Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!usc!elroy.jpl.nasa.gov!decwrl!world!iecc!compilers-sender
From: rfg@ncd.com (Ron Guilmette)
Newsgroups: comp.compilers
Subject: Re: intermediate representation
Keywords: question, optimize, design, analysis
Message-ID: <4106@lupine.NCD.COM>
Date: 26 Feb 91 00:06:13 GMT
References: <9102210223.AA15327@cyan.cs.rochester.edu>
Sender: compilers-sender@iecc.cambridge.ma.us
Reply-To: rfg@ncd.com (Ron Guilmette)
Organization: Network Computing Devices, Inc., Mt. View, CA
Lines: 23
Approved: compilers@iecc.cambridge.ma.us


In article <9102210223.AA15327@cyan.cs.rochester.edu> han@cs.rochester.edu writes:
>
>   Actually, my fellow students and I have taken a look at the RTL dump
>produced by GCC.   To our surprise, the RTL dump showed strong dependency
>on the target machine even at the earliest stage of optimization.  

For what it's worth, I thought that I should mention that GCC actually
deals with two different internal representations of a program.  There
is a high level representation (TREEs) and also the low-level representation
that you mentioned (i.e. RTL).

The high level representation is fairly machine independent, however it is
*not* something that you would want to try applying optimizations to.
(Well, at least GCC doesn't try to do that.  Some people may say that
the high level tree representation would be a good thing to be munging
for certain "high level" optimizations.  I won't argue that point either
way.)

-- 

// Ron Guilmette
// Internet: rfg@ncd.com      uucp: ...uunet!lupine!rfg
-- 
Send compilers articles to compilers@iecc.cambridge.ma.us or
{ima | spdcc | world}!iecc!compilers.  Meta-mail to compilers-request.


Brought to you by Super Global Mega Corp .com