> > Actually, it's not overblown for "our current needs," if you consider
> > other ports besides the	i386.  And, i've explained this in detail
> > several times before, when justifying the need for real MI bounce
> > buffers.
> 
> This still sounds very sad to me.  As long as we don't have MI bounce
> buffers, it's forbidden to have them on any arch.  As long as we don't
> have a MI serial port driver which can keep up with high speeds, it's
> illegal to have one for one arch only.

BTW:

there's nothing that prevents the i386 port from copying
e.g. dev/isa/com.c to arch/i386/isa/com.c, and maintaining a
machine-dependent version of the driver...  (except maybe good
software engineering practices.)

But it's a REALLY bad idea to break dev/isa/com.c for other ports,
since it's explicitly intended (by its placement) to be machine
indepedent, and is used (on some ports for critical functions) by
other ports.


cgd