Merge tag 'drm-intel-gt-next-2021-01-21-1' of git://anongit.freedesktop.org/drm/drm-intel into drm-next

Cross-subsystem Changes:

- Includes gvt-gt-next-2021-01-18 + header check fix for GVT

Driver Changes:

- Fix for #2955: Clear potentially malicious register state before
  executing clear residuals security mitigation (Chris)
- Fixes that lead to marking per-engine-reset as supported on Gen7
  (Chris)
- Remove per-client stats from debugfs/i915_gem_objects) (Tvrtko, Chris)
- Add arbitration check before semaphore wait (Chris)
- Apply interactive priority to explicit flip fences (Chris)
- Make GEM errors non-fatal by default to help capturing logs during
  development (Chris)
- Fix object page offset within a region in error capture (CQ, Matt A)
- Close race between enable_breadcrumbs and cancel_breadcrumbs (Chris)
- Almagamate clflushes on suspend/freeze to speed up S/R (Chris)
- Protect used framebuffers from casual eviction (Chris)

- Fix the sgt.pfn sanity check (Kui, Matt A)
- Reduce locking around i915_request.lock and ctx->engines_mutex (Chris)
- Simplify tracking for engine->fw_active and stats.active (Chris)
- Constrain pool objects by mapping type (Chris, Matt A)
- Use shrinkable status for unknown swizzle quirks (Chris)
- Do not suspend bonded requests if one hangs (Chris)
- Restore "Skip over completed active execlists" optimization (Chris)

- Move stolen node into GEM object union (Chris)
. Split gem_create into own file (Matt A)
- Convert object_create into object_init in LMEM region code (Matt A)
- Reduce test_and_set_bit to set_bit in i915_request_submit() (Chris)
- Mark up protected uses of 'i915_request_completed' (Chris)
- Remove extraneous inline modifiers (Chris)
- Add function to define defaults for GuC/HuC enable (John)

- Improve code locality by moving closer to single user (Matt A, Chris)
- Compiler warning fixes (Matt A, Chris)
- Selftest / CI improvements (Chris)

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210121150747.GA58732@jlahtine-mobl.ger.corp.intel.com

drivers/gpu/drm/i915/Kconfig.debug

@@ -31,10 +31,13 @@ config DRM_I915_DEBUG
 	select DRM_DEBUG_SELFTEST
 	select DMABUF_SELFTESTS
 	select SW_SYNC # signaling validation framework (igt/syncobj*)
+	select DRM_I915_WERROR
+	select DRM_I915_DEBUG_GEM
+	select DRM_I915_DEBUG_GEM_ONCE
+	select DRM_I915_DEBUG_MMIO
+	select DRM_I915_DEBUG_RUNTIME_PM
 	select DRM_I915_SW_FENCE_DEBUG_OBJECTS
 	select DRM_I915_SELFTEST
-	select DRM_I915_DEBUG_RUNTIME_PM
-	select DRM_I915_DEBUG_MMIO
 	default n
 	help
 	  Choose this option to turn on extra driver debugging that may affect
@@ -69,6 +72,21 @@ config DRM_I915_DEBUG_GEM
 
 	  If in doubt, say "N".
 
+config DRM_I915_DEBUG_GEM_ONCE
+	bool "Make a GEM debug failure fatal"
+	default n
+	depends on DRM_I915_DEBUG_GEM
+	help
+	  During development, we often only want the very first failure
+	  as that would otherwise be lost in the deluge of subsequent
+	  failures. However, more casual testers may not want to trigger
+	  a hard BUG_ON and hope that the system remains sufficiently usable
+	  to capture a bug report in situ.
+
+	  Recommended for driver developers only.
+
+	  If in doubt, say "N".
+
 config DRM_I915_ERRLOG_GEM
 	bool "Insert extra logging (very verbose) for common GEM errors"
 	default n

drivers/gpu/drm/i915/Makefile

@@ -136,6 +136,7 @@ gem-y += \
 	gem/i915_gem_clflush.o \
 	gem/i915_gem_client_blt.o \
 	gem/i915_gem_context.o \
+	gem/i915_gem_create.o \
 	gem/i915_gem_dmabuf.o \
 	gem/i915_gem_domain.o \
 	gem/i915_gem_execbuffer.o \

drivers/gpu/drm/i915/display/intel_display.c

@@ -2247,7 +2247,7 @@ intel_pin_and_fence_fb_obj(struct drm_framebuffer *fb,
 		 */
 		ret = i915_vma_pin_fence(vma);
 		if (ret != 0 && INTEL_GEN(dev_priv) < 4) {
-			i915_gem_object_unpin_from_display_plane(vma);
+			i915_vma_unpin(vma);
 			vma = ERR_PTR(ret);
 			goto err;
 		}
@@ -2265,12 +2265,9 @@ err:
 
 void intel_unpin_fb_vma(struct i915_vma *vma, unsigned long flags)
 {
-	i915_gem_object_lock(vma->obj, NULL);
 	if (flags & PLANE_HAS_FENCE)
 		i915_vma_unpin_fence(vma);
-	i915_gem_object_unpin_from_display_plane(vma);
-	i915_gem_object_unlock(vma->obj);
-
+	i915_vma_unpin(vma);
 	i915_vma_put(vma);
 }
 
@@ -15631,15 +15628,6 @@ void intel_plane_unpin_fb(struct intel_plane_state *old_plane_state)
 		intel_unpin_fb_vma(vma, old_plane_state->flags);
 }
 
-static void fb_obj_bump_render_priority(struct drm_i915_gem_object *obj)
-{
-	struct i915_sched_attr attr = {
-		.priority = I915_USER_PRIORITY(I915_PRIORITY_DISPLAY),
-	};
-
-	i915_gem_object_wait_priority(obj, 0, &attr);
-}
-
 /**
  * intel_prepare_plane_fb - Prepare fb for usage on plane
  * @_plane: drm plane to prepare for
@@ -15656,6 +15644,9 @@ int
 intel_prepare_plane_fb(struct drm_plane *_plane,
 		       struct drm_plane_state *_new_plane_state)
 {
+	struct i915_sched_attr attr = {
+		.priority = I915_USER_PRIORITY(I915_PRIORITY_DISPLAY),
+	};
 	struct intel_plane *plane = to_intel_plane(_plane);
 	struct intel_plane_state *new_plane_state =
 		to_intel_plane_state(_new_plane_state);
@@ -15695,6 +15686,8 @@ intel_prepare_plane_fb(struct drm_plane *_plane,
 	}
 
 	if (new_plane_state->uapi.fence) { /* explicit fencing */
+		i915_gem_fence_wait_priority(new_plane_state->uapi.fence,
+					     &attr);
 		ret = i915_sw_fence_await_dma_fence(&state->commit_ready,
 						    new_plane_state->uapi.fence,
 						    i915_fence_timeout(dev_priv),
@@ -15716,7 +15709,7 @@ intel_prepare_plane_fb(struct drm_plane *_plane,
 	if (ret)
 		return ret;
 
-	fb_obj_bump_render_priority(obj);
+	i915_gem_object_wait_priority(obj, 0, &attr);
 	i915_gem_object_flush_frontbuffer(obj, ORIGIN_DIRTYFB);
 
 	if (!new_plane_state->uapi.fence) { /* implicit fencing */

drivers/gpu/drm/i915/display/intel_fbdev.c

@@ -256,7 +256,7 @@ static int intelfb_create(struct drm_fb_helper *helper,
 	 * If the object is stolen however, it will be full of whatever
 	 * garbage was left in there.
 	 */
-	if (vma->obj->stolen && !prealloc)
+	if (!i915_gem_object_is_shmem(vma->obj) && !prealloc)
 		memset_io(info->screen_base, 0, info->screen_size);
 
 	/* Use default scratch pixmap (info->pixmap.flags = FB_PIXMAP_SYSTEM) */
@@ -595,7 +595,7 @@ void intel_fbdev_set_suspend(struct drm_device *dev, int state, bool synchronous
 	 * full of whatever garbage was left in there.
 	 */
 	if (state == FBINFO_STATE_RUNNING &&
-	    intel_fb_obj(&ifbdev->fb->base)->stolen)
+	    !i915_gem_object_is_shmem(intel_fb_obj(&ifbdev->fb->base)))
 		memset_io(info->screen_base, 0, info->screen_size);
 
 	drm_fb_helper_set_suspend(&ifbdev->helper, state);

drivers/gpu/drm/i915/display/intel_frontbuffer.c

@@ -225,8 +225,10 @@ static void frontbuffer_release(struct kref *ref)
 	struct i915_vma *vma;
 
 	spin_lock(&obj->vma.lock);
-	for_each_ggtt_vma(vma, obj)
+	for_each_ggtt_vma(vma, obj) {
+		i915_vma_clear_scanout(vma);
 		vma->display_alignment = I915_GTT_MIN_ALIGNMENT;
+	}
 	spin_unlock(&obj->vma.lock);
 
 	RCU_INIT_POINTER(obj->frontbuffer, NULL);

drivers/gpu/drm/i915/display/intel_overlay.c

@@ -360,7 +360,7 @@ static void intel_overlay_release_old_vma(struct intel_overlay *overlay)
 	intel_frontbuffer_flip_complete(overlay->i915,
 					INTEL_FRONTBUFFER_OVERLAY(overlay->crtc->pipe));
 
-	i915_gem_object_unpin_from_display_plane(vma);
+	i915_vma_unpin(vma);
 	i915_vma_put(vma);
 }
 
@@ -861,7 +861,7 @@ static int intel_overlay_do_put_image(struct intel_overlay *overlay,
 	return 0;
 
 out_unpin:
-	i915_gem_object_unpin_from_display_plane(vma);
+	i915_vma_unpin(vma);
 out_pin_section:
 	atomic_dec(&dev_priv->gpu_error.pending_fb_pin);
 

drivers/gpu/drm/i915/gem/i915_gem_context.c

@@ -408,7 +408,7 @@ __active_engine(struct i915_request *rq, struct intel_engine_cs **active)
 	}
 
 	if (i915_request_is_active(rq)) {
-		if (!i915_request_completed(rq))
+		if (!__i915_request_is_complete(rq))
 			*active = locked;
 		ret = true;
 	}
@@ -717,7 +717,8 @@ err_free:
 }
 
 static inline struct i915_gem_engines *
-__context_engines_await(const struct i915_gem_context *ctx)
+__context_engines_await(const struct i915_gem_context *ctx,
+			bool *user_engines)
 {
 	struct i915_gem_engines *engines;
 
@@ -726,6 +727,10 @@ __context_engines_await(const struct i915_gem_context *ctx)
 		engines = rcu_dereference(ctx->engines);
 		GEM_BUG_ON(!engines);
 
+		if (user_engines)
+			*user_engines = i915_gem_context_user_engines(ctx);
+
+		/* successful await => strong mb */
 		if (unlikely(!i915_sw_fence_await(&engines->fence)))
 			continue;
 
@@ -749,7 +754,7 @@ context_apply_all(struct i915_gem_context *ctx,
 	struct intel_context *ce;
 	int err = 0;
 
-	e = __context_engines_await(ctx);
+	e = __context_engines_await(ctx, NULL);
 	for_each_gem_engine(ce, e, it) {
 		err = fn(ce, data);
 		if (err)
@@ -1075,7 +1080,7 @@ static int context_barrier_task(struct i915_gem_context *ctx,
 		return err;
 	}
 
-	e = __context_engines_await(ctx);
+	e = __context_engines_await(ctx, NULL);
 	if (!e) {
 		i915_active_release(&cb->base);
 		return -ENOENT;
@@ -1838,27 +1843,6 @@ replace:
 	return 0;
 }
 
-static struct i915_gem_engines *
-__copy_engines(struct i915_gem_engines *e)
-{
-	struct i915_gem_engines *copy;
-	unsigned int n;
-
-	copy = alloc_engines(e->num_engines);
-	if (!copy)
-		return ERR_PTR(-ENOMEM);
-
-	for (n = 0; n < e->num_engines; n++) {
-		if (e->engines[n])
-			copy->engines[n] = intel_context_get(e->engines[n]);
-		else
-			copy->engines[n] = NULL;
-	}
-	copy->num_engines = n;
-
-	return copy;
-}
-
 static int
 get_engines(struct i915_gem_context *ctx,
 	    struct drm_i915_gem_context_param *args)
@@ -1866,19 +1850,17 @@ get_engines(struct i915_gem_context *ctx,
 	struct i915_context_param_engines __user *user;
 	struct i915_gem_engines *e;
 	size_t n, count, size;
+	bool user_engines;
 	int err = 0;
 
-	err = mutex_lock_interruptible(&ctx->engines_mutex);
-	if (err)
-		return err;
+	e = __context_engines_await(ctx, &user_engines);
+	if (!e)
+		return -ENOENT;
 
-	e = NULL;
-	if (i915_gem_context_user_engines(ctx))
-		e = __copy_engines(i915_gem_context_engines(ctx));
-	mutex_unlock(&ctx->engines_mutex);
-	if (IS_ERR_OR_NULL(e)) {
+	if (!user_engines) {
+		i915_sw_fence_complete(&e->fence);
 		args->size = 0;
-		return PTR_ERR_OR_ZERO(e);
+		return 0;
 	}
 
 	count = e->num_engines;
@@ -1929,7 +1911,7 @@ get_engines(struct i915_gem_context *ctx,
 	args->size = size;
 
 err_free:
-	free_engines(e);
+	i915_sw_fence_complete(&e->fence);
 	return err;
 }
 
@@ -2095,11 +2077,14 @@ static int copy_ring_size(struct intel_context *dst,
 static int clone_engines(struct i915_gem_context *dst,
 			 struct i915_gem_context *src)
 {
-	struct i915_gem_engines *e = i915_gem_context_lock_engines(src);
-	struct i915_gem_engines *clone;
+	struct i915_gem_engines *clone, *e;
 	bool user_engines;
 	unsigned long n;
 
+	e = __context_engines_await(src, &user_engines);
+	if (!e)
+		return -ENOENT;
+
 	clone = alloc_engines(e->num_engines);
 	if (!clone)
 		goto err_unlock;
@@ -2141,9 +2126,7 @@ static int clone_engines(struct i915_gem_context *dst,
 		}
 	}
 	clone->num_engines = n;
-
-	user_engines = i915_gem_context_user_engines(src);
-	i915_gem_context_unlock_engines(src);
+	i915_sw_fence_complete(&e->fence);
 
 	/* Serialised by constructor */
 	engines_idle_release(dst, rcu_replace_pointer(dst->engines, clone, 1));
@@ -2154,7 +2137,7 @@ static int clone_engines(struct i915_gem_context *dst,
 	return 0;
 
 err_unlock:
-	i915_gem_context_unlock_engines(src);
+	i915_sw_fence_complete(&e->fence);
 	return -ENOMEM;
 }
 

drivers/gpu/drm/i915/gem/i915_gem_create.c

@@ -0,0 +1,113 @@
+// SPDX-License-Identifier: MIT
+/*
+ * Copyright © 2020 Intel Corporation
+ */
+
+#include "gem/i915_gem_ioctls.h"
+#include "gem/i915_gem_region.h"
+
+#include "i915_drv.h"
+
+static int
+i915_gem_create(struct drm_file *file,
+		struct intel_memory_region *mr,
+		u64 *size_p,
+		u32 *handle_p)
+{
+	struct drm_i915_gem_object *obj;
+	u32 handle;
+	u64 size;
+	int ret;
+
+	GEM_BUG_ON(!is_power_of_2(mr->min_page_size));
+	size = round_up(*size_p, mr->min_page_size);
+	if (size == 0)
+		return -EINVAL;
+
+	/* For most of the ABI (e.g. mmap) we think in system pages */
+	GEM_BUG_ON(!IS_ALIGNED(size, PAGE_SIZE));
+
+	/* Allocate the new object */
+	obj = i915_gem_object_create_region(mr, size, 0);
+	if (IS_ERR(obj))
+		return PTR_ERR(obj);
+
+	GEM_BUG_ON(size != obj->base.size);
+
+	ret = drm_gem_handle_create(file, &obj->base, &handle);
+	/* drop reference from allocate - handle holds it now */
+	i915_gem_object_put(obj);
+	if (ret)
+		return ret;
+
+	*handle_p = handle;
+	*size_p = size;
+	return 0;
+}
+
+int
+i915_gem_dumb_create(struct drm_file *file,
+		     struct drm_device *dev,
+		     struct drm_mode_create_dumb *args)
+{
+	enum intel_memory_type mem_type;
+	int cpp = DIV_ROUND_UP(args->bpp, 8);
+	u32 format;
+
+	switch (cpp) {
+	case 1:
+		format = DRM_FORMAT_C8;
+		break;
+	case 2:
+		format = DRM_FORMAT_RGB565;
+		break;
+	case 4:
+		format = DRM_FORMAT_XRGB8888;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	/* have to work out size/pitch and return them */
+	args->pitch = ALIGN(args->width * cpp, 64);
+
+	/* align stride to page size so that we can remap */
+	if (args->pitch > intel_plane_fb_max_stride(to_i915(dev), format,
+						    DRM_FORMAT_MOD_LINEAR))
+		args->pitch = ALIGN(args->pitch, 4096);
+
+	if (args->pitch < args->width)
+		return -EINVAL;
+
+	args->size = mul_u32_u32(args->pitch, args->height);
+
+	mem_type = INTEL_MEMORY_SYSTEM;
+	if (HAS_LMEM(to_i915(dev)))
+		mem_type = INTEL_MEMORY_LOCAL;
+
+	return i915_gem_create(file,
+			       intel_memory_region_by_type(to_i915(dev),
+							   mem_type),
+			       &args->size, &args->handle);
+}
+
+/**
+ * Creates a new mm object and returns a handle to it.
+ * @dev: drm device pointer
+ * @data: ioctl data blob
+ * @file: drm file pointer
+ */
+int
+i915_gem_create_ioctl(struct drm_device *dev, void *data,
+		      struct drm_file *file)
+{
+	struct drm_i915_private *i915 = to_i915(dev);
+	struct drm_i915_gem_create *args = data;
+
+	i915_gem_flush_free_objects(i915);
+
+	return i915_gem_create(file,
+			       intel_memory_region_by_type(i915,
+							   INTEL_MEMORY_SYSTEM),
+			       &args->size, &args->handle);
+}

drivers/gpu/drm/i915/gem/i915_gem_domain.c

@@ -5,6 +5,7 @@
  */
 
 #include "display/intel_frontbuffer.h"
+#include "gt/intel_gt.h"
 
 #include "i915_drv.h"
 #include "i915_gem_clflush.h"
@@ -15,13 +16,58 @@
 #include "i915_gem_lmem.h"
 #include "i915_gem_mman.h"
 
+static bool gpu_write_needs_clflush(struct drm_i915_gem_object *obj)
+{
+	return !(obj->cache_level == I915_CACHE_NONE ||
+		 obj->cache_level == I915_CACHE_WT);
+}
+
+static void
+flush_write_domain(struct drm_i915_gem_object *obj, unsigned int flush_domains)
+{
+	struct i915_vma *vma;
+
+	assert_object_held(obj);
+
+	if (!(obj->write_domain & flush_domains))
+		return;
+
+	switch (obj->write_domain) {
+	case I915_GEM_DOMAIN_GTT:
+		spin_lock(&obj->vma.lock);
+		for_each_ggtt_vma(vma, obj) {
+			if (i915_vma_unset_ggtt_write(vma))
+				intel_gt_flush_ggtt_writes(vma->vm->gt);
+		}
+		spin_unlock(&obj->vma.lock);
+
+		i915_gem_object_flush_frontbuffer(obj, ORIGIN_CPU);
+		break;
+
+	case I915_GEM_DOMAIN_WC:
+		wmb();
+		break;
+
+	case I915_GEM_DOMAIN_CPU:
+		i915_gem_clflush_object(obj, I915_CLFLUSH_SYNC);
+		break;
+
+	case I915_GEM_DOMAIN_RENDER:
+		if (gpu_write_needs_clflush(obj))
+			obj->cache_dirty = true;
+		break;
+	}
+
+	obj->write_domain = 0;
+}
+
 static void __i915_gem_object_flush_for_display(struct drm_i915_gem_object *obj)
 {
 	/*
 	 * We manually flush the CPU domain so that we can override and
 	 * force the flush for the display, and perform it asyncrhonously.
 	 */
-	i915_gem_object_flush_write_domain(obj, ~I915_GEM_DOMAIN_CPU);
+	flush_write_domain(obj, ~I915_GEM_DOMAIN_CPU);
 	if (obj->cache_dirty)
 		i915_gem_clflush_object(obj, I915_CLFLUSH_FORCE);
 	obj->write_domain = 0;
@@ -80,7 +126,7 @@ i915_gem_object_set_to_wc_domain(struct drm_i915_gem_object *obj, bool write)
 	if (ret)
 		return ret;
 
-	i915_gem_object_flush_write_domain(obj, ~I915_GEM_DOMAIN_WC);
+	flush_write_domain(obj, ~I915_GEM_DOMAIN_WC);
 
 	/* Serialise direct access to this object with the barriers for
 	 * coherent writes from the GPU, by effectively invalidating the
@@ -141,7 +187,7 @@ i915_gem_object_set_to_gtt_domain(struct drm_i915_gem_object *obj, bool write)
 	if (ret)
 		return ret;
 
-	i915_gem_object_flush_write_domain(obj, ~I915_GEM_DOMAIN_GTT);
+	flush_write_domain(obj, ~I915_GEM_DOMAIN_GTT);
 
 	/* Serialise direct access to this object with the barriers for
 	 * coherent writes from the GPU, by effectively invalidating the
@@ -370,6 +416,7 @@ retry:
 	}
 
 	vma->display_alignment = max_t(u64, vma->display_alignment, alignment);
+	i915_vma_mark_scanout(vma);
 
 	i915_gem_object_flush_if_display_locked(obj);
 
@@ -387,48 +434,6 @@ err:
 	return vma;
 }
 
-static void i915_gem_object_bump_inactive_ggtt(struct drm_i915_gem_object *obj)
-{
-	struct drm_i915_private *i915 = to_i915(obj->base.dev);
-	struct i915_vma *vma;
-
-	if (list_empty(&obj->vma.list))
-		return;
-
-	mutex_lock(&i915->ggtt.vm.mutex);
-	spin_lock(&obj->vma.lock);
-	for_each_ggtt_vma(vma, obj) {
-		if (!drm_mm_node_allocated(&vma->node))
-			continue;
-
-		GEM_BUG_ON(vma->vm != &i915->ggtt.vm);
-		list_move_tail(&vma->vm_link, &vma->vm->bound_list);
-	}
-	spin_unlock(&obj->vma.lock);
-	mutex_unlock(&i915->ggtt.vm.mutex);
-
-	if (i915_gem_object_is_shrinkable(obj)) {
-		unsigned long flags;
-
-		spin_lock_irqsave(&i915->mm.obj_lock, flags);
-
-		if (obj->mm.madv == I915_MADV_WILLNEED &&
-		    !atomic_read(&obj->mm.shrink_pin))
-			list_move_tail(&obj->mm.link, &i915->mm.shrink_list);
-
-		spin_unlock_irqrestore(&i915->mm.obj_lock, flags);
-	}
-}
-
-void
-i915_gem_object_unpin_from_display_plane(struct i915_vma *vma)
-{
-	/* Bump the LRU to try and avoid premature eviction whilst flipping  */
-	i915_gem_object_bump_inactive_ggtt(vma->obj);
-
-	i915_vma_unpin(vma);
-}
-
 /**
  * Moves a single object to the CPU read, and possibly write domain.
  * @obj: object to act on
@@ -451,7 +456,7 @@ i915_gem_object_set_to_cpu_domain(struct drm_i915_gem_object *obj, bool write)
 	if (ret)
 		return ret;
 
-	i915_gem_object_flush_write_domain(obj, ~I915_GEM_DOMAIN_CPU);
+	flush_write_domain(obj, ~I915_GEM_DOMAIN_CPU);
 
 	/* Flush the CPU cache if it's still invalid. */
 	if ((obj->read_domains & I915_GEM_DOMAIN_CPU) == 0) {
@@ -569,9 +574,6 @@ i915_gem_set_domain_ioctl(struct drm_device *dev, void *data,
 	else
 		err = i915_gem_object_set_to_cpu_domain(obj, write_domain);
 
-	/* And bump the LRU for this access */
-	i915_gem_object_bump_inactive_ggtt(obj);
-
 	i915_gem_object_unlock(obj);
 
 	if (write_domain)
@@ -619,7 +621,7 @@ int i915_gem_object_prepare_read(struct drm_i915_gem_object *obj,
 			goto out;
 	}
 
-	i915_gem_object_flush_write_domain(obj, ~I915_GEM_DOMAIN_CPU);
+	flush_write_domain(obj, ~I915_GEM_DOMAIN_CPU);
 
 	/* If we're not in the cpu read domain, set ourself into the gtt
 	 * read domain and manually flush cachelines (if required). This
@@ -670,7 +672,7 @@ int i915_gem_object_prepare_write(struct drm_i915_gem_object *obj,
 			goto out;
 	}
 
-	i915_gem_object_flush_write_domain(obj, ~I915_GEM_DOMAIN_CPU);
+	flush_write_domain(obj, ~I915_GEM_DOMAIN_CPU);
 
 	/* If we're not in the cpu write domain, set ourself into the
 	 * gtt write domain and manually flush cachelines (as required).

drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c

@@ -1276,7 +1276,10 @@ static int __reloc_gpu_alloc(struct i915_execbuffer *eb,
 	int err;
 
 	if (!pool) {
-		pool = intel_gt_get_buffer_pool(engine->gt, PAGE_SIZE);
+		pool = intel_gt_get_buffer_pool(engine->gt, PAGE_SIZE,
+						cache->has_llc ?
+						I915_MAP_WB :
+						I915_MAP_WC);
 		if (IS_ERR(pool))
 			return PTR_ERR(pool);
 	}
@@ -1286,10 +1289,7 @@ static int __reloc_gpu_alloc(struct i915_execbuffer *eb,
 	if (err)
 		goto err_pool;
 
-	cmd = i915_gem_object_pin_map(pool->obj,
-				      cache->has_llc ?
-				      I915_MAP_FORCE_WB :
-				      I915_MAP_FORCE_WC);
+	cmd = i915_gem_object_pin_map(pool->obj, pool->type);
 	if (IS_ERR(cmd)) {
 		err = PTR_ERR(cmd);
 		goto err_pool;
@@ -2458,7 +2458,8 @@ static int eb_parse(struct i915_execbuffer *eb)
 		return -EINVAL;
 
 	if (!pool) {
-		pool = intel_gt_get_buffer_pool(eb->engine->gt, len);
+		pool = intel_gt_get_buffer_pool(eb->engine->gt, len,
+						I915_MAP_WB);
 		if (IS_ERR(pool))
 			return PTR_ERR(pool);
 		eb->batch_pool = pool;

drivers/gpu/drm/i915/gem/i915_gem_lmem.c

@@ -31,18 +31,13 @@ i915_gem_object_create_lmem(struct drm_i915_private *i915,
 					     size, flags);
 }
 
-struct drm_i915_gem_object *
-__i915_gem_lmem_object_create(struct intel_memory_region *mem,
-			      resource_size_t size,
-			      unsigned int flags)
+int __i915_gem_lmem_object_init(struct intel_memory_region *mem,
+				struct drm_i915_gem_object *obj,
+				resource_size_t size,
+				unsigned int flags)
 {
 	static struct lock_class_key lock_class;
 	struct drm_i915_private *i915 = mem->i915;
-	struct drm_i915_gem_object *obj;
-
-	obj = i915_gem_object_alloc();
-	if (!obj)
-		return ERR_PTR(-ENOMEM);
 
 	drm_gem_private_object_init(&i915->drm, &obj->base, size);
 	i915_gem_object_init(obj, &i915_gem_lmem_obj_ops, &lock_class);
@@ -53,5 +48,5 @@ __i915_gem_lmem_object_create(struct intel_memory_region *mem,
 
 	i915_gem_object_init_memory_region(obj, mem, flags);
 
-	return obj;
+	return 0;
 }

drivers/gpu/drm/i915/gem/i915_gem_lmem.h

@@ -21,9 +21,9 @@ i915_gem_object_create_lmem(struct drm_i915_private *i915,
 			    resource_size_t size,
 			    unsigned int flags);
 
-struct drm_i915_gem_object *
-__i915_gem_lmem_object_create(struct intel_memory_region *mem,
-			      resource_size_t size,
-			      unsigned int flags);
+int __i915_gem_lmem_object_init(struct intel_memory_region *mem,
+				struct drm_i915_gem_object *obj,
+				resource_size_t size,
+				unsigned int flags);
 
 #endif /* !__I915_GEM_LMEM_H */

drivers/gpu/drm/i915/gem/i915_gem_object.c

@@ -25,7 +25,6 @@
 #include <linux/sched/mm.h>
 
 #include "display/intel_frontbuffer.h"
-#include "gt/intel_gt.h"
 #include "i915_drv.h"
 #include "i915_gem_clflush.h"
 #include "i915_gem_context.h"
@@ -313,52 +312,6 @@ static void i915_gem_free_object(struct drm_gem_object *gem_obj)
 		queue_work(i915->wq, &i915->mm.free_work);
 }
 
-static bool gpu_write_needs_clflush(struct drm_i915_gem_object *obj)
-{
-	return !(obj->cache_level == I915_CACHE_NONE ||
-		 obj->cache_level == I915_CACHE_WT);
-}
-
-void
-i915_gem_object_flush_write_domain(struct drm_i915_gem_object *obj,
-				   unsigned int flush_domains)
-{
-	struct i915_vma *vma;
-
-	assert_object_held(obj);
-
-	if (!(obj->write_domain & flush_domains))
-		return;
-
-	switch (obj->write_domain) {
-	case I915_GEM_DOMAIN_GTT:
-		spin_lock(&obj->vma.lock);
-		for_each_ggtt_vma(vma, obj) {
-			if (i915_vma_unset_ggtt_write(vma))
-				intel_gt_flush_ggtt_writes(vma->vm->gt);
-		}
-		spin_unlock(&obj->vma.lock);
-
-		i915_gem_object_flush_frontbuffer(obj, ORIGIN_CPU);
-		break;
-
-	case I915_GEM_DOMAIN_WC:
-		wmb();
-		break;
-
-	case I915_GEM_DOMAIN_CPU:
-		i915_gem_clflush_object(obj, I915_CLFLUSH_SYNC);
-		break;
-
-	case I915_GEM_DOMAIN_RENDER:
-		if (gpu_write_needs_clflush(obj))
-			obj->cache_dirty = true;
-		break;
-	}
-
-	obj->write_domain = 0;
-}
-
 void __i915_gem_object_flush_frontbuffer(struct drm_i915_gem_object *obj,
 					 enum fb_op_origin origin)
 {

drivers/gpu/drm/i915/gem/i915_gem_object.h

@@ -187,6 +187,24 @@ i915_gem_object_set_volatile(struct drm_i915_gem_object *obj)
 	obj->flags |= I915_BO_ALLOC_VOLATILE;
 }
 
+static inline bool
+i915_gem_object_has_tiling_quirk(struct drm_i915_gem_object *obj)
+{
+	return test_bit(I915_TILING_QUIRK_BIT, &obj->flags);
+}
+
+static inline void
+i915_gem_object_set_tiling_quirk(struct drm_i915_gem_object *obj)
+{
+	set_bit(I915_TILING_QUIRK_BIT, &obj->flags);
+}
+
+static inline void
+i915_gem_object_clear_tiling_quirk(struct drm_i915_gem_object *obj)
+{
+	clear_bit(I915_TILING_QUIRK_BIT, &obj->flags);
+}
+
 static inline bool
 i915_gem_object_type_has(const struct drm_i915_gem_object *obj,
 			 unsigned long flags)
@@ -384,14 +402,6 @@ int __i915_gem_object_put_pages(struct drm_i915_gem_object *obj);
 void i915_gem_object_truncate(struct drm_i915_gem_object *obj);
 void i915_gem_object_writeback(struct drm_i915_gem_object *obj);
 
-enum i915_map_type {
-	I915_MAP_WB = 0,
-	I915_MAP_WC,
-#define I915_MAP_OVERRIDE BIT(31)
-	I915_MAP_FORCE_WB = I915_MAP_WB | I915_MAP_OVERRIDE,
-	I915_MAP_FORCE_WC = I915_MAP_WC | I915_MAP_OVERRIDE,
-};
-
 /**
  * i915_gem_object_pin_map - return a contiguous mapping of the entire object
  * @obj: the object to map into kernel address space
@@ -435,10 +445,6 @@ static inline void i915_gem_object_unpin_map(struct drm_i915_gem_object *obj)
 
 void __i915_gem_object_release_map(struct drm_i915_gem_object *obj);
 
-void
-i915_gem_object_flush_write_domain(struct drm_i915_gem_object *obj,
-				   unsigned int flush_domains);
-
 int i915_gem_object_prepare_read(struct drm_i915_gem_object *obj,
 				 unsigned int *needs_clflush);
 int i915_gem_object_prepare_write(struct drm_i915_gem_object *obj,
@@ -486,7 +492,6 @@ i915_gem_object_pin_to_display_plane(struct drm_i915_gem_object *obj,
 				     u32 alignment,
 				     const struct i915_ggtt_view *view,
 				     unsigned int flags);
-void i915_gem_object_unpin_from_display_plane(struct i915_vma *vma);
 
 void i915_gem_object_make_unshrinkable(struct drm_i915_gem_object *obj);
 void i915_gem_object_make_shrinkable(struct drm_i915_gem_object *obj);
@@ -512,6 +517,9 @@ static inline void __start_cpu_write(struct drm_i915_gem_object *obj)
 		obj->cache_dirty = true;
 }
 
+void i915_gem_fence_wait_priority(struct dma_fence *fence,
+				  const struct i915_sched_attr *attr);
+
 int i915_gem_object_wait(struct drm_i915_gem_object *obj,
 			 unsigned int flags,
 			 long timeout);
@@ -540,4 +548,6 @@ i915_gem_object_invalidate_frontbuffer(struct drm_i915_gem_object *obj,
 		__i915_gem_object_invalidate_frontbuffer(obj, origin);
 }
 
+bool i915_gem_object_is_shmem(const struct drm_i915_gem_object *obj);
+
 #endif

drivers/gpu/drm/i915/gem/i915_gem_object_blt.c

@@ -35,7 +35,7 @@ struct i915_vma *intel_emit_vma_fill_blt(struct intel_context *ce,
 	count = div_u64(round_up(vma->size, block_size), block_size);
 	size = (1 + 8 * count) * sizeof(u32);
 	size = round_up(size, PAGE_SIZE);
-	pool = intel_gt_get_buffer_pool(ce->engine->gt, size);
+	pool = intel_gt_get_buffer_pool(ce->engine->gt, size, I915_MAP_WC);
 	if (IS_ERR(pool)) {
 		err = PTR_ERR(pool);
 		goto out_pm;
@@ -55,7 +55,7 @@ struct i915_vma *intel_emit_vma_fill_blt(struct intel_context *ce,
 	if (unlikely(err))
 		goto out_put;
 
-	cmd = i915_gem_object_pin_map(pool->obj, I915_MAP_WC);
+	cmd = i915_gem_object_pin_map(pool->obj, pool->type);
 	if (IS_ERR(cmd)) {
 		err = PTR_ERR(cmd);
 		goto out_unpin;
@@ -257,7 +257,7 @@ struct i915_vma *intel_emit_vma_copy_blt(struct intel_context *ce,
 	count = div_u64(round_up(dst->size, block_size), block_size);
 	size = (1 + 11 * count) * sizeof(u32);
 	size = round_up(size, PAGE_SIZE);
-	pool = intel_gt_get_buffer_pool(ce->engine->gt, size);
+	pool = intel_gt_get_buffer_pool(ce->engine->gt, size, I915_MAP_WC);
 	if (IS_ERR(pool)) {
 		err = PTR_ERR(pool);
 		goto out_pm;
@@ -277,7 +277,7 @@ struct i915_vma *intel_emit_vma_copy_blt(struct intel_context *ce,
 	if (unlikely(err))
 		goto out_put;
 
-	cmd = i915_gem_object_pin_map(pool->obj, I915_MAP_WC);
+	cmd = i915_gem_object_pin_map(pool->obj, pool->type);
 	if (IS_ERR(cmd)) {
 		err = PTR_ERR(cmd);
 		goto out_unpin;

drivers/gpu/drm/i915/gem/i915_gem_object_types.h

@@ -67,6 +67,14 @@ struct drm_i915_gem_object_ops {
 	const char *name; /* friendly name for debug, e.g. lockdep classes */
 };
 
+enum i915_map_type {
+	I915_MAP_WB = 0,
+	I915_MAP_WC,
+#define I915_MAP_OVERRIDE BIT(31)
+	I915_MAP_FORCE_WB = I915_MAP_WB | I915_MAP_OVERRIDE,
+	I915_MAP_FORCE_WC = I915_MAP_WC | I915_MAP_OVERRIDE,
+};
+
 enum i915_mmap_type {
 	I915_MMAP_TYPE_GTT = 0,
 	I915_MMAP_TYPE_WC,
@@ -142,8 +150,6 @@ struct drm_i915_gem_object {
 	 */
 	struct list_head obj_link;
 
-	/** Stolen memory for this object, instead of being backed by shmem. */
-	struct drm_mm_node *stolen;
 	union {
 		struct rcu_head rcu;
 		struct llist_node freed;
@@ -167,6 +173,7 @@ struct drm_i915_gem_object {
 #define I915_BO_ALLOC_VOLATILE   BIT(1)
 #define I915_BO_ALLOC_FLAGS (I915_BO_ALLOC_CONTIGUOUS | I915_BO_ALLOC_VOLATILE)
 #define I915_BO_READONLY         BIT(2)
+#define I915_TILING_QUIRK_BIT    3 /* unknown swizzling; do not release! */
 
 	/*
 	 * Is the object to be mapped as read-only to the GPU
@@ -275,12 +282,6 @@ struct drm_i915_gem_object {
 		 * pages were last acquired.
 		 */
 		bool dirty:1;
-
-		/**
-		 * This is set if the object has been pinned due to unknown
-		 * swizzling.
-		 */
-		bool quirked:1;
 	} mm;
 
 	/** Record of address bit 17 of each page at last unbind. */
@@ -295,6 +296,8 @@ struct drm_i915_gem_object {
 			struct work_struct *work;
 		} userptr;
 
+		struct drm_mm_node *stolen;
+
 		unsigned long scratch;
 		u64 encode;
 

drivers/gpu/drm/i915/gem/i915_gem_pages.c

@@ -16,6 +16,7 @@ void __i915_gem_object_set_pages(struct drm_i915_gem_object *obj,
 {
 	struct drm_i915_private *i915 = to_i915(obj->base.dev);
 	unsigned long supported = INTEL_INFO(i915)->page_sizes;
+	bool shrinkable;
 	int i;
 
 	lockdep_assert_held(&obj->mm.lock);
@@ -38,13 +39,6 @@ void __i915_gem_object_set_pages(struct drm_i915_gem_object *obj,
 
 	obj->mm.pages = pages;
 
-	if (i915_gem_object_is_tiled(obj) &&
-	    i915->quirks & QUIRK_PIN_SWIZZLED_PAGES) {
-		GEM_BUG_ON(obj->mm.quirked);
-		__i915_gem_object_pin_pages(obj);
-		obj->mm.quirked = true;
-	}
-
 	GEM_BUG_ON(!sg_page_sizes);
 	obj->mm.page_sizes.phys = sg_page_sizes;
 
@@ -63,7 +57,16 @@ void __i915_gem_object_set_pages(struct drm_i915_gem_object *obj,
 	}
 	GEM_BUG_ON(!HAS_PAGE_SIZES(i915, obj->mm.page_sizes.sg));
 
-	if (i915_gem_object_is_shrinkable(obj)) {
+	shrinkable = i915_gem_object_is_shrinkable(obj);
+
+	if (i915_gem_object_is_tiled(obj) &&
+	    i915->quirks & QUIRK_PIN_SWIZZLED_PAGES) {
+		GEM_BUG_ON(i915_gem_object_has_tiling_quirk(obj));
+		i915_gem_object_set_tiling_quirk(obj);
+		shrinkable = false;
+	}
+
+	if (shrinkable) {
 		struct list_head *list;
 		unsigned long flags;
 

drivers/gpu/drm/i915/gem/i915_gem_phys.c

@@ -213,7 +213,7 @@ int i915_gem_object_attach_phys(struct drm_i915_gem_object *obj, int align)
 	if (obj->ops == &i915_gem_phys_ops)
 		return 0;
 
-	if (obj->ops != &i915_gem_shmem_ops)
+	if (!i915_gem_object_is_shmem(obj))
 		return -EINVAL;
 
 	err = i915_gem_object_unbind(obj, I915_GEM_OBJECT_UNBIND_ACTIVE);
@@ -227,7 +227,7 @@ int i915_gem_object_attach_phys(struct drm_i915_gem_object *obj, int align)
 		goto err_unlock;
 	}
 
-	if (obj->mm.quirked) {
+	if (i915_gem_object_has_tiling_quirk(obj)) {
 		err = -EFAULT;
 		goto err_unlock;
 	}

drivers/gpu/drm/i915/gem/i915_gem_pm.c

@@ -11,6 +11,13 @@
 
 #include "i915_drv.h"
 
+#if defined(CONFIG_X86)
+#include <asm/smp.h>
+#else
+#define wbinvd_on_all_cpus() \
+	pr_warn(DRIVER_NAME ": Missing cache flush in %s\n", __func__)
+#endif
+
 void i915_gem_suspend(struct drm_i915_private *i915)
 {
 	GEM_TRACE("%s\n", dev_name(i915->drm.dev));
@@ -32,13 +39,6 @@ void i915_gem_suspend(struct drm_i915_private *i915)
 	i915_gem_drain_freed_objects(i915);
 }
 
-static struct drm_i915_gem_object *first_mm_object(struct list_head *list)
-{
-	return list_first_entry_or_null(list,
-					struct drm_i915_gem_object,
-					mm.link);
-}
-
 void i915_gem_suspend_late(struct drm_i915_private *i915)
 {
 	struct drm_i915_gem_object *obj;
@@ -48,6 +48,7 @@ void i915_gem_suspend_late(struct drm_i915_private *i915)
 		NULL
 	}, **phase;
 	unsigned long flags;
+	bool flush = false;
 
 	/*
 	 * Neither the BIOS, ourselves or any other kernel
@@ -73,29 +74,15 @@ void i915_gem_suspend_late(struct drm_i915_private *i915)
 
 	spin_lock_irqsave(&i915->mm.obj_lock, flags);
 	for (phase = phases; *phase; phase++) {
-		LIST_HEAD(keep);
-
-		while ((obj = first_mm_object(*phase))) {
-			list_move_tail(&obj->mm.link, &keep);
-
-			/* Beware the background _i915_gem_free_objects */
-			if (!kref_get_unless_zero(&obj->base.refcount))
-				continue;
-
-			spin_unlock_irqrestore(&i915->mm.obj_lock, flags);
-
-			i915_gem_object_lock(obj, NULL);
-			drm_WARN_ON(&i915->drm,
-			    i915_gem_object_set_to_gtt_domain(obj, false));
-			i915_gem_object_unlock(obj);
-			i915_gem_object_put(obj);
-
-			spin_lock_irqsave(&i915->mm.obj_lock, flags);
+		list_for_each_entry(obj, *phase, mm.link) {
+			if (!(obj->cache_coherent & I915_BO_CACHE_COHERENT_FOR_READ))
+				flush |= (obj->read_domains & I915_GEM_DOMAIN_CPU) == 0;
+			__start_cpu_write(obj); /* presume auto-hibernate */
 		}
-
-		list_splice_tail(&keep, *phase);
 	}
 	spin_unlock_irqrestore(&i915->mm.obj_lock, flags);
+	if (flush)
+		wbinvd_on_all_cpus();
 }
 
 void i915_gem_resume(struct drm_i915_private *i915)

drivers/gpu/drm/i915/gem/i915_gem_region.c

@@ -143,6 +143,7 @@ i915_gem_object_create_region(struct intel_memory_region *mem,
 			      unsigned int flags)
 {
 	struct drm_i915_gem_object *obj;
+	int err;
 
 	/*
 	 * NB: Our use of resource_size_t for the size stems from using struct
@@ -173,9 +174,18 @@ i915_gem_object_create_region(struct intel_memory_region *mem,
 	if (overflows_type(size, obj->base.size))
 		return ERR_PTR(-E2BIG);
 
-	obj = mem->ops->create_object(mem, size, flags);
-	if (!IS_ERR(obj))
-		trace_i915_gem_object_create(obj);
+	obj = i915_gem_object_alloc();
+	if (!obj)
+		return ERR_PTR(-ENOMEM);
 
+	err = mem->ops->init_object(mem, obj, size, flags);
+	if (err)
+		goto err_object_free;
+
+	trace_i915_gem_object_create(obj);
 	return obj;
+
+err_object_free:
+	i915_gem_object_free(obj);
+	return ERR_PTR(err);
 }

drivers/gpu/drm/i915/gem/i915_gem_shmem.c

@@ -464,26 +464,21 @@ static int __create_shmem(struct drm_i915_private *i915,
 	return 0;
 }
 
-static struct drm_i915_gem_object *
-create_shmem(struct intel_memory_region *mem,
-	     resource_size_t size,
-	     unsigned int flags)
+static int shmem_object_init(struct intel_memory_region *mem,
+			     struct drm_i915_gem_object *obj,
+			     resource_size_t size,
+			     unsigned int flags)
 {
 	static struct lock_class_key lock_class;
 	struct drm_i915_private *i915 = mem->i915;
-	struct drm_i915_gem_object *obj;
 	struct address_space *mapping;
 	unsigned int cache_level;
 	gfp_t mask;
 	int ret;
 
-	obj = i915_gem_object_alloc();
-	if (!obj)
-		return ERR_PTR(-ENOMEM);
-
 	ret = __create_shmem(i915, &obj->base, size);
 	if (ret)
-		goto fail;
+		return ret;
 
 	mask = GFP_HIGHUSER | __GFP_RECLAIMABLE;
 	if (IS_I965GM(i915) || IS_I965G(i915)) {
@@ -522,11 +517,7 @@ create_shmem(struct intel_memory_region *mem,
 
 	i915_gem_object_init_memory_region(obj, mem, 0);
 
-	return obj;
-
-fail:
-	i915_gem_object_free(obj);
-	return ERR_PTR(ret);
+	return 0;
 }
 
 struct drm_i915_gem_object *
@@ -611,7 +602,7 @@ static void release_shmem(struct intel_memory_region *mem)
 static const struct intel_memory_region_ops shmem_region_ops = {
 	.init = init_shmem,
 	.release = release_shmem,
-	.create_object = create_shmem,
+	.init_object = shmem_object_init,
 };
 
 struct intel_memory_region *i915_gem_shmem_setup(struct drm_i915_private *i915)
@@ -621,3 +612,8 @@ struct intel_memory_region *i915_gem_shmem_setup(struct drm_i915_private *i915)
 					  PAGE_SIZE, 0,
 					  &shmem_region_ops);
 }
+
+bool i915_gem_object_is_shmem(const struct drm_i915_gem_object *obj)
+{
+	return obj->ops == &i915_gem_shmem_ops;
+}

drivers/gpu/drm/i915/gem/i915_gem_stolen.c

@@ -621,18 +621,13 @@ static const struct drm_i915_gem_object_ops i915_gem_object_stolen_ops = {
 	.release = i915_gem_object_release_stolen,
 };
 
-static struct drm_i915_gem_object *
-__i915_gem_object_create_stolen(struct intel_memory_region *mem,
-				struct drm_mm_node *stolen)
+static int __i915_gem_object_create_stolen(struct intel_memory_region *mem,
+					   struct drm_i915_gem_object *obj,
+					   struct drm_mm_node *stolen)
 {
 	static struct lock_class_key lock_class;
-	struct drm_i915_gem_object *obj;
 	unsigned int cache_level;
-	int err = -ENOMEM;
-
-	obj = i915_gem_object_alloc();
-	if (!obj)
-		goto err;
+	int err;
 
 	drm_gem_private_object_init(&mem->i915->drm, &obj->base, stolen->size);
 	i915_gem_object_init(obj, &i915_gem_object_stolen_ops, &lock_class);
@@ -644,55 +639,47 @@ __i915_gem_object_create_stolen(struct intel_memory_region *mem,
 
 	err = i915_gem_object_pin_pages(obj);
 	if (err)
-		goto cleanup;
+		return err;
 
 	i915_gem_object_init_memory_region(obj, mem, 0);
 
-	return obj;
-
-cleanup:
-	i915_gem_object_free(obj);
-err:
-	return ERR_PTR(err);
+	return 0;
 }
 
-static struct drm_i915_gem_object *
-_i915_gem_object_create_stolen(struct intel_memory_region *mem,
-			       resource_size_t size,
-			       unsigned int flags)
+static int _i915_gem_object_stolen_init(struct intel_memory_region *mem,
+					struct drm_i915_gem_object *obj,
+					resource_size_t size,
+					unsigned int flags)
 {
 	struct drm_i915_private *i915 = mem->i915;
-	struct drm_i915_gem_object *obj;
 	struct drm_mm_node *stolen;
 	int ret;
 
 	if (!drm_mm_initialized(&i915->mm.stolen))
-		return ERR_PTR(-ENODEV);
+		return -ENODEV;
 
 	if (size == 0)
-		return ERR_PTR(-EINVAL);
+		return -EINVAL;
 
 	stolen = kzalloc(sizeof(*stolen), GFP_KERNEL);
 	if (!stolen)
-		return ERR_PTR(-ENOMEM);
+		return -ENOMEM;
 
 	ret = i915_gem_stolen_insert_node(i915, stolen, size, 4096);
-	if (ret) {
-		obj = ERR_PTR(ret);
+	if (ret)
 		goto err_free;
-	}
 
-	obj = __i915_gem_object_create_stolen(mem, stolen);
-	if (IS_ERR(obj))
+	ret = __i915_gem_object_create_stolen(mem, obj, stolen);
+	if (ret)
 		goto err_remove;
 
-	return obj;
+	return 0;
 
 err_remove:
 	i915_gem_stolen_remove_node(i915, stolen);
 err_free:
 	kfree(stolen);
-	return obj;
+	return ret;
 }
 
 struct drm_i915_gem_object *
@@ -722,7 +709,7 @@ static void release_stolen(struct intel_memory_region *mem)
 static const struct intel_memory_region_ops i915_region_stolen_ops = {
 	.init = init_stolen,
 	.release = release_stolen,
-	.create_object = _i915_gem_object_create_stolen,
+	.init_object = _i915_gem_object_stolen_init,
 };
 
 struct intel_memory_region *i915_gem_stolen_setup(struct drm_i915_private *i915)
@@ -771,16 +758,31 @@ i915_gem_object_create_stolen_for_preallocated(struct drm_i915_private *i915,
 		goto err_free;
 	}
 
-	obj = __i915_gem_object_create_stolen(mem, stolen);
-	if (IS_ERR(obj))
+	obj = i915_gem_object_alloc();
+	if (!obj) {
+		obj = ERR_PTR(-ENOMEM);
 		goto err_stolen;
+	}
+
+	ret = __i915_gem_object_create_stolen(mem, obj, stolen);
+	if (ret) {
+		obj = ERR_PTR(ret);
+		goto err_object_free;
+	}
 
 	i915_gem_object_set_cache_coherency(obj, I915_CACHE_NONE);
 	return obj;
 
+err_object_free:
+	i915_gem_object_free(obj);
 err_stolen:
 	i915_gem_stolen_remove_node(i915, stolen);
 err_free:
 	kfree(stolen);
 	return obj;
 }
+
+bool i915_gem_object_is_stolen(const struct drm_i915_gem_object *obj)
+{
+	return obj->ops == &i915_gem_object_stolen_ops;
+}

drivers/gpu/drm/i915/gem/i915_gem_stolen.h

@@ -30,6 +30,8 @@ i915_gem_object_create_stolen_for_preallocated(struct drm_i915_private *dev_priv
 					       resource_size_t stolen_offset,
 					       resource_size_t size);
 
+bool i915_gem_object_is_stolen(const struct drm_i915_gem_object *obj);
+
 #define I915_GEM_STOLEN_BIAS SZ_128K
 
 #endif /* __I915_GEM_STOLEN_H__ */

drivers/gpu/drm/i915/gem/i915_gem_tiling.c

@@ -270,14 +270,14 @@ i915_gem_object_set_tiling(struct drm_i915_gem_object *obj,
 	    obj->mm.madv == I915_MADV_WILLNEED &&
 	    i915->quirks & QUIRK_PIN_SWIZZLED_PAGES) {
 		if (tiling == I915_TILING_NONE) {
-			GEM_BUG_ON(!obj->mm.quirked);
-			__i915_gem_object_unpin_pages(obj);
-			obj->mm.quirked = false;
+			GEM_BUG_ON(!i915_gem_object_has_tiling_quirk(obj));
+			i915_gem_object_clear_tiling_quirk(obj);
+			i915_gem_object_make_shrinkable(obj);
 		}
 		if (!i915_gem_object_is_tiled(obj)) {
-			GEM_BUG_ON(obj->mm.quirked);
-			__i915_gem_object_pin_pages(obj);
-			obj->mm.quirked = true;
+			GEM_BUG_ON(i915_gem_object_has_tiling_quirk(obj));
+			i915_gem_object_make_unshrinkable(obj);
+			i915_gem_object_set_tiling_quirk(obj);
 		}
 	}
 	mutex_unlock(&obj->mm.lock);

drivers/gpu/drm/i915/gem/i915_gem_wait.c

@@ -5,6 +5,7 @@
  */
 
 #include <linux/dma-fence-array.h>
+#include <linux/dma-fence-chain.h>
 #include <linux/jiffies.h>
 
 #include "gt/intel_engine.h"
@@ -44,8 +45,7 @@ i915_gem_object_wait_reservation(struct dma_resv *resv,
 		unsigned int count, i;
 		int ret;
 
-		ret = dma_resv_get_fences_rcu(resv,
-							&excl, &count, &shared);
+		ret = dma_resv_get_fences_rcu(resv, &excl, &count, &shared);
 		if (ret)
 			return ret;
 
@@ -91,8 +91,8 @@ i915_gem_object_wait_reservation(struct dma_resv *resv,
 	return timeout;
 }
 
-static void __fence_set_priority(struct dma_fence *fence,
-				 const struct i915_sched_attr *attr)
+static void fence_set_priority(struct dma_fence *fence,
+			       const struct i915_sched_attr *attr)
 {
 	struct i915_request *rq;
 	struct intel_engine_cs *engine;
@@ -103,27 +103,47 @@ static void __fence_set_priority(struct dma_fence *fence,
 	rq = to_request(fence);
 	engine = rq->engine;
 
-	local_bh_disable();
 	rcu_read_lock(); /* RCU serialisation for set-wedged protection */
 	if (engine->schedule)
 		engine->schedule(rq, attr);
 	rcu_read_unlock();
-	local_bh_enable(); /* kick the tasklets if queues were reprioritised */
 }
 
-static void fence_set_priority(struct dma_fence *fence,
-			       const struct i915_sched_attr *attr)
+static inline bool __dma_fence_is_chain(const struct dma_fence *fence)
 {
+	return fence->ops == &dma_fence_chain_ops;
+}
+
+void i915_gem_fence_wait_priority(struct dma_fence *fence,
+				  const struct i915_sched_attr *attr)
+{
+	if (dma_fence_is_signaled(fence))
+		return;
+
+	local_bh_disable();
+
 	/* Recurse once into a fence-array */
 	if (dma_fence_is_array(fence)) {
 		struct dma_fence_array *array = to_dma_fence_array(fence);
 		int i;
 
 		for (i = 0; i < array->num_fences; i++)
-			__fence_set_priority(array->fences[i], attr);
+			fence_set_priority(array->fences[i], attr);
+	} else if (__dma_fence_is_chain(fence)) {
+		struct dma_fence *iter;
+
+		/* The chain is ordered; if we boost the last, we boost all */
+		dma_fence_chain_for_each(iter, fence) {
+			fence_set_priority(to_dma_fence_chain(iter)->fence,
+					   attr);
+			break;
+		}
+		dma_fence_put(iter);
 	} else {
-		__fence_set_priority(fence, attr);
+		fence_set_priority(fence, attr);
 	}
+
+	local_bh_enable(); /* kick the tasklets if queues were reprioritised */
 }
 
 int
@@ -139,12 +159,12 @@ i915_gem_object_wait_priority(struct drm_i915_gem_object *obj,
 		int ret;
 
 		ret = dma_resv_get_fences_rcu(obj->base.resv,
-							&excl, &count, &shared);
+					      &excl, &count, &shared);
 		if (ret)
 			return ret;
 
 		for (i = 0; i < count; i++) {
-			fence_set_priority(shared[i], attr);
+			i915_gem_fence_wait_priority(shared[i], attr);
 			dma_fence_put(shared[i]);
 		}
 
@@ -154,7 +174,7 @@ i915_gem_object_wait_priority(struct drm_i915_gem_object *obj,
 	}
 
 	if (excl) {
-		fence_set_priority(excl, attr);
+		i915_gem_fence_wait_priority(excl, attr);
 		dma_fence_put(excl);
 	}
 	return 0;

drivers/gpu/drm/i915/gt/gen6_ppgtt.c

@@ -12,9 +12,9 @@
 #include "intel_gt.h"
 
 /* Write pde (index) from the page directory @pd to the page table @pt */
-static inline void gen6_write_pde(const struct gen6_ppgtt *ppgtt,
-				  const unsigned int pde,
-				  const struct i915_page_table *pt)
+static void gen6_write_pde(const struct gen6_ppgtt *ppgtt,
+			   const unsigned int pde,
+			   const struct i915_page_table *pt)
 {
 	dma_addr_t addr = pt ? px_dma(pt) : px_dma(ppgtt->base.vm.scratch[1]);
 
@@ -27,8 +27,6 @@ void gen7_ppgtt_enable(struct intel_gt *gt)
 {
 	struct drm_i915_private *i915 = gt->i915;
 	struct intel_uncore *uncore = gt->uncore;
-	struct intel_engine_cs *engine;
-	enum intel_engine_id id;
 	u32 ecochk;
 
 	intel_uncore_rmw(uncore, GAC_ECO_BITS, 0, ECOBITS_PPGTT_CACHE64B);
@@ -41,13 +39,6 @@ void gen7_ppgtt_enable(struct intel_gt *gt)
 		ecochk &= ~ECOCHK_PPGTT_GFDT_IVB;
 	}
 	intel_uncore_write(uncore, GAM_ECOCHK, ecochk);
-
-	for_each_engine(engine, gt, id) {
-		/* GFX_MODE is per-ring on gen7+ */
-		ENGINE_WRITE(engine,
-			     RING_MODE_GEN7,
-			     _MASKED_BIT_ENABLE(GFX_PPGTT_ENABLE));
-	}
 }
 
 void gen6_ppgtt_enable(struct intel_gt *gt)

drivers/gpu/drm/i915/gt/gen7_renderclear.c

@@ -40,7 +40,7 @@ struct batch_vals {
 	u32 size;
 };
 
-static inline int num_primitives(const struct batch_vals *bv)
+static int num_primitives(const struct batch_vals *bv)
 {
 	/*
 	 * We need to saturate the GPU with work in order to dispatch
@@ -353,19 +353,21 @@ static void gen7_emit_pipeline_flush(struct batch_chunk *batch)
 
 static void gen7_emit_pipeline_invalidate(struct batch_chunk *batch)
 {
-	u32 *cs = batch_alloc_items(batch, 0, 8);
+	u32 *cs = batch_alloc_items(batch, 0, 10);
 
 	/* ivb: Stall before STATE_CACHE_INVALIDATE */
-	*cs++ = GFX_OP_PIPE_CONTROL(4);
+	*cs++ = GFX_OP_PIPE_CONTROL(5);
 	*cs++ = PIPE_CONTROL_STALL_AT_SCOREBOARD |
 		PIPE_CONTROL_CS_STALL;
 	*cs++ = 0;
 	*cs++ = 0;
+	*cs++ = 0;
 
-	*cs++ = GFX_OP_PIPE_CONTROL(4);
+	*cs++ = GFX_OP_PIPE_CONTROL(5);
 	*cs++ = PIPE_CONTROL_STATE_CACHE_INVALIDATE;
 	*cs++ = 0;
 	*cs++ = 0;
+	*cs++ = 0;
 
 	batch_advance(batch, cs);
 }
@@ -390,6 +392,17 @@ static void emit_batch(struct i915_vma * const vma,
 						     &cb_kernel_ivb,
 						     desc_count);
 
+	/* Reset inherited context registers */
+	gen7_emit_pipeline_invalidate(&cmds);
+	batch_add(&cmds, MI_LOAD_REGISTER_IMM(2));
+	batch_add(&cmds, i915_mmio_reg_offset(CACHE_MODE_0_GEN7));
+	batch_add(&cmds, 0xffff0000);
+	batch_add(&cmds, i915_mmio_reg_offset(CACHE_MODE_1));
+	batch_add(&cmds, 0xffff0000 | PIXEL_SUBSPAN_COLLECT_OPT_DISABLE);
+	gen7_emit_pipeline_invalidate(&cmds);
+	gen7_emit_pipeline_flush(&cmds);
+
+	/* Switch to the media pipeline and our base address */
 	gen7_emit_pipeline_invalidate(&cmds);
 	batch_add(&cmds, PIPELINE_SELECT | PIPELINE_SELECT_MEDIA);
 	batch_add(&cmds, MI_NOOP);
@@ -399,9 +412,11 @@ static void emit_batch(struct i915_vma * const vma,
 	gen7_emit_state_base_address(&cmds, descriptors);
 	gen7_emit_pipeline_invalidate(&cmds);
 
+	/* Set the clear-residual kernel state */
 	gen7_emit_vfe_state(&cmds, bv, urb_size - 1, 0, 0);
 	gen7_emit_interface_descriptor_load(&cmds, descriptors, desc_count);
 
+	/* Execute the kernel on all HW threads */
 	for (i = 0; i < num_primitives(bv); i++)
 		gen7_emit_media_object(&cmds, i);
 

drivers/gpu/drm/i915/gt/gen8_engine_cs.c

@@ -330,7 +330,7 @@ int gen12_emit_flush_xcs(struct i915_request *rq, u32 mode)
 	return 0;
 }
 
-static inline u32 preempt_address(struct intel_engine_cs *engine)
+static u32 preempt_address(struct intel_engine_cs *engine)
 {
 	return (i915_ggtt_offset(engine->status_page.vma) +
 		I915_GEM_HWS_PREEMPT_ADDR);
@@ -488,6 +488,7 @@ static u32 *gen8_emit_wa_tail(struct i915_request *rq, u32 *cs)
 
 static u32 *emit_preempt_busywait(struct i915_request *rq, u32 *cs)
 {
+	*cs++ = MI_ARB_CHECK; /* trigger IDLE->ACTIVE first */
 	*cs++ = MI_SEMAPHORE_WAIT |
 		MI_SEMAPHORE_GLOBAL_GTT |
 		MI_SEMAPHORE_POLL |
@@ -495,6 +496,7 @@ static u32 *emit_preempt_busywait(struct i915_request *rq, u32 *cs)
 	*cs++ = 0;
 	*cs++ = preempt_address(rq->engine);
 	*cs++ = 0;
+	*cs++ = MI_NOOP;
 
 	return cs;
 }

drivers/gpu/drm/i915/gt/gen8_ppgtt.c

@@ -109,7 +109,7 @@ static void gen8_ppgtt_notify_vgt(struct i915_ppgtt *ppgtt, bool create)
 
 #define as_pd(x) container_of((x), typeof(struct i915_page_directory), pt)
 
-static inline unsigned int
+static unsigned int
 gen8_pd_range(u64 start, u64 end, int lvl, unsigned int *idx)
 {
 	const int shift = gen8_pd_shift(lvl);
@@ -125,7 +125,7 @@ gen8_pd_range(u64 start, u64 end, int lvl, unsigned int *idx)
 		return i915_pde_index(end, shift) - *idx;
 }
 
-static inline bool gen8_pd_contains(u64 start, u64 end, int lvl)
+static bool gen8_pd_contains(u64 start, u64 end, int lvl)
 {
 	const u64 mask = ~0ull << gen8_pd_shift(lvl + 1);
 
@@ -133,7 +133,7 @@ static inline bool gen8_pd_contains(u64 start, u64 end, int lvl)
 	return (start ^ end) & mask && (start & ~mask) == 0;
 }
 
-static inline unsigned int gen8_pt_count(u64 start, u64 end)
+static unsigned int gen8_pt_count(u64 start, u64 end)
 {
 	GEM_BUG_ON(start >= end);
 	if ((start ^ end) >> gen8_pd_shift(1))
@@ -142,14 +142,13 @@ static inline unsigned int gen8_pt_count(u64 start, u64 end)
 		return end - start;
 }
 
-static inline unsigned int
-gen8_pd_top_count(const struct i915_address_space *vm)
+static unsigned int gen8_pd_top_count(const struct i915_address_space *vm)
 {
 	unsigned int shift = __gen8_pte_shift(vm->top);
 	return (vm->total + (1ull << shift) - 1) >> shift;
 }
 
-static inline struct i915_page_directory *
+static struct i915_page_directory *
 gen8_pdp_for_page_index(struct i915_address_space * const vm, const u64 idx)
 {
 	struct i915_ppgtt * const ppgtt = i915_vm_to_ppgtt(vm);
@@ -160,7 +159,7 @@ gen8_pdp_for_page_index(struct i915_address_space * const vm, const u64 idx)
 		return i915_pd_entry(ppgtt->pd, gen8_pd_index(idx, vm->top));
 }
 
-static inline struct i915_page_directory *
+static struct i915_page_directory *
 gen8_pdp_for_page_address(struct i915_address_space * const vm, const u64 addr)
 {
 	return gen8_pdp_for_page_index(vm, addr >> GEN8_PTE_SHIFT);

drivers/gpu/drm/i915/gt/intel_breadcrumbs.c

@@ -453,16 +453,17 @@ void i915_request_cancel_breadcrumb(struct i915_request *rq)
 {
 	struct intel_breadcrumbs *b = READ_ONCE(rq->engine)->breadcrumbs;
 	struct intel_context *ce = rq->context;
-	unsigned long flags;
 	bool release;
 
-	if (!test_and_clear_bit(I915_FENCE_FLAG_SIGNAL, &rq->fence.flags))
+	spin_lock(&ce->signal_lock);
+	if (!test_and_clear_bit(I915_FENCE_FLAG_SIGNAL, &rq->fence.flags)) {
+		spin_unlock(&ce->signal_lock);
 		return;
+	}
 
-	spin_lock_irqsave(&ce->signal_lock, flags);
 	list_del_rcu(&rq->signal_link);
 	release = remove_signaling_context(b, ce);
-	spin_unlock_irqrestore(&ce->signal_lock, flags);
+	spin_unlock(&ce->signal_lock);
 	if (release)
 		intel_context_put(ce);
 
@@ -517,8 +518,8 @@ static void print_signals(struct intel_breadcrumbs *b, struct drm_printer *p)
 		list_for_each_entry_rcu(rq, &ce->signals, signal_link)
 			drm_printf(p, "\t[%llx:%llx%s] @ %dms\n",
 				   rq->fence.context, rq->fence.seqno,
-				   i915_request_completed(rq) ? "!" :
-				   i915_request_started(rq) ? "*" :
+				   __i915_request_is_complete(rq) ? "!" :
+				   __i915_request_has_started(rq) ? "*" :
 				   "",
 				   jiffies_to_msecs(jiffies - rq->emitted_jiffies));
 	}

drivers/gpu/drm/i915/gt/intel_engine_cs.c

@@ -342,7 +342,7 @@ static int intel_engine_setup(struct intel_gt *gt, enum intel_engine_id id)
 	engine->schedule = NULL;
 
 	ewma__engine_latency_init(&engine->latency);
-	seqlock_init(&engine->stats.lock);
+	seqcount_init(&engine->stats.lock);
 
 	ATOMIC_INIT_NOTIFIER_HEAD(&engine->context_status_notifier);
 
@@ -1676,7 +1676,7 @@ void intel_engine_dump(struct intel_engine_cs *engine,
 			   ktime_to_ms(intel_engine_get_busy_time(engine,
 								  &dummy)));
 	drm_printf(m, "\tForcewake: %x domains, %d active\n",
-		   engine->fw_domain, atomic_read(&engine->fw_active));
+		   engine->fw_domain, READ_ONCE(engine->fw_active));
 
 	rcu_read_lock();
 	rq = READ_ONCE(engine->heartbeat.systole);
@@ -1754,7 +1754,7 @@ static ktime_t __intel_engine_get_busy_time(struct intel_engine_cs *engine,
 	 * add it to the total.
 	 */
 	*now = ktime_get();
-	if (atomic_read(&engine->stats.active))
+	if (READ_ONCE(engine->stats.active))
 		total = ktime_add(total, ktime_sub(*now, engine->stats.start));
 
 	return total;
@@ -1773,9 +1773,9 @@ ktime_t intel_engine_get_busy_time(struct intel_engine_cs *engine, ktime_t *now)
 	ktime_t total;
 
 	do {
-		seq = read_seqbegin(&engine->stats.lock);
+		seq = read_seqcount_begin(&engine->stats.lock);
 		total = __intel_engine_get_busy_time(engine, now);
-	} while (read_seqretry(&engine->stats.lock, seq));
+	} while (read_seqcount_retry(&engine->stats.lock, seq));
 
 	return total;
 }
@@ -1811,7 +1811,7 @@ intel_engine_find_active_request(struct intel_engine_cs *engine)
 		struct intel_timeline *tl = request->context->timeline;
 
 		list_for_each_entry_from_reverse(request, &tl->requests, link) {
-			if (i915_request_completed(request))
+			if (__i915_request_is_complete(request))
 				break;
 
 			active = request;
@@ -1822,10 +1822,10 @@ intel_engine_find_active_request(struct intel_engine_cs *engine)
 		return active;
 
 	list_for_each_entry(request, &engine->active.requests, sched.link) {
-		if (i915_request_completed(request))
+		if (__i915_request_is_complete(request))
 			continue;
 
-		if (!i915_request_started(request))
+		if (!__i915_request_has_started(request))
 			continue;
 
 		/* More than one preemptible request may match! */

drivers/gpu/drm/i915/gt/intel_engine_pm.c

@@ -79,7 +79,7 @@ static int __engine_unpark(struct intel_wakeref *wf)
 
 #if IS_ENABLED(CONFIG_LOCKDEP)
 
-static inline unsigned long __timeline_mark_lock(struct intel_context *ce)
+static unsigned long __timeline_mark_lock(struct intel_context *ce)
 {
 	unsigned long flags;
 
@@ -89,8 +89,8 @@ static inline unsigned long __timeline_mark_lock(struct intel_context *ce)
 	return flags;
 }
 
-static inline void __timeline_mark_unlock(struct intel_context *ce,
-					  unsigned long flags)
+static void __timeline_mark_unlock(struct intel_context *ce,
+				   unsigned long flags)
 {
 	mutex_release(&ce->timeline->mutex.dep_map, _THIS_IP_);
 	local_irq_restore(flags);
@@ -98,13 +98,13 @@ static inline void __timeline_mark_unlock(struct intel_context *ce,
 
 #else
 
-static inline unsigned long __timeline_mark_lock(struct intel_context *ce)
+static unsigned long __timeline_mark_lock(struct intel_context *ce)
 {
 	return 0;
 }
 
-static inline void __timeline_mark_unlock(struct intel_context *ce,
-					  unsigned long flags)
+static void __timeline_mark_unlock(struct intel_context *ce,
+				   unsigned long flags)
 {
 }
 

drivers/gpu/drm/i915/gt/intel_engine_stats.h

@@ -0,0 +1,60 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Copyright © 2020 Intel Corporation
+ */
+
+#ifndef __INTEL_ENGINE_STATS_H__
+#define __INTEL_ENGINE_STATS_H__
+
+#include <linux/atomic.h>
+#include <linux/ktime.h>
+#include <linux/seqlock.h>
+
+#include "i915_gem.h" /* GEM_BUG_ON */
+#include "intel_engine.h"
+
+static inline void intel_engine_context_in(struct intel_engine_cs *engine)
+{
+	unsigned long flags;
+
+	if (engine->stats.active) {
+		engine->stats.active++;
+		return;
+	}
+
+	/* The writer is serialised; but the pmu reader may be from hardirq */
+	local_irq_save(flags);
+	write_seqcount_begin(&engine->stats.lock);
+
+	engine->stats.start = ktime_get();
+	engine->stats.active++;
+
+	write_seqcount_end(&engine->stats.lock);
+	local_irq_restore(flags);
+
+	GEM_BUG_ON(!engine->stats.active);
+}
+
+static inline void intel_engine_context_out(struct intel_engine_cs *engine)
+{
+	unsigned long flags;
+
+	GEM_BUG_ON(!engine->stats.active);
+	if (engine->stats.active > 1) {
+		engine->stats.active--;
+		return;
+	}
+
+	local_irq_save(flags);
+	write_seqcount_begin(&engine->stats.lock);
+
+	engine->stats.active--;
+	engine->stats.total =
+		ktime_add(engine->stats.total,
+			  ktime_sub(ktime_get(), engine->stats.start));
+
+	write_seqcount_end(&engine->stats.lock);
+	local_irq_restore(flags);
+}
+
+#endif /* __INTEL_ENGINE_STATS_H__ */

drivers/gpu/drm/i915/gt/intel_engine_types.h

@@ -319,7 +319,7 @@ struct intel_engine_cs {
 	 * as possible.
 	 */
 	enum forcewake_domains fw_domain;
-	atomic_t fw_active;
+	unsigned int fw_active;
 
 	unsigned long context_tag;
 
@@ -516,12 +516,12 @@ struct intel_engine_cs {
 		/**
 		 * @active: Number of contexts currently scheduled in.
 		 */
-		atomic_t active;
+		unsigned int active;
 
 		/**
 		 * @lock: Lock protecting the below fields.
 		 */
-		seqlock_t lock;
+		seqcount_t lock;
 
 		/**
 		 * @total: Total time this engine was busy.

drivers/gpu/drm/i915/gt/intel_execlists_submission.c

@@ -115,6 +115,7 @@
 #include "intel_breadcrumbs.h"
 #include "intel_context.h"
 #include "intel_engine_pm.h"
+#include "intel_engine_stats.h"
 #include "intel_execlists_submission.h"
 #include "intel_gt.h"
 #include "intel_gt_pm.h"
@@ -230,8 +231,7 @@ active_request(const struct intel_timeline * const tl, struct i915_request *rq)
 	return __active_request(tl, rq, 0);
 }
 
-static inline void
-ring_set_paused(const struct intel_engine_cs *engine, int state)
+static void ring_set_paused(const struct intel_engine_cs *engine, int state)
 {
 	/*
 	 * We inspect HWS_PREEMPT with a semaphore inside
@@ -244,12 +244,12 @@ ring_set_paused(const struct intel_engine_cs *engine, int state)
 		wmb();
 }
 
-static inline struct i915_priolist *to_priolist(struct rb_node *rb)
+static struct i915_priolist *to_priolist(struct rb_node *rb)
 {
 	return rb_entry(rb, struct i915_priolist, node);
 }
 
-static inline int rq_prio(const struct i915_request *rq)
+static int rq_prio(const struct i915_request *rq)
 {
 	return READ_ONCE(rq->sched.attr.priority);
 }
@@ -299,8 +299,8 @@ static int virtual_prio(const struct intel_engine_execlists *el)
 	return rb ? rb_entry(rb, struct ve_node, rb)->prio : INT_MIN;
 }
 
-static inline bool need_preempt(const struct intel_engine_cs *engine,
-				const struct i915_request *rq)
+static bool need_preempt(const struct intel_engine_cs *engine,
+			 const struct i915_request *rq)
 {
 	int last_prio;
 
@@ -351,7 +351,7 @@ static inline bool need_preempt(const struct intel_engine_cs *engine,
 		   queue_prio(&engine->execlists)) > last_prio;
 }
 
-__maybe_unused static inline bool
+__maybe_unused static bool
 assert_priority_queue(const struct i915_request *prev,
 		      const struct i915_request *next)
 {
@@ -418,7 +418,7 @@ execlists_unwind_incomplete_requests(struct intel_engine_execlists *execlists)
 	return __unwind_incomplete_requests(engine);
 }
 
-static inline void
+static void
 execlists_context_status_change(struct i915_request *rq, unsigned long status)
 {
 	/*
@@ -432,39 +432,6 @@ execlists_context_status_change(struct i915_request *rq, unsigned long status)
 				   status, rq);
 }
 
-static void intel_engine_context_in(struct intel_engine_cs *engine)
-{
-	unsigned long flags;
-
-	if (atomic_add_unless(&engine->stats.active, 1, 0))
-		return;
-
-	write_seqlock_irqsave(&engine->stats.lock, flags);
-	if (!atomic_add_unless(&engine->stats.active, 1, 0)) {
-		engine->stats.start = ktime_get();
-		atomic_inc(&engine->stats.active);
-	}
-	write_sequnlock_irqrestore(&engine->stats.lock, flags);
-}
-
-static void intel_engine_context_out(struct intel_engine_cs *engine)
-{
-	unsigned long flags;
-
-	GEM_BUG_ON(!atomic_read(&engine->stats.active));
-
-	if (atomic_add_unless(&engine->stats.active, -1, 1))
-		return;
-
-	write_seqlock_irqsave(&engine->stats.lock, flags);
-	if (atomic_dec_and_test(&engine->stats.active)) {
-		engine->stats.total =
-			ktime_add(engine->stats.total,
-				  ktime_sub(ktime_get(), engine->stats.start));
-	}
-	write_sequnlock_irqrestore(&engine->stats.lock, flags);
-}
-
 static void reset_active(struct i915_request *rq,
 			 struct intel_engine_cs *engine)
 {
@@ -503,7 +470,7 @@ static void reset_active(struct i915_request *rq,
 	ce->lrc.lrca = lrc_update_regs(ce, engine, head);
 }
 
-static inline struct intel_engine_cs *
+static struct intel_engine_cs *
 __execlists_schedule_in(struct i915_request *rq)
 {
 	struct intel_engine_cs * const engine = rq->engine;
@@ -539,7 +506,7 @@ __execlists_schedule_in(struct i915_request *rq)
 	ce->lrc.ccid |= engine->execlists.ccid;
 
 	__intel_gt_pm_get(engine->gt);
-	if (engine->fw_domain && !atomic_fetch_inc(&engine->fw_active))
+	if (engine->fw_domain && !engine->fw_active++)
 		intel_uncore_forcewake_get(engine->uncore, engine->fw_domain);
 	execlists_context_status_change(rq, INTEL_CONTEXT_SCHEDULE_IN);
 	intel_engine_context_in(engine);
@@ -549,7 +516,7 @@ __execlists_schedule_in(struct i915_request *rq)
 	return engine;
 }
 
-static inline void execlists_schedule_in(struct i915_request *rq, int idx)
+static void execlists_schedule_in(struct i915_request *rq, int idx)
 {
 	struct intel_context * const ce = rq->context;
 	struct intel_engine_cs *old;
@@ -608,9 +575,9 @@ static void kick_siblings(struct i915_request *rq, struct intel_context *ce)
 		tasklet_hi_schedule(&ve->base.execlists.tasklet);
 }
 
-static inline void __execlists_schedule_out(struct i915_request *rq)
+static void __execlists_schedule_out(struct i915_request * const rq,
+				     struct intel_context * const ce)
 {
-	struct intel_context * const ce = rq->context;
 	struct intel_engine_cs * const engine = rq->engine;
 	unsigned int ccid;
 
@@ -621,6 +588,7 @@ static inline void __execlists_schedule_out(struct i915_request *rq)
 	 */
 
 	CE_TRACE(ce, "schedule-out, ccid:%x\n", ce->lrc.ccid);
+	GEM_BUG_ON(ce->inflight != engine);
 
 	if (IS_ENABLED(CONFIG_DRM_I915_DEBUG_GEM))
 		lrc_check_regs(ce, engine, "after");
@@ -645,7 +613,7 @@ static inline void __execlists_schedule_out(struct i915_request *rq)
 	lrc_update_runtime(ce);
 	intel_engine_context_out(engine);
 	execlists_context_status_change(rq, INTEL_CONTEXT_SCHEDULE_OUT);
-	if (engine->fw_domain && !atomic_dec_return(&engine->fw_active))
+	if (engine->fw_domain && !--engine->fw_active)
 		intel_uncore_forcewake_put(engine->uncore, engine->fw_domain);
 	intel_gt_pm_put_async(engine->gt);
 
@@ -660,10 +628,12 @@ static inline void __execlists_schedule_out(struct i915_request *rq)
 	 */
 	if (ce->engine != engine)
 		kick_siblings(rq, ce);
+
+	WRITE_ONCE(ce->inflight, NULL);
+	intel_context_put(ce);
 }
 
-static inline void
-execlists_schedule_out(struct i915_request *rq)
+static inline void execlists_schedule_out(struct i915_request *rq)
 {
 	struct intel_context * const ce = rq->context;
 
@@ -671,12 +641,8 @@ execlists_schedule_out(struct i915_request *rq)
 
 	GEM_BUG_ON(!ce->inflight);
 	ce->inflight = ptr_dec(ce->inflight);
-	if (!__intel_context_inflight_count(ce->inflight)) {
-		GEM_BUG_ON(ce->inflight != rq->engine);
-		__execlists_schedule_out(rq);
-		WRITE_ONCE(ce->inflight, NULL);
-		intel_context_put(ce);
-	}
+	if (!__intel_context_inflight_count(ce->inflight))
+		__execlists_schedule_out(rq, ce);
 
 	i915_request_put(rq);
 }
@@ -728,7 +694,7 @@ static u64 execlists_update_context(struct i915_request *rq)
 	return desc;
 }
 
-static inline void write_desc(struct intel_engine_execlists *execlists, u64 desc, u32 port)
+static void write_desc(struct intel_engine_execlists *execlists, u64 desc, u32 port)
 {
 	if (execlists->ctrl_reg) {
 		writel(lower_32_bits(desc), execlists->submit_reg + port * 2);
@@ -757,7 +723,7 @@ dump_port(char *buf, int buflen, const char *prefix, struct i915_request *rq)
 	return buf;
 }
 
-static __maybe_unused void
+static __maybe_unused noinline void
 trace_ports(const struct intel_engine_execlists *execlists,
 	    const char *msg,
 	    struct i915_request * const *ports)
@@ -774,13 +740,13 @@ trace_ports(const struct intel_engine_execlists *execlists,
 		     dump_port(p1, sizeof(p1), ", ", ports[1]));
 }
 
-static inline bool
+static bool
 reset_in_progress(const struct intel_engine_execlists *execlists)
 {
 	return unlikely(!__tasklet_is_enabled(&execlists->tasklet));
 }
 
-static __maybe_unused bool
+static __maybe_unused noinline bool
 assert_pending_valid(const struct intel_engine_execlists *execlists,
 		     const char *msg)
 {
@@ -1258,12 +1224,20 @@ static void set_preempt_timeout(struct intel_engine_cs *engine,
 		     active_preempt_timeout(engine, rq));
 }
 
+static bool completed(const struct i915_request *rq)
+{
+	if (i915_request_has_sentinel(rq))
+		return false;
+
+	return __i915_request_is_complete(rq);
+}
+
 static void execlists_dequeue(struct intel_engine_cs *engine)
 {
 	struct intel_engine_execlists * const execlists = &engine->execlists;
 	struct i915_request **port = execlists->pending;
 	struct i915_request ** const last_port = port + execlists->port_mask;
-	struct i915_request *last = *execlists->active;
+	struct i915_request *last, * const *active;
 	struct virtual_engine *ve;
 	struct rb_node *rb;
 	bool submit = false;
@@ -1300,21 +1274,13 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 	 * i.e. we will retrigger preemption following the ack in case
 	 * of trouble.
 	 *
-	 * In theory we can skip over completed contexts that have not
-	 * yet been processed by events (as those events are in flight):
-	 *
-	 * while ((last = *active) && i915_request_completed(last))
-	 *	active++;
-	 *
-	 * However, the GPU cannot handle this as it will ultimately
-	 * find itself trying to jump back into a context it has just
-	 * completed and barf.
 	 */
+	active = execlists->active;
+	while ((last = *active) && completed(last))
+		active++;
 
 	if (last) {
-		if (__i915_request_is_complete(last)) {
-			goto check_secondary;
-		} else if (need_preempt(engine, last)) {
+		if (need_preempt(engine, last)) {
 			ENGINE_TRACE(engine,
 				     "preempting last=%llx:%lld, prio=%d, hint=%d\n",
 				     last->fence.context,
@@ -1393,9 +1359,7 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 			 * we hopefully coalesce several updates into a single
 			 * submission.
 			 */
-check_secondary:
-			if (!list_is_last(&last->sched.link,
-					  &engine->active.requests)) {
+			if (active[1]) {
 				/*
 				 * Even if ELSP[1] is occupied and not worthy
 				 * of timeslices, our queue might be.
@@ -1596,7 +1560,7 @@ done:
 	 * of ordered contexts.
 	 */
 	if (submit &&
-	    memcmp(execlists->active,
+	    memcmp(active,
 		   execlists->pending,
 		   (port - execlists->pending) * sizeof(*port))) {
 		*port = NULL;
@@ -1604,7 +1568,7 @@ done:
 			execlists_schedule_in(*port, port - execlists->pending);
 
 		WRITE_ONCE(execlists->yield, -1);
-		set_preempt_timeout(engine, *execlists->active);
+		set_preempt_timeout(engine, *active);
 		execlists_submit_ports(engine);
 	} else {
 		ring_set_paused(engine, 0);
@@ -1621,12 +1585,12 @@ static void execlists_dequeue_irq(struct intel_engine_cs *engine)
 	local_irq_enable(); /* flush irq_work (e.g. breadcrumb enabling) */
 }
 
-static inline void clear_ports(struct i915_request **ports, int count)
+static void clear_ports(struct i915_request **ports, int count)
 {
 	memset_p((void **)ports, NULL, count);
 }
 
-static inline void
+static void
 copy_ports(struct i915_request **dst, struct i915_request **src, int count)
 {
 	/* A memcpy_p() would be very useful here! */
@@ -1660,8 +1624,7 @@ cancel_port_requests(struct intel_engine_execlists * const execlists,
 	return inactive;
 }
 
-static inline void
-invalidate_csb_entries(const u64 *first, const u64 *last)
+static void invalidate_csb_entries(const u64 *first, const u64 *last)
 {
 	clflush((void *)first);
 	clflush((void *)last);
@@ -1693,7 +1656,7 @@ invalidate_csb_entries(const u64 *first, const u64 *last)
  *     bits 47-57: sw context id of the lrc the GT switched away from
  *     bits 58-63: sw counter of the lrc the GT switched away from
  */
-static inline bool gen12_csb_parse(const u64 csb)
+static bool gen12_csb_parse(const u64 csb)
 {
 	bool ctx_away_valid = GEN12_CSB_CTX_VALID(upper_32_bits(csb));
 	bool new_queue =
@@ -1720,7 +1683,7 @@ static inline bool gen12_csb_parse(const u64 csb)
 	return false;
 }
 
-static inline bool gen8_csb_parse(const u64 csb)
+static bool gen8_csb_parse(const u64 csb)
 {
 	return csb & (GEN8_CTX_STATUS_IDLE_ACTIVE | GEN8_CTX_STATUS_PREEMPTED);
 }
@@ -1759,8 +1722,7 @@ wa_csb_read(const struct intel_engine_cs *engine, u64 * const csb)
 	return entry;
 }
 
-static inline u64
-csb_read(const struct intel_engine_cs *engine, u64 * const csb)
+static u64 csb_read(const struct intel_engine_cs *engine, u64 * const csb)
 {
 	u64 entry = READ_ONCE(*csb);
 
@@ -2026,6 +1988,9 @@ static void __execlists_hold(struct i915_request *rq)
 			struct i915_request *w =
 				container_of(p->waiter, typeof(*w), sched);
 
+			if (p->flags & I915_DEPENDENCY_WEAK)
+				continue;
+
 			/* Leave semaphores spinning on the other engines */
 			if (w->engine != rq->engine)
 				continue;
@@ -2124,6 +2089,9 @@ static void __execlists_unhold(struct i915_request *rq)
 			struct i915_request *w =
 				container_of(p->waiter, typeof(*w), sched);
 
+			if (p->flags & I915_DEPENDENCY_WEAK)
+				continue;
+
 			/* Propagate any change in error status */
 			if (rq->fence.error)
 				i915_request_set_error_once(w, rq->fence.error);
@@ -3180,8 +3148,7 @@ logical_ring_default_vfuncs(struct intel_engine_cs *engine)
 	}
 }
 
-static inline void
-logical_ring_default_irqs(struct intel_engine_cs *engine)
+static void logical_ring_default_irqs(struct intel_engine_cs *engine)
 {
 	unsigned int shift = 0;
 
@@ -3296,7 +3263,7 @@ static void rcu_virtual_context_destroy(struct work_struct *wrk)
 
 		old = fetch_and_zero(&ve->request);
 		if (old) {
-			GEM_BUG_ON(!i915_request_completed(old));
+			GEM_BUG_ON(!__i915_request_is_complete(old));
 			__i915_request_submit(old);
 			i915_request_put(old);
 		}
@@ -3573,7 +3540,7 @@ static void virtual_submit_request(struct i915_request *rq)
 	}
 
 	if (ve->request) { /* background completion from preempt-to-busy */
-		GEM_BUG_ON(!i915_request_completed(ve->request));
+		GEM_BUG_ON(!__i915_request_is_complete(ve->request));
 		__i915_request_submit(ve->request);
 		i915_request_put(ve->request);
 	}

drivers/gpu/drm/i915/gt/intel_gt_buffer_pool.c

@@ -145,7 +145,8 @@ static void pool_retire(struct i915_active *ref)
 }
 
 static struct intel_gt_buffer_pool_node *
-node_create(struct intel_gt_buffer_pool *pool, size_t sz)
+node_create(struct intel_gt_buffer_pool *pool, size_t sz,
+	    enum i915_map_type type)
 {
 	struct intel_gt *gt = to_gt(pool);
 	struct intel_gt_buffer_pool_node *node;
@@ -169,12 +170,14 @@ node_create(struct intel_gt_buffer_pool *pool, size_t sz)
 
 	i915_gem_object_set_readonly(obj);
 
+	node->type = type;
 	node->obj = obj;
 	return node;
 }
 
 struct intel_gt_buffer_pool_node *
-intel_gt_get_buffer_pool(struct intel_gt *gt, size_t size)
+intel_gt_get_buffer_pool(struct intel_gt *gt, size_t size,
+			 enum i915_map_type type)
 {
 	struct intel_gt_buffer_pool *pool = &gt->buffer_pool;
 	struct intel_gt_buffer_pool_node *node;
@@ -191,6 +194,9 @@ intel_gt_get_buffer_pool(struct intel_gt *gt, size_t size)
 		if (node->obj->base.size < size)
 			continue;
 
+		if (node->type != type)
+			continue;
+
 		age = READ_ONCE(node->age);
 		if (!age)
 			continue;
@@ -205,7 +211,7 @@ intel_gt_get_buffer_pool(struct intel_gt *gt, size_t size)
 	rcu_read_unlock();
 
 	if (&node->link == list) {
-		node = node_create(pool, size);
+		node = node_create(pool, size, type);
 		if (IS_ERR(node))
 			return node;
 	}

drivers/gpu/drm/i915/gt/intel_gt_buffer_pool.h

@@ -15,7 +15,8 @@ struct intel_gt;
 struct i915_request;
 
 struct intel_gt_buffer_pool_node *
-intel_gt_get_buffer_pool(struct intel_gt *gt, size_t size);
+intel_gt_get_buffer_pool(struct intel_gt *gt, size_t size,
+			 enum i915_map_type type);
 
 static inline int
 intel_gt_buffer_pool_mark_active(struct intel_gt_buffer_pool_node *node,

drivers/gpu/drm/i915/gt/intel_gt_buffer_pool_types.h

@@ -11,10 +11,9 @@
 #include <linux/spinlock.h>
 #include <linux/workqueue.h>
 
+#include "gem/i915_gem_object_types.h"
 #include "i915_active_types.h"
 
-struct drm_i915_gem_object;
-
 struct intel_gt_buffer_pool {
 	spinlock_t lock;
 	struct list_head cache_list[4];
@@ -31,6 +30,7 @@ struct intel_gt_buffer_pool_node {
 		struct rcu_head rcu;
 	};
 	unsigned long age;
+	enum i915_map_type type;
 };
 
 #endif /* INTEL_GT_BUFFER_POOL_TYPES_H */

drivers/gpu/drm/i915/gt/intel_lrc.c

@@ -1035,7 +1035,7 @@ gen12_emit_indirect_ctx_xcs(const struct intel_context *ce, u32 *cs)
 	return cs;
 }
 
-static inline u32 context_wa_bb_offset(const struct intel_context *ce)
+static u32 context_wa_bb_offset(const struct intel_context *ce)
 {
 	return PAGE_SIZE * ce->wa_bb_page;
 }
@@ -1098,7 +1098,7 @@ setup_indirect_ctx_bb(const struct intel_context *ce,
  * engine info, SW context ID and SW counter need to form a unique number
  * (Context ID) per lrc.
  */
-static inline u32 lrc_descriptor(const struct intel_context *ce)
+static u32 lrc_descriptor(const struct intel_context *ce)
 {
 	u32 desc;
 

drivers/gpu/drm/i915/gt/intel_mocs.c

@@ -472,7 +472,7 @@ static u16 get_entry_l3cc(const struct drm_i915_mocs_table *table,
 	return table->table[I915_MOCS_PTE].l3cc_value;
 }
 
-static inline u32 l3cc_combine(u16 low, u16 high)
+static u32 l3cc_combine(u16 low, u16 high)
 {
 	return low | (u32)high << 16;
 }

drivers/gpu/drm/i915/gt/intel_ppgtt.c

@@ -80,7 +80,7 @@ void free_px(struct i915_address_space *vm, struct i915_page_table *pt, int lvl)
 	kfree(pt);
 }
 
-static inline void
+static void
 write_dma_entry(struct drm_i915_gem_object * const pdma,
 		const unsigned short idx,
 		const u64 encoded_entry)

drivers/gpu/drm/i915/gt/intel_rc6.c

@@ -49,7 +49,7 @@ static struct drm_i915_private *rc6_to_i915(struct intel_rc6 *rc)
 	return rc6_to_gt(rc)->i915;
 }
 
-static inline void set(struct intel_uncore *uncore, i915_reg_t reg, u32 val)
+static void set(struct intel_uncore *uncore, i915_reg_t reg, u32 val)
 {
 	intel_uncore_write_fw(uncore, reg, val);
 }

drivers/gpu/drm/i915/gt/intel_region_lmem.c

@@ -98,7 +98,7 @@ region_lmem_init(struct intel_memory_region *mem)
 static const struct intel_memory_region_ops intel_region_lmem_ops = {
 	.init = region_lmem_init,
 	.release = region_lmem_release,
-	.create_object = __i915_gem_lmem_object_create,
+	.init_object = __i915_gem_lmem_object_init,
 };
 
 struct intel_memory_region *

drivers/gpu/drm/i915/gt/intel_reset.c

@@ -151,8 +151,7 @@ static void mark_innocent(struct i915_request *rq)
 void __i915_request_reset(struct i915_request *rq, bool guilty)
 {
 	RQ_TRACE(rq, "guilty? %s\n", yesno(guilty));
-
-	GEM_BUG_ON(i915_request_completed(rq));
+	GEM_BUG_ON(__i915_request_is_complete(rq));
 
 	rcu_read_lock(); /* protect the GEM context */
 	if (guilty) {
@@ -1110,7 +1109,7 @@ error:
 	goto finish;
 }
 
-static inline int intel_gt_reset_engine(struct intel_engine_cs *engine)
+static int intel_gt_reset_engine(struct intel_engine_cs *engine)
 {
 	return __intel_gt_reset(engine->gt, engine->mask);
 }

drivers/gpu/drm/i915/gt/intel_ring.c

@@ -42,7 +42,7 @@ int intel_ring_pin(struct intel_ring *ring, struct i915_gem_ww_ctx *ww)
 	/* Ring wraparound at offset 0 sometimes hangs. No idea why. */
 	flags = PIN_OFFSET_BIAS | i915_ggtt_pin_bias(vma);
 
-	if (vma->obj->stolen)
+	if (i915_gem_object_is_stolen(vma->obj))
 		flags |= PIN_MAPPABLE;
 	else
 		flags |= PIN_HIGH;

drivers/gpu/drm/i915/gt/intel_ring_submission.c

@@ -122,31 +122,27 @@ static void set_hwsp(struct intel_engine_cs *engine, u32 offset)
 		hwsp = RING_HWS_PGA(engine->mmio_base);
 	}
 
-	intel_uncore_write(engine->uncore, hwsp, offset);
-	intel_uncore_posting_read(engine->uncore, hwsp);
+	intel_uncore_write_fw(engine->uncore, hwsp, offset);
+	intel_uncore_posting_read_fw(engine->uncore, hwsp);
 }
 
 static void flush_cs_tlb(struct intel_engine_cs *engine)
 {
-	struct drm_i915_private *dev_priv = engine->i915;
-
-	if (!IS_GEN_RANGE(dev_priv, 6, 7))
+	if (!IS_GEN_RANGE(engine->i915, 6, 7))
 		return;
 
 	/* ring should be idle before issuing a sync flush*/
-	drm_WARN_ON(&dev_priv->drm,
-		    (ENGINE_READ(engine, RING_MI_MODE) & MODE_IDLE) == 0);
+	GEM_DEBUG_WARN_ON((ENGINE_READ(engine, RING_MI_MODE) & MODE_IDLE) == 0);
 
-	ENGINE_WRITE(engine, RING_INSTPM,
-		     _MASKED_BIT_ENABLE(INSTPM_TLB_INVALIDATE |
-					INSTPM_SYNC_FLUSH));
-	if (intel_wait_for_register(engine->uncore,
-				    RING_INSTPM(engine->mmio_base),
-				    INSTPM_SYNC_FLUSH, 0,
-				    1000))
-		drm_err(&dev_priv->drm,
-			"%s: wait for SyncFlush to complete for TLB invalidation timed out\n",
-			engine->name);
+	ENGINE_WRITE_FW(engine, RING_INSTPM,
+			_MASKED_BIT_ENABLE(INSTPM_TLB_INVALIDATE |
+					   INSTPM_SYNC_FLUSH));
+	if (__intel_wait_for_register_fw(engine->uncore,
+					 RING_INSTPM(engine->mmio_base),
+					 INSTPM_SYNC_FLUSH, 0,
+					 2000, 0, NULL))
+		ENGINE_TRACE(engine,
+			     "wait for SyncFlush to complete for TLB invalidation timed out\n");
 }
 
 static void ring_setup_status_page(struct intel_engine_cs *engine)
@@ -157,21 +153,6 @@ static void ring_setup_status_page(struct intel_engine_cs *engine)
 	flush_cs_tlb(engine);
 }
 
-static bool stop_ring(struct intel_engine_cs *engine)
-{
-	intel_engine_stop_cs(engine);
-
-	ENGINE_WRITE(engine, RING_HEAD, ENGINE_READ(engine, RING_TAIL));
-
-	ENGINE_WRITE(engine, RING_HEAD, 0);
-	ENGINE_WRITE(engine, RING_TAIL, 0);
-
-	/* The ring must be empty before it is disabled */
-	ENGINE_WRITE(engine, RING_CTL, 0);
-
-	return (ENGINE_READ(engine, RING_HEAD) & HEAD_ADDR) == 0;
-}
-
 static struct i915_address_space *vm_alias(struct i915_address_space *vm)
 {
 	if (i915_is_ggtt(vm))
@@ -189,9 +170,16 @@ static void set_pp_dir(struct intel_engine_cs *engine)
 {
 	struct i915_address_space *vm = vm_alias(engine->gt->vm);
 
-	if (vm) {
-		ENGINE_WRITE(engine, RING_PP_DIR_DCLV, PP_DIR_DCLV_2G);
-		ENGINE_WRITE(engine, RING_PP_DIR_BASE, pp_dir(vm));
+	if (!vm)
+		return;
+
+	ENGINE_WRITE_FW(engine, RING_PP_DIR_DCLV, PP_DIR_DCLV_2G);
+	ENGINE_WRITE_FW(engine, RING_PP_DIR_BASE, pp_dir(vm));
+
+	if (INTEL_GEN(engine->i915) >= 7) {
+		ENGINE_WRITE_FW(engine,
+				RING_MODE_GEN7,
+				_MASKED_BIT_ENABLE(GFX_PPGTT_ENABLE));
 	}
 }
 
@@ -199,38 +187,10 @@ static int xcs_resume(struct intel_engine_cs *engine)
 {
 	struct drm_i915_private *dev_priv = engine->i915;
 	struct intel_ring *ring = engine->legacy.ring;
-	int ret = 0;
 
 	ENGINE_TRACE(engine, "ring:{HEAD:%04x, TAIL:%04x}\n",
 		     ring->head, ring->tail);
 
-	intel_uncore_forcewake_get(engine->uncore, FORCEWAKE_ALL);
-
-	/* WaClearRingBufHeadRegAtInit:ctg,elk */
-	if (!stop_ring(engine)) {
-		/* G45 ring initialization often fails to reset head to zero */
-		drm_dbg(&dev_priv->drm, "%s head not reset to zero "
-			"ctl %08x head %08x tail %08x start %08x\n",
-			engine->name,
-			ENGINE_READ(engine, RING_CTL),
-			ENGINE_READ(engine, RING_HEAD),
-			ENGINE_READ(engine, RING_TAIL),
-			ENGINE_READ(engine, RING_START));
-
-		if (!stop_ring(engine)) {
-			drm_err(&dev_priv->drm,
-				"failed to set %s head to zero "
-				"ctl %08x head %08x tail %08x start %08x\n",
-				engine->name,
-				ENGINE_READ(engine, RING_CTL),
-				ENGINE_READ(engine, RING_HEAD),
-				ENGINE_READ(engine, RING_TAIL),
-				ENGINE_READ(engine, RING_START));
-			ret = -EIO;
-			goto out;
-		}
-	}
-
 	if (HWS_NEEDS_PHYSICAL(dev_priv))
 		ring_setup_phys_status_page(engine);
 	else
@@ -247,7 +207,7 @@ static int xcs_resume(struct intel_engine_cs *engine)
 	 * also enforces ordering), otherwise the hw might lose the new ring
 	 * register values.
 	 */
-	ENGINE_WRITE(engine, RING_START, i915_ggtt_offset(ring->vma));
+	ENGINE_WRITE_FW(engine, RING_START, i915_ggtt_offset(ring->vma));
 
 	/* Check that the ring offsets point within the ring! */
 	GEM_BUG_ON(!intel_ring_offset_valid(ring, ring->head));
@@ -257,46 +217,44 @@ static int xcs_resume(struct intel_engine_cs *engine)
 	set_pp_dir(engine);
 
 	/* First wake the ring up to an empty/idle ring */
-	ENGINE_WRITE(engine, RING_HEAD, ring->head);
-	ENGINE_WRITE(engine, RING_TAIL, ring->head);
+	ENGINE_WRITE_FW(engine, RING_HEAD, ring->head);
+	ENGINE_WRITE_FW(engine, RING_TAIL, ring->head);
 	ENGINE_POSTING_READ(engine, RING_TAIL);
 
-	ENGINE_WRITE(engine, RING_CTL, RING_CTL_SIZE(ring->size) | RING_VALID);
+	ENGINE_WRITE_FW(engine, RING_CTL,
+			RING_CTL_SIZE(ring->size) | RING_VALID);
 
 	/* If the head is still not zero, the ring is dead */
-	if (intel_wait_for_register(engine->uncore,
-				    RING_CTL(engine->mmio_base),
-				    RING_VALID, RING_VALID,
-				    50)) {
-		drm_err(&dev_priv->drm, "%s initialization failed "
-			  "ctl %08x (valid? %d) head %08x [%08x] tail %08x [%08x] start %08x [expected %08x]\n",
-			  engine->name,
-			  ENGINE_READ(engine, RING_CTL),
-			  ENGINE_READ(engine, RING_CTL) & RING_VALID,
-			  ENGINE_READ(engine, RING_HEAD), ring->head,
-			  ENGINE_READ(engine, RING_TAIL), ring->tail,
-			  ENGINE_READ(engine, RING_START),
-			  i915_ggtt_offset(ring->vma));
-		ret = -EIO;
-		goto out;
+	if (__intel_wait_for_register_fw(engine->uncore,
+					 RING_CTL(engine->mmio_base),
+					 RING_VALID, RING_VALID,
+					 5000, 0, NULL)) {
+		drm_err(&dev_priv->drm,
+			"%s initialization failed; "
+			"ctl %08x (valid? %d) head %08x [%08x] tail %08x [%08x] start %08x [expected %08x]\n",
+			engine->name,
+			ENGINE_READ(engine, RING_CTL),
+			ENGINE_READ(engine, RING_CTL) & RING_VALID,
+			ENGINE_READ(engine, RING_HEAD), ring->head,
+			ENGINE_READ(engine, RING_TAIL), ring->tail,
+			ENGINE_READ(engine, RING_START),
+			i915_ggtt_offset(ring->vma));
+		return -EIO;
 	}
 
 	if (INTEL_GEN(dev_priv) > 2)
-		ENGINE_WRITE(engine,
-			     RING_MI_MODE, _MASKED_BIT_DISABLE(STOP_RING));
+		ENGINE_WRITE_FW(engine,
+				RING_MI_MODE, _MASKED_BIT_DISABLE(STOP_RING));
 
 	/* Now awake, let it get started */
 	if (ring->tail != ring->head) {
-		ENGINE_WRITE(engine, RING_TAIL, ring->tail);
+		ENGINE_WRITE_FW(engine, RING_TAIL, ring->tail);
 		ENGINE_POSTING_READ(engine, RING_TAIL);
 	}
 
 	/* Papering over lost _interrupts_ immediately following the restart */
 	intel_engine_signal_breadcrumbs(engine);
-out:
-	intel_uncore_forcewake_put(engine->uncore, FORCEWAKE_ALL);
-
-	return ret;
+	return 0;
 }
 
 static void sanitize_hwsp(struct intel_engine_cs *engine)
@@ -332,11 +290,25 @@ static void xcs_sanitize(struct intel_engine_cs *engine)
 	clflush_cache_range(engine->status_page.addr, PAGE_SIZE);
 }
 
+static bool stop_ring(struct intel_engine_cs *engine)
+{
+	/* Empty the ring by skipping to the end */
+	ENGINE_WRITE_FW(engine, RING_HEAD, ENGINE_READ_FW(engine, RING_TAIL));
+	ENGINE_POSTING_READ(engine, RING_HEAD);
+
+	/* The ring must be empty before it is disabled */
+	ENGINE_WRITE_FW(engine, RING_CTL, 0);
+	ENGINE_POSTING_READ(engine, RING_CTL);
+
+	/* Then reset the disabled ring */
+	ENGINE_WRITE_FW(engine, RING_HEAD, 0);
+	ENGINE_WRITE_FW(engine, RING_TAIL, 0);
+
+	return (ENGINE_READ_FW(engine, RING_HEAD) & HEAD_ADDR) == 0;
+}
+
 static void reset_prepare(struct intel_engine_cs *engine)
 {
-	struct intel_uncore *uncore = engine->uncore;
-	const u32 base = engine->mmio_base;
-
 	/*
 	 * We stop engines, otherwise we might get failed reset and a
 	 * dead gpu (on elk). Also as modern gpu as kbl can suffer
@@ -348,30 +320,35 @@ static void reset_prepare(struct intel_engine_cs *engine)
 	 * WaKBLVECSSemaphoreWaitPoll:kbl (on ALL_ENGINES)
 	 *
 	 * WaMediaResetMainRingCleanup:ctg,elk (presumably)
+	 * WaClearRingBufHeadRegAtInit:ctg,elk
 	 *
 	 * FIXME: Wa for more modern gens needs to be validated
 	 */
 	ENGINE_TRACE(engine, "\n");
+	intel_engine_stop_cs(engine);
 
-	if (intel_engine_stop_cs(engine))
-		ENGINE_TRACE(engine, "timed out on STOP_RING\n");
+	if (!stop_ring(engine)) {
+		/* G45 ring initialization often fails to reset head to zero */
+		drm_dbg(&engine->i915->drm,
+			"%s head not reset to zero "
+			"ctl %08x head %08x tail %08x start %08x\n",
+			engine->name,
+			ENGINE_READ_FW(engine, RING_CTL),
+			ENGINE_READ_FW(engine, RING_HEAD),
+			ENGINE_READ_FW(engine, RING_TAIL),
+			ENGINE_READ_FW(engine, RING_START));
+	}
 
-	intel_uncore_write_fw(uncore,
-			      RING_HEAD(base),
-			      intel_uncore_read_fw(uncore, RING_TAIL(base)));
-	intel_uncore_posting_read_fw(uncore, RING_HEAD(base)); /* paranoia */
-
-	intel_uncore_write_fw(uncore, RING_HEAD(base), 0);
-	intel_uncore_write_fw(uncore, RING_TAIL(base), 0);
-	intel_uncore_posting_read_fw(uncore, RING_TAIL(base));
-
-	/* The ring must be empty before it is disabled */
-	intel_uncore_write_fw(uncore, RING_CTL(base), 0);
-
-	/* Check acts as a post */
-	if (intel_uncore_read_fw(uncore, RING_HEAD(base)))
-		ENGINE_TRACE(engine, "ring head [%x] not parked\n",
-			     intel_uncore_read_fw(uncore, RING_HEAD(base)));
+	if (!stop_ring(engine)) {
+		drm_err(&engine->i915->drm,
+			"failed to set %s head to zero "
+			"ctl %08x head %08x tail %08x start %08x\n",
+			engine->name,
+			ENGINE_READ_FW(engine, RING_CTL),
+			ENGINE_READ_FW(engine, RING_HEAD),
+			ENGINE_READ_FW(engine, RING_TAIL),
+			ENGINE_READ_FW(engine, RING_START));
+	}
 }
 
 static void reset_rewind(struct intel_engine_cs *engine, bool stalled)
@@ -382,12 +359,14 @@ static void reset_rewind(struct intel_engine_cs *engine, bool stalled)
 
 	rq = NULL;
 	spin_lock_irqsave(&engine->active.lock, flags);
+	rcu_read_lock();
 	list_for_each_entry(pos, &engine->active.requests, sched.link) {
-		if (!i915_request_completed(pos)) {
+		if (!__i915_request_is_complete(pos)) {
 			rq = pos;
 			break;
 		}
 	}
+	rcu_read_unlock();
 
 	/*
 	 * The guilty request will get skipped on a hung engine.
@@ -663,9 +642,9 @@ static int load_pd_dir(struct i915_request *rq,
 	return rq->engine->emit_flush(rq, EMIT_FLUSH);
 }
 
-static inline int mi_set_context(struct i915_request *rq,
-				 struct intel_context *ce,
-				 u32 flags)
+static int mi_set_context(struct i915_request *rq,
+			  struct intel_context *ce,
+			  u32 flags)
 {
 	struct intel_engine_cs *engine = rq->engine;
 	struct drm_i915_private *i915 = engine->i915;

drivers/gpu/drm/i915/gt/intel_rps.c

@@ -43,7 +43,7 @@ static u32 rps_pm_sanitize_mask(struct intel_rps *rps, u32 mask)
 	return mask & ~rps->pm_intrmsk_mbz;
 }
 
-static inline void set(struct intel_uncore *uncore, i915_reg_t reg, u32 val)
+static void set(struct intel_uncore *uncore, i915_reg_t reg, u32 val)
 {
 	intel_uncore_write_fw(uncore, reg, val);
 }

drivers/gpu/drm/i915/gt/intel_timeline.c

@@ -582,11 +582,11 @@ int intel_timeline_read_hwsp(struct i915_request *from,
 
 	rcu_read_lock();
 	cl = rcu_dereference(from->hwsp_cacheline);
-	if (i915_request_completed(from)) /* confirm cacheline is valid */
+	if (i915_request_signaled(from)) /* confirm cacheline is valid */
 		goto unlock;
 	if (unlikely(!i915_active_acquire_if_busy(&cl->active)))
 		goto unlock; /* seqno wrapped and completed! */
-	if (unlikely(i915_request_completed(from)))
+	if (unlikely(__i915_request_is_complete(from)))
 		goto release;
 	rcu_read_unlock();
 

drivers/gpu/drm/i915/gt/intel_workarounds.c

@@ -1304,7 +1304,7 @@ bool intel_gt_verify_workarounds(struct intel_gt *gt, const char *from)
 }
 
 __maybe_unused
-static inline bool is_nonpriv_flags_valid(u32 flags)
+static bool is_nonpriv_flags_valid(u32 flags)
 {
 	/* Check only valid flag bits are set */
 	if (flags & ~RING_FORCE_TO_NONPRIV_MASK_VALID)

drivers/gpu/drm/i915/gt/selftest_hangcheck.c

@@ -704,6 +704,7 @@ static int __igt_reset_engine(struct intel_gt *gt, bool active)
 
 	for_each_engine(engine, gt, id) {
 		unsigned int reset_count, reset_engine_count;
+		unsigned long count;
 		IGT_TIMEOUT(end_time);
 
 		if (active && !intel_engine_can_store_dword(engine))
@@ -721,6 +722,7 @@ static int __igt_reset_engine(struct intel_gt *gt, bool active)
 
 		st_engine_heartbeat_disable(engine);
 		set_bit(I915_RESET_ENGINE + id, &gt->reset.flags);
+		count = 0;
 		do {
 			if (active) {
 				struct i915_request *rq;
@@ -770,9 +772,13 @@ static int __igt_reset_engine(struct intel_gt *gt, bool active)
 				err = -EINVAL;
 				break;
 			}
+
+			count++;
 		} while (time_before(jiffies, end_time));
 		clear_bit(I915_RESET_ENGINE + id, &gt->reset.flags);
 		st_engine_heartbeat_enable(engine);
+		pr_info("%s: Completed %lu %s resets\n",
+			engine->name, count, active ? "active" : "idle");
 
 		if (err)
 			break;
@@ -1623,7 +1629,8 @@ static int igt_reset_queue(void *arg)
 			prev = rq;
 			count++;
 		} while (time_before(jiffies, end_time));
-		pr_info("%s: Completed %d resets\n", engine->name, count);
+		pr_info("%s: Completed %d queued resets\n",
+			engine->name, count);
 
 		*h.batch = MI_BATCH_BUFFER_END;
 		intel_gt_chipset_flush(engine->gt);
@@ -1720,7 +1727,8 @@ static int __igt_atomic_reset_engine(struct intel_engine_cs *engine,
 	GEM_TRACE("i915_reset_engine(%s:%s) under %s\n",
 		  engine->name, mode, p->name);
 
-	tasklet_disable(t);
+	if (t->func)
+		tasklet_disable(t);
 	if (strcmp(p->name, "softirq"))
 		local_bh_disable();
 	p->critical_section_begin();
@@ -1730,8 +1738,10 @@ static int __igt_atomic_reset_engine(struct intel_engine_cs *engine,
 	p->critical_section_end();
 	if (strcmp(p->name, "softirq"))
 		local_bh_enable();
-	tasklet_enable(t);
-	tasklet_hi_schedule(t);
+	if (t->func) {
+		tasklet_enable(t);
+		tasklet_hi_schedule(t);
+	}
 
 	if (err)
 		pr_err("i915_reset_engine(%s:%s) failed under %s\n",

drivers/gpu/drm/i915/gt/selftest_reset.c

@@ -321,7 +321,10 @@ static int igt_atomic_engine_reset(void *arg)
 		goto out_unlock;
 
 	for_each_engine(engine, gt, id) {
-		tasklet_disable(&engine->execlists.tasklet);
+		struct tasklet_struct *t = &engine->execlists.tasklet;
+
+		if (t->func)
+			tasklet_disable(t);
 		intel_engine_pm_get(engine);
 
 		for (p = igt_atomic_phases; p->name; p++) {
@@ -345,8 +348,10 @@ static int igt_atomic_engine_reset(void *arg)
 		}
 
 		intel_engine_pm_put(engine);
-		tasklet_enable(&engine->execlists.tasklet);
-		tasklet_hi_schedule(&engine->execlists.tasklet);
+		if (t->func) {
+			tasklet_enable(t);
+			tasklet_hi_schedule(t);
+		}
 		if (err)
 			break;
 	}

drivers/gpu/drm/i915/gt/shmem_utils.c

@@ -33,7 +33,7 @@ struct file *shmem_create_from_object(struct drm_i915_gem_object *obj)
 	struct file *file;
 	void *ptr;
 
-	if (obj->ops == &i915_gem_shmem_ops) {
+	if (i915_gem_object_is_shmem(obj)) {
 		file = obj->base.filp;
 		atomic_long_inc(&file->f_count);
 		return file;

drivers/gpu/drm/i915/gt/uc/intel_uc.c

@@ -15,6 +15,29 @@
 static const struct intel_uc_ops uc_ops_off;
 static const struct intel_uc_ops uc_ops_on;
 
+static void uc_expand_default_options(struct intel_uc *uc)
+{
+	struct drm_i915_private *i915 = uc_to_gt(uc)->i915;
+
+	if (i915->params.enable_guc != -1)
+		return;
+
+	/* Don't enable GuC/HuC on pre-Gen12 */
+	if (INTEL_GEN(i915) < 12) {
+		i915->params.enable_guc = 0;
+		return;
+	}
+
+	/* Don't enable GuC/HuC on older Gen12 platforms */
+	if (IS_TIGERLAKE(i915) || IS_ROCKETLAKE(i915)) {
+		i915->params.enable_guc = 0;
+		return;
+	}
+
+	/* Default: enable HuC authentication only */
+	i915->params.enable_guc = ENABLE_GUC_LOAD_HUC;
+}
+
 /* Reset GuC providing us with fresh state for both GuC and HuC.
  */
 static int __intel_uc_reset_hw(struct intel_uc *uc)
@@ -52,9 +75,6 @@ static void __confirm_options(struct intel_uc *uc)
 		yesno(intel_uc_wants_guc_submission(uc)),
 		yesno(intel_uc_wants_huc(uc)));
 
-	if (i915->params.enable_guc == -1)
-		return;
-
 	if (i915->params.enable_guc == 0) {
 		GEM_BUG_ON(intel_uc_wants_guc(uc));
 		GEM_BUG_ON(intel_uc_wants_guc_submission(uc));
@@ -79,8 +99,7 @@ static void __confirm_options(struct intel_uc *uc)
 			 "Incompatible option enable_guc=%d - %s\n",
 			 i915->params.enable_guc, "GuC submission is N/A");
 
-	if (i915->params.enable_guc & ~(ENABLE_GUC_SUBMISSION |
-					  ENABLE_GUC_LOAD_HUC))
+	if (i915->params.enable_guc & ~ENABLE_GUC_MASK)
 		drm_info(&i915->drm,
 			 "Incompatible option enable_guc=%d - %s\n",
 			 i915->params.enable_guc, "undocumented flag");
@@ -88,6 +107,8 @@ static void __confirm_options(struct intel_uc *uc)
 
 void intel_uc_init_early(struct intel_uc *uc)
 {
+	uc_expand_default_options(uc);
+
 	intel_guc_init_early(&uc->guc);
 	intel_huc_init_early(&uc->huc);
 

drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c

@@ -152,16 +152,11 @@ __uc_fw_auto_select(struct drm_i915_private *i915, struct intel_uc_fw *uc_fw)
 			uc_fw->path = NULL;
 		}
 	}
-
-	/* We don't want to enable GuC/HuC on pre-Gen11 by default */
-	if (i915->params.enable_guc == -1 && p < INTEL_ICELAKE)
-		uc_fw->path = NULL;
 }
 
 static const char *__override_guc_firmware_path(struct drm_i915_private *i915)
 {
-	if (i915->params.enable_guc & (ENABLE_GUC_SUBMISSION |
-				       ENABLE_GUC_LOAD_HUC))
+	if (i915->params.enable_guc & ENABLE_GUC_MASK)
 		return i915->params.guc_firmware_path;
 	return "";
 }

drivers/gpu/drm/i915/gvt/cmd_parser.c

@@ -38,11 +38,17 @@
 
 #include "i915_drv.h"
 #include "gt/intel_gpu_commands.h"
+#include "gt/intel_lrc.h"
 #include "gt/intel_ring.h"
+#include "gt/intel_gt_requests.h"
 #include "gvt.h"
 #include "i915_pvinfo.h"
 #include "trace.h"
 
+#include "gem/i915_gem_context.h"
+#include "gem/i915_gem_pm.h"
+#include "gt/intel_context.h"
+
 #define INVALID_OP    (~0U)
 
 #define OP_LEN_MI           9
@@ -455,6 +461,7 @@ enum {
 	RING_BUFFER_INSTRUCTION,
 	BATCH_BUFFER_INSTRUCTION,
 	BATCH_BUFFER_2ND_LEVEL,
+	RING_BUFFER_CTX,
 };
 
 enum {
@@ -496,6 +503,7 @@ struct parser_exec_state {
 	 */
 	int saved_buf_addr_type;
 	bool is_ctx_wa;
+	bool is_init_ctx;
 
 	const struct cmd_info *info;
 
@@ -709,6 +717,11 @@ static inline u32 cmd_val(struct parser_exec_state *s, int index)
 	return *cmd_ptr(s, index);
 }
 
+static inline bool is_init_ctx(struct parser_exec_state *s)
+{
+	return (s->buf_type == RING_BUFFER_CTX && s->is_init_ctx);
+}
+
 static void parser_exec_state_dump(struct parser_exec_state *s)
 {
 	int cnt = 0;
@@ -722,7 +735,8 @@ static void parser_exec_state_dump(struct parser_exec_state *s)
 
 	gvt_dbg_cmd("  %s %s ip_gma(%08lx) ",
 			s->buf_type == RING_BUFFER_INSTRUCTION ?
-			"RING_BUFFER" : "BATCH_BUFFER",
+			"RING_BUFFER" : ((s->buf_type == RING_BUFFER_CTX) ?
+				"CTX_BUFFER" : "BATCH_BUFFER"),
 			s->buf_addr_type == GTT_BUFFER ?
 			"GTT" : "PPGTT", s->ip_gma);
 
@@ -757,7 +771,8 @@ static inline void update_ip_va(struct parser_exec_state *s)
 	if (WARN_ON(s->ring_head == s->ring_tail))
 		return;
 
-	if (s->buf_type == RING_BUFFER_INSTRUCTION) {
+	if (s->buf_type == RING_BUFFER_INSTRUCTION ||
+			s->buf_type == RING_BUFFER_CTX) {
 		unsigned long ring_top = s->ring_start + s->ring_size;
 
 		if (s->ring_head > s->ring_tail) {
@@ -821,68 +836,12 @@ static inline int cmd_length(struct parser_exec_state *s)
 	*addr = val; \
 } while (0)
 
-static bool is_shadowed_mmio(unsigned int offset)
-{
-	bool ret = false;
-
-	if ((offset == 0x2168) || /*BB current head register UDW */
-	    (offset == 0x2140) || /*BB current header register */
-	    (offset == 0x211c) || /*second BB header register UDW */
-	    (offset == 0x2114)) { /*second BB header register UDW */
-		ret = true;
-	}
-	return ret;
-}
-
-static inline bool is_force_nonpriv_mmio(unsigned int offset)
-{
-	return (offset >= 0x24d0 && offset < 0x2500);
-}
-
-static int force_nonpriv_reg_handler(struct parser_exec_state *s,
-		unsigned int offset, unsigned int index, char *cmd)
-{
-	struct intel_gvt *gvt = s->vgpu->gvt;
-	unsigned int data;
-	u32 ring_base;
-	u32 nopid;
-
-	if (!strcmp(cmd, "lri"))
-		data = cmd_val(s, index + 1);
-	else {
-		gvt_err("Unexpected forcenonpriv 0x%x write from cmd %s\n",
-			offset, cmd);
-		return -EINVAL;
-	}
-
-	ring_base = s->engine->mmio_base;
-	nopid = i915_mmio_reg_offset(RING_NOPID(ring_base));
-
-	if (!intel_gvt_in_force_nonpriv_whitelist(gvt, data) &&
-			data != nopid) {
-		gvt_err("Unexpected forcenonpriv 0x%x LRI write, value=0x%x\n",
-			offset, data);
-		patch_value(s, cmd_ptr(s, index), nopid);
-		return 0;
-	}
-	return 0;
-}
-
 static inline bool is_mocs_mmio(unsigned int offset)
 {
 	return ((offset >= 0xc800) && (offset <= 0xcff8)) ||
 		((offset >= 0xb020) && (offset <= 0xb0a0));
 }
 
-static int mocs_cmd_reg_handler(struct parser_exec_state *s,
-				unsigned int offset, unsigned int index)
-{
-	if (!is_mocs_mmio(offset))
-		return -EINVAL;
-	vgpu_vreg(s->vgpu, offset) = cmd_val(s, index + 1);
-	return 0;
-}
-
 static int is_cmd_update_pdps(unsigned int offset,
 			      struct parser_exec_state *s)
 {
@@ -930,6 +889,7 @@ static int cmd_reg_handler(struct parser_exec_state *s,
 	struct intel_vgpu *vgpu = s->vgpu;
 	struct intel_gvt *gvt = vgpu->gvt;
 	u32 ctx_sr_ctl;
+	u32 *vreg, vreg_old;
 
 	if (offset + 4 > gvt->device_info.mmio_size) {
 		gvt_vgpu_err("%s access to (%x) outside of MMIO range\n",
@@ -937,34 +897,101 @@ static int cmd_reg_handler(struct parser_exec_state *s,
 		return -EFAULT;
 	}
 
+	if (is_init_ctx(s)) {
+		struct intel_gvt_mmio_info *mmio_info;
+
+		intel_gvt_mmio_set_cmd_accessible(gvt, offset);
+		mmio_info = intel_gvt_find_mmio_info(gvt, offset);
+		if (mmio_info && mmio_info->write)
+			intel_gvt_mmio_set_cmd_write_patch(gvt, offset);
+		return 0;
+	}
+
 	if (!intel_gvt_mmio_is_cmd_accessible(gvt, offset)) {
 		gvt_vgpu_err("%s access to non-render register (%x)\n",
 				cmd, offset);
 		return -EBADRQC;
 	}
 
-	if (is_shadowed_mmio(offset)) {
-		gvt_vgpu_err("found access of shadowed MMIO %x\n", offset);
+	if (!strncmp(cmd, "srm", 3) ||
+			!strncmp(cmd, "lrm", 3)) {
+		if (offset != i915_mmio_reg_offset(GEN8_L3SQCREG4) &&
+				offset != 0x21f0) {
+			gvt_vgpu_err("%s access to register (%x)\n",
+					cmd, offset);
+			return -EPERM;
+		} else
+			return 0;
+	}
+
+	if (!strncmp(cmd, "lrr-src", 7) ||
+			!strncmp(cmd, "lrr-dst", 7)) {
+		gvt_vgpu_err("not allowed cmd %s\n", cmd);
+		return -EPERM;
+	}
+
+	if (!strncmp(cmd, "pipe_ctrl", 9)) {
+		/* TODO: add LRI POST logic here */
 		return 0;
 	}
 
-	if (is_mocs_mmio(offset) &&
-	    mocs_cmd_reg_handler(s, offset, index))
-		return -EINVAL;
-
-	if (is_force_nonpriv_mmio(offset) &&
-		force_nonpriv_reg_handler(s, offset, index, cmd))
+	if (strncmp(cmd, "lri", 3))
 		return -EPERM;
 
+	/* below are all lri handlers */
+	vreg = &vgpu_vreg(s->vgpu, offset);
+	if (!intel_gvt_mmio_is_cmd_accessible(gvt, offset)) {
+		gvt_vgpu_err("%s access to non-render register (%x)\n",
+				cmd, offset);
+		return -EBADRQC;
+	}
+
+	if (is_cmd_update_pdps(offset, s) &&
+	    cmd_pdp_mmio_update_handler(s, offset, index))
+		return -EINVAL;
+
 	if (offset == i915_mmio_reg_offset(DERRMR) ||
 		offset == i915_mmio_reg_offset(FORCEWAKE_MT)) {
 		/* Writing to HW VGT_PVINFO_PAGE offset will be discarded */
 		patch_value(s, cmd_ptr(s, index), VGT_PVINFO_PAGE);
 	}
 
-	if (is_cmd_update_pdps(offset, s) &&
-	    cmd_pdp_mmio_update_handler(s, offset, index))
-		return -EINVAL;
+	if (is_mocs_mmio(offset))
+		*vreg = cmd_val(s, index + 1);
+
+	vreg_old = *vreg;
+
+	if (intel_gvt_mmio_is_cmd_write_patch(gvt, offset)) {
+		u32 cmdval_new, cmdval;
+		struct intel_gvt_mmio_info *mmio_info;
+
+		cmdval = cmd_val(s, index + 1);
+
+		mmio_info = intel_gvt_find_mmio_info(gvt, offset);
+		if (!mmio_info) {
+			cmdval_new = cmdval;
+		} else {
+			u64 ro_mask = mmio_info->ro_mask;
+			int ret;
+
+			if (likely(!ro_mask))
+				ret = mmio_info->write(s->vgpu, offset,
+						&cmdval, 4);
+			else {
+				gvt_vgpu_err("try to write RO reg %x\n",
+						offset);
+				ret = -EBADRQC;
+			}
+			if (ret)
+				return ret;
+			cmdval_new = *vreg;
+		}
+		if (cmdval_new != cmdval)
+			patch_value(s, cmd_ptr(s, index+1), cmdval_new);
+	}
+
+	/* only patch cmd. restore vreg value if changed in mmio write handler*/
+	*vreg = vreg_old;
 
 	/* TODO
 	 * In order to let workload with inhibit context to generate
@@ -1216,6 +1243,8 @@ static int cmd_handler_mi_batch_buffer_end(struct parser_exec_state *s)
 		s->buf_type = BATCH_BUFFER_INSTRUCTION;
 		ret = ip_gma_set(s, s->ret_ip_gma_bb);
 		s->buf_addr_type = s->saved_buf_addr_type;
+	} else if (s->buf_type == RING_BUFFER_CTX) {
+		ret = ip_gma_set(s, s->ring_tail);
 	} else {
 		s->buf_type = RING_BUFFER_INSTRUCTION;
 		s->buf_addr_type = GTT_BUFFER;
@@ -2764,7 +2793,8 @@ static int command_scan(struct parser_exec_state *s,
 	gma_bottom = rb_start +  rb_len;
 
 	while (s->ip_gma != gma_tail) {
-		if (s->buf_type == RING_BUFFER_INSTRUCTION) {
+		if (s->buf_type == RING_BUFFER_INSTRUCTION ||
+				s->buf_type == RING_BUFFER_CTX) {
 			if (!(s->ip_gma >= rb_start) ||
 				!(s->ip_gma < gma_bottom)) {
 				gvt_vgpu_err("ip_gma %lx out of ring scope."
@@ -3057,6 +3087,171 @@ int intel_gvt_scan_and_shadow_wa_ctx(struct intel_shadow_wa_ctx *wa_ctx)
 	return 0;
 }
 
+/* generate dummy contexts by sending empty requests to HW, and let
+ * the HW to fill Engine Contexts. This dummy contexts are used for
+ * initialization purpose (update reg whitelist), so referred to as
+ * init context here
+ */
+void intel_gvt_update_reg_whitelist(struct intel_vgpu *vgpu)
+{
+	struct intel_gvt *gvt = vgpu->gvt;
+	struct drm_i915_private *dev_priv = gvt->gt->i915;
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+	const unsigned long start = LRC_STATE_PN * PAGE_SIZE;
+	struct i915_request *rq;
+	struct intel_vgpu_submission *s = &vgpu->submission;
+	struct i915_request *requests[I915_NUM_ENGINES] = {};
+	bool is_ctx_pinned[I915_NUM_ENGINES] = {};
+	int ret;
+
+	if (gvt->is_reg_whitelist_updated)
+		return;
+
+	for_each_engine(engine, &dev_priv->gt, id) {
+		ret = intel_context_pin(s->shadow[id]);
+		if (ret) {
+			gvt_vgpu_err("fail to pin shadow ctx\n");
+			goto out;
+		}
+		is_ctx_pinned[id] = true;
+
+		rq = i915_request_create(s->shadow[id]);
+		if (IS_ERR(rq)) {
+			gvt_vgpu_err("fail to alloc default request\n");
+			ret = -EIO;
+			goto out;
+		}
+		requests[id] = i915_request_get(rq);
+		i915_request_add(rq);
+	}
+
+	if (intel_gt_wait_for_idle(&dev_priv->gt,
+				I915_GEM_IDLE_TIMEOUT) == -ETIME) {
+		ret = -EIO;
+		goto out;
+	}
+
+	/* scan init ctx to update cmd accessible list */
+	for_each_engine(engine, &dev_priv->gt, id) {
+		int size = engine->context_size - PAGE_SIZE;
+		void *vaddr;
+		struct parser_exec_state s;
+		struct drm_i915_gem_object *obj;
+		struct i915_request *rq;
+
+		rq = requests[id];
+		GEM_BUG_ON(!i915_request_completed(rq));
+		GEM_BUG_ON(!intel_context_is_pinned(rq->context));
+		obj = rq->context->state->obj;
+
+		if (!obj) {
+			ret = -EIO;
+			goto out;
+		}
+
+		i915_gem_object_set_cache_coherency(obj,
+						    I915_CACHE_LLC);
+
+		vaddr = i915_gem_object_pin_map(obj, I915_MAP_WB);
+		if (IS_ERR(vaddr)) {
+			gvt_err("failed to pin init ctx obj, ring=%d, err=%lx\n",
+				id, PTR_ERR(vaddr));
+			goto out;
+		}
+
+		s.buf_type = RING_BUFFER_CTX;
+		s.buf_addr_type = GTT_BUFFER;
+		s.vgpu = vgpu;
+		s.engine = engine;
+		s.ring_start = 0;
+		s.ring_size = size;
+		s.ring_head = 0;
+		s.ring_tail = size;
+		s.rb_va = vaddr + start;
+		s.workload = NULL;
+		s.is_ctx_wa = false;
+		s.is_init_ctx = true;
+
+		/* skipping the first RING_CTX_SIZE(0x50) dwords */
+		ret = ip_gma_set(&s, RING_CTX_SIZE);
+		if (ret) {
+			i915_gem_object_unpin_map(obj);
+			goto out;
+		}
+
+		ret = command_scan(&s, 0, size, 0, size);
+		if (ret)
+			gvt_err("Scan init ctx error\n");
+
+		i915_gem_object_unpin_map(obj);
+	}
+
+out:
+	if (!ret)
+		gvt->is_reg_whitelist_updated = true;
+
+	for (id = 0; id < I915_NUM_ENGINES ; id++) {
+		if (requests[id])
+			i915_request_put(requests[id]);
+
+		if (is_ctx_pinned[id])
+			intel_context_unpin(s->shadow[id]);
+	}
+}
+
+int intel_gvt_scan_engine_context(struct intel_vgpu_workload *workload)
+{
+	struct intel_vgpu *vgpu = workload->vgpu;
+	unsigned long gma_head, gma_tail, gma_start, ctx_size;
+	struct parser_exec_state s;
+	int ring_id = workload->engine->id;
+	struct intel_context *ce = vgpu->submission.shadow[ring_id];
+	int ret;
+
+	GEM_BUG_ON(atomic_read(&ce->pin_count) < 0);
+
+	ctx_size = workload->engine->context_size - PAGE_SIZE;
+
+	/* Only ring contxt is loaded to HW for inhibit context, no need to
+	 * scan engine context
+	 */
+	if (is_inhibit_context(ce))
+		return 0;
+
+	gma_start = i915_ggtt_offset(ce->state) + LRC_STATE_PN*PAGE_SIZE;
+	gma_head = 0;
+	gma_tail = ctx_size;
+
+	s.buf_type = RING_BUFFER_CTX;
+	s.buf_addr_type = GTT_BUFFER;
+	s.vgpu = workload->vgpu;
+	s.engine = workload->engine;
+	s.ring_start = gma_start;
+	s.ring_size = ctx_size;
+	s.ring_head = gma_start + gma_head;
+	s.ring_tail = gma_start + gma_tail;
+	s.rb_va = ce->lrc_reg_state;
+	s.workload = workload;
+	s.is_ctx_wa = false;
+	s.is_init_ctx = false;
+
+	/* don't scan the first RING_CTX_SIZE(0x50) dwords, as it's ring
+	 * context
+	 */
+	ret = ip_gma_set(&s, gma_start + gma_head + RING_CTX_SIZE);
+	if (ret)
+		goto out;
+
+	ret = command_scan(&s, gma_head, gma_tail,
+		gma_start, ctx_size);
+out:
+	if (ret)
+		gvt_vgpu_err("scan shadow ctx error\n");
+
+	return ret;
+}
+
 static int init_cmd_table(struct intel_gvt *gvt)
 {
 	unsigned int gen_type = intel_gvt_get_device_type(gvt);

drivers/gpu/drm/i915/gvt/cmd_parser.h

@@ -40,6 +40,7 @@
 
 struct intel_gvt;
 struct intel_shadow_wa_ctx;
+struct intel_vgpu;
 struct intel_vgpu_workload;
 
 void intel_gvt_clean_cmd_parser(struct intel_gvt *gvt);
@@ -50,4 +51,8 @@ int intel_gvt_scan_and_shadow_ringbuffer(struct intel_vgpu_workload *workload);
 
 int intel_gvt_scan_and_shadow_wa_ctx(struct intel_shadow_wa_ctx *wa_ctx);
 
+void intel_gvt_update_reg_whitelist(struct intel_vgpu *vgpu);
+
+int intel_gvt_scan_engine_context(struct intel_vgpu_workload *workload);
+
 #endif

drivers/gpu/drm/i915/gvt/gvt.h

@@ -248,7 +248,7 @@ struct gvt_mmio_block {
 #define INTEL_GVT_MMIO_HASH_BITS 11
 
 struct intel_gvt_mmio {
-	u8 *mmio_attribute;
+	u16 *mmio_attribute;
 /* Register contains RO bits */
 #define F_RO		(1 << 0)
 /* Register contains graphics address */
@@ -267,6 +267,8 @@ struct intel_gvt_mmio {
  * logical context image
  */
 #define F_SR_IN_CTX	(1 << 7)
+/* Value of command write of this reg needs to be patched */
+#define F_CMD_WRITE_PATCH	(1 << 8)
 
 	struct gvt_mmio_block *mmio_block;
 	unsigned int num_mmio_block;
@@ -333,6 +335,7 @@ struct intel_gvt {
 		u32 *mocs_mmio_offset_list;
 		u32 mocs_mmio_offset_list_cnt;
 	} engine_mmio_list;
+	bool is_reg_whitelist_updated;
 
 	struct dentry *debugfs_root;
 };
@@ -416,6 +419,9 @@ int intel_gvt_load_firmware(struct intel_gvt *gvt);
 #define vgpu_fence_base(vgpu) (vgpu->fence.base)
 #define vgpu_fence_sz(vgpu) (vgpu->fence.size)
 
+/* ring context size i.e. the first 0x50 dwords*/
+#define RING_CTX_SIZE 320
+
 struct intel_vgpu_creation_params {
 	__u64 handle;
 	__u64 low_gm_sz;  /* in MB */
@@ -687,6 +693,35 @@ static inline void intel_gvt_mmio_set_sr_in_ctx(
 }
 
 void intel_gvt_debugfs_add_vgpu(struct intel_vgpu *vgpu);
+/**
+ * intel_gvt_mmio_set_cmd_write_patch -
+ *				mark an MMIO if its cmd write needs to be
+ *				patched
+ * @gvt: a GVT device
+ * @offset: register offset
+ *
+ */
+static inline void intel_gvt_mmio_set_cmd_write_patch(
+			struct intel_gvt *gvt, unsigned int offset)
+{
+	gvt->mmio.mmio_attribute[offset >> 2] |= F_CMD_WRITE_PATCH;
+}
+
+/**
+ * intel_gvt_mmio_is_cmd_write_patch - check if an mmio's cmd access needs to
+ * be patched
+ * @gvt: a GVT device
+ * @offset: register offset
+ *
+ * Returns:
+ * True if GPU commmand write to an MMIO should be patched
+ */
+static inline bool intel_gvt_mmio_is_cmd_write_patch(
+			struct intel_gvt *gvt, unsigned int offset)
+{
+	return gvt->mmio.mmio_attribute[offset >> 2] & F_CMD_WRITE_PATCH;
+}
+
 void intel_gvt_debugfs_remove_vgpu(struct intel_vgpu *vgpu);
 void intel_gvt_debugfs_init(struct intel_gvt *gvt);
 void intel_gvt_debugfs_clean(struct intel_gvt *gvt);

drivers/gpu/drm/i915/gvt/handlers.c

@@ -83,7 +83,7 @@ static void write_vreg(struct intel_vgpu *vgpu, unsigned int offset,
 	memcpy(&vgpu_vreg(vgpu, offset), p_data, bytes);
 }
 
-static struct intel_gvt_mmio_info *find_mmio_info(struct intel_gvt *gvt,
+struct intel_gvt_mmio_info *intel_gvt_find_mmio_info(struct intel_gvt *gvt,
 						  unsigned int offset)
 {
 	struct intel_gvt_mmio_info *e;
@@ -96,7 +96,7 @@ static struct intel_gvt_mmio_info *find_mmio_info(struct intel_gvt *gvt,
 }
 
 static int new_mmio_info(struct intel_gvt *gvt,
-		u32 offset, u8 flags, u32 size,
+		u32 offset, u16 flags, u32 size,
 		u32 addr_mask, u32 ro_mask, u32 device,
 		gvt_mmio_func read, gvt_mmio_func write)
 {
@@ -118,7 +118,7 @@ static int new_mmio_info(struct intel_gvt *gvt,
 			return -ENOMEM;
 
 		info->offset = i;
-		p = find_mmio_info(gvt, info->offset);
+		p = intel_gvt_find_mmio_info(gvt, info->offset);
 		if (p) {
 			WARN(1, "dup mmio definition offset %x\n",
 				info->offset);
@@ -1965,7 +1965,8 @@ static int init_generic_mmio_info(struct intel_gvt *gvt)
 
 	/* RING MODE */
 #define RING_REG(base) _MMIO((base) + 0x29c)
-	MMIO_RING_DFH(RING_REG, D_ALL, F_MODE_MASK | F_CMD_ACCESS, NULL,
+	MMIO_RING_DFH(RING_REG, D_ALL,
+		F_MODE_MASK | F_CMD_ACCESS | F_CMD_WRITE_PATCH, NULL,
 		ring_mode_mmio_write);
 #undef RING_REG
 
@@ -2885,8 +2886,8 @@ static int init_bdw_mmio_info(struct intel_gvt *gvt)
 	MMIO_DFH(_MMIO(0xb10c), D_BDW, F_CMD_ACCESS, NULL, NULL);
 	MMIO_D(_MMIO(0xb110), D_BDW);
 
-	MMIO_F(_MMIO(0x24d0), 48, F_CMD_ACCESS, 0, 0, D_BDW_PLUS,
-		NULL, force_nonpriv_write);
+	MMIO_F(_MMIO(0x24d0), 48, F_CMD_ACCESS | F_CMD_WRITE_PATCH, 0, 0,
+		D_BDW_PLUS, NULL, force_nonpriv_write);
 
 	MMIO_D(_MMIO(0x44484), D_BDW_PLUS);
 	MMIO_D(_MMIO(0x4448c), D_BDW_PLUS);
@@ -3626,7 +3627,7 @@ int intel_vgpu_mmio_reg_rw(struct intel_vgpu *vgpu, unsigned int offset,
 	/*
 	 * Normal tracked MMIOs.
 	 */
-	mmio_info = find_mmio_info(gvt, offset);
+	mmio_info = intel_gvt_find_mmio_info(gvt, offset);
 	if (!mmio_info) {
 		gvt_dbg_mmio("untracked MMIO %08x len %d\n", offset, bytes);
 		goto default_rw;

drivers/gpu/drm/i915/gvt/mmio.h

@@ -80,6 +80,9 @@ int intel_gvt_for_each_tracked_mmio(struct intel_gvt *gvt,
 	int (*handler)(struct intel_gvt *gvt, u32 offset, void *data),
 	void *data);
 
+struct intel_gvt_mmio_info *intel_gvt_find_mmio_info(struct intel_gvt *gvt,
+						  unsigned int offset);
+
 int intel_vgpu_init_mmio(struct intel_vgpu *vgpu);
 void intel_vgpu_reset_mmio(struct intel_vgpu *vgpu, bool dmlr);
 void intel_vgpu_clean_mmio(struct intel_vgpu *vgpu);

drivers/gpu/drm/i915/gvt/reg.h

@@ -133,4 +133,6 @@
 #define RING_GFX_MODE(base)	_MMIO((base) + 0x29c)
 #define VF_GUARDBAND		_MMIO(0x83a4)
 
+
+#define BCS_TILE_REGISTER_VAL_OFFSET (0x43*4)
 #endif

drivers/gpu/drm/i915/gvt/scheduler.c

@@ -137,6 +137,7 @@ static int populate_shadow_context(struct intel_vgpu_workload *workload)
 	int i;
 	bool skip = false;
 	int ring_id = workload->engine->id;
+	int ret;
 
 	GEM_BUG_ON(!intel_context_is_pinned(ctx));
 
@@ -163,16 +164,24 @@ static int populate_shadow_context(struct intel_vgpu_workload *workload)
 		COPY_REG(bb_per_ctx_ptr);
 		COPY_REG(rcs_indirect_ctx);
 		COPY_REG(rcs_indirect_ctx_offset);
-	}
+	} else if (workload->engine->id == BCS0)
+		intel_gvt_hypervisor_read_gpa(vgpu,
+				workload->ring_context_gpa +
+				BCS_TILE_REGISTER_VAL_OFFSET,
+				(void *)shadow_ring_context +
+				BCS_TILE_REGISTER_VAL_OFFSET, 4);
 #undef COPY_REG
 #undef COPY_REG_MASKED
 
+	/* don't copy Ring Context (the first 0x50 dwords),
+	 * only copy the Engine Context part from guest
+	 */
 	intel_gvt_hypervisor_read_gpa(vgpu,
 			workload->ring_context_gpa +
-			sizeof(*shadow_ring_context),
+			RING_CTX_SIZE,
 			(void *)shadow_ring_context +
-			sizeof(*shadow_ring_context),
-			I915_GTT_PAGE_SIZE - sizeof(*shadow_ring_context));
+			RING_CTX_SIZE,
+			I915_GTT_PAGE_SIZE - RING_CTX_SIZE);
 
 	sr_oa_regs(workload, (u32 *)shadow_ring_context, false);
 
@@ -239,6 +248,11 @@ read:
 		gpa_size = I915_GTT_PAGE_SIZE;
 		dst = context_base + (i << I915_GTT_PAGE_SHIFT);
 	}
+	ret = intel_gvt_scan_engine_context(workload);
+	if (ret) {
+		gvt_vgpu_err("invalid cmd found in guest context pages\n");
+		return ret;
+	}
 	s->last_ctx[ring_id].valid = true;
 	return 0;
 }

drivers/gpu/drm/i915/gvt/vgpu.c

@@ -500,9 +500,11 @@ struct intel_vgpu *intel_gvt_create_vgpu(struct intel_gvt *gvt,
 
 	mutex_lock(&gvt->lock);
 	vgpu = __intel_gvt_create_vgpu(gvt, &param);
-	if (!IS_ERR(vgpu))
+	if (!IS_ERR(vgpu)) {
 		/* calculate left instance change for types */
 		intel_gvt_update_vgpu_types(gvt);
+		intel_gvt_update_reg_whitelist(vgpu);
+	}
 	mutex_unlock(&gvt->lock);
 
 	return vgpu;

drivers/gpu/drm/i915/i915_cmd_parser.c

@@ -1143,7 +1143,7 @@ static u32 *copy_batch(struct drm_i915_gem_object *dst_obj,
 	void *dst, *src;
 	int ret;
 
-	dst = i915_gem_object_pin_map(dst_obj, I915_MAP_FORCE_WB);
+	dst = i915_gem_object_pin_map(dst_obj, I915_MAP_WB);
 	if (IS_ERR(dst))
 		return dst;
 

drivers/gpu/drm/i915/i915_debugfs.c

@@ -210,7 +210,7 @@ i915_debugfs_describe_obj(struct seq_file *m, struct drm_i915_gem_object *obj)
 	spin_unlock(&obj->vma.lock);
 
 	seq_printf(m, " (pinned x %d)", pin_count);
-	if (obj->stolen)
+	if (i915_gem_object_is_stolen(obj))
 		seq_printf(m, " (stolen: %08llx)", obj->stolen->start);
 	if (i915_gem_object_is_framebuffer(obj))
 		seq_printf(m, " (fb)");
@@ -220,145 +220,6 @@ i915_debugfs_describe_obj(struct seq_file *m, struct drm_i915_gem_object *obj)
 		seq_printf(m, " (%s)", engine->name);
 }
 
-struct file_stats {
-	struct i915_address_space *vm;
-	unsigned long count;
-	u64 total;
-	u64 active, inactive;
-	u64 closed;
-};
-
-static int per_file_stats(int id, void *ptr, void *data)
-{
-	struct drm_i915_gem_object *obj = ptr;
-	struct file_stats *stats = data;
-	struct i915_vma *vma;
-
-	if (IS_ERR_OR_NULL(obj) || !kref_get_unless_zero(&obj->base.refcount))
-		return 0;
-
-	stats->count++;
-	stats->total += obj->base.size;
-
-	spin_lock(&obj->vma.lock);
-	if (!stats->vm) {
-		for_each_ggtt_vma(vma, obj) {
-			if (!drm_mm_node_allocated(&vma->node))
-				continue;
-
-			if (i915_vma_is_active(vma))
-				stats->active += vma->node.size;
-			else
-				stats->inactive += vma->node.size;
-
-			if (i915_vma_is_closed(vma))
-				stats->closed += vma->node.size;
-		}
-	} else {
-		struct rb_node *p = obj->vma.tree.rb_node;
-
-		while (p) {
-			long cmp;
-
-			vma = rb_entry(p, typeof(*vma), obj_node);
-			cmp = i915_vma_compare(vma, stats->vm, NULL);
-			if (cmp == 0) {
-				if (drm_mm_node_allocated(&vma->node)) {
-					if (i915_vma_is_active(vma))
-						stats->active += vma->node.size;
-					else
-						stats->inactive += vma->node.size;
-
-					if (i915_vma_is_closed(vma))
-						stats->closed += vma->node.size;
-				}
-				break;
-			}
-			if (cmp < 0)
-				p = p->rb_right;
-			else
-				p = p->rb_left;
-		}
-	}
-	spin_unlock(&obj->vma.lock);
-
-	i915_gem_object_put(obj);
-	return 0;
-}
-
-#define print_file_stats(m, name, stats) do { \
-	if (stats.count) \
-		seq_printf(m, "%s: %lu objects, %llu bytes (%llu active, %llu inactive, %llu closed)\n", \
-			   name, \
-			   stats.count, \
-			   stats.total, \
-			   stats.active, \
-			   stats.inactive, \
-			   stats.closed); \
-} while (0)
-
-static void print_context_stats(struct seq_file *m,
-				struct drm_i915_private *i915)
-{
-	struct file_stats kstats = {};
-	struct i915_gem_context *ctx, *cn;
-
-	spin_lock(&i915->gem.contexts.lock);
-	list_for_each_entry_safe(ctx, cn, &i915->gem.contexts.list, link) {
-		struct i915_gem_engines_iter it;
-		struct intel_context *ce;
-
-		if (!kref_get_unless_zero(&ctx->ref))
-			continue;
-
-		spin_unlock(&i915->gem.contexts.lock);
-
-		for_each_gem_engine(ce,
-				    i915_gem_context_lock_engines(ctx), it) {
-			if (intel_context_pin_if_active(ce)) {
-				rcu_read_lock();
-				if (ce->state)
-					per_file_stats(0,
-						       ce->state->obj, &kstats);
-				per_file_stats(0, ce->ring->vma->obj, &kstats);
-				rcu_read_unlock();
-				intel_context_unpin(ce);
-			}
-		}
-		i915_gem_context_unlock_engines(ctx);
-
-		mutex_lock(&ctx->mutex);
-		if (!IS_ERR_OR_NULL(ctx->file_priv)) {
-			struct file_stats stats = {
-				.vm = rcu_access_pointer(ctx->vm),
-			};
-			struct drm_file *file = ctx->file_priv->file;
-			struct task_struct *task;
-			char name[80];
-
-			rcu_read_lock();
-			idr_for_each(&file->object_idr, per_file_stats, &stats);
-			rcu_read_unlock();
-
-			rcu_read_lock();
-			task = pid_task(ctx->pid ?: file->pid, PIDTYPE_PID);
-			snprintf(name, sizeof(name), "%s",
-				 task ? task->comm : "<unknown>");
-			rcu_read_unlock();
-
-			print_file_stats(m, name, stats);
-		}
-		mutex_unlock(&ctx->mutex);
-
-		spin_lock(&i915->gem.contexts.lock);
-		list_safe_reset_next(ctx, cn, link);
-		i915_gem_context_put(ctx);
-	}
-	spin_unlock(&i915->gem.contexts.lock);
-
-	print_file_stats(m, "[k]contexts", kstats);
-}
-
 static int i915_gem_object_info(struct seq_file *m, void *data)
 {
 	struct drm_i915_private *i915 = node_to_i915(m->private);
@@ -372,9 +233,6 @@ static int i915_gem_object_info(struct seq_file *m, void *data)
 	for_each_memory_region(mr, i915, id)
 		seq_printf(m, "%s: total:%pa, available:%pa bytes\n",
 			   mr->name, &mr->total, &mr->avail);
-	seq_putc(m, '\n');
-
-	print_context_stats(m, i915);
 
 	return 0;
 }

drivers/gpu/drm/i915/i915_gem.c

@@ -179,108 +179,6 @@ try_again:
 	return ret;
 }
 
-static int
-i915_gem_create(struct drm_file *file,
-		struct intel_memory_region *mr,
-		u64 *size_p,
-		u32 *handle_p)
-{
-	struct drm_i915_gem_object *obj;
-	u32 handle;
-	u64 size;
-	int ret;
-
-	GEM_BUG_ON(!is_power_of_2(mr->min_page_size));
-	size = round_up(*size_p, mr->min_page_size);
-	if (size == 0)
-		return -EINVAL;
-
-	/* For most of the ABI (e.g. mmap) we think in system pages */
-	GEM_BUG_ON(!IS_ALIGNED(size, PAGE_SIZE));
-
-	/* Allocate the new object */
-	obj = i915_gem_object_create_region(mr, size, 0);
-	if (IS_ERR(obj))
-		return PTR_ERR(obj);
-
-	ret = drm_gem_handle_create(file, &obj->base, &handle);
-	/* drop reference from allocate - handle holds it now */
-	i915_gem_object_put(obj);
-	if (ret)
-		return ret;
-
-	*handle_p = handle;
-	*size_p = size;
-	return 0;
-}
-
-int
-i915_gem_dumb_create(struct drm_file *file,
-		     struct drm_device *dev,
-		     struct drm_mode_create_dumb *args)
-{
-	enum intel_memory_type mem_type;
-	int cpp = DIV_ROUND_UP(args->bpp, 8);
-	u32 format;
-
-	switch (cpp) {
-	case 1:
-		format = DRM_FORMAT_C8;
-		break;
-	case 2:
-		format = DRM_FORMAT_RGB565;
-		break;
-	case 4:
-		format = DRM_FORMAT_XRGB8888;
-		break;
-	default:
-		return -EINVAL;
-	}
-
-	/* have to work out size/pitch and return them */
-	args->pitch = ALIGN(args->width * cpp, 64);
-
-	/* align stride to page size so that we can remap */
-	if (args->pitch > intel_plane_fb_max_stride(to_i915(dev), format,
-						    DRM_FORMAT_MOD_LINEAR))
-		args->pitch = ALIGN(args->pitch, 4096);
-
-	if (args->pitch < args->width)
-		return -EINVAL;
-
-	args->size = mul_u32_u32(args->pitch, args->height);
-
-	mem_type = INTEL_MEMORY_SYSTEM;
-	if (HAS_LMEM(to_i915(dev)))
-		mem_type = INTEL_MEMORY_LOCAL;
-
-	return i915_gem_create(file,
-			       intel_memory_region_by_type(to_i915(dev),
-							   mem_type),
-			       &args->size, &args->handle);
-}
-
-/**
- * Creates a new mm object and returns a handle to it.
- * @dev: drm device pointer
- * @data: ioctl data blob
- * @file: drm file pointer
- */
-int
-i915_gem_create_ioctl(struct drm_device *dev, void *data,
-		      struct drm_file *file)
-{
-	struct drm_i915_private *i915 = to_i915(dev);
-	struct drm_i915_gem_create *args = data;
-
-	i915_gem_flush_free_objects(i915);
-
-	return i915_gem_create(file,
-			       intel_memory_region_by_type(i915,
-							   INTEL_MEMORY_SYSTEM),
-			       &args->size, &args->handle);
-}
-
 static int
 shmem_pread(struct page *page, int offset, int len, char __user *user_data,
 	    bool needs_clflush)
@@ -1059,14 +957,14 @@ i915_gem_madvise_ioctl(struct drm_device *dev, void *data,
 	    i915_gem_object_is_tiled(obj) &&
 	    i915->quirks & QUIRK_PIN_SWIZZLED_PAGES) {
 		if (obj->mm.madv == I915_MADV_WILLNEED) {
-			GEM_BUG_ON(!obj->mm.quirked);
-			__i915_gem_object_unpin_pages(obj);
-			obj->mm.quirked = false;
+			GEM_BUG_ON(!i915_gem_object_has_tiling_quirk(obj));
+			i915_gem_object_clear_tiling_quirk(obj);
+			i915_gem_object_make_shrinkable(obj);
 		}
 		if (args->madv == I915_MADV_WILLNEED) {
-			GEM_BUG_ON(obj->mm.quirked);
-			__i915_gem_object_pin_pages(obj);
-			obj->mm.quirked = true;
+			GEM_BUG_ON(i915_gem_object_has_tiling_quirk(obj));
+			i915_gem_object_make_unshrinkable(obj);
+			i915_gem_object_set_tiling_quirk(obj);
 		}
 	}
 
@@ -1277,19 +1175,13 @@ int i915_gem_freeze_late(struct drm_i915_private *i915)
 	 * the objects as well, see i915_gem_freeze()
 	 */
 
-	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
-
-	i915_gem_shrink(i915, -1UL, NULL, ~0);
+	with_intel_runtime_pm(&i915->runtime_pm, wakeref)
+		i915_gem_shrink(i915, -1UL, NULL, ~0);
 	i915_gem_drain_freed_objects(i915);
 
-	list_for_each_entry(obj, &i915->mm.shrink_list, mm.link) {
-		i915_gem_object_lock(obj, NULL);
-		drm_WARN_ON(&i915->drm,
-			    i915_gem_object_set_to_cpu_domain(obj, true));
-		i915_gem_object_unlock(obj);
-	}
-
-	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
+	wbinvd_on_all_cpus();
+	list_for_each_entry(obj, &i915->mm.shrink_list, mm.link)
+		__start_cpu_write(obj);
 
 	return 0;
 }

drivers/gpu/drm/i915/i915_gem.h

@@ -38,11 +38,18 @@ struct drm_i915_private;
 
 #define GEM_SHOW_DEBUG() drm_debug_enabled(DRM_UT_DRIVER)
 
+#ifdef CONFIG_DRM_I915_DEBUG_GEM_ONCE
+#define __GEM_BUG(cond) BUG()
+#else
+#define __GEM_BUG(cond) \
+	WARN(1, "%s:%d GEM_BUG_ON(%s)\n", __func__, __LINE__, __stringify(cond))
+#endif
+
 #define GEM_BUG_ON(condition) do { if (unlikely((condition))) {	\
 		GEM_TRACE_ERR("%s:%d GEM_BUG_ON(%s)\n", \
 			      __func__, __LINE__, __stringify(condition)); \
 		GEM_TRACE_DUMP(); \
-		BUG(); \
+		__GEM_BUG(condition); \
 		} \
 	} while(0)
 #define GEM_WARN_ON(expr) WARN_ON(expr)

drivers/gpu/drm/i915/i915_gem_evict.c

@@ -61,6 +61,17 @@ mark_free(struct drm_mm_scan *scan,
 	return drm_mm_scan_add_block(scan, &vma->node);
 }
 
+static bool defer_evict(struct i915_vma *vma)
+{
+	if (i915_vma_is_active(vma))
+		return true;
+
+	if (i915_vma_is_scanout(vma))
+		return true;
+
+	return false;
+}
+
 /**
  * i915_gem_evict_something - Evict vmas to make room for binding a new one
  * @vm: address space to evict from
@@ -150,7 +161,7 @@ search_again:
 		 * To notice when we complete one full cycle, we record the
 		 * first active element seen, before moving it to the tail.
 		 */
-		if (active != ERR_PTR(-EAGAIN) && i915_vma_is_active(vma)) {
+		if (active != ERR_PTR(-EAGAIN) && defer_evict(vma)) {
 			if (!active)
 				active = vma;
 

drivers/gpu/drm/i915/i915_gpu_error.c

@@ -1051,7 +1051,9 @@ i915_vma_coredump_create(const struct intel_gt *gt,
 		for_each_sgt_daddr(dma, iter, vma->pages) {
 			void __iomem *s;
 
-			s = io_mapping_map_wc(&mem->iomap, dma, PAGE_SIZE);
+			s = io_mapping_map_wc(&mem->iomap,
+					      dma - mem->region.start,
+					      PAGE_SIZE);
 			ret = compress_page(compress,
 					    (void __force *)s, dst,
 					    true);

drivers/gpu/drm/i915/i915_mm.c

@@ -62,7 +62,7 @@ static int remap_sg(pte_t *pte, unsigned long addr, void *data)
 {
 	struct remap_pfn *r = data;
 
-	if (GEM_WARN_ON(!r->sgt.pfn))
+	if (GEM_WARN_ON(!r->sgt.sgp))
 		return -EINVAL;
 
 	/* Special PTE are not associated with any struct page */

drivers/gpu/drm/i915/i915_params.h

@@ -32,6 +32,7 @@ struct drm_printer;
 
 #define ENABLE_GUC_SUBMISSION		BIT(0)
 #define ENABLE_GUC_LOAD_HUC		BIT(1)
+#define ENABLE_GUC_MASK			GENMASK(1, 0)
 
 /*
  * Invoke param, a function-like macro, for each i915 param, with arguments:

drivers/gpu/drm/i915/i915_pci.c

@@ -455,6 +455,7 @@ static const struct intel_device_info snb_m_gt2_info = {
 	.has_llc = 1, \
 	.has_rc6 = 1, \
 	.has_rc6p = 1, \
+	.has_reset_engine = true, \
 	.has_rps = true, \
 	.dma_mask_size = 40, \
 	.ppgtt_type = INTEL_PPGTT_ALIASING, \
@@ -513,6 +514,7 @@ static const struct intel_device_info vlv_info = {
 	.cpu_transcoder_mask = BIT(TRANSCODER_A) | BIT(TRANSCODER_B),
 	.has_runtime_pm = 1,
 	.has_rc6 = 1,
+	.has_reset_engine = true,
 	.has_rps = true,
 	.display.has_gmch = 1,
 	.display.has_hotplug = 1,
@@ -571,8 +573,7 @@ static const struct intel_device_info hsw_gt3_info = {
 	.dma_mask_size = 39, \
 	.ppgtt_type = INTEL_PPGTT_FULL, \
 	.ppgtt_size = 48, \
-	.has_64bit_reloc = 1, \
-	.has_reset_engine = 1
+	.has_64bit_reloc = 1
 
 #define BDW_PLATFORM \
 	GEN8_FEATURES, \

drivers/gpu/drm/i915/i915_request.c

@@ -276,7 +276,7 @@ static void remove_from_engine(struct i915_request *rq)
 
 bool i915_request_retire(struct i915_request *rq)
 {
-	if (!i915_request_completed(rq))
+	if (!__i915_request_is_complete(rq))
 		return false;
 
 	RQ_TRACE(rq, "\n");
@@ -342,8 +342,7 @@ void i915_request_retire_upto(struct i915_request *rq)
 	struct i915_request *tmp;
 
 	RQ_TRACE(rq, "\n");
-
-	GEM_BUG_ON(!i915_request_completed(rq));
+	GEM_BUG_ON(!__i915_request_is_complete(rq));
 
 	do {
 		tmp = list_first_entry(&tl->requests, typeof(*tmp), link);
@@ -552,8 +551,10 @@ bool __i915_request_submit(struct i915_request *request)
 	 * dropped upon retiring. (Otherwise if resubmit a *retired*
 	 * request, this would be a horrible use-after-free.)
 	 */
-	if (i915_request_completed(request))
-		goto xfer;
+	if (__i915_request_is_complete(request)) {
+		list_del_init(&request->sched.link);
+		goto active;
+	}
 
 	if (unlikely(intel_context_is_banned(request->context)))
 		i915_request_set_error_once(request, -EIO);
@@ -588,11 +589,11 @@ bool __i915_request_submit(struct i915_request *request)
 	engine->serial++;
 	result = true;
 
-xfer:
-	if (!test_and_set_bit(I915_FENCE_FLAG_ACTIVE, &request->fence.flags)) {
-		list_move_tail(&request->sched.link, &engine->active.requests);
-		clear_bit(I915_FENCE_FLAG_PQUEUE, &request->fence.flags);
-	}
+	GEM_BUG_ON(test_bit(I915_FENCE_FLAG_ACTIVE, &request->fence.flags));
+	list_move_tail(&request->sched.link, &engine->active.requests);
+active:
+	clear_bit(I915_FENCE_FLAG_PQUEUE, &request->fence.flags);
+	set_bit(I915_FENCE_FLAG_ACTIVE, &request->fence.flags);
 
 	/*
 	 * XXX Rollback bonded-execution on __i915_request_unsubmit()?
@@ -652,7 +653,7 @@ void __i915_request_unsubmit(struct i915_request *request)
 		i915_request_cancel_breadcrumb(request);
 
 	/* We've already spun, don't charge on resubmitting. */
-	if (request->sched.semaphores && i915_request_started(request))
+	if (request->sched.semaphores && __i915_request_has_started(request))
 		request->sched.semaphores = 0;
 
 	/*
@@ -864,7 +865,7 @@ __i915_request_create(struct intel_context *ce, gfp_t gfp)
 	RCU_INIT_POINTER(rq->timeline, tl);
 	RCU_INIT_POINTER(rq->hwsp_cacheline, tl->hwsp_cacheline);
 	rq->hwsp_seqno = tl->hwsp_seqno;
-	GEM_BUG_ON(i915_request_completed(rq));
+	GEM_BUG_ON(__i915_request_is_complete(rq));
 
 	rq->rcustate = get_state_synchronize_rcu(); /* acts as smp_mb() */
 
@@ -970,15 +971,22 @@ i915_request_await_start(struct i915_request *rq, struct i915_request *signal)
 	if (i915_request_started(signal))
 		return 0;
 
+	/*
+	 * The caller holds a reference on @signal, but we do not serialise
+	 * against it being retired and removed from the lists.
+	 *
+	 * We do not hold a reference to the request before @signal, and
+	 * so must be very careful to ensure that it is not _recycled_ as
+	 * we follow the link backwards.
+	 */
 	fence = NULL;
 	rcu_read_lock();
-	spin_lock_irq(&signal->lock);
 	do {
 		struct list_head *pos = READ_ONCE(signal->link.prev);
 		struct i915_request *prev;
 
 		/* Confirm signal has not been retired, the link is valid */
-		if (unlikely(i915_request_started(signal)))
+		if (unlikely(__i915_request_has_started(signal)))
 			break;
 
 		/* Is signal the earliest request on its timeline? */
@@ -1003,7 +1011,6 @@ i915_request_await_start(struct i915_request *rq, struct i915_request *signal)
 
 		fence = &prev->fence;
 	} while (0);
-	spin_unlock_irq(&signal->lock);
 	rcu_read_unlock();
 	if (!fence)
 		return 0;
@@ -1520,7 +1527,7 @@ __i915_request_add_to_timeline(struct i915_request *rq)
 	 */
 	prev = to_request(__i915_active_fence_set(&timeline->last_request,
 						  &rq->fence));
-	if (prev && !i915_request_completed(prev)) {
+	if (prev && !__i915_request_is_complete(prev)) {
 		/*
 		 * The requests are supposed to be kept in order. However,
 		 * we need to be wary in case the timeline->last_request
@@ -1897,10 +1904,10 @@ static char queue_status(const struct i915_request *rq)
 
 static const char *run_status(const struct i915_request *rq)
 {
-	if (i915_request_completed(rq))
+	if (__i915_request_is_complete(rq))
 		return "!";
 
-	if (i915_request_started(rq))
+	if (__i915_request_has_started(rq))
 		return "*";
 
 	if (!i915_sw_fence_signaled(&rq->semaphore))

drivers/gpu/drm/i915/i915_scheduler.c

@@ -520,7 +520,7 @@ void i915_request_show_with_schedule(struct drm_printer *m,
 		if (signaler->timeline == rq->timeline)
 			continue;
 
-		if (i915_request_completed(signaler))
+		if (__i915_request_is_complete(signaler))
 			continue;
 
 		i915_request_show(m, signaler, prefix, indent + 2);

drivers/gpu/drm/i915/i915_vma.h

@@ -363,6 +363,21 @@ i915_vma_unpin_fence(struct i915_vma *vma)
 
 void i915_vma_parked(struct intel_gt *gt);
 
+static inline bool i915_vma_is_scanout(const struct i915_vma *vma)
+{
+	return test_bit(I915_VMA_SCANOUT_BIT, __i915_vma_flags(vma));
+}
+
+static inline void i915_vma_mark_scanout(struct i915_vma *vma)
+{
+	set_bit(I915_VMA_SCANOUT_BIT, __i915_vma_flags(vma));
+}
+
+static inline void i915_vma_clear_scanout(struct i915_vma *vma)
+{
+	clear_bit(I915_VMA_SCANOUT_BIT, __i915_vma_flags(vma));
+}
+
 #define for_each_until(cond) if (cond) break; else
 
 /**

drivers/gpu/drm/i915/i915_vma_types.h

@@ -249,6 +249,9 @@ struct i915_vma {
 #define I915_VMA_USERFAULT	((int)BIT(I915_VMA_USERFAULT_BIT))
 #define I915_VMA_GGTT_WRITE	((int)BIT(I915_VMA_GGTT_WRITE_BIT))
 
+#define I915_VMA_SCANOUT_BIT	18
+#define I915_VMA_SCANOUT	((int)BIT(I915_VMA_SCANOUT_BIT))
+
 	struct i915_active active;
 
 #define I915_VMA_PAGES_BIAS 24

drivers/gpu/drm/i915/intel_memory_region.h

@@ -57,10 +57,10 @@ struct intel_memory_region_ops {
 	int (*init)(struct intel_memory_region *mem);
 	void (*release)(struct intel_memory_region *mem);
 
-	struct drm_i915_gem_object *
-	(*create_object)(struct intel_memory_region *mem,
-			 resource_size_t size,
-			 unsigned int flags);
+	int (*init_object)(struct intel_memory_region *mem,
+			   struct drm_i915_gem_object *obj,
+			   resource_size_t size,
+			   unsigned int flags);
 };
 
 struct intel_memory_region {

drivers/gpu/drm/i915/selftests/i915_gem_evict.c

@@ -38,8 +38,8 @@ static void quirk_add(struct drm_i915_gem_object *obj,
 		      struct list_head *objects)
 {
 	/* quirk is only for live tiled objects, use it to declare ownership */
-	GEM_BUG_ON(obj->mm.quirked);
-	obj->mm.quirked = true;
+	GEM_BUG_ON(i915_gem_object_has_tiling_quirk(obj));
+	i915_gem_object_set_tiling_quirk(obj);
 	list_add(&obj->st_link, objects);
 }
 
@@ -85,7 +85,7 @@ static void unpin_ggtt(struct i915_ggtt *ggtt)
 	struct i915_vma *vma;
 
 	list_for_each_entry(vma, &ggtt->vm.bound_list, vm_link)
-		if (vma->obj->mm.quirked)
+		if (i915_gem_object_has_tiling_quirk(vma->obj))
 			i915_vma_unpin(vma);
 }
 
@@ -94,8 +94,8 @@ static void cleanup_objects(struct i915_ggtt *ggtt, struct list_head *list)
 	struct drm_i915_gem_object *obj, *on;
 
 	list_for_each_entry_safe(obj, on, list, st_link) {
-		GEM_BUG_ON(!obj->mm.quirked);
-		obj->mm.quirked = false;
+		GEM_BUG_ON(!i915_gem_object_has_tiling_quirk(obj));
+		i915_gem_object_set_tiling_quirk(obj);
 		i915_gem_object_put(obj);
 	}
 

drivers/gpu/drm/i915/selftests/mock_region.c

@@ -15,21 +15,16 @@ static const struct drm_i915_gem_object_ops mock_region_obj_ops = {
 	.release = i915_gem_object_release_memory_region,
 };
 
-static struct drm_i915_gem_object *
-mock_object_create(struct intel_memory_region *mem,
-		   resource_size_t size,
-		   unsigned int flags)
+static int mock_object_init(struct intel_memory_region *mem,
+			    struct drm_i915_gem_object *obj,
+			    resource_size_t size,
+			    unsigned int flags)
 {
 	static struct lock_class_key lock_class;
 	struct drm_i915_private *i915 = mem->i915;
-	struct drm_i915_gem_object *obj;
 
 	if (size > mem->mm.size)
-		return ERR_PTR(-E2BIG);
-
-	obj = i915_gem_object_alloc();
-	if (!obj)
-		return ERR_PTR(-ENOMEM);
+		return -E2BIG;
 
 	drm_gem_private_object_init(&i915->drm, &obj->base, size);
 	i915_gem_object_init(obj, &mock_region_obj_ops, &lock_class);
@@ -40,13 +35,13 @@ mock_object_create(struct intel_memory_region *mem,
 
 	i915_gem_object_init_memory_region(obj, mem, flags);
 
-	return obj;
+	return 0;
 }
 
 static const struct intel_memory_region_ops mock_region_ops = {
 	.init = intel_memory_region_init_buddy,
 	.release = intel_memory_region_release_buddy,
-	.create_object = mock_object_create,
+	.init_object = mock_object_init,
 };
 
 struct intel_memory_region *