[Jason R. Thorpe]

> So, what I was thinking of doing is having the MI layer stash the LUN
> in the cdb in the case of the scsibus's "maxlun" property being 7 ...

[Justin T. Gibbs]

> I think we should only send it if the device IDs itself as SCSI I
> compliant.  Even the spec suggests leaving it as 0 by default.

This sounds like a much better solution than my suggestion of adding a
new quirk to the SCSI subsystem.  Unless I am much mistaken, this is a
one-liner in /sys/scsi/scsi_base.c, and can't hurt anything so long as
it's done only for SCSI 1 devices.  Get it in there before 1.2 ships!

-tih
-- 
Tom Ivar Helbekkmo
tih@Hamartun.Priv.NO