Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!henry
From: henry@utzoo.UUCP (Henry Spencer)
Newsgroups: net.arch
Subject: Re: Delayed Branches
Message-ID: <7090@utzoo.UUCP>
Date: Sat, 6-Sep-86 21:30:44 EDT
Article-I.D.: utzoo.7090
Posted: Sat Sep  6 21:30:44 1986
Date-Received: Sat, 6-Sep-86 21:30:44 EDT
References: <304@sei.cmu.edu>
Organization: U of Toronto Zoology
Lines: 15

> First, I don't believe the Assembler should
> reorganise code.  The purpose of assembler code
> is to allow the user total control over what
> the machine will do...

John Mashey has already rebutted this in some detail.  I'd like to add one
thing:  I think we have an unrecognized cultural difference here.  The Unix
community assumes that compilers generate assembler, while a lot of other
people assume the compiler goes direct to binary.  Reorganization at the
assembler level makes rather more sense when you realize that *all* code
goes through the assembler.  I.e., most of the code going into the assembler
was *not* written directly by humans.
-- 
				Henry Spencer @ U of Toronto Zoology
				{allegra,ihnp4,decvax,pyramid}!utzoo!henry