Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1a 12/4/83; site rlgvax.UUCP
Path: utzoo!watmath!clyde!burl!hou3c!hocda!houxm!houxz!vax135!floyd!cmcl2!seismo!rlgvax!guy
From: guy@rlgvax.UUCP (Guy Harris)
Newsgroups: net.micro,net.arch,net.followup
Subject: Re: What's in a 3b2?
Message-ID: <2006@rlgvax.UUCP>
Date: Sat, 9-Jun-84 04:53:39 EDT
Article-I.D.: rlgvax.2006
Posted: Sat Jun  9 04:53:39 1984
Date-Received: Sun, 10-Jun-84 01:07:42 EDT
References: <2680@ecsvax.UUCP>
Organization: CCI Office Systems Group, Reston, VA
Lines: 25

Read the article "The Operating System and Language Support Features of
the BELLMAC-32", in the Proceedings of the Symposium on Architectural
Support for Programming Languages and Operating Systems, published as
ACM SIGARCH Computer Architecture News Volume 10, Number 2, March 1982
and ACM SIGPLAN Notices, Volume 17, Number 4, April 1982.  It doesn't contain
an opcode map and description, but it does give a number of details -
it's got a sort of post-PDP-11 architecture (general registers, instructions
consisting of opcode+operands, the latter being specified by "addressing
modes", etc.).  10 general purpose registers, and the PC, SP, FP, and AP
in the VAX-11 fashion.  So read the article to the AT&T person - it'll
be fun if it makes their head explode...  (Admittedly, HP is even more
closed-mouthed about their Focus chip, so AT&T isn't the only offender
here.)

One little tidbit - I've seen the disassembler code for the 3B20, and it's
one of the biggest I've ever seen.  I get the impression from it that
it's a lot of complexity in the instruction set to no great benefit -
it's slightly less space-efficient than VAX-11 code (the VAX-11 disassembler
consists of about two pages of code and a gigantic table) and the 3B20
performs at about the same level as an 11/780.  I don't know how similar
the 3B20 and the WE32000 are; I've heard that they are and I've heard that
they really aren't.

	Guy Harris
	{seismo,ihnp4,allegra}!rlgvax!guy