On Monday 14 March 2011 1:38:11 am Mark Goldstein wrote:
It might be the same issue I've seen before with oS 11.1 and newer kernels (from Evergeen repository). Somehow the KDM is started early and X later and KDM times out before X is ready. The simple workaround was to change KDM timeout in /opt/kde3/share/config/kdm/kdmrc The default was 15 seconds. I set it to 100 and it solved the issue.
Thanks, Mark. Although it didn't directly fix the problem, it enabled me to stumble opon the solution! There were so many timeout parms there, that I decided to change the Local Displays section and, while I was there, Server Attempts (from 1 to 2). A reboot produced the same commandline though, and while I was sitting there pondering the issue, without logging in, (after 15 seconds, I guess) it automatically retried and went straight to the KDM login screen! So I started looking at all the logs (warn, X, messages, etc) and came across this (just a snippet from "/var/log/messages"): ++++++++++++++++++++++++++++++++++++++++++++++ Mar 14 08:48:24 s113kde3 kdm_config[2370]: Multiple occurrences of key 'AutoLoginEnable' in section [X-:0-Core] of /opt/kde3/share/config/kdm/kdmrc Mar 14 08:48:24 s113kde3 acpid: 1 client rule loaded Mar 14 09:48:23 s113kde3 ntpd[2417]: ntpd 4.2.4p8@1.1612-o Mon Jul 5 13:44:05 UTC 2010 (1) Mar 14 09:48:23 s113kde3 ntpd[2418]: precision = 1.000 usec Mar 14 09:48:23 s113kde3 ntpd[2418]: ntp_io: estimated max descriptors: 1024, initial socket boundary: 16 Mar 14 09:48:23 s113kde3 ntpd[2418]: Listening on interface #0 wildcard, 0.0.0.0#123 Disabled .. .. Mar 14 09:48:25 s113kde3 kdm[2369]: X server startup timeout, terminating +++++++++++++++++++++++++++++++++++++++++++++ See the one hour upshift? At that point some process must think there's a one-hour gap since the Xserver was asked to fire up ... perhaps this is the issue? Bug?? Dunno, beyond me. That got me thinking ... this had all started Sunday morning (DST commencement), and I was on this 11.3 system from about 11 pm till 1:30 am, pre-DST. So I rebooted, checked the BIOS and, sure enough, the BIOS clock was still at 8:57, one hour behind actual DST (9:57 actual). Is this normal? I advanced the clock one hour manually, rebooted, and everything was back to normal, inluding the other openSuse partition that was displaying the same problem. I've now reset kdmrc back to it's original state and all is still well/normal. The two other gens/partitions that were NOT displaying the same problem were 11.4 gens, one a KDE3 and the other a KDE4, They still boot fine. Obviously the 11.4 ones seem to have handled the problem more gracefully than 11.3, but if all have NTP running, why did none of them not adjust the clock? Anyway, I'll look into it further but, in the meantime, many thanks to all for the assistance. -- To unsubscribe, e-mail: opensuse-kde3+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-kde3+help@opensuse.org