Hi Umesh, Thanks for taking a look. On Friday, 3 February 2023 20:21:38 CET Umesh Nerlige Ramappa wrote: > On Tue, Jan 31, 2023 at 10:17:29AM +0100, Janusz Krzysztofik wrote: > >Users reported oopses on list corruptions when using i915 perf with a > >number of concurrently running graphics applications. That indicates we > >are currently missing some important tests for such scenarios. Cover > >that gap. > > Any bug ids that were filed for the issue? Yes, https://gitlab.freedesktop.org/drm/intel/issues/6333. > Also what were the parameters > passed to perf OA when the issues were reported? This could help > actually root cause the issue and have a test closer to the actual use > case. Root cause has been already identified, and it lays not in perf, only in barriers handling code, see https://patchwork.freedesktop.org/series/113636/ (submitted to trybot so far, soon to the public list). Thanks, Janusz > > Regards, > Umesh > > > >Chris Wilson (1): > > i915/perf: Stress opening of new perf streams against existing > > contexts > > > >Janusz Krzysztofik (1): > > tests/i915/perf: Exercise barrier race > > > > tests/i915/perf.c | 111 ++++++++++++++++++++++++++++++++++++++++++++++ > > 1 file changed, 111 insertions(+) > > >
