> From: "Chris G. Demetriou" <cgd@cs.cmu.edu>
> Date: Sat, 25 Jan 1997 21:47:14 -0500

[names...]
> Actually, on the alpha, i'd call it 'asc'...  8-)

In the version I uploaded, the MD code can call it whatever it wants.

> > I do _not_ like the notion of <machine/espvar.h>, or whatever. [...]
[ I've already responded to this. -gwr ]

> It's not just a matter of this...
> 
> If every driver using MI code starts requiring a header in <machine>,
> things quickly get out of hand...

True, you need one for each port, but almost all of the traditional
xxxvar.h stuff would be in the common <dev/ic/xxxvar.h> file.
To make this more clear, I've attached a copy of the file I would
propose as <arch/sparc/include/espvar.h> below.  (Only 58 lines!)

> > The functionality that is provided by the current macros in the header
> > files could be provided by function pointers.
> 
> "yes."

They could, yes, but that is not the ONLY way, and I've yet to see
any explanation as to why it SHOULD be the only way permitted.

> > I'd also like to see softc layering a'la the MI 5380 driver.  (That
> > even allows ports to arbitrarily rename it :-)
> 
> "yes."

Arbitrary naming is there now.

Gordon