Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!seismo!brl-adm!rutgers!sri-unix!hplabs!hp-sdd!ncr-sd!ncrcae!ece-csc!mcnc!unc!rentsch
From: rentsch@unc.UUCP (Tim Rentsch)
Newsgroups: comp.lang.smalltalk
Subject: BS on sun-3's, fix for multiplication primitive
Message-ID: <255@unc.unc.UUCP>
Date: Tue, 18-Nov-86 15:29:20 EST
Article-I.D.: unc.255
Posted: Tue Nov 18 15:29:20 1986
Date-Received: Wed, 19-Nov-86 21:29:36 EST
Distribution: net
Organization: CS Dept, U. of N. Carolina, Chapel Hill
Lines: 33

(OK, just one more note on net.lang.st80....)

I know some people out there are wondering about running BS on
sun-3's.  For those that haven't been able to get this going, here
is a hint or two.

We got BS running on a sun-3 mostly just by recompiling the sources
(sun-3 a.out format is different, remember?).  My recollection is
that this is straightforward, i.e., no strange machine dependencies.

There is one thing that had to be changed.  Several of the BS
modules are compiled to produce assembly output, then an 'awk'
script is used on the intermediate assembly (before running it
through the assembler) to do some form of peephole optimization.  To
make a long story short, the 'awk' peephole optimization doesn't
work;  the reason the same thing worked on the sun-2's is that the
particular optimization case never happened.  (In fairness to the
Berkeley people, I should say that it is my opinion that the 'awk'
script does not work.  But BS ran fine after I removed that step.)

So, change the makefile to delete the 'awk' step on those modules
that formerly used it, and things should work fine.  

Also, those with BS should know that the multiplication primitive
sometimes fails, due to the way it is done, and how overflow is
detected.  (Also true on sun-2, as I recall.)  The fix is to
multiply A times B to give C, then see if C/B == A, since if it does
then overflow will not have occurred.  (Maybe I posted this before,
I have forgotten.)

cheers,

txr