Bug 539303

Summary: Xorg ati radeon segmentation faults
Product: [Fedora] Fedora Reporter: Steve Holland <sh1>
Component: xorg-x11-drv-atiAssignee: Jérôme Glisse <jglisse>
Status: CLOSED WORKSFORME QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: medium Docs Contact:
Priority: low    
Version: 12CC: airlied, alan.ernhart, angavrilov, campbecg, clancy.kieran+redhat, vedran, xgl-maint
Target Milestone: ---Keywords: Triaged
Target Release: ---   
Hardware: All   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2010-05-24 20:32:12 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Attachments:
Description Flags
Xorg log file
none
xorg.conf none

Description Steve Holland 2009-11-19 20:25:55 UTC
Description of problem:
The X server fails with a segmentation fault at random times, dies, and
is restarted by GDM. 

Version-Release number of selected component (if applicable):
xorg-x11-drv-ati-6.13.0-0.10.20091006git457646d73.fc12.i686
xorg-x11-server-Xorg-1.7.1-7.fc12.i686

How reproducible:
Happens randomly, especially when viewing graphic files.

Steps to Reproduce:
1. Use computer, log in
  
Actual results:
Screen goes black and is replaced by the login screen

Expected results:
No X.org server crashes


Additional info:
IBM Thinkpad T60. nomodeset kernel option. 
01:00.0 VGA compatible controller: ATI Technologies Inc Radeon Mobility X1400 (prog-if 00 [VGA controller])

Backtrace:
0: /usr/bin/Xorg (xorg_backtrace+0x3c) [0x80e587c]
1: /usr/bin/Xorg (0x8047000+0x5fb66) [0x80a6b66]
2: (vdso) (__kernel_rt_sigreturn+0x0) [0x96940c]
3: /usr/lib/xorg/modules/drivers/radeon_drv.so (0x257000+0xcdf4a) [0x324f4a]
4: /usr/lib/xorg/modules/drivers/radeon_drv.so (0x257000+0x98a1a) [0x2efa1a]
5: /usr/lib/xorg/modules/libexa.so (0x21f000+0xc7c1) [0x22b7c1]
6: /usr/lib/xorg/modules/libexa.so (0x21f000+0xd36c) [0x22c36c]
7: /usr/bin/Xorg (0x8047000+0xd4a07) [0x811ba07]
8: /usr/bin/Xorg (CompositePicture+0x290) [0x810eef0]
9: /usr/bin/Xorg (miCompositeRects+0x290) [0x81b6710]
10: /usr/bin/Xorg (CompositeRects+0x74) [0x810ec44]
11: /usr/bin/Xorg (0x8047000+0xceb8d) [0x8115b8d]
12: /usr/bin/Xorg (0x8047000+0xcacb4) [0x8111cb4]
13: /usr/bin/Xorg (0x8047000+0x271f7) [0x806e1f7]
14: /usr/bin/Xorg (0x8047000+0x1b8c5) [0x80628c5]
15: /lib/libc.so.6 (__libc_start_main+0xe6) [0x7eabb6]
16: /usr/bin/Xorg (0x8047000+0x1b4b1) [0x80624b1]
Segmentation fault at address 0x14

Probably a NULL structure pointer dereference?

Comment 1 Chris Campbell 2009-11-20 01:00:33 UTC
Thanks for the bug report.  We have reviewed the information you have provided above, and there is some additional information we require that will be helpful in our diagnosis of this issue.

Please attach your X server config file (/etc/X11/xorg.conf, if available), output of the dmesg command, and X server log file (/var/log/Xorg.*.log) to the bug report as individual uncompressed file attachments using the bugzilla file attachment link below.

We will review this issue again once you've had a chance to attach this information.

Thanks in advance.

This bug has been triaged

-- 
Fedora Bugzappers volunteer triage team
https://fedoraproject.org/wiki/BugZappers

Comment 2 Steve Holland 2009-11-20 01:57:04 UTC
Created attachment 372399 [details]
Xorg log file

Comment 3 Steve Holland 2009-11-20 01:59:21 UTC
Created attachment 372400 [details]
xorg.conf

No xorg.conf was actually used. This is the output of Xorg -configure :1

Comment 4 Dave Airlie 2009-11-20 03:25:20 UTC
why are you using nomodeset? please don't (unless you've got a bug filed for it not working).

There is a fixed -ati package in koji, 6.13.0-0.11 or higher. please test and provide feedback.

Comment 5 Steve Holland 2009-11-20 05:49:14 UTC
I'm trying xorg-x11-drv-ati-6.13.0-0.12.20091119git437113124.fc12.i686.rpm. Will let you know how it turns out. This is a random occurrence bug so it may take a while to know for certain that it has been fixed. 

nomodeset is because of suspend/resume (Thinkpad T60). Resume just yields lots of flashing colors without it.

Comment 6 Alexander Gavrilov 2009-11-21 13:03:07 UTC
In my case it was so bad that I had to force the use of XAA via xorg.conf before I could even find this bug. However, the rpm mentioned above seems to have helped.

Comment 7 Dave Airlie 2009-11-24 09:31:20 UTC
btw T60 nomodeset s/r flashing should be fixed, please retest with a kernel > 140 from koji

Comment 8 Alan Ernhart 2009-11-24 15:14:42 UTC
I can vouch for the value of xorg-x11-drv-ati-6.13.0-0.12.20091119git437113124.fc12.i686.rpm over that which shipped in F12.

Thank you for the fix and the suggestion to try it!

Some details in case it's of value:

Running a Lenovo T60p, model 2007AD1 with ATI Technologies Inc M56GL [Mobility FireGL V5200] and external monitor.

I booted off the F12 KDE Live CD (final bits) and then installed from there to a partition, formatting it in the process. I did `yum update` to get current, and all seemed OK, so I installed some corporate packages. Somewhere in this, reboots hung *very* early in the process. I had earlier removed the quiet and rhgb options from grub.conf, so I could see the process and in the hang could confirm it was very early. Not having finding any solid clues, I tried the boot option nomodeset as the hang seemed very early when such stuff goes on. This allowed booting again.

I'm running KDE with only whatever gnome bits are in as dependencies for other packages. I noticed the random X crashes after I had most of this installed and in each case, X restarts and takes me back to GUI login. KDE Failsafe option does not prevent the crashes. After this restart, Xorg.0.log.old reports:

Backtrace:
0: /usr/bin/X (xorg_backtrace+0x3c) [0x80e587c]
1: /usr/bin/X (0x8047000+0x5fb66) [0x80a6b66]
2: (vdso) (__kernel_rt_sigreturn+0x0) [0xfe740c]
3: /usr/lib/xorg/modules/drivers/radeon_drv.so (0x38f000+0xcdf4a) [0x45cf4a]
4: /usr/lib/xorg/modules/drivers/radeon_drv.so (0x38f000+0x98a1a) [0x427a1a]
5: /usr/lib/xorg/modules/libexa.so (0x525000+0xc7c1) [0x5317c1]
6: /usr/lib/xorg/modules/libexa.so (0x525000+0xd36c) [0x53236c]
7: /usr/bin/X (0x8047000+0xd4a07) [0x811ba07]
8: /usr/bin/X (CompositePicture+0x290) [0x810eef0]
9: /usr/bin/X (0x8047000+0xce125) [0x8115125]
10: /usr/bin/X (0x8047000+0xcacb4) [0x8111cb4]
11: /usr/bin/X (0x8047000+0x271f7) [0x806e1f7]
12: /usr/bin/X (0x8047000+0x1b8c5) [0x80628c5]
13: /lib/libc.so.6 (__libc_start_main+0xe6) [0x5c1bb6]
14: /usr/bin/X (0x8047000+0x1b4b1) [0x80624b1]
Segmentation fault at address 0x14

Fatal server error:
Caught signal 11 (Segmentation fault). Server aborting

I all this I found steps certain to crash X: Start kate and click on the New document button.
NX Client also seems to trigger the crash shortly after connection to a (remote) system.

From http://koji.fedoraproject.org/koji/buildinfo?buildID=142121, I downloaded and updated to xorg-x11-drv-ati-6.13.0-0.12.20091119git437113124.fc12.i686.rpm.

This has fixed all my X crashes, including the reproducible cases. I've been working nearly a day now without issue, so can recommend this if you're experiencing such crashes in F12.

I'm still booting with nomodeset and might remove it later, but I don't have much use for bootup bling, preferring the old-school scroll of dmesg stuff (I've removed quiet and rhgb also).

Comment 9 Jérôme Glisse 2009-11-24 19:21:18 UTC
KMS is not only about graphical boot. With KMS kernel is now in charge and things works differently than with UMS. We are focussing on fixing KMS bugs. Initial reporter can you confirm that the ddx pointed out in comment #8 helps ?

Comment 10 Alan Ernhart 2009-11-25 12:04:26 UTC
> KMS is not only about graphical boot. With KMS kernel is now in charge and
> things works differently than with UMS. We are focussing on fixing KMS bugs.

Thanks for the info. I've now removed nomodeset and booted w/o error using:

kernel /boot/vmlinuz-2.6.31.5-127.fc12.i686 ro root=UUID=b74b4bad-aa7b-4c42-9fbc-6c67f1821da1 noiswmd LANG=en_US.UTF-8 SYSFONT=latarcyrheb-sun16 KEYBOARDTYPE=pc KEYTABLE=us
initrd /boot/initramfs-2.6.31.5-127.fc12.i686.img

Comment 11 Steve Holland 2010-01-06 16:09:08 UTC
Confirmation that xorg-x11-drv-ati-6.13.0-0.12.20091119git437113124.fc12.i686.rpm solves the problem. 

In addition  the KMS fix listed above works, but the following message is displayed after each resume: NMI received for unknown reason a1 on CPU 0.

Comment 12 Vedran Miletić 2010-05-24 20:32:12 UTC
Closing per comment 11.

---

Fedora Bugzappers volunteer triage team
https://fedoraproject.org/wiki/BugZappers

[This triage is part of collective effort done by students of University of
Rijeka Department of Informatics.]