Quoting Tvrtko Ursulin (2019-01-10 09:46:56)
>
> On 10/01/2019 01:32, Daniele Ceraolo Spurio wrote:
> > -static bool verify_gt_engine_wa(struct drm_i915_private *i915, const char *str)
> > +static bool verify_gt_engine_wa(struct drm_i915_private *i915,
> > + struct wa_lists *lists, const char *str)
> > {
> > struct intel_engine_cs *engine;
> > enum intel_engine_id id;
> > bool ok = true;
> >
> > - ok &= intel_gt_verify_workarounds(i915, str);
> > + ok &= wa_list_verify(i915, &lists->gt_wa_list, str);
> > +
> > for_each_engine(engine, i915, id)
> > - ok &= intel_engine_verify_workarounds(engine, str);
> > + ok &= wa_list_verify(i915, &lists->engine[id].wa_list, str);
>
> Any point in checking that the two copies of each list also match?
> (Separate subtest, for easy problem detection.) Say rename
> wa_list_verify to wa_list_verify_mmio and add a new wa_list_compare
> which would compare the two lists.
The same thought crossed my mind, as it gives us a sanitycheck useful if
the readback fails.
-Chris
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
