On Fri, 21 Mar 1997 14:57:26 +1030 (CST) 
 Michael Smith <msmith@atrad.adelaide.edu.au> wrote:

 > Of course, if it _does_ implement ISA properly (ie. provides both
 > I/O mapped and memory mapped space, as well as busmaster DMA), then there's
 > no end to the evil things that could be done with it.  8)

...well, remember, the whole notion of i/o-mapped is sort of tied to the
old Intel family... there was an additional line that was raised to
indicate "i/o port" rather than "memory" addresses.

In any case, i/o-mapped access is emulated on other CPUs... it's actually
memory-mapped, typically with a mutant address decoding.

 > > ...it's all memory-mapped on the hp300.  We've already dealt with these
 > > issues... it's called "bus.h" :-)
 > 
 > I'll go read it before I make more of a fool of myself 8)

No fools here :-)  bus.h isn't implemented on the hp300 yes, but it really
ought to be.  That's on my medium-term TODO list, but I may have to bump
it up if you guys keep making progress :-)

 > > ...wow, more reasons to visit Australia!  :-)  (Off topic, but are you
 > > near Melbourne?)
 > 
 > Uhh, about 750km by road.

Ah ... I was going to suggest hooking up w/ some other NetBSD hackers
in Melbourne... beer-by-proxy :-)

Jason R. Thorpe                                       thorpej@nas.nasa.gov
NASA Ames Research Center                               Home: 408.866.1912
NAS: M/S 258-6                                          Work: 415.604.0935
Moffett Field, CA 94035                                Pager: 415.428.6939