Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!mhuxl!ihnp4!zehntel!hplabs!sri-unix!salkind@nyu
From: salkind%nyu@sri-unix.UUCP
Newsgroups: net.unix-wizards
Subject: Re: 4.2bsd & nexus 0x6c on 11/780
Message-ID: <17343@sri-arpa.UUCP>
Date: Fri, 9-Mar-84 06:47:00 EST
Article-I.D.: sri-arpa.17343
Posted: Fri Mar  9 06:47:00 1984
Date-Received: Tue, 13-Mar-84 08:05:33 EST
Lines: 21

From:  Lou Salkind 

Although the fix handles the usual case (internally interleaved
controllers), it doesn't handle some "pathological" configurations
(for example, what happens if only the upper controller of the MS780E
is enabled?).

I also fixed the MS780E problem, but in a different and slightly more
general way.

Note the 4.2 bsd memory handling code makes the following assumption:
	There is a 1-1 correspondence between cpu type and memory controller
	type.  (Note, for example, that the case tests in the memory
	controller routines are by CPU type.)
To my way of thinking, this is not a valid assumption.

My fix involves setting the memory controller type in autoconf.c, and
then changing the memory code to test by controller type, not cpu type.
The code works fine on our 780's.

	Lou