> I get the slot 9 thing, too.  Hmmm...I guess we'll have to figure out a
> clean way not to probe slots which don't exist :-)

Yeah. I mentioned this to Allen. We shouldn't be looking at slots that
don't exist especially because other things reside in those spaces on
some machines.

I figure we can some stuff to the tables in machdep.c and have nubus.c
use those values rather than the hard-coded ones it's using now.