FS#36107 - [linux] 3.10.x corruption/graphical artifacts on resume (intel IVB)

Attached to Project: Arch Linux
Opened by Brandon Watkins (bwat47) - Thursday, 11 July 2013, 19:26 GMT
Last edited by Tobias Powalowski (tpowa) - Thursday, 10 October 2013, 10:38 GMT
Task Type Bug Report
Category Packages: Testing
Status Closed
Assigned To Tobias Powalowski (tpowa)
Thomas Bächler (brain0)
Architecture x86_64
Severity Medium
Priority Normal
Reported Version
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 11
Private No

Details

Upon resuming from suspend on intel HD4000 with kernel 3.10 from testing, the screen is just a mess of colors and artifacts.

Upstream discussion of the bug: http://www.gossamer-threads.com/lists/linux/kernel/1743038

I also noticed other times suspend will just seemingly totally hang, just a black screen, no blinking cursor or anything (don't have either of these issues with 3.9.9) not sure if its related. But right now kernel 3.10 is severely broken when it comes to suspend/resume on intel hd4000.
This task depends upon

Closed by  Tobias Powalowski (tpowa)
Thursday, 10 October 2013, 10:38 GMT
Reason for closing:  Fixed
Comment by Brandon Watkins (bwat47) - Thursday, 11 July 2013, 19:30 GMT
and here's the bugzilla report for the graphical corruption: https://bugzilla.kernel.org/show_bug.cgi?id=60530
Comment by Andrea Fagiani (Hador) - Sunday, 21 July 2013, 08:58 GMT
And this is the patch submitted to intel-drm-fixes : http://cgit.freedesktop.org/~danvet/drm-intel/commit/?id=94a335dba34ff47cad3d6d0c29b452d43a1be3c8 .
Apparently, it won't make it to 3.10.2 (you can see the patch series here https://lkml.org/lkml/2013/7/19/147); it might be a good idea to apply it in the PKGBUILD though?
Comment by Bjoern Bidar (Thaodan) - Sunday, 21 July 2013, 13:54 GMT
there's a patch commited that fixes the issue see here: https://bugzilla.kernel.org/attachment.cgi?id=106910&action=diff
Comment by John (graysky) - Monday, 22 July 2013, 21:35 GMT
@Thaodan - Is this[1] the patch you're using and does it resolve the issue for you? Also, do you see anything odd in your dmesg when running this patch?

1. http://cgit.freedesktop.org/~danvet/drm-intel/commit/?id=94a335dba34ff47cad3d6d0c29b452d43a1be3c8
Comment by Bjoern Bidar (Thaodan) - Monday, 22 July 2013, 21:49 GMT
Yes it fixes the issue for my (Thinkpad Edge E530, Intel Core 3rd gen HD4000). I don't see anything odd in kernel output (using Linux wit pf patchset 3.10.1)
Comment by John (graysky) - Monday, 22 July 2013, 21:52 GMT
OK, so I applied it to linux built from ABS (3.10.2-1-ARCH) as a vanilla test and found two odd entries in my dmesg shown below. i7-3770k using onboard graphics.

[ 7114.087816] [drm:intel_dp_complete_link_train] *ERROR* failed to train DP, aborting
[ 7174.521130] [drm:intel_dp_complete_link_train] *ERROR* failed to train DP, aborting
Comment by Bjoern Bidar (Thaodan) - Monday, 22 July 2013, 21:57 GMT
Hmm, I'll look again at my notebook when I get it back from warranty support.
Comment by Bjoern Bidar (Thaodan) - Monday, 22 July 2013, 21:58 GMT
When do you got it after resume or from fresh boot?
Comment by John (graysky) - Monday, 22 July 2013, 22:04 GMT
Nothing on a fresh boot, but I get the error I posted above consistently when I wake up the monitor after it enters power savings mode. For example, the most recent entry around 11591 sec:

% dmesg | grep intel_dp
[ 7114.087816] [drm:intel_dp_complete_link_train] *ERROR* failed to train DP, aborting
[ 7174.521130] [drm:intel_dp_complete_link_train] *ERROR* failed to train DP, aborting
[11591.343093] [drm:intel_dp_complete_link_train] *ERROR* failed to train DP, aborting

My setup is little unique in that I am using a display port connection. No idea if it is related or not.
Comment by John (graysky) - Monday, 22 July 2013, 22:09 GMT
From a fresh boot, I get nothing; I disabled suspend from the kernel at compilation[1] to keep my wife from accidentally suspending my workstation so no idea :p

1. Power management and ACPI options ---> Suspend to RAM and standby (unchecked) and Power management and ACPI options ---> Hibernation (aka 'suspend to disk') (unchecked)
Comment by Bjoern Bidar (Thaodan) - Monday, 22 July 2013, 22:17 GMT
Hmm, it works for me. I don't know if I get any messsage currently. I think I can check it in about a week.
Comment by John (graysky) - Monday, 22 July 2013, 23:54 GMT
It is not the patch. I removed the patch and compiled an otherwise identical 3.10.2-1-ARCH from ABS: same messages in dmesg. Sometimes it pays to run the positive control ;)
Comment by Bjoern Bidar (Thaodan) - Thursday, 01 August 2013, 21:42 GMT
The patch has been queued see: https://git.kernel.org/cgit/linux/kernel/git/stable/stable-queue.git/tree/queue-3.10/
Comment by Tobias Powalowski (tpowa) - Tuesday, 17 September 2013, 08:02 GMT
Status on 3.11.1?

Loading...