Merge tag 'drm-next-2019-09-18' of git://anongit.freedesktop.org/drm/drm

Pull drm updates from Dave Airlie:
 "This is the main pull request for 5.4-rc1 merge window. I don't think
  there is anything outstanding so next week should just be fixes, but
  we'll see if I missed anything. I landed some fixes earlier in the
  week but got delayed writing summary and sending it out, due to a mix
  of sick kid and jetlag!

  There are some fixes pending, but I'd rather get the main merge out of
  the way instead of delaying it longer.

  It's also pretty large in commit count and new amd header file size.
  The largest thing is four new amdgpu products (navi12/14, arcturus and
  renoir APU support).

  Otherwise it's pretty much lots of work across the board, i915 has
  started landing tigerlake support, lots of icelake fixes and lots of
  locking reworking for future gpu support, lots of header file rework
  (drmP.h is nearly gone), some old legacy hacks (DRM_WAIT_ON) have been
  put into the places they are needed.

  uapi:
   - content protection type property for HDCP

  core:
   - rework include dependencies
   - lots of drmP.h removals
   - link rate calculation robustness fix
   - make fb helper map only when required
   - add connector->DDC adapter link
   - DRM_WAIT_ON removed
   - drop DRM_AUTH usage from drivers

  dma-buf:
   - reservation object fence helper

  dma-fence:
   - shrink dma_fence struct
   - merge signal functions
   - store timestamps in dma_fence
   - selftests

  ttm:
   - embed drm_get_object struct into ttm_buffer_object
   - release_notify callback

  bridges:
   - sii902x - audio graph card support
   - tc358767 - aux data handling rework
   - ti-snd64dsi86 - debugfs support, DSI mode flags support

  panels:
   - Support for GiantPlus GPM940B0, Sharp LQ070Y3DG3B, Ortustech
     COM37H3M, Novatek NT39016, Sharp LS020B1DD01D, Raydium RM67191, Boe
     Himax8279d, Sharp LD-D5116Z01B
   - TI nspire, NEC NL8048HL11, LG Philips LB035Q02, Sharp LS037V7DW01,
     Sony ACX565AKM, Toppoly TD028TTEC1 Toppoly TD043MTEA1

  i915:
   - Initial tigerlake platform support
   - Locking simplification work, general all over refactoring.
   - Selftests
   - HDCP debug info improvements
   - DSI properties
   - Icelake display PLL fixes, colorspace fixes, bandwidth fixes, DSI
     suspend/resume
   - GuC fixes
   - Perf fixes
   - ElkhartLake enablement
   - DP MST fixes
   - GVT - command parser enhancements

  amdgpu:
   - add wipe memory on release flag for buffer creation
   - Navi12/14 support (may be marked experimental)
   - Arcturus support
   - Renoir APU support
   - mclk DPM for Navi
   - DC display fixes
   - Raven scatter/gather support
   - RAS support for GFX
   - Navi12 + Arcturus power features
   - GPU reset for Picasso
   - smu11 i2c controller support

  amdkfd:
   - navi12/14 support
   - Arcturus support

  radeon:
   - kexec fix

  nouveau:
   - improved display color management
   - detect lack of GPU power cables

  vmwgfx:
   - evicition priority support
   - remove unused security feature

  msm:
   - msm8998 display support
   - better async commit support for cursor updates

  etnaviv:
   - per-process address space support
   - performance counter fixes
   - softpin support

  mcde:
   - DCS transfers fix

  exynos:
   - drmP.h cleanup

  lima:
   - reduce logging

  kirin:
   - misc clenaups

  komeda:
   - dual-link support
   - DT memory regions

  hisilicon:
   - misc fixes

  imx:
   - IPUv3 image converter fixes
   - 32-bit RGB V4L2 pixel format support

  ingenic:
   - more support for panel related cases

  mgag200:
   - cursor support fix

  panfrost:
   - export GPU features register to userspace
   - gpu heap allocations
   - per-fd address space support

  pl111:
   - CLD pads wiring support removed from DT

  rockchip:
   - rework to use DRM PSR helpers
   - fix bug in VOP_WIN_GET macro
   - DSI DT binding rework

  sun4i:
   - improve support for color encoding and range
   - DDC enabled GPIO

  tinydrm:
   - rework SPI support
   - improve MIPI-DBI support
   - moved to drm/tiny

  vkms:
   - rework CRC tracking

  dw-hdmi:
   - get_eld and i2s improvements

  gm12u320:
   - misc fixes

  meson:
   - global code cleanup
   - vpu feature detect

  omap:
   - alpha/pixel blend mode properties

  rcar-du:
   - misc fixes"

* tag 'drm-next-2019-09-18' of git://anongit.freedesktop.org/drm/drm: (2112 commits)
  drm/nouveau/bar/gm20b: Avoid BAR1 teardown during init
  drm/nouveau: Fix ordering between TTM and GEM release
  drm/nouveau/prime: Extend DMA reservation object lock
  drm/nouveau: Fix fallout from reservation object rework
  drm/nouveau/kms/nv50-: Don't create MSTMs for eDP connectors
  drm/i915: Use NOEVICT for first pass on attemping to pin a GGTT mmap
  drm/i915: to make vgpu ppgtt notificaiton as atomic operation
  drm/i915: Flush the existing fence before GGTT read/write
  drm/i915: Hold irq-off for the entire fake lock period
  drm/i915/gvt: update RING_START reg of vGPU when the context is submitted to i915
  drm/i915/gvt: update vgpu workload head pointer correctly
  drm/mcde: Fix DSI transfers
  drm/msm: Use the correct dma_sync calls harder
  drm/msm: remove unlikely() from WARN_ON() conditions
  drm/msm/dsi: Fix return value check for clk_get_parent
  drm/msm: add atomic traces
  drm/msm/dpu: async commit support
  drm/msm: async commit support
  drm/msm: split power control from prepare/complete_commit
  drm/msm: add kms->flush_commit()
  ...

1 files changed, 198 insertions, 183 deletions

diff --git a/drivers/gpu/drm/i915/i915_request.c b/drivers/gpu/drm/i915/i915_request.c
index a195a92d0105..a53777dd371c 100644
--- a/drivers/gpu/drm/i915/i915_request.c
+++ b/drivers/gpu/drm/i915/i915_request.c
@@ -35,6 +35,7 @@
 #include "i915_active.h"
 #include "i915_drv.h"
 #include "i915_globals.h"
+#include "i915_trace.h"
 #include "intel_pm.h"
 
 struct execute_cb {
@@ -119,12 +120,56 @@ const struct dma_fence_ops i915_fence_ops = {
 	.release = i915_fence_release,
 };
 
+static void irq_execute_cb(struct irq_work *wrk)
+{
+	struct execute_cb *cb = container_of(wrk, typeof(*cb), work);
+
+	i915_sw_fence_complete(cb->fence);
+	kmem_cache_free(global.slab_execute_cbs, cb);
+}
+
+static void irq_execute_cb_hook(struct irq_work *wrk)
+{
+	struct execute_cb *cb = container_of(wrk, typeof(*cb), work);
+
+	cb->hook(container_of(cb->fence, struct i915_request, submit),
+		 &cb->signal->fence);
+	i915_request_put(cb->signal);
+
+	irq_execute_cb(wrk);
+}
+
+static void __notify_execute_cb(struct i915_request *rq)
+{
+	struct execute_cb *cb;
+
+	lockdep_assert_held(&rq->lock);
+
+	if (list_empty(&rq->execute_cb))
+		return;
+
+	list_for_each_entry(cb, &rq->execute_cb, link)
+		irq_work_queue(&cb->work);
+
+	/*
+	 * XXX Rollback on __i915_request_unsubmit()
+	 *
+	 * In the future, perhaps when we have an active time-slicing scheduler,
+	 * it will be interesting to unsubmit parallel execution and remove
+	 * busywaits from the GPU until their master is restarted. This is
+	 * quite hairy, we have to carefully rollback the fence and do a
+	 * preempt-to-idle cycle on the target engine, all the while the
+	 * master execute_cb may refire.
+	 */
+	INIT_LIST_HEAD(&rq->execute_cb);
+}
+
 static inline void
-i915_request_remove_from_client(struct i915_request *request)
+remove_from_client(struct i915_request *request)
 {
 	struct drm_i915_file_private *file_priv;
 
-	file_priv = request->file_priv;
+	file_priv = READ_ONCE(request->file_priv);
 	if (!file_priv)
 		return;
 
@@ -136,40 +181,6 @@ i915_request_remove_from_client(struct i915_request *request)
 	spin_unlock(&file_priv->mm.lock);
 }
 
-static void advance_ring(struct i915_request *request)
-{
-	struct intel_ring *ring = request->ring;
-	unsigned int tail;
-
-	/*
-	 * We know the GPU must have read the request to have
-	 * sent us the seqno + interrupt, so use the position
-	 * of tail of the request to update the last known position
-	 * of the GPU head.
-	 *
-	 * Note this requires that we are always called in request
-	 * completion order.
-	 */
-	GEM_BUG_ON(!list_is_first(&request->ring_link, &ring->request_list));
-	if (list_is_last(&request->ring_link, &ring->request_list)) {
-		/*
-		 * We may race here with execlists resubmitting this request
-		 * as we retire it. The resubmission will move the ring->tail
-		 * forwards (to request->wa_tail). We either read the
-		 * current value that was written to hw, or the value that
-		 * is just about to be. Either works, if we miss the last two
-		 * noops - they are safe to be replayed on a reset.
-		 */
-		tail = READ_ONCE(request->tail);
-		list_del(&ring->active_link);
-	} else {
-		tail = request->postfix;
-	}
-	list_del_init(&request->ring_link);
-
-	ring->head = tail;
-}
-
 static void free_capture_list(struct i915_request *request)
 {
 	struct i915_capture_list *capture;
@@ -187,7 +198,7 @@ static bool i915_request_retire(struct i915_request *rq)
 {
 	struct i915_active_request *active, *next;
 
-	lockdep_assert_held(&rq->i915->drm.struct_mutex);
+	lockdep_assert_held(&rq->timeline->mutex);
 	if (!i915_request_completed(rq))
 		return false;
 
@@ -199,7 +210,17 @@ static bool i915_request_retire(struct i915_request *rq)
 	GEM_BUG_ON(!i915_sw_fence_signaled(&rq->submit));
 	trace_i915_request_retire(rq);
 
-	advance_ring(rq);
+	/*
+	 * We know the GPU must have read the request to have
+	 * sent us the seqno + interrupt, so use the position
+	 * of tail of the request to update the last known position
+	 * of the GPU head.
+	 *
+	 * Note this requires that we are always called in request
+	 * completion order.
+	 */
+	GEM_BUG_ON(!list_is_first(&rq->link, &rq->timeline->requests));
+	rq->ring->head = rq->postfix;
 
 	/*
 	 * Walk through the active list, calling retire on each. This allows
@@ -232,6 +253,12 @@ static bool i915_request_retire(struct i915_request *rq)
 
 	local_irq_disable();
 
+	/*
+	 * We only loosely track inflight requests across preemption,
+	 * and so we may find ourselves attempting to retire a _completed_
+	 * request that we have removed from the HW and put back on a run
+	 * queue.
+	 */
 	spin_lock(&rq->engine->active.lock);
 	list_del(&rq->sched.link);
 	spin_unlock(&rq->engine->active.lock);
@@ -242,20 +269,25 @@ static bool i915_request_retire(struct i915_request *rq)
 		dma_fence_signal_locked(&rq->fence);
 	if (test_bit(DMA_FENCE_FLAG_ENABLE_SIGNAL_BIT, &rq->fence.flags))
 		i915_request_cancel_breadcrumb(rq);
-	if (rq->waitboost) {
+	if (i915_request_has_waitboost(rq)) {
 		GEM_BUG_ON(!atomic_read(&rq->i915->gt_pm.rps.num_waiters));
 		atomic_dec(&rq->i915->gt_pm.rps.num_waiters);
 	}
+	if (!test_bit(I915_FENCE_FLAG_ACTIVE, &rq->fence.flags)) {
+		set_bit(I915_FENCE_FLAG_ACTIVE, &rq->fence.flags);
+		__notify_execute_cb(rq);
+	}
+	GEM_BUG_ON(!list_empty(&rq->execute_cb));
 	spin_unlock(&rq->lock);
 
 	local_irq_enable();
 
+	remove_from_client(rq);
+	list_del(&rq->link);
+
 	intel_context_exit(rq->hw_context);
 	intel_context_unpin(rq->hw_context);
 
-	i915_request_remove_from_client(rq);
-	list_del(&rq->link);
-
 	free_capture_list(rq);
 	i915_sched_node_fini(&rq->sched);
 	i915_request_put(rq);
@@ -265,7 +297,7 @@ static bool i915_request_retire(struct i915_request *rq)
 
 void i915_request_retire_upto(struct i915_request *rq)
 {
-	struct intel_ring *ring = rq->ring;
+	struct intel_timeline * const tl = rq->timeline;
 	struct i915_request *tmp;
 
 	GEM_TRACE("%s fence %llx:%lld, current %d\n",
@@ -273,62 +305,14 @@ void i915_request_retire_upto(struct i915_request *rq)
 		  rq->fence.context, rq->fence.seqno,
 		  hwsp_seqno(rq));
 
-	lockdep_assert_held(&rq->i915->drm.struct_mutex);
+	lockdep_assert_held(&tl->mutex);
 	GEM_BUG_ON(!i915_request_completed(rq));
 
-	if (list_empty(&rq->ring_link))
-		return;
-
 	do {
-		tmp = list_first_entry(&ring->request_list,
-				       typeof(*tmp), ring_link);
+		tmp = list_first_entry(&tl->requests, typeof(*tmp), link);
 	} while (i915_request_retire(tmp) && tmp != rq);
 }
 
-static void irq_execute_cb(struct irq_work *wrk)
-{
-	struct execute_cb *cb = container_of(wrk, typeof(*cb), work);
-
-	i915_sw_fence_complete(cb->fence);
-	kmem_cache_free(global.slab_execute_cbs, cb);
-}
-
-static void irq_execute_cb_hook(struct irq_work *wrk)
-{
-	struct execute_cb *cb = container_of(wrk, typeof(*cb), work);
-
-	cb->hook(container_of(cb->fence, struct i915_request, submit),
-		 &cb->signal->fence);
-	i915_request_put(cb->signal);
-
-	irq_execute_cb(wrk);
-}
-
-static void __notify_execute_cb(struct i915_request *rq)
-{
-	struct execute_cb *cb;
-
-	lockdep_assert_held(&rq->lock);
-
-	if (list_empty(&rq->execute_cb))
-		return;
-
-	list_for_each_entry(cb, &rq->execute_cb, link)
-		irq_work_queue(&cb->work);
-
-	/*
-	 * XXX Rollback on __i915_request_unsubmit()
-	 *
-	 * In the future, perhaps when we have an active time-slicing scheduler,
-	 * it will be interesting to unsubmit parallel execution and remove
-	 * busywaits from the GPU until their master is restarted. This is
-	 * quite hairy, we have to carefully rollback the fence and do a
-	 * preempt-to-idle cycle on the target engine, all the while the
-	 * master execute_cb may refire.
-	 */
-	INIT_LIST_HEAD(&rq->execute_cb);
-}
-
 static int
 __i915_request_await_execution(struct i915_request *rq,
 			       struct i915_request *signal,
@@ -512,6 +496,10 @@ submit_notify(struct i915_sw_fence *fence, enum i915_sw_fence_notify state)
 	switch (state) {
 	case FENCE_COMPLETE:
 		trace_i915_request_submit(request);
+
+		if (unlikely(fence->error))
+			i915_request_skip(request, fence->error);
+
 		/*
 		 * We need to serialize use of the submit_request() callback
 		 * with its hotplugging performed during an emergency
@@ -552,29 +540,28 @@ semaphore_notify(struct i915_sw_fence *fence, enum i915_sw_fence_notify state)
 	return NOTIFY_DONE;
 }
 
-static void ring_retire_requests(struct intel_ring *ring)
+static void retire_requests(struct intel_timeline *tl)
 {
 	struct i915_request *rq, *rn;
 
-	list_for_each_entry_safe(rq, rn, &ring->request_list, ring_link)
+	list_for_each_entry_safe(rq, rn, &tl->requests, link)
 		if (!i915_request_retire(rq))
 			break;
 }
 
 static noinline struct i915_request *
-request_alloc_slow(struct intel_context *ce, gfp_t gfp)
+request_alloc_slow(struct intel_timeline *tl, gfp_t gfp)
 {
-	struct intel_ring *ring = ce->ring;
 	struct i915_request *rq;
 
-	if (list_empty(&ring->request_list))
+	if (list_empty(&tl->requests))
 		goto out;
 
 	if (!gfpflags_allow_blocking(gfp))
 		goto out;
 
 	/* Move our oldest request to the slab-cache (if not in use!) */
-	rq = list_first_entry(&ring->request_list, typeof(*rq), ring_link);
+	rq = list_first_entry(&tl->requests, typeof(*rq), link);
 	i915_request_retire(rq);
 
 	rq = kmem_cache_alloc(global.slab_requests,
@@ -583,11 +570,11 @@ request_alloc_slow(struct intel_context *ce, gfp_t gfp)
 		return rq;
 
 	/* Ratelimit ourselves to prevent oom from malicious clients */
-	rq = list_last_entry(&ring->request_list, typeof(*rq), ring_link);
+	rq = list_last_entry(&tl->requests, typeof(*rq), link);
 	cond_synchronize_rcu(rq->rcustate);
 
 	/* Retire our old requests in the hope that we free some */
-	ring_retire_requests(ring);
+	retire_requests(tl);
 
 out:
 	return kmem_cache_alloc(global.slab_requests, gfp);
@@ -596,7 +583,7 @@ out:
 struct i915_request *
 __i915_request_create(struct intel_context *ce, gfp_t gfp)
 {
-	struct i915_timeline *tl = ce->ring->timeline;
+	struct intel_timeline *tl = ce->timeline;
 	struct i915_request *rq;
 	u32 seqno;
 	int ret;
@@ -638,14 +625,14 @@ __i915_request_create(struct intel_context *ce, gfp_t gfp)
 	rq = kmem_cache_alloc(global.slab_requests,
 			      gfp | __GFP_RETRY_MAYFAIL | __GFP_NOWARN);
 	if (unlikely(!rq)) {
-		rq = request_alloc_slow(ce, gfp);
+		rq = request_alloc_slow(tl, gfp);
 		if (!rq) {
 			ret = -ENOMEM;
 			goto err_unreserve;
 		}
 	}
 
-	ret = i915_timeline_get_seqno(tl, rq, &seqno);
+	ret = intel_timeline_get_seqno(tl, rq, &seqno);
 	if (ret)
 		goto err_free;
 
@@ -673,7 +660,7 @@ __i915_request_create(struct intel_context *ce, gfp_t gfp)
 	rq->file_priv = NULL;
 	rq->batch = NULL;
 	rq->capture_list = NULL;
-	rq->waitboost = false;
+	rq->flags = 0;
 	rq->execution_mask = ALL_ENGINES;
 
 	INIT_LIST_HEAD(&rq->active_list);
@@ -730,15 +717,15 @@ struct i915_request *
 i915_request_create(struct intel_context *ce)
 {
 	struct i915_request *rq;
-	int err;
+	struct intel_timeline *tl;
 
-	err = intel_context_timeline_lock(ce);
-	if (err)
-		return ERR_PTR(err);
+	tl = intel_context_timeline_lock(ce);
+	if (IS_ERR(tl))
+		return ERR_CAST(tl);
 
 	/* Move our oldest request to the slab-cache (if not in use!) */
-	rq = list_first_entry(&ce->ring->request_list, typeof(*rq), ring_link);
-	if (!list_is_last(&rq->ring_link, &ce->ring->request_list))
+	rq = list_first_entry(&tl->requests, typeof(*rq), link);
+	if (!list_is_last(&rq->link, &tl->requests))
 		i915_request_retire(rq);
 
 	intel_context_enter(ce);
@@ -748,23 +735,23 @@ i915_request_create(struct intel_context *ce)
 		goto err_unlock;
 
 	/* Check that we do not interrupt ourselves with a new request */
-	rq->cookie = lockdep_pin_lock(&ce->ring->timeline->mutex);
+	rq->cookie = lockdep_pin_lock(&tl->mutex);
 
 	return rq;
 
 err_unlock:
-	intel_context_timeline_unlock(ce);
+	intel_context_timeline_unlock(tl);
 	return rq;
 }
 
 static int
 i915_request_await_start(struct i915_request *rq, struct i915_request *signal)
 {
-	if (list_is_first(&signal->ring_link, &signal->ring->request_list))
+	if (list_is_first(&signal->link, &signal->timeline->requests))
 		return 0;
 
-	signal = list_prev_entry(signal, ring_link);
-	if (i915_timeline_sync_is_later(rq->timeline, &signal->fence))
+	signal = list_prev_entry(signal, link);
+	if (intel_timeline_sync_is_later(rq->timeline, &signal->fence))
 		return 0;
 
 	return i915_sw_fence_await_dma_fence(&rq->submit,
@@ -818,7 +805,7 @@ emit_semaphore_wait(struct i915_request *to,
 		return err;
 
 	/* We need to pin the signaler's HWSP until we are finished reading. */
-	err = i915_timeline_read_hwsp(from, to, &hwsp_offset);
+	err = intel_timeline_read_hwsp(from, to, &hwsp_offset);
 	if (err)
 		return err;
 
@@ -928,8 +915,8 @@ i915_request_await_dma_fence(struct i915_request *rq, struct dma_fence *fence)
 			continue;
 
 		/* Squash repeated waits to the same timelines */
-		if (fence->context != rq->i915->mm.unordered_timeline &&
-		    i915_timeline_sync_is_later(rq->timeline, fence))
+		if (fence->context &&
+		    intel_timeline_sync_is_later(rq->timeline, fence))
 			continue;
 
 		if (dma_fence_is_i915(fence))
@@ -942,8 +929,8 @@ i915_request_await_dma_fence(struct i915_request *rq, struct dma_fence *fence)
 			return ret;
 
 		/* Record the latest fence used against each timeline */
-		if (fence->context != rq->i915->mm.unordered_timeline)
-			i915_timeline_sync_set(rq->timeline, fence);
+		if (fence->context)
+			intel_timeline_sync_set(rq->timeline, fence);
 	} while (--nchild);
 
 	return 0;
@@ -1027,7 +1014,7 @@ i915_request_await_object(struct i915_request *to,
 		struct dma_fence **shared;
 		unsigned int count, i;
 
-		ret = reservation_object_get_fences_rcu(obj->base.resv,
+		ret = dma_resv_get_fences_rcu(obj->base.resv,
 							&excl, &count, &shared);
 		if (ret)
 			return ret;
@@ -1044,7 +1031,7 @@ i915_request_await_object(struct i915_request *to,
 			dma_fence_put(shared[i]);
 		kfree(shared);
 	} else {
-		excl = reservation_object_get_excl_rcu(obj->base.resv);
+		excl = dma_resv_get_excl_rcu(obj->base.resv);
 	}
 
 	if (excl) {
@@ -1065,6 +1052,9 @@ void i915_request_skip(struct i915_request *rq, int error)
 	GEM_BUG_ON(!IS_ERR_VALUE((long)error));
 	dma_fence_set_error(&rq->fence, error);
 
+	if (rq->infix == rq->postfix)
+		return;
+
 	/*
 	 * As this request likely depends on state from the lost
 	 * context, clear out all the user operations leaving the
@@ -1076,12 +1066,13 @@ void i915_request_skip(struct i915_request *rq, int error)
 		head = 0;
 	}
 	memset(vaddr + head, 0, rq->postfix - head);
+	rq->infix = rq->postfix;
 }
 
 static struct i915_request *
 __i915_request_add_to_timeline(struct i915_request *rq)
 {
-	struct i915_timeline *timeline = rq->timeline;
+	struct intel_timeline *timeline = rq->timeline;
 	struct i915_request *prev;
 
 	/*
@@ -1104,7 +1095,8 @@ __i915_request_add_to_timeline(struct i915_request *rq)
 	 * precludes optimising to use semaphores serialisation of a single
 	 * timeline across engines.
 	 */
-	prev = rcu_dereference_protected(timeline->last_request.request, 1);
+	prev = rcu_dereference_protected(timeline->last_request.request,
+					 lockdep_is_held(&timeline->mutex));
 	if (prev && !i915_request_completed(prev)) {
 		if (is_power_of_2(prev->engine->mask | rq->engine->mask))
 			i915_sw_fence_await_sw_fence(&rq->submit,
@@ -1143,7 +1135,6 @@ struct i915_request *__i915_request_commit(struct i915_request *rq)
 {
 	struct intel_engine_cs *engine = rq->engine;
 	struct intel_ring *ring = rq->ring;
-	struct i915_request *prev;
 	u32 *cs;
 
 	GEM_TRACE("%s fence %llx:%lld\n",
@@ -1156,6 +1147,7 @@ struct i915_request *__i915_request_commit(struct i915_request *rq)
 	 */
 	GEM_BUG_ON(rq->reserved_space > ring->space);
 	rq->reserved_space = 0;
+	rq->emitted_jiffies = jiffies;
 
 	/*
 	 * Record the position of the start of the breadcrumb so that
@@ -1167,13 +1159,12 @@ struct i915_request *__i915_request_commit(struct i915_request *rq)
 	GEM_BUG_ON(IS_ERR(cs));
 	rq->postfix = intel_ring_offset(rq, cs);
 
-	prev = __i915_request_add_to_timeline(rq);
-
-	list_add_tail(&rq->ring_link, &ring->request_list);
-	if (list_is_first(&rq->ring_link, &ring->request_list))
-		list_add(&ring->active_link, &rq->i915->gt.active_rings);
-	rq->emitted_jiffies = jiffies;
+	return __i915_request_add_to_timeline(rq);
+}
 
+void __i915_request_queue(struct i915_request *rq,
+			  const struct i915_sched_attr *attr)
+{
 	/*
 	 * Let the backend know a new request has arrived that may need
 	 * to adjust the existing execution schedule due to a high priority
@@ -1185,57 +1176,54 @@ struct i915_request *__i915_request_commit(struct i915_request *rq)
 	 * decide whether to preempt the entire chain so that it is ready to
 	 * run at the earliest possible convenience.
 	 */
-	local_bh_disable();
 	i915_sw_fence_commit(&rq->semaphore);
-	rcu_read_lock(); /* RCU serialisation for set-wedged protection */
-	if (engine->schedule) {
-		struct i915_sched_attr attr = rq->gem_context->sched;
-
-		/*
-		 * Boost actual workloads past semaphores!
-		 *
-		 * With semaphores we spin on one engine waiting for another,
-		 * simply to reduce the latency of starting our work when
-		 * the signaler completes. However, if there is any other
-		 * work that we could be doing on this engine instead, that
-		 * is better utilisation and will reduce the overall duration
-		 * of the current work. To avoid PI boosting a semaphore
-		 * far in the distance past over useful work, we keep a history
-		 * of any semaphore use along our dependency chain.
-		 */
-		if (!(rq->sched.flags & I915_SCHED_HAS_SEMAPHORE_CHAIN))
-			attr.priority |= I915_PRIORITY_NOSEMAPHORE;
-
-		/*
-		 * Boost priorities to new clients (new request flows).
-		 *
-		 * Allow interactive/synchronous clients to jump ahead of
-		 * the bulk clients. (FQ_CODEL)
-		 */
-		if (list_empty(&rq->sched.signalers_list))
-			attr.priority |= I915_PRIORITY_WAIT;
-
-		engine->schedule(rq, &attr);
-	}
-	rcu_read_unlock();
+	if (attr && rq->engine->schedule)
+		rq->engine->schedule(rq, attr);
 	i915_sw_fence_commit(&rq->submit);
-	local_bh_enable(); /* Kick the execlists tasklet if just scheduled */
-
-	return prev;
 }
 
 void i915_request_add(struct i915_request *rq)
 {
+	struct i915_sched_attr attr = rq->gem_context->sched;
+	struct intel_timeline * const tl = rq->timeline;
 	struct i915_request *prev;
 
-	lockdep_assert_held(&rq->timeline->mutex);
-	lockdep_unpin_lock(&rq->timeline->mutex, rq->cookie);
+	lockdep_assert_held(&tl->mutex);
+	lockdep_unpin_lock(&tl->mutex, rq->cookie);
 
 	trace_i915_request_add(rq);
 
 	prev = __i915_request_commit(rq);
 
 	/*
+	 * Boost actual workloads past semaphores!
+	 *
+	 * With semaphores we spin on one engine waiting for another,
+	 * simply to reduce the latency of starting our work when
+	 * the signaler completes. However, if there is any other
+	 * work that we could be doing on this engine instead, that
+	 * is better utilisation and will reduce the overall duration
+	 * of the current work. To avoid PI boosting a semaphore
+	 * far in the distance past over useful work, we keep a history
+	 * of any semaphore use along our dependency chain.
+	 */
+	if (!(rq->sched.flags & I915_SCHED_HAS_SEMAPHORE_CHAIN))
+		attr.priority |= I915_PRIORITY_NOSEMAPHORE;
+
+	/*
+	 * Boost priorities to new clients (new request flows).
+	 *
+	 * Allow interactive/synchronous clients to jump ahead of
+	 * the bulk clients. (FQ_CODEL)
+	 */
+	if (list_empty(&rq->sched.signalers_list))
+		attr.priority |= I915_PRIORITY_WAIT;
+
+	local_bh_disable();
+	__i915_request_queue(rq, &attr);
+	local_bh_enable(); /* Kick the execlists tasklet if just scheduled */
+
+	/*
 	 * In typical scenarios, we do not expect the previous request on
 	 * the timeline to be still tracked by timeline->last_request if it
 	 * has been completed. If the completed request is still here, that
@@ -1252,10 +1240,10 @@ void i915_request_add(struct i915_request *rq)
 	 * work on behalf of others -- but instead we should benefit from
 	 * improved resource management. (Well, that's the theory at least.)
 	 */
-	if (prev && i915_request_completed(prev))
+	if (prev && i915_request_completed(prev) && prev->timeline == tl)
 		i915_request_retire_upto(prev);
 
-	mutex_unlock(&rq->timeline->mutex);
+	mutex_unlock(&tl->mutex);
 }
 
 static unsigned long local_clock_us(unsigned int *cpu)
@@ -1390,8 +1378,7 @@ long i915_request_wait(struct i915_request *rq,
 	 * serialise wait/reset with an explicit lock, we do want
 	 * lockdep to detect potential dependency cycles.
 	 */
-	mutex_acquire(&rq->i915->gpu_error.wedge_mutex.dep_map,
-		      0, 0, _THIS_IP_);
+	mutex_acquire(&rq->engine->gt->reset.mutex.dep_map, 0, 0, _THIS_IP_);
 
 	/*
 	 * Optimistic spin before touching IRQs.
@@ -1447,8 +1434,10 @@ long i915_request_wait(struct i915_request *rq,
 	for (;;) {
 		set_current_state(state);
 
-		if (i915_request_completed(rq))
+		if (i915_request_completed(rq)) {
+			dma_fence_signal(&rq->fence);
 			break;
+		}
 
 		if (signal_pending_state(state, current)) {
 			timeout = -ERESTARTSYS;
@@ -1467,25 +1456,51 @@ long i915_request_wait(struct i915_request *rq,
 	dma_fence_remove_callback(&rq->fence, &wait.cb);
 
 out:
-	mutex_release(&rq->i915->gpu_error.wedge_mutex.dep_map, 0, _THIS_IP_);
+	mutex_release(&rq->engine->gt->reset.mutex.dep_map, 0, _THIS_IP_);
 	trace_i915_request_wait_end(rq);
 	return timeout;
 }
 
 bool i915_retire_requests(struct drm_i915_private *i915)
 {
-	struct intel_ring *ring, *tmp;
+	struct intel_gt_timelines *timelines = &i915->gt.timelines;
+	struct intel_timeline *tl, *tn;
+	unsigned long flags;
+	LIST_HEAD(free);
+
+	spin_lock_irqsave(&timelines->lock, flags);
+	list_for_each_entry_safe(tl, tn, &timelines->active_list, link) {
+		if (!mutex_trylock(&tl->mutex))
+			continue;
+
+		intel_timeline_get(tl);
+		GEM_BUG_ON(!tl->active_count);
+		tl->active_count++; /* pin the list element */
+		spin_unlock_irqrestore(&timelines->lock, flags);
 
-	lockdep_assert_held(&i915->drm.struct_mutex);
+		retire_requests(tl);
 
-	list_for_each_entry_safe(ring, tmp,
-				 &i915->gt.active_rings, active_link) {
-		intel_ring_get(ring); /* last rq holds reference! */
-		ring_retire_requests(ring);
-		intel_ring_put(ring);
+		spin_lock_irqsave(&timelines->lock, flags);
+
+		/* Resume iteration after dropping lock */
+		list_safe_reset_next(tl, tn, link);
+		if (!--tl->active_count)
+			list_del(&tl->link);
+
+		mutex_unlock(&tl->mutex);
+
+		/* Defer the final release to after the spinlock */
+		if (refcount_dec_and_test(&tl->kref.refcount)) {
+			GEM_BUG_ON(tl->active_count);
+			list_add(&tl->link, &free);
+		}
 	}
+	spin_unlock_irqrestore(&timelines->lock, flags);
+
+	list_for_each_entry_safe(tl, tn, &free, link)
+		__intel_timeline_free(&tl->kref);
 
-	return !list_empty(&i915->gt.active_rings);
+	return !list_empty(&timelines->active_list);
 }
 
 #if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)