Bug 111108

The CI Bug Log issue associated to this bug has been updated.

### New filters associated

* GVT-d: igt@i915_selftest@live_execlists - dmesg-fail - live_suppress_self_preempt failed with error -22
  - https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_13596/fi-skl-gvtdvm/igt@i915_selftest@live_execlists.html
  - https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_13597/fi-skl-gvtdvm/igt@i915_selftest@live_execlists.html
  - https://intel-gfx-ci.01.org/tree/drm-tip/IGTPW_3253/fi-skl-gvtdvm/igt@i915_selftest@live_execlists.html
  - https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_6450/fi-skl-gvtdvm/igt@i915_selftest@live_execlists.html
  - https://intel-gfx-ci.01.org/tree/drm-tip/Trybot_4587/fi-skl-gvtdvm/igt@i915_selftest@live_execlists.html
  - https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_6453/fi-skl-gvtdvm/igt@i915_selftest@live_execlists.html
  - https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_13613/fi-skl-gvtdvm/igt@i915_selftest@live_execlists.html

commit 506927ec8bbbdd99261cec4cd28cd0fd54e02218 (HEAD -> drm-intel-next-queued, drm-intel/for-linux-next, drm-intel/drm-intel-next-queued)
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date:   Fri Jul 12 09:25:49 2019 +0100

    drm/i915/selftests: Ignore self-preemption suppression under gvt
    
    GVT forces single port submission of individual requests. We do not
    enjoy the context amalgamation that the test depends upon for setting up
    the test (where port 0 has a large number of requests with a priority
    change somewhere in the middle). Under single request submission of gvt
    it is quite able for the preemption event to occur while another context
    is active and so there be a real need to act upon that preemption.
    
    Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=111108
    Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
    Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
    Cc: Zhenyu Wang <zhenyuw@linux.intel.com>
    Acked-by: Zhenyu Wang <zhenyuw@linux.intel.com>
    Link: https://patchwork.freedesktop.org/patch/msgid/20190712082549.25053-1-chris@chris-wilson.co.uk

(In reply to Chris Wilson from comment #2)
> commit 506927ec8bbbdd99261cec4cd28cd0fd54e02218 (HEAD ->
> drm-intel-next-queued, drm-intel/for-linux-next,
> drm-intel/drm-intel-next-queued)
> Author: Chris Wilson <chris@chris-wilson.co.uk>
> Date:   Fri Jul 12 09:25:49 2019 +0100
> 
>     drm/i915/selftests: Ignore self-preemption suppression under gvt
>     
>     GVT forces single port submission of individual requests. We do not
>     enjoy the context amalgamation that the test depends upon for setting up
>     the test (where port 0 has a large number of requests with a priority
>     change somewhere in the middle). Under single request submission of gvt
>     it is quite able for the preemption event to occur while another context
>     is active and so there be a real need to act upon that preemption.
>     
>     Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=111108
>     Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
>     Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
>     Cc: Zhenyu Wang <zhenyuw@linux.intel.com>
>     Acked-by: Zhenyu Wang <zhenyuw@linux.intel.com>
>     Link:
> https://patchwork.freedesktop.org/patch/msgid/20190712082549.25053-1-
> chris@chris-wilson.co.uk

This is the lastest failure which happened on CI_DRM_6651 (11 hours, 45 minutes old)
https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_6651/fi-skl-gvtdvm/igt@i915_selftest@live_execlists.html

A CI Bug Log filter associated to this bug has been updated:

{- GVT-d: igt@i915_selftest@live_execlists - dmesg-fail - live_suppress_self_preempt failed with error -22 -}
{+ GVT-d fi-icl-y: igt@i915_selftest@live_execlists - dmesg-fail - live_suppress_self_preempt failed with error -22 +}

New failures caught by the filter:

  * https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_6653/fi-icl-y/igt@i915_selftest@live_execlists.html

(In reply to CI Bug Log from comment #4)
> A CI Bug Log filter associated to this bug has been updated:
> 
> {- GVT-d: igt@i915_selftest@live_execlists - dmesg-fail -
> live_suppress_self_preempt failed with error -22 -}
> {+ GVT-d fi-icl-y: igt@i915_selftest@live_execlists - dmesg-fail -
> live_suppress_self_preempt failed with error -22 +}
> 
> New failures caught by the filter:
> 
>   *
> https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_6653/fi-icl-y/
> igt@i915_selftest@live_execlists.html

Also seen on fi-icl-y.

Hopefully,

commit acb9488dcad866987406aefc4f8298f2bf971575
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date:   Mon Aug 12 10:10:39 2019 +0100

    drm/i915/selftests: Prevent the timeslice expiring during suppression tests
    
    When testing whether we prevent suppressing preemption, it helps to
    avoid a time slice expiring prematurely.
    
    Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=111108
    Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
    Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
    Link: https://patchwork.freedesktop.org/patch/msgid/20190812091045.29587-2-chris@chris-wilson.co.uk

Nope!

*** Bug 111422 has been marked as a duplicate of this bug. ***

A CI Bug Log filter associated to this bug has been updated:

{- GVT-d fi-icl-y: igt@i915_selftest@live_execlists - dmesg-fail - live_suppress_self_preempt failed with error -22 -}
{+ KBL ICL GVT-d: igt@i915_selftest@live_execlists - dmesg-fail - live_suppress_self_preempt failed with error -22 +}

New failures caught by the filter:

  * https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_6811/fi-kbl-soraka/igt@i915_selftest@live_execlists.html

A CI Bug Log filter associated to this bug has been updated:

{- KBL ICL GVT-d: igt@i915_selftest@live_execlists - dmesg-fail - live_suppress_self_preempt failed with error -22 -}
{+ SKL KBL ICL GVT-d: igt@i915_selftest@live_execlists - dmesg-fail - live_suppress_self_preempt failed with error -22 +}


  No new failures caught with the new filter

Chris says the remaining issue is minor, lowering priority.

To be clear, the annoyance is high as it is a flip-flop in BAT. The impact is low, just an internal mechanics test that should not lead to any problems in execution (only less than ideal as we wakeup the tasklet for no reason). I still think it is an uncontrolled variable in the test, as opposed to a logic bug in scheduling; nothing to worry about. I've never trapped the bug locally to be able to dissect it, so I guess I need to think about what tracing I can put into the test for CI to solve the mystery.

-- GitLab Migration Automatic Message --

This bug has been migrated to freedesktop.org's GitLab instance and has been closed from further activity.

You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.freedesktop.org/drm/intel/issues/329.