On Wed, 20 Oct 1999, Jonathan Stone wrote:

> Great news!
> 
> Where does this put us vis-a-vis shipping MI scsi for 1.5?
> My understanding is that there's two efforts:
> 	a) Nisimura-san's branch which uses MI scsi ncr53c9x.cK
>         b) michael hitch's code which uses the `old' Mach-derived
> 	   asc driver, but interfacing to the MI scsi subsystem

  My code uses Nishimura-san's initial code for the ncr53c9x driver, which
originally didn't work completely.  I got it working for the IOASIC and TC
option cards, but got stuck on kernel dumps using the IOASIC SCSI.
[Which, by the way, I finally was able to get working :-)]  I also added
the buffer alignment corrections that Nisimura-san did.  Those corrections
addressed even-byte aligned buffers, but not odd-byte aligned buffers.
I've been working on getting odd-aligned buffers to work (reads are now
working, but writes still have some trouble).

> Neither of which addresses the SII in the 2100/3100/5100.  So
> reworking the SII driver to talk to the MI scsi layer is the main
> hurdle. Is that right? 
> 
> Michael?  Want to take on another 4.4BSD-to-scspi rewrite?

  I keep thinking about looking at the SII driver and seeing if it might
be possible to warp the driver into working with the MI SCSI stuff, but
never seem to find the incentive to get started.  I did finally get the
3100 set back up after my last office move, so I do have the hardware to
work on ready.  If I can get another 3100 working, I may be able to take
it home so I can work on it there, which gives me more debugging time.

--
Michael L. Hitch			mhitch@montana.edu
Computer Consultant
Information Technology Center
Montana State University	Bozeman, MT	USA