Bug 95521 - DP unplug and undocking ThinkPad produces dmesg errors
Summary: DP unplug and undocking ThinkPad produces dmesg errors
Status: CLOSED WORKSFORME
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/Intel (show other bugs)
Version: unspecified
Hardware: x86-64 (AMD64) Linux (All)
: medium minor
Assignee: Intel GFX Bugs mailing list
QA Contact: Intel GFX Bugs mailing list
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2016-05-22 13:55 UTC by Martin Jørgensen
Modified: 2019-11-17 23:17 UTC (History)
6 users (show)

See Also:
i915 platform: HSW
i915 features: display/DP


Attachments
(almost) full dmesg (101.56 KB, text/plain)
2016-05-24 12:06 UTC, Martin Jørgensen
no flags Details
full dmesg for kernel 4.6 (97.17 KB, text/plain)
2016-06-14 10:45 UTC, Martin Jørgensen
no flags Details
full dmesg (98.42 KB, text/plain)
2016-06-27 12:36 UTC, Martin Jørgensen
no flags Details

Description Martin Jørgensen 2016-05-22 13:55:49 UTC
I'm running latest Debian Stretch on a ThinkPad X1 Carbon 20A7 "Linux mkjlap 4.5.0-2-amd64 #1 SMP Debian 4.5.3-2 (2016-05-08) x86_64 GNU/Linux"


I often get the following dmesg error when unplugging a DP cable or undocking from the dock:

i915 0000:00:02.0: BAR 6: [??? 0x00000000 flags 0x2] has bogus alignment

Sometimes its accompanied with a:

[drm:intel_mst_disable_dp [i915]] *ERROR* failed to update payload -22

The system seems to work properly mostly, but sometimes, when using a cable, my monitor doesn't seem to detect any visual and goes to standby, but xrandr claims that the monitor is enabled. To fix this I "--off" the monitor manually and "--auto" the monitor right after.
Comment 1 yann 2016-05-24 12:02:30 UTC
Can you attach full dmesg?
Comment 2 Martin Jørgensen 2016-05-24 12:06:40 UTC
Created attachment 124021 [details]
(almost) full dmesg
Comment 3 Kristijan Vrban 2016-06-02 15:45:47 UTC
+1 from me. Same issue here with Intel® HD Graphics 5500 (Broadwell GT2) (Lenovo X250) running Ubuntu 14.04 with Linux 4.4.12 from kernel.ubuntu.com/~kernel-ppa/mainline/
Comment 4 Martin Jørgensen 2016-06-14 10:44:41 UTC
Just upgraded to kernel 4.6, and two new messages new occurs:
[14633.660425] [drm:intel_set_cpu_fifo_underrun_reporting [i915]] *ERROR* uncleared fifo underrun on pipe B
[14633.660437] [drm:ironlake_irq_handler [i915]] *ERROR* CPU pipe B FIFO underrun

Full dmesg attached: dmesg-4.6
Comment 5 Martin Jørgensen 2016-06-14 10:45:09 UTC
Created attachment 124524 [details]
full dmesg for kernel 4.6
Comment 6 Dhinakaran Pandiyan 2016-06-24 00:42:38 UTC
Are you still having trouble with the monitor going to standby?
Comment 7 Martin Jørgensen 2016-06-24 05:47:26 UTC
(In reply to Dhinakaran Pandiyan from comment #6)
> Are you still having trouble with the monitor going to standby?

No I haven't got issues with any monitors going to standby. The initial issue still applies.
Comment 8 Martin Jørgensen 2016-06-24 06:25:54 UTC
(In reply to Dhinakaran Pandiyan from comment #6)
> Are you still having trouble with the monitor going to standby?

Well, I guess you can call it a standby issue :-) I just don't know where the hardware or software stack issue comes from.
Comment 9 Martin Jørgensen 2016-06-27 11:20:03 UTC
i've upgraded to 4.6.2 kernel and latest xf86-video-intel from git.
i still get the error below: 

[21419.173879] [drm:haswell_crtc_disable [i915]] *ERROR* failed to update payload -22

... but the rest seems to have disappeared, and monitor hotplugin seems to be working properly now. i will test more over the coming days.
Comment 10 Martin Jørgensen 2016-06-27 12:36:37 UTC
well, it doesn't :-) first issue found. after unplugin and plugin the laptop a few times through the day, I suddenly came across this issue in dmesg:

[25925.587641] [drm] HPD interrupt storm detected on connector DP-2: switching from hotplug detection to polling

it this point the monitor didn't show any image, and xrandr didn't report any other monitors than the eDP1. only after turning the monitor OFF and ON again, the monitor became visible in xrandr, and everything was working again.

attaching full dmesg later.
Comment 11 Martin Jørgensen 2016-06-27 12:36:57 UTC
Created attachment 124741 [details]
full dmesg
Comment 12 Martin Jørgensen 2016-06-30 11:48:12 UTC
... and right now it happened again. Image was turning on and off on monitor, and after I xrandr --output DP2-2 --off the machine froze for a few seconds and I got this in the dmesg:


[21714.659267] [drm:intel_set_cpu_fifo_underrun_reporting [i915]] *ERROR* uncleared fifo underrun on pipe B
[21714.659281] [drm:ironlake_irq_handler [i915]] *ERROR* CPU pipe B FIFO underrun
[21715.477955] [drm:intel_dp_link_training_clock_recovery [i915]] *ERROR* too many voltage retries, give up
[21716.181167] [drm:intel_dp_link_training_clock_recovery [i915]] *ERROR* too many voltage retries, give up
[21716.882401] [drm:intel_dp_link_training_clock_recovery [i915]] *ERROR* too many voltage retries, give up
[21717.581145] [drm:intel_dp_link_training_clock_recovery [i915]] *ERROR* too many voltage retries, give up
[21718.279735] [drm:intel_dp_link_training_clock_recovery [i915]] *ERROR* too many voltage retries, give up
[21718.978615] [drm:intel_dp_link_training_clock_recovery [i915]] *ERROR* too many voltage retries, give up
[21718.979114] [drm:intel_dp_start_link_train [i915]] *ERROR* failed to train DP, aborting
[21718.989251] [drm:intel_dp_start_link_train [i915]] *ERROR* Timed out waiting for DP idle patterns
Comment 13 Jani Saarinen 2016-12-09 10:57:12 UTC
Reporter, still issue seen?
Comment 14 Martin Jørgensen 2016-12-17 13:38:00 UTC
I haven't used the dock in a long time. Don't even know if I still got the dock .... :-/
Comment 15 Pierre-Yves David 2017-01-21 15:44:33 UTC
I've been having similar issue for multiple months, I've using a Think pad X260 with a dock and 2 monitor plugged to the dock through display port. The monitor fails to fails to wake up when get back to the laptop (and other strange behavior like X freezing for a couple of minute when I managed to get them detected again. I've seen many of the dmesg shown in this thread. Today I got:

  [Sat Jan 21 14:47:49 2017] [drm:intel_encoders_disable.isra.107 [i915]] *ERROR* failed to update payload -22
  [Sat Jan 21 14:47:49 2017] [drm:intel_encoders_disable.isra.107 [i915]] *ERROR* failed to update payload -22
  [Sat Jan 21 16:02:01 2017] [drm:intel_pipe_update_end [i915]] *ERROR* Atomic update failure on pipe A (start=292 end=293) time 127 us, min 763, max 767, scanline start 761, end 769
  [Sat Jan 21 16:15:46 2017] [drm:intel_pipe_update_end [i915]] *ERROR* Atomic update failure on pipe A (start=49791 end=49792) time 125 us, min 763, max 767, scanline start 761, end 769

I'm using debian unstable: kernel is" Linux 4.9.0-1-amd64 #1 SMP Debian 4.9.2-2 (2017-01-12) x86_64 GNU/Linux.
How can I help you debug the problem further?
Comment 16 Jani Saarinen 2017-03-08 07:50:46 UTC
Are you able to test if latest drm-tip helps on this at all?
Comment 17 Jani Saarinen 2017-05-22 09:45:38 UTC
Resolving as no comments received. If still issue please re-open.
Comment 18 Paul Menzel 2019-05-16 12:09:05 UTC
Martin, Pierre, is this happening only with MST (multi-stream transport) monitors?

I have the same problem with a Fujitsu desktop system and the MST HiDPI monitor Dell UP3214Q. Turning the monitor off, and trying to turn it back on does not work.

    [ 2295.887039] [drm:intel_mst_disable_dp [i915]] *ERROR* failed to update payload -22
    [12678.562632] [drm:intel_cpu_fifo_underrun_irq_handler [i915]] *ERROR* CPU pipe A FIFO underrun
Comment 19 Paul Menzel 2019-05-16 12:09:24 UTC
Just experienced this with Linux 5.1.2.
Comment 20 Martin Jørgensen 2019-06-01 07:52:59 UTC
(In reply to Paul Menzel from comment #18)
> Martin, Pierre, is this happening only with MST (multi-stream transport)
> monitors?
> 
> I have the same problem with a Fujitsu desktop system and the MST HiDPI
> monitor Dell UP3214Q. Turning the monitor off, and trying to turn it back on
> does not work.
> 
>     [ 2295.887039] [drm:intel_mst_disable_dp [i915]] *ERROR* failed to
> update payload -22
>     [12678.562632] [drm:intel_cpu_fifo_underrun_irq_handler [i915]] *ERROR*
> CPU pipe A FIFO underrun

Sorry haven't been able to test this much, since I've switched hardware in the meantime


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.