Bug 84310 - [BSW] Link training error (black screen) while executing xrandr rotation
Summary: [BSW] Link training error (black screen) while executing xrandr rotation
Status: CLOSED FIXED
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/Intel (show other bugs)
Version: unspecified
Hardware: Other All
: medium normal
Assignee: Intel GFX Bugs mailing list
QA Contact: Intel GFX Bugs mailing list
URL:
Whiteboard:
Keywords:
: 82236 83873 (view as bug list)
Depends on:
Blocks:
 
Reported: 2014-09-25 06:48 UTC by Guo Jinxian
Modified: 2016-09-28 13:38 UTC (History)
2 users (show)

See Also:
i915 platform:
i915 features:


Attachments
dmesg (123.82 KB, text/plain)
2014-09-25 06:48 UTC, Guo Jinxian
no flags Details
left (123.06 KB, text/plain)
2014-10-13 09:12 UTC, Guo Jinxian
no flags Details
xrandr --output eDP1 --rotation normal (123.07 KB, text/plain)
2014-10-13 09:13 UTC, Guo Jinxian
no flags Details

Description Guo Jinxian 2014-09-25 06:48:19 UTC
Created attachment 106831 [details]
dmesg

==System Environment==
--------------------------
Regression: No.
I don't find good commit on kernel

Non-working platforms: BDW

==kernel==
--------------------------
origin/drm-intel-nightly: 0f7cc12c94e3a3033a46ce41bed55e8b6b35561b(fails)
    drm-intel-nightly: 2014y-09m-24d-19h-04m-51s UTC integration manifest
origin/drm-intel-next-queued: 7ca5a41f4da201371e131fc0641033652f76bf30(fails)
    drm/i915: Avoid re-configure panel on every PSR re-enable.
origin/drm-intel-fixes: c84db77010877da6c5da119868ed54c43d59e726(fails)
    drm/i915/hdmi: fix hdmi audio state readout

==Bug detailed description==
-----------------------------
Black screen shows while executing xrandr rotation.
I only find the issue on one BDW device.

Hardware environment:
root@x-bdw05:~# lspci -nnn
00:00.0 Host bridge [0600]: Intel Corporation Device [8086:1604] (rev 08)
00:02.0 VGA compatible controller [0300]: Intel Corporation Device [8086:1616] (rev 08)
00:03.0 Audio device [0403]: Intel Corporation Device [8086:160c] (rev 08)
00:14.0 USB controller [0c03]: Intel Corporation Device [8086:9cb1] (rev 03)
00:16.0 Communication controller [0780]: Intel Corporation Device [8086:9cba] (rev 03)
00:19.0 Ethernet controller [0200]: Intel Corporation Ethernet Connection (3) I218-LM [8086:15a2] (rev 03)
00:1f.0 ISA bridge [0601]: Intel Corporation Device [8086:9cc3] (rev 03)
00:1f.2 SATA controller [0106]: Intel Corporation Device [8086:9c83] (rev 03)
00:1f.3 SMBus [0c05]: Intel Corporation Device [8086:9ca2] (rev 03)
00:1f.6 Signal processing controller [1180]: Intel Corporation Device [8086:9ca4] (rev 03)



==Reproduce steps==
---------------------------- 
1. xinit &
2. xrandr --output eDP1 --rotation left
Comment 1 Chris Wilson 2014-09-25 07:05:47 UTC
Just a random:

[   60.264167] [drm:intel_dp_set_signal_levels] Using signal levels 00000000
[   60.270946] [drm:drm_dp_dpcd_access] too many retries, giving up
[   60.270947] [drm:intel_dp_start_link_train] *ERROR* failed to enable link training
[   60.275600] [drm:intel_dp_set_signal_levels] Using signal levels 00000000
[   60.282323] [drm:drm_dp_dpcd_access] too many retries, giving up
[   60.282324] [drm:intel_dp_complete_link_train] *ERROR* failed to update link training

and no further mode changes after that to reset the screen. (Failure on our part to propagate the error and a failure to never try again.)

If it is deterministic, that would be very useful for debugging.
Comment 2 Rodrigo Vivi 2014-09-25 21:14:28 UTC
*** Bug 82236 has been marked as a duplicate of this bug. ***
Comment 3 Rodrigo Vivi 2014-10-08 23:45:25 UTC
*** Bug 83873 has been marked as a duplicate of this bug. ***
Comment 4 Rodrigo Vivi 2014-10-08 23:53:12 UTC
Hi Guo,

Rotation is working fine on my bdw here with recent -nightly. Could you please verify if it is still happening on your side?
Comment 5 Guo Jinxian 2014-10-13 02:51:07 UTC
(In reply to Rodrigo Vivi from comment #4)
> Hi Guo,
> 
> Rotation is working fine on my bdw here with recent -nightly. Could you
> please verify if it is still happening on your side?

The failure still able to reproduce on latest -nightly(ead7d71357628dc1231535886ec09853e9164d6d).

Only "xrandr --output eDP1 --rotation normal" works well. Rotation left, rotation right and rotation inverted all unable to work.
Comment 6 Chris Wilson 2014-10-13 06:34:12 UTC
(In reply to Guo Jinxian from comment #5)
> (In reply to Rodrigo Vivi from comment #4)
> > Hi Guo,
> > 
> > Rotation is working fine on my bdw here with recent -nightly. Could you
> > please verify if it is still happening on your side?
> 
> The failure still able to reproduce on latest
> -nightly(ead7d71357628dc1231535886ec09853e9164d6d).
> 
> Only "xrandr --output eDP1 --rotation normal" works well. Rotation left,
> rotation right and rotation inverted all unable to work.

Could you update the dmesg? In particular does:

xrandr --output eDP1 --rotation left
xrandr --output eDP1 --rotation normal

work?
Comment 7 Guo Jinxian 2014-10-13 09:12:45 UTC
Created attachment 107765 [details]
left

(In reply to Chris Wilson from comment #6)
> (In reply to Guo Jinxian from comment #5)
> > (In reply to Rodrigo Vivi from comment #4)
> > > Hi Guo,
> > > 
> > > Rotation is working fine on my bdw here with recent -nightly. Could you
> > > please verify if it is still happening on your side?
> > 
> > The failure still able to reproduce on latest
> > -nightly(ead7d71357628dc1231535886ec09853e9164d6d).
> > 
> > Only "xrandr --output eDP1 --rotation normal" works well. Rotation left,
> > rotation right and rotation inverted all unable to work.
> 
> Could you update the dmesg? In particular does:
> 
> xrandr --output eDP1 --rotation left
> xrandr --output eDP1 --rotation normal
> 
> work?

Please check the dmesg while rotate left.
Comment 8 Guo Jinxian 2014-10-13 09:13:07 UTC
Created attachment 107766 [details]
xrandr --output eDP1 --rotation normal
Comment 9 zhaodan 2014-10-28 01:09:50 UTC
I can reproduce the issue on the SNB and use the kernel on 3.18.0-rc1_drm-intel-testing_a08ea5.
reproduce step:
xrandr --output eDP2 --rotation inverted
Comment 10 jinliangx.wang 2014-12-26 01:22:45 UTC
This bug also appear on BSW, when change rotation normal or inverted, have the chance to be blank screen.

kernel version: drm-intel-testing c3e107ab06274aa47f082df40ee4365e480e7b2a
test steps:
    xrandr --output eDP1 --rotation normal
    xrandr --output eDP1 --rotation inverted
Comment 11 Chris Wilson 2014-12-26 11:15:50 UTC
(In reply to jinliangx.wang from comment #10)
> This bug also appear on BSW, when change rotation normal or inverted, have
> the chance to be blank screen.
> 
> kernel version: drm-intel-testing c3e107ab06274aa47f082df40ee4365e480e7b2a
> test steps:
>     xrandr --output eDP1 --rotation normal
>     xrandr --output eDP1 --rotation inverted

inverted is likely to be a separate bug (a more recent kernel regression), see bug 86548.
Comment 12 Chris Wilson 2014-12-27 09:50:54 UTC
This bug should only be about the link training error in the original report. The common failure is bug 86548.
Comment 13 Jani Nikula 2015-01-30 07:08:58 UTC
Please retest with current drm-intel-nightly and http://patchwork.freedesktop.org/patch/41360
Comment 14 Ding Heng 2015-02-02 02:45:48 UTC
I had reproduce this issue with nightly branch(2014-09-24), and it is unable to reproduce on latest nightly branch. Change state to verified.
Comment 15 Jari Tahvanainen 2016-09-28 13:38:16 UTC
Closing verified+fixed.


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.