Well, after checking the syslog entries generated by kdm I realized that the problem was an I/O error on /var/run/xdmctl [...] In fact, I have had to reinstall my X server some time ago, as similar disk I/O related problems did occur.
Has anybody experienced similar problems? What should be done?
Yes. Replace hard disk. Simple ;) Use package smartmontools and run smartctl -s on -a /dev/hda to get information on the disk's internal selftest status. Run disk selftests with smartctl -t short /dev/hda and smartctl -t long /dev/hda. If either stops with a block error the disk is dead. If the number of reallocated sectors is very high, say
100: disk dead. If it keeps increasing: disk dieing fast.
As you had a few of these disk read errors now, chances of the disk not being *&@*#$(ed are very small. Volker -- Volker Kuhlmann is possibly list0570 with the domain in header http://volker.dnsalias.net/ Please do not CC list postings to me.