On Thu, Apr 05, 2007 at 12:25:29AM +0200, Brice Goglin wrote:
> You should be able to check this by downgrading your nvidia driver with
> http://snapshot.debian.net/package/nvidia-graphics-drivers
Looks like this is not nvidia issue.
> I would be good to get a backtrace with debugging enabled (you would
> need to rebuild a debugging xserver-xorg-core as described on
> http://wiki.debian.org/HowToGetABacktrace). Then, you attach 'gdb
> $(pidof X)' and wait for the crash to occur to get a backtrace (with bt
> full). But, this gdb can't be run from inside X since it would die
> during the crash. And, sometimes, the text console dies too. So we
> usually recommend to log by ssh to attach gdb, but you can't do that
> during suspend/resume.
> So, if the text console does not die when X crashes, you could just
> attach a gdb from a text console before suspending. If not, maybe it is
> possible to put the gdb in a screen shell and log by ssh after the crash
> and get the screen shell back to debug in gdb?
So I have just recompiled xserver-xorg-core and installed recompiled
versions of packages. After this i decide to use gdb a in screen shell.
After crashing there was followed messages in screen session:
Program received signal SIGUSR1, User defined signal 1.
0xffffe410 in __kernel_vsyscall ()
(gdb)
Continuing.
Program received signal SIGSEGV, Segmentation fault.
0x08195b73 in WaitForSomething (
pClientsReady=0xbfea0a70)
at ../../os/WaitFor.c:220
220 if (timeout < 0)
(gdb)
Continuing.
Program received signal SIGABRT, Aborted.
0xffffe410 in __kernel_vsyscall ()
(gdb)
Continuing.
Program terminated with signal SIGABRT, Aborted.
The program no longer exists.
(gdb)
The program is not being run.
I'm not sure why gdb continue executing process after SIGSEGV. Will try to
find something on monday.
--
WBR, Dmitry
Attachment:
signature.asc
Description: Digital signature