Quoting Andi Shyti (2019-12-07 17:42:39)
> Hi Chris,
>
> On Sat, Dec 07, 2019 at 01:08:35AM +0000, Chris Wilson wrote:
> > This reverts commit 343aae776a58a67fa153825385e6fe90e3185c5b.
> >
> > __for_each_physical_engine() reprograms the context, invalidating the
> > use of e->flags to select engines, necessitating e->index instead.
> > Withot also fixing up the engine selection, the result is that random
> > engines were being used to read registers from the intended engine.
> > This does not end well.
>
> So, the problem here is that with __for_each_physical_engine() we
> "reprogram the context" which means that there is a re-mapping of
> engines in it. Have I understood correctly?
>
> Doesn't that happen only in the case when the context has no
> engines in it?
iirc, it does something like init_engine_map and sets a consistent
engine layout on the context -- but switches execbuf over to index mode.
> > @@ -877,7 +876,8 @@ igt_main
> > igt_skip_on(gen > LAST_KNOWN_GEN);
> > }
> >
> > - __for_each_physical_engine(fd, e) {
> > + for (const struct intel_execution_engine2 *e = intel_execution_engines2;
> > + e->name; e++) {
>
> __for_each_static_engine() ?
I'm just doing a simple revert so we can try again :)
-Chris
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
