gem_get_gtt_alignment(dev,
+						   obj->base.size,
+						   obj->tiling_mode,
+						   false);
+		size = flags & PIN_MAPPABLE ? fence_size : obj->base.size;
+	}
 
 	if (alignment == 0)
 		alignment = flags & PIN_MAPPABLE ? fence_alignment :
 						unfenced_alignment;
 	if (flags & PIN_MAPPABLE && alignment & (fence_alignment - 1)) {
-		DRM_DEBUG("Invalid object alignment requested %u\n", alignment);
+		DRM_DEBUG("Invalid object (view type=%u) alignment requested %u\n",
+			  ggtt_view ? ggtt_view->type : 0,
+			  alignment);
 		return ERR_PTR(-EINVAL);
 	}
 
-	size = flags & PIN_MAPPABLE ? fence_size : obj->base.size;
-
-	/* If the object is bigger than the entire aperture, reject it early
-	 * before evicting everything in a vain attempt to find space.
+	/* If binding the object/GGTT view requires more space than the entire
+	 * aperture has, reject it early before evicting everything in a vain
+	 * attempt to find space.
 	 */
-	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,
+	if (size > end) {
+		DRM_DEBUG("Attempting to bind an object (view type=%u) larger than the aperture: size=%u > %s aperture=%lu\n",
+			  ggtt_view ? ggtt_view->type : 0,
+			  size,
 			  flags & PIN_MAPPABLE ? "mappable" : "total",
 			  end);
 		return ERR_PTR(-E2BIG);
@@ -3515,20 +3772,8 @@ search_free:
 	if (ret)
 		goto err_remove_node;
 
-	/*  allocate before insert / bind */
-	if (vma->vm->allocate_va_range) {
-		trace_i915_va_alloc(vma->vm, vma->node.start, vma->node.size,
-				VM_TO_TRACE_NAME(vma->vm));
-		ret = vma->vm->allocate_va_range(vma->vm,
-						vma->node.start,
-						vma->node.size);
-		if (ret)
-			goto err_remove_node;
-	}
-
 	trace_i915_vma_bind(vma, flags);
-	ret = i915_vma_bind(vma, obj->cache_level,
-			    flags & PIN_GLOBAL ? GLOBAL_BIND : 0);
+	ret = i915_vma_bind(vma, obj->cache_level, flags);
 	if (ret)
 		goto err_finish_gtt;
 
@@ -3658,8 +3903,6 @@ i915_gem_object_set_to_gtt_domain(struct drm_i915_gem_object *obj, bool write)
 	if (ret)
 		return ret;
 
-	i915_gem_object_retire(obj);
-
 	/* Flush and acquire obj->pages so that we are coherent through
 	 * direct access in memory with previous cached writes through
 	 * shmemfs and that our cache domain tracking remains valid.
@@ -3735,7 +3978,7 @@ int i915_gem_object_set_cache_level(struct drm_i915_gem_object *obj,
 	}
 
 	if (i915_gem_obj_bound_any(obj)) {
-		ret = i915_gem_object_finish_gpu(obj);
+		ret = i915_gem_object_wait_rendering(obj, false);
 		if (ret)
 			return ret;
 
@@ -3754,7 +3997,7 @@ int i915_gem_object_set_cache_level(struct drm_i915_gem_object *obj,
 		list_for_each_entry(vma, &obj->vma_list, vma_link)
 			if (drm_mm_node_allocated(&vma->node)) {
 				ret = i915_vma_bind(vma, cache_level,
-						    vma->bound & GLOBAL_BIND);
+						    PIN_UPDATE);
 				if (ret)
 					return ret;
 			}
@@ -3779,17 +4022,10 @@ int i915_gem_get_caching_ioctl(struct drm_device *dev, void *data,
 {
 	struct drm_i915_gem_caching *args = data;
 	struct drm_i915_gem_object *obj;
-	int ret;
-
-	ret = i915_mutex_lock_interruptible(dev);
-	if (ret)
-		return ret;
 
 	obj = to_intel_bo(drm_gem_object_lookup(dev, file, args->handle));
-	if (&obj->base == NULL) {
-		ret = -ENOENT;
-		goto unlock;
-	}
+	if (&obj->base == NULL)
+		return -ENOENT;
 
 	switch (obj->cache_level) {
 	case I915_CACHE_LLC:
@@ -3806,10 +4042,8 @@ int i915_gem_get_caching_ioctl(struct drm_device *dev, void *data,
 		break;
 	}
 
-	drm_gem_object_unreference(&obj->base);
-unlock:
-	mutex_unlock(&dev->struct_mutex);
-	return ret;
+	drm_gem_object_unreference_unlocked(&obj->base);
+	return 0;
 }
 
 int i915_gem_set_caching_ioctl(struct drm_device *dev, void *data,
@@ -3852,24 +4086,6 @@ unlock:
 	return ret;
 }
 
-static bool is_pin_display(struct drm_i915_gem_object *obj)
-{
-	struct i915_vma *vma;
-
-	vma = i915_gem_obj_to_ggtt(obj);
-	if (!vma)
-		return false;
-
-	/* There are 2 sources that pin objects:
-	 *   1. The display engine (scanouts, sprites, cursors);
-	 *   2. Reservations for execbuffer;
-	 *
-	 * We can ignore reservations as we hold the struct_mutex and
-	 * are only called outside of the reservation path.
-	 */
-	return vma->pin_count;
-}
-
 /*
  * Prepare buffer for display plane (scanout, cursors, etc).
  * Can be called from an uninterruptible phase (modesetting) and allows
@@ -3882,20 +4098,16 @@ i915_gem_object_pin_to_display_plane(struct drm_i915_gem_object *obj,
 				     const struct i915_ggtt_view *view)
 {
 	u32 old_read_domains, old_write_domain;
-	bool was_pin_display;
 	int ret;
 
-	if (pipelined != i915_gem_request_get_ring(obj->last_read_req)) {
-		ret = i915_gem_object_sync(obj, pipelined);
-		if (ret)
-			return ret;
-	}
+	ret = i915_gem_object_sync(obj, pipelined);
+	if (ret)
+		return ret;
 
 	/* Mark the pin_display early so that we account for the
 	 * display coherency whilst setting up the cache domains.
 	 */
-	was_pin_display = obj->pin_display;
-	obj->pin_display = true;
+	obj->pin_display++;
 
 	/* The display engine is not coherent with the LLC cache on gen6.  As
 	 * a result, we make sure that the pinning that is about to occur is
@@ -3939,8 +4151,7 @@ i915_gem_object_pin_to_display_plane(struct drm_i915_gem_object *obj,
 	return 0;
 
 err_unpin_display:
-	WARN_ON(was_pin_display != is_pin_display(obj));
-	obj->pin_display = was_pin_display;
+	obj->pin_display--;
 	return ret;
 }
 
@@ -3948,26 +4159,12 @@ void
 i915_gem_object_unpin_from_display_plane(struct drm_i915_gem_object *obj,
 					 const struct i915_ggtt_view *view)
 {
-	i915_gem_object_ggtt_unpin_view(obj, view);
-
-	obj->pin_display = is_pin_display(obj);
-}
-
-int
-i915_gem_object_finish_gpu(struct drm_i915_gem_object *obj)
-{
-	int ret;
-
-	if ((obj->base.read_domains & I915_GEM_GPU_DOMAINS) == 0)
-		return 0;
+	if (WARN_ON(obj->pin_display == 0))
+		return;
 
-	ret = i915_gem_object_wait_rendering(obj, false);
-	if (ret)
-		return ret;
+	i915_gem_object_ggtt_unpin_view(obj, view);
 
-	/* Ensure that we invalidate the GPU's caches and TLBs. */
-	obj->base.read_domains &= ~I915_GEM_GPU_DOMAINS;
-	return 0;
+	obj->pin_display--;
 }
 
 /**
@@ -3989,7 +4186,6 @@ i915_gem_object_set_to_cpu_domain(struct drm_i915_gem_object *obj, bool write)
 	if (ret)
 		return ret;
 
-	i915_gem_object_retire(obj);
 	i915_gem_object_flush_gtt_write_domain(obj);
 
 	old_write_domain = obj->base.write_domain;
@@ -4040,7 +4236,7 @@ i915_gem_ring_throttle(struct drm_device *dev, struct drm_file *file)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	struct drm_i915_file_private *file_priv = file->driver_priv;
-	unsigned long recent_enough = jiffies - msecs_to_jiffies(20);
+	unsigned long recent_enough = jiffies - DRM_I915_THROTTLE_JIFFIES;
 	struct drm_i915_gem_request *request, *target = NULL;
 	unsigned reset_counter;
 	int ret;
@@ -4072,9 +4268,7 @@ i915_gem_ring_throttle(struct drm_device *dev, struct drm_file *file)
 	if (ret == 0)
 		queue_delayed_work(dev_priv->wq, &dev_priv->mm.retire_work, 0);
 
-	mutex_lock(&dev->struct_mutex);
-	i915_gem_request_unreference(target);
-	mutex_unlock(&dev->struct_mutex);
+	i915_gem_request_unreference__unlocked(target);
 
 	return ret;
 }
@@ -4155,23 +4349,18 @@ i915_gem_object_do_pin(struct drm_i915_gem_object *obj,
 
 	bound = vma ? vma->bound : 0;
 	if (vma == NULL || !drm_mm_node_allocated(&vma->node)) {
-		/* In true PPGTT, bind has possibly changed PDEs, which
-		 * means we must do a context switch before the GPU can
-		 * accurately read some of the VMAs.
-		 */
 		vma = i915_gem_object_bind_to_vm(obj, vm, ggtt_view, alignment,
 						 flags);
 		if (IS_ERR(vma))
 			return PTR_ERR(vma);
-	}
-
-	if (flags & PIN_GLOBAL && !(vma->bound & GLOBAL_BIND)) {
-		ret = i915_vma_bind(vma, obj->cache_level, GLOBAL_BIND);
+	} else {
+		ret = i915_vma_bind(vma, obj->cache_level, flags);
 		if (ret)
 			return ret;
 	}
 
-	if ((bound ^ vma->bound) & GLOBAL_BIND) {
+	if (ggtt_view && ggtt_view->type == I915_GGTT_VIEW_NORMAL &&
+	    (bound ^ vma->bound) & GLOBAL_BIND) {
 		bool mappable, fenceable;
 		u32 fence_size, fence_alignment;
 
@@ -4190,14 +4379,11 @@ i915_gem_object_do_pin(struct drm_i915_gem_object *obj,
 			    dev_priv->gtt.mappable_end);
 
 		obj->map_and_fenceable = mappable && fenceable;
-	}
 
-	WARN_ON(flags & PIN_MAPPABLE && !obj->map_and_fenceable);
+		WARN_ON(flags & PIN_MAPPABLE && !obj->map_and_fenceable);
+	}
 
 	vma->pin_count++;
-	if (flags & PIN_MAPPABLE)
-		obj->pin_mappable |= true;
-
 	return 0;
 }
 
@@ -4235,8 +4421,7 @@ i915_gem_object_ggtt_unpin_view(struct drm_i915_gem_object *obj,
 	WARN_ON(vma->pin_count == 0);
 	WARN_ON(!i915_gem_obj_ggtt_bound_view(obj, view));
 
-	if (--vma->pin_count == 0 && view->type == I915_GGTT_VIEW_NORMAL)
-		obj->pin_mappable = false;
+	--vma->pin_count;
 }
 
 bool
@@ -4289,15 +4474,15 @@ i915_gem_busy_ioctl(struct drm_device *dev, void *data,
 	 * necessary flushes here.
 	 */
 	ret = i915_gem_object_flush_active(obj);
+	if (ret)
+		goto unref;
 
-	args->busy = obj->active;
-	if (obj->last_read_req) {
-		struct intel_engine_cs *ring;
-		BUILD_BUG_ON(I915_NUM_RINGS > 16);
-		ring = i915_gem_request_get_ring(obj->last_read_req);
-		args->busy |= intel_ring_flag(ring) << 16;
-	}
+	BUILD_BUG_ON(I915_NUM_RINGS > 16);
+	args->busy = obj->active << 16;
+	if (obj->last_write_req)
+		args->busy |= obj->last_write_req->ring->id;
 
+unref:
 	drm_gem_object_unreference(&obj->base);
 unlock:
 	mutex_unlock(&dev->struct_mutex);
@@ -4371,11 +4556,14 @@ unlock:
 void i915_gem_object_init(struct drm_i915_gem_object *obj,
 			  const struct drm_i915_gem_object_ops *ops)
 {
+	int i;
+
 	INIT_LIST_HEAD(&obj->global_list);
-	INIT_LIST_HEAD(&obj->ring_list);
+	for (i = 0; i < I915_NUM_RINGS; i++)
+		INIT_LIST_HEAD(&obj->ring_list[i]);
 	INIT_LIST_HEAD(&obj->obj_exec_link);
 	INIT_LIST_HEAD(&obj->vma_list);
-	INIT_LIST_HEAD(&obj->batch_pool_list);
+	INIT_LIST_HEAD(&obj->batch_pool_link);
 
 	obj->ops = ops;
 
@@ -4577,7 +4765,7 @@ void i915_gem_vma_destroy(struct i915_vma *vma)
 
 	list_del(&vma->vma_link);
 
-	kfree(vma);
+	kmem_cache_free(to_i915(vma->obj->base.dev)->vmas, vma);
 }
 
 static void
@@ -4864,12 +5052,12 @@ int i915_gem_init(struct drm_device *dev)
 	}
 
 	if (!i915.enable_execlists) {
-		dev_priv->gt.do_execbuf = i915_gem_ringbuffer_submission;
+		dev_priv->gt.execbuf_submit = i915_gem_ringbuffer_submission;
 		dev_priv->gt.init_rings = i915_gem_init_rings;
 		dev_priv->gt.cleanup_ring = intel_cleanup_ring_buffer;
 		dev_priv->gt.stop_ring = intel_stop_ring_buffer;
 	} else {
-		dev_priv->gt.do_execbuf = intel_execlists_submission;
+		dev_priv->gt.execbuf_submit = intel_execlists_submission;
 		dev_priv->gt.init_rings = intel_logical_rings_init;
 		dev_priv->gt.cleanup_ring = intel_logical_ring_cleanup;
 		dev_priv->gt.stop_ring = intel_logical_ring_stop;
@@ -4951,11 +5139,21 @@ i915_gem_load(struct drm_device *dev)
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	int i;
 
-	dev_priv->slab =
+	dev_priv->objects =
 		kmem_cache_create("i915_gem_object",
 				  sizeof(struct drm_i915_gem_object), 0,
 				  SLAB_HWCACHE_ALIGN,
 				  NULL);
+	dev_priv->vmas =
+		kmem_cache_create("i915_gem_vma",
+				  sizeof(struct i915_vma), 0,
+				  SLAB_HWCACHE_ALIGN,
+				  NULL);
+	dev_priv->requests =
+		kmem_cache_create("i915_gem_request",
+				  sizeof(struct drm_i915_gem_request), 0,
+				  SLAB_HWCACHE_ALIGN,
+				  NULL);
 
 	INIT_LIST_HEAD(&dev_priv->vm_list);
 	i915_init_vm(dev_priv, &dev_priv->gtt.base);
@@ -4998,8 +5196,6 @@ i915_gem_load(struct drm_device *dev)
 
 	i915_gem_shrinker_init(dev_priv);
 
-	i915_gem_batch_pool_init(dev, &dev_priv->mm.batch_pool);
-
 	mutex_init(&dev_priv->fb_tracking.lock);
 }
 
@@ -5007,8 +5203,6 @@ void i915_gem_release(struct drm_device *dev, struct drm_file *file)
 {
 	struct drm_i915_file_private *file_priv = file->driver_priv;
 
-	cancel_delayed_work_sync(&file_priv->mm.idle_work);
-
 	/* Clean up our request list when the client is going away, so that
 	 * later retire_requests won't dereference our soon-to-be-gone
 	 * file_priv.
@@ -5024,15 +5218,12 @@ void i915_gem_release(struct drm_device *dev, struct drm_file *file)
 		request->file_priv = NULL;
 	}
 	spin_unlock(&file_priv->mm.lock);
-}
 
-static void
-i915_gem_file_idle_work_handler(struct work_struct *work)
-{
-	struct drm_i915_file_private *file_priv =
-		container_of(work, typeof(*file_priv), mm.idle_work.work);
-
-	atomic_set(&file_priv->rps_wait_boost, false);
+	if (!list_empty(&file_priv->rps.link)) {
+		spin_lock(&to_i915(dev)->rps.client_lock);
+		list_del(&file_priv->rps.link);
+		spin_unlock(&to_i915(dev)->rps.client_lock);
+	}
 }
 
 int i915_gem_open(struct drm_device *dev, struct drm_file *file)
@@ -5049,11 +5240,10 @@ int i915_gem_open(struct drm_device *dev, struct drm_file *file)
 	file->driver_priv = file_priv;
 	file_priv->dev_priv = dev->dev_private;
 	file_priv->file = file;
+	INIT_LIST_HEAD(&file_priv->rps.link);
 
 	spin_lock_init(&file_priv->mm.lock);
 	INIT_LIST_HEAD(&file_priv->mm.request_list);
-	INIT_DELAYED_WORK(&file_priv->mm.idle_work,
-			  i915_gem_file_idle_work_handler);
 
 	ret = i915_gem_context_open(dev, file);
 	if (ret)
@@ -5123,7 +5313,7 @@ i915_gem_obj_ggtt_offset_view(struct drm_i915_gem_object *o,
 		    i915_ggtt_view_equal(&vma->ggtt_view, view))
 			return vma->node.start;
 
-	WARN(1, "global vma for this object not found.\n");
+	WARN(1, "global vma for this object not found. (view=%u)\n", view->type);
 	return -1;
 }
 
@@ -5192,13 +5382,10 @@ unsigned long i915_gem_obj_size(struct drm_i915_gem_object *o,
 bool i915_gem_obj_is_pinned(struct drm_i915_gem_object *obj)
 {
 	struct i915_vma *vma;
-	list_for_each_entry(vma, &obj->vma_list, vma_link) {
-		if (i915_is_ggtt(vma->vm) &&
-		    vma->ggtt_view.type != I915_GGTT_VIEW_NORMAL)
-			continue;
+	list_for_each_entry(vma, &obj->vma_list, vma_link)
 		if (vma->pin_count > 0)
 			return true;
-	}
+
 	return false;
 }