On Wed, 2008-04-30 at 18:48 -0400, fred smith wrote: > Just had a weird thing happen that I've never encountered before. > > Running Centos 5.x (updated) on 32-bit (Athlon XP). > > Was going along in Firefox when suddenly my screen went black and the hard > drive light came on mostly steady for several seconds. After a little bit > I got back the X login screen. It normally comes up on F7, but F7 was dead > (with some text I'll paste in below in case it means anything to someone) > and (since I have another dedicated app on F8) it came up on F9. > > The last few entries in /var/log/messages, which are about the right > time, are: > > Apr 30 18:34:44 hostname restorecond: Will not restore a file with more than one hard link (/etc/resolv.conf) No such file or directory > Apr 30 18:34:52 hostname kernel: NVRM: not using NVAGP, an AGPGART backend is loaded! > Apr 30 18:35:10 hostname gconfd (username-32302): GConf server is not in use, shutting down. > Apr 30 18:35:10 hostname gconfd (username-32302): Exiting > Apr 30 18:36:59 hostname restorecond: Will not restore a file with more than one hard link (/etc/resolv.conf) No such file or directory > Apr 30 18:37:22 hostname gconfd (username-8990): starting (version 2.14.0), pid 8990 user 'username' > Apr 30 18:37:22 hostname gconfd (username-8990): Resolved address "xml:readonly:/etc/gconf/gconf.xml.mandatory" to a read-only configuration source at position 0 > Apr 30 18:37:22 hostname gconfd (username-8990): Resolved address "xml:readwrite:/home/username/.gconf" to a writable configuration source at position 1 > Apr 30 18:37:22 hostname gconfd (username-8990): Resolved address "xml:readonly:/etc/gconf/gconf.xml.defaults" to a read-only configuration source at position 2 > Apr 30 18:37:27 hostname pcscd: winscard.c:219:SCardConnect() Reader E-Gate 0 0 Not Found > Apr 30 18:37:27 hostname last message repeated 4 times > Apr 30 18:37:30 hostname gconfd (username-8990): Resolved address "xml:readwrite:/home/username/.gconf" to a writable configuration source at position 0 > > > and the stuff on the F7 screen is (hard to read and surely incomplete): > > > b7913000-b7a52000 rw-p b7913000 00:00 0 > b7a52000-b7a7c000 rw-s 00000000 00:08 192315392 /SYSV00000000 (deleted) > b7a7c000-b7abc000 rw-s d3f90000 00:10 6900 /dev/nvidia0 > b7abc000-b7ac0000 rw-s d0500000 00:10 6900 /dev/nvidia0 > b7ac0000-b7fc0000 rw-s d0000000 00:10 6900 /dev/nvidia0 > b7fc0000-b7fd0000 rw-s e4800000 00:10 6900 /dev/nvidia0 > b7fd0000-b7fd4000 rw-p b7fd0000 00:00 0 > b7fd4000-b7fd5000 rw-s e4682000 00:10 6900 /dev/nvidia0 > b7fd5000-b7fd6000 rw-s e4680000 00:10 6900 /dev/nvidia0 > b7fd6000-b7fde000 rw-s 240bb000 00:10 6900 /dev/nvidia0 > b7fde000-b7fdf000 rw-s 1dc43000 00:10 6900 > /dev/nvidia0 > b7fdf000-b7fef000 rw-s 24730000 00:10 6900 /dev/nvidia0 > bf864000-bf87f000 rwxp bf864000 00:00 0 [stack] > bf87f000-bf881000 rw-p bf87f000 00:00 0 > > > So since the stuff from /var/log/messages mentions agpgart and this > stuff from F7 mentions /dev/nvidia, I'm guessing the nvidia drivers had > some kind of brain fart (:) ?? > > FWIW, I'm running the Nvidia 96.43.01 drivers, direct from Nvidia > and not any repository. > > Any ideas? Yes don't force selection of the "agpart" Systems that can't run it will give you that message. It should use the defauilt provided by the driver. So if you put it in the xorg.conf file take it out. Check your /etc/xorg/xorg.conf to make sure your using the "nvidia" driver and not the "nv" driver. > > Thanks! > > _______________________________________________ > CentOS mailing list > CentOS at centos.org > http://lists.centos.org/mailman/listinfo/centos -- ~/john OpenPGP Sig:BA91F079