I posted my previous message too quickly! I didn't notice the following
message in /var/adm/syslog:
Kernel panic: EXT2-fs panic (device 22/1): ext2_find_entry: buffer head pointer is NULL
This was the last message logged before the lockup. I hope this will
help fix the bug. Thanks!
How this message got logged is still not clear for me - panic() does an
infinite loop so syslogd didn't have a chance to run after the panic...
Just curious - could someone explain this to me?
Marek