Re: [igt-dev] [PATCH igt 5/5] igt/gem_exec_fence: Exercise merging fences

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Quoting Antonio Argenziano (2018-02-28 22:44:31)
> On 28/02/18 07:51, Chris Wilson wrote:
> > +     igt_assert(!gem_bo_busy(fd, obj.handle));
> > +     igt_assert_eq(sync_fence_status(all),
> > +                   flags & HANG ? -EIO : SYNC_FENCE_OK);
> 
> Do you get -EIO also if only one engine hangs?

Yes. If the hang occurs on the fence (single request/engine, or many
requests/engines merged sync_file) we expect to be notified with -EIO.
That part of the contract is checked in gem_eio and should be checked in
gem_reset_stats (pretty much all that test should be doing, injecting a
hang on each ring, observing fence states and reset-stats for itself and
concurrent contexts).
-Chris
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/intel-gfx




[Index of Archives]     [Linux USB Devel]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]
  Powered by Linux