Re: [Intel-gfx] [PATCH v5 1/4] drm/i915/perf: break OA config buffer object in 2

2020-04-14 Thread Lionel Landwerlin
Yeah it's only filtering. We need to pin the context to a particular HW tag so that filtering can work properly. -Lionel On 14/04/2020 02:24, Umesh Nerlige Ramappa wrote: Hi Lionel, What's the implication of using separate contexts for 3d and compute on perf OA? Is it only

Re: [Intel-gfx] [PATCH v5 1/4] drm/i915/perf: break OA config buffer object in 2

2020-04-13 Thread Umesh Nerlige Ramappa
Hi Lionel, What's the implication of using separate contexts for 3d and compute on perf OA? Is it only context-filtering? If so, have you considered disabling context filtering with a parameter instead of actually filtering for specific contexts? Is this privileged use case? Thanks, Umesh

[Intel-gfx] [PATCH v5 1/4] drm/i915/perf: break OA config buffer object in 2

2020-04-09 Thread Lionel Landwerlin
We want to enable performance monitoring on multiple contexts to cover the Iris use case of using 2 GEM contexts (3D & compute). So start by breaking the OA configuration BO which contains global & per context register writes. NOA muxes & OA configurations are global, while FLEXEU register