Hello,
Running kernel 2.6.18-8.1.4.el5.centos.plus on centos5, on shutdown sometimes I get:
Starting killall: Sending all processes the TERM signal...scim-panel-gtk ...scim-panel-gtk[4065]- segfault at... error 4. (sometimes there is one segfault error and sometimes 7+]
Also, [and this may be unrelated] I occasionally can't boot into X and need to: #rm /tmp/.X0-lock Then I can startx.
I have noticed sometimes that I get the segfault, but the next time I boot, I can boot into X. I really don't know enough to guess whether sometimes it is finishing X and removing the lock before the segfault and sometimes it is not.
I am not so unhappy with the situation but being unknowledable of course worry about what effects a segfault during shutdown may have. I do notice that after the segfaults, it goes on to unmount the filesystems and other things that it does normally (but I can't say everything normally as I don't really understand the normal steps of it shutting down).
I'd like to:
1) figure out how to stop the segfault especially since sometimes the scim panel won't appear after a reboot (and requires another reboot to appear again).
2) see if the scim segfault is causing the trouble with X starting and if not, figure out why the lock is not being removed.
If you have suggestions as to what to look into, I am happy to hear them.
The only thing I have tried is googling scim-panel-gtk segfault and found
scim-lib 0.99.6 is available (13 Aug 2004) Fixed a bug which causes segfault on x86_64 platform.
But that must be a different bug as: rpm -qa | grep scim-lib scim-libs-1.4.4-39.el5.x86_64
I saw an RPM someplace for scim-debug but wouldn't really know how to go about trying to debug it.
Shawn
P.S. Note, I do not believe this is in anyway connected to being a centos plus kernel as on a stock kernel had the trouble.