[Intel-gfx] [PATCH 3/9] drm/i915/selftests: Enable selftesting of busy-stats

[Chris Wilson]

Quoting Mika Kuoppala (2020-06-16 10:03:19)
> Chris Wilson <chris at chris-wilson.co.uk> writes:
> 
> > A couple of very simple tests to ensure that the basic properties of
> > per-engine busyness accounting [0% and 100% busy] are faithful.
> >
> > Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
> > ---
> >  drivers/gpu/drm/i915/gt/selftest_engine_pm.c | 94 ++++++++++++++++++++
> >  drivers/gpu/drm/i915/gt/selftest_rps.c       |  5 ++
> >  2 files changed, 99 insertions(+)
> >
> > diff --git a/drivers/gpu/drm/i915/gt/selftest_engine_pm.c b/drivers/gpu/drm/i915/gt/selftest_engine_pm.c
> > index cbf6b0735272..fb0fd8a7db9a 100644
> > --- a/drivers/gpu/drm/i915/gt/selftest_engine_pm.c
> > +++ b/drivers/gpu/drm/i915/gt/selftest_engine_pm.c
> > @@ -7,6 +7,99 @@
> >  #include "i915_selftest.h"
> >  #include "selftest_engine.h"
> >  #include "selftests/igt_atomic.h"
> > +#include "selftests/igt_flush_test.h"
> > +#include "selftests/igt_spinner.h"
> > +
> > +static int live_engine_busy_stats(void *arg)
> > +{
> > +     struct intel_gt *gt = arg;
> > +     struct intel_engine_cs *engine;
> > +     enum intel_engine_id id;
> > +     struct igt_spinner spin;
> > +     int err = 0;
> > +
> > +     /*
> > +      * Check that if an engine supports busy-stats, they tell the truth.
> > +      */
> > +
> > +     if (igt_spinner_init(&spin, gt))
> > +             return -ENOMEM;
> > +
> > +     GEM_BUG_ON(intel_gt_pm_is_awake(gt));
> > +     for_each_engine(engine, gt, id) {
> > +             struct i915_request *rq;
> > +             ktime_t de;
> > +             u64 dt;
> > +
> > +             if (!intel_engine_supports_stats(engine))
> > +                     continue;
> > +
> > +             if (!intel_engine_can_store_dword(engine))
> > +                     continue;
> > +
> > +             if (intel_gt_pm_wait_for_idle(gt)) {
> > +                     err = -EBUSY;
> > +                     break;
> > +             }
> > +
> > +             preempt_disable();
> > +             dt = ktime_to_ns(ktime_get());
> > +             de = intel_engine_get_busy_time(engine);
> > +             udelay(100);
> > +             de = ktime_sub(intel_engine_get_busy_time(engine), de);
> > +             dt = ktime_to_ns(ktime_get()) - dt;
> > +             preempt_enable();
> > +             if (de > 10) {
> 
> 10 is from stetson?
> 
> Well I would say it is strict enough.
> 
> The signed de just makes me nervous, so de < 0 too?

de < 0 would be nasty, monotonic going backwards, so yeah.
-Chris