In article <199402200910.BAA01324@sun-lamp.cs.berkeley.edu> Adam Glass <glass@sun-lamp.cs.berkeley.edu> writes:
> 
> losing output, not taking input etc.  I've revised it considerably vs
> the version thats in the tree and am now willing to throw it away for
> a ported-'zs' driver.  If anyone has made any progress porting the
> 'zs' driver, let me know ASAP, as I'll port the sparc one sunday
> sometime unless i hear otherwise.

For anyone interested in working on the zs driver, I found a source for
the zscc chip series manual.  Call Zilog at (408) 370-8000 and ask for
your local sales office's number (or, they may allow you to order one
directly).  The Dallas office is (214) 987-9987.  Sorry, I don't have
the others.  The manual is free, BTW. 

I also found out more about speeding up the zs driver. First, the Z8530
has a two-byte FIFO, not a three-byte one. Also, it's essentially
useless since the chip generates an interrupt for every byte anyway.
(However, the "lockup" problem may still be solvable by reading two
bytes from the FIFO if availible).

More interesting to me, I found out some of what I wanted to know about
the Z85230.  It does have a larger FIFO (though I couldn't find the
exact size durring a unfortunately brief look through someone else's
manual).  The '230s are pin-compatible with the '30s, assuming the '230
is availible in the same packaging/type/speed as the '30 as used in the
Sun 3 (40 pin DIP, CMOS, 8MHz).  The chip also has the ability to
interrupt only on the first received character instead of every one. 
So, it looks like it may be a winner for squeezing more serial
performance out of the Sun 3.

-- 
Ty Sarna                 "As you know, Joel, children have always looked
tsarna@endicor.com        up to cowboys as role models. And vice versa."

------------------------------------------------------------------------------