Hello Brice,<br><br>I'm happy to announce a complete backtrace. No idea what was wrong the last time. But this trace looks really "nice", at least if a segmentation fault can look nice.<br><br>Greetings,<br> Andreas<br>
<br><div class="gmail_quote">2009/2/23 Brice Goglin <span dir="ltr"><<a href="mailto:Brice.Goglin@ens-lyon.org">Brice.Goglin@ens-lyon.org</a>></span><br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<div class="im">Andreas Cord-Landwehr wrote:<br>
> Hi, I started the x-Server by statx. Then I connected by ssh, attatched<br>
> gdm. After this I waited until the server freezes/crashes. Then I<br>
> executed "bt full" in gdm and quit gdm.<br>
<br>
</div>Oh, you mean "gdb", not "gdm" :)<br>
<div class="im"><br>
> After this the Xorg process was<br>
> gone. I assume the gdm does not kill any processes it is attached to?<br>
><br>
<br>
</div>No. gdb catches the process before it dies because of this SIGABRT. When<br>
you detach gdb, the process behave as if gdb wasn't there, i.e. it dies<br>
from SIGABRT for real.<br>
<div class="im"><br>
> After considering when these crashes do occured, I think they can be<br>
> related to the new kwin effects for KDE 4.2. Especially on switching the<br>
> kde-desktops, loosing scope of window aso.<br>
><br>
<br>
</div>I doesn't justify the missing backtrace.<br>
<br>
I am puzzled...<br>
<font color="#888888"><br>
Brice<br>
<br>
</font></blockquote></div><br>