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

### New filters associated

* ILK: suspend tests - dmesg-warn - i915->runtime_pm.wakeref_count=1 on cleanup
  - https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_5446/fi-ilk-m540/igt@kms_pipe_crc_basic@suspend-read-crc-pipe-a.html

* ILK: igt@runner@aborted - fail - Previous test: *(suspend|s3)*
  - https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_5446/fi-ilk-m540/igt@runner@aborted.html

This likely got introduced by the fix for https://bugs.freedesktop.org/show_bug.cgi?id=107588 .

First glance suggested a spurious warning, the code assumed it was running from a serial context and probably wasn't (i.e. the wakeref_count wasn't stable).

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

### New filters associated

* KBL: igt@pm_rpm@module-reload - dmesg-fail - i915->runtime_pm.wakeref_count=1 on cleanup
  - https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_5466/fi-kbl-7500u/igt@pm_rpm@module-reload.html
  - https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_5449_214/fi-kbl-7500u/igt@pm_rpm@module-reload.html

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

{- ILK: igt@runner@aborted - fail - Previous test: *(suspend|s3)* -}
{+ ILK KBL: igt@runner@aborted - fail - Previous test: *(suspend|s3|module-reload)* +}

New failures caught by the filter:

* https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_5466/fi-kbl-7500u/igt@runner@aborted.html
* https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_5449_214/fi-kbl-7500u/igt@runner@aborted.html

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

{- ILK: suspend tests - dmesg-warn - i915->runtime_pm.wakeref_count=1 on cleanup -}
{+ ILK APL: suspend tests - dmesg-warn - i915->runtime_pm.wakeref_count=1 on cleanup +}

New failures caught by the filter:

* https://intel-gfx-ci.01.org/tree/drm-tip/drmtip_200/fi-apl-guc/igt@kms_pipe_crc_basic@suspend-read-crc-pipe-a.html
* https://intel-gfx-ci.01.org/tree/drm-tip/drmtip_200/fi-apl-guc/igt@i915_suspend@sysfs-reader.html

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

{- ILK KBL: igt@runner@aborted - fail - Previous test: *(suspend|s3|module-reload)* -}
{+ ILK KBL APL: igt@runner@aborted - fail - Previous test: *(suspend|s3|module-reload|sysfs-reader)* +}

New failures caught by the filter:

* https://intel-gfx-ci.01.org/tree/drm-tip/drmtip_200/fi-apl-guc/igt@runner@aborted.html
* https://intel-gfx-ci.01.org/tree/drm-tip/drmtip_200/fi-apl-guc/igt@runner@aborted.html

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

{- ILK APL: suspend tests - dmesg-warn - i915->runtime_pm.wakeref_count=1 on cleanup -}
{+ ILK APL: suspend tests - dmesg-warn - i915->runtime_pm.wakeref_count=\d on cleanup +}

New failures caught by the filter:

* https://intel-gfx-ci.01.org/tree/drm-tip/drmtip_204/fi-apl-guc/igt@kms_cursor_crc@cursor-64x64-suspend.html

Steve, any updates on this bug?

I'm unable to reproduce this bug and the last sighting was on 2019-02-08.
If I understand Chris' Comment #3, this warning might be just the result of a race condition and so has negligible customer impact. 

I'm dropping the severity to medium and will continue to monitor. Please let me know if you disagree.

-don

commit 4547c255f4420e20c6cda2ee4172ae68b323e695
Author: Imre Deak <imre.deak@intel.com>
Date:   Thu May 9 20:34:36 2019 +0300

    drm/i915: Add support for tracking wakerefs w/o power-on guarantee
    
    It's useful to track runtime PM refs that don't guarantee a device
    power-on state to the rest of the driver. One such case is holding a
    reference that will be put asynchronously, during which normal users
    without their own reference shouldn't access the HW. A follow-up patch
    will add support for disabling display power domains asynchronously
    which needs this.
    
    For this we can split wakeref_count into a low half-word tracking
    all references (raw-wakerefs) and a high half-word tracking
    references guaranteeing a power-on state (wakelocks).
    
    Follow-up patches will make use of the API added here.
    
    While at it add the missing docbook header for the unchecked
    display-power and runtime_pm put functions.
    
    No functional changes, except for printing leaked raw-wakerefs
    and wakelocks separately in intel_runtime_pm_cleanup().
    
    v2:
    - Track raw wakerefs/wakelocks in the low/high half-word of
      wakeref_count, instead of adding a new counter. (Chris)
    v3:
    - Add a struct_member(T, m) helper instead of open-coding it. (Chris)
    - Checkpatch indentation formatting fix.
    
    Cc: Chris Wilson <chris@chris-wilson.co.uk>
    Signed-off-by: Imre Deak <imre.deak@intel.com>
    Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
    Link: https://patchwork.freedesktop.org/patch/msgid/20190509173446.31095-2-imre.deak@intel.com