[PATCH 1/6] drm/i915: Complete the fences as they are cancelled due to wedging

[PATCH 1/6] drm/i915: Complete the fences as they are cancelled due to wedging

[Chris Wilson]

We inspect the requests under the assumption that they will be marked as
completed when they are removed from the queue. Currently however, in the
process of wedging the requests will be removed from the queue before they
are completed, so rearrange the code to complete the fences before the
locks are dropped.

<1>[  354.473346] BUG: unable to handle kernel NULL pointer dereference at 0000000000000250
<6>[  354.473363] PGD 0 P4D 0
<4>[  354.473370] Oops: 0000 [#1] PREEMPT SMP PTI
<4>[  354.473380] CPU: 0 PID: 4470 Comm: gem_eio Tainted: G     U            4.20.0-rc4-CI-CI_DRM_5216+ #1
<4>[  354.473393] Hardware name: Intel Corporation NUC7CJYH/NUC7JYB, BIOS JYGLKCPX.86A.0027.2018.0125.1347 01/25/2018
<4>[  354.473480] RIP: 0010:__i915_schedule+0x311/0x5e0 [i915]
<4>[  354.473490] Code: 49 89 44 24 20 4d 89 4c 24 28 4d 89 29 44 39 b3 a0 04 00 00 7d 3a 41 8b 44 24 78 85 c0 74 13 48 8b 93 78 04 00 00 48 83 e2 fc <39> 82 50 02 00 00 79 1e 44 89 b3 a0 04 00 00 48 8d bb d0 03 00 00
<4>[  354.473515] RSP: 0018:ffffc900001bba90 EFLAGS: 00010046
<4>[  354.473524] RAX: 0000000000000003 RBX: ffff8882624c8008 RCX: f34a737800000000
<4>[  354.473535] RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffff8882624c8048
<4>[  354.473545] RBP: ffffc900001bbab0 R08: 000000005963f1f1 R09: 0000000000000000
<4>[  354.473556] R10: ffffc900001bba10 R11: ffff8882624c8060 R12: ffff88824fdd7b98
<4>[  354.473567] R13: ffff88824fdd7bb8 R14: 0000000000000001 R15: ffff88824fdd7750
<4>[  354.473578] FS:  00007f44b4b5b980(0000) GS:ffff888277e00000(0000) knlGS:0000000000000000
<4>[  354.473590] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
<4>[  354.473599] CR2: 0000000000000250 CR3: 000000026976e000 CR4: 0000000000340ef0
<4>[  354.473611] Call Trace:
<4>[  354.473622]  ? lock_acquire+0xa6/0x1c0
<4>[  354.473677]  ? i915_schedule_bump_priority+0x57/0xd0 [i915]
<4>[  354.473736]  i915_schedule_bump_priority+0x72/0xd0 [i915]
<4>[  354.473792]  i915_request_wait+0x4db/0x840 [i915]
<4>[  354.473804]  ? get_pwq.isra.4+0x2c/0x50
<4>[  354.473813]  ? ___preempt_schedule+0x16/0x18
<4>[  354.473824]  ? wake_up_q+0x70/0x70
<4>[  354.473831]  ? wake_up_q+0x70/0x70
<4>[  354.473882]  ? gen6_rps_boost+0x118/0x120 [i915]
<4>[  354.473936]  i915_gem_object_wait_fence+0x8a/0x110 [i915]
<4>[  354.473991]  i915_gem_object_wait+0x113/0x500 [i915]
<4>[  354.474047]  i915_gem_wait_ioctl+0x11c/0x2f0 [i915]
<4>[  354.474101]  ? i915_gem_unset_wedged+0x210/0x210 [i915]
<4>[  354.474113]  drm_ioctl_kernel+0x81/0xf0
<4>[  354.474123]  drm_ioctl+0x2de/0x390
<4>[  354.474175]  ? i915_gem_unset_wedged+0x210/0x210 [i915]
<4>[  354.474187]  ? finish_task_switch+0x95/0x260
<4>[  354.474197]  ? lock_acquire+0xa6/0x1c0
<4>[  354.474207]  do_vfs_ioctl+0xa0/0x6e0
<4>[  354.474217]  ? __fget+0xfc/0x1e0
<4>[  354.474225]  ksys_ioctl+0x35/0x60
<4>[  354.474233]  __x64_sys_ioctl+0x11/0x20
<4>[  354.474241]  do_syscall_64+0x55/0x190
<4>[  354.474251]  entry_SYSCALL_64_after_hwframe+0x49/0xbe
<4>[  354.474260] RIP: 0033:0x7f44b3de65d7
<4>[  354.474267] Code: b3 66 90 48 8b 05 b1 48 2d 00 64 c7 00 26 00 00 00 48 c7 c0 ff ff ff ff c3 66 2e 0f 1f 84 00 00 00 00 00 b8 10 00 00 00 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d 81 48 2d 00 f7 d8 64 89 01 48
<4>[  354.474293] RSP: 002b:00007fff974948e8 EFLAGS: 00000246 ORIG_RAX: 0000000000000010
<4>[  354.474305] RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007f44b3de65d7
<4>[  354.474316] RDX: 00007fff97494940 RSI: 00000000c010646c RDI: 0000000000000007
<4>[  354.474327] RBP: 00007fff97494940 R08: 0000000000000000 R09: 00007f44b40bbc40
<4>[  354.474337] R10: 0000000000000000 R11: 0000000000000246 R12: 00000000c010646c
<4>[  354.474348] R13: 0000000000000007 R14: 0000000000000000 R15: 0000000000000000

v2: Avoid floating requests.
v3: Can't call dma_fence_signal() under the timeline lock!

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/i915_gem.c         | 56 ++++++-------------------
 drivers/gpu/drm/i915/intel_lrc.c        | 13 ++++--
 drivers/gpu/drm/i915/intel_ringbuffer.c | 13 +++++-
 3 files changed, 33 insertions(+), 49 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index c55b1f75c980..0eaa8a6f0423 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -3308,16 +3308,6 @@ void i915_gem_reset_finish(struct drm_i915_private *dev_priv)
 }
 
 static void nop_submit_request(struct i915_request *request)
-{
-	GEM_TRACE("%s fence %llx:%d -> -EIO\n",
-		  request->engine->name,
-		  request->fence.context, request->fence.seqno);
-	dma_fence_set_error(&request->fence, -EIO);
-
-	i915_request_submit(request);
-}
-
-static void nop_complete_submit_request(struct i915_request *request)
 {
 	unsigned long flags;
 
@@ -3330,6 +3320,8 @@ static void nop_complete_submit_request(struct i915_request *request)
 	__i915_request_submit(request);
 	intel_engine_init_global_seqno(request->engine, request->global_seqno);
 	spin_unlock_irqrestore(&request->engine->timeline.lock, flags);
+
+	dma_fence_signal(&request->fence);
 }
 
 void i915_gem_set_wedged(struct drm_i915_private *i915)
@@ -3354,57 +3346,33 @@ void i915_gem_set_wedged(struct drm_i915_private *i915)
 	 * rolling the global seqno forward (since this would complete requests
 	 * for which we haven't set the fence error to EIO yet).
 	 */
-	for_each_engine(engine, i915, id) {
+	for_each_engine(engine, i915, id)
 		i915_gem_reset_prepare_engine(engine);
 
-		engine->submit_request = nop_submit_request;
-		engine->schedule = NULL;
-	}
-	i915->caps.scheduler = 0;
-
 	/* Even if the GPU reset fails, it should still stop the engines */
 	if (INTEL_GEN(i915) >= 5)
 		intel_gpu_reset(i915, ALL_ENGINES);
 
-	/*
-	 * Make sure no one is running the old callback before we proceed with
-	 * cancelling requests and resetting the completion tracking. Otherwise
-	 * we might submit a request to the hardware which never completes.
-	 */
-	synchronize_rcu();
-
 	for_each_engine(engine, i915, id) {
-		/* Mark all executing requests as skipped */
-		engine->cancel_requests(engine);
-
-		/*
-		 * Only once we've force-cancelled all in-flight requests can we
-		 * start to complete all requests.
-		 */
-		engine->submit_request = nop_complete_submit_request;
+		engine->submit_request = nop_submit_request;
+		engine->schedule = NULL;
 	}
+	i915->caps.scheduler = 0;
 
 	/*
 	 * Make sure no request can slip through without getting completed by
 	 * either this call here to intel_engine_init_global_seqno, or the one
-	 * in nop_complete_submit_request.
+	 * in nop_submit_request.
 	 */
 	synchronize_rcu();
 
-	for_each_engine(engine, i915, id) {
-		unsigned long flags;
-
-		/*
-		 * Mark all pending requests as complete so that any concurrent
-		 * (lockless) lookup doesn't try and wait upon the request as we
-		 * reset it.
-		 */
-		spin_lock_irqsave(&engine->timeline.lock, flags);
-		intel_engine_init_global_seqno(engine,
-					       intel_engine_last_submit(engine));
-		spin_unlock_irqrestore(&engine->timeline.lock, flags);
+	/* Mark all executing requests as skipped */
+	for_each_engine(engine, i915, id)
+		engine->cancel_requests(engine);
 
+	for_each_engine(engine, i915, id) {
 		i915_gem_reset_finish_engine(engine);
+		intel_engine_wakeup(engine);
 	}
 
 out:
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index 11f4e6148557..b5511a054f30 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -818,8 +818,11 @@ static void execlists_cancel_requests(struct intel_engine_cs *engine)
 	/* Mark all executing requests as skipped. */
 	list_for_each_entry(rq, &engine->timeline.requests, link) {
 		GEM_BUG_ON(!rq->global_seqno);
-		if (!i915_request_completed(rq))
-			dma_fence_set_error(&rq->fence, -EIO);
+
+		if (test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &rq->fence.flags))
+			continue;
+
+		dma_fence_set_error(&rq->fence, -EIO);
 	}
 
 	/* Flush the queued requests to the timeline list (for retiring). */
@@ -830,8 +833,8 @@ static void execlists_cancel_requests(struct intel_engine_cs *engine)
 		priolist_for_each_request_consume(rq, rn, p, i) {
 			list_del_init(&rq->sched.link);
 
-			dma_fence_set_error(&rq->fence, -EIO);
 			__i915_request_submit(rq);
+			dma_fence_set_error(&rq->fence, -EIO);
 		}
 
 		rb_erase_cached(&p->node, &execlists->queue);
@@ -839,6 +842,10 @@ static void execlists_cancel_requests(struct intel_engine_cs *engine)
 			kmem_cache_free(engine->i915->priorities, p);
 	}
 
+	intel_write_status_page(engine,
+				I915_GEM_HWS_INDEX,
+				intel_engine_last_submit(engine));
+
 	/* Remaining _unready_ requests will be nop'ed when submitted */
 
 	execlists->queue_priority = INT_MIN;
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index e18a64d41843..6c9a67ae3d26 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -750,9 +750,18 @@ static void cancel_requests(struct intel_engine_cs *engine)
 	/* Mark all submitted requests as skipped. */
 	list_for_each_entry(request, &engine->timeline.requests, link) {
 		GEM_BUG_ON(!request->global_seqno);
-		if (!i915_request_completed(request))
-			dma_fence_set_error(&request->fence, -EIO);
+
+		if (test_bit(DMA_FENCE_FLAG_SIGNALED_BIT,
+			     &request->fence.flags))
+			continue;
+
+		dma_fence_set_error(&request->fence, -EIO);
 	}
+
+	intel_write_status_page(engine,
+				I915_GEM_HWS_INDEX,
+				intel_engine_last_submit(engine));
+
 	/* Remaining _unready_ requests will be nop'ed when submitted */
 
 	spin_unlock_irqrestore(&engine->timeline.lock, flags);
-- 
2.20.0.rc1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 2/6] drm/i915/ringbuffer: Clear semaphore sync registers on ring init

[Chris Wilson]

Ensure that the sync registers are cleared every time we restart the
ring to avoid stale values from creeping in from random neutrinos.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/intel_ringbuffer.c | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index 6c9a67ae3d26..9875ee474b83 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -529,6 +529,13 @@ static int init_ring_common(struct intel_engine_cs *engine)
 
 	intel_engine_reset_breadcrumbs(engine);
 
+	if (HAS_LEGACY_SEMAPHORES(engine->i915)) {
+		I915_WRITE(RING_SYNC_0(engine->mmio_base), 0);
+		I915_WRITE(RING_SYNC_1(engine->mmio_base), 0);
+		if (HAS_VEBOX(dev_priv))
+			I915_WRITE(RING_SYNC_2(engine->mmio_base), 0);
+	}
+
 	/* Enforce ordering by reading HEAD register back */
 	I915_READ_HEAD(engine);
 
-- 
2.20.0.rc1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 3/6] drm/i915: Allocate a common scratch page

[Chris Wilson]

Currently we allocate a scratch page for each engine, but since we only
ever write into it for post-sync operations, it is not exposed to
userspace nor do we care for coherency. As we then do not care about its
contents, we can use one page for all, reducing our allocations and
avoid complications by not assuming per-engine isolation.

For later use, it simplifies engine initialisation (by removing the
allocation that required struct_mutex!) and means that we can always rely
on there being a scratch page.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
---
 drivers/gpu/drm/i915/i915_drv.h         |  7 ++++
 drivers/gpu/drm/i915/i915_gem.c         | 50 ++++++++++++++++++++++++-
 drivers/gpu/drm/i915/i915_gpu_error.c   |  2 +-
 drivers/gpu/drm/i915/intel_engine_cs.c  | 42 ---------------------
 drivers/gpu/drm/i915/intel_lrc.c        | 17 +++------
 drivers/gpu/drm/i915/intel_ringbuffer.c | 33 +++++-----------
 drivers/gpu/drm/i915/intel_ringbuffer.h |  5 ---
 7 files changed, 71 insertions(+), 85 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 43ac6873a2bb..6cd576424ec0 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -1995,6 +1995,8 @@ struct drm_i915_private {
 		struct delayed_work idle_work;
 
 		ktime_t last_init_time;
+
+		struct i915_vma *scratch;
 	} gt;
 
 	/* perform PHY state sanity checks? */
@@ -3723,4 +3725,9 @@ static inline int intel_hws_csb_write_index(struct drm_i915_private *i915)
 		return I915_HWS_CSB_WRITE_INDEX;
 }
 
+static inline u32 i915_scratch_offset(const struct drm_i915_private *i915)
+{
+	return i915_ggtt_offset(i915->gt.scratch);
+}
+
 #endif
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 0eaa8a6f0423..a6f798333e24 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -5497,6 +5497,44 @@ static int __intel_engines_record_defaults(struct drm_i915_private *i915)
 	goto out_ctx;
 }
 
+static int
+i915_gem_init_scratch(struct drm_i915_private *i915, unsigned int size)
+{
+	struct drm_i915_gem_object *obj;
+	struct i915_vma *vma;
+	int ret;
+
+	obj = i915_gem_object_create_stolen(i915, size);
+	if (!obj)
+		obj = i915_gem_object_create_internal(i915, size);
+	if (IS_ERR(obj)) {
+		DRM_ERROR("Failed to allocate scratch page\n");
+		return PTR_ERR(obj);
+	}
+
+	vma = i915_vma_instance(obj, &i915->ggtt.vm, NULL);
+	if (IS_ERR(vma)) {
+		ret = PTR_ERR(vma);
+		goto err_unref;
+	}
+
+	ret = i915_vma_pin(vma, 0, 0, PIN_GLOBAL | PIN_HIGH);
+	if (ret)
+		goto err_unref;
+
+	i915->gt.scratch = vma;
+	return 0;
+
+err_unref:
+	i915_gem_object_put(obj);
+	return ret;
+}
+
+static void i915_gem_fini_scratch(struct drm_i915_private *i915)
+{
+	i915_vma_unpin_and_release(&i915->gt.scratch, 0);
+}
+
 int i915_gem_init(struct drm_i915_private *dev_priv)
 {
 	int ret;
@@ -5543,12 +5581,19 @@ int i915_gem_init(struct drm_i915_private *dev_priv)
 		goto err_unlock;
 	}
 
-	ret = i915_gem_contexts_init(dev_priv);
+	ret = i915_gem_init_scratch(dev_priv,
+				    IS_GEN2(dev_priv) ? SZ_256K : PAGE_SIZE);
 	if (ret) {
 		GEM_BUG_ON(ret == -EIO);
 		goto err_ggtt;
 	}
 
+	ret = i915_gem_contexts_init(dev_priv);
+	if (ret) {
+		GEM_BUG_ON(ret == -EIO);
+		goto err_scratch;
+	}
+
 	ret = intel_engines_init(dev_priv);
 	if (ret) {
 		GEM_BUG_ON(ret == -EIO);
@@ -5621,6 +5666,8 @@ int i915_gem_init(struct drm_i915_private *dev_priv)
 err_context:
 	if (ret != -EIO)
 		i915_gem_contexts_fini(dev_priv);
+err_scratch:
+	i915_gem_fini_scratch(dev_priv);
 err_ggtt:
 err_unlock:
 	intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL);
@@ -5672,6 +5719,7 @@ void i915_gem_fini(struct drm_i915_private *dev_priv)
 	intel_uc_fini(dev_priv);
 	i915_gem_cleanup_engines(dev_priv);
 	i915_gem_contexts_fini(dev_priv);
+	i915_gem_fini_scratch(dev_priv);
 	mutex_unlock(&dev_priv->drm.struct_mutex);
 
 	intel_cleanup_gt_powersave(dev_priv);
diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c
index a6885a59568b..07465123c166 100644
--- a/drivers/gpu/drm/i915/i915_gpu_error.c
+++ b/drivers/gpu/drm/i915/i915_gpu_error.c
@@ -1571,7 +1571,7 @@ static void gem_record_rings(struct i915_gpu_state *error)
 			if (HAS_BROKEN_CS_TLB(i915))
 				ee->wa_batchbuffer =
 					i915_error_object_create(i915,
-								 engine->scratch);
+								 i915->gt.scratch);
 			request_record_user_bo(request, ee);
 
 			ee->ctx =
diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c
index 759c0fd58f8c..2390985384d6 100644
--- a/drivers/gpu/drm/i915/intel_engine_cs.c
+++ b/drivers/gpu/drm/i915/intel_engine_cs.c
@@ -493,46 +493,6 @@ void intel_engine_setup_common(struct intel_engine_cs *engine)
 	intel_engine_init_cmd_parser(engine);
 }
 
-int intel_engine_create_scratch(struct intel_engine_cs *engine,
-				unsigned int size)
-{
-	struct drm_i915_gem_object *obj;
-	struct i915_vma *vma;
-	int ret;
-
-	WARN_ON(engine->scratch);
-
-	obj = i915_gem_object_create_stolen(engine->i915, size);
-	if (!obj)
-		obj = i915_gem_object_create_internal(engine->i915, size);
-	if (IS_ERR(obj)) {
-		DRM_ERROR("Failed to allocate scratch page\n");
-		return PTR_ERR(obj);
-	}
-
-	vma = i915_vma_instance(obj, &engine->i915->ggtt.vm, NULL);
-	if (IS_ERR(vma)) {
-		ret = PTR_ERR(vma);
-		goto err_unref;
-	}
-
-	ret = i915_vma_pin(vma, 0, 0, PIN_GLOBAL | PIN_HIGH);
-	if (ret)
-		goto err_unref;
-
-	engine->scratch = vma;
-	return 0;
-
-err_unref:
-	i915_gem_object_put(obj);
-	return ret;
-}
-
-void intel_engine_cleanup_scratch(struct intel_engine_cs *engine)
-{
-	i915_vma_unpin_and_release(&engine->scratch, 0);
-}
-
 static void cleanup_status_page(struct intel_engine_cs *engine)
 {
 	if (HWS_NEEDS_PHYSICAL(engine->i915)) {
@@ -707,8 +667,6 @@ void intel_engine_cleanup_common(struct intel_engine_cs *engine)
 {
 	struct drm_i915_private *i915 = engine->i915;
 
-	intel_engine_cleanup_scratch(engine);
-
 	cleanup_status_page(engine);
 
 	intel_engine_fini_breadcrumbs(engine);
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index b5511a054f30..de070dca4033 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -1286,9 +1286,10 @@ static int execlists_request_alloc(struct i915_request *request)
 static u32 *
 gen8_emit_flush_coherentl3_wa(struct intel_engine_cs *engine, u32 *batch)
 {
+	/* NB no one else is allowed to scribble over scratch + 256! */
 	*batch++ = MI_STORE_REGISTER_MEM_GEN8 | MI_SRM_LRM_GLOBAL_GTT;
 	*batch++ = i915_mmio_reg_offset(GEN8_L3SQCREG4);
-	*batch++ = i915_ggtt_offset(engine->scratch) + 256;
+	*batch++ = i915_scratch_offset(engine->i915) + 256;
 	*batch++ = 0;
 
 	*batch++ = MI_LOAD_REGISTER_IMM(1);
@@ -1302,7 +1303,7 @@ gen8_emit_flush_coherentl3_wa(struct intel_engine_cs *engine, u32 *batch)
 
 	*batch++ = MI_LOAD_REGISTER_MEM_GEN8 | MI_SRM_LRM_GLOBAL_GTT;
 	*batch++ = i915_mmio_reg_offset(GEN8_L3SQCREG4);
-	*batch++ = i915_ggtt_offset(engine->scratch) + 256;
+	*batch++ = i915_scratch_offset(engine->i915) + 256;
 	*batch++ = 0;
 
 	return batch;
@@ -1339,7 +1340,7 @@ static u32 *gen8_init_indirectctx_bb(struct intel_engine_cs *engine, u32 *batch)
 				       PIPE_CONTROL_GLOBAL_GTT_IVB |
 				       PIPE_CONTROL_CS_STALL |
 				       PIPE_CONTROL_QW_WRITE,
-				       i915_ggtt_offset(engine->scratch) +
+				       i915_scratch_offset(engine->i915) +
 				       2 * CACHELINE_BYTES);
 
 	*batch++ = MI_ARB_ON_OFF | MI_ARB_ENABLE;
@@ -1969,7 +1970,7 @@ static int gen8_emit_flush_render(struct i915_request *request,
 {
 	struct intel_engine_cs *engine = request->engine;
 	u32 scratch_addr =
-		i915_ggtt_offset(engine->scratch) + 2 * CACHELINE_BYTES;
+		i915_scratch_offset(engine->i915) + 2 * CACHELINE_BYTES;
 	bool vf_flush_wa = false, dc_flush_wa = false;
 	u32 *cs, flags = 0;
 	int len;
@@ -2306,10 +2307,6 @@ int logical_render_ring_init(struct intel_engine_cs *engine)
 	if (ret)
 		return ret;
 
-	ret = intel_engine_create_scratch(engine, PAGE_SIZE);
-	if (ret)
-		goto err_cleanup_common;
-
 	ret = intel_init_workaround_bb(engine);
 	if (ret) {
 		/*
@@ -2322,10 +2319,6 @@ int logical_render_ring_init(struct intel_engine_cs *engine)
 	}
 
 	return 0;
-
-err_cleanup_common:
-	intel_engine_cleanup_common(engine);
-	return ret;
 }
 
 int logical_xcs_ring_init(struct intel_engine_cs *engine)
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index 9875ee474b83..1b7141a171b6 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -150,8 +150,7 @@ gen4_render_ring_flush(struct i915_request *rq, u32 mode)
 	 */
 	if (mode & EMIT_INVALIDATE) {
 		*cs++ = GFX_OP_PIPE_CONTROL(4) | PIPE_CONTROL_QW_WRITE;
-		*cs++ = i915_ggtt_offset(rq->engine->scratch) |
-			PIPE_CONTROL_GLOBAL_GTT;
+		*cs++ = i915_scratch_offset(rq->i915) | PIPE_CONTROL_GLOBAL_GTT;
 		*cs++ = 0;
 		*cs++ = 0;
 
@@ -159,8 +158,7 @@ gen4_render_ring_flush(struct i915_request *rq, u32 mode)
 			*cs++ = MI_FLUSH;
 
 		*cs++ = GFX_OP_PIPE_CONTROL(4) | PIPE_CONTROL_QW_WRITE;
-		*cs++ = i915_ggtt_offset(rq->engine->scratch) |
-			PIPE_CONTROL_GLOBAL_GTT;
+		*cs++ = i915_scratch_offset(rq->i915) | PIPE_CONTROL_GLOBAL_GTT;
 		*cs++ = 0;
 		*cs++ = 0;
 	}
@@ -212,8 +210,7 @@ gen4_render_ring_flush(struct i915_request *rq, u32 mode)
 static int
 intel_emit_post_sync_nonzero_flush(struct i915_request *rq)
 {
-	u32 scratch_addr =
-		i915_ggtt_offset(rq->engine->scratch) + 2 * CACHELINE_BYTES;
+	u32 scratch_addr = i915_scratch_offset(rq->i915) + 2 * CACHELINE_BYTES;
 	u32 *cs;
 
 	cs = intel_ring_begin(rq, 6);
@@ -246,8 +243,7 @@ intel_emit_post_sync_nonzero_flush(struct i915_request *rq)
 static int
 gen6_render_ring_flush(struct i915_request *rq, u32 mode)
 {
-	u32 scratch_addr =
-		i915_ggtt_offset(rq->engine->scratch) + 2 * CACHELINE_BYTES;
+	u32 scratch_addr = i915_scratch_offset(rq->i915) + 2 * CACHELINE_BYTES;
 	u32 *cs, flags = 0;
 	int ret;
 
@@ -316,8 +312,7 @@ gen7_render_ring_cs_stall_wa(struct i915_request *rq)
 static int
 gen7_render_ring_flush(struct i915_request *rq, u32 mode)
 {
-	u32 scratch_addr =
-		i915_ggtt_offset(rq->engine->scratch) + 2 * CACHELINE_BYTES;
+	u32 scratch_addr = i915_scratch_offset(rq->i915) + 2 * CACHELINE_BYTES;
 	u32 *cs, flags = 0;
 
 	/*
@@ -1004,7 +999,7 @@ i830_emit_bb_start(struct i915_request *rq,
 		   u64 offset, u32 len,
 		   unsigned int dispatch_flags)
 {
-	u32 *cs, cs_offset = i915_ggtt_offset(rq->engine->scratch);
+	u32 *cs, cs_offset = i915_scratch_offset(rq->i915);
 
 	cs = intel_ring_begin(rq, 6);
 	if (IS_ERR(cs))
@@ -1461,7 +1456,6 @@ static int intel_init_ring_buffer(struct intel_engine_cs *engine)
 {
 	struct i915_timeline *timeline;
 	struct intel_ring *ring;
-	unsigned int size;
 	int err;
 
 	intel_engine_setup_common(engine);
@@ -1486,21 +1480,12 @@ static int intel_init_ring_buffer(struct intel_engine_cs *engine)
 	GEM_BUG_ON(engine->buffer);
 	engine->buffer = ring;
 
-	size = PAGE_SIZE;
-	if (HAS_BROKEN_CS_TLB(engine->i915))
-		size = I830_WA_SIZE;
-	err = intel_engine_create_scratch(engine, size);
-	if (err)
-		goto err_unpin;
-
 	err = intel_engine_init_common(engine);
 	if (err)
-		goto err_scratch;
+		goto err_unpin;
 
 	return 0;
 
-err_scratch:
-	intel_engine_cleanup_scratch(engine);
 err_unpin:
 	intel_ring_unpin(ring);
 err_ring:
@@ -1574,7 +1559,7 @@ static int flush_pd_dir(struct i915_request *rq)
 	/* Stall until the page table load is complete */
 	*cs++ = MI_STORE_REGISTER_MEM | MI_SRM_LRM_GLOBAL_GTT;
 	*cs++ = i915_mmio_reg_offset(RING_PP_DIR_BASE(engine));
-	*cs++ = i915_ggtt_offset(engine->scratch);
+	*cs++ = i915_scratch_offset(rq->i915);
 	*cs++ = MI_NOOP;
 
 	intel_ring_advance(rq, cs);
@@ -1683,7 +1668,7 @@ static inline int mi_set_context(struct i915_request *rq, u32 flags)
 			/* Insert a delay before the next switch! */
 			*cs++ = MI_STORE_REGISTER_MEM | MI_SRM_LRM_GLOBAL_GTT;
 			*cs++ = i915_mmio_reg_offset(last_reg);
-			*cs++ = i915_ggtt_offset(engine->scratch);
+			*cs++ = i915_scratch_offset(rq->i915);
 			*cs++ = MI_NOOP;
 		}
 		*cs++ = MI_ARB_ON_OFF | MI_ARB_ENABLE;
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index 8a2270b209b0..970fb5c05c36 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -451,7 +451,6 @@ struct intel_engine_cs {
 
 	struct intel_hw_status_page status_page;
 	struct i915_ctx_workarounds wa_ctx;
-	struct i915_vma *scratch;
 
 	u32             irq_keep_mask; /* always keep these interrupts */
 	u32		irq_enable_mask; /* bitmask to enable ring interrupt */
@@ -908,10 +907,6 @@ void intel_engine_setup_common(struct intel_engine_cs *engine);
 int intel_engine_init_common(struct intel_engine_cs *engine);
 void intel_engine_cleanup_common(struct intel_engine_cs *engine);
 
-int intel_engine_create_scratch(struct intel_engine_cs *engine,
-				unsigned int size);
-void intel_engine_cleanup_scratch(struct intel_engine_cs *engine);
-
 int intel_init_render_ring_buffer(struct intel_engine_cs *engine);
 int intel_init_bsd_ring_buffer(struct intel_engine_cs *engine);
 int intel_init_blt_ring_buffer(struct intel_engine_cs *engine);
-- 
2.20.0.rc1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 4/6] drm/i915/selftests: Terminate hangcheck sanitycheck forcibly

[Chris Wilson]

If all else fails and we are stuck eternally waiting for the undying
request, abandon all hope.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/selftests/intel_hangcheck.c | 12 +++++++++---
 1 file changed, 9 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/i915/selftests/intel_hangcheck.c b/drivers/gpu/drm/i915/selftests/intel_hangcheck.c
index defe671130ab..a48fbe2557ea 100644
--- a/drivers/gpu/drm/i915/selftests/intel_hangcheck.c
+++ b/drivers/gpu/drm/i915/selftests/intel_hangcheck.c
@@ -308,6 +308,7 @@ static int igt_hang_sanitycheck(void *arg)
 		goto unlock;
 
 	for_each_engine(engine, i915, id) {
+		struct igt_wedge_me w;
 		long timeout;
 
 		if (!intel_engine_can_store_dword(engine))
@@ -328,9 +329,14 @@ static int igt_hang_sanitycheck(void *arg)
 
 		i915_request_add(rq);
 
-		timeout = i915_request_wait(rq,
-					    I915_WAIT_LOCKED,
-					    MAX_SCHEDULE_TIMEOUT);
+		timeout = 0;
+		igt_wedge_on_timeout(&w, i915, HZ / 10 /* 100ms timeout*/)
+			timeout = i915_request_wait(rq,
+						    I915_WAIT_LOCKED,
+						    MAX_SCHEDULE_TIMEOUT);
+		if (i915_terminally_wedged(&i915->gpu_error))
+			timeout = -EIO;
+
 		i915_request_put(rq);
 
 		if (timeout < 0) {
-- 
2.20.0.rc1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 5/6] drm/i915/selftests: Reorder request allocation vs vma pinning

[Chris Wilson]

Impose a restraint that we have all vma pinned for a request prior to
its allocation. This is to simplify request construction, and should
facilitate unravelling the lock interdependencies later.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/selftests/huge_pages.c   |  31 +++--
 .../gpu/drm/i915/selftests/intel_hangcheck.c  | 123 +++++++++---------
 drivers/gpu/drm/i915/selftests/intel_lrc.c    |  86 ++++++------
 3 files changed, 119 insertions(+), 121 deletions(-)

diff --git a/drivers/gpu/drm/i915/selftests/huge_pages.c b/drivers/gpu/drm/i915/selftests/huge_pages.c
index 26c065c8d2c0..a0c7cbc212ba 100644
--- a/drivers/gpu/drm/i915/selftests/huge_pages.c
+++ b/drivers/gpu/drm/i915/selftests/huge_pages.c
@@ -972,7 +972,6 @@ static int gpu_write(struct i915_vma *vma,
 {
 	struct i915_request *rq;
 	struct i915_vma *batch;
-	int flags = 0;
 	int err;
 
 	GEM_BUG_ON(!intel_engine_can_store_dword(engine));
@@ -981,14 +980,14 @@ static int gpu_write(struct i915_vma *vma,
 	if (err)
 		return err;
 
-	rq = i915_request_alloc(engine, ctx);
-	if (IS_ERR(rq))
-		return PTR_ERR(rq);
-
 	batch = gpu_write_dw(vma, dword * sizeof(u32), value);
-	if (IS_ERR(batch)) {
-		err = PTR_ERR(batch);
-		goto err_request;
+	if (IS_ERR(batch))
+		return PTR_ERR(batch);
+
+	rq = i915_request_alloc(engine, ctx);
+	if (IS_ERR(rq)) {
+		err = PTR_ERR(rq);
+		goto err_batch;
 	}
 
 	err = i915_vma_move_to_active(batch, rq, 0);
@@ -996,21 +995,21 @@ static int gpu_write(struct i915_vma *vma,
 		goto err_request;
 
 	i915_gem_object_set_active_reference(batch->obj);
-	i915_vma_unpin(batch);
-	i915_vma_close(batch);
 
-	err = engine->emit_bb_start(rq,
-				    batch->node.start, batch->node.size,
-				    flags);
+	err = i915_vma_move_to_active(vma, rq, EXEC_OBJECT_WRITE);
 	if (err)
 		goto err_request;
 
-	err = i915_vma_move_to_active(vma, rq, EXEC_OBJECT_WRITE);
+	err = engine->emit_bb_start(rq,
+				    batch->node.start, batch->node.size,
+				    0);
+err_request:
 	if (err)
 		i915_request_skip(rq, err);
-
-err_request:
 	i915_request_add(rq);
+err_batch:
+	i915_vma_unpin(batch);
+	i915_vma_close(batch);
 
 	return err;
 }
diff --git a/drivers/gpu/drm/i915/selftests/intel_hangcheck.c b/drivers/gpu/drm/i915/selftests/intel_hangcheck.c
index a48fbe2557ea..b767fab9ce1f 100644
--- a/drivers/gpu/drm/i915/selftests/intel_hangcheck.c
+++ b/drivers/gpu/drm/i915/selftests/intel_hangcheck.c
@@ -102,52 +102,87 @@ static u64 hws_address(const struct i915_vma *hws,
 	return hws->node.start + offset_in_page(sizeof(u32)*rq->fence.context);
 }
 
-static int emit_recurse_batch(struct hang *h,
-			      struct i915_request *rq)
+static int move_to_active(struct i915_vma *vma,
+			  struct i915_request *rq,
+			  unsigned int flags)
+{
+	int err;
+
+	err = i915_vma_move_to_active(vma, rq, 0);
+	if (err)
+		return err;
+
+	if (!i915_gem_object_has_active_reference(vma->obj)) {
+		i915_gem_object_get(vma->obj);
+		i915_gem_object_set_active_reference(vma->obj);
+	}
+
+	return 0;
+}
+
+static struct i915_request *
+hang_create_request(struct hang *h, struct intel_engine_cs *engine)
 {
 	struct drm_i915_private *i915 = h->i915;
 	struct i915_address_space *vm =
-		rq->gem_context->ppgtt ?
-		&rq->gem_context->ppgtt->vm :
-		&i915->ggtt.vm;
+		h->ctx->ppgtt ? &h->ctx->ppgtt->vm : &i915->ggtt.vm;
+	struct i915_request *rq = NULL;
 	struct i915_vma *hws, *vma;
 	unsigned int flags;
 	u32 *batch;
 	int err;
 
+	if (i915_gem_object_is_active(h->obj)) {
+		struct drm_i915_gem_object *obj;
+		void *vaddr;
+
+		obj = i915_gem_object_create_internal(h->i915, PAGE_SIZE);
+		if (IS_ERR(obj))
+			return ERR_CAST(obj);
+
+		vaddr = i915_gem_object_pin_map(obj,
+						i915_coherent_map_type(h->i915));
+		if (IS_ERR(vaddr)) {
+			i915_gem_object_put(obj);
+			return ERR_CAST(vaddr);
+		}
+
+		i915_gem_object_unpin_map(h->obj);
+		i915_gem_object_put(h->obj);
+
+		h->obj = obj;
+		h->batch = vaddr;
+	}
+
 	vma = i915_vma_instance(h->obj, vm, NULL);
 	if (IS_ERR(vma))
-		return PTR_ERR(vma);
+		return ERR_CAST(vma);
 
 	hws = i915_vma_instance(h->hws, vm, NULL);
 	if (IS_ERR(hws))
-		return PTR_ERR(hws);
+		return ERR_CAST(hws);
 
 	err = i915_vma_pin(vma, 0, 0, PIN_USER);
 	if (err)
-		return err;
+		return ERR_PTR(err);
 
 	err = i915_vma_pin(hws, 0, 0, PIN_USER);
 	if (err)
 		goto unpin_vma;
 
-	err = i915_vma_move_to_active(vma, rq, 0);
-	if (err)
+	rq = i915_request_alloc(engine, h->ctx);
+	if (IS_ERR(rq)) {
+		err = PTR_ERR(rq);
 		goto unpin_hws;
-
-	if (!i915_gem_object_has_active_reference(vma->obj)) {
-		i915_gem_object_get(vma->obj);
-		i915_gem_object_set_active_reference(vma->obj);
 	}
 
-	err = i915_vma_move_to_active(hws, rq, 0);
+	err = move_to_active(vma, rq, 0);
 	if (err)
-		goto unpin_hws;
+		goto cancel_rq;
 
-	if (!i915_gem_object_has_active_reference(hws->obj)) {
-		i915_gem_object_get(hws->obj);
-		i915_gem_object_set_active_reference(hws->obj);
-	}
+	err = move_to_active(hws, rq, 0);
+	if (err)
+		goto cancel_rq;
 
 	batch = h->batch;
 	if (INTEL_GEN(i915) >= 8) {
@@ -212,52 +247,16 @@ static int emit_recurse_batch(struct hang *h,
 
 	err = rq->engine->emit_bb_start(rq, vma->node.start, PAGE_SIZE, flags);
 
+cancel_rq:
+	if (err) {
+		i915_request_skip(rq, err);
+		i915_request_add(rq);
+	}
 unpin_hws:
 	i915_vma_unpin(hws);
 unpin_vma:
 	i915_vma_unpin(vma);
-	return err;
-}
-
-static struct i915_request *
-hang_create_request(struct hang *h, struct intel_engine_cs *engine)
-{
-	struct i915_request *rq;
-	int err;
-
-	if (i915_gem_object_is_active(h->obj)) {
-		struct drm_i915_gem_object *obj;
-		void *vaddr;
-
-		obj = i915_gem_object_create_internal(h->i915, PAGE_SIZE);
-		if (IS_ERR(obj))
-			return ERR_CAST(obj);
-
-		vaddr = i915_gem_object_pin_map(obj,
-						i915_coherent_map_type(h->i915));
-		if (IS_ERR(vaddr)) {
-			i915_gem_object_put(obj);
-			return ERR_CAST(vaddr);
-		}
-
-		i915_gem_object_unpin_map(h->obj);
-		i915_gem_object_put(h->obj);
-
-		h->obj = obj;
-		h->batch = vaddr;
-	}
-
-	rq = i915_request_alloc(engine, h->ctx);
-	if (IS_ERR(rq))
-		return rq;
-
-	err = emit_recurse_batch(h, rq);
-	if (err) {
-		i915_request_add(rq);
-		return ERR_PTR(err);
-	}
-
-	return rq;
+	return err ? ERR_PTR(err) : rq;
 }
 
 static u32 hws_seqno(const struct hang *h, const struct i915_request *rq)
diff --git a/drivers/gpu/drm/i915/selftests/intel_lrc.c b/drivers/gpu/drm/i915/selftests/intel_lrc.c
index 94fc0e5c8766..80f881423c41 100644
--- a/drivers/gpu/drm/i915/selftests/intel_lrc.c
+++ b/drivers/gpu/drm/i915/selftests/intel_lrc.c
@@ -80,48 +80,65 @@ static u64 hws_address(const struct i915_vma *hws,
 	return hws->node.start + seqno_offset(rq->fence.context);
 }
 
-static int emit_recurse_batch(struct spinner *spin,
-			      struct i915_request *rq,
-			      u32 arbitration_command)
+static int move_to_active(struct i915_vma *vma,
+			  struct i915_request *rq,
+			  unsigned int flags)
 {
-	struct i915_address_space *vm = &rq->gem_context->ppgtt->vm;
+	int err;
+
+	err = i915_vma_move_to_active(vma, rq, 0);
+	if (err)
+		return err;
+
+	if (!i915_gem_object_has_active_reference(vma->obj)) {
+		i915_gem_object_get(vma->obj);
+		i915_gem_object_set_active_reference(vma->obj);
+	}
+
+	return 0;
+}
+
+static struct i915_request *
+spinner_create_request(struct spinner *spin,
+		       struct i915_gem_context *ctx,
+		       struct intel_engine_cs *engine,
+		       u32 arbitration_command)
+{
+	struct i915_address_space *vm = &ctx->ppgtt->vm;
+	struct i915_request *rq = NULL;
 	struct i915_vma *hws, *vma;
 	u32 *batch;
 	int err;
 
 	vma = i915_vma_instance(spin->obj, vm, NULL);
 	if (IS_ERR(vma))
-		return PTR_ERR(vma);
+		return ERR_CAST(vma);
 
 	hws = i915_vma_instance(spin->hws, vm, NULL);
 	if (IS_ERR(hws))
-		return PTR_ERR(hws);
+		return ERR_CAST(hws);
 
 	err = i915_vma_pin(vma, 0, 0, PIN_USER);
 	if (err)
-		return err;
+		return ERR_PTR(err);
 
 	err = i915_vma_pin(hws, 0, 0, PIN_USER);
 	if (err)
 		goto unpin_vma;
 
-	err = i915_vma_move_to_active(vma, rq, 0);
-	if (err)
+	rq = i915_request_alloc(engine, ctx);
+	if (IS_ERR(rq)) {
+		err = PTR_ERR(rq);
 		goto unpin_hws;
-
-	if (!i915_gem_object_has_active_reference(vma->obj)) {
-		i915_gem_object_get(vma->obj);
-		i915_gem_object_set_active_reference(vma->obj);
 	}
 
-	err = i915_vma_move_to_active(hws, rq, 0);
+	err = move_to_active(vma, rq, 0);
 	if (err)
-		goto unpin_hws;
+		goto cancel_rq;
 
-	if (!i915_gem_object_has_active_reference(hws->obj)) {
-		i915_gem_object_get(hws->obj);
-		i915_gem_object_set_active_reference(hws->obj);
-	}
+	err = move_to_active(hws, rq, 0);
+	if (err)
+		goto cancel_rq;
 
 	batch = spin->batch;
 
@@ -139,35 +156,18 @@ static int emit_recurse_batch(struct spinner *spin,
 
 	i915_gem_chipset_flush(spin->i915);
 
-	err = rq->engine->emit_bb_start(rq, vma->node.start, PAGE_SIZE, 0);
+	err = engine->emit_bb_start(rq, vma->node.start, PAGE_SIZE, 0);
 
+cancel_rq:
+	if (err) {
+		i915_request_skip(rq, err);
+		i915_request_add(rq);
+	}
 unpin_hws:
 	i915_vma_unpin(hws);
 unpin_vma:
 	i915_vma_unpin(vma);
-	return err;
-}
-
-static struct i915_request *
-spinner_create_request(struct spinner *spin,
-		       struct i915_gem_context *ctx,
-		       struct intel_engine_cs *engine,
-		       u32 arbitration_command)
-{
-	struct i915_request *rq;
-	int err;
-
-	rq = i915_request_alloc(engine, ctx);
-	if (IS_ERR(rq))
-		return rq;
-
-	err = emit_recurse_batch(spin, rq, arbitration_command);
-	if (err) {
-		i915_request_add(rq);
-		return ERR_PTR(err);
-	}
-
-	return rq;
+	return err ? ERR_PTR(err) : rq;
 }
 
 static u32 hws_seqno(const struct spinner *spin, const struct i915_request *rq)
-- 
2.20.0.rc1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH 6/6] drm/i915: Pipeline PDP updates for Braswell

[Chris Wilson]

Currently we face a severe problem on Braswell that manifests as invalid
ppGTT accesses. The code tries to maintain the PDP (page directory
pointers) inside the context in two ways, direct write into the context
and a pipelined LRI update. The direct write into the context is
fundamentally racy as it is unserialised with any access (read or write)
the GPU is doing. By asserting that Braswell is not used with vGPU
(currently an unsupported platform) we can eliminate the dangerous
direct write into the context image and solely use the pipelined update.

However, the LRI of the PDP fouls up the GPU, causing it to freeze and
take out the machine with "forcewake ack timeouts". This seems possible
to workaround by preventing the GPU from sleeping (via means of
disabling the power-state management interface, i.e. forcing each ring
to remain awake) around the update.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=108656
References: https://bugs.freedesktop.org/show_bug.cgi?id=108714
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
---
 drivers/gpu/drm/i915/i915_gem_gtt.c     |   2 -
 drivers/gpu/drm/i915/i915_request.c     |   5 -
 drivers/gpu/drm/i915/intel_lrc.c        | 137 +++++++++++-------------
 drivers/gpu/drm/i915/intel_ringbuffer.c |   5 +-
 4 files changed, 68 insertions(+), 81 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index add1fe7aeb93..62bde517d383 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -1423,8 +1423,6 @@ static int gen8_ppgtt_alloc_pdp(struct i915_address_space *vm,
 			gen8_initialize_pd(vm, pd);
 			gen8_ppgtt_set_pdpe(vm, pdp, pd, pdpe);
 			GEM_BUG_ON(pdp->used_pdpes > i915_pdpes_per_pdp(vm));
-
-			mark_tlbs_dirty(i915_vm_to_ppgtt(vm));
 		}
 
 		ret = gen8_ppgtt_alloc_pd(vm, pd, start, length);
diff --git a/drivers/gpu/drm/i915/i915_request.c b/drivers/gpu/drm/i915/i915_request.c
index ca95ab2f4cfa..8ab8e8e6a086 100644
--- a/drivers/gpu/drm/i915/i915_request.c
+++ b/drivers/gpu/drm/i915/i915_request.c
@@ -719,11 +719,6 @@ i915_request_alloc(struct intel_engine_cs *engine, struct i915_gem_context *ctx)
 	 */
 	rq->head = rq->ring->emit;
 
-	/* Unconditionally invalidate GPU caches and TLBs. */
-	ret = engine->emit_flush(rq, EMIT_INVALIDATE);
-	if (ret)
-		goto err_unwind;
-
 	ret = engine->request_alloc(rq);
 	if (ret)
 		goto err_unwind;
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index de070dca4033..0a690c557113 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -363,31 +363,12 @@ execlists_context_schedule_out(struct i915_request *rq, unsigned long status)
 	trace_i915_request_out(rq);
 }
 
-static void
-execlists_update_context_pdps(struct i915_hw_ppgtt *ppgtt, u32 *reg_state)
-{
-	ASSIGN_CTX_PDP(ppgtt, reg_state, 3);
-	ASSIGN_CTX_PDP(ppgtt, reg_state, 2);
-	ASSIGN_CTX_PDP(ppgtt, reg_state, 1);
-	ASSIGN_CTX_PDP(ppgtt, reg_state, 0);
-}
-
 static u64 execlists_update_context(struct i915_request *rq)
 {
-	struct i915_hw_ppgtt *ppgtt = rq->gem_context->ppgtt;
 	struct intel_context *ce = rq->hw_context;
-	u32 *reg_state = ce->lrc_reg_state;
-
-	reg_state[CTX_RING_TAIL+1] = intel_ring_set_tail(rq->ring, rq->tail);
 
-	/*
-	 * True 32b PPGTT with dynamic page allocation: update PDP
-	 * registers and point the unallocated PDPs to scratch page.
-	 * PML4 is allocated during ppgtt init, so this is not needed
-	 * in 48-bit mode.
-	 */
-	if (!i915_vm_is_48bit(&ppgtt->vm))
-		execlists_update_context_pdps(ppgtt, reg_state);
+	ce->lrc_reg_state[CTX_RING_TAIL + 1] =
+		intel_ring_set_tail(rq->ring, rq->tail);
 
 	/*
 	 * Make sure the context image is complete before we submit it to HW.
@@ -1240,29 +1221,80 @@ execlists_context_pin(struct intel_engine_cs *engine,
 	return __execlists_context_pin(engine, ctx, ce);
 }
 
+static int emit_pdps(struct i915_request *rq)
+{
+	const struct intel_engine_cs * const engine = rq->engine;
+	struct i915_hw_ppgtt * const ppgtt = rq->gem_context->ppgtt;
+	int err, i;
+	u32 *cs;
+
+	err = engine->emit_flush(rq, EMIT_INVALIDATE);
+	if (err)
+		return err;
+
+	cs = intel_ring_begin(rq, 4 * GEN8_3LVL_PDPES + 2);
+	if (IS_ERR(cs))
+		return PTR_ERR(cs);
+
+	/*
+	 * Force the GPU (not just the local engine/powerwell!) to remain awake,
+	 * or else we may kill the machine with "timed out waiting for
+	 * forcewake ack request".
+	 */
+
+	*cs++ = MI_LOAD_REGISTER_IMM(2 * GEN8_3LVL_PDPES);
+	for (i = GEN8_3LVL_PDPES; i--; ) {
+		const dma_addr_t pd_daddr = i915_page_dir_dma_addr(ppgtt, i);
+
+		*cs++ = i915_mmio_reg_offset(GEN8_RING_PDP_UDW(engine, i));
+		*cs++ = upper_32_bits(pd_daddr);
+		*cs++ = i915_mmio_reg_offset(GEN8_RING_PDP_LDW(engine, i));
+		*cs++ = lower_32_bits(pd_daddr);
+	}
+	*cs++ = MI_NOOP;
+
+	intel_ring_advance(rq, cs);
+
+	err = engine->emit_flush(rq, EMIT_INVALIDATE);
+	if (err)
+		return err;
+
+	return 0;
+}
+
 static int execlists_request_alloc(struct i915_request *request)
 {
 	int ret;
 
 	GEM_BUG_ON(!request->hw_context->pin_count);
 
-	/* Flush enough space to reduce the likelihood of waiting after
+	/*
+	 * Flush enough space to reduce the likelihood of waiting after
 	 * we start building the request - in which case we will just
 	 * have to repeat work.
 	 */
 	request->reserved_space += EXECLISTS_REQUEST_SIZE;
 
-	ret = intel_ring_wait_for_space(request->ring, request->reserved_space);
-	if (ret)
-		return ret;
-
-	/* Note that after this point, we have committed to using
+	/*
+	 * Note that after this point, we have committed to using
 	 * this request as it is being used to both track the
 	 * state of engine initialisation and liveness of the
 	 * golden renderstate above. Think twice before you try
 	 * to cancel/unwind this request now.
 	 */
 
+	/* Unconditionally invalidate GPU caches and TLBs. */
+	if (i915_vm_is_48bit(&request->gem_context->ppgtt->vm)) {
+		ret = request->engine->emit_flush(request, EMIT_INVALIDATE);
+		if (ret)
+			return ret;
+	} else {
+		GEM_BUG_ON(intel_vgpu_active(request->i915));
+		ret = emit_pdps(request);
+		if (ret)
+			return ret;
+	}
+
 	request->reserved_space -= EXECLISTS_REQUEST_SIZE;
 	return 0;
 }
@@ -1832,56 +1864,11 @@ static void execlists_reset_finish(struct intel_engine_cs *engine)
 		  atomic_read(&execlists->tasklet.count));
 }
 
-static int intel_logical_ring_emit_pdps(struct i915_request *rq)
-{
-	struct i915_hw_ppgtt *ppgtt = rq->gem_context->ppgtt;
-	struct intel_engine_cs *engine = rq->engine;
-	const int num_lri_cmds = GEN8_3LVL_PDPES * 2;
-	u32 *cs;
-	int i;
-
-	cs = intel_ring_begin(rq, num_lri_cmds * 2 + 2);
-	if (IS_ERR(cs))
-		return PTR_ERR(cs);
-
-	*cs++ = MI_LOAD_REGISTER_IMM(num_lri_cmds);
-	for (i = GEN8_3LVL_PDPES - 1; i >= 0; i--) {
-		const dma_addr_t pd_daddr = i915_page_dir_dma_addr(ppgtt, i);
-
-		*cs++ = i915_mmio_reg_offset(GEN8_RING_PDP_UDW(engine, i));
-		*cs++ = upper_32_bits(pd_daddr);
-		*cs++ = i915_mmio_reg_offset(GEN8_RING_PDP_LDW(engine, i));
-		*cs++ = lower_32_bits(pd_daddr);
-	}
-
-	*cs++ = MI_NOOP;
-	intel_ring_advance(rq, cs);
-
-	return 0;
-}
-
 static int gen8_emit_bb_start(struct i915_request *rq,
 			      u64 offset, u32 len,
 			      const unsigned int flags)
 {
 	u32 *cs;
-	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 ((intel_engine_flag(rq->engine) & rq->gem_context->ppgtt->pd_dirty_rings) &&
-	    !i915_vm_is_48bit(&rq->gem_context->ppgtt->vm) &&
-	    !intel_vgpu_active(rq->i915)) {
-		ret = intel_logical_ring_emit_pdps(rq);
-		if (ret)
-			return ret;
-
-		rq->gem_context->ppgtt->pd_dirty_rings &= ~intel_engine_flag(rq->engine);
-	}
 
 	cs = intel_ring_begin(rq, 6);
 	if (IS_ERR(cs))
@@ -1914,6 +1901,7 @@ static int gen8_emit_bb_start(struct i915_request *rq,
 
 	*cs++ = MI_ARB_ON_OFF | MI_ARB_DISABLE;
 	*cs++ = MI_NOOP;
+
 	intel_ring_advance(rq, cs);
 
 	return 0;
@@ -2544,6 +2532,11 @@ static void execlists_init_reg_state(u32 *regs,
 		 * other PDP Descriptors are ignored.
 		 */
 		ASSIGN_CTX_PML4(ctx->ppgtt, regs);
+	} else {
+		ASSIGN_CTX_PDP(ctx->ppgtt, regs, 3);
+		ASSIGN_CTX_PDP(ctx->ppgtt, regs, 2);
+		ASSIGN_CTX_PDP(ctx->ppgtt, regs, 1);
+		ASSIGN_CTX_PDP(ctx->ppgtt, regs, 0);
 	}
 
 	if (rcs) {
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index 1b7141a171b6..d21abeb3dc6a 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -1826,11 +1826,12 @@ static int ring_request_alloc(struct i915_request *request)
 	 */
 	request->reserved_space += LEGACY_REQUEST_SIZE;
 
-	ret = intel_ring_wait_for_space(request->ring, request->reserved_space);
+	ret = switch_context(request);
 	if (ret)
 		return ret;
 
-	ret = switch_context(request);
+	/* Unconditionally invalidate GPU caches and TLBs. */
+	ret = request->engine->emit_flush(request, EMIT_INVALIDATE);
 	if (ret)
 		return ret;
 
-- 
2.20.0.rc1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

✗ Fi.CI.CHECKPATCH: warning for series starting with [1/6] drm/i915: Complete the fences as they are cancelled due to wedging

[Patchwork]

== Series Details ==

Series: series starting with [1/6] drm/i915: Complete the fences as they are cancelled due to wedging
URL   : https://patchwork.freedesktop.org/series/53308/
State : warning

== Summary ==

$ dim checkpatch origin/drm-tip
1826a0f0e99f drm/i915: Complete the fences as they are cancelled due to wedging
-:13: WARNING:COMMIT_LOG_LONG_LINE: Possible unwrapped commit description (prefer a maximum 75 chars per line)
#13: 
<1>[  354.473346] BUG: unable to handle kernel NULL pointer dereference at 0000000000000250

total: 0 errors, 1 warnings, 0 checks, 143 lines checked
7457cba9434a drm/i915/ringbuffer: Clear semaphore sync registers on ring init
e0e3ac8bdd61 drm/i915: Allocate a common scratch page
e5832e383cc9 drm/i915/selftests: Terminate hangcheck sanitycheck forcibly
725e7b22dc50 drm/i915/selftests: Reorder request allocation vs vma pinning
c19b34b6c7fe drm/i915: Pipeline PDP updates for Braswell

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

✗ Fi.CI.SPARSE: warning for series starting with [1/6] drm/i915: Complete the fences as they are cancelled due to wedging

[Patchwork]

== Series Details ==

Series: series starting with [1/6] drm/i915: Complete the fences as they are cancelled due to wedging
URL   : https://patchwork.freedesktop.org/series/53308/
State : warning

== Summary ==

$ dim sparse origin/drm-tip
Sparse version: v0.5.2
Commit: drm/i915: Complete the fences as they are cancelled due to wedging
Okay!

Commit: drm/i915/ringbuffer: Clear semaphore sync registers on ring init
Okay!

Commit: drm/i915: Allocate a common scratch page
-drivers/gpu/drm/i915/selftests/../i915_drv.h:3570:16: warning: expression using sizeof(void)
+drivers/gpu/drm/i915/selftests/../i915_drv.h:3572:16: warning: expression using sizeof(void)

Commit: drm/i915/selftests: Terminate hangcheck sanitycheck forcibly
Okay!

Commit: drm/i915/selftests: Reorder request allocation vs vma pinning
Okay!

Commit: drm/i915: Pipeline PDP updates for Braswell
Okay!

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

✓ Fi.CI.BAT: success for series starting with [1/6] drm/i915: Complete the fences as they are cancelled due to wedging

[Patchwork]

== Series Details ==

Series: series starting with [1/6] drm/i915: Complete the fences as they are cancelled due to wedging
URL   : https://patchwork.freedesktop.org/series/53308/
State : success

== Summary ==

CI Bug Log - changes from CI_DRM_5230 -> Patchwork_10976
====================================================

Summary
-------

  **SUCCESS**

  No regressions found.

  External URL: https://patchwork.freedesktop.org/api/1.0/series/53308/revisions/1/mbox/

Possible new issues
-------------------

  Here are the unknown changes that may have been introduced in Patchwork_10976:

### IGT changes ###

#### Possible regressions ####

  * igt@kms_flip@basic-flip-vs-dpms:
    - {fi-icl-u3}:        PASS -> DMESG-WARN

  * {igt@runner@aborted}:
    - {fi-icl-u3}:        NOTRUN -> FAIL

  
Known issues
------------

  Here are the changes found in Patchwork_10976 that come from known issues:

### IGT changes ###

#### Issues hit ####

  * igt@gem_exec_suspend@basic-s4-devices:
    - fi-ivb-3520m:       PASS -> FAIL [fdo#108880]

  * {igt@runner@aborted}:
    - {fi-icl-y}:         NOTRUN -> FAIL [fdo#108070]

  
#### Possible fixes ####

  * igt@gem_ctx_create@basic-files:
    - fi-kbl-7560u:       INCOMPLETE [fdo#103665] -> PASS

  * igt@kms_pipe_crc_basic@suspend-read-crc-pipe-b:
    - fi-apl-guc:         DMESG-WARN [fdo#108566] -> PASS

  
  {name}: This element is suppressed. This means it is ignored when computing
          the status of the difference (SUCCESS, WARNING, or FAILURE).

  [fdo#103665]: https://bugs.freedesktop.org/show_bug.cgi?id=103665
  [fdo#108070]: https://bugs.freedesktop.org/show_bug.cgi?id=108070
  [fdo#108566]: https://bugs.freedesktop.org/show_bug.cgi?id=108566
  [fdo#108880]: https://bugs.freedesktop.org/show_bug.cgi?id=108880


Participating hosts (50 -> 45)
------------------------------

  Additional (1): fi-icl-y 
  Missing    (6): fi-kbl-soraka fi-ilk-m540 fi-hsw-4200u fi-byt-squawks fi-bsw-cyan fi-ctg-p8600 


Build changes
-------------

    * Linux: CI_DRM_5230 -> Patchwork_10976

  CI_DRM_5230: b0a2de64f8969163f6e01071d5e05748f18a8bab @ git://anongit.freedesktop.org/gfx-ci/linux
  IGT_4736: 285ebfb3b7adc56586031afa5150c4e5ad40c229 @ git://anongit.freedesktop.org/xorg/app/intel-gpu-tools
  Patchwork_10976: c19b34b6c7fe784029535a8743bd098eb06cdadd @ git://anongit.freedesktop.org/gfx-ci/linux


== Linux commits ==

c19b34b6c7fe drm/i915: Pipeline PDP updates for Braswell
725e7b22dc50 drm/i915/selftests: Reorder request allocation vs vma pinning
e5832e383cc9 drm/i915/selftests: Terminate hangcheck sanitycheck forcibly
e0e3ac8bdd61 drm/i915: Allocate a common scratch page
7457cba9434a drm/i915/ringbuffer: Clear semaphore sync registers on ring init
1826a0f0e99f drm/i915: Complete the fences as they are cancelled due to wedging

== Logs ==

For more details see: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_10976/
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

✗ Fi.CI.IGT: failure for series starting with [1/6] drm/i915: Complete the fences as they are cancelled due to wedging

[Patchwork]

== Series Details ==

Series: series starting with [1/6] drm/i915: Complete the fences as they are cancelled due to wedging
URL   : https://patchwork.freedesktop.org/series/53308/
State : failure

== Summary ==

CI Bug Log - changes from CI_DRM_5230_full -> Patchwork_10976_full
====================================================

Summary
-------

  **FAILURE**

  Serious unknown changes coming with Patchwork_10976_full absolutely need to be
  verified manually.
  
  If you think the reported changes have nothing to do with the changes
  introduced in Patchwork_10976_full, please notify your bug team to allow them
  to document this new failure mode, which will reduce false positives in CI.

  

Possible new issues
-------------------

  Here are the unknown changes that may have been introduced in Patchwork_10976_full:

### IGT changes ###

#### Possible regressions ####

  * igt@gem_eio@in-flight-internal-10ms:
    - shard-apl:          PASS -> DMESG-WARN +1
    - shard-snb:          PASS -> DMESG-WARN +1

  * igt@gem_eio@in-flight-internal-1us:
    - shard-skl:          PASS -> DMESG-WARN +1

  * igt@gem_eio@in-flight-internal-immediate:
    - shard-glk:          PASS -> DMESG-WARN +1
    - {shard-iclb}:       PASS -> DMESG-WARN +1
    - shard-skl:          NOTRUN -> DMESG-WARN
    - shard-kbl:          PASS -> DMESG-WARN +2

  
#### Warnings ####

  * igt@pm_rc6_residency@rc6-accuracy:
    - shard-snb:          PASS -> SKIP

  
Known issues
------------

  Here are the changes found in Patchwork_10976_full that come from known issues:

### IGT changes ###

#### Issues hit ####

  * igt@drm_import_export@import-close-race-flink:
    - shard-skl:          PASS -> TIMEOUT [fdo#108667]

  * igt@gem_softpin@noreloc-s3:
    - {shard-iclb}:       PASS -> INCOMPLETE [fdo#107713]

  * igt@kms_busy@extended-modeset-hang-newfb-render-a:
    - {shard-iclb}:       NOTRUN -> DMESG-WARN [fdo#107956]

  * igt@kms_busy@extended-pageflip-modeset-hang-oldfb-render-c:
    - {shard-iclb}:       PASS -> DMESG-WARN [fdo#107956]

  * igt@kms_ccs@pipe-a-crc-sprite-planes-basic:
    - shard-glk:          PASS -> FAIL [fdo#108145] +1

  * igt@kms_cursor_crc@cursor-64x21-random:
    - shard-glk:          PASS -> FAIL [fdo#103232]

  * igt@kms_cursor_crc@cursor-64x64-suspend:
    - {shard-iclb}:       NOTRUN -> FAIL [fdo#103232] +2

  * igt@kms_draw_crc@draw-method-xrgb2101010-mmap-cpu-untiled:
    - shard-skl:          PASS -> FAIL [fdo#103184]

  * igt@kms_fbcon_fbt@psr:
    - shard-skl:          NOTRUN -> FAIL [fdo#107882]

  * igt@kms_flip@2x-flip-vs-expired-vblank:
    - shard-glk:          PASS -> FAIL [fdo#105363]

  * igt@kms_frontbuffer_tracking@fbc-1p-primscrn-cur-indfb-draw-mmap-cpu:
    - shard-glk:          PASS -> FAIL [fdo#103167] +1

  * igt@kms_frontbuffer_tracking@fbcpsr-1p-primscrn-spr-indfb-draw-pwrite:
    - {shard-iclb}:       PASS -> FAIL [fdo#103167] +2

  * igt@kms_plane_alpha_blend@pipe-a-coverage-7efc:
    - shard-skl:          NOTRUN -> FAIL [fdo#107815] / [fdo#108145]

  * igt@kms_plane_alpha_blend@pipe-b-alpha-transparant-fb:
    - shard-kbl:          NOTRUN -> FAIL [fdo#108145]

  * igt@kms_plane_multiple@atomic-pipe-a-tiling-x:
    - {shard-iclb}:       PASS -> FAIL [fdo#103166]

  * igt@kms_plane_multiple@atomic-pipe-b-tiling-y:
    - shard-apl:          PASS -> FAIL [fdo#103166] +2

  * igt@kms_plane_multiple@atomic-pipe-c-tiling-y:
    - shard-glk:          PASS -> FAIL [fdo#103166] +1

  * igt@kms_plane_scaling@pipe-c-scaler-with-rotation:
    - {shard-iclb}:       NOTRUN -> DMESG-WARN [fdo#107724]

  * igt@kms_psr@no_drrs:
    - {shard-iclb}:       PASS -> FAIL [fdo#108341]

  * igt@kms_setmode@basic:
    - {shard-iclb}:       NOTRUN -> FAIL [fdo#99912]

  * igt@pm_backlight@fade_with_suspend:
    - {shard-iclb}:       NOTRUN -> FAIL [fdo#107847]

  * igt@pm_rpm@modeset-lpsp:
    - {shard-iclb}:       PASS -> INCOMPLETE [fdo#108840]

  * igt@pm_rpm@pm-caching:
    - shard-skl:          PASS -> INCOMPLETE [fdo#107807]

  
#### Possible fixes ####

  * igt@gem_ppgtt@blt-vs-render-ctx0:
    - shard-skl:          TIMEOUT [fdo#108039] -> PASS

  * igt@gem_ppgtt@blt-vs-render-ctxn:
    - shard-kbl:          INCOMPLETE [fdo#103665] / [fdo#106023] / [fdo#106887] -> PASS

  * igt@kms_atomic_transition@plane-all-transition-fencing:
    - {shard-iclb}:       DMESG-WARN [fdo#107724] -> PASS +2

  * igt@kms_busy@extended-modeset-hang-newfb-with-reset-render-c:
    - shard-hsw:          DMESG-WARN [fdo#107956] -> PASS

  * igt@kms_chv_cursor_fail@pipe-a-128x128-bottom-edge:
    - shard-skl:          FAIL [fdo#104671] -> PASS

  * igt@kms_cursor_crc@cursor-128x128-offscreen:
    - shard-skl:          FAIL [fdo#103232] -> PASS +1

  * igt@kms_cursor_crc@cursor-64x21-sliding:
    - shard-apl:          FAIL [fdo#103232] -> PASS +1

  * igt@kms_flip@plain-flip-fb-recreate:
    - shard-skl:          FAIL [fdo#100368] -> PASS +1

  * igt@kms_frontbuffer_tracking@fbc-1p-primscrn-spr-indfb-draw-mmap-gtt:
    - shard-apl:          FAIL [fdo#103167] -> PASS +2

  * igt@kms_frontbuffer_tracking@fbc-2p-primscrn-spr-indfb-onoff:
    - shard-glk:          FAIL [fdo#103167] -> PASS +2

  * igt@kms_frontbuffer_tracking@fbcpsr-rgb565-draw-mmap-wc:
    - {shard-iclb}:       DMESG-FAIL [fdo#107724] -> PASS

  * igt@kms_frontbuffer_tracking@fbcpsr-tilingchange:
    - shard-skl:          FAIL [fdo#105682] -> PASS +1

  * igt@kms_frontbuffer_tracking@psr-1p-primscrn-spr-indfb-fullscreen:
    - {shard-iclb}:       FAIL [fdo#103167] -> PASS +2

  * igt@kms_frontbuffer_tracking@psr-rgb101010-draw-blt:
    - shard-skl:          FAIL [fdo#103167] -> PASS +1

  * igt@kms_plane@plane-panning-bottom-right-suspend-pipe-a-planes:
    - {shard-iclb}:       INCOMPLETE [fdo#107713] -> PASS

  * igt@kms_plane_multiple@atomic-pipe-a-tiling-x:
    - shard-glk:          FAIL [fdo#103166] -> PASS

  * igt@kms_plane_multiple@atomic-pipe-b-tiling-yf:
    - {shard-iclb}:       FAIL [fdo#103166] -> PASS +3

  * igt@kms_plane_multiple@atomic-pipe-c-tiling-y:
    - shard-apl:          FAIL [fdo#103166] -> PASS

  * igt@kms_rotation_crc@primary-rotation-180:
    - shard-skl:          FAIL [fdo#103925] / [fdo#107815] -> PASS

  * igt@pm_rpm@legacy-planes:
    - {shard-iclb}:       DMESG-WARN [fdo#108654] -> PASS

  * igt@pm_rpm@legacy-planes-dpms:
    - shard-skl:          INCOMPLETE [fdo#105959] / [fdo#107807] -> PASS

  * igt@pm_rpm@system-suspend-devices:
    - shard-skl:          INCOMPLETE [fdo#107807] -> PASS +1

  * igt@pm_rpm@system-suspend-execbuf:
    - shard-skl:          INCOMPLETE [fdo#104108] / [fdo#107807] -> PASS

  * igt@sw_sync@sync_busy_fork_unixsocket:
    - {shard-iclb}:       INCOMPLETE [fdo#108889] -> PASS

  
#### Warnings ####

  * igt@i915_selftest@live_contexts:
    - {shard-iclb}:       DMESG-FAIL [fdo#108569] -> INCOMPLETE [fdo#108315]

  
  {name}: This element is suppressed. This means it is ignored when computing
          the status of the difference (SUCCESS, WARNING, or FAILURE).

  [fdo#100368]: https://bugs.freedesktop.org/show_bug.cgi?id=100368
  [fdo#103166]: https://bugs.freedesktop.org/show_bug.cgi?id=103166
  [fdo#103167]: https://bugs.freedesktop.org/show_bug.cgi?id=103167
  [fdo#103184]: https://bugs.freedesktop.org/show_bug.cgi?id=103184
  [fdo#103232]: https://bugs.freedesktop.org/show_bug.cgi?id=103232
  [fdo#103665]: https://bugs.freedesktop.org/show_bug.cgi?id=103665
  [fdo#103925]: https://bugs.freedesktop.org/show_bug.cgi?id=103925
  [fdo#104108]: https://bugs.freedesktop.org/show_bug.cgi?id=104108
  [fdo#104671]: https://bugs.freedesktop.org/show_bug.cgi?id=104671
  [fdo#105363]: https://bugs.freedesktop.org/show_bug.cgi?id=105363
  [fdo#105682]: https://bugs.freedesktop.org/show_bug.cgi?id=105682
  [fdo#105959]: https://bugs.freedesktop.org/show_bug.cgi?id=105959
  [fdo#106023]: https://bugs.freedesktop.org/show_bug.cgi?id=106023
  [fdo#106887]: https://bugs.freedesktop.org/show_bug.cgi?id=106887
  [fdo#107713]: https://bugs.freedesktop.org/show_bug.cgi?id=107713
  [fdo#107724]: https://bugs.freedesktop.org/show_bug.cgi?id=107724
  [fdo#107807]: https://bugs.freedesktop.org/show_bug.cgi?id=107807
  [fdo#107815]: https://bugs.freedesktop.org/show_bug.cgi?id=107815
  [fdo#107847]: https://bugs.freedesktop.org/show_bug.cgi?id=107847
  [fdo#107882]: https://bugs.freedesktop.org/show_bug.cgi?id=107882
  [fdo#107956]: https://bugs.freedesktop.org/show_bug.cgi?id=107956
  [fdo#108039]: https://bugs.freedesktop.org/show_bug.cgi?id=108039
  [fdo#108145]: https://bugs.freedesktop.org/show_bug.cgi?id=108145
  [fdo#108315]: https://bugs.freedesktop.org/show_bug.cgi?id=108315
  [fdo#108341]: https://bugs.freedesktop.org/show_bug.cgi?id=108341
  [fdo#108569]: https://bugs.freedesktop.org/show_bug.cgi?id=108569
  [fdo#108654]: https://bugs.freedesktop.org/show_bug.cgi?id=108654
  [fdo#108667]: https://bugs.freedesktop.org/show_bug.cgi?id=108667
  [fdo#108840]: https://bugs.freedesktop.org/show_bug.cgi?id=108840
  [fdo#108889]: https://bugs.freedesktop.org/show_bug.cgi?id=108889
  [fdo#99912]: https://bugs.freedesktop.org/show_bug.cgi?id=99912


Participating hosts (7 -> 7)
------------------------------

  No changes in participating hosts


Build changes
-------------

    * Linux: CI_DRM_5230 -> Patchwork_10976

  CI_DRM_5230: b0a2de64f8969163f6e01071d5e05748f18a8bab @ git://anongit.freedesktop.org/gfx-ci/linux
  IGT_4736: 285ebfb3b7adc56586031afa5150c4e5ad40c229 @ git://anongit.freedesktop.org/xorg/app/intel-gpu-tools
  Patchwork_10976: c19b34b6c7fe784029535a8743bd098eb06cdadd @ git://anongit.freedesktop.org/gfx-ci/linux
  piglit_4509: fdc5a4ca11124ab8413c7988896eec4c97336694 @ git://anongit.freedesktop.org/piglit

== Logs ==

For more details see: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_10976/
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

[PATCH] drm/i915: Complete the fences as they are cancelled due to wedging

[Chris Wilson]

We inspect the requests under the assumption that they will be marked as
completed when they are removed from the queue. Currently however, in the
process of wedging the requests will be removed from the queue before they
are completed, so rearrange the code to complete the fences before the
locks are dropped.

<1>[  354.473346] BUG: unable to handle kernel NULL pointer dereference at 0000000000000250
<6>[  354.473363] PGD 0 P4D 0
<4>[  354.473370] Oops: 0000 [#1] PREEMPT SMP PTI
<4>[  354.473380] CPU: 0 PID: 4470 Comm: gem_eio Tainted: G     U            4.20.0-rc4-CI-CI_DRM_5216+ #1
<4>[  354.473393] Hardware name: Intel Corporation NUC7CJYH/NUC7JYB, BIOS JYGLKCPX.86A.0027.2018.0125.1347 01/25/2018
<4>[  354.473480] RIP: 0010:__i915_schedule+0x311/0x5e0 [i915]
<4>[  354.473490] Code: 49 89 44 24 20 4d 89 4c 24 28 4d 89 29 44 39 b3 a0 04 00 00 7d 3a 41 8b 44 24 78 85 c0 74 13 48 8b 93 78 04 00 00 48 83 e2 fc <39> 82 50 02 00 00 79 1e 44 89 b3 a0 04 00 00 48 8d bb d0 03 00 00
<4>[  354.473515] RSP: 0018:ffffc900001bba90 EFLAGS: 00010046
<4>[  354.473524] RAX: 0000000000000003 RBX: ffff8882624c8008 RCX: f34a737800000000
<4>[  354.473535] RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffff8882624c8048
<4>[  354.473545] RBP: ffffc900001bbab0 R08: 000000005963f1f1 R09: 0000000000000000
<4>[  354.473556] R10: ffffc900001bba10 R11: ffff8882624c8060 R12: ffff88824fdd7b98
<4>[  354.473567] R13: ffff88824fdd7bb8 R14: 0000000000000001 R15: ffff88824fdd7750
<4>[  354.473578] FS:  00007f44b4b5b980(0000) GS:ffff888277e00000(0000) knlGS:0000000000000000
<4>[  354.473590] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
<4>[  354.473599] CR2: 0000000000000250 CR3: 000000026976e000 CR4: 0000000000340ef0
<4>[  354.473611] Call Trace:
<4>[  354.473622]  ? lock_acquire+0xa6/0x1c0
<4>[  354.473677]  ? i915_schedule_bump_priority+0x57/0xd0 [i915]
<4>[  354.473736]  i915_schedule_bump_priority+0x72/0xd0 [i915]
<4>[  354.473792]  i915_request_wait+0x4db/0x840 [i915]
<4>[  354.473804]  ? get_pwq.isra.4+0x2c/0x50
<4>[  354.473813]  ? ___preempt_schedule+0x16/0x18
<4>[  354.473824]  ? wake_up_q+0x70/0x70
<4>[  354.473831]  ? wake_up_q+0x70/0x70
<4>[  354.473882]  ? gen6_rps_boost+0x118/0x120 [i915]
<4>[  354.473936]  i915_gem_object_wait_fence+0x8a/0x110 [i915]
<4>[  354.473991]  i915_gem_object_wait+0x113/0x500 [i915]
<4>[  354.474047]  i915_gem_wait_ioctl+0x11c/0x2f0 [i915]
<4>[  354.474101]  ? i915_gem_unset_wedged+0x210/0x210 [i915]
<4>[  354.474113]  drm_ioctl_kernel+0x81/0xf0
<4>[  354.474123]  drm_ioctl+0x2de/0x390
<4>[  354.474175]  ? i915_gem_unset_wedged+0x210/0x210 [i915]
<4>[  354.474187]  ? finish_task_switch+0x95/0x260
<4>[  354.474197]  ? lock_acquire+0xa6/0x1c0
<4>[  354.474207]  do_vfs_ioctl+0xa0/0x6e0
<4>[  354.474217]  ? __fget+0xfc/0x1e0
<4>[  354.474225]  ksys_ioctl+0x35/0x60
<4>[  354.474233]  __x64_sys_ioctl+0x11/0x20
<4>[  354.474241]  do_syscall_64+0x55/0x190
<4>[  354.474251]  entry_SYSCALL_64_after_hwframe+0x49/0xbe
<4>[  354.474260] RIP: 0033:0x7f44b3de65d7
<4>[  354.474267] Code: b3 66 90 48 8b 05 b1 48 2d 00 64 c7 00 26 00 00 00 48 c7 c0 ff ff ff ff c3 66 2e 0f 1f 84 00 00 00 00 00 b8 10 00 00 00 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d 81 48 2d 00 f7 d8 64 89 01 48
<4>[  354.474293] RSP: 002b:00007fff974948e8 EFLAGS: 00000246 ORIG_RAX: 0000000000000010
<4>[  354.474305] RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007f44b3de65d7
<4>[  354.474316] RDX: 00007fff97494940 RSI: 00000000c010646c RDI: 0000000000000007
<4>[  354.474327] RBP: 00007fff97494940 R08: 0000000000000000 R09: 00007f44b40bbc40
<4>[  354.474337] R10: 0000000000000000 R11: 0000000000000246 R12: 00000000c010646c
<4>[  354.474348] R13: 0000000000000007 R14: 0000000000000000 R15: 0000000000000000

v2: Avoid floating requests.
v3: Can't call dma_fence_signal() under the timeline lock!
v4: Can't call dma_fence_signal() from inside another fence either.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/i915_gem.c         | 54 +++++--------------------
 drivers/gpu/drm/i915/intel_lrc.c        | 13 ++++--
 drivers/gpu/drm/i915/intel_ringbuffer.c | 13 +++++-
 3 files changed, 31 insertions(+), 49 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index c55b1f75c980..834240a9b262 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -3308,16 +3308,6 @@ void i915_gem_reset_finish(struct drm_i915_private *dev_priv)
 }
 
 static void nop_submit_request(struct i915_request *request)
-{
-	GEM_TRACE("%s fence %llx:%d -> -EIO\n",
-		  request->engine->name,
-		  request->fence.context, request->fence.seqno);
-	dma_fence_set_error(&request->fence, -EIO);
-
-	i915_request_submit(request);
-}
-
-static void nop_complete_submit_request(struct i915_request *request)
 {
 	unsigned long flags;
 
@@ -3354,57 +3344,33 @@ void i915_gem_set_wedged(struct drm_i915_private *i915)
 	 * rolling the global seqno forward (since this would complete requests
 	 * for which we haven't set the fence error to EIO yet).
 	 */
-	for_each_engine(engine, i915, id) {
+	for_each_engine(engine, i915, id)
 		i915_gem_reset_prepare_engine(engine);
 
-		engine->submit_request = nop_submit_request;
-		engine->schedule = NULL;
-	}
-	i915->caps.scheduler = 0;
-
 	/* Even if the GPU reset fails, it should still stop the engines */
 	if (INTEL_GEN(i915) >= 5)
 		intel_gpu_reset(i915, ALL_ENGINES);
 
-	/*
-	 * Make sure no one is running the old callback before we proceed with
-	 * cancelling requests and resetting the completion tracking. Otherwise
-	 * we might submit a request to the hardware which never completes.
-	 */
-	synchronize_rcu();
-
 	for_each_engine(engine, i915, id) {
-		/* Mark all executing requests as skipped */
-		engine->cancel_requests(engine);
-
-		/*
-		 * Only once we've force-cancelled all in-flight requests can we
-		 * start to complete all requests.
-		 */
-		engine->submit_request = nop_complete_submit_request;
+		engine->submit_request = nop_submit_request;
+		engine->schedule = NULL;
 	}
+	i915->caps.scheduler = 0;
 
 	/*
 	 * Make sure no request can slip through without getting completed by
 	 * either this call here to intel_engine_init_global_seqno, or the one
-	 * in nop_complete_submit_request.
+	 * in nop_submit_request.
 	 */
 	synchronize_rcu();
 
-	for_each_engine(engine, i915, id) {
-		unsigned long flags;
-
-		/*
-		 * Mark all pending requests as complete so that any concurrent
-		 * (lockless) lookup doesn't try and wait upon the request as we
-		 * reset it.
-		 */
-		spin_lock_irqsave(&engine->timeline.lock, flags);
-		intel_engine_init_global_seqno(engine,
-					       intel_engine_last_submit(engine));
-		spin_unlock_irqrestore(&engine->timeline.lock, flags);
+	/* Mark all executing requests as skipped */
+	for_each_engine(engine, i915, id)
+		engine->cancel_requests(engine);
 
+	for_each_engine(engine, i915, id) {
 		i915_gem_reset_finish_engine(engine);
+		intel_engine_wakeup(engine);
 	}
 
 out:
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index 11f4e6148557..b5511a054f30 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -818,8 +818,11 @@ static void execlists_cancel_requests(struct intel_engine_cs *engine)
 	/* Mark all executing requests as skipped. */
 	list_for_each_entry(rq, &engine->timeline.requests, link) {
 		GEM_BUG_ON(!rq->global_seqno);
-		if (!i915_request_completed(rq))
-			dma_fence_set_error(&rq->fence, -EIO);
+
+		if (test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &rq->fence.flags))
+			continue;
+
+		dma_fence_set_error(&rq->fence, -EIO);
 	}
 
 	/* Flush the queued requests to the timeline list (for retiring). */
@@ -830,8 +833,8 @@ static void execlists_cancel_requests(struct intel_engine_cs *engine)
 		priolist_for_each_request_consume(rq, rn, p, i) {
 			list_del_init(&rq->sched.link);
 
-			dma_fence_set_error(&rq->fence, -EIO);
 			__i915_request_submit(rq);
+			dma_fence_set_error(&rq->fence, -EIO);
 		}
 
 		rb_erase_cached(&p->node, &execlists->queue);
@@ -839,6 +842,10 @@ static void execlists_cancel_requests(struct intel_engine_cs *engine)
 			kmem_cache_free(engine->i915->priorities, p);
 	}
 
+	intel_write_status_page(engine,
+				I915_GEM_HWS_INDEX,
+				intel_engine_last_submit(engine));
+
 	/* Remaining _unready_ requests will be nop'ed when submitted */
 
 	execlists->queue_priority = INT_MIN;
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index 28ae1e436ea6..992889f9e0ff 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -748,9 +748,18 @@ static void cancel_requests(struct intel_engine_cs *engine)
 	/* Mark all submitted requests as skipped. */
 	list_for_each_entry(request, &engine->timeline.requests, link) {
 		GEM_BUG_ON(!request->global_seqno);
-		if (!i915_request_completed(request))
-			dma_fence_set_error(&request->fence, -EIO);
+
+		if (test_bit(DMA_FENCE_FLAG_SIGNALED_BIT,
+			     &request->fence.flags))
+			continue;
+
+		dma_fence_set_error(&request->fence, -EIO);
 	}
+
+	intel_write_status_page(engine,
+				I915_GEM_HWS_INDEX,
+				intel_engine_last_submit(engine));
+
 	/* Remaining _unready_ requests will be nop'ed when submitted */
 
 	spin_unlock_irqrestore(&engine->timeline.lock, flags);
-- 
2.20.0.rc1

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

✗ Fi.CI.BAT: failure for series starting with drm/i915: Complete the fences as they are cancelled due to wedging (rev2)

[Patchwork]

== Series Details ==

Series: series starting with drm/i915: Complete the fences as they are cancelled due to wedging (rev2)
URL   : https://patchwork.freedesktop.org/series/53308/
State : failure

== Summary ==

Applying: drm/i915: Complete the fences as they are cancelled due to wedging
Applying: drm/i915/ringbuffer: Clear semaphore sync registers on ring init
Applying: drm/i915: Allocate a common scratch page
Applying: drm/i915/selftests: Terminate hangcheck sanitycheck forcibly
Applying: drm/i915/selftests: Reorder request allocation vs vma pinning
Using index info to reconstruct a base tree...
M	drivers/gpu/drm/i915/selftests/intel_lrc.c
Falling back to patching base and 3-way merge...
Auto-merging drivers/gpu/drm/i915/selftests/intel_lrc.c
CONFLICT (content): Merge conflict in drivers/gpu/drm/i915/selftests/intel_lrc.c
error: Failed to merge in the changes.
Patch failed at 0005 drm/i915/selftests: Reorder request allocation vs vma pinning
Use 'git am --show-current-patch' to see the failed patch
When you have resolved this problem, run "git am --continue".
If you prefer to skip this patch, run "git am --skip" instead.
To restore the original branch and stop patching, run "git am --abort".

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx