drm/i915: Update reset path to fix incomplete requests

Update reset path in preparation for engine reset which requires
identification of incomplete requests and associated context and fixing
their state so that engine can resume correctly after reset.

The request that caused the hang will be skipped and head is reset to the
start of breadcrumb. This allows us to resume from where we left-off.
Since this request didn't complete normally we also need to cleanup elsp
queue manually. This is vital if we employ nonblocking request
submission where we may have a web of dependencies upon the hung request
and so advancing the seqno manually is no longer trivial.

ABI: gem_reset_stats / DRM_IOCTL_I915_GET_RESET_STATS

We change the way we count pending batches. Only the active context
involved in the reset is marked as either innocent or guilty, and not
mark the entire world as pending. By inspection this only affects
igt/gem_reset_stats (which assumes implementation details) and not
piglit.

ARB_robustness gives this guide on how we expect the user of this
interface to behave:

 * Provide a mechanism for an OpenGL application to learn about
   graphics resets that affect the context.  When a graphics reset
   occurs, the OpenGL context becomes unusable and the application
   must create a new context to continue operation. Detecting a
   graphics reset happens through an inexpensive query.

And with regards to the actual meaning of the reset values:

   Certain events can result in a reset of the GL context. Such a reset
   causes all context state to be lost. Recovery from such events
   requires recreation of all objects in the affected context. The
   current status of the graphics reset state is returned by

	enum GetGraphicsResetStatusARB();

   The symbolic constant returned indicates if the GL context has been
   in a reset state at any point since the last call to
   GetGraphicsResetStatusARB. NO_ERROR indicates that the GL context
   has not been in a reset state since the last call.
   GUILTY_CONTEXT_RESET_ARB indicates that a reset has been detected
   that is attributable to the current GL context.
   INNOCENT_CONTEXT_RESET_ARB indicates a reset has been detected that
   is not attributable to the current GL context.
   UNKNOWN_CONTEXT_RESET_ARB indicates a detected graphics reset whose
   cause is unknown.

The language here is explicit in that we must mark up the guilty batch,
but is loose enough for us to relax the innocent (i.e. pending)
accounting as only the active batches are involved with the reset.

In the future, we are looking towards single engine resetting (with
minimal locking), where it seems inappropriate to mark the entire world
as innocent since the reset occurred on a different engine. Reducing the
information available means we only have to encounter the pain once, and
also reduces the information leaking from one context to another.

v2: Legacy ringbuffer submission required a reset following hibernation,
or else we restore stale values to the RING_HEAD and walked over
stolen garbage.

v3: GuC requires replaying the requests after a reset.

v4: Restore engine IRQ after reset (so waiters will be woken!)
    Rearm hangcheck if resetting with a waiter.

Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@intel.com>
Cc: Arun Siluvery <arun.siluvery@linux.intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20160909131201.16673-13-chris@chris-wilson.co.uk

1 files changed, 75 insertions, 50 deletions

diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 23069a2d2850..674e0eaf39ea 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -2555,29 +2555,85 @@ i915_gem_find_active_request(struct intel_engine_cs *engine)
 	return NULL;
 }
 
-static void i915_gem_reset_engine_status(struct intel_engine_cs *engine)
+static void reset_request(struct drm_i915_gem_request *request)
+{
+	void *vaddr = request->ring->vaddr;
+	u32 head;
+
+	/* As this request likely depends on state from the lost
+	 * context, clear out all the user operations leaving the
+	 * breadcrumb at the end (so we get the fence notifications).
+	 */
+	head = request->head;
+	if (request->postfix < head) {
+		memset(vaddr + head, 0, request->ring->size - head);
+		head = 0;
+	}
+	memset(vaddr + head, 0, request->postfix - head);
+}
+
+static void i915_gem_reset_engine(struct intel_engine_cs *engine)
 {
 	struct drm_i915_gem_request *request;
+	struct i915_gem_context *incomplete_ctx;
 	bool ring_hung;
 
+	/* Ensure irq handler finishes, and not run again. */
+	tasklet_kill(&engine->irq_tasklet);
+	if (engine->irq_seqno_barrier)
+		engine->irq_seqno_barrier(engine);
+
 	request = i915_gem_find_active_request(engine);
-	if (request == NULL)
+	if (!request)
 		return;
 
 	ring_hung = engine->hangcheck.score >= HANGCHECK_SCORE_RING_HUNG;
-
 	i915_set_reset_status(request->ctx, ring_hung);
+	if (!ring_hung)
+		return;
+
+	DRM_DEBUG_DRIVER("resetting %s to restart from tail of request 0x%x\n",
+			 engine->name, request->fence.seqno);
+
+	/* Setup the CS to resume from the breadcrumb of the hung request */
+	engine->reset_hw(engine, request);
+
+	/* Users of the default context do not rely on logical state
+	 * preserved between batches. They have to emit full state on
+	 * every batch and so it is safe to execute queued requests following
+	 * the hang.
+	 *
+	 * Other contexts preserve state, now corrupt. We want to skip all
+	 * queued requests that reference the corrupt context.
+	 */
+	incomplete_ctx = request->ctx;
+	if (i915_gem_context_is_default(incomplete_ctx))
+		return;
+
 	list_for_each_entry_continue(request, &engine->request_list, link)
-		i915_set_reset_status(request->ctx, false);
+		if (request->ctx == incomplete_ctx)
+			reset_request(request);
 }
 
-static void i915_gem_reset_engine_cleanup(struct intel_engine_cs *engine)
+void i915_gem_reset(struct drm_i915_private *dev_priv)
 {
-	struct drm_i915_gem_request *request;
-	struct intel_ring *ring;
+	struct intel_engine_cs *engine;
 
-	/* Ensure irq handler finishes, and not run again. */
-	tasklet_kill(&engine->irq_tasklet);
+	i915_gem_retire_requests(dev_priv);
+
+	for_each_engine(engine, dev_priv)
+		i915_gem_reset_engine(engine);
+
+	i915_gem_restore_fences(&dev_priv->drm);
+}
+
+static void nop_submit_request(struct drm_i915_gem_request *request)
+{
+}
+
+static void i915_gem_cleanup_engine(struct intel_engine_cs *engine)
+{
+	engine->submit_request = nop_submit_request;
 
 	/* Mark all pending requests as complete so that any concurrent
 	 * (lockless) lookup doesn't try and wait upon the request as we
@@ -2600,54 +2656,22 @@ static void i915_gem_reset_engine_cleanup(struct intel_engine_cs *engine)
 		spin_unlock(&engine->execlist_lock);
 	}
 
-	/*
-	 * We must free the requests after all the corresponding objects have
-	 * been moved off active lists. Which is the same order as the normal
-	 * retire_requests function does. This is important if object hold
-	 * implicit references on things like e.g. ppgtt address spaces through
-	 * the request.
-	 */
-	request = i915_gem_active_raw(&engine->last_request,
-				      &engine->i915->drm.struct_mutex);
-	if (request)
-		i915_gem_request_retire_upto(request);
-	GEM_BUG_ON(intel_engine_is_active(engine));
-
-	/* Having flushed all requests from all queues, we know that all
-	 * ringbuffers must now be empty. However, since we do not reclaim
-	 * all space when retiring the request (to prevent HEADs colliding
-	 * with rapid ringbuffer wraparound) the amount of available space
-	 * upon reset is less than when we start. Do one more pass over
-	 * all the ringbuffers to reset last_retired_head.
-	 */
-	list_for_each_entry(ring, &engine->buffers, link) {
-		ring->last_retired_head = ring->tail;
-		intel_ring_update_space(ring);
-	}
-
 	engine->i915->gt.active_engines &= ~intel_engine_flag(engine);
 }
 
-void i915_gem_reset(struct drm_device *dev)
+void i915_gem_set_wedged(struct drm_i915_private *dev_priv)
 {
-	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct intel_engine_cs *engine;
 
-	/*
-	 * Before we free the objects from the requests, we need to inspect
-	 * them for finding the guilty party. As the requests only borrow
-	 * their reference to the objects, the inspection must be done first.
-	 */
-	for_each_engine(engine, dev_priv)
-		i915_gem_reset_engine_status(engine);
+	lockdep_assert_held(&dev_priv->drm.struct_mutex);
+	set_bit(I915_WEDGED, &dev_priv->gpu_error.flags);
 
+	i915_gem_context_lost(dev_priv);
 	for_each_engine(engine, dev_priv)
-		i915_gem_reset_engine_cleanup(engine);
+		i915_gem_cleanup_engine(engine);
 	mod_delayed_work(dev_priv->wq, &dev_priv->gt.idle_work, 0);
 
-	i915_gem_context_reset(dev);
-
-	i915_gem_restore_fences(dev);
+	i915_gem_retire_requests(dev_priv);
 }
 
 static void
@@ -4343,8 +4367,7 @@ void i915_gem_resume(struct drm_device *dev)
 	 * guarantee that the context image is complete. So let's just reset
 	 * it and start again.
 	 */
-	if (i915.enable_execlists)
-		intel_lr_context_reset(dev_priv, dev_priv->kernel_context);
+	dev_priv->gt.resume(dev_priv);
 
 	mutex_unlock(&dev->struct_mutex);
 }
@@ -4496,8 +4519,10 @@ int i915_gem_init(struct drm_device *dev)
 	mutex_lock(&dev->struct_mutex);
 
 	if (!i915.enable_execlists) {
+		dev_priv->gt.resume = intel_legacy_submission_resume;
 		dev_priv->gt.cleanup_engine = intel_engine_cleanup;
 	} else {
+		dev_priv->gt.resume = intel_lr_context_resume;
 		dev_priv->gt.cleanup_engine = intel_logical_ring_cleanup;
 	}
 
@@ -4530,7 +4555,7 @@ int i915_gem_init(struct drm_device *dev)
 		 * for all other failure, such as an allocation failure, bail.
 		 */
 		DRM_ERROR("Failed to initialize GPU, declaring it wedged\n");
-		set_bit(I915_WEDGED, &dev_priv->gpu_error.flags);
+		i915_gem_set_wedged(dev_priv);
 		ret = 0;
 	}