[Intel-gfx] [igt-dev] [PATCH i-g-t] tests/perf_pmu: Improve accuracy by waiting on spinner to start

[Chris Wilson]

Quoting Tvrtko Ursulin (2018-03-15 14:53:08)
> 
> On 15/03/2018 14:46, Chris Wilson wrote:
> > Quoting Tvrtko Ursulin (2018-03-15 14:37:59)
> >>
> >> On 15/03/2018 13:45, Chris Wilson wrote:
> >>> As we are making changes to igt_spin_t, one of the ideas was that we put
> >>> the obj[] array there (with the offsets and flags setup correctly) so
> >>> that we could just feed that in again later without having to worry
> >>> about the relocations.
> >>
> >> I tried that before but we couldn't agree on resubmit semantics.
> >>
> >> My patch had igt_spin_batch_restart(fd, spin) - so emitting the exact
> >> same batch, including the dependency. That would actually work well for
> >> this use case.
> >>
> >> So if you are happy with that, I can resurrect that patch, add one more
> >> to implement stuff from this patch, and rebase perf_pmu changes to follow.
> > 
> > Honestly, best to do here first, as we will probably take forever to come
> > up with something we both like and applies to more test cases.
> 
> Don't quite get what you mean by "best to do here first" - where is 
> here?

Fix perf_pmu, then worry about API. We're still waiting for kasan
results, we may have more work ahead of us yet.

> Still locally to perf_pmu so no generic spin batch resubmit yet? 
> But I can cache the obj array under spin_batch_t as a shortcut for time 
> being?

I'd take igt_spin_t.obj[] :) But I don't insist on it, I'd like to get
the wait-for-spin working before tackling the resubmit API. There's a
few other places that either have a open-coded wait-for-submit, or need
one.
-Chris