Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!utgpu!water!watnot!watmath!clyde!rutgers!lll-lcc!pyramid!voder!jay
From: jay@voder.UUCP
Newsgroups: comp.arch
Subject: Re: register window machine questions
Message-ID: <2857@voder.UUCP>
Date: Fri, 27-Feb-87 15:02:42 EST
Article-I.D.: voder.2857
Posted: Fri Feb 27 15:02:42 1987
Date-Received: Sun, 1-Mar-87 11:03:56 EST
References: <4376@columbia.UUCP> <7284@boring.mcvax.cwi.nl> <14986@amdcad.UUCP>
Organization: National Semiconductor, Santa Clara
Lines: 18
Summary: register window machine questions

In article <14986@amdcad.UUCP>, bcase@amdcad.UUCP (Brian Case) writes:
> In article <7284@boring.mcvax.cwi.nl> jack@boring.UUCP (Jack Jansen) writes:
> >
> >Something that I would be very interested in is wether there
> >is any research done into machines with an unlimited register
> >file.
> 
> Close, but not quite, the C Machine Stack Cache.  See SIGARCH/SIGPLAN
> notices, vol. 17, no. 4, April 1982.

It seems like a waste to have a machine with unlimited registers. 
At any given point there is only one set of registers active and other
registers are just sitting there. Instead, a better idea would be to 
limit the number of windows to, say two or three and do loads and stores
if you exceed the number of available windows. It might be a wothwhile
experiment to find out how many windows are optimum for running UNIX.

Jay