I think I've figured out exactly what my problem is with talking to
my scanner causing ext2 panics. My scanner is rather trigger-happy
causing aborts, which kill commands that are currently on my hard
drive's queue. This trickles up to ext2, which panics.
Where in the ext2 filesystem are things panicing? While it would be
good see if we can fix things on the SCSI side of things, making the
ext2 code more robust would also be a good thing.