drm/i915: Protect against request freeing during cancellation on wedging

As soon as we mark a request as completed, it may be retired. So when
cancelling a request and marking it complete, make sure we first keep a
reference to the request.

Signed-off-by: Chris Wilson <[email protected]>
Reviewed-by: Mika Kuoppala <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
Signed-off-by: Daniel Vetter <[email protected]>

Author: ickle

drivers/gpu/drm/i915/gt/intel_execlists_submission.c

@@ -2954,16 +2954,18 @@ static void execlists_reset_cancel(struct intel_engine_cs *engine)
 
 	/* Mark all executing requests as skipped. */
 	list_for_each_entry(rq, &engine->active.requests, sched.link)
-		i915_request_mark_eio(rq);
+		i915_request_put(i915_request_mark_eio(rq));
 	intel_engine_signal_breadcrumbs(engine);
 
 	/* Flush the queued requests to the timeline list (for retiring). */
 	while ((rb = rb_first_cached(&execlists->queue))) {
 		struct i915_priolist *p = to_priolist(rb);
 
 		priolist_for_each_request_consume(rq, rn, p) {
-			i915_request_mark_eio(rq);
-			__i915_request_submit(rq);
+			if (i915_request_mark_eio(rq)) {
+				__i915_request_submit(rq);
+				i915_request_put(rq);
+			}
 		}
 
 		rb_erase_cached(&p->node, &execlists->queue);
@@ -2972,7 +2974,7 @@ static void execlists_reset_cancel(struct intel_engine_cs *engine)
 
 	/* On-hold requests will be flushed to timeline upon their release */
 	list_for_each_entry(rq, &engine->active.hold, sched.link)
-		i915_request_mark_eio(rq);
+		i915_request_put(i915_request_mark_eio(rq));
 
 	/* Cancel all attached virtual engines */
 	while ((rb = rb_first_cached(&execlists->virtual))) {
@@ -2985,10 +2987,11 @@ static void execlists_reset_cancel(struct intel_engine_cs *engine)
 		spin_lock(&ve->base.active.lock);
 		rq = fetch_and_zero(&ve->request);
 		if (rq) {
-			i915_request_mark_eio(rq);
-
-			rq->engine = engine;
-			__i915_request_submit(rq);
+			if (i915_request_mark_eio(rq)) {
+				rq->engine = engine;
+				__i915_request_submit(rq);
+				i915_request_put(rq);
+			}
 			i915_request_put(rq);
 
 			ve->base.execlists.queue_priority_hint = INT_MIN;

drivers/gpu/drm/i915/gt/intel_reset.c

@@ -786,18 +786,15 @@ static void reset_finish(struct intel_gt *gt, intel_engine_mask_t awake)
 
 static void nop_submit_request(struct i915_request *request)
 {
-	struct intel_engine_cs *engine = request->engine;
-	unsigned long flags;
-
 	RQ_TRACE(request, "-EIO\n");
-	i915_request_set_error_once(request, -EIO);
 
-	spin_lock_irqsave(&engine->active.lock, flags);
-	__i915_request_submit(request);
-	i915_request_mark_complete(request);
-	spin_unlock_irqrestore(&engine->active.lock, flags);
+	request = i915_request_mark_eio(request);
+	if (request) {
+		i915_request_submit(request);
+		intel_engine_signal_breadcrumbs(request->engine);
 
-	intel_engine_signal_breadcrumbs(engine);
+		i915_request_put(request);
+	}
 }
 
 static void __intel_gt_set_wedged(struct intel_gt *gt)

drivers/gpu/drm/i915/gt/intel_ring_submission.c

@@ -411,7 +411,7 @@ static void reset_cancel(struct intel_engine_cs *engine)
 
 	/* Mark all submitted requests as skipped. */
 	list_for_each_entry(request, &engine->active.requests, sched.link)
-		i915_request_mark_eio(request);
+		i915_request_put(i915_request_mark_eio(request));
 	intel_engine_signal_breadcrumbs(engine);
 
 	/* Remaining _unready_ requests will be nop'ed when submitted */

drivers/gpu/drm/i915/gt/mock_engine.c

@@ -239,13 +239,15 @@ static void mock_reset_cancel(struct intel_engine_cs *engine)
 
 	/* Mark all submitted requests as skipped. */
 	list_for_each_entry(rq, &engine->active.requests, sched.link)
-		i915_request_mark_eio(rq);
+		i915_request_put(i915_request_mark_eio(rq));
 	intel_engine_signal_breadcrumbs(engine);
 
 	/* Cancel and submit all pending requests. */
 	list_for_each_entry(rq, &mock->hw_queue, mock.link) {
-		i915_request_mark_eio(rq);
-		__i915_request_submit(rq);
+		if (i915_request_mark_eio(rq)) {
+			__i915_request_submit(rq);
+			i915_request_put(rq);
+		}
 	}
 	INIT_LIST_HEAD(&mock->hw_queue);
 

drivers/gpu/drm/i915/i915_request.c

@@ -514,15 +514,20 @@ void i915_request_set_error_once(struct i915_request *rq, int error)
 	} while (!try_cmpxchg(&rq->fence.error, &old, error));
 }
 
-void i915_request_mark_eio(struct i915_request *rq)
+struct i915_request *i915_request_mark_eio(struct i915_request *rq)
 {
 	if (__i915_request_is_complete(rq))
-		return;
+		return NULL;
 
 	GEM_BUG_ON(i915_request_signaled(rq));
 
+	/* As soon as the request is completed, it may be retired */
+	rq = i915_request_get(rq);
+
 	i915_request_set_error_once(rq, -EIO);
 	i915_request_mark_complete(rq);
+
+	return rq;
 }
 
 bool __i915_request_submit(struct i915_request *request)

drivers/gpu/drm/i915/i915_request.h

@@ -311,7 +311,7 @@ i915_request_create(struct intel_context *ce);
 
 void __i915_request_skip(struct i915_request *rq);
 void i915_request_set_error_once(struct i915_request *rq, int error);
-void i915_request_mark_eio(struct i915_request *rq);
+struct i915_request *i915_request_mark_eio(struct i915_request *rq);
 
 struct i915_request *__i915_request_commit(struct i915_request *request);
 void __i915_request_queue(struct i915_request *rq,