On Sun, 3 Dec 1995, Christopher Eric Hanson wrote:

> In article <NETBSDList.1vnl@arcticus.burner.com> John Shardlow <john@iceberg.demon.co.uk> writes:
> > > Actually, the timeline for Sun was: Sun1 and Sun2 (68000) then Sun3
> > > (68020 and a few 030).  Then Sun made some i386 boxes.  Then Sun4
> > > (sparc).
> > How did they run Unix on a 68000 processor? Did it have no virtual
> > memory? If so maybe old versions of SunOS could run on any Amiga?
> 
>   Nope. I believe the Sun 68000 boxes had a Sun-proprietary mini-MMU.
> 
>   Didn't the AT&T 7300 Unix PC have something similar?
> 
> > John
> Chris - Xenon

>From everything I've read about the 68000-series, you have to have a 
68010 in order for any kind of MMU to work.  So the Sun-1 must be at 
least a 68010, or else they would have to be very careful because the 
68000 isn't VM-safe, namely if a page fault occurs in the middle of an 
instruction, there isn't enough data to know how far the instruction had
progressed (especially re: post-increment and pre-decrement instructions)..

As for the AT&T, that is just an Intel 80386 running SVR3.2, I should 
know because we have a number of these in our CS Lab, but the stupid 
sysadmin couldn't hack Unix so he just put DOS on all of them... :-(

------------------------------------------------------------------------------
     Jake Hamby                         |   E-Mail:  jehamby@lightside.com
  Student, Cal Poly University, Pomona  |   System Administrator, JPL
------------------------------------------------------------------------------