[PATCH 03/55] drm/i915/gt: Restore ce->signal flush before releasing virtual engine

Chris Wilson chris at chris-wilson.co.uk
Fri Jan 1 22:00:23 UTC 2021


Before we mark the virtual engine as no longer inflight, flush any
ongoing signaling that may be using the ce->signal_link along the
previous breadcrumbs. On switch to a new physical engine, that link will
be inserted into the new set of breadcrumbs, causing confusion to an
ongoing iterator.

This patch undoes a last minute mistake introduced into commit
bab0557c8dca ("drm/i915/gt: Remove virtual breadcrumb before transfer"),
whereby instead of unconditionally applying the flush, it was only
applied if the request itself was going to be reused.

Fixes: bab0557c8dca ("drm/i915/gt: Remove virtual breadcrumb before transfer")
Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/gt/intel_execlists_submission.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/i915/gt/intel_execlists_submission.c b/drivers/gpu/drm/i915/gt/intel_execlists_submission.c
index f3ecfa24432f..e57d8dd0ecc1 100644
--- a/drivers/gpu/drm/i915/gt/intel_execlists_submission.c
+++ b/drivers/gpu/drm/i915/gt/intel_execlists_submission.c
@@ -592,8 +592,6 @@ resubmit_virtual_request(struct i915_request *rq, struct virtual_engine *ve)
 		 * ce->signal_link.
 		 */
 		i915_request_cancel_breadcrumb(rq);
-		while (atomic_read(&engine->breadcrumbs->signaler_active))
-			cpu_relax();
 	}
 
 	spin_lock_irq(&engine->active.lock);
@@ -609,6 +607,7 @@ static void kick_siblings(struct i915_request *rq, struct intel_context *ce)
 {
 	struct virtual_engine *ve = container_of(ce, typeof(*ve), context);
 	struct intel_engine_cs *engine = rq->engine;
+	bool signals = !list_empty(&ce->signals);
 
 	/*
 	 * This engine is now too busy to run this virtual request, so
@@ -622,6 +621,10 @@ static void kick_siblings(struct i915_request *rq, struct intel_context *ce)
 
 	if (READ_ONCE(ve->request))
 		tasklet_hi_schedule(&ve->base.execlists.tasklet);
+
+	/* Flush concurrent signal_irq_work before we reuse the link */
+	while (signals && atomic_read(&engine->breadcrumbs->signaler_active))
+		cpu_relax();
 }
 
 static inline void __execlists_schedule_out(struct i915_request *rq)
-- 
2.20.1



More information about the Intel-gfx-trybot mailing list