>> What's really involved in moving to MI SCSI?  At a _very_ quick glance,
>> it looks like the Alpha uses the MI stuff and also support PMAZ cards,
>> so there's going to be a lot of stuff to work from.  You're right - it's
>> been talked about for a while now :)
>
> The TC Alphas use a tcds chip which provides two SCSI buses (one for
> disks and one supposeldy dedicated for a multimedia CD-ROM).
>
> Which is completely different from the ioasic and associated DMA
> hardware on the TC Decstations....  I havent looked but I wonder if
> the Alpha TC PMAZ (PMAZ-C?) support uses real DMA rather than the
> onboard SRAM hardware-bounce-buffer on the PMAZ-A.

Builtin dual channel SCSI adaptor found in TC Alpha is driven by TCDS
asic.  There are versions of TC option card of TCDS named PMAZB-ZZ and
PMAZC-ZZ, which exactly work like builtin TCDS.

NetBSD/pmax has one obstracle in PMAZ-AA prevents MI ncr53c94 glued
together.  It handles two different SCSI adaptors DMA-capable IOASIC
asc and PMAZ-AA TC option card in the same manner.  Cuting-pasting MI
ncr53c94 code for IOASIC would be little problem, but the code cannot
cope with PMAZ-AA.  So, simultaneous support for IOASIC/PMAZ-AA with
MI ncr53c94 needs significant work, probably right support of bus_*
things and glues in NetBSD/pmax. 

Tohru Nishimura
Nara Institute of Science and Technology