X crashes in FindGlyphByHash

In karmic X session exits, kills all applications that are running and returns to the login prompt.
Description: Ubuntu karmic (development branch)
Release: 9.10

Xorg.0.log.old gives:

Backtrace:
0: /usr/bin/X(xorg_backtrace+0x3b) [0x81353db]
1: /usr/bin/X(xf86SigHandler+0x55) [0x80c3185]
2: [0x718400]
3: /usr/bin/X(FindGlyphByHash+0x3c) [0x816f9dc]
4: /usr/bin/X [0x817c30c]
5: /usr/bin/X [0x8175455]
6: /usr/bin/X(Dispatch+0x33f) [0x808d67f]
7: /usr/bin/X(main+0x3bd) [0x80723dd]
8: /lib/tls/i686/cmov/libc.so.6(__libc_start_main+0xe5) [0xb2c7a5]
9: /usr/bin/X [0x8071891]
Saw signal 11. Server aborting.

I cannot find a way to reliably cause this crash. It has happen a second time now within 1 hour.

You're using a mixture of jaunty and karmic packages here, if you're going to use edgers change the sources to karmic. If you dont want to use xorg-edgers any longer, then remove it completely from your sources, and do a dpkg -l | grep sarvatt (or ~jaunty perhaps) and for every package it lists add /karmic to the end of apt-get (sudo apt-get install packagename/karmic packagename2/karmic for example)

After doing all Robert Hooker suggested, it still crashes.
Is this really Invalid ?

Summary of Xorg.0.log file
Backtrace:
0: /usr/bin/X(xorg_backtrace+0x3b) [0x8133a3b]
1: /usr/bin/X(xf86SigHandler+0x55) [0x80c7945]
2: [0x79a400]
3: /usr/bin/X(FindGlyphRef+0x11) [0x816daf1]
4: /usr/bin/X(FindGlyphByHash+0x3c) [0x816e1dc]
5: /usr/bin/X [0x817abfc]
6: /usr/bin/X [0x8173da5]
7: /usr/bin/X(Dispatch+0x347) [0x808ce27]
8: /usr/bin/X(main+0x395) [0x80723d5]
9: /lib/tls/i686/cmov/libc.so.6(__libc_start_main+0xe5) [0x1b97c5]
10: /usr/bin/X [0x8071881]
Saw signal 11. Server aborting.

Please collect a full backtrace on this issue - see http://wiki.ubuntu.com/X/Backtracing for guidance on how to collect this info.

We're closing this bug since it is has been some time with no response from the original reporter. However, if the issue still exists please feel free to reopen with the requested information. Also, if you could, please test against the latest development version of Ubuntu, since this confirms the bug is one we may be able to pass upstream for help.

same crash here with release version of karmic. Sporadic crash of xserver.

Backtrace:
0: /usr/bin/X(xorg_backtrace+0x3b) [0x8133d6b]
1: /usr/bin/X(xf86SigHandler+0x55) [0x80c7d35]
2: [0x508400]
3: /usr/bin/X(FindGlyphByHash+0x3c) [0x816e71c]
4: /usr/bin/X [0x817b13c]
5: /usr/bin/X [0x81742e5]
6: /usr/bin/X(Dispatch+0x35f) [0x808d17f]
7: /usr/bin/X(main+0x395) [0x8072515]
8: /lib/tls/i686/cmov/libc.so.6(__libc_start_main+0xe6) [0x801b56]
9: /usr/bin/X [0x80719c1]
Saw signal 11. Server aborting.

I applied dbg packages to my system. Hopefully the next crash will provide a better backtrace

Klaus, what applications and desktop environment were you using when it crashed?

I am using KDE desktop. The crash seems to be related to kmail. Everytime a crash occured I was working in kmail

Thanks, can you please attach your Xorg.0.log.old after a crash?

I just had another crash. I was displaying a message in kmail.

Please install the xserver-xorg-video-ati-dbg, xserver-xorg-video-radeon-dbg, xserver-xorg-core-dbg packages.

The last Xorg.0.log.old is from a system with the xserver-xorg-video-ati-dbg, xserver-xorg-video-radeon-dbg, xserver-xorg-core-dbg packages installed. I checked the https://wiki.ubuntu.com/X/Backtracing. Am I missing something?

You must be missing some other -dbg packages then, but I don't see which ones.

I have the same problem but I have Nvidia graphic card.
It happens working with QTCreator 1.2.1 opening a *.ui file after a full project recompilation.
My system is a fresh Ubuntu 9.10, graphic driver Nvidia 190 (but also happens with 185)

A piece of my /var/log/Xorg.0.log.old :

Backtrace:
0: /usr/bin/X(xorg_backtrace+0x3b) [0x8133d6b]
1: /usr/bin/X(xf86SigHandler+0x55) [0x80c7d35]
2: [0xa79400]
3: /usr/bin/X(FindGlyphByHash+0x3c) [0x816e71c]
4: /usr/bin/X [0x817b13c]
5: /usr/bin/X [0x81742e5]
6: /usr/bin/X(Dispatch+0x35f) [0x808d17f]
7: /usr/bin/X(main+0x395) [0x8072515]
8: /lib/tls/i686/cmov/libc.so.6(__libc_start_main+0xe6) [0x24db56]
9: /usr/bin/X [0x80719c1]
Saw signal 11. Server aborting.

after several weeks without a crash on my system, I just got a new crash

Almost the same here - however I only get the first three lines of the stack trace (and nothing at all in detail) - s. attachment.
I'm using a GeForce 6700 XL and a karmic that I auto-upgraded from jaunty.

Untested patch for xorg-server which should fix this issue. It exists as far back as 2007 it looks like. Sorry I do not have packages including it ready for testing at the moment but I will provide them in a PPA as soon as I can.

Taken from:
http://lists.x.org/archives/xorg-devel/2010-March/006436.html

I have uploaded a fixed xorg-server to this PPA for karmic. Can anyone experiencing the problem currently give it a shot and report your findings?

https://edge.launchpad.net/~sarvatt/+archive/bugfixes

Since the crash occurs sporadically it might be hard for people to definitively test and prove it solves the issue. But in examining the patch it looks perfectly sane to me; it is essentially just a check against dereferencing a null pointer. Upstream may prefer a different fix but this looks quite fine for Lucid. We can drop it if/when upstream has it in a release.

This bug was fixed in the package xorg-server - 2:1.7.6-1ubuntu1

---------------
xorg-server (2:1.7.6-1ubuntu1) lucid; urgency=low

  [Timo Aaltonen]
  * Merge from Debian unstable.
  * Drop patch 107, included in Debian.
  * Drop patch 108, included upstream.
  * control: Drop the udeb, we don't need it for lucid.
  * control: Relax/drop some build-deps caused by the udeb.

  [Bryce Harrington]
  * Add 110_findglyphbyhash-fix.patch from upstream to fix a sporadic
    crash in FindGlyphByHash.
    (LP: #401045)
  * Renumber patch 201_armel-drv-fallbacks.patch to 111

xorg-server (2:1.7.6-1) unstable; urgency=low

  [ Brice Goglin ]
  * Add 06_dont_trap_access_to_timer_and_keyboard.diff,
    thanks Samuel Thibault.

  [ Timo Aaltonen ]
  * New upstream release, closes: #574354.

  [ Julien Cristau ]
  * Don't build xserver-xorg-core-udeb on sparc. The linker fails because
    relocations have to be truncated when building statically against
    libnettle.

xorg-server (2:1.7.5.902-1) unstable; urgency=low

  [ Julien Cristau ]
  * config/udev: fix adding unnamed devices.
  * Build two flavours, one for the main package and one for the udeb.
  * Add patch to fix Xorg build with XV disabled.

  [ Timo Aaltonen ]
  * Add 16-config-dont-filter-input-subsys.diff so for instance serial
    wacom devices are initialized by the udev backend (LP: #522318,
    closes: #568236).

  [ Brice Goglin ]
  * New upstream release.
    + Allow for missing or disabled compat_output, closes: #572268, #554450.
    + Reenable RECORD extension, closes: #570680.
    + dix: try to ring the bell even if the current device doesn't have one,
      closes: #564200.
  * Refresh patches.
  * Fix typo in xvfb-run.1, thanks Joey Hess, closes: #527490.
  * Add 06_dont_trap_access_to_timer_and_keyboard.diff,
    thanks Samuel Thibault.

  [ Cyril Brulebois ]
  * Add udeb needed for the graphical installer: xserver-xorg-core-udeb.
  * Version/Bump some B-D to make sure xserver-xorg-core-udeb gets its
    dependencies on the (recently-added) udebs rather than on the
    libraries:
     - libpciaccess-dev
     - libudev-dev
     - libxau-dev
     - libxfont-dev
  * Replace 02_Add-libgcrypt-as-an-option-for-sha1.diff with
    02_Add-libgcrypt-and-libnettle-as-options-for-sha1.diff so that it's
    also possible to link against libnettle. Link (unconditionally)
    statically against libnettle.a to avoid an extra udeb for a few bytes.
  * Add nettle-dev to Build-Depends; and pass --with-sha1=libnettle for
    the udeb build (and --with-sha1=libgcrypt for the main build).
  * Add myself to Uploaders.
  * Tweak builderstring to use the name of the person responsible for the
    upload instead of an host-specific one (Closes: #574017).
 -- Bryce Harrington <email address hidden> Wed, 24 Mar 2010 12:04:20 -0700

I just added this ppa for the patch, if I ever get this problem again, I'll post it here

I just had the crash again...