drm/i915: Refactor activity tracking for requests

With the introduction of requests, we amplified the number of atomic
refcounted objects we use and update every execbuffer; from none to
several references, and a set of references that need to be changed. We
also introduced interesting side-effects in the order of retiring
requests and objects.

Instead of independently tracking the last request for an object, track
the active objects for each request. The object will reside in the
buffer list of its most recent active request and so we reduce the kref
interchange to a list_move. Now retirements are entirely driven by the
request, dramatically simplifying activity tracking on the object
themselves, and removing the ambiguity between retiring objects and
retiring requests.

Furthermore with the consolidation of managing the activity tracking
centrally, we can look forward to using RCU to enable lockless lookup of
the current active requests for an object. In the future, we will be
able to query the status or wait upon rendering to an object without
even touching the struct_mutex BKL.

All told, less code, simpler and faster, and more extensible.

v2: Add a typedef for the function pointer for convenience later.
v3: Make the noop retirement callback explicit. Allow passing NULL to
the init_request_active() which is expanded to a common noop function.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/1470293567-10811-16-git-send-email-chris@chris-wilson.co.uk

9 files changed, 138 insertions, 243 deletions

diff --git a/drivers/gpu/drm/i915/Makefile b/drivers/gpu/drm/i915/Makefile
index 6092f0ea24df..dda724f04445 100644
--- a/drivers/gpu/drm/i915/Makefile
+++ b/drivers/gpu/drm/i915/Makefile
@@ -25,7 +25,6 @@ i915-$(CONFIG_DEBUG_FS) += i915_debugfs.o
 i915-y += i915_cmd_parser.o \
 	  i915_gem_batch_pool.o \
 	  i915_gem_context.o \
-	  i915_gem_debug.o \
 	  i915_gem_dmabuf.o \
 	  i915_gem_evict.o \
 	  i915_gem_execbuffer.o \
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 17a206f19fcb..6b57d8ff7218 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -432,8 +432,6 @@ void intel_link_compute_m_n(int bpp, int nlanes,
 #define DRIVER_MINOR		6
 #define DRIVER_PATCHLEVEL	0
 
-#define WATCH_LISTS	0
-
 struct opregion_header;
 struct opregion_acpi;
 struct opregion_swsci;
@@ -2153,7 +2151,6 @@ struct drm_i915_gem_object {
 	struct drm_mm_node *stolen;
 	struct list_head global_list;
 
-	struct list_head engine_list[I915_NUM_ENGINES];
 	/** Used in execbuf to temporarily hold a ref */
 	struct list_head obj_exec_link;
 
@@ -3463,13 +3460,6 @@ static inline bool i915_gem_object_needs_bit17_swizzle(struct drm_i915_gem_objec
 		obj->tiling_mode != I915_TILING_NONE;
 }
 
-/* i915_gem_debug.c */
-#if WATCH_LISTS
-int i915_verify_lists(struct drm_device *dev);
-#else
-#define i915_verify_lists(dev) 0
-#endif
-
 /* i915_debugfs.c */
 #ifdef CONFIG_DEBUG_FS
 int i915_debugfs_register(struct drm_i915_private *dev_priv);
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 54732ff6b551..a2eaa442d294 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -43,10 +43,6 @@
 
 static void i915_gem_object_flush_gtt_write_domain(struct drm_i915_gem_object *obj);
 static void i915_gem_object_flush_cpu_write_domain(struct drm_i915_gem_object *obj);
-static void
-i915_gem_object_retire__write(struct drm_i915_gem_object *obj);
-static void
-i915_gem_object_retire__read(struct drm_i915_gem_object *obj, int engine);
 
 static bool cpu_cache_is_coherent(struct drm_device *dev,
 				  enum i915_cache_level level)
@@ -141,7 +137,6 @@ int i915_mutex_lock_interruptible(struct drm_device *dev)
 	if (ret)
 		return ret;
 
-	WARN_ON(i915_verify_lists(dev));
 	return 0;
 }
 
@@ -1339,23 +1334,6 @@ put_rpm:
 	return ret;
 }
 
-static void
-i915_gem_object_retire_request(struct drm_i915_gem_object *obj,
-			       struct drm_i915_gem_request *req)
-{
-	int idx = req->engine->id;
-
-	if (i915_gem_active_peek(&obj->last_read[idx],
-				 &obj->base.dev->struct_mutex) == req)
-		i915_gem_object_retire__read(obj, idx);
-	else if (i915_gem_active_peek(&obj->last_write,
-				      &obj->base.dev->struct_mutex) == req)
-		i915_gem_object_retire__write(obj);
-
-	if (!i915_reset_in_progress(&req->i915->gpu_error))
-		i915_gem_request_retire_upto(req);
-}
-
 /**
  * Ensures that all rendering to the object has completed and the object is
  * safe to unbind from the GTT or access from the CPU.
@@ -1382,18 +1360,10 @@ i915_gem_object_wait_rendering(struct drm_i915_gem_object *obj,
 	}
 
 	for_each_active(active_mask, idx) {
-		struct drm_i915_gem_request *request;
-
-		request = i915_gem_active_peek(&active[idx],
-					       &obj->base.dev->struct_mutex);
-		if (!request)
-			continue;
-
-		ret = i915_wait_request(request);
+		ret = i915_gem_active_wait(&active[idx],
+					   &obj->base.dev->struct_mutex);
 		if (ret)
 			return ret;
-
-		i915_gem_object_retire_request(obj, request);
 	}
 
 	resv = i915_gem_object_get_dmabuf_resv(obj);
@@ -1453,11 +1423,8 @@ i915_gem_object_wait_rendering__nonblocking(struct drm_i915_gem_object *obj,
 		ret = __i915_wait_request(requests[i], true, NULL, rps);
 	mutex_lock(&dev->struct_mutex);
 
-	for (i = 0; i < n; i++) {
-		if (ret == 0)
-			i915_gem_object_retire_request(obj, requests[i]);
+	for (i = 0; i < n; i++)
 		i915_gem_request_put(requests[i]);
-	}
 
 	return ret;
 }
@@ -2376,40 +2343,31 @@ void i915_vma_move_to_active(struct i915_vma *vma,
 		i915_gem_object_get(obj);
 	obj->active |= intel_engine_flag(engine);
 
-	list_move_tail(&obj->engine_list[engine->id], &engine->active_list);
 	i915_gem_active_set(&obj->last_read[engine->id], req);
 
 	list_move_tail(&vma->vm_link, &vma->vm->active_list);
 }
 
 static void
-i915_gem_object_retire__write(struct drm_i915_gem_object *obj)
+i915_gem_object_retire__write(struct i915_gem_active *active,
+			      struct drm_i915_gem_request *request)
 {
-	GEM_BUG_ON(!i915_gem_active_isset(&obj->last_write));
-	GEM_BUG_ON(!(obj->active &
-		     intel_engine_flag(i915_gem_active_get_engine(&obj->last_write,
-								  &obj->base.dev->struct_mutex))));
+	struct drm_i915_gem_object *obj =
+		container_of(active, struct drm_i915_gem_object, last_write);
 
-	i915_gem_active_set(&obj->last_write, NULL);
 	intel_fb_obj_flush(obj, true, ORIGIN_CS);
 }
 
 static void
-i915_gem_object_retire__read(struct drm_i915_gem_object *obj, int idx)
+i915_gem_object_retire__read(struct i915_gem_active *active,
+			     struct drm_i915_gem_request *request)
 {
-	struct intel_engine_cs *engine;
+	int idx = request->engine->id;
+	struct drm_i915_gem_object *obj =
+		container_of(active, struct drm_i915_gem_object, last_read[idx]);
 	struct i915_vma *vma;
 
-	GEM_BUG_ON(!i915_gem_active_isset(&obj->last_read[idx]));
-	GEM_BUG_ON(!(obj->active & (1 << idx)));
-
-	list_del_init(&obj->engine_list[idx]);
-	i915_gem_active_set(&obj->last_read[idx], NULL);
-
-	engine = i915_gem_active_get_engine(&obj->last_write,
-					    &obj->base.dev->struct_mutex);
-	if (engine && engine->id == idx)
-		i915_gem_object_retire__write(obj);
+	GEM_BUG_ON((obj->active & (1 << idx)) == 0);
 
 	obj->active &= ~(1 << idx);
 	if (obj->active)
@@ -2419,15 +2377,13 @@ i915_gem_object_retire__read(struct drm_i915_gem_object *obj, int idx)
 	 * so that we don't steal from recently used but inactive objects
 	 * (unless we are forced to ofc!)
 	 */
-	list_move_tail(&obj->global_list,
-		       &to_i915(obj->base.dev)->mm.bound_list);
+	list_move_tail(&obj->global_list, &request->i915->mm.bound_list);
 
 	list_for_each_entry(vma, &obj->vma_list, obj_link) {
 		if (!list_empty(&vma->vm_link))
 			list_move_tail(&vma->vm_link, &vma->vm->inactive_list);
 	}
 
-	i915_gem_active_set(&obj->last_fence, NULL);
 	i915_gem_object_put(obj);
 }
 
@@ -2505,16 +2461,6 @@ static void i915_gem_reset_engine_cleanup(struct intel_engine_cs *engine)
 {
 	struct intel_ring *ring;
 
-	while (!list_empty(&engine->active_list)) {
-		struct drm_i915_gem_object *obj;
-
-		obj = list_first_entry(&engine->active_list,
-				       struct drm_i915_gem_object,
-				       engine_list[engine->id]);
-
-		i915_gem_object_retire__read(obj, engine->id);
-	}
-
 	/* Mark all pending requests as complete so that any concurrent
 	 * (lockless) lookup doesn't try and wait upon the request as we
 	 * reset it.
@@ -2586,8 +2532,6 @@ void i915_gem_reset(struct drm_device *dev)
 	i915_gem_context_reset(dev);
 
 	i915_gem_restore_fences(dev);
-
-	WARN_ON(i915_verify_lists(dev));
 }
 
 /**
@@ -2597,13 +2541,6 @@ void i915_gem_reset(struct drm_device *dev)
 void
 i915_gem_retire_requests_ring(struct intel_engine_cs *engine)
 {
-	WARN_ON(i915_verify_lists(engine->dev));
-
-	/* Retire requests first as we use it above for the early return.
-	 * If we retire requests last, we may use a later seqno and so clear
-	 * the requests lists without clearing the active list, leading to
-	 * confusion.
-	 */
 	while (!list_empty(&engine->request_list)) {
 		struct drm_i915_gem_request *request;
 
@@ -2616,26 +2553,6 @@ i915_gem_retire_requests_ring(struct intel_engine_cs *engine)
 
 		i915_gem_request_retire_upto(request);
 	}
-
-	/* Move any buffers on the active list that are no longer referenced
-	 * by the ringbuffer to the flushing/inactive lists as appropriate,
-	 * before we free the context associated with the requests.
-	 */
-	while (!list_empty(&engine->active_list)) {
-		struct drm_i915_gem_object *obj;
-
-		obj = list_first_entry(&engine->active_list,
-				       struct drm_i915_gem_object,
-				       engine_list[engine->id]);
-
-		if (!list_empty(&i915_gem_active_peek(&obj->last_read[engine->id],
-						      &obj->base.dev->struct_mutex)->link))
-			break;
-
-		i915_gem_object_retire__read(obj, engine->id);
-	}
-
-	WARN_ON(i915_verify_lists(engine->dev));
 }
 
 void i915_gem_retire_requests(struct drm_i915_private *dev_priv)
@@ -2818,8 +2735,7 @@ out:
 }
 
 static int
-__i915_gem_object_sync(struct drm_i915_gem_object *obj,
-		       struct drm_i915_gem_request *to,
+__i915_gem_object_sync(struct drm_i915_gem_request *to,
 		       struct drm_i915_gem_request *from)
 {
 	int ret;
@@ -2827,9 +2743,6 @@ __i915_gem_object_sync(struct drm_i915_gem_object *obj,
 	if (to->engine == from->engine)
 		return 0;
 
-	if (i915_gem_request_completed(from))
-		return 0;
-
 	if (!i915.semaphores) {
 		ret = __i915_wait_request(from,
 					  from->i915->mm.interruptible,
@@ -2837,8 +2750,6 @@ __i915_gem_object_sync(struct drm_i915_gem_object *obj,
 					  NO_WAITBOOST);
 		if (ret)
 			return ret;
-
-		i915_gem_object_retire_request(obj, from);
 	} else {
 		int idx = intel_engine_sync_index(from->engine, to->engine);
 		if (from->fence.seqno <= from->engine->semaphore.sync_seqno[idx])
@@ -2905,7 +2816,7 @@ i915_gem_object_sync(struct drm_i915_gem_object *obj,
 		if (!request)
 			continue;
 
-		ret = __i915_gem_object_sync(obj, to, request);
+		ret = __i915_gem_object_sync(to, request);
 		if (ret)
 			return ret;
 	}
@@ -3041,7 +2952,6 @@ int i915_gem_wait_for_idle(struct drm_i915_private *dev_priv)
 			return ret;
 	}
 
-	WARN_ON(i915_verify_lists(dev));
 	return 0;
 }
 
@@ -4081,7 +3991,11 @@ void i915_gem_object_init(struct drm_i915_gem_object *obj,
 
 	INIT_LIST_HEAD(&obj->global_list);
 	for (i = 0; i < I915_NUM_ENGINES; i++)
-		INIT_LIST_HEAD(&obj->engine_list[i]);
+		init_request_active(&obj->last_read[i],
+				    i915_gem_object_retire__read);
+	init_request_active(&obj->last_write,
+			    i915_gem_object_retire__write);
+	init_request_active(&obj->last_fence, NULL);
 	INIT_LIST_HEAD(&obj->obj_exec_link);
 	INIT_LIST_HEAD(&obj->vma_list);
 	INIT_LIST_HEAD(&obj->batch_pool_link);
@@ -4574,7 +4488,6 @@ i915_gem_cleanup_engines(struct drm_device *dev)
 static void
 init_engine_lists(struct intel_engine_cs *engine)
 {
-	INIT_LIST_HEAD(&engine->active_list);
 	INIT_LIST_HEAD(&engine->request_list);
 }
 
diff --git a/drivers/gpu/drm/i915/i915_gem_debug.c b/drivers/gpu/drm/i915/i915_gem_debug.c
deleted file mode 100644
index a56516482394..000000000000
--- a/drivers/gpu/drm/i915/i915_gem_debug.c
+++ /dev/null
@@ -1,70 +0,0 @@
-/*
- * Copyright © 2008 Intel Corporation
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice (including the next
- * paragraph) shall be included in all copies or substantial portions of the
- * Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
- * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
- * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
- * IN THE SOFTWARE.
- *
- * Authors:
- *    Keith Packard <keithp@keithp.com>
- *
- */
-
-#include <drm/drmP.h>
-#include <drm/i915_drm.h>
-#include "i915_drv.h"
-
-#if WATCH_LISTS
-int
-i915_verify_lists(struct drm_device *dev)
-{
-	static int warned;
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct drm_i915_gem_object *obj;
-	struct intel_engine_cs *engine;
-	int err = 0;
-
-	if (warned)
-		return 0;
-
-	for_each_engine(engine, dev_priv) {
-		list_for_each_entry(obj, &engine->active_list,
-				    engine_list[engine->id]) {
-			if (obj->base.dev != dev ||
-			    !atomic_read(&obj->base.refcount.refcount)) {
-				DRM_ERROR("%s: freed active obj %p\n",
-					  engine->name, obj);
-				err++;
-				break;
-			} else if (!obj->active ||
-				   obj->last_read_req[engine->id] == NULL) {
-				DRM_ERROR("%s: invalid active obj %p\n",
-					  engine->name, obj);
-				err++;
-			} else if (obj->base.write_domain) {
-				DRM_ERROR("%s: invalid write obj %p (w %x)\n",
-					  engine->name,
-					  obj, obj->base.write_domain);
-				err++;
-			}
-		}
-	}
-
-	return warned = err;
-}
-#endif /* WATCH_LIST */
diff --git a/drivers/gpu/drm/i915/i915_gem_fence.c b/drivers/gpu/drm/i915/i915_gem_fence.c
index a4ec4fecbea0..dbaab9ce29c9 100644
--- a/drivers/gpu/drm/i915/i915_gem_fence.c
+++ b/drivers/gpu/drm/i915/i915_gem_fence.c
@@ -261,15 +261,8 @@ static inline void i915_gem_object_fence_lost(struct drm_i915_gem_object *obj)
 static int
 i915_gem_object_wait_fence(struct drm_i915_gem_object *obj)
 {
-	int ret;
-
-	ret = i915_gem_active_wait(&obj->last_fence,
-				   &obj->base.dev->struct_mutex);
-	if (ret)
-		return ret;
-
-	i915_gem_active_set(&obj->last_fence, NULL);
-	return 0;
+	return i915_gem_active_retire(&obj->last_fence,
+				      &obj->base.dev->struct_mutex);
 }
 
 /**
diff --git a/drivers/gpu/drm/i915/i915_gem_request.c b/drivers/gpu/drm/i915/i915_gem_request.c
index 7802156d481a..cdaaeb6dd913 100644
--- a/drivers/gpu/drm/i915/i915_gem_request.c
+++ b/drivers/gpu/drm/i915/i915_gem_request.c
@@ -22,6 +22,8 @@
  *
  */
 
+#include <linux/prefetch.h>
+
 #include "i915_drv.h"
 
 static const char *i915_fence_get_driver_name(struct fence *fence)
@@ -157,8 +159,16 @@ i915_gem_request_remove_from_client(struct drm_i915_gem_request *request)
 	request->pid = NULL;
 }
 
+void i915_gem_retire_noop(struct i915_gem_active *active,
+			  struct drm_i915_gem_request *request)
+{
+	/* Space left intentionally blank */
+}
+
 static void i915_gem_request_retire(struct drm_i915_gem_request *request)
 {
+	struct i915_gem_active *active, *next;
+
 	trace_i915_gem_request_retire(request);
 	list_del_init(&request->link);
 
@@ -172,6 +182,33 @@ static void i915_gem_request_retire(struct drm_i915_gem_request *request)
 	 */
 	request->ring->last_retired_head = request->postfix;
 
+	/* Walk through the active list, calling retire on each. This allows
+	 * objects to track their GPU activity and mark themselves as idle
+	 * when their *last* active request is completed (updating state
+	 * tracking lists for eviction, active references for GEM, etc).
+	 *
+	 * As the ->retire() may free the node, we decouple it first and
+	 * pass along the auxiliary information (to avoid dereferencing
+	 * the node after the callback).
+	 */
+	list_for_each_entry_safe(active, next, &request->active_list, link) {
+		/* In microbenchmarks or focusing upon time inside the kernel,
+		 * we may spend an inordinate amount of time simply handling
+		 * the retirement of requests and processing their callbacks.
+		 * Of which, this loop itself is particularly hot due to the
+		 * cache misses when jumping around the list of i915_gem_active.
+		 * So we try to keep this loop as streamlined as possible and
+		 * also prefetch the next i915_gem_active to try and hide
+		 * the likely cache miss.
+		 */
+		prefetchw(next);
+
+		INIT_LIST_HEAD(&active->link);
+		active->request = NULL;
+
+		active->retire(active, request);
+	}
+
 	i915_gem_request_remove_from_client(request);
 
 	if (request->previous_context) {
@@ -200,8 +237,6 @@ void i915_gem_request_retire_upto(struct drm_i915_gem_request *req)
 
 		i915_gem_request_retire(tmp);
 	} while (tmp != req);
-
-	WARN_ON(i915_verify_lists(engine->dev));
 }
 
 static int i915_gem_check_wedge(unsigned int reset_counter, bool interruptible)
@@ -336,6 +371,7 @@ i915_gem_request_alloc(struct intel_engine_cs *engine,
 		   engine->fence_context,
 		   seqno);
 
+	INIT_LIST_HEAD(&req->active_list);
 	req->i915 = dev_priv;
 	req->engine = engine;
 	req->ctx = i915_gem_context_get(ctx);
@@ -570,9 +606,6 @@ int __i915_wait_request(struct drm_i915_gem_request *req,
 
 	might_sleep();
 
-	if (list_empty(&req->link))
-		return 0;
-
 	if (i915_gem_request_completed(req))
 		return 0;
 
@@ -705,10 +738,13 @@ int i915_wait_request(struct drm_i915_gem_request *req)
 {
 	int ret;
 
-	GEM_BUG_ON(!req);
 	lockdep_assert_held(&req->i915->drm.struct_mutex);
+	GEM_BUG_ON(list_empty(&req->link));
 
-	ret = __i915_wait_request(req, req->i915->mm.interruptible, NULL, NULL);
+	ret = __i915_wait_request(req,
+				  req->i915->mm.interruptible,
+				  NULL,
+				  NULL);
 	if (ret)
 		return ret;
 
diff --git a/drivers/gpu/drm/i915/i915_gem_request.h b/drivers/gpu/drm/i915/i915_gem_request.h
index 3e406613cdc9..6cfae2079e19 100644
--- a/drivers/gpu/drm/i915/i915_gem_request.h
+++ b/drivers/gpu/drm/i915/i915_gem_request.h
@@ -101,6 +101,7 @@ struct drm_i915_gem_request {
 	 * error state dump only).
 	 */
 	struct drm_i915_gem_object *batch_obj;
+	struct list_head active_list;
 
 	/** Time at which this request was emitted, in jiffies. */
 	unsigned long emitted_jiffies;
@@ -213,8 +214,12 @@ struct intel_rps_client;
 int __i915_wait_request(struct drm_i915_gem_request *req,
 			bool interruptible,
 			s64 *timeout,
-			struct intel_rps_client *rps);
-int __must_check i915_wait_request(struct drm_i915_gem_request *req);
+			struct intel_rps_client *rps)
+	__attribute__((nonnull(1)));
+
+int __must_check
+i915_wait_request(struct drm_i915_gem_request *req)
+	__attribute__((nonnull));
 
 static inline u32 intel_engine_get_seqno(struct intel_engine_cs *engine);
 
@@ -276,10 +281,39 @@ static inline bool i915_spin_request(const struct drm_i915_gem_request *request,
  * can then perform any action, such as delayed freeing of an active
  * resource including itself.
  */
+struct i915_gem_active;
+
+typedef void (*i915_gem_retire_fn)(struct i915_gem_active *,
+				   struct drm_i915_gem_request *);
+
 struct i915_gem_active {
 	struct drm_i915_gem_request *request;
+	struct list_head link;
+	i915_gem_retire_fn retire;
 };
 
+void i915_gem_retire_noop(struct i915_gem_active *,
+			  struct drm_i915_gem_request *request);
+
+/**
+ * init_request_active - prepares the activity tracker for use
+ * @active - the active tracker
+ * @func - a callback when then the tracker is retired (becomes idle),
+ *         can be NULL
+ *
+ * init_request_active() prepares the embedded @active struct for use as
+ * an activity tracker, that is for tracking the last known active request
+ * associated with it. When the last request becomes idle, when it is retired
+ * after completion, the optional callback @func is invoked.
+ */
+static inline void
+init_request_active(struct i915_gem_active *active,
+		    i915_gem_retire_fn retire)
+{
+	INIT_LIST_HEAD(&active->link);
+	active->retire = retire ?: i915_gem_retire_noop;
+}
+
 /**
  * i915_gem_active_set - updates the tracker to watch the current request
  * @active - the active tracker
@@ -293,7 +327,8 @@ static inline void
 i915_gem_active_set(struct i915_gem_active *active,
 		    struct drm_i915_gem_request *request)
 {
-	i915_gem_request_assign(&active->request, request);
+	list_move(&active->link, &request->active_list);
+	active->request = request;
 }
 
 static inline struct drm_i915_gem_request *
@@ -303,17 +338,23 @@ __i915_gem_active_peek(const struct i915_gem_active *active)
 }
 
 /**
- * i915_gem_active_peek - report the request being monitored
+ * i915_gem_active_peek - report the active request being monitored
  * @active - the active tracker
  *
- * i915_gem_active_peek() returns the current request being tracked, or NULL.
- * It does not obtain a reference on the request for the caller, so the
- * caller must hold struct_mutex.
+ * i915_gem_active_peek() returns the current request being tracked if
+ * still active, or NULL. It does not obtain a reference on the request
+ * for the caller, so the caller must hold struct_mutex.
  */
 static inline struct drm_i915_gem_request *
 i915_gem_active_peek(const struct i915_gem_active *active, struct mutex *mutex)
 {
-	return active->request;
+	struct drm_i915_gem_request *request;
+
+	request = active->request;
+	if (!request || i915_gem_request_completed(request))
+		return NULL;
+
+	return request;
 }
 
 /**
@@ -326,13 +367,7 @@ i915_gem_active_peek(const struct i915_gem_active *active, struct mutex *mutex)
 static inline struct drm_i915_gem_request *
 i915_gem_active_get(const struct i915_gem_active *active, struct mutex *mutex)
 {
-	struct drm_i915_gem_request *request;
-
-	request = i915_gem_active_peek(active, mutex);
-	if (!request || i915_gem_request_completed(request))
-		return NULL;
-
-	return i915_gem_request_get(request);
+	return i915_gem_request_get(i915_gem_active_peek(active, mutex));
 }
 
 /**
@@ -361,13 +396,7 @@ static inline bool
 i915_gem_active_is_idle(const struct i915_gem_active *active,
 			struct mutex *mutex)
 {
-	struct drm_i915_gem_request *request;
-
-	request = i915_gem_active_peek(active, mutex);
-	if (!request || i915_gem_request_completed(request))
-		return true;
-
-	return false;
+	return !i915_gem_active_peek(active, mutex);
 }
 
 /**
@@ -377,6 +406,9 @@ i915_gem_active_is_idle(const struct i915_gem_active *active,
  * i915_gem_active_wait() waits until the request is completed before
  * returning. Note that it does not guarantee that the request is
  * retired first, see i915_gem_active_retire().
+ *
+ * i915_gem_active_wait() returns immediately if the active
+ * request is already complete.
  */
 static inline int __must_check
 i915_gem_active_wait(const struct i915_gem_active *active, struct mutex *mutex)
@@ -387,7 +419,7 @@ i915_gem_active_wait(const struct i915_gem_active *active, struct mutex *mutex)
 	if (!request)
 		return 0;
 
-	return i915_wait_request(request);
+	return __i915_wait_request(request, true, NULL, NULL);
 }
 
 /**
@@ -400,10 +432,25 @@ i915_gem_active_wait(const struct i915_gem_active *active, struct mutex *mutex)
  * tracker is idle, the function returns immediately.
  */
 static inline int __must_check
-i915_gem_active_retire(const struct i915_gem_active *active,
+i915_gem_active_retire(struct i915_gem_active *active,
 		       struct mutex *mutex)
 {
-	return i915_gem_active_wait(active, mutex);
+	struct drm_i915_gem_request *request;
+	int ret;
+
+	request = active->request;
+	if (!request)
+		return 0;
+
+	ret = __i915_wait_request(request, true, NULL, NULL);
+	if (ret)
+		return ret;
+
+	list_del_init(&active->link);
+	active->request = NULL;
+	active->retire(active, request);
+
+	return 0;
 }
 
 /* Convenience functions for peeking at state inside active's request whilst
diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c
index 4ec914ed9104..202ad83b3dd1 100644
--- a/drivers/gpu/drm/i915/intel_engine_cs.c
+++ b/drivers/gpu/drm/i915/intel_engine_cs.c
@@ -177,7 +177,6 @@ void intel_engine_init_hangcheck(struct intel_engine_cs *engine)
  */
 void intel_engine_setup_common(struct intel_engine_cs *engine)
 {
-	INIT_LIST_HEAD(&engine->active_list);
 	INIT_LIST_HEAD(&engine->request_list);
 	INIT_LIST_HEAD(&engine->buffers);
 	INIT_LIST_HEAD(&engine->execlist_queue);
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index 170624171887..236e7a2f5689 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -293,18 +293,6 @@ struct intel_engine_cs {
 	u32 ctx_desc_template;
 
 	/**
-	 * List of objects currently involved in rendering from the
-	 * ringbuffer.
-	 *
-	 * Includes buffers having the contents of their GPU caches
-	 * flushed, not necessarily primitives.  last_read_req
-	 * represents when the rendering involved will be completed.
-	 *
-	 * A reference is held on the buffer while on this list.
-	 */
-	struct list_head active_list;
-
-	/**
 	 * List of breadcrumbs associated with GPU requests currently
 	 * outstanding.
 	 */