Re: [Intel-gfx] [PATCH 2/2] drm/i915/lrc: Prevent preemption when lite-restore is disabled

2015-09-14 Thread Daniel Vetter
On Thu, Sep 10, 2015 at 06:37:17PM +0100, Arun Siluvery wrote: > On 04/09/2015 12:59, Michel Thierry wrote: > >When WaEnableForceRestoreInCtxtDescForVCS is required, it is only > >safe to send new contexts if the last reported event is "active to > >idle". Otherwise the same context can fully

Re: [Intel-gfx] [PATCH 2/2] drm/i915/lrc: Prevent preemption when lite-restore is disabled

2015-09-10 Thread Arun Siluvery
On 04/09/2015 12:59, Michel Thierry wrote: When WaEnableForceRestoreInCtxtDescForVCS is required, it is only safe to send new contexts if the last reported event is "active to idle". Otherwise the same context can fully preempt itself because lite-restore is disabled. Testcase:

[Intel-gfx] [PATCH 2/2] drm/i915/lrc: Prevent preemption when lite-restore is disabled

2015-09-04 Thread Michel Thierry
When WaEnableForceRestoreInCtxtDescForVCS is required, it is only safe to send new contexts if the last reported event is "active to idle". Otherwise the same context can fully preempt itself because lite-restore is disabled. Testcase: igt/gem_concurrent_blit Reported-by: Daniele Ceraolo Spurio

Re: [Intel-gfx] [PATCH 2/2] drm/i915/lrc: Prevent preemption when lite-restore is disabled

2015-09-04 Thread Daniele Ceraolo Spurio
On 04/09/15 12:59, Michel Thierry wrote: When WaEnableForceRestoreInCtxtDescForVCS is required, it is only safe to send new contexts if the last reported event is "active to idle". Otherwise the same context can fully preempt itself because lite-restore is disabled. Testcase: