[Intel-gfx] ✗ Fi.CI.BAT: failure for drm/i915/execlists: Move the assertion we have the rpm wakeref down

[Chris Wilson]

Quoting Tvrtko Ursulin (2018-07-19 12:42:06)
> 
> On 19/07/2018 09:33, Patchwork wrote:
> > == Series Details ==
> > 
> > Series: drm/i915/execlists: Move the assertion we have the rpm wakeref down
> > URL   : https://patchwork.freedesktop.org/series/46837/
> > State : failure
> > 
> > == Summary ==
> > 
> > = CI Bug Log - changes from CI_DRM_4507 -> Patchwork_9713 =
> > 
> > == Summary - FAILURE ==
> > 
> >    Serious unknown changes coming with Patchwork_9713 absolutely need to be
> >    verified manually.
> >    
> >    If you think the reported changes have nothing to do with the changes
> >    introduced in Patchwork_9713, please notify your bug team to allow them
> >    to document this new failure mode, which will reduce false positives in CI.
> > 
> >    External URL: https://patchwork.freedesktop.org/api/1.0/series/46837/revisions/1/mbox/
> > 
> > == Possible new issues ==
> > 
> >    Here are the unknown changes that may have been introduced in Patchwork_9713:
> > 
> >    === IGT changes ===
> > 
> >      ==== Possible regressions ====
> > 
> >      igt at drv_selftest@live_evict:
> >        fi-cnl-psr:         NOTRUN -> DMESG-WARN +12
> > 
> >      igt at drv_selftest@live_workarounds:
> >        {fi-cfl-8109u}:     NOTRUN -> DMESG-FAIL
> >        fi-cnl-psr:         NOTRUN -> DMESG-FAIL
> 
> How come these failure are not in, copy & paste from end of email:

Postulating it's because fi-cnl-psr spontaneously reappeared and wasn't
in the baseline. Fwiw, it's all

[drm:intel_print_wm_latency [i915]] ERROR Gen9 Plane WM1 latency not provided

above and beyond the expected (reset recovery fails).
-Chris