> Hmm.  I haven't looked at this particular program, but you might check
> for endian problems.  It was probably written for the i386 and assumed
> little-endian bit ordering for the data fields.

Thanks, John and Allen.  I got an "illegal request" message from the
driver (looks like this: "cd0(:6:0): illegal request, data = 00 00 00
00 24 00 00 00 00 00"), so I'll try to follow the WorkMan code with
a SCSI book and driver source, looking for endian problems.

Thanks!

kenn