I found a reliable way to crash 2.2.17-pre1, aic7xxx:
Use a cd-writer that sometimes just hangs after a writing error occured.
Then, issue anoter SCSI command to the writer. That leads to scsi resets
because the drive doesn't answer.
While the driver is still trying to get the command through, rmmod the
aic7xxx driver. This results in the following messages:
Jun 13 20:53:05 anduin kernel: SCSI device not inactive - rq_status=1, target=6, pid=24255, state=4099, owner=258.
Jun 13 20:53:05 anduin kernel: Device busy???
After that, lsmod doesn't show aic7xxx any more, so I try to reload it:
Jun 13 20:53:20 anduin kernel: aic7xxx: <Adaptec AHA-294X SCSI host adapterJ> at
PCI 0/15/0
Jun 13 20:53:20 anduin kernel: aic7xxx: I/O ports already in use, ignoring.
This is the last message syslog wrote - although the driver notices there
is something wrong, the machine oopses (but the oops doesn't get written
to disk).
I know, this is not a normal situation, but I'd like to have a safer reaction.
Just failing the rmmod would be an option. I might have to reboot to get scsi
working, but this is still better than a crash.
Jan
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Thu Jun 15 2000 - 21:00:29 EST