compiz.real crashed with SIGSEGV in glXCreateContext()

StacktraceTop:?? () from /usr/lib/libGL.so.1
?? () from /usr/lib/libGL.so.1
glXCreateContext () from /usr/lib/libGL.so.1
addScreen (display=0x95da4a8, screenNum=0,
addDisplay (name=0x0) at ../../src/display.c:2373

Martin - this bug report doesn't have a NonfreeKernelModules: bit in the description yet "fglrx" does show up in ProcModules.txt. Do you know why it might not show up in this report?

There isn't an fglrx driver for Jaunty yet, and the one you are trying to use is not appropriate for the X server you are running.

[atiddxSetup] X version mismatch - detected X.org 7.1.-1.902, required X.org 7.4.-1.906

Subsequently, X falls back to using the ati driver, which should be good for your video card, but you have the libGl from the fglrx driver which is causing another failure and the resulting crash in compiz.

I have to reinstall de package xserver-xorg-core, because some file as /usr/lib/libGL.so.1 maybe not are present.

This bug is for request and show a message of error or exception when compiz crash with SIGSEGV in glXCreateContext(). These should be useful.

Now this is the ouput:

$ compiz --replace
Checking for Xgl: not present.
xset q doesn't reveal the location of the log file. Using fallback /var/log/Xorg.0.log
Detected PCI ID for VGA:
Checking for texture_from_pixmap: Aborted (core dumped)
not present.
Trying again with indirect rendering:
Checking for texture_from_pixmap: present.
Checking for non power of two support: present.
Checking for Composite extension: present.
Checking screen 1Comparing resolution (1440x900) to maximum 3D texture size (2048): Passed.
Checking for Software Rasterizer: Not present.
Checking for nVidia: not present.
Checking for FBConfig: present.
running under gnome seesion, checking for gnomecompat
Checking for Xgl: not present.
/usr/bin/compiz.real (cube) - Warn: Failed to load slide: /usr/share/gdm/themes/Human/ubuntu.png

Bryan, thanks for pointing this out. This will happen if "modinfo fglrx" does not print anything or fails; thus I guess the module is not really available? It also has this strange "(P)" after it, I newer saw that before.

For the apport side of things, let's err on the side of caution and consider an invalid module as nonfree.

apport trunk r1498

This bug was fixed in the package apport - 1.6-0ubuntu1

---------------
apport (1.6-0ubuntu1) karmic; urgency=low

  * New upstream release:
    - Add support for kernel crashes, thanks to Michael Vogt!
    - apport/ui.py, run_crash(): Do not re-collect information if we already
      have a Dependencies field. This happens when calling apport on an already
      pre-processed .crash file with -c. (LP: #394497)
    - apport/hookutils.py, pci_devices(): Deliver all matching devices, not
      just the last one. (LP: #398906)
    - hookutils.py, _get_module_license(): Return "invalid" if modinfo fails,
      so that they do not count as "free". (LP: #341720)
    - packaging-apt-dpkg.py: Support additional custom native origins in
      /etc/apport/native-origins.d/ . (LP: #386052)
    - packaging-apt-dpkg.py: Drop PPA origin hack, launchpad behaves properly
      now
    - apport-gtk: Avoid focus stealing when being called without arguments (i.
      e. auto-launched). LP: #396243)
    - apport-kde: Use standard gettext again
    - Fix handling of PC lacking disassembly due to invalid memory location.
  * debian/local/apport-collect: Tag bugs with "apport-collected" on success.
    (LP: #391392)

 -- Martin Pitt <email address hidden> Wed, 15 Jul 2009 18:02:59 +0200

It appears this was caused by installing an incompatible driver and it was crashing in that driver so compiz couldn't catch it ahead of time.