Re: XFree crashing on kernel 2.4.28
BTW. Has anyone got a coredump from the X server crash? I don't remember
the exact rules for when a coredump is produced but /usr/X11R6/bin/X is
setuid and setgid so that probably stops it producing a core. Maybe try
chmod u-s /usr/X11R6/bin/X
chmod g-s /usr/X11R6/bin/X
then start X as root and see if that gives a coredump.
Get a backtrace with
gdb /usr/X11R6/bin/X core
using the "where" command (and probably lots more). That will probably
help to identify what is happening (of course the non-debug version of X
probably doesn't have symbols but it might give something of use)
On Wed, 2004-12-01 at 09:55, Richard Mortimer wrote:
> This looks very similar to Debian bug 280384 the crash happens in the
> same place.
> On Wed, 2004-12-01 at 06:29, Jurzitza, Dieter wrote:
> > Dear listmembers,
> > the latest information on my side comparing an strace - output of XFree86. Thanks to Admar who gave the hint moving XFree86 to XFree86.bin and doing a strace /usr/bin/X11/XFree86.bin out of the script /usr/bin/X11/XFree86.
> > XFree86 segfaults during startup. That's about all you get out of the straces - but someone of you might see even more.
> > As always, U60 - SMP @ 360 MHz, 512 MB of ram, Debian Sarge booting a vanilla kernel.
> > Thank you for any feedback.
> > The strace of the functional startup is really large, so I compressed it.
> > The file strace is a small excerpt out of the segfaulting 2.4.28 log.
> > I found out that even "normal" network communication is somewhat out of order with 2.4.28. A ssh may be successful - may be not. Something must have gone wrong along the migration - path from 2.4.27 to 2.4.28.
> > The .config files for compilation between 2.4.27 and 2.4.28 are identical. This cannot be the root cause for the problems I see.
> > Thanks for any inputs.
> > Take care
> > Dieter