linux-kernel - Video screen corruption after resume from suspend2ram

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [day] [month] [year] [list]

Message-ID: <54CD5B18.3060102@nod.at>
Date:	Sat, 31 Jan 2015 23:45:44 +0100
From:	Richard Weinberger <richard@....at>
To:	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
	linux-pm@...r.kernel.org, intel-gfx@...ts.freedesktop.org,
	dri-devel@...ts.freedesktop.org
CC:	airlied@...hat.com, daniel.vetter@...el.com
Subject: Video screen corruption after resume from suspend2ram

Hi!

I'm facing a strange issue on my Laptop (Lenovo T430, UEFI boot).
Sometimes after resuming from suspend2ram the system freezes after 0 to 2 minutes
and my screen corrupts as shown on this photo:
http://git.infradead.org/~rw/crash_suspend.jpg

The screen corruption always shows the same pattern, little read dots forming more or less
squares all across the screen.

Facts I know so far:

a) It happens with all kernels (tested with any kernel between 3.7 and Linus' as of today)

b) It happens only sometimes, every one or two weeks, I suspend my laptop about 2-10 times a day

c) It happens after I've upgraded my Laptop's memory from 4GiB to 16GiB. The ram is fine, so far it passed
all tests and I've never had any kind of crash/corruption except the suspend thing. And I actually
use all of it because I run many KVM guests on my Laptop, etc...

d) /var/log/messages contains a lot of 0x00 exactly at the time of the crash. Maybe the page cache faces
also an corruption, i.e.
---cut---
00000000  32 30 31 35 2d 30 31 2d  33 31 54 31 36 3a 34 32  |2015-01-31T16:42|
00000010  3a 34 37 2e 38 31 30 31  35 30 2b 30 31 3a 30 30  |:47.810150+01:00|
00000020  20 73 61 6e 64 70 75 70  70 79 20 63 6f 6c 6c 65  | sandpuppy colle|
00000030  63 74 64 5b 31 37 30 32  5d 3a 20 72 72 64 74 6f  |ctd[1702]: rrdto|
00000040  6f 6c 20 70 6c 75 67 69  6e 3a 20 72 72 64 5f 75  |ol plugin: rrd_u|
00000050  70 64 61 74 65 5f 72 20  28 73 61 6e 64 70 75 70  |pdate_r (sandpup|
00000060  70 79 2f 69 72 71 2f 69  72 71 2d 54 48 52 2e 72  |py/irq/irq-THR.r|
00000070  72 64 29 20 66 61 69 6c  65 64 3a 20 73 61 6e 64  |rd) failed: sand|
00000080  70 75 70 70 79 2f 69 72  71 2f 69 72 71 2d 54 48  |puppy/irq/irq-TH|
00000090  52 2e 72 72 64 3a 20 69  6c 6c 65 67 61 6c 20 61  |R.rrd: illegal a|
000000a0  74 74 65 6d 70 74 20 74  6f 20 75 70 64 61 74 65  |ttempt to update|
000000b0  20 75 73 69 6e 67 20 74  69 6d 65 20 31 34 32 32  | using time 1422|
000000c0  37 31 38 39 36 37 20 77  68 65 6e 20 6c 61 73 74  |718967 when last|
000000d0  20 75 70 64 61 74 65 20  74 69 6d 65 20 69 73 20  | update time is |
000000e0  31 34 32 32 37 31 38 39  36 37 20 28 6d 69 6e 69  |1422718967 (mini|
000000f0  6d 75 6d 20 6f 6e 65 20  73 65 63 6f 6e 64 20 73  |mum one second s|
00000100  74 65 70 29 0a 00 00 00  00 00 00 00 00 00 00 00  |tep)............|
00000110  00 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00  |................|
*
00000410  00 00 00 00 32 30 31 35  2d 30 31 2d 33 31 54 31  |....2015-01-31T1|
00000420  36 3a 34 35 3a 35 31 2e  38 31 32 30 35 30 2b 30  |6:45:51.812050+0|
00000430  31 3a 30 30 20 73 61 6e  64 70 75 70 70 79 20 72  |1:00 sandpuppy r|
00000440  73 79 73 6c 6f 67 64 3a  20 5b 6f 72 69 67 69 6e  |syslogd: [origin|
00000450  20 73 6f 66 74 77 61 72  65 3d 22 72 73 79 73 6c  | software="rsysl|
00000460  6f 67 64 22 20 73 77 56  65 72 73 69 6f 6e 3d 22  |ogd" swVersion="|
00000470  38 2e 34 2e 32 22 20 78  2d 70 69 64 3d 22 31 35  |8.4.2" x-pid="15|
00000480  36 38 22 20 78 2d 69 6e  66 6f 3d 22 68 74 74 70  |68" x-info="http|
00000490  3a 2f 2f 77 77 77 2e 72  73 79 73 6c 6f 67 2e 63  |://www.rsyslog.c|
000004a0  6f 6d 22 5d 20 73 74 61  72 74 0a                 |om"] start.|
000004ab
---cut---

The laptop crashed at 16:42 after resume, I had to power cycle it, at 16:45 it was up
again and rsyslog began logging. Between this log lines are 0x00s.

Can it be that something like that is hitting us again?

commit 3fa016a0b5c5237e9c387fc3249592b2cb5391c6
Author: Dave Airlie <airlied@...hat.com>
Date:   Wed Mar 28 10:48:49 2012 +0100

    drm/i915: suspend fbdev device around suspend/hibernate

    Looking at hibernate overwriting I though it looked like a cursor,
    so I tracked down this missing piece to stop the cursor blink
    timer. I've no idea if this is sufficient to fix the hibernate
    problems people are seeing, but please test it.

    Both radeon and nouveau have done this for a long time.

    I've run this personally all night hib/resume cycles with no fails.

dmesg and lspci output are attached.
Attached suspend.log is the content of /var/log/messages between suspend -> resume and crash.

Thanks,
//richard


View attachment "dmesg_good.txt" of type "text/plain" (78799 bytes)

View attachment "lspci-v.txt" of type "text/plain" (6268 bytes)

View attachment "suspend.log" of type "text/x-log" (20844 bytes)