Quoting Mika Kuoppala (2020-01-31 13:22:38)
> Chris Wilson <chris@xxxxxxxxxxxxxxxxxx> writes:
>
> > If the heartbeat fires in the middle of the preempt-hang test, it
> > consumes our forced hang disrupting the test.
> >
> > Reported-by: Daniel Vetter <daniel.vetter@xxxxxxxx>
> > Signed-off-by: Chris Wilson <chris@xxxxxxxxxxxxxxxxxx>
> > ---
> > drivers/gpu/drm/i915/gt/selftest_lrc.c | 10 ++++++++++
> > 1 file changed, 10 insertions(+)
> >
> > diff --git a/drivers/gpu/drm/i915/gt/selftest_lrc.c b/drivers/gpu/drm/i915/gt/selftest_lrc.c
> > index 5e04ecb61dcc..f5214a374fb7 100644
> > --- a/drivers/gpu/drm/i915/gt/selftest_lrc.c
> > +++ b/drivers/gpu/drm/i915/gt/selftest_lrc.c
> > @@ -2452,15 +2452,19 @@ static int live_preempt_hang(void *arg)
> > I915_USER_PRIORITY(I915_CONTEXT_MIN_USER_PRIORITY);
> >
> > for_each_engine(engine, gt, id) {
> > + unsigned long heartbeat;
> > struct i915_request *rq;
> >
> > if (!intel_engine_has_preemption(engine))
> > continue;
> >
> > + engine_heartbeat_disable(engine, &heartbeat);
> > +
> > rq = spinner_create_request(&spin_lo, ctx_lo, engine,
> > MI_ARB_CHECK);
> > if (IS_ERR(rq)) {
> > err = PTR_ERR(rq);
> > + engine_heartbeat_enable(engine, heartbeat);
>
> Was thinking that you could grab the engine and abuse then the error
> path to enable.
>
> But too hairy.
Yeah, we've done staggered error paths, but both look ugly, So I took
the easy approach. Move the inner loop to its own test function usually
helps though.
-Chris
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
