[Intel-gfx] [PATCH] drm/i915/selftests: Always use an active engine while resetting

[Chris Wilson]

Quoting Mika Kuoppala (2019-02-15 09:09:46)
> Chris Wilson <chris at chris-wilson.co.uk> writes:
> 
> > Currently, we only try to reset a live engine for checking the whitelist
> > retention across a per-engine reset. For safety, it appears we need to
> > prime the system with a hanging spinner before performing a full-device
> > reset. (Figuring out the root cause behind the instability with handling
> > a reset during a no-op request is a challenge for another test, the
> > whitelist test has its own purpose.)
> >
> 
> Agreed on the sentiment and it does what it says on the tin.
> Reviewed-by: Mika Kuoppala <mika.kuoppala at linux.intel.com>

Fwiw, I'm working on a live_hangcheck pass to try and hit the same issue
(although CI hints that we may have fixed it already with "drm/i915:
Only try to park engines after a failed reset")
-Chris