[K8M800]with kernel 2.6.31-14-generic my X server crashes during KDE 4.3.2 startup, _mesa_reference_renderbuffer() assertion

Bug #471646 reported by jyio
82
This bug affects 13 people
Affects Status Importance Assigned to Milestone
openchrome
New
Unknown
Debian
Fix Released
Unknown
Gentoo Linux
New
Undecided
Unassigned
linux (Ubuntu)
Fix Released
High
Unassigned
mesa (Ubuntu)
Invalid
Undecided
Unassigned

Bug Description

I just upgraded my Jaunty machine running Linux 2.6.28-16-generic to Karmic running Linux 2.6.31-14-generic, and noticed that KDE crashes back to the KDM login screen during startup. I can run Fluxbox just fine, and KDE works if I boot into Linux 2.6.28-16-generic. I can also startx remotely over Xephyr/SSH. Locally, however, I get the following error in kdm.log (and output during startx):

=====[ CUT ]=====
X.Org X Server 1.6.4
Release Date: 2009-9-27
X Protocol Version 11, Revision 0
Build Operating System: Linux 2.6.24-23-server i686 Ubuntu
Current Operating System: Linux phoenix 2.6.31-14-generic #48-Ubuntu SMP Fri Oct 16 14:04:26 UTC 2009 i686
Kernel command line: root=UUID=c29b60cc-e6ef-450c-b3c5-ea761c6e3416 ro single
Build Date: 26 October 2009 05:15:02PM
xorg-server 2:1.6.4-2ubuntu4 (buildd@)
        Before reporting problems, check http://wiki.x.org
        to make sure that you have the latest version.
Markers: (--) probed, (**) from config file, (==) default setting,
        (++) from command line, (!!) notice, (II) informational,
        (WW) warning, (EE) error, (NI) not implemented, (??) unknown.
(==) Log file: "/var/log/Xorg.0.log", Time: Mon Nov 2 14:39:52 2009
(==) Using config file: "/etc/X11/xorg.conf"
Fulfilled via DRI at 20976640
Freed 20976640 (pool 2)
X: main/renderbuffer.c:2159: _mesa_reference_renderbuffer: Assertion `oldRb->Magic == 0xaabbccdd' failed.
xinit: connection to X server lost.
=====[ CUT ]=====

I'm using a VIA Technologies, Inc. K8M800/K8N800/K8N800A [S3 UniChrome Pro] (rev 01) video chipset with the xserver-xorg-video-openchrome driver, and the crash occurs whether I use an xorg.conf or not. The xorg.conf, if used, is

=====[ CUT ]=====
Section "Device"
        Identifier "Configured Video Device"
        Option "UseFBDev" "true"
EndSection

Section "Monitor"
        Identifier "Configured Monitor"
        Option "DPMS"
        HorizSync 31.5-65.2 # 31.5-37.9
EndSection

Section "Screen"
        Identifier "Default Screen"
        Monitor "Configured Monitor"
        Device "Configured Video Device"
        DefaultDepth 24
        SubSection "Display"
                Depth 24
                Modes "1280x1024"
        EndSubSection
EndSection
=====[ CUT ]=====

I'd appreciate any help with this matter, and advice about other information that I should provide. Thanks!

ProblemType: Bug
Architecture: i386
AudioDevicesInUse: Error: command ['fuser', '-v', '/dev/dsp', '/dev/snd/by-path', '/dev/snd/controlC0', '/dev/snd/pcmC0D0c', '/dev/snd/pcmC0D1c', '/dev/snd/pcmC0D1p', '/dev/snd/pcmC0D0p', '/dev/snd/seq', '/dev/snd/timer', '/dev/sequencer', '/dev/sequencer2'] failed with exit code 1:
CRDA: Error: [Errno 2] No such file or directory
Card0.Amixer.info:
 Card hw:0 'V8237'/'VIA 8237 with ALC655 at 0xd400, irq 22'
   Mixer name : 'Realtek ALC655 rev 0'
   Components : 'AC97a:414c4760'
   Controls : 48
   Simple ctrls : 29
Date: Mon Nov 2 15:29:22 2009
DistroRelease: Ubuntu 9.10
HibernationDevice: RESUME=UUID=f4ed7091-30d8-4c93-9e4b-e7ae7eb77d34
Package: linux-image-2.6.31-14-generic 2.6.31-14.48
ProcCmdLine: BOOT_IMAGE=/boot/vmlinuz-2.6.31-14-generic root=UUID=c29b60cc-e6ef-450c-b3c5-ea761c6e3416 ro quiet splash
ProcEnviron:
 LANGUAGE=
 LANG=en_US.UTF-8
 SHELL=/bin/bash
ProcVersionSignature: Ubuntu 2.6.31-14.48-generic
RelatedPackageVersions:
 linux-backports-modules-2.6.31-14-generic N/A
 linux-firmware 1.24
RfKill:

SourcePackage: linux
Uname: Linux 2.6.31-14-generic i686
WpaSupplicantLog:

XsessionErrors:
 (process:3009): Gtk-CRITICAL **: gtk_clipboard_get_for_display: assertion `display != NULL' failed
 (<unknown>:3009): Gdk-CRITICAL **: gdk_window_get_origin: assertion `GDK_IS_WINDOW (window)' failed
dmi.bios.date: 12/29/2006
dmi.bios.vendor: Phoenix Technologies, LTD
dmi.bios.version: 6.00 PG
dmi.board.name: K8M800-8237
dmi.chassis.type: 3
dmi.modalias: dmi:bvnPhoenixTechnologies,LTD:bvr6.00PG:bd12/29/2006:svn:pn:pvr:rvn:rnK8M800-8237:rvr:cvn:ct3:cvr:

Revision history for this message
jyio (inportb) wrote :
Revision history for this message
jyio (inportb) wrote :

I noticed that making the Xorg device option "NoAccel" "true" allows KDE to start, but screen updates are exceptionally slow. I got this idea from the OpenChrome Trac entry http://www.openchrome.org/trac/ticket/322

Changed in openchrome:
status: Unknown → New
summary: - with kernel 2.6.31-14-generic my X server crashes during KDE 4.3.2
- startup
+ [K8M800]with kernel 2.6.31-14-generic my X server crashes during KDE
+ 4.3.2 startup
Changed in xserver-xorg-video-openchrome (Ubuntu):
importance: Undecided → High
Changed in linux (Ubuntu):
importance: Undecided → High
Bryce Harrington (bryce)
tags: added: crash
Changed in xserver-xorg-video-openchrome (Ubuntu):
status: New → Confirmed
Revision history for this message
libv (libv) wrote : Re: [K8M800]with kernel 2.6.31-14-generic my X server crashes during KDE 4.3.2 startup

relevant line:
X: main/renderbuffer.c:2159: _mesa_reference_renderbuffer: Assertion `oldRb->Magic == 0xaabbccdd' failed.

Revision history for this message
Bartosz Kosiorek (gang65) wrote :

Attach your /var/log/Xorg.0.log file.
If you can from after reproducing this issue (or /var/log/Xorg.0.log.old).

Revision history for this message
Bartosz Kosiorek (gang65) wrote :
Changed in debian:
status: Unknown → New
Revision history for this message
Bartosz Kosiorek (gang65) wrote :

Try to disable DRI.

I created the patch which disable DRI (it use swrast instead)

Check if it works for you.
The detail build instruction for Ubuntu/Debian is available at:
https://help.ubuntu.com/community/OpenChrome

Attach your /var/log/Xorg.0.log file.

Revision history for this message
Bartosz Kosiorek (gang65) wrote :

You can also disable DRI by add to your xorg.conf file, following lines:
    Section "Module"
       ...
       Disable "dri"
       ...
    EndSection

Revision history for this message
Octavian Petre (octavsly) wrote :

Disabling DRI results in X being horribly slow, therefore I cannot use it on my VIA EPIA EN12000EG.

Best regards,
Octavian

Revision history for this message
jyio (inportb) wrote :

Would disabling DRI be similar to enabling NoAccel? Actually I cannot test this at the moment because I won't have physical access to the machine until Thanksgiving break or so, but I will let you know ASAP how this turns out. Thanks so much!

Revision history for this message
jyio (inportb) wrote :

Thanks for the suggestion. I have removed the "NoAccel" line and added the directive to disable DRI. I now have, in xorg.conf,

Section "Device"
        Identifier "Configured Video Device"
EndSection
...
Section "Module"
        Disable "dri"
EndSection

My Xorg.0.log, after successfully starting a KDE session, is attached. This is much more performant than "NoAccel", I guess because 2D rendering still works. By the way, is there any way to test these things remotely? I'd like to be able to start a desktop session and check the Xorg logs for errors.

Revision history for this message
jyio (inportb) wrote :

Incidentally, XRender works and I could get some basic (and slow) compositing. Well, now we're sure DRI's the culprit.

Changed in openchrome:
status: New → Unknown
Changed in openchrome:
status: Unknown → New
Revision history for this message
Bartosz Kosiorek (gang65) wrote :

Hi.
Try to enable DRI and disable AIGLX.

You can disable AIGLX and the compositing extension by amending your /etc/X11/xorg.conf like this:

Section "ServerFlags"
Option "AIGLX" "off"
EndSection

Section "Extensions"
Option "Composite" "Disable"
EndSection

Changed in xserver-xorg-video-openchrome (Ubuntu):
status: Confirmed → Triaged
assignee: nobody → Bartosz (gang65)
Revision history for this message
Bartosz Kosiorek (gang65) wrote :

Try also set 16 bit depth and check if it helps

Revision history for this message
Steve Cayford (cayfo001) wrote : Re: [Bug 471646] Re: [K8M800]with kernel 2.6.31-14-generic my X server crashes during KDE 4.3.2 startup

Bartosz wrote:
> Hi.
> Try to enable DRI and disable AIGLX.
>

This worked for me. I can log into KDE again.

Revision history for this message
Bartosz Kosiorek (gang65) wrote : Re: [K8M800]with kernel 2.6.31-14-generic my X server crashes during KDE 4.3.2 startup

Try to enable AIGLX and change color depth (for example to 16 bit, 24bit or to 32bit)

Is it helps?

Revision history for this message
Andrey Yurkovsky (andrey.yurkovsky) wrote :

I have the same problem with P4M890T-M. Changin color depth with enabled AIGLX had no effect.

http://www.openchrome.org/trac/ticket/340

Revision history for this message
jyio (inportb) wrote :

I re-enabled DRI and disabled AIGLX. I could still login. Thanks.

Revision history for this message
jyio (inportb) wrote :

As previously reported, changing the color depth with AIGLX enabled does not alleviate the problem.

However... I just installed KDE 4.4 beta 2 and can no longer reproduce this issue. At least, KDE just says that "compositing is not supported on your system" and does not try to enable compositing at all, which was the case before anyway. I'm not sure what would happen on a system that does support compositing. I know many things were changed between KDE versions, but if there's any other information I could provide, I'd be happy to do so.

Revision history for this message
Bartosz Kosiorek (gang65) wrote :

@jyio
Attach the ~/.xsession-errors file to this bug.

Is there any other information added to "compositing is not supported on your system"?

Revision history for this message
Daniel60 (webmaster-epanouissance) wrote :

When I boot the karmic Koala Desktop CD, the KDE splash screen loops.
During the first loop all icons appear. Then the screen turns black and the KDE splash screen starts over, displaying the first icon. From now on it loops endlessly: 1st icon - black screen - 1st icon ...
Probably the X server crashes again and again.
No response to keyboard or mouse except the following keys:
Ctrl-Alt-F1 (does not display text terminal #1 as expected, but weird blocks of color).
Ctrl-Alt-F7 returns me to the KDE splash screen loop"

I think KDE won't work with my computer 'standard' chipset which is Via s3g Unichrome IGP KM400/KN400/P4M800[S3 Unichrome]
Note that is ok with Gnome version.

Revision history for this message
Bartosz Kosiorek (gang65) wrote :

Try to turn on the Ubuntu Backports and upgrade KDE to 4.3.4

Here you have information how to enable backports:
http://www.addictivetips.com/ubuntu-linux-tips/how-to-enable-backports-in-ubuntu-linux/

Don't forget about attach your ~/.xsession-errors file.

Revision history for this message
Daniel60 (webmaster-epanouissance) wrote :

Sorry, but there is NO WAY to do such operations when booting on a CD LIVE engraved from the last .ISO file.
The display still returns on splash screen every 2 seconds and there is no possibility to access a console with the keyboard.
I whish this issue will be settled within Kubuntu 10.04.

Regards

Revision history for this message
Daniel60 (webmaster-epanouissance) wrote :

New info :
When starting with "graphic" mode I've successed to get this attached file :

Revision history for this message
Chris (csherrmann) wrote :

hello everyone,

i have the same problem with the kde splash screen appearing at installation bootup and then restarting again and again..... i have made it through the installation via the text base install and some weird xorg changes ( which i don't recall atm)... when running the computer now, the xorg.conf log files indicate glibc crashing at kde startup.... the xorg openchrome driver seems to be the culprit.

i have attached my verbose xorg output, sincerley hoping someone can fix this!
btw, I can start kde, if i disable composite in xorg.conf, so the error might be somewhere around there...

lastly, i'm able to run other window managers & kdm itself just fine with composite enabled in xorg.conf. however, i attribute that to the non usage of composition by my blackbox wm :)

anyway, i hope this is somewhat helpfull...

summary: [K8M800]with kernel 2.6.31-14-generic my X server crashes during KDE
- 4.3.2 startup
+ 4.3.2 startup, _mesa_reference_renderbuffer() assertion
Bryce Harrington (bryce)
tags: added: kubuntu
tags: added: review-request
Revision history for this message
vandyswa (ajv-cauriumbin) wrote :

Can confirm this happens on LTSP thin clients from NTA. AIGLX -> off, and Composite -> disable avoids the assertion failure. This is latest Karmic with corresponding LTSP /opt/ltsp/i386 tree. FYI on a thin client it simply freezes the client, requiring a power cycle. I was lucky enough to be able to add an RS-232 terminal w. getty to be able to see the X server's assertion failure.

Changed in debian:
status: New → Incomplete
Revision history for this message
Jeremy Foshee (jeremyfoshee) wrote :

Hi jyio,

If you could also please test the latest upstream kernel available that would be great. It will allow additional upstream developers to examine the issue. Refer to https://wiki.ubuntu.com/KernelMainlineBuilds . Once you've tested the upstream kernel, please remove the 'needs-upstream-testing' tag. This can be done by clicking on the yellow pencil icon next to the tag located at the bottom of the bug description and deleting the 'needs-upstream-testing' text. Please let us know your results.

Thanks in advance.

    [This is an automated message. Apologies if it has reached you inappropriately; please just reply to this message indicating so.]

tags: added: needs-upstream-testing
tags: added: kj-triage
Changed in linux (Ubuntu):
status: New → Incomplete
Ameen Arabi (ameenarabi)
Changed in linux (Ubuntu):
status: Incomplete → Fix Committed
Bryce Harrington (bryce)
Changed in mesa (Ubuntu):
status: New → Confirmed
Changed in linux (Ubuntu):
status: Fix Committed → Fix Released
bugbot (bugbot)
tags: added: karmic
Revision history for this message
Timo Aaltonen (tjaalton) wrote :

there is no DRI driver for openchrome in precise, since DRI1 drivers got purged upstream.

Changed in mesa (Ubuntu):
status: Confirmed → Invalid
Changed in debian:
status: Incomplete → Fix Released
Changed in xserver-xorg-video-openchrome (Ubuntu):
status: Triaged → Fix Released
no longer affects: xserver-xorg-video-openchrome (Ubuntu)
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.