Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!uunet!world!iecc!compilers-sender
From: han@cs.rochester.edu
Newsgroups: comp.compilers
Subject: intermediate representation
Keywords: question, optimize, design, analysis
Message-ID: <9102210223.AA15327@cyan.cs.rochester.edu>
Date: 21 Feb 91 02:23:33 GMT
Sender: compilers-sender@iecc.cambridge.ma.us
Reply-To: han@cs.rochester.edu
Organization: Compilers Central
Lines: 23
Approved: compilers@iecc.cambridge.ma.us


   I am taking a compiler seminar course that emphasizes the back end
optimization issues.   In order to get hands-on experience, I would like to
practice applying various optimization techniques on the intermediate
representation produced by the front end.   (I have not decided on the
front end, yet.  I might write one or pick up one and modify.)   At this
point, I am looking for intermediate representations that I can play with.
So, my first question is:  Are there well documented (and, hopefully,
state-of-the-art) definitions of intermediate representations available?
How can I get them?  (Does the RTL that GCC uses count as one such IR?)
   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.  
The RTL dump on SPARC looked very different from that on Sun3 workstation.
We had expected that the machine dependency would show up at much later
stage and that we would be able to try optimization with RTL without worrying
about machine dependencies.
So, my second question is: Is it generally expected for an intermediate
representation to be machine dependent?   How far toward the end of code
generation phase can machine dependencies be prevented from showing up?
-- 
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