Bug 62791 - [ILK regression] Change from X-window to text terminal causes Call Trace
Summary: [ILK regression] Change from X-window to text terminal causes Call Trace
Status: CLOSED DUPLICATE of bug 54687
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/Intel (show other bugs)
Version: unspecified
Hardware: Other All
: medium major
Assignee: Intel GFX Bugs mailing list
QA Contact: Intel GFX Bugs mailing list
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2013-03-27 06:07 UTC by cancan,feng
Modified: 2016-10-28 09:55 UTC (History)
1 user (show)

See Also:
i915 platform:
i915 features:


Attachments
dmesg (122.60 KB, text/plain)
2013-03-27 06:07 UTC, cancan,feng
no flags Details

Description cancan,feng 2013-03-27 06:07:36 UTC
Created attachment 77084 [details]
dmesg

Environment:
--------------------
The Linux Kernel, the operating system core itself
Kernel: (drm-intel-next-queued)e3dff585508636c8d2915cc1595e04f16ccd66ba
Some additional commit info:
Author: Ben Widawsky <ben@bwidawsk.net>
Date:   Wed Mar 20 14:49:14 2013 -0700

    drm/i915: Implement WaSwitchSolVfFArbitrationPriority

Bug Details:
---------------------
Call Trace appears while doing chvt on Ironlake. After booting system, start X, then change to text terminal with chvt, Call Trace will be found. This issue only happens after starting X. I attached dmesg log which you can find in the attachment.

Reproduce steps:
---------------------
1.boot or reboot
2.xinit &
3.chvt 1 --> Call Trace
Comment 1 cancan,feng 2013-03-27 08:19:35 UTC
Additional information:
-----------------------------
Platform: Calpella

This issue happens when eDP and DP both plugged in, but it will go away if unplug DP. I never tested this with DP and eDP both plugged in before. So I'm sorry I can't find a good commit though I have back to kernel 3.8..
Comment 2 Daniel Vetter 2013-03-27 10:23:46 UTC
There seem to be two things: The WARN about pipe_wait_off timeout is probably old and a dupe of bug #54687.

But the link train failure is a bit more ugly. Can you please test whether reverting

commit 3b4f819d5eac94ba8fe5e8c061f6dabfe8d7b22c
Author: Takashi Iwai <tiwai@suse.de>
Date:   Mon Mar 11 18:40:16 2013 +0100

    Revert "drm/i915: try to train DP even harder"

improves the situation?
Comment 3 cancan,feng 2013-03-28 02:19:17 UTC
(In reply to comment #2)
> There seem to be two things: The WARN about pipe_wait_off timeout is
> probably old and a dupe of bug #54687.

But the link train failure is a bit
> more ugly. Can you please test whether reverting

commit
> 3b4f819d5eac94ba8fe5e8c061f6dabfe8d7b22c
Author: Takashi Iwai
> <tiwai@suse.de>
Date:   Mon Mar 11 18:40:16 2013 +0100

    Revert
> "drm/i915: try to train DP even harder"

improves the situation?

I test what you said, but it doesn't improve the situation, Call Trace still there. :'( 
If you need any special information, just to let me know.
Comment 4 Daniel Vetter 2013-03-28 09:52:51 UTC
Just to check: Is there anything broken (DP screen not working) or is the issue just the WARN backtrace?
Comment 5 cancan,feng 2013-03-29 03:16:47 UTC
(In reply to comment #4)
> Just to check: Is there anything broken (DP screen not working) or is the
> issue just the WARN backtrace?

After call trace, I can chvt to X window, then do glxgears, then chvt to text terminal, all of these seem no problem but call trace. So is this just a  WARN backtrace?
Comment 6 Chris Wilson 2013-06-12 14:39:37 UTC
If the training failure is just a red herring, let's just treat this as a simple dupe.

*** This bug has been marked as a duplicate of bug 54686 ***
Comment 7 Milan Bouchet-Valat 2013-06-12 15:55:46 UTC
Sorry, Chris, but I really don't think you intended to mark this bug as a duplicate of bug 54686. Or is there something with percent signs in LibreOffice that the Intel driver does not like? ;-)
Comment 8 Milan Bouchet-Valat 2013-06-12 15:56:18 UTC

*** This bug has been marked as a duplicate of bug 54687 ***
Comment 9 Jari Tahvanainen 2016-10-28 09:55:17 UTC
Closing verified+duplicate as duplicate of closed+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.