diff options
| author | Tvrtko Ursulin <tvrtko.ursulin@intel.com> | 2017-11-23 09:24:01 +0000 |
|---|---|---|
| committer | Tvrtko Ursulin <tvrtko.ursulin@intel.com> | 2017-11-23 12:29:25 +0000 |
| commit | a1e444f4c8178acb590d41c21e921c6447668be4 (patch) | |
| tree | 35b8e3775999fdd6c9e40b450d4d81136c1389c3 /tests | |
| parent | 8cf978bb55c312fb585364451c5aaf9f4ef9781a (diff) | |
tests/perf_pmu: Bump measuring duration for semaphores as well
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 <tvrtko.ursulin@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Diffstat (limited to 'tests')
| -rw-r--r-- | tests/perf_pmu.c | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/tests/perf_pmu.c b/tests/perf_pmu.c index bf862b2f..a97f5900 100644 --- a/tests/perf_pmu.c +++ b/tests/perf_pmu.c @@ -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; |