I've two questions:
1. Could the driver somehow have sent the device off to lala-land?
2. What state is the kernel in after the reset? A reboot resulted in lots of
messages indicating that the buffer cache was confused to some degree (the
"Weird...." messages about finding clean buffers where it shouldn't). Was the
reboot necessary? There was no OOPS or anything like that... but it just
didn't look like it could be too healthy after reporting an I/O error on the
swap-partition device immediately following the SCSI reset.
-- Alan Eldridge (alane@wozzle.york.cuny.edu, also alane@libnet.com)