tests/perf_pmu: Bump measuring duration for semaphores as well
PATCH 1/1 tests/perf_pmu: Bump measuring duration for semaphores as well
From: Tvrtko Ursulin <[email protected]>
As Chris has discover 100ms is not long enough to cover the sampling error in general, fix the semaphore subtest as well to measure for 500ms.
--- tests/perf_pmu.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
tests/perf_pmu.c +1 -1
@@ -416,7 +416,7 @@ sema_wait(int gem_fd, const struct intel_execution_engine2 *e)gem_execbuf (gem_fd, &eb); - slept =measured_usleep (100e3 ); + slept =measured_usleep (batch_duration_ns /1000 ); *obj_ptr =1 ;
Signature · 6 lines
2.14.1 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Quoting Tvrtko Ursulin (2017-11-23 09:24:01)
Quoted 8 lines
From: Tvrtko Ursulin <[email protected]>
As Chris has discover 100ms is not long enough to cover the sampling error in general, fix the semaphore subtest as well to measure for 500ms.
Signed-off-by Tvrtko Ursulin <[email protected]>Cc Chris Wilson <[email protected]>
And one value for all batches makes it easier for us to reason about,
At least any deviation from that rule deserves comment. -Chris
_______________________________________________
Intel-gfx mailing list [email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Patchwork
CI✗ Fi.CI.BAT: failure for tests/perf_pmu: Bump measuring duration for semaphores as well
In reply to Tvrtko Ursulin== Series Details ==
== Series Details ==
Series: tests/perf_pmu: Bump measuring duration for semaphores as well
URL : https://patchwork.freedesktop.org/series/34285/
State : failure
== Summary ==
Series 34285 revision 1 was fully merged or fully failed: no git log
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx