[SPAM]Log out from gnome makes machine freeze

Mark Howell mhowell at northlink.com
Thu Apr 22 08:00:21 UTC 2004


[snip]
> 
> Hi !
> 
> I have a Dell Optiplex GX240 with ATI Rage 128 Pro Ultra TF 
> using the r128 driver in X.
> 
> Running 2.6.5-1.332 and the rest is also updated to latest.
> 
> When i logout from Gnome the screen goes black and the 
> machine freeze completely as far as I can see, it does not 
> even answer to ping anymore. Choosing shutdown or restart works.
> 
> kernel args from grub.conf: kernel /vmlinuz-2.6.5-1.332 ro 
> root=LABEL=/ quiet
> 
> selinux is disabled in /etc/sysconfig/selinux.
> 
> Any ideas ?
> 
> regards
> -- 
> Karl-Olov Serrander kase at cntw.com

I have a similar problem with my Dell Inspiron 7500 notebook with ATI Rage
Mobility P video, except it goes black and hangs whenever I log out,
shutdown, or restart (whether from GNOME or from a vt or from an ssh
session). The xorg ati driver doesn't work with my video card, so I've been
using the vesa driver. What about you, Karl-Olov: ati or vesa?

Haven't gotten around to debugging it, but it has been an issue since FC2T2
and with every rawhide kernel and xorg since, as I've updated. If someone
can give us some suggestions on debugging or tracing this behavior I'd
appreciate it. My level of sophistication at this point is ssh'ing in and
tail -f /var/log/messages, which yields this:

Broadcast message from root (pts/3) (Wed Apr 21 23:51:38 2004):

The system is going down for reboot NOW!
Apr 21 23:51:38 marx shutdown: shutting down for system reboot
Apr 21 23:51:38 marx init: Switching to runlevel: 6
Apr 21 23:51:39 marx gconfd (markh-3114): Received signal 15, shutting down
clea
nly
Apr 21 23:51:39 marx gconfd (markh-3114): Exiting
Apr 21 23:51:40 marx su(pam_unix)[26719]: session closed for user root
Apr 21 23:51:41 marx gdm(pam_unix)[2896]: session closed for user markh
Apr 21 23:51:41 marx udev[26888]: removing device node '/udev/vcsa2'
Apr 21 23:51:41 marx udev[26887]: removing device node '/udev/vcs2'
Apr 21 23:51:41 marx udev[26892]: removing device node '/udev/vcsa4'
Apr 21 23:51:41 marx udev[26898]: removing device node '/udev/vcsa6'
Apr 21 23:51:41 marx udev[26895]: removing device node '/udev/vcs5'
Apr 21 23:51:41 marx udev[26890]: removing device node '/udev/vcsa3'
Apr 21 23:51:41 marx udev[26896]: removing device node '/udev/vcsa5'
Apr 21 23:51:41 marx udev[26889]: removing device node '/udev/vcs3'
Apr 21 23:51:41 marx udev[26897]: removing device node '/udev/vcs6'
Apr 21 23:51:41 marx udev[26891]: removing device node '/udev/vcs4'
Apr 21 23:51:43 marx bonobo-activation-server (markh-3119): iid
OAFIID:BrokenNoT
ype:20000808 has a NULL type
Apr 21 23:51:43 marx bonobo-activation-server (markh-3119): invalid
character '#
' in iid 'OAFIID:This#!!%$iid%^$%_|~!OAFIID_ContainsBadChars'

... And then the machine is in la-la land until I cycle power. All of these
messages looked benign, I don't think I'm getting a view of what is really
breaking yet. This is with latest (332) kernel.

I'll go get a null modem cable if what I really need is serial debug... But
I'd appreciate some pointers on stuff to try, and I will report back.

Cheers,

Mark.





More information about the fedora-test-list mailing list