[Bug 96428] New: [IVB bisected] [drm:intel_dp_aux_ch] *ERROR* dp aux hw did not signal timeout (has irq: 1)!

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=96428

            Bug ID: 96428
           Summary: [IVB bisected] [drm:intel_dp_aux_ch] *ERROR* dp aux hw
                    did not signal timeout (has irq: 1)!
           Product: DRI
           Version: DRI git
          Hardware: Other
                OS: All
            Status: NEW
          Severity: normal
          Priority: medium
         Component: DRM/Intel
          Assignee: intel-gfx-bugs at lists.freedesktop.org
          Reporter: chris.bainbridge at gmail.com
        QA Contact: intel-gfx-bugs at lists.freedesktop.org
                CC: intel-gfx-bugs at lists.freedesktop.org

Created attachment 124389
  --> https://bugs.freedesktop.org/attachment.cgi?id=124389&action=edit
0001-Revert-d1d70677e165-drm-i915-make-fbdev-initializati.patch

Boot error intermittently appears around 6% of the time (with external displays
and active DP adaptors):

[    1.447236] [drm:intel_dp_aux_ch] *ERROR* dp aux hw did not signal timeout
(has irq: 1)!

The error does not seem to cause any real problems as all displays still work
fine.

A lengthy and interesting bisect shows the cause of this is two interacting
commits, reverting either of these commits results in the error disappearing
entirely (auto revert of d1d706 on current git master is not clean, equivalent
patch attached):


commit 2ed903c5485bad0eafdd3d59ff993598736e4f31
Author: Chuansheng Liu <chuansheng.liu at intel.com>
Date:   Thu Sep 4 15:17:55 2014 +0800

    cpuidle: Use wake_up_all_idle_cpus() to wake up all idle cpus

    Currently kick_all_cpus_sync() or smp_call_function() can not
    break the polling idle cpu immediately.

    Instead using wake_up_all_idle_cpus() which can wake up the polling idle
    cpu quickly is much more helpful for power.


commit d1d70677e165826f3fa9966e1b7ec3765d7c0fb7
Author: Jesse Barnes <jbarnes at virtuousgeek.org>
Date:   Wed May 28 14:39:03 2014 -0700

    drm/i915: make fbdev initialization asynchronous v2

    This gets us out of our init code and out to userspace quite a bit
    faster, but does open us up to some bugs given the state of our init
    time locking.

    v2: switch to async_schedule (Chris)
        check with lockdep, seems happy (Jesse)
        move hotplug enable flag set to fbdev_initial_config (Jesse)


This bug might be related to bug #92685

-- 
You are receiving this mail because:
You are on the CC list for the bug.
You are the QA Contact for the bug.
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/intel-gfx-bugs/attachments/20160607/f3cbcdd2/attachment.html>