RE: [PATCH v5 2/3] drm/i915/guc: Close deregister-context race against CT-loss

2023-10-18 Thread Gupta, Anshuman
C > ; Gupta, Anshuman > ; Ursulin, Tvrtko ; > Jana, Mousumi > Subject: [PATCH v5 2/3] drm/i915/guc: Close deregister-context race against > CT-loss > > If we are at the end of suspend or very early in resume its possible an async > fence signal (via rcu_call) is trigge

[PATCH v5 2/3] drm/i915/guc: Close deregister-context race against CT-loss

2023-10-13 Thread Alan Previn
If we are at the end of suspend or very early in resume its possible an async fence signal (via rcu_call) is triggered to free_engines which could lead us to the execution of the context destruction worker (after a prior worker flush). Thus, when suspending, insert rcu_barriers at the start of i91