Grab execlist lock when cleaning up execlist queues after GPU reset to avoid
concurrency problems between the context event interrupt handler and the reset
path immediately following a GPU reset.
* v2 (Chris Wilson):
Do execlist check and use simpler form of spinlock functions.
* v3 (Tvrtko Ursulin):
Don't hold the execlist_lock while dereferencing the requests since that might
lead to sleep while IRQs are disabled.
Signed-off-by: Tomas Elf <tomas.elf@xxxxxxxxx>
Cc: Tvrtko Ursulin <tvrtko.ursulin@xxxxxxxxx>
Cc: Chris Wilson <chris@xxxxxxxxxxxxxxxxxx>
---
drivers/gpu/drm/i915/i915_gem.c | 16 +++++++---------
1 file changed, 7 insertions(+), 9 deletions(-)
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index e57061a..b492603 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -2753,18 +2753,16 @@ static void i915_gem_reset_ring_cleanup(struct drm_i915_private *dev_priv,
* are the ones that keep the context and ringbuffer backing objects
* pinned in place.
*/
- while (!list_empty(&ring->execlist_queue)) {
- struct drm_i915_gem_request *submit_req;
- submit_req = list_first_entry(&ring->execlist_queue,
- struct drm_i915_gem_request,
- execlist_link);
- list_del(&submit_req->execlist_link);
+ if (i915.enable_execlists) {
+ spin_lock_irq(&ring->execlist_lock);
- if (submit_req->ctx != ring->default_context)
- intel_lr_context_unpin(submit_req);
+ /* list_splice_tail_init checks for empty lists */
+ list_splice_tail_init(&ring->execlist_queue,
+ &ring->execlist_retired_req_list);
- i915_gem_request_unreference(submit_req);
+ spin_unlock_irq(&ring->execlist_lock);
+ intel_execlists_retire_requests(ring);
}
/*
--
1.9.1
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
