Subject: Re: smparc - success with mixed modules?!
To: Herb Peyerl <hpeyerl@beer.org>
From: Paul Kranenburg <pk@cs.few.eur.nl>
List: port-sparc
Date: 01/20/2003 23:25:01
> Anyway, I just powered up the machine with my 100Mhz module and my
> 150Mhz module.  Here's what I get:
> 
> <#0> ok cd Ross,RT625@f,fafffffc
> <#0> ok cd Ross,RT626@f,f8fffffc

> So, it looks like the 100Mhz has a 256K cache and the 150 has a 512K cache.
> 
> Guess I can't use these two together with NetBSD yet.
 

I'm curious how far it gets though. I've removed a couple of the
all-caches-are-equal assumptions.

At least one of the remaining (in fact, the only one I can spot right now)
can be circumvented by putting the module with the largest cache in 
MBus slot #0.  It appears you have already done that.

So, modulo all of the other hypersparc bugs left, that would not be
a showstopper.

-pk