Bug 112282

Summary: [CI][DRMTIP]Random tests - dmesg-warn - *ERROR* Link Training Unsuccessful
Product: DRI Reporter: Lakshmi <lakshminarayana.vudum>
Component: DRM/IntelAssignee: Intel GFX Bugs mailing list <intel-gfx-bugs>
Status: RESOLVED MOVED QA Contact: Intel GFX Bugs mailing list <intel-gfx-bugs>
Severity: not set    
Priority: not set CC: intel-gfx-bugs, matthew.d.roper
Version: DRI git   
Hardware: Other   
OS: All   
Whiteboard:
i915 platform: SKL i915 features:

Description Lakshmi 2019-11-14 13:14:28 UTC
https://intel-gfx-ci.01.org/tree/drm-tip/drmtip_402/fi-skl-guc/igt@gem_eio@kms.html
<3> [205.576972] [drm:intel_dp_get_link_train_fallback_values [i915]] *ERROR* Link Training Unsuccessful
<3> [205.980474] [drm:intel_dp_get_link_train_fallback_values [i915]] *ERROR* Link Training Unsuccessful
<3> [206.569543] [drm:intel_dp_get_link_train_fallback_values [i915]] *ERROR* Link Training Unsuccessful
<3> [206.855115] [drm:intel_dp_get_link_train_fallback_values [i915]] *ERROR* Link Training Unsuccessful
<3> [207.031355] [drm:intel_dp_get_link_train_fallback_values [i915]] *ERROR* Link Training Unsuccessful
<3> [207.208430] [drm:intel_dp_get_link_train_fallback_values [i915]] *ERROR* Link Training Unsuccessful
Comment 1 CI Bug Log 2019-11-14 13:32:41 UTC
The CI Bug Log issue associated to this bug has been updated.

### New filters associated

* SKL: igt@gem_eio@kms - fail - *ERROR* Link Training Unsuccessful
  - https://intel-gfx-ci.01.org/tree/drm-tip/drmtip_402/fi-skl-guc/igt@gem_eio@kms.html
Comment 2 Matt Roper 2019-11-19 17:58:01 UTC
This looks like it's probably an issue with the physical link.  The system initially links trains successfully at 162000 with 2 lanes, but after running for a while that configuration fails and it has to downgrade to 1 lane.  Even with 1 lane there are a few sporadic failures and since we're already at the lowest rate and lowest lane count we get the error messages from the log here.

We may need to double check that the cable hasn't come loose on this machine and/or replace it with a new cable.  As far as I can see, this doesn't appear to be a software issue.
Comment 3 Tomi Sarvela 2019-11-20 08:03:31 UTC
SKL-GUC is running with two dummy displays: HDMI and VGA dongles. Failing DP probe is probably the VGA dongle, looking at the 1024x resolution.

As this VGA dongle is not a good representative of real hardware, we could consider this transient issue with this particular dongle and continue without it. Would this be good resolution for this issue?
Comment 4 Matt Roper 2019-11-20 16:09:23 UTC
(In reply to Tomi Sarvela from comment #3)
> SKL-GUC is running with two dummy displays: HDMI and VGA dongles. Failing DP
> probe is probably the VGA dongle, looking at the 1024x resolution.
> 
> As this VGA dongle is not a good representative of real hardware, we could
> consider this transient issue with this particular dongle and continue
> without it. Would this be good resolution for this issue?

Makes sense.  That sounds reasonable to me.
Comment 5 CI Bug Log 2019-11-21 08:35:27 UTC
A CI Bug Log filter associated to this bug has been updated:

{- SKL: igt@gem_eio@kms - fail - *ERROR* Link Training Unsuccessful -}
{+ SKL: Random tests - fail - *ERROR* Link Training Unsuccessful +}

New failures caught by the filter:

  * https://intel-gfx-ci.01.org/tree/drm-tip/drmtip_407/fi-skl-guc/igt@kms_plane_lowres@pipe-b-tiling-x.html
Comment 6 CI Bug Log 2019-11-27 12:11:45 UTC
A CI Bug Log filter associated to this bug has been updated:

{- SKL: Random tests - fail - *ERROR* Link Training Unsuccessful -}
{+ SKL: Random tests - fail - *ERROR* Link Training Unsuccessful +}

New failures caught by the filter:

  * https://intel-gfx-ci.01.org/tree/drm-tip/drmtip_408/fi-skl-guc/igt@kms_universal_plane@cursor-fb-leak-pipe-a.html
Comment 7 Martin Peres 2019-11-29 19:47:42 UTC
-- 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/603.

Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.