Forwarding a bug from Ubuntu user lykwydchykyn:
Display freezes during the KDE login process. GPU hung and batchbuffer dump attached.Bug was originally reported for Ubuntu Lucid which has kernel 2.6.32 and intel driver 2.9.1, but the bachbuffer dump and log here are from kernel 2.6.33-rc5 and new git-snapshots of xorg from xorg-edgers.
Using Ubuntu Lucid (10.04) with latest updates 1/20/10. Graphics chip is intel 845G in a Dell GX60.
In KDE, enabling compositing under System Settings causes X to freeze up. System does not respond to anything except alt+sysrq commands.
After reboot, X freezes during user login, presumably when Kwin is being started but that's just a guess (it locks about 3/4 of the way
through the login splash animation).
[Automatic info from original report]
Date: Thu Jan 21 13:34:36 2010
DistroRelease: Ubuntu 10.04
MachineType: Dell Computer Corporation OptiPlex GX60
Package: xserver-xorg-video-intel 2:2.9.1-1ubuntu1
ProcCmdLine: BOOT_IMAGE=/boot/vmlinuz-2.6.32-11-generic root=UUID=2dd1c926-1c4f-42ad-9bb0-a86bd0adb63b ro nosetmode
ProcVersionSignature: Ubuntu 2.6.32-11.15-generic
Uname: Linux 2.6.32-11-generic i686
XorgConf: Error: [Errno 2] No such file or directory: '/etc/X11/xorg.conf'
Xrandr: Error: command ['xrandr', '--verbose'] failed with exit code 1: Can't open display
dmi.bios.vendor: Dell Computer Corporation
dmi.board.vendor: Dell Computer Corp.
dmi.chassis.vendor: Dell Computer Corporation
dmi.product.name: OptiPlex GX60
dmi.sys.vendor: Dell Computer Corporation
fglrx: Not loaded
glxinfo: Error: [Errno 2] No such file or directory
setxkbmap: Error: command ['setxkbmap', '-print'] failed with exit code 255: Cannot open display "default display"
architecture: i686kernel: 2.6.32-11-generic
Created attachment 32814 [details]
Batchbuffer dump with DebugFlushCaches set
Created attachment 32815 [details]
Created attachment 32816 [details]
Created attachment 32817 [details]
Output of lspci -vvnn
Spontaneous lockup on i845, I wonder what that reminds me of...
Sadly the dump is afflicted by the usual post-hang-check lack of batch buffer information, so I can't say for sure that is a true dup, but it is so similarly to several of the reported cases, that I suspect it is one and the same bug.
*** This bug has been marked as a duplicate of bug 26345 ***