Aborted Command, ahd_timeout, panic: no dir

[Edgar Fuß <ef%math.uni-bonn.de@localhost>]

Can anyone make sense out of the following sequence of events (on 4.0, I have
to admit)?

1. A lengthy ahd(4) Dump Card State
2. sd0(ahd0:0:0:0): Check Condition on CDB 0x2a ...
   SENSE KEY: Aborted Command
   ASC/ASCQ: ASC 0x4d ASCQ 0x19
(ASC 4D is TAGGED OVERLAPPED COMMANDS)
3. ahd0: ahd_timeout
   ahd0:SCB 0x2 - timed out
4. Another lengthy ahd Dump Card State
5. A message from postfix "fatal: open /etc/postfix/main.cf: no such file or 
directory"
6. Another Check Condition (with ASC 4D)
7. Another postfix message
8. Another ahd_timout message
9. A third ahd Dump Card State
10. /usr: bad dir ino xxx at offset 0: mangled entry
11. panic: bad dir

Both /etc and /usr are on a RAIDframe Level 1 RAID (on sd0 and sd1), so I would
have expected some RAIDframe message had there been an error reading some
block.

After a reboot, sd0 didn't respond fully (i.e. the SCSI BIOS reported the 
drive model, but not the capacity and one couldn't boot from it) until I 
pulled out its SCA tray and re-plugged it.
After that, the machine came up fine.