Quoting Tvrtko Ursulin (2017-11-22 20:12:54) > > On 22/11/2017 18:05, Chris Wilson wrote: > > Make sure the HW is idle before we start sampling the GPU for busyness. > > If we do not rest for long enough between tests, we may carry the > > sampling over. > > Hm.. not sure about this one. New PMU client is supposed to start with a > counter relative to the current value at counter enable. How did you > come to this being a problem? I was just staring at the legacy submission inaccuracy, thinking it was due to sampling earlier state and the knee-jerk reaction is to put quiescent gpu between failed subtests. (Which is due to not clearing the prev across parked :) I haven't tested without this part, but with the two parts combined, perf_pmu on byt has been stable (passing). -Chris _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/intel-gfx
