[reply to multiple messages here]

> Waitaminute - are you saying that you can make a multiple-cpu UVII?

Yes.  I worked with someone who was doing robot control on a uV2.
Problem was, the CPU didn't have enough computrons to do the control
loop *and* run an OS.  So we slapped in a 620 (as I mentioned, we
couldn't find 630s that didn't come with MicroVAXen around them) to do
the robot control.  I wrote a little microkernel for it, just barely
enough to support what we needed....

> I happen to have access to another KA630 - of course the questions
> loom: Does NetBSD actually support this?

Heh.  AFAIK NetBSD supports multiprocessor machines in only a few
cases, and only in the sense that it doesn't break in the presence of
multiple processors.  Last I heard it doesn't actually use the
additional processors for anything, ever.

I'm sure efforts are underway to change this, but given the NetBSD
tendency to "do it right", they are unlikely to appear soon.

> Would I just put the two CPUs in slots 1 and 2, the two 8meg boards
> in 3 and 4, and the memory interconnect across all four?

I'm pretty sure that wouldn't work.  I don't know much about the memory
interconnect on the C/D slots, but I imagine it wasn't designed for
this (and shoppa's note seems to confirm this).

> What about two KA, two mems, say cpu1 has 16MB board and cpu2 has
> 8MB.  in a ba123 of course....I haven't tried it yet but I have been
> tempted.  What do I hose up?

Probably the memory interconnect on the C/D slots.  Unless you cut a
bunch of etch runs....

> assuming that the memory addresses on the qbus will overlap (the
> first 8 MB anyhow).

The Q-bus is only 4M big (IIRC).  Each CPU is responsible for setting
up mapping for whatever fraction of its memory it chooses to export to
the Q-bus....

					der Mouse

			       mouse@rodents.montreal.qc.ca
		     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B