>> How does it affect the MI pci subsystem if sparc64 chooses to do
>> things otherwise, ie, not use it?  I must be missing something.
> If the MI subsystem is deficient, then that should be addressed,
> rather than adding machine-dependent kludges.

Yes, it should.  But since when has this become dogma strong enough to
justify overriding a port's portmaster, and thereby presumably breaking
that port until someone figures out how to fix the MI stuff, and does
it?  And, then, why aren't you also backing out, say, all graphics-head
console support for ports that don't use wscons?  Or any of the other
various cases where a port doesn't yet use the MI version of something?

Or will you be, and this is just where you started?

/~\ The ASCII				der Mouse
\ / Ribbon Campaign
 X  Against HTML	       mouse@rodents.montreal.qc.ca
/ \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B