I can't explain it, but this did improve its first and only test run. Maybe it was a fluke, maybe I need to understand the barrier (engine-pm and timelines->lock) and how that is interacting with engine_retire and intel_gt_retire_requests()... Hopefully the last hurdle. Also wouldn't mind if it failed so I didn't have to find the explanation... --- drivers/gpu/drm/i915/i915_request.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_request.c b/drivers/gpu/drm/i915/i915_request.c index a558f64186fa..1b8e6827461a 100644 --- a/drivers/gpu/drm/i915/i915_request.c +++ b/drivers/gpu/drm/i915/i915_request.c @@ -269,12 +269,12 @@ bool i915_request_retire(struct i915_request *rq) GEM_BUG_ON(!list_empty(&rq->execute_cb)); spin_unlock_irq(&rq->lock); - remove_from_client(rq); - list_del(&rq->link); - intel_context_exit(rq->hw_context); intel_context_unpin(rq->hw_context); + remove_from_client(rq); + list_del(&rq->link); + free_capture_list(rq); i915_sched_node_fini(&rq->sched); i915_request_put(rq); -- 2.24.0 _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/intel-gfx
