In article <1992Dec14.221159.21177@leland.Stanford.EDU>,
dhinds@leland.Stanford.EDU (David Hinds) wrote:
>
> We just got a Maxtor Tahiti II optical drive from Parity Systems, and
> I'm having trouble getting it going. I remember that a while back,
> someone posted about this, and gave the DIP switch settings he used,
> but alas I did not save it. The #^%$&* instructions from Parity are
> not helpful, to be generous; they only mention the device ID switches.
> I've set the ID, and properly hooked up the drive (plus terminator).
> When I boot, 'hinv' reports an optical drive in the appropriate place.
> I remembered reading that it was possible to avoid using the Parity
> drivers and just use the regular SCSI disk drivers, so I thought I'd
> give that a try. I tried running 'fx' with a new cartridge, and it
> seemed to successfully identify the drive, and gave some messages
> about creating default partitions and so on. The disk label looked
> OK. But when I went to exit from 'fx', I got an I/O error with the
> following messages in /usr/adm/SYSLOG:
>
> > Dec 14 13:50:46 hyper unix: sc1,5,0: cmd=0x8 Unexpected info phase 46, state 49. Resetting SCSI bus
> > Dec 14 13:50:46 hyper unix: PMO87: read error on block #0.
> > Dec 14 13:50:46 hyper unix: Error detected on PMO87 due to command <dis/allow medium removal>: unit attention : Power on, reset or bus device reset occurred
>
> So, I thought I'd go ahead and install the Parity driver and rebuild
> the kernel. That seemed to work, but when I run the 'pmoinfo' thing,
> even with just '-c INQUIRY', I get an I/O error the first time, and
> the second time seems to block in the kernel. I suppose I'll log a
> call with Parity, but if anyone could send me their DIP switch
> settings, it might save me some time...
>
We bought our Tahiti II directly from MaxOptix. After much
difficulty we finally found out that S1 #7, labelled
"disable unit attention" must be ON when the drive is
connected to a Macintosh. Since the error messages you
quote mention "unit attention" I suspect this is also
true for SGI drives. The Unit Attention state occurs
when a SCSI bus reset takes place; it is a way for the
drive to tell the controller "check my status BEFORE you
do anything else." Many controllers ignore this
request, and try to acces the drive anyway.
Perhaps somebody from SGI can tell us whether I am correct
about all this as it applies to SGI hardware; I've never
tried the optical drive on our SGI so I only know what