WebSVN – Kolibri OS – Diff – /drivers/video/drm/i915/intel_lrc.c

  */
 #include
 #include
 #include "intel_drv.h"
-#include "i915_drv.h"
+#include "intel_mocs.h"
 #define GEN9_LR_CONTEXT_RENDER_SIZE (22 * PAGE_SIZE)
 #define GEN8_CTX_VALID (1<<0)
 #define GEN8_CTX_FORCE_PD_RESTORE (1<<1)
 #define GEN8_CTX_FORCE_RESTORE (1<<2)
 #define GEN8_CTX_L3LLC_COHERENT (1<<5)
 #define GEN8_CTX_PRIVILEGE (1<<8)
+#define ASSIGN_CTX_PDP(ppgtt, reg_state, n) { \
+	const u64 _addr = i915_page_dir_dma_addr((ppgtt), (n));	\
+	reg_state[CTX_PDP ## n ## _UDW+1] = upper_32_bits(_addr); \
+	reg_state[CTX_PDP ## n ## _LDW+1] = lower_32_bits(_addr); \
+}
+#define ASSIGN_CTX_PML4(ppgtt, reg_state) { \
+	reg_state[CTX_PDP0_UDW + 1] = upper_32_bits(px_dma(&ppgtt->pml4)); \
+	reg_state[CTX_PDP0_LDW + 1] = lower_32_bits(px_dma(&ppgtt->pml4)); \
+}
 enum {
 	ADVANCED_CONTEXT = 0,
-	LEGACY_CONTEXT,
+	LEGACY_32B_CONTEXT,
 	ADVANCED_AD_CONTEXT,
 	LEGACY_64B_CONTEXT
 };
-#define GEN8_CTX_MODE_SHIFT 3
+#define GEN8_CTX_ADDRESSING_MODE_SHIFT 3
+#define GEN8_CTX_ADDRESSING_MODE(dev)  (USES_FULL_48BIT_PPGTT(dev) ?\
+		LEGACY_64B_CONTEXT :\
+		LEGACY_32B_CONTEXT)
 enum {
 	FAULT_AND_HANG = 0,
 	FAULT_AND_HALT, /* Debug only */
 	FAULT_AND_STREAM,
 	FAULT_AND_CONTINUE /* Unsupported */
 };
 #define GEN8_CTX_ID_SHIFT 32
+#define CTX_RCS_INDIRECT_CTX_OFFSET_DEFAULT  0x17
+static int intel_lr_context_pin(struct drm_i915_gem_request *rq);
+static void lrc_setup_hardware_status_page(struct intel_engine_cs *ring,
+		struct drm_i915_gem_object *default_ctx_obj);
-static int intel_lr_context_pin(struct intel_engine_cs *ring,
 		struct intel_context *ctx);
 /**
  * intel_sanitize_enable_execlists() - sanitize i915.enable_execlists
  * @dev: DRM device.
  * @enable_execlists: value of i915.enable_execlists module parameter.
  *
- * Only certain platforms support Execlists (the prerequisites being
- * support for Logical Ring Contexts and Aliasing PPGTT or better),
+ * Only certain platforms support Execlists (the prerequisites being
- * and only when enabled via module parameter.
+ * support for Logical Ring Contexts and Aliasing PPGTT or better).
+ *
  * Return: 1 if Execlists is supported and has to be enabled.
  */
 int intel_sanitize_enable_execlists(struct drm_device *dev, int enable_execlists)
+{
+	WARN_ON(i915.enable_ppgtt == -1);
+	/* On platforms with execlist available, vGPU will only
+	 * support execlist mode, no ring buffer mode.
+	 */
 	if (HAS_LOGICAL_RING_CONTEXTS(dev) && intel_vgpu_active(dev))
-	WARN_ON(i915.enable_ppgtt == -1);
+		return 1;
 	if (INTEL_INFO(dev)->gen >= 9)
+ *
  * Return: 20-bits globally unique context ID.
  */
 u32 intel_execlists_ctx_id(struct drm_i915_gem_object *ctx_obj)
+{
-	u32 lrca = i915_gem_obj_ggtt_offset(ctx_obj);
+	u32 lrca = i915_gem_obj_ggtt_offset(ctx_obj) +
+			LRC_PPHWSP_PN * PAGE_SIZE;
 	/* LRCA is required to be 4K aligned so the more significant 20 bits
 	 * are globally unique */
 	return lrca >> 12;
 }
+static bool disable_lite_restore_wa(struct intel_engine_cs *ring)
+{
+	struct drm_device *dev = ring->dev;
+	return ((IS_SKYLAKE(dev) && INTEL_REVID(dev) <= SKL_REVID_B0) ||
+		(IS_BROXTON(dev) && INTEL_REVID(dev) == BXT_REVID_A0)) &&
+	       (ring->id == VCS || ring->id == VCS2);
+}
+uint64_t intel_lr_context_descriptor(struct intel_context *ctx,
+				     struct intel_engine_cs *ring)
 static uint64_t execlists_ctx_descriptor(struct drm_i915_gem_object *ctx_obj)
-{
+	struct drm_i915_gem_object *ctx_obj = ctx->engine[ring->id].state;
+	uint64_t desc;
-	uint64_t desc;
+	uint64_t lrca = i915_gem_obj_ggtt_offset(ctx_obj) +
-	uint64_t lrca = i915_gem_obj_ggtt_offset(ctx_obj);
+			LRC_PPHWSP_PN * PAGE_SIZE;
+	WARN_ON(lrca & 0xFFFFFFFF00000FFFULL);
 	WARN_ON(lrca & 0xFFFFFFFF00000FFFULL);
 	desc = GEN8_CTX_VALID;
-	desc = GEN8_CTX_VALID;
+	desc |= GEN8_CTX_ADDRESSING_MODE(dev) << GEN8_CTX_ADDRESSING_MODE_SHIFT;
-	desc |= LEGACY_CONTEXT << GEN8_CTX_MODE_SHIFT;
+	if (IS_GEN8(ctx_obj->base.dev))
-	desc |= GEN8_CTX_L3LLC_COHERENT;
+		desc |= GEN8_CTX_L3LLC_COHERENT;
 	desc |= GEN8_CTX_PRIVILEGE;
 	desc |= lrca;
+	desc |= (u64)intel_execlists_ctx_id(ctx_obj) << GEN8_CTX_ID_SHIFT;
+	/* TODO: WaDisableLiteRestore when we start using semaphore
+	 * signalling between Command Streamers */
+	/* desc |= GEN8_CTX_FORCE_RESTORE; */
 	desc |= (u64)intel_execlists_ctx_id(ctx_obj) << GEN8_CTX_ID_SHIFT;
 	/* WaEnableForceRestoreInCtxtDescForVCS:skl */
-	/* TODO: WaDisableLiteRestore when we start using semaphore
+	/* WaEnableForceRestoreInCtxtDescForVCS:bxt */
-	 * signalling between Command Streamers */
-	/* desc |= GEN8_CTX_FORCE_RESTORE; */
+	if (disable_lite_restore_wa(ring))
 		desc |= GEN8_CTX_FORCE_RESTORE;
+	return desc;
 	return desc;
-static void execlists_elsp_write(struct intel_engine_cs *ring,
+static void execlists_elsp_write(struct drm_i915_gem_request *rq0,
-				 struct drm_i915_gem_object *ctx_obj0,
-				 struct drm_i915_gem_object *ctx_obj1)
-{
+				 struct drm_i915_gem_request *rq1)
-	struct drm_device *dev = ring->dev;
+{
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	uint64_t temp = 0;
-	uint32_t desc[4];
-	unsigned long flags;
-	/* XXX: You must always write both descriptors in the order below. */
-	if (ctx_obj1)
-		temp = execlists_ctx_descriptor(ctx_obj1);
-	else
-		temp = 0;
-	desc[1] = (u32)(temp >> 32);
-	desc[0] = (u32)temp;
-	temp = execlists_ctx_descriptor(ctx_obj0);
-	desc[3] = (u32)(temp >> 32);
-	desc[2] = (u32)temp;
-	/* Set Force Wakeup bit to prevent GT from entering C6 while ELSP writes
-	 * are in progress.
-	 *
-	 * The other problem is that we can't just call gen6_gt_force_wake_get()
-	 * because that function calls intel_runtime_pm_get(), which might sleep.
-	 * Instead, we do the runtime_pm_get/put when creating/destroying requests.
-	 */
-	spin_lock_irqsave(&dev_priv->uncore.lock, flags);
-	if (IS_CHERRYVIEW(dev) || INTEL_INFO(dev)->gen >= 9) {
-		if (dev_priv->uncore.fw_rendercount++ == 0)
-			dev_priv->uncore.funcs.force_wake_get(dev_priv,
-							      FORCEWAKE_RENDER);
-		if (dev_priv->uncore.fw_mediacount++ == 0)
+	struct intel_engine_cs *ring = rq0->ring;
-			dev_priv->uncore.funcs.force_wake_get(dev_priv,
-							      FORCEWAKE_MEDIA);
-		if (INTEL_INFO(dev)->gen >= 9) {
+	struct drm_device *dev = ring->dev;
-			if (dev_priv->uncore.fw_blittercount++ == 0)
+	struct drm_i915_private *dev_priv = dev->dev_private;
-				dev_priv->uncore.funcs.force_wake_get(dev_priv,
-							FORCEWAKE_BLITTER);
-		}
-	} else {
-		if (dev_priv->uncore.forcewake_count++ == 0)
-			dev_priv->uncore.funcs.force_wake_get(dev_priv,
-							      FORCEWAKE_ALL);
-	}
+	uint64_t desc[2];
-	spin_unlock_irqrestore(&dev_priv->uncore.lock, flags);
+	if (rq1) {
-	I915_WRITE(RING_ELSP(ring), desc[1]);
+		desc[1] = intel_lr_context_descriptor(rq1->ctx, rq1->ring);
-	I915_WRITE(RING_ELSP(ring), desc[0]);
-	I915_WRITE(RING_ELSP(ring), desc[3]);
-	/* The context is automatically loaded after the following */
-	I915_WRITE(RING_ELSP(ring), desc[2]);
-	/* ELSP is a wo register, so use another nearby reg for posting instead */
+		rq1->elsp_submitted++;
-	POSTING_READ(RING_EXECLIST_STATUS(ring));
-	/* Release Force Wakeup (see the big comment above). */
-	spin_lock_irqsave(&dev_priv->uncore.lock, flags);
+	} else {
-	if (IS_CHERRYVIEW(dev) || INTEL_INFO(dev)->gen >= 9) {
-		if (--dev_priv->uncore.fw_rendercount == 0)
-			dev_priv->uncore.funcs.force_wake_put(dev_priv,
-							      FORCEWAKE_RENDER);
-		if (--dev_priv->uncore.fw_mediacount == 0)
+		desc[1] = 0;
-			dev_priv->uncore.funcs.force_wake_put(dev_priv,
-							      FORCEWAKE_MEDIA);
+	}
+	desc[0] = intel_lr_context_descriptor(rq0->ctx, rq0->ring);
+	rq0->elsp_submitted++;
+	/* You must always write both descriptors in the order below. */
+	spin_lock(&dev_priv->uncore.lock);
-		if (INTEL_INFO(dev)->gen >= 9) {
+	intel_uncore_forcewake_get__locked(dev_priv, FORCEWAKE_ALL);
-			if (--dev_priv->uncore.fw_blittercount == 0)
+	I915_WRITE_FW(RING_ELSP(ring), upper_32_bits(desc[1]));
-				dev_priv->uncore.funcs.force_wake_put(dev_priv,
+	I915_WRITE_FW(RING_ELSP(ring), lower_32_bits(desc[1]));
-							FORCEWAKE_BLITTER);
-		}
 	} else {
+	I915_WRITE_FW(RING_ELSP(ring), upper_32_bits(desc[0]));
+	/* The context is automatically loaded after the following */
+	I915_WRITE_FW(RING_ELSP(ring), lower_32_bits(desc[0]));
-		if (--dev_priv->uncore.forcewake_count == 0)
+	/* ELSP is a wo register, use another nearby reg for posting */
-			dev_priv->uncore.funcs.force_wake_put(dev_priv,
+	POSTING_READ_FW(RING_EXECLIST_STATUS_LO(ring));
+	intel_uncore_forcewake_put__locked(dev_priv, FORCEWAKE_ALL);
+	spin_unlock(&dev_priv->uncore.lock);
+}
-							      FORCEWAKE_ALL);
+static int execlists_update_context(struct drm_i915_gem_request *rq)
+{
+	struct intel_engine_cs *ring = rq->ring;
-	spin_unlock_irqrestore(&dev_priv->uncore.lock, flags);
+	struct i915_hw_ppgtt *ppgtt = rq->ctx->ppgtt;
+	struct drm_i915_gem_object *ctx_obj = rq->ctx->engine[ring->id].state;
+	struct drm_i915_gem_object *rb_obj = rq->ringbuf->obj;
+	struct page *page;
+	uint32_t *reg_state;
+	BUG_ON(!ctx_obj);
+	WARN_ON(!i915_gem_obj_is_pinned(ctx_obj));
+	WARN_ON(!i915_gem_obj_is_pinned(rb_obj));
+	page = i915_gem_object_get_page(ctx_obj, LRC_STATE_PN);
+	reg_state = kmap_atomic(page);
 	reg_state[CTX_RING_TAIL+1] = rq->tail;
 	reg_state[CTX_RING_BUFFER_START+1] = i915_gem_obj_ggtt_offset(rb_obj);
 static int execlists_update_context(struct drm_i915_gem_object *ctx_obj,
-				    struct drm_i915_gem_object *ring_obj,
+	if (ppgtt && !USES_FULL_48BIT_PPGTT(ppgtt->base.dev)) {
-				    u32 tail)
-{
+		/* True 32b PPGTT with dynamic page allocation: update PDP
-	struct page *page;
+		 * registers and point the unallocated PDPs to scratch page.
-	uint32_t *reg_state;
-	page = i915_gem_object_get_page(ctx_obj, 1);
-	reg_state = kmap_atomic(page);
-	reg_state[CTX_RING_TAIL+1] = tail;
-	reg_state[CTX_RING_BUFFER_START+1] = i915_gem_obj_ggtt_offset(ring_obj);
-	kunmap_atomic(reg_state);
+		 * PML4 is allocated during ppgtt init, so this is not needed
-	return 0;
+		 * in 48-bit mode.
-}
-static void execlists_submit_contexts(struct intel_engine_cs *ring,
+		 */
-				      struct intel_context *to0, u32 tail0,
-				      struct intel_context *to1, u32 tail1)
-{
-	struct drm_i915_gem_object *ctx_obj0 = to0->engine[ring->id].state;
-	struct intel_ringbuffer *ringbuf0 = to0->engine[ring->id].ringbuf;
-	struct drm_i915_gem_object *ctx_obj1 = NULL;
+		ASSIGN_CTX_PDP(ppgtt, reg_state, 3);
-	struct intel_ringbuffer *ringbuf1 = NULL;
+		ASSIGN_CTX_PDP(ppgtt, reg_state, 2);
 		ASSIGN_CTX_PDP(ppgtt, reg_state, 1);
-	BUG_ON(!ctx_obj0);
+		ASSIGN_CTX_PDP(ppgtt, reg_state, 0);
-	WARN_ON(!i915_gem_obj_is_pinned(ctx_obj0));
+	}
-	WARN_ON(!i915_gem_obj_is_pinned(ringbuf0->obj));
 	kunmap_atomic(reg_state);
+	return 0;
+}
+static void execlists_submit_requests(struct drm_i915_gem_request *rq0,
+				      struct drm_i915_gem_request *rq1)
 	execlists_update_context(ctx_obj0, ringbuf0->obj, tail0);
 	execlists_update_context(rq0);
 	if (to1) {
-		ringbuf1 = to1->engine[ring->id].ringbuf;
+	if (rq1)
 			req1 = cursor;
 			break;
+		}
+	}
+	if (IS_GEN8(ring->dev) || IS_GEN9(ring->dev)) {
+		/*
+		 * WaIdleLiteRestore: make sure we never cause a lite
+		 * restore with HEAD==TAIL
+		 */
+		if (req0->elsp_submitted) {
+			/*
+			 * Apply the wa NOOPS to prevent ring:HEAD == req:TAIL
+			 * as we resubmit the request. See gen8_emit_request()
+			 * for where we prepare the padding after the end of the
+			 * request.
+			 */
+			struct intel_ringbuffer *ringbuf;
+			ringbuf = req0->ctx->engine[ring->id].ringbuf;
+			req0->tail += 8;
+			req0->tail &= ringbuf->size - 1;
+		}
+	}
 	WARN_ON(req1 && req1->elsp_submitted);
-	execlists_submit_contexts(ring, req0->ctx, req0->tail,
-				  req1 ? req1->ctx : NULL,
-				  req1 ? req1->tail : 0);
-	req0->elsp_submitted++;
 	if (req1)
-		req1->elsp_submitted++;
+	execlists_submit_requests(req0, req1);
+}
 static bool execlists_check_remove_request(struct intel_engine_cs *ring,
 					   u32 request_id)
+{
-	struct intel_ctx_submit_request *head_req;
+	struct drm_i915_gem_request *head_req;
 	assert_spin_locked(&ring->execlist_lock);
 	head_req = list_first_entry_or_null(&ring->execlist_queue,
 	return false;
+}
 /**
- * intel_execlists_handle_ctx_events() - handle Context Switch interrupts
+ * intel_lrc_irq_handler() - handle Context Switch interrupts
  * @ring: Engine Command Streamer to handle.
+ *
  * Check the unread Context Status Buffers and manage the submission of new
  * contexts to the ELSP accordingly.
  */
-void intel_execlists_handle_ctx_events(struct intel_engine_cs *ring)
+void intel_lrc_irq_handler(struct intel_engine_cs *ring)
+{
 	struct drm_i915_private *dev_priv = ring->dev->dev_private;
 	u32 status_pointer;
 	u8 read_pointer;
 	u8 write_pointer;
-	u32 status;
+	u32 status = 0;
 	u32 status_id;
 	u32 submit_contexts = 0;
 	status_pointer = I915_READ(RING_CONTEXT_STATUS_PTR(ring));
 	read_pointer = ring->next_context_status_buffer;
-	write_pointer = status_pointer & 0x07;
+	write_pointer = status_pointer & GEN8_CSB_PTR_MASK;
 	if (read_pointer > write_pointer)
-		write_pointer += 6;
+		write_pointer += GEN8_CSB_ENTRIES;
+	spin_lock(&ring->execlist_lock);
-	spin_lock(&ring->execlist_lock);
+	while (read_pointer < write_pointer) {
-	while (read_pointer < write_pointer) {
+		read_pointer++;
-		read_pointer++;
+		status = I915_READ(RING_CONTEXT_STATUS_BUF_LO(ring, read_pointer % GEN8_CSB_ENTRIES));
-		status = I915_READ(RING_CONTEXT_STATUS_BUF(ring) +
+		status_id = I915_READ(RING_CONTEXT_STATUS_BUF_HI(ring, read_pointer % GEN8_CSB_ENTRIES));
 				(read_pointer % 6) * 8);
 			if (execlists_check_remove_request(ring, status_id))
 				submit_contexts++;
+		}
+	}
+	if (disable_lite_restore_wa(ring)) {
+		/* Prevent a ctx to preempt itself */
+		if ((status & GEN8_CTX_STATUS_ACTIVE_IDLE) &&
-	if (submit_contexts != 0)
+		    (submit_contexts != 0))
+			execlists_context_unqueue(ring);
+	} else if (submit_contexts != 0) {
+		execlists_context_unqueue(ring);
 		execlists_context_unqueue(ring);
 	spin_unlock(&ring->execlist_lock);
+	WARN(submit_contexts > 2, "More than two context complete events?\n");
-	WARN(submit_contexts > 2, "More than two context complete events?\n");
+	ring->next_context_status_buffer = write_pointer % GEN8_CSB_ENTRIES;
-	ring->next_context_status_buffer = write_pointer % 6;
+	I915_WRITE(RING_CONTEXT_STATUS_PTR(ring),
+		   _MASKED_FIELD(GEN8_CSB_PTR_MASK << 8,
-	I915_WRITE(RING_CONTEXT_STATUS_PTR(ring),
-		   ((u32)ring->next_context_status_buffer & 0x07) << 8);
 				 ((u32)ring->next_context_status_buffer &
+				  GEN8_CSB_PTR_MASK) << 8));
-static int execlists_context_queue(struct intel_engine_cs *ring,
+}
-				   struct intel_context *to,
 				   u32 tail)
-{
-	struct intel_ctx_submit_request *req = NULL, *cursor;
-	struct drm_i915_private *dev_priv = ring->dev->dev_private;
-	unsigned long flags;
-	int num_elements = 0;
-	req = kzalloc(sizeof(*req), GFP_KERNEL);
+static int execlists_context_queue(struct drm_i915_gem_request *request)
-	if (req == NULL)
+{
-		return -ENOMEM;
-	req->ctx = to;
-	i915_gem_context_reference(req->ctx);
+	struct intel_engine_cs *ring = request->ring;
-	if (to != ring->default_context)
+	struct drm_i915_gem_request *cursor;
-		intel_lr_context_pin(ring, to);
+	int num_elements = 0;
-	req->ring = ring;
+	if (request->ctx != ring->default_context)
-	req->tail = tail;
+		intel_lr_context_pin(request);
-	intel_runtime_pm_get(dev_priv);
+	i915_gem_request_reference(request);
-	spin_lock_irqsave(&ring->execlist_lock, flags);
+	spin_lock_irq(&ring->execlist_lock);
 	list_for_each_entry(cursor, &ring->execlist_queue, execlist_link)
 		if (++num_elements > 2)
 			break;
 	if (num_elements > 2) {
-		struct intel_ctx_submit_request *tail_req;
+		struct drm_i915_gem_request *tail_req;
 		tail_req = list_last_entry(&ring->execlist_queue,
-					   struct intel_ctx_submit_request,
+					   struct drm_i915_gem_request,
 					   execlist_link);
-		if (to == tail_req->ctx) {
+		if (request->ctx == tail_req->ctx) {
 			WARN(tail_req->elsp_submitted != 0,
 				"More than 2 already-submitted reqs queued\n");
 			list_del(&tail_req->execlist_link);
 			list_add_tail(&tail_req->execlist_link,
 				&ring->execlist_retired_req_list);
+		}
+	}
-	list_add_tail(&req->execlist_link, &ring->execlist_queue);
+	list_add_tail(&request->execlist_link, &ring->execlist_queue);
 	if (num_elements == 0)
 		execlists_context_unqueue(ring);
-	spin_unlock_irqrestore(&ring->execlist_lock, flags);
+	spin_unlock_irq(&ring->execlist_lock);
 	return 0;
 }
-static int logical_ring_invalidate_all_caches(struct intel_ringbuffer *ringbuf)
+static int logical_ring_invalidate_all_caches(struct drm_i915_gem_request *req)
 {
-	struct intel_engine_cs *ring = ringbuf->ring;
+	struct intel_engine_cs *ring = req->ring;
 	uint32_t flush_domains;
 	int ret;
 	flush_domains = 0;
 	if (ring->gpu_caches_dirty)
+		flush_domains = I915_GEM_GPU_DOMAINS;
-		flush_domains = I915_GEM_GPU_DOMAINS;
 	ret = ring->emit_flush(req, I915_GEM_GPU_DOMAINS, flush_domains);
-	ret = ring->emit_flush(ringbuf, I915_GEM_GPU_DOMAINS, flush_domains);
+	if (ret)
+		return ret;
 	if (ret)
-		return ret;
+	ring->gpu_caches_dirty = false;
 	return 0;
 		wmb();
 	/* Unconditionally invalidate gpu caches and ensure that we do flush
 	 * any residual writes from the previous batch.
 	 */
+	return logical_ring_invalidate_all_caches(req);
+}
+int intel_logical_ring_alloc_request_extras(struct drm_i915_gem_request *request)
+{
+	int ret;
+	request->ringbuf = request->ctx->engine[request->ring->id].ringbuf;
+	if (request->ctx != request->ring->default_context) {
+		ret = intel_lr_context_pin(request);
+		if (ret)
+			return ret;
+	}
+	return 0;
+}
+static int logical_ring_wait_for_space(struct drm_i915_gem_request *req,
+				       int bytes)
+{
+	struct intel_ringbuffer *ringbuf = req->ringbuf;
+	struct intel_engine_cs *ring = req->ring;
+	struct drm_i915_gem_request *target;
+	unsigned space;
+	int ret;
+	if (intel_ring_space(ringbuf) >= bytes)
+		return 0;
+	/* The whole point of reserving space is to not wait! */
+	WARN_ON(ringbuf->reserved_in_use);
+	list_for_each_entry(target, &ring->request_list, list) {
+		/*
+		 * The request queue is per-engine, so can contain requests
+		 * from multiple ringbuffers. Here, we must ignore any that
+		 * aren't from the ringbuffer we're considering.
+		 */
+		if (target->ringbuf != ringbuf)
+			continue;
+		/* Would completion of this request free enough space? */
+		space = __intel_ring_space(target->postfix, ringbuf->tail,
+					   ringbuf->size);
+		if (space >= bytes)
+			break;
+	}
+	if (WARN_ON(&target->list == &ring->request_list))
+		return -ENOSPC;
+	ret = i915_wait_request(target);
+	if (ret)
+		return ret;
+	ringbuf->space = space;
+	return 0;
+}
+/*
+ * intel_logical_ring_advance_and_submit() - advance the tail and submit the workload
+ * @request: Request to advance the logical ringbuffer of.
+ *
+ * The tail is updated in our logical ringbuffer struct, not in the actual context. What
+ * really happens during submission is that the context and current tail will be placed
+ * on a queue waiting for the ELSP to be ready to accept a new context submission. At that
+ * point, the tail *inside* the context is updated and the ELSP written to.
+ */
+static void
+intel_logical_ring_advance_and_submit(struct drm_i915_gem_request *request)
+{
+	struct intel_engine_cs *ring = request->ring;
+	struct drm_i915_private *dev_priv = request->i915;
+	intel_logical_ring_advance(request->ringbuf);
+	request->tail = request->ringbuf->tail;
+	if (intel_ring_stopped(ring))
+		return;
+	if (dev_priv->guc.execbuf_client)
+		i915_guc_submit(dev_priv->guc.execbuf_client, request);
+	else
+		execlists_context_queue(request);
+}
+static void __wrap_ring_buffer(struct intel_ringbuffer *ringbuf)
+{
+	uint32_t __iomem *virt;
+	int rem = ringbuf->size - ringbuf->tail;
+	virt = ringbuf->virtual_start + ringbuf->tail;
+	rem /= 4;
+	while (rem--)
+		iowrite32(MI_NOOP, virt++);
+	ringbuf->tail = 0;
+	intel_ring_update_space(ringbuf);
+}
+static int logical_ring_prepare(struct drm_i915_gem_request *req, int bytes)
+{
+	struct intel_ringbuffer *ringbuf = req->ringbuf;
+	int remain_usable = ringbuf->effective_size - ringbuf->tail;
+	int remain_actual = ringbuf->size - ringbuf->tail;
+	int ret, total_bytes, wait_bytes = 0;
+	bool need_wrap = false;
+	if (ringbuf->reserved_in_use)
+		total_bytes = bytes;
+	else
+		total_bytes = bytes + ringbuf->reserved_size;
+	if (unlikely(bytes > remain_usable)) {
+		/*
+		 * Not enough space for the basic request. So need to flush
+		 * out the remainder and then wait for base + reserved.
+		 */
+		wait_bytes = remain_actual + total_bytes;
+		need_wrap = true;
+	} else {
+		if (unlikely(total_bytes > remain_usable)) {
+			/*
+			 * The base request will fit but the reserved space
+			 * falls off the end. So only need to to wait for the
+			 * reserved size after flushing out the remainder.
+			 */
+			wait_bytes = remain_actual + ringbuf->reserved_size;
+			need_wrap = true;
+		} else if (total_bytes > ringbuf->space) {
+			/* No wrapping required, just waiting. */
+			wait_bytes = total_bytes;
+		}
+	}
+	if (wait_bytes) {
+		ret = logical_ring_wait_for_space(req, wait_bytes);
+		if (unlikely(ret))
+			return ret;
+		if (need_wrap)
+			__wrap_ring_buffer(ringbuf);
+	}
+	return 0;
+}
+/**
+ * intel_logical_ring_begin() - prepare the logical ringbuffer to accept some commands
+ *
+ * @req: The request to start some new work for
+ * @num_dwords: number of DWORDs that we plan to write to the ringbuffer.
+ *
+ * The ringbuffer might not be ready to accept the commands right away (maybe it needs to
+ * be wrapped, or wait a bit for the tail to be updated). This function takes care of that
+ * and also preallocates a request (every workload submission is still mediated through
+ * requests, same as it did with legacy ringbuffer submission).
+ *
+ * Return: non-zero if the ringbuffer is not ready to be written to.
+ */
+int intel_logical_ring_begin(struct drm_i915_gem_request *req, int num_dwords)
+{
+	struct drm_i915_private *dev_priv;
+	int ret;
+	WARN_ON(req == NULL);
+	dev_priv = req->ring->dev->dev_private;
+	ret = i915_gem_check_wedge(&dev_priv->gpu_error,
+				   dev_priv->mm.interruptible);
+	if (ret)
+		return ret;
+	ret = logical_ring_prepare(req, num_dwords * sizeof(uint32_t));
+	if (ret)
+		return ret;
+	req->ringbuf->space -= num_dwords * sizeof(uint32_t);
+	return 0;
+}
+int intel_logical_ring_reserve_space(struct drm_i915_gem_request *request)
+{
+	/*
+	 * The first call merely notes the reserve request and is common for
+	 * all back ends. The subsequent localised _begin() call actually
+	 * ensures that the reservation is available. Without the begin, if
+	 * the request creator immediately submitted the request without
+	 * adding any commands to it then there might not actually be
+	 * sufficient room for the submission commands.
+	 */
+	intel_ring_reserved_space_reserve(request->ringbuf, MIN_SPACE_FOR_ADD_REQUEST);
-	return logical_ring_invalidate_all_caches(ringbuf);
+	return intel_logical_ring_begin(request, 0);
+}
 /**
  * @ctx: Context to employ for this submission.
  * @args: execbuffer call arguments.
  * @vmas: list of vmas.
  * @batch_obj: the batchbuffer to submit.
  * @exec_start: batchbuffer start virtual address pointer.
- * @flags: translated execbuffer call flags.
+ * @dispatch_flags: translated execbuffer call flags.
+ *
  * This is the evil twin version of i915_gem_ringbuffer_submission. It abstracts
  * away the submission details of the execbuffer ioctl call.
+ *
  * Return: non-zero if the submission fails.
  */
-int intel_execlists_submission(struct drm_device *dev, struct drm_file *file,
+int intel_execlists_submission(struct i915_execbuffer_params *params,
-			       struct intel_engine_cs *ring,
-			       struct intel_context *ctx,
 			       struct drm_i915_gem_execbuffer2 *args,
-			       struct list_head *vmas,
+			       struct list_head *vmas)
-			       struct drm_i915_gem_object *batch_obj,
-			       u64 exec_start, u32 flags)
+{
+	struct drm_device       *dev = params->dev;
+	struct intel_engine_cs  *ring = params->ring;
 	struct drm_i915_private *dev_priv = dev->dev_private;
-	struct intel_ringbuffer *ringbuf = ctx->engine[ring->id].ringbuf;
+	struct intel_ringbuffer *ringbuf = params->ctx->engine[ring->id].ringbuf;
+	u64 exec_start;
 	int instp_mode;
 	u32 instp_mask;
 	int ret;
 	default:
 		DRM_DEBUG("execbuf with unknown constants: %d\n", instp_mode);
 		return -EINVAL;
+	}
-	if (args->num_cliprects != 0) {
-		DRM_DEBUG("clip rectangles are only valid on pre-gen5\n");
-		return -EINVAL;
-	} else {
-		if (args->DR4 == 0xffffffff) {
-			DRM_DEBUG("UXA submitting garbage DR4, fixing up\n");
-			args->DR4 = 0;
-		}
-		if (args->DR1 || args->DR4 || args->cliprects_ptr) {
-			DRM_DEBUG("0 cliprects but dirt in cliprects fields\n");
-			return -EINVAL;
-		}
-	}
 	if (args->flags & I915_EXEC_GEN7_SOL_RESET) {
 		DRM_DEBUG("sol reset is gen7 only\n");
 		return -EINVAL;
 	}
-	ret = execlists_move_to_gpu(ringbuf, vmas);
+	ret = execlists_move_to_gpu(params->request, vmas);
 	if (ret)
 		return ret;
 	if (ring == &dev_priv->ring[RCS] &&
 	    instp_mode != dev_priv->relative_constants_mode) {
-		ret = intel_logical_ring_begin(ringbuf, 4);
+		ret = intel_logical_ring_begin(params->request, 4);
 		if (ret)
 		intel_logical_ring_advance(ringbuf);
 		dev_priv->relative_constants_mode = instp_mode;
+	}
+	exec_start = params->batch_obj_vm_offset +
-	}
+		     args->batch_start_offset;
-	ret = ring->emit_bb_start(ringbuf, exec_start, flags);
+	ret = ring->emit_bb_start(params->request, exec_start, params->dispatch_flags);
+	if (ret)
+		return ret;
-	if (ret)
-		return ret;
+	trace_i915_gem_ring_dispatch(params->request, params->dispatch_flags);
-	i915_gem_execbuffer_move_to_active(vmas, ring);
+	i915_gem_execbuffer_move_to_active(vmas, params->request);
-	i915_gem_execbuffer_retire_commands(dev, file, ring, batch_obj);
+	i915_gem_execbuffer_retire_commands(params);
 	return 0;
-}
 void intel_execlists_retire_requests(struct intel_engine_cs *ring)
-	struct intel_ctx_submit_request *req, *tmp;
+void intel_execlists_retire_requests(struct intel_engine_cs *ring)
 	struct drm_i915_private *dev_priv = ring->dev->dev_private;
-	unsigned long flags;
+	struct drm_i915_gem_request *req, *tmp;
 	struct list_head retired_list;
 	WARN_ON(!mutex_is_locked(&ring->dev->struct_mutex));
 	if (list_empty(&ring->execlist_retired_req_list))
 		return;
 	INIT_LIST_HEAD(&retired_list);
-	spin_lock_irqsave(&ring->execlist_lock, flags);
+	spin_lock_irq(&ring->execlist_lock);
 	list_replace_init(&ring->execlist_retired_req_list, &retired_list);
-	spin_unlock_irqrestore(&ring->execlist_lock, flags);
-	list_for_each_entry_safe(req, tmp, &retired_list, execlist_link) {
+	spin_unlock_irq(&ring->execlist_lock);
-		struct intel_context *ctx = req->ctx;
-		struct drm_i915_gem_object *ctx_obj =
+	list_for_each_entry_safe(req, tmp, &retired_list, execlist_link) {
-				ctx->engine[ring->id].state;
+		struct intel_context *ctx = req->ctx;
 		struct drm_i915_gem_object *ctx_obj =
-		if (ctx_obj && (ctx != ring->default_context))
+				ctx->engine[ring->id].state;
 		return;
+	}
 	I915_WRITE_MODE(ring, _MASKED_BIT_DISABLE(STOP_RING));
+}
-int logical_ring_flush_all_caches(struct intel_ringbuffer *ringbuf)
+int logical_ring_flush_all_caches(struct drm_i915_gem_request *req)
 {
-	struct intel_engine_cs *ring = ringbuf->ring;
+	struct intel_engine_cs *ring = req->ring;
 	int ret;
 	if (!ring->gpu_caches_dirty)
 		return 0;
-	ret = ring->emit_flush(ringbuf, 0, I915_GEM_GPU_DOMAINS);
+	ret = ring->emit_flush(req, 0, I915_GEM_GPU_DOMAINS);
 	if (ret)
 		return ret;
-	ring->gpu_caches_dirty = false;
-	return 0;
-}
-/**
- * intel_logical_ring_advance_and_submit() - advance the tail and submit the workload
- * @ringbuf: Logical Ringbuffer to advance.
- *
- * The tail is updated in our logical ringbuffer struct, not in the actual context. What
- * really happens during submission is that the context and current tail will be placed
- * on a queue waiting for the ELSP to be ready to accept a new context submission. At that
+	ring->gpu_caches_dirty = false;
- * point, the tail *inside* the context is updated and the ELSP written to.
- */
-void intel_logical_ring_advance_and_submit(struct intel_ringbuffer *ringbuf)
-{
-	struct intel_engine_cs *ring = ringbuf->ring;
-	struct intel_context *ctx = ringbuf->FIXME_lrc_ctx;
-	intel_logical_ring_advance(ringbuf);
-	if (intel_ring_stopped(ring))
-		return;
+	return 0;
 	execlists_context_queue(ring, ctx, ringbuf->tail);
-}
+static int intel_lr_context_do_pin(struct intel_engine_cs *ring,
-static int intel_lr_context_pin(struct intel_engine_cs *ring,
+		struct drm_i915_gem_object *ctx_obj,
-		struct intel_context *ctx)
+		struct intel_ringbuffer *ringbuf)
-{
-	struct drm_i915_gem_object *ctx_obj = ctx->engine[ring->id].state;
+{
-	struct intel_ringbuffer *ringbuf = ctx->engine[ring->id].ringbuf;
+	struct drm_device *dev = ring->dev;
-	int ret = 0;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	int ret = 0;
 	WARN_ON(!mutex_is_locked(&ring->dev->struct_mutex));
-	if (ctx->engine[ring->id].unpin_count++ == 0) {
+	WARN_ON(!mutex_is_locked(&ring->dev->struct_mutex));
-		ret = i915_gem_obj_ggtt_pin(ctx_obj,
+	ret = i915_gem_obj_ggtt_pin(ctx_obj, GEN8_LR_CONTEXT_ALIGN,
+			PIN_OFFSET_BIAS | GUC_WOPCM_TOP);
+	if (ret)
-				GEN8_LR_CONTEXT_ALIGN, 0);
+		return ret;
+	ret = intel_pin_and_map_ringbuffer_obj(ring->dev, ringbuf);
+	if (ret)
-		if (ret)
+		goto unpin_ctx_obj;
 			goto reset_unpin_count;
 	ctx_obj->dirty = true;
-		ret = intel_pin_and_map_ringbuffer_obj(ring->dev, ringbuf);
-		if (ret)
 			goto unpin_ctx_obj;
 	/* Invalidate GuC TLB. */
+	if (i915.enable_guc_submission)
-	return ret;
 		I915_WRITE(GEN8_GTCR, GEN8_GTCR_INVALIDATE);
+	return ret;
-unpin_ctx_obj:
-	i915_gem_object_ggtt_unpin(ctx_obj);
+unpin_ctx_obj:
+	i915_gem_object_ggtt_unpin(ctx_obj);
+	return ret;
+}
+static int intel_lr_context_pin(struct drm_i915_gem_request *rq)
+{
+	int ret = 0;
+	struct intel_engine_cs *ring = rq->ring;
+	struct drm_i915_gem_object *ctx_obj = rq->ctx->engine[ring->id].state;
+	struct intel_ringbuffer *ringbuf = rq->ringbuf;
+	if (rq->ctx->engine[ring->id].pin_count++ == 0) {
+		ret = intel_lr_context_do_pin(ring, ctx_obj, ringbuf);
+		if (ret)
+			goto reset_pin_count;
+	}
+	return ret;
 reset_unpin_count:
-	ctx->engine[ring->id].unpin_count = 0;
+reset_pin_count:
+	rq->ctx->engine[ring->id].pin_count = 0;
 	return ret;
+}
-void intel_lr_context_unpin(struct intel_engine_cs *ring,
+void intel_lr_context_unpin(struct drm_i915_gem_request *rq)
 		struct intel_context *ctx)
-{
+	struct intel_engine_cs *ring = rq->ring;
-	struct drm_i915_gem_object *ctx_obj = ctx->engine[ring->id].state;
-	struct intel_ringbuffer *ringbuf = ctx->engine[ring->id].ringbuf;
+	struct drm_i915_gem_object *ctx_obj = rq->ctx->engine[ring->id].state;
+	struct intel_ringbuffer *ringbuf = rq->ringbuf;
+	if (ctx_obj) {
+		WARN_ON(!mutex_is_locked(&ring->dev->struct_mutex));
+		if (--rq->ctx->engine[ring->id].pin_count == 0) {
+			intel_unpin_ringbuffer_obj(ringbuf);
-	if (ctx_obj) {
+			i915_gem_object_ggtt_unpin(ctx_obj);
 		WARN_ON(!mutex_is_locked(&ring->dev->struct_mutex));
-		if (--ctx->engine[ring->id].unpin_count == 0) {
+	}
-			intel_unpin_ringbuffer_obj(ringbuf);
+}
-			i915_gem_object_ggtt_unpin(ctx_obj);
-		}
-	}
-}
+static int intel_logical_ring_workarounds_emit(struct drm_i915_gem_request *req)
-static int logical_ring_alloc_seqno(struct intel_engine_cs *ring,
+{
-				    struct intel_context *ctx)
+	int ret, i;
-{
-	int ret;
+	struct intel_engine_cs *ring = req->ring;
+	struct intel_ringbuffer *ringbuf = req->ringbuf;
+	struct drm_device *dev = ring->dev;
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct i915_workarounds *w = &dev_priv->workarounds;
 	if (WARN_ON_ONCE(w->count == 0))
-	if (ring->outstanding_lazy_seqno)
+		return 0;
-		return 0;
-	if (ring->preallocated_lazy_request == NULL) {
-		struct drm_i915_gem_request *request;
-		request = kmalloc(sizeof(*request), GFP_KERNEL);
-		if (request == NULL)
+	ring->gpu_caches_dirty = true;
+	ret = logical_ring_flush_all_caches(req);
-			return -ENOMEM;
+	if (ret)
+		return ret;
-		if (ctx != ring->default_context) {
+	ret = intel_logical_ring_begin(req, w->count * 2 + 2);
-			ret = intel_lr_context_pin(ring, ctx);
+	if (ret)
-			if (ret) {
+		return ret;
-				kfree(request);
-				return ret;
+	intel_logical_ring_emit(ringbuf, MI_LOAD_REGISTER_IMM(w->count));
-			}
+	for (i = 0; i < w->count; i++) {
-		}
+		intel_logical_ring_emit(ringbuf, w->reg[i].addr);
+		intel_logical_ring_emit(ringbuf, w->reg[i].value);
+	}
-		/* Hold a reference to the context this request belongs to
-		 * (we will need it when the time comes to emit/retire the
-		 * request).
+	intel_logical_ring_emit(ringbuf, MI_NOOP);
+	intel_logical_ring_advance(ringbuf);
+	ring->gpu_caches_dirty = true;
+	ret = logical_ring_flush_all_caches(req);
+	if (ret)
+		return ret;
+	return 0;
+}
+#define wa_ctx_emit(batch, index, cmd)					\
-		 */
+	do {								\
+		int __index = (index)++;				\
+		if (WARN_ON(__index >= (PAGE_SIZE / sizeof(uint32_t)))) { \
+			return -ENOSPC;					\
-		request->ctx = ctx;
+		}							\
-		i915_gem_context_reference(request->ctx);
+		batch[__index] = (cmd);					\
+	} while (0)
+/*
+ * In this WA we need to set GEN8_L3SQCREG4[21:21] and reset it after
+ * PIPE_CONTROL instruction. This is required for the flush to happen correctly
+ * but there is a slight complication as this is applied in WA batch where the
+ * values are only initialized once so we cannot take register value at the
+ * beginning and reuse it further; hence we save its value to memory, upload a
+ * constant value with bit21 set and then we restore it back with the saved value.
+ * To simplify the WA, a constant value is formed by using the default value
-		ring->preallocated_lazy_request = request;
+ * of this register. This shouldn't be a problem because we are only modifying
-	}
+ * it for a short period and this batch in non-premptible. We can ofcourse
+ * use additional instructions that read the actual value of the register
-	return i915_gem_get_seqno(ring->dev, &ring->outstanding_lazy_seqno);
+ * at that time and set our bit of interest but it makes the WA complicated.
-}
+ *
-static int logical_ring_wait_request(struct intel_ringbuffer *ringbuf,
+ * This WA is also required for Gen9 so extracting as a function avoids
-				     int bytes)
+ * code duplication.
-{
+ */
+static inline int gen8_emit_flush_coherentl3_wa(struct intel_engine_cs *ring,
+						uint32_t *const batch,
+						uint32_t index)
-	struct intel_engine_cs *ring = ringbuf->ring;
+{
-	struct drm_i915_gem_request *request;
+	uint32_t l3sqc4_flush = (0x40400000 | GEN8_LQSC_FLUSH_COHERENT_LINES);
-	u32 seqno = 0;
+	/*
+	 * WaDisableLSQCROPERFforOCL:skl
+	 * This WA is implemented in skl_init_clock_gating() but since
-	int ret;
+	 * this batch updates GEN8_L3SQCREG4 with default value we need to
+	 * set this bit here to retain the WA during flush.
-	if (ringbuf->last_retired_head != -1) {
+	 */
-		ringbuf->head = ringbuf->last_retired_head;
-		ringbuf->last_retired_head = -1;
+	if (IS_SKYLAKE(ring->dev) && INTEL_REVID(ring->dev) <= SKL_REVID_E0)
 		l3sqc4_flush |= GEN8_LQSC_RO_PERF_DIS;
-		ringbuf->space = intel_ring_space(ringbuf);
-		if (ringbuf->space >= bytes)
+	wa_ctx_emit(batch, index, (MI_STORE_REGISTER_MEM_GEN8 |
+				   MI_SRM_LRM_GLOBAL_GTT));
-			return 0;
+	wa_ctx_emit(batch, index, GEN8_L3SQCREG4);
-	}
-	list_for_each_entry(request, &ring->request_list, list) {
+	wa_ctx_emit(batch, index, ring->scratch.gtt_offset + 256);
-		if (__intel_ring_space(request->tail, ringbuf->tail,
-				       ringbuf->size) >= bytes) {
+	wa_ctx_emit(batch, index, 0);
-			seqno = request->seqno;
-			break;
+	wa_ctx_emit(batch, index, MI_LOAD_REGISTER_IMM(1));
-		}
+	wa_ctx_emit(batch, index, GEN8_L3SQCREG4);
+	wa_ctx_emit(batch, index, l3sqc4_flush);
+	wa_ctx_emit(batch, index, GFX_OP_PIPE_CONTROL(6));
+	wa_ctx_emit(batch, index, (PIPE_CONTROL_CS_STALL |
+				   PIPE_CONTROL_DC_FLUSH_ENABLE));
+	wa_ctx_emit(batch, index, 0);
+	wa_ctx_emit(batch, index, 0);
+	wa_ctx_emit(batch, index, 0);
+	wa_ctx_emit(batch, index, 0);
+	wa_ctx_emit(batch, index, (MI_LOAD_REGISTER_MEM_GEN8 |
+				   MI_SRM_LRM_GLOBAL_GTT));
+	wa_ctx_emit(batch, index, GEN8_L3SQCREG4);
+	wa_ctx_emit(batch, index, ring->scratch.gtt_offset + 256);
+	wa_ctx_emit(batch, index, 0);
+	return index;
-	}
+}
+static inline uint32_t wa_ctx_start(struct i915_wa_ctx_bb *wa_ctx,
+				    uint32_t offset,
+				    uint32_t start_alignment)
+{
+	return wa_ctx->offset = ALIGN(offset, start_alignment);
+}
+static inline int wa_ctx_end(struct i915_wa_ctx_bb *wa_ctx,
+			     uint32_t offset,
+			     uint32_t size_alignment)
+{
+	wa_ctx->size = offset - wa_ctx->offset;
+	WARN(wa_ctx->size % size_alignment,
+	     "wa_ctx_bb failed sanity checks: size %d is not aligned to %d\n",
+	     wa_ctx->size, size_alignment);
-	if (seqno == 0)
+	return 0;
+}
+/**
+ * gen8_init_indirectctx_bb() - initialize indirect ctx batch with WA
+ *
-		return -ENOSPC;
+ * @ring: only applicable for RCS
+ * @wa_ctx: structure representing wa_ctx
+ *  offset: specifies start of the batch, should be cache-aligned. This is updated
+ *    with the offset value received as input.
+ *  size: size of the batch in DWORDS but HW expects in terms of cachelines
+ * @batch: page in which WA are loaded
+ * @offset: This field specifies the start of the batch, it should be
-	ret = i915_wait_seqno(ring, seqno);
+ *  cache-aligned otherwise it is adjusted accordingly.
+ *  Typically we only have one indirect_ctx and per_ctx batch buffer which are
+ *  initialized at the beginning and shared across all contexts but this field
+ *  helps us to have multiple batches at different offsets and select them based
+ *  on a criteria. At the moment this batch always start at the beginning of the page
+ *  and at this point we don't have multiple wa_ctx batch buffers.
+ *
-	if (ret)
+ *  The number of WA applied are not known at the beginning; we use this field
+ *  to return the no of DWORDS written.
+ *
+ *  It is to be noted that this batch does not contain MI_BATCH_BUFFER_END
+ *  so it adds NOOPs as padding to make it cacheline aligned.
+ *  MI_BATCH_BUFFER_END will be added to perctx batch and both of them together
+ *  makes a complete batch buffer.
+ *
+ * Return: non-zero if we exceed the PAGE_SIZE limit.
+ */
+static int gen8_init_indirectctx_bb(struct intel_engine_cs *ring,
+				    struct i915_wa_ctx_bb *wa_ctx,
-		return ret;
+				    uint32_t *const batch,
+				    uint32_t *offset)
+{
+	uint32_t scratch_addr;
+	uint32_t index = wa_ctx_start(wa_ctx, *offset, CACHELINE_DWORDS);
+	/* WaDisableCtxRestoreArbitration:bdw,chv */
+	wa_ctx_emit(batch, index, MI_ARB_ON_OFF | MI_ARB_DISABLE);
-	i915_gem_retire_requests_ring(ring);
-	ringbuf->head = ringbuf->last_retired_head;
+	/* WaFlushCoherentL3CacheLinesAtContextSwitch:bdw */
-	ringbuf->last_retired_head = -1;
-	ringbuf->space = intel_ring_space(ringbuf);
-	return 0;
+	if (IS_BROADWELL(ring->dev)) {
+		int rc = gen8_emit_flush_coherentl3_wa(ring, batch, index);
+		if (rc < 0)
+			return rc;
+		index = rc;
+	}
+	/* WaClearSlmSpaceAtContextSwitch:bdw,chv */
+	/* Actual scratch location is at 128 bytes offset */
+	scratch_addr = ring->scratch.gtt_offset + 2*CACHELINE_BYTES;
+	wa_ctx_emit(batch, index, GFX_OP_PIPE_CONTROL(6));
+	wa_ctx_emit(batch, index, (PIPE_CONTROL_FLUSH_L3 |
+				   PIPE_CONTROL_GLOBAL_GTT_IVB |
+				   PIPE_CONTROL_CS_STALL |
+				   PIPE_CONTROL_QW_WRITE));
+	wa_ctx_emit(batch, index, scratch_addr);
+	wa_ctx_emit(batch, index, 0);
+	wa_ctx_emit(batch, index, 0);
+	wa_ctx_emit(batch, index, 0);
-}
+	/* Pad to end of cacheline */
+	while (index % CACHELINE_DWORDS)
+		wa_ctx_emit(batch, index, MI_NOOP);
-static int logical_ring_wait_for_space(struct intel_ringbuffer *ringbuf,
+	/*
-				       int bytes)
-{
-	struct intel_engine_cs *ring = ringbuf->ring;
+	 * MI_BATCH_BUFFER_END is not required in Indirect ctx BB because
-	struct drm_device *dev = ring->dev;
-	struct drm_i915_private *dev_priv = dev->dev_private;
-	unsigned long end;
-	int ret;
+	 * execution depends on the length specified in terms of cache lines
-	ret = logical_ring_wait_request(ringbuf, bytes);
+	 * in the register CTX_RCS_INDIRECT_CTX
-	if (ret != -ENOSPC)
+	 */
+	return wa_ctx_end(wa_ctx, *offset = index, CACHELINE_DWORDS);
+}
 		return ret;
+/**
-	/* Force the context submission in case we have been skipping it */
+ * gen8_init_perctx_bb() - initialize per ctx batch with WA
+ *
-	intel_logical_ring_advance_and_submit(ringbuf);
+ * @ring: only applicable for RCS
+ * @wa_ctx: structure representing wa_ctx
+ *  offset: specifies start of the batch, should be cache-aligned.
+ *  size: size of the batch in DWORDS but HW expects in terms of cachelines
+ * @batch: page in which WA are loaded
+ * @offset: This field specifies the start of this batch.
-	/* With GEM the hangcheck timer should kick us out of the loop,
+ *   This batch is started immediately after indirect_ctx batch. Since we ensure
-	 * leaving it early runs the risk of corrupting GEM state (due
+ *   that indirect_ctx ends on a cacheline this batch is aligned automatically.
+ *
+ *   The number of DWORDS written are returned using this field.
+ *
+ *  This batch is terminated with MI_BATCH_BUFFER_END and so we need not add padding
+ *  to align it with cacheline as padding after MI_BATCH_BUFFER_END is redundant.
+ */
+static int gen8_init_perctx_bb(struct intel_engine_cs *ring,
-	 * to running on almost untested codepaths). But on resume
+			       struct i915_wa_ctx_bb *wa_ctx,
-	 * timers don't work yet, so prevent a complete hang in that
+			       uint32_t *const batch,
+			       uint32_t *offset)
-	 * case by choosing an insanely large timeout. */
+{
-	end = jiffies + 60 * HZ;
+	uint32_t index = wa_ctx_start(wa_ctx, *offset, CACHELINE_DWORDS);
+	/* WaDisableCtxRestoreArbitration:bdw,chv */
+	wa_ctx_emit(batch, index, MI_ARB_ON_OFF | MI_ARB_ENABLE);
+	wa_ctx_emit(batch, index, MI_BATCH_BUFFER_END);
+	return wa_ctx_end(wa_ctx, *offset = index, 1);
+}
-	do {
+static int gen9_init_indirectctx_bb(struct intel_engine_cs *ring,
-		ringbuf->head = I915_READ_HEAD(ring);
+				    struct i915_wa_ctx_bb *wa_ctx,
+				    uint32_t *const batch,
+				    uint32_t *offset)
+{
+	int ret;
+	struct drm_device *dev = ring->dev;
+	uint32_t index = wa_ctx_start(wa_ctx, *offset, CACHELINE_DWORDS);
+	/* WaDisableCtxRestoreArbitration:skl,bxt */
-		ringbuf->space = intel_ring_space(ringbuf);
+	if ((IS_SKYLAKE(dev) && (INTEL_REVID(dev) <= SKL_REVID_D0)) ||
-		if (ringbuf->space >= bytes) {
+	    (IS_BROXTON(dev) && (INTEL_REVID(dev) == BXT_REVID_A0)))
-			ret = 0;
+		wa_ctx_emit(batch, index, MI_ARB_ON_OFF | MI_ARB_DISABLE);
 			break;
 	/* WaFlushCoherentL3CacheLinesAtContextSwitch:skl,bxt */
+	ret = gen8_emit_flush_coherentl3_wa(ring, batch, index);
-		msleep(1);
+	if (ret < 0)
+		return ret;
-		ret = i915_gem_check_wedge(&dev_priv->gpu_error,
+	index = ret;
 					   dev_priv->mm.interruptible);
-		if (ret)
+	/* Pad to end of cacheline */
+	while (index % CACHELINE_DWORDS)
-			break;
+		wa_ctx_emit(batch, index, MI_NOOP);
+	return wa_ctx_end(wa_ctx, *offset = index, CACHELINE_DWORDS);
 		if (time_after(jiffies, end)) {
 			ret = -EBUSY;
-			break;
+static int gen9_init_perctx_bb(struct intel_engine_cs *ring,
 			       struct i915_wa_ctx_bb *wa_ctx,
-	} while (1);
-	return ret;
-}
-static int logical_ring_wrap_buffer(struct intel_ringbuffer *ringbuf)
-{
-	uint32_t __iomem *virt;
-	int rem = ringbuf->size - ringbuf->tail;
-	if (ringbuf->space < rem) {
-		int ret = logical_ring_wait_for_space(ringbuf, rem);
-		if (ret)
+			       uint32_t *const batch,
-			return ret;
+			       uint32_t *offset)
-	}
+{
+	struct drm_device *dev = ring->dev;
-	virt = ringbuf->virtual_start + ringbuf->tail;
+	uint32_t index = wa_ctx_start(wa_ctx, *offset, CACHELINE_DWORDS);
-	rem /= 4;
-	while (rem--)
+	/* WaSetDisablePixMaskCammingAndRhwoInCommonSliceChicken:skl,bxt */
-		iowrite32(MI_NOOP, virt++);
-	ringbuf->tail = 0;
+	if ((IS_SKYLAKE(dev) && (INTEL_REVID(dev) <= SKL_REVID_B0)) ||
-	ringbuf->space = intel_ring_space(ringbuf);
+	    (IS_BROXTON(dev) && (INTEL_REVID(dev) == BXT_REVID_A0))) {
+		wa_ctx_emit(batch, index, MI_LOAD_REGISTER_IMM(1));
-	return 0;
+		wa_ctx_emit(batch, index, GEN9_SLICE_COMMON_ECO_CHICKEN0);
+		wa_ctx_emit(batch, index,
-}
+			    _MASKED_BIT_ENABLE(DISABLE_PIXEL_MASK_CAMMING));
+		wa_ctx_emit(batch, index, MI_NOOP);
+	}
-static int logical_ring_prepare(struct intel_ringbuffer *ringbuf, int bytes)
-{
-	int ret;
+	/* WaDisableCtxRestoreArbitration:skl,bxt */
+	if ((IS_SKYLAKE(dev) && (INTEL_REVID(dev) <= SKL_REVID_D0)) ||
+	    (IS_BROXTON(dev) && (INTEL_REVID(dev) == BXT_REVID_A0)))
+		wa_ctx_emit(batch, index, MI_ARB_ON_OFF | MI_ARB_ENABLE);
 	if (unlikely(ringbuf->tail + bytes > ringbuf->effective_size)) {
-		ret = logical_ring_wrap_buffer(ringbuf);
+	wa_ctx_emit(batch, index, MI_BATCH_BUFFER_END);
-		if (unlikely(ret))
-			return ret;
+	return wa_ctx_end(wa_ctx, *offset = index, 1);
-	}
-	if (unlikely(ringbuf->space < bytes)) {
+}
-		ret = logical_ring_wait_for_space(ringbuf, bytes);
-		if (unlikely(ret))
-			return ret;
+static int lrc_setup_wa_ctx_obj(struct intel_engine_cs *ring, u32 size)
-	}
+{
+	int ret;
+	ring->wa_ctx.obj = i915_gem_alloc_object(ring->dev, PAGE_ALIGN(size));
+	if (!ring->wa_ctx.obj) {
+		DRM_DEBUG_DRIVER("alloc LRC WA ctx backing obj failed.\n");
-	return 0;
+		return -ENOMEM;
+	}
+	ret = i915_gem_obj_ggtt_pin(ring->wa_ctx.obj, PAGE_SIZE, 0);
-}
+	if (ret) {
+		DRM_DEBUG_DRIVER("pin LRC WA ctx backing obj failed: %d\n",
+				 ret);
+		drm_gem_object_unreference(&ring->wa_ctx.obj->base);
 		return ret;
-/**
+	}
- * intel_logical_ring_begin() - prepare the logical ringbuffer to accept some commands
+	return 0;
+}
 static void lrc_destroy_wa_ctx_obj(struct intel_engine_cs *ring)
- * @ringbuf: Logical ringbuffer.
+{
+	if (ring->wa_ctx.obj) {
+		i915_gem_object_ggtt_unpin(ring->wa_ctx.obj);
+		drm_gem_object_unreference(&ring->wa_ctx.obj->base);
+		ring->wa_ctx.obj = NULL;
+	}
+}
- * @num_dwords: number of DWORDs that we plan to write to the ringbuffer.
+static int intel_init_workaround_bb(struct intel_engine_cs *ring)
- *
+{
+	int ret;
- * The ringbuffer might not be ready to accept the commands right away (maybe it needs to
+	uint32_t *batch;
+	uint32_t offset;
- * be wrapped, or wait a bit for the tail to be updated). This function takes care of that
+	struct page *page;
- * and also preallocates a request (every workload submission is still mediated through
+	struct i915_ctx_workarounds *wa_ctx = &ring->wa_ctx;
- * requests, same as it did with legacy ringbuffer submission).
- *
- * Return: non-zero if the ringbuffer is not ready to be written to.
- */
+	WARN_ON(ring->id != RCS);
-int intel_logical_ring_begin(struct intel_ringbuffer *ringbuf, int num_dwords)
-	struct intel_engine_cs *ring = ringbuf->ring;
+	/* update this when WA for higher Gen are added */
-	struct drm_device *dev = ring->dev;
+	if (INTEL_INFO(ring->dev)->gen > 9) {
-	struct drm_i915_private *dev_priv = dev->dev_private;
+		DRM_ERROR("WA batch buffer is not initialized for Gen%d\n",
-	int ret;
+			  INTEL_INFO(ring->dev)->gen);
 		return 0;
 	ret = i915_gem_check_wedge(&dev_priv->gpu_error,
 				   dev_priv->mm.interruptible);
+	/* some WA perform writes to scratch page, ensure it is valid */
+	if (ring->scratch.obj == NULL) {
+		DRM_ERROR("scratch page not allocated for %s\n", ring->name);
+		return -EINVAL;
 	if (ret)
 		return ret;
+	ret = lrc_setup_wa_ctx_obj(ring, PAGE_SIZE);
+	if (ret) {
+		DRM_DEBUG_DRIVER("Failed to setup context WA page: %d\n", ret);
+		return ret;
+	}
 	page = i915_gem_object_get_page(wa_ctx->obj, 0);
-	ret = logical_ring_prepare(ringbuf, num_dwords * sizeof(uint32_t));
+	batch = kmap_atomic(page);
-	if (ret)
+	offset = 0;
 		return ret;
+	if (INTEL_INFO(ring->dev)->gen == 8) {
+		ret = gen8_init_indirectctx_bb(ring,
+					       &wa_ctx->indirect_ctx,
+					       batch,
+					       &offset);
+		if (ret)
+			goto out;
+		ret = gen8_init_perctx_bb(ring,
+					  &wa_ctx->per_ctx,
+					  batch,
+					  &offset);
+		if (ret)
+			goto out;
+	} else if (INTEL_INFO(ring->dev)->gen == 9) {
+		ret = gen9_init_indirectctx_bb(ring,
+					       &wa_ctx->indirect_ctx,
+					       batch,
+					       &offset);
+		if (ret)
+			goto out;
+		ret = gen9_init_perctx_bb(ring,
 					  &wa_ctx->per_ctx,
-	/* Preallocate the olr before touching the ring */
+					  batch,
-	ret = logical_ring_alloc_seqno(ring, ringbuf->FIXME_lrc_ctx);
+					  &offset);
+	 *
 	 * WaDisableAsyncFlipPerfMode:snb,ivb,hsw,vlv,bdw,chv
 	 */
 	I915_WRITE(MI_MODE, _MASKED_BIT_ENABLE(ASYNC_FLIP_PERF_DISABLE));
+	I915_WRITE(INSTPM, _MASKED_BIT_ENABLE(INSTPM_FORCE_ORDERING));
+	return init_workarounds_ring(ring);
+}
+static int gen9_init_render_ring(struct intel_engine_cs *ring)
+{
+	int ret;
-	ret = intel_init_pipe_control(ring);
+	ret = gen8_init_common_ring(ring);
 	if (ret)
-		return ret;
-	I915_WRITE(INSTPM, _MASKED_BIT_ENABLE(INSTPM_FORCE_ORDERING));
+		return ret;
 	return init_workarounds_ring(ring);
-}
+}
+static int intel_logical_ring_emit_pdps(struct drm_i915_gem_request *req)
+{
+	struct i915_hw_ppgtt *ppgtt = req->ctx->ppgtt;
+	struct intel_engine_cs *ring = req->ring;
+	struct intel_ringbuffer *ringbuf = req->ringbuf;
+	const int num_lri_cmds = GEN8_LEGACY_PDPES * 2;
+	int i, ret;
+	ret = intel_logical_ring_begin(req, num_lri_cmds * 2 + 2);
+	if (ret)
+		return ret;
+	intel_logical_ring_emit(ringbuf, MI_LOAD_REGISTER_IMM(num_lri_cmds));
+	for (i = GEN8_LEGACY_PDPES - 1; i >= 0; i--) {
+		const dma_addr_t pd_daddr = i915_page_dir_dma_addr(ppgtt, i);
+		intel_logical_ring_emit(ringbuf, GEN8_RING_PDP_UDW(ring, i));
+		intel_logical_ring_emit(ringbuf, upper_32_bits(pd_daddr));
+		intel_logical_ring_emit(ringbuf, GEN8_RING_PDP_LDW(ring, i));
+		intel_logical_ring_emit(ringbuf, lower_32_bits(pd_daddr));
+	}
+	intel_logical_ring_emit(ringbuf, MI_NOOP);
+	intel_logical_ring_advance(ringbuf);
+	return 0;
+}
+static int gen8_emit_bb_start(struct drm_i915_gem_request *req,
-static int gen8_emit_bb_start(struct intel_ringbuffer *ringbuf,
+			      u64 offset, unsigned dispatch_flags)
 			      u64 offset, unsigned flags)
+	struct intel_ringbuffer *ringbuf = req->ringbuf;
+	bool ppgtt = !(dispatch_flags & I915_DISPATCH_SECURE);
+	int ret;
+	/* Don't rely in hw updating PDPs, specially in lite-restore.
+	 * Ideally, we should set Force PD Restore in ctx descriptor,
+	 * but we can't. Force Restore would be a second option, but
+	 * it is unsafe in case of lite-restore (because the ctx is
+	 * not idle). PML4 is allocated during ppgtt init so this is
+	 * not needed in 48-bit.*/
+	if (req->ctx->ppgtt &&
+	    (intel_ring_flag(req->ring) & req->ctx->ppgtt->pd_dirty_rings)) {
+		if (!USES_FULL_48BIT_PPGTT(req->i915) &&
+		    !intel_vgpu_active(req->i915->dev)) {
+			ret = intel_logical_ring_emit_pdps(req);
+			if (ret)
+				return ret;
+		}
-{
-	bool ppgtt = !(flags & I915_DISPATCH_SECURE);
+		req->ctx->ppgtt->pd_dirty_rings &= ~intel_ring_flag(req->ring);
 	int ret;
-	ret = intel_logical_ring_begin(ringbuf, 4);
+	ret = intel_logical_ring_begin(req, 4);
+	if (ret)
+		return ret;
-	if (ret)
+	/* FIXME(BDW): Address space and security selectors. */
-		return ret;
+	intel_logical_ring_emit(ringbuf, MI_BATCH_BUFFER_START_GEN8 |
 				(ppgtt<<8) |
-	/* FIXME(BDW): Address space and security selectors. */
+				(dispatch_flags & I915_DISPATCH_RS ?
 		POSTING_READ(RING_IMR(ring->mmio_base));
+	}
 	spin_unlock_irqrestore(&dev_priv->irq_lock, flags);
+}
-static int gen8_emit_flush(struct intel_ringbuffer *ringbuf,
+static int gen8_emit_flush(struct drm_i915_gem_request *request,
 			   u32 invalidate_domains,
 			   u32 unused)
+{
 	struct intel_ringbuffer *ringbuf = request->ringbuf;
 	struct intel_engine_cs *ring = ringbuf->ring;
 	struct drm_device *dev = ring->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
 	uint32_t cmd;
 	int ret;
-	ret = intel_logical_ring_begin(ringbuf, 4);
+	ret = intel_logical_ring_begin(request, 4);
 	if (ret)
 		return ret;
 	cmd = MI_FLUSH_DW + 1;
+	/* We always require a command barrier so that subsequent
-	if (ring == &dev_priv->ring[VCS]) {
+	 * commands, such as breadcrumb interrupts, are strictly ordered
-		if (invalidate_domains & I915_GEM_GPU_DOMAINS)
+	 * wrt the contents of the write cache being flushed to memory
-			cmd |= MI_INVALIDATE_TLB | MI_INVALIDATE_BSD |
+	 * (and thus being coherent from the CPU).
-				MI_FLUSH_DW_STORE_INDEX |
+	 */
+	cmd |= MI_FLUSH_DW_STORE_INDEX | MI_FLUSH_DW_OP_STOREDW;
-				MI_FLUSH_DW_OP_STOREDW;
-	} else {
+	if (invalidate_domains & I915_GEM_GPU_DOMAINS) {
-		if (invalidate_domains & I915_GEM_DOMAIN_RENDER)
+		cmd |= MI_INVALIDATE_TLB;
-			cmd |= MI_INVALIDATE_TLB | MI_FLUSH_DW_STORE_INDEX |
+		if (ring == &dev_priv->ring[VCS])
-				MI_FLUSH_DW_OP_STOREDW;
+			cmd |= MI_INVALIDATE_BSD;
 	intel_logical_ring_advance(ringbuf);
 	return 0;
 }
-static int gen8_emit_flush_render(struct intel_ringbuffer *ringbuf,
+static int gen8_emit_flush_render(struct drm_i915_gem_request *request,
 				  u32 invalidate_domains,
+				  u32 flush_domains)
 				  u32 flush_domains)
 	struct intel_ringbuffer *ringbuf = request->ringbuf;
+	struct intel_engine_cs *ring = ringbuf->ring;
-	struct intel_engine_cs *ring = ringbuf->ring;
+	u32 scratch_addr = ring->scratch.gtt_offset + 2 * CACHELINE_BYTES;
-	u32 scratch_addr = ring->scratch.gtt_offset + 2 * CACHELINE_BYTES;
+	bool vf_flush_wa;
 	u32 flags = 0;
 	int ret;
 	flags |= PIPE_CONTROL_CS_STALL;
 	if (flush_domains) {
-	if (flush_domains) {
+		flags |= PIPE_CONTROL_RENDER_TARGET_CACHE_FLUSH;
-		flags |= PIPE_CONTROL_RENDER_TARGET_CACHE_FLUSH;
+		flags |= PIPE_CONTROL_DEPTH_CACHE_FLUSH;
-		flags |= PIPE_CONTROL_DEPTH_CACHE_FLUSH;
+		flags |= PIPE_CONTROL_FLUSH_ENABLE;
 		flags |= PIPE_CONTROL_STATE_CACHE_INVALIDATE;
 		flags |= PIPE_CONTROL_QW_WRITE;
 		flags |= PIPE_CONTROL_GLOBAL_GTT_IVB;
+	}
+	/*
+	 * On GEN9+ Before VF_CACHE_INVALIDATE we need to emit a NULL pipe
+	 * control.
+	 */
+	vf_flush_wa = INTEL_INFO(ring->dev)->gen >= 9 &&
+		      flags & PIPE_CONTROL_VF_CACHE_INVALIDATE;
-	ret = intel_logical_ring_begin(ringbuf, 6);
+	ret = intel_logical_ring_begin(request, vf_flush_wa ? 12 : 6);
 	if (ret)
+		return ret;
+	if (vf_flush_wa) {
+		intel_logical_ring_emit(ringbuf, GFX_OP_PIPE_CONTROL(6));
+		intel_logical_ring_emit(ringbuf, 0);
+		intel_logical_ring_emit(ringbuf, 0);
+		intel_logical_ring_emit(ringbuf, 0);
+		intel_logical_ring_emit(ringbuf, 0);
+		intel_logical_ring_emit(ringbuf, 0);
 		return ret;
 	intel_logical_ring_emit(ringbuf, GFX_OP_PIPE_CONTROL(6));
 	intel_logical_ring_emit(ringbuf, flags);
 	intel_logical_ring_emit(ringbuf, scratch_addr);
 static void gen8_set_seqno(struct intel_engine_cs *ring, u32 seqno)
+{
 	intel_write_status_page(ring, I915_GEM_HWS_INDEX, seqno);
+}
+static u32 bxt_a_get_seqno(struct intel_engine_cs *ring, bool lazy_coherency)
+{
+	/*
+	 * On BXT A steppings there is a HW coherency issue whereby the
+	 * MI_STORE_DATA_IMM storing the completed request's seqno
+	 * occasionally doesn't invalidate the CPU cache. Work around this by
+	 * clflushing the corresponding cacheline whenever the caller wants
+	 * the coherency to be guaranteed. Note that this cacheline is known
+	 * to be clean at this point, since we only write it in
+	 * bxt_a_set_seqno(), where we also do a clflush after the write. So
+	 * this clflush in practice becomes an invalidate operation.
+	 */
+	if (!lazy_coherency)
+		intel_flush_status_page(ring, I915_GEM_HWS_INDEX);
+	return intel_read_status_page(ring, I915_GEM_HWS_INDEX);
+}
+static void bxt_a_set_seqno(struct intel_engine_cs *ring, u32 seqno)
+{
+	intel_write_status_page(ring, I915_GEM_HWS_INDEX, seqno);
+	/* See bxt_a_get_seqno() explaining the reason for the clflush. */
+	intel_flush_status_page(ring, I915_GEM_HWS_INDEX);
+}
-static int gen8_emit_request(struct intel_ringbuffer *ringbuf)
+static int gen8_emit_request(struct drm_i915_gem_request *request)
+{
 	struct intel_ringbuffer *ringbuf = request->ringbuf;
 	struct intel_engine_cs *ring = ringbuf->ring;
 	u32 cmd;
+	int ret;
+	/*
+	 * Reserve space for 2 NOOPs at the end of each request to be
+	 * used as a workaround for not being allowed to do lite
-	int ret;
+	 * restore with HEAD==TAIL (WaIdleLiteRestore).
 	 */
-	ret = intel_logical_ring_begin(ringbuf, 6);
+	ret = intel_logical_ring_begin(request, 8);
 	if (ret)
 		return ret;
-	cmd = MI_STORE_DWORD_IMM_GEN8;
+	cmd = MI_STORE_DWORD_IMM_GEN4;
 	cmd |= MI_GLOBAL_GTT;
 	intel_logical_ring_emit(ringbuf, cmd);
 	intel_logical_ring_emit(ringbuf,
 				(ring->status_page.gfx_addr +
 				(I915_GEM_HWS_INDEX << MI_STORE_DWORD_INDEX_SHIFT)));
 	intel_logical_ring_emit(ringbuf, 0);
+	intel_logical_ring_emit(ringbuf, i915_gem_request_get_seqno(request));
+	intel_logical_ring_emit(ringbuf, MI_USER_INTERRUPT);
+	intel_logical_ring_emit(ringbuf, MI_NOOP);
+	intel_logical_ring_advance_and_submit(request);
+	/*
+	 * Here we add two extra NOOPs as padding to avoid
+	 * lite restore of a context with HEAD==TAIL.
+	 */
+	intel_logical_ring_emit(ringbuf, MI_NOOP);
+	intel_logical_ring_emit(ringbuf, MI_NOOP);
+	intel_logical_ring_advance(ringbuf);
+	return 0;
+}
+static int intel_lr_context_render_state_init(struct drm_i915_gem_request *req)
+{
+	struct render_state so;
+	int ret;
-	intel_logical_ring_emit(ringbuf, ring->outstanding_lazy_seqno);
+	ret = i915_gem_render_state_prepare(req->ring, &so);
+	if (ret)
+		return ret;
+	if (so.rodata == NULL)
+		return 0;
+	ret = req->ring->emit_bb_start(req, so.ggtt_offset,
+				       I915_DISPATCH_SECURE);
+	if (ret)
+		goto out;
+	ret = req->ring->emit_bb_start(req,
+				       (so.ggtt_offset + so.aux_batch_offset),
+				       I915_DISPATCH_SECURE);
+	if (ret)
+		goto out;
+	i915_vma_move_to_active(i915_gem_obj_to_ggtt(so.obj), req);
+out:
+	i915_gem_render_state_fini(&so);
+	return ret;
+}
+static int gen8_init_rcs_context(struct drm_i915_gem_request *req)
+{
+	int ret;
+	ret = intel_logical_ring_workarounds_emit(req);
+	if (ret)
+		return ret;
+	ret = intel_rcs_context_init_mocs(req);
+	/*
+	 * Failing to program the MOCS is non-fatal.The system will not
+	 * run at peak performance. So generate an error and carry on.
-	intel_logical_ring_emit(ringbuf, MI_USER_INTERRUPT);
+	 */
-	intel_logical_ring_emit(ringbuf, MI_NOOP);
+	if (ret)
-	intel_logical_ring_advance_and_submit(ringbuf);
+		DRM_ERROR("MOCS failed to program: expect performance issues.\n");
 	dev_priv = ring->dev->dev_private;
-	intel_logical_ring_stop(ring);
-	WARN_ON((I915_READ_MODE(ring) & MODE_IDLE) == 0);
-	ring->preallocated_lazy_request = NULL;
+	intel_logical_ring_stop(ring);
-	ring->outstanding_lazy_seqno = 0;
+	WARN_ON((I915_READ_MODE(ring) & MODE_IDLE) == 0);
+	if (ring->cleanup)
-	if (ring->cleanup)
+		ring->cleanup(ring);
 		ring->cleanup(ring);
 	i915_cmd_parser_fini_ring(ring);
-	i915_cmd_parser_fini_ring(ring);
+	i915_gem_batch_pool_fini(&ring->batch_pool);
+	if (ring->status_page.obj) {
 		kunmap(sg_page(ring->status_page.obj->pages->sgl));
-	if (ring->status_page.obj) {
+		ring->status_page.obj = NULL;
 		kunmap(sg_page(ring->status_page.obj->pages->sgl));
 		ring->status_page.obj = NULL;
 	ring->buffer = NULL;
 	ring->dev = dev;
 	INIT_LIST_HEAD(&ring->active_list);
+	INIT_LIST_HEAD(&ring->request_list);
-	INIT_LIST_HEAD(&ring->request_list);
+	i915_gem_batch_pool_init(dev, &ring->batch_pool);
 	init_waitqueue_head(&ring->irq_queue);
 	INIT_LIST_HEAD(&ring->execlist_queue);
-	INIT_LIST_HEAD(&ring->execlist_retired_req_list);
-	spin_lock_init(&ring->execlist_lock);
+	INIT_LIST_HEAD(&ring->execlist_retired_req_list);
-	ring->next_context_status_buffer = 0;
+	spin_lock_init(&ring->execlist_lock);
-	ret = i915_cmd_parser_init_ring(ring);
-	if (ret)
+	ret = i915_cmd_parser_init_ring(ring);
-		return ret;
+	if (ret)
 		return ret;
-	if (ring->init) {
+	ret = intel_lr_context_deferred_alloc(ring->default_context, ring);
+	if (ret)
+		return ret;
-		ret = ring->init(ring);
+	/* As this is the default context, always pin it */
+	ret = intel_lr_context_do_pin(
+			ring,
+			ring->default_context->engine[ring->id].state,
+			ring->default_context->engine[ring->id].ringbuf);
+	if (ret) {
-		if (ret)
+		DRM_ERROR(
-			return ret;
+			"Failed to pin and map ringbuffer %s: %d\n",
 			ring->name, ret);
 		return ret;
 	ret = intel_lr_context_deferred_create(ring->default_context, ring);
+	return ret;
 	return ret;
 static int logical_render_ring_init(struct drm_device *dev)
 static int logical_render_ring_init(struct drm_device *dev)
 	ring->irq_keep_mask =
 		GT_CONTEXT_SWITCH_INTERRUPT << GEN8_RCS_IRQ_SHIFT;
 	if (HAS_L3_DPF(dev))
 		ring->irq_keep_mask |= GT_RENDER_L3_PARITY_ERROR_INTERRUPT;
+	if (INTEL_INFO(dev)->gen >= 9)
+		ring->init_hw = gen9_init_render_ring;
+	else
-	ring->init = gen8_init_render_ring;
+		ring->init_hw = gen8_init_render_ring;
-	ring->init_context = intel_logical_ring_workarounds_emit;
+	ring->init_context = gen8_init_rcs_context;
+	ring->cleanup = intel_fini_pipe_control;
+	if (IS_BROXTON(dev) && INTEL_REVID(dev) < BXT_REVID_B0) {
+		ring->get_seqno = bxt_a_get_seqno;
+		ring->set_seqno = bxt_a_set_seqno;
-	ring->cleanup = intel_fini_pipe_control;
+	} else {
-	ring->get_seqno = gen8_get_seqno;
+		ring->get_seqno = gen8_get_seqno;
+		ring->set_seqno = gen8_set_seqno;
 	ring->set_seqno = gen8_set_seqno;
 	ring->emit_request = gen8_emit_request;
 	ring->emit_flush = gen8_emit_flush_render;
 	ring->irq_get = gen8_logical_ring_get_irq;
 	ring->irq_put = gen8_logical_ring_put_irq;
+	ring->emit_bb_start = gen8_emit_bb_start;
+	ring->dev = dev;
+	ret = intel_init_pipe_control(ring);
+	if (ret)
+		return ret;
+	ret = intel_init_workaround_bb(ring);
+	if (ret) {
+		/*
+		 * We continue even if we fail to initialize WA batch
+		 * because we only expect rare glitches but nothing
+		 * critical to prevent us from using GPU
+		 */
+		DRM_ERROR("WA batch buffer initialization failed: %d\n",
+			  ret);
-	ring->emit_bb_start = gen8_emit_bb_start;
+	}
+	ret = logical_ring_init(dev, ring);
+	if (ret) {
+		lrc_destroy_wa_ctx_obj(ring);
+	}
-	return logical_ring_init(dev, ring);
+	return ret;
+}
 	ring->irq_enable_mask =
 		GT_RENDER_USER_INTERRUPT << GEN8_VCS1_IRQ_SHIFT;
 	ring->irq_keep_mask =
 		GT_CONTEXT_SWITCH_INTERRUPT << GEN8_VCS1_IRQ_SHIFT;
+	ring->init_hw = gen8_init_common_ring;
+	if (IS_BROXTON(dev) && INTEL_REVID(dev) < BXT_REVID_B0) {
+		ring->get_seqno = bxt_a_get_seqno;
+		ring->set_seqno = bxt_a_set_seqno;
-	ring->init = gen8_init_common_ring;
+	} else {
-	ring->get_seqno = gen8_get_seqno;
+		ring->get_seqno = gen8_get_seqno;
+		ring->set_seqno = gen8_set_seqno;
 	ring->set_seqno = gen8_set_seqno;
 	ring->emit_request = gen8_emit_request;
 	ring->emit_flush = gen8_emit_flush;
 	ring->irq_get = gen8_logical_ring_get_irq;
 	ring->irq_put = gen8_logical_ring_put_irq;
 	ring->irq_enable_mask =
 		GT_RENDER_USER_INTERRUPT << GEN8_VCS2_IRQ_SHIFT;
 	ring->irq_keep_mask =
 		GT_CONTEXT_SWITCH_INTERRUPT << GEN8_VCS2_IRQ_SHIFT;
-	ring->init = gen8_init_common_ring;
+	ring->init_hw = gen8_init_common_ring;
 	ring->get_seqno = gen8_get_seqno;
 	ring->set_seqno = gen8_set_seqno;
 	ring->emit_request = gen8_emit_request;
 	ring->emit_flush = gen8_emit_flush;
 	ring->irq_enable_mask =
 		GT_RENDER_USER_INTERRUPT << GEN8_BCS_IRQ_SHIFT;
 	ring->irq_keep_mask =
 		GT_CONTEXT_SWITCH_INTERRUPT << GEN8_BCS_IRQ_SHIFT;
+	ring->init_hw = gen8_init_common_ring;
+	if (IS_BROXTON(dev) && INTEL_REVID(dev) < BXT_REVID_B0) {
+		ring->get_seqno = bxt_a_get_seqno;
+		ring->set_seqno = bxt_a_set_seqno;
-	ring->init = gen8_init_common_ring;
+	} else {
-	ring->get_seqno = gen8_get_seqno;
+		ring->get_seqno = gen8_get_seqno;
+		ring->set_seqno = gen8_set_seqno;
 	ring->set_seqno = gen8_set_seqno;
 	ring->emit_request = gen8_emit_request;
 	ring->emit_flush = gen8_emit_flush;
 	ring->irq_get = gen8_logical_ring_get_irq;
 	ring->irq_put = gen8_logical_ring_put_irq;
 	ring->irq_enable_mask =
 		GT_RENDER_USER_INTERRUPT << GEN8_VECS_IRQ_SHIFT;
 	ring->irq_keep_mask =
 		GT_CONTEXT_SWITCH_INTERRUPT << GEN8_VECS_IRQ_SHIFT;
+	ring->init_hw = gen8_init_common_ring;
+	if (IS_BROXTON(dev) && INTEL_REVID(dev) < BXT_REVID_B0) {
+		ring->get_seqno = bxt_a_get_seqno;
+		ring->set_seqno = bxt_a_set_seqno;
-	ring->init = gen8_init_common_ring;
+	} else {
-	ring->get_seqno = gen8_get_seqno;
+		ring->get_seqno = gen8_get_seqno;
+		ring->set_seqno = gen8_set_seqno;
 	ring->set_seqno = gen8_set_seqno;
 	ring->emit_request = gen8_emit_request;
 	ring->emit_flush = gen8_emit_flush;
 	ring->irq_get = gen8_logical_ring_get_irq;
 	ring->irq_put = gen8_logical_ring_put_irq;
 		ret = logical_bsd2_ring_init(dev);
 		if (ret)
 			goto cleanup_vebox_ring;
+	}
-	ret = i915_gem_set_seqno(dev, ((u32)~0 - 0x1000));
-	if (ret)
-		goto cleanup_bsd2_ring;
-	return 0;
-cleanup_bsd2_ring:
+	return 0;
 	intel_logical_ring_cleanup(&dev_priv->ring[VCS2]);
 cleanup_vebox_ring:
 	intel_logical_ring_cleanup(&dev_priv->ring[VECS]);
 cleanup_blt_ring:
 	intel_logical_ring_cleanup(&dev_priv->ring[RCS]);
 	return ret;
 }
-int intel_lr_context_render_state_init(struct intel_engine_cs *ring,
+static u32
-				       struct intel_context *ctx)
-{
-	struct intel_ringbuffer *ringbuf = ctx->engine[ring->id].ringbuf;
-	struct render_state so;
-	struct drm_i915_file_private *file_priv = ctx->file_priv;
+make_rpcs(struct drm_device *dev)
+{
-	struct drm_file *file = file_priv ? file_priv->file : NULL;
+	u32 rpcs = 0;
-	int ret;
-	ret = i915_gem_render_state_prepare(ring, &so);
+	/*
-	if (ret)
+	 * No explicit RPCS request is needed to ensure full
-		return ret;
+	 * slice/subslice/EU enablement prior to Gen9.
+	*/
+	if (INTEL_INFO(dev)->gen < 9)
+		return 0;
-	if (so.rodata == NULL)
+	/*
+	 * Starting in Gen9, render power gating can leave
+	 * slice/subslice/EU in a partially enabled state. We
+	 * must make an explicit request through RPCS for full
+	 * enablement.
-		return 0;
+	*/
+	if (INTEL_INFO(dev)->has_slice_pg) {
-	ret = ring->emit_bb_start(ringbuf,
+		rpcs |= GEN8_RPCS_S_CNT_ENABLE;
+		rpcs |= INTEL_INFO(dev)->slice_total <<
+			GEN8_RPCS_S_CNT_SHIFT;
-			so.ggtt_offset,
+		rpcs |= GEN8_RPCS_ENABLE;
+	}
+	if (INTEL_INFO(dev)->has_subslice_pg) {
+		rpcs |= GEN8_RPCS_SS_CNT_ENABLE;
-			I915_DISPATCH_SECURE);
+		rpcs |= INTEL_INFO(dev)->subslice_per_slice <<
+			GEN8_RPCS_SS_CNT_SHIFT;
-	if (ret)
+		rpcs |= GEN8_RPCS_ENABLE;
+	}
-		goto out;
+	if (INTEL_INFO(dev)->has_eu_pg) {
-	i915_vma_move_to_active(i915_gem_obj_to_ggtt(so.obj), ring);
+		rpcs |= INTEL_INFO(dev)->eu_per_subslice <<
+			GEN8_RPCS_EU_MIN_SHIFT;
-	ret = __i915_add_request(ring, file, so.obj, NULL);
+		rpcs |= INTEL_INFO(dev)->eu_per_subslice <<
-	/* intel_logical_ring_add_request moves object to inactive if it
+			GEN8_RPCS_EU_MAX_SHIFT;
-	 * fails */
+		rpcs |= GEN8_RPCS_ENABLE;
 out:
 	i915_gem_object_pin_pages(ctx_obj);
 	/* The second page of the context object contains some fields which must
 	 * be set up prior to the first execution. */
-	page = i915_gem_object_get_page(ctx_obj, 1);
+	page = i915_gem_object_get_page(ctx_obj, LRC_STATE_PN);
 	reg_state = kmap_atomic(page);
 	else
 		reg_state[CTX_LRI_HEADER_0] = MI_LOAD_REGISTER_IMM(11);
 	reg_state[CTX_LRI_HEADER_0] |= MI_LRI_FORCE_POSTED;
 	reg_state[CTX_CONTEXT_CONTROL] = RING_CONTEXT_CONTROL(ring);
 	reg_state[CTX_CONTEXT_CONTROL+1] =
-			_MASKED_BIT_ENABLE((1<<3) | MI_RESTORE_INHIBIT);
+		_MASKED_BIT_ENABLE(CTX_CTRL_INHIBIT_SYN_CTX_SWITCH |
+				   CTX_CTRL_ENGINE_CTX_RESTORE_INHIBIT |
+				   CTX_CTRL_RS_CTX_ENABLE);
 	reg_state[CTX_RING_HEAD] = RING_HEAD(ring->mmio_base);
 	reg_state[CTX_RING_HEAD+1] = 0;
 	reg_state[CTX_RING_TAIL] = RING_TAIL(ring->mmio_base);
 	reg_state[CTX_RING_TAIL+1] = 0;
 	reg_state[CTX_RING_BUFFER_START] = RING_START(ring->mmio_base);
 	reg_state[CTX_SECOND_BB_HEAD_L] = ring->mmio_base + 0x114;
 	reg_state[CTX_SECOND_BB_HEAD_L+1] = 0;
 	reg_state[CTX_SECOND_BB_STATE] = ring->mmio_base + 0x118;
 	reg_state[CTX_SECOND_BB_STATE+1] = 0;
 	if (ring->id == RCS) {
-		/* TODO: according to BSpec, the register state context
-		 * for CHV does not have these. OTOH, these registers do
-		 * exist in CHV. I'm waiting for a clarification */
 		reg_state[CTX_BB_PER_CTX_PTR] = ring->mmio_base + 0x1c0;
 		reg_state[CTX_BB_PER_CTX_PTR+1] = 0;
 		reg_state[CTX_RCS_INDIRECT_CTX] = ring->mmio_base + 0x1c4;
 		reg_state[CTX_RCS_INDIRECT_CTX+1] = 0;
 		reg_state[CTX_RCS_INDIRECT_CTX_OFFSET] = ring->mmio_base + 0x1c8;
 		reg_state[CTX_RCS_INDIRECT_CTX_OFFSET+1] = 0;
+		if (ring->wa_ctx.obj) {
+			struct i915_ctx_workarounds *wa_ctx = &ring->wa_ctx;
+			uint32_t ggtt_offset = i915_gem_obj_ggtt_offset(wa_ctx->obj);
+			reg_state[CTX_RCS_INDIRECT_CTX+1] =
+				(ggtt_offset + wa_ctx->indirect_ctx.offset * sizeof(uint32_t)) |
+				(wa_ctx->indirect_ctx.size / CACHELINE_DWORDS);
+			reg_state[CTX_RCS_INDIRECT_CTX_OFFSET+1] =
+				CTX_RCS_INDIRECT_CTX_OFFSET_DEFAULT << 6;
+			reg_state[CTX_BB_PER_CTX_PTR+1] =
+				(ggtt_offset + wa_ctx->per_ctx.offset * sizeof(uint32_t)) |
+x01;
+		}
+	}
 	reg_state[CTX_LRI_HEADER_1] = MI_LOAD_REGISTER_IMM(9);
 	reg_state[CTX_LRI_HEADER_1] |= MI_LRI_FORCE_POSTED;
 	reg_state[CTX_CTX_TIMESTAMP] = ring->mmio_base + 0x3a8;
 	reg_state[CTX_CTX_TIMESTAMP+1] = 0;
 	reg_state[CTX_PDP2_LDW] = GEN8_RING_PDP_LDW(ring, 2);
 	reg_state[CTX_PDP1_UDW] = GEN8_RING_PDP_UDW(ring, 1);
 	reg_state[CTX_PDP1_LDW] = GEN8_RING_PDP_LDW(ring, 1);
 	reg_state[CTX_PDP0_UDW] = GEN8_RING_PDP_UDW(ring, 0);
 	reg_state[CTX_PDP0_LDW] = GEN8_RING_PDP_LDW(ring, 0);
-	reg_state[CTX_PDP3_UDW+1] = upper_32_bits(ppgtt->pd_dma_addr[3]);
+	if (USES_FULL_48BIT_PPGTT(ppgtt->base.dev)) {
+		/* 64b PPGTT (48bit canonical)
+		 * PDP0_DESCRIPTOR contains the base address to PML4 and
+		 * other PDP Descriptors are ignored.
+		 */
+		ASSIGN_CTX_PML4(ppgtt, reg_state);
+	} else {
+		/* 32b PPGTT
-	reg_state[CTX_PDP3_LDW+1] = lower_32_bits(ppgtt->pd_dma_addr[3]);
+		 * PDP*_DESCRIPTOR contains the base address of space supported.
-	reg_state[CTX_PDP2_UDW+1] = upper_32_bits(ppgtt->pd_dma_addr[2]);
+		 * With dynamic page allocation, PDPs may not be allocated at
-	reg_state[CTX_PDP2_LDW+1] = lower_32_bits(ppgtt->pd_dma_addr[2]);
+		 * this point. Point the unallocated PDPs to the scratch page
+		 */
-	reg_state[CTX_PDP1_UDW+1] = upper_32_bits(ppgtt->pd_dma_addr[1]);
+		ASSIGN_CTX_PDP(ppgtt, reg_state, 3);
-	reg_state[CTX_PDP1_LDW+1] = lower_32_bits(ppgtt->pd_dma_addr[1]);
+		ASSIGN_CTX_PDP(ppgtt, reg_state, 2);
-	reg_state[CTX_PDP0_UDW+1] = upper_32_bits(ppgtt->pd_dma_addr[0]);
+		ASSIGN_CTX_PDP(ppgtt, reg_state, 1);
-	reg_state[CTX_PDP0_LDW+1] = lower_32_bits(ppgtt->pd_dma_addr[0]);
+		ASSIGN_CTX_PDP(ppgtt, reg_state, 0);
+	}
 	if (ring->id == RCS) {
 		reg_state[CTX_LRI_HEADER_2] = MI_LOAD_REGISTER_IMM(1);
-		reg_state[CTX_R_PWR_CLK_STATE] = 0x20c8;
+		reg_state[CTX_R_PWR_CLK_STATE] = GEN8_R_PWR_CLK_STATE;
-		reg_state[CTX_R_PWR_CLK_STATE+1] = 0;
+		reg_state[CTX_R_PWR_CLK_STATE+1] = make_rpcs(dev);
+	}
 	kunmap_atomic(reg_state);
 	ctx_obj->dirty = 1;
-//   set_page_dirty(page);
+	ctx_obj->dirty = 1;
-	i915_gem_object_unpin_pages(ctx_obj);
+    i915_gem_object_unpin_pages(ctx_obj);
 			if (ctx == ring->default_context) {
 				intel_unpin_ringbuffer_obj(ringbuf);
 				i915_gem_object_ggtt_unpin(ctx_obj);
 			}
-			intel_destroy_ringbuffer_obj(ringbuf);
+			WARN_ON(ctx->engine[ring->id].pin_count);
-			kfree(ringbuf);
+			intel_ringbuffer_free(ringbuf);
 			drm_gem_object_unreference(&ctx_obj->base);
+		}
+	}
 static void lrc_setup_hardware_status_page(struct intel_engine_cs *ring,
 		struct drm_i915_gem_object *default_ctx_obj)
+{
+	struct drm_i915_private *dev_priv = ring->dev->dev_private;
-	struct drm_i915_private *dev_priv = ring->dev->dev_private;
+	struct page *page;
-	/* The status page is offset 0 from the default context object
-	 * in LRC mode. */
+	/* The HWSP is part of the default context object in LRC mode. */
-	ring->status_page.gfx_addr = i915_gem_obj_ggtt_offset(default_ctx_obj);
+	ring->status_page.gfx_addr = i915_gem_obj_ggtt_offset(default_ctx_obj)
+			+ LRC_PPHWSP_PN * PAGE_SIZE;
-	ring->status_page.page_addr =
+	page = i915_gem_object_get_page(default_ctx_obj, LRC_PPHWSP_PN);
-			kmap(sg_page(default_ctx_obj->pages->sgl));
+	ring->status_page.page_addr = kmap(page);
 	ring->status_page.obj = default_ctx_obj;
 	I915_WRITE(RING_HWS_PGA(ring->mmio_base),
 			(u32)ring->status_page.gfx_addr);
 	POSTING_READ(RING_HWS_PGA(ring->mmio_base));
+}
 /**
- * intel_lr_context_deferred_create() - create the LRC specific bits of a context
+ * intel_lr_context_deferred_alloc() - create the LRC specific bits of a context
  * @ctx: LR context to create.
  * the creation is a deferred call: it's better to make sure first that we need to use
  * a given ring with the context.
+ *
  * Return: non-zero on error.
  */
-int intel_lr_context_deferred_create(struct intel_context *ctx,
+int intel_lr_context_deferred_alloc(struct intel_context *ctx,
 				     struct intel_engine_cs *ring)
+{
-	const bool is_global_default_ctx = (ctx == ring->default_context);
 	struct drm_device *dev = ring->dev;
 	struct drm_i915_gem_object *ctx_obj;
 	uint32_t context_size;
 	struct intel_ringbuffer *ringbuf;
 	int ret;
 	WARN_ON(ctx->legacy_hw_ctx.rcs_state != NULL);
-	if (ctx->engine[ring->id].state)
-		return 0;
+	WARN_ON(ctx->engine[ring->id].state);
-	context_size = round_up(get_lr_context_size(ring), 4096);
+	context_size = round_up(get_lr_context_size(ring), 4096);
-	ctx_obj = i915_gem_alloc_context_obj(dev, context_size);
-	if (IS_ERR(ctx_obj)) {
-		ret = PTR_ERR(ctx_obj);
-		DRM_DEBUG_DRIVER("Alloc LRC backing obj failed: %d\n", ret);
-		return ret;
-	}
+	/* One extra page as the sharing data between driver and GuC */
+	context_size += PAGE_SIZE * LRC_PPHWSP_PN;
-	if (is_global_default_ctx) {
-		ret = i915_gem_obj_ggtt_pin(ctx_obj, GEN8_LR_CONTEXT_ALIGN, 0);
-		if (ret) {
-			DRM_DEBUG_DRIVER("Pin LRC backing obj failed: %d\n",
+	ctx_obj = i915_gem_alloc_object(dev, context_size);
+	if (!ctx_obj) {
+		DRM_DEBUG_DRIVER("Alloc LRC backing obj failed.\n");
+		return -ENOMEM;
+	}
-					ret);
+	ringbuf = intel_engine_create_ringbuffer(ring, 4 * PAGE_SIZE);
-			drm_gem_object_unreference(&ctx_obj->base);
+	if (IS_ERR(ringbuf)) {
-			return ret;
+		ret = PTR_ERR(ringbuf);
-		}
+		goto error_deref_obj;
-	}
-	ringbuf = kzalloc(sizeof(*ringbuf), GFP_KERNEL);
+	}
 	if (!ringbuf) {
-		DRM_DEBUG_DRIVER("Failed to allocate ringbuffer %s\n",
+	ret = populate_lr_context(ctx, ctx_obj, ring, ringbuf);
-				ring->name);
+	if (ret) {
-		ret = -ENOMEM;
-		goto error_unpin_ctx;
+		DRM_DEBUG_DRIVER("Failed to populate LRC: %d\n", ret);
-	}
-	ringbuf->ring = ring;
-	ringbuf->FIXME_lrc_ctx = ctx;
+		goto error_ringbuf;
+	}
-	ringbuf->size = 32 * PAGE_SIZE;
-	ringbuf->effective_size = ringbuf->size;
+	ctx->engine[ring->id].ringbuf = ringbuf;
-	ringbuf->head = 0;
-	ringbuf->tail = 0;
+	ctx->engine[ring->id].state = ctx_obj;
-	ringbuf->space = ringbuf->size;
-	ringbuf->last_retired_head = -1;
+	if (ctx != ring->default_context && ring->init_context) {
 		struct drm_i915_gem_request *req;
-	if (ringbuf->obj == NULL) {
-		ret = intel_alloc_ringbuffer_obj(dev, ringbuf);
-		if (ret) {
+		ret = i915_gem_request_alloc(ring,
-			DRM_DEBUG_DRIVER(
-				"Failed to allocate ringbuffer obj %s: %d\n",
+			ctx, &req);
-				ring->name, ret);
+		if (ret) {
+			DRM_ERROR("ring create req: %d\n",
-			goto error_free_rbuf;
+				ret);
 			goto error_ringbuf;
+		}
+		ret = ring->init_context(req);
+		if (ret) {
+			DRM_ERROR("ring init context: %d\n",
+				ret);
+			i915_gem_request_cancel(req);
+			goto error_ringbuf;
+		}
+		i915_add_request_no_flush(req);
 		if (is_global_default_ctx) {
-			ret = intel_pin_and_map_ringbuffer_obj(dev, ringbuf);
+	return 0;
-			if (ret) {
+error_ringbuf:
-				DRM_ERROR(
+	intel_ringbuffer_free(ringbuf);
-					"Failed to pin and map ringbuffer %s: %d\n",
+error_deref_obj:
-					ring->name, ret);
+	drm_gem_object_unreference(&ctx_obj->base);
+	ctx->engine[ring->id].ringbuf = NULL;
+	ctx->engine[ring->id].state = NULL;
-				goto error_destroy_rbuf;
+	return ret;
+}
-			}
+void intel_lr_context_reset(struct drm_device *dev,
+			struct intel_context *ctx)
-		}
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
-	}
+	struct intel_engine_cs *ring;
-	ret = populate_lr_context(ctx, ctx_obj, ring, ringbuf);
+	int i;
-	if (ret) {
-		DRM_DEBUG_DRIVER("Failed to populate LRC: %d\n", ret);
-		goto error;
+	for_each_ring(ring, dev_priv, i) {
+		struct drm_i915_gem_object *ctx_obj =
+				ctx->engine[ring->id].state;
-	}
-	ctx->engine[ring->id].ringbuf = ringbuf;
-	ctx->engine[ring->id].state = ctx_obj;
+		struct intel_ringbuffer *ringbuf =
+				ctx->engine[ring->id].ringbuf;
-	if (ctx == ring->default_context)
-		lrc_setup_hardware_status_page(ring, ctx_obj);
-	if (ring->id == RCS && !ctx->rcs_initialized) {
-		if (ring->init_context) {
+		uint32_t *reg_state;
-			ret = ring->init_context(ring, ctx);
-			if (ret)
-				DRM_ERROR("ring init context: %d\n", ret);
-		}
-		ret = intel_lr_context_render_state_init(ring, ctx);
-		if (ret) {
+		struct page *page;
-			DRM_ERROR("Init render state failed: %d\n", ret);
-			ctx->engine[ring->id].ringbuf = NULL;
-			ctx->engine[ring->id].state = NULL;
-			goto error;
-		}
+		if (!ctx_obj)
-		ctx->rcs_initialized = true;
+			continue;

Subversion Repositories Kolibri OS

(root)/drivers/video/drm/i915/intel_lrc.c – Rev 5354 → 6084