[Intel-gfx] [PATCH v2] drm/i915/pmu: Reconstruct active state on starting busy-stats

Chris Wilson chris at chris-wilson.co.uk
Wed Jan 10 16:27:56 UTC 2018


Quoting Tvrtko Ursulin (2018-01-10 16:17:50)
> 
> On 10/01/2018 16:11, Chris Wilson wrote:
> > We have a hole in our busy-stat accounting if the pmu is enabled during
> > a long running batch, the pmu will not start accumulating busy-time
> > until the next context switch. This then fails tests that are only
> > sampling a single batch.
> > 
> > v2: Count each active port just once (context in/out events are only on
> > the first and last assigment to a port).
> > 
> > Testcase: igt/perf_pmu/busy-start
> > Testcase: igt/perf_pmu/busy-double-start
> > Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
> > Cc: Tvrtko Ursulin <tvrtko.ursulin at intel.com>
> > ---
> Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin at intel.com>

And Tvrkto also pointed out,
Fixes: 30e17b7847f5 ("drm/i915: Engine busy time tracking")

Though we are unlikely to be affected outside of the confines of igt
testing single batches and expecting perfection.
-Chris


More information about the Intel-gfx mailing list