Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/18/84; site peora.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!mhuxr!mhuxt!houxm!mtuxo!mtunh!mtung!mtunf!ariel!vax135!petsd!peora!jer
From: jer@peora.UUCP (J. Eric Roskos)
Newsgroups: net.arch
Subject: Re: Re: RISC (really on multiplication d
Message-ID: <1376@peora.UUCP>
Date: Thu, 25-Jul-85 11:24:40 EDT
Article-I.D.: peora.1376
Posted: Thu Jul 25 11:24:40 1985
Date-Received: Sat, 27-Jul-85 01:33:39 EDT
References: <149@mips.UUCP> <600005@pbear.UUCP> <493@mmintl.UUCP>
Organization: Perkin-Elmer SDC, Orlando, Fl.
Lines: 24

> Incidently, re the subscript computation problem, why not a command which
> multiplies a word by a one-byte constant?  This would deal cleanly with
> most two-dimensional arrays and arrays of structures, yet should be
> implementable as a reasonably efficient instruction.

Why not have instructions that contain small constants built into the
opcode?  What?  Never heard of that?  I guess it's because it's not a RISC...

Seriously, though, the above sorts of things are quite common in current
CISC research... in fact, there are some good papers (e.g., by Flynn) on
optimally encoding instructions; this is where the recently-much-maligned
ideas on having machines with different instruction sets for different
applications/compilers come from (whether or not different COMPILERS need
different instruction sets, or whether it's really the APPLICATION that
determines it, depends on who you talk to.).  Actually it's much more
general than that, though, since you use an optimization scheme to minimize
the length of instructions for some desired set of programs; and a result
of this is that short constants come out encoded in the opcodes for some
common classes of programs.
-- 
Shyy-Anzr:  J. Eric Roskos
UUCP:       ..!{decvax,ucbvax,ihnp4}!vax135!petsd!peora!jer
US Mail:    MS 795; Perkin-Elmer SDC;
	    2486 Sand Lake Road, Orlando, FL 32809-7642