Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!usc!julius.cs.uiuc.edu!coolidge
From: coolidge@cs.uiuc.edu (John Coolidge)
Newsgroups: comp.unix.aux
Subject: Re: A/UX cc compiler
Message-ID: <1990Nov15.031150.3997@julius.cs.uiuc.edu>
Date: 15 Nov 90 03:11:50 GMT
References: <1281@ucl-cs.uucp>
Sender: news@julius.cs.uiuc.edu (USENet News)
Reply-To: coolidge@cs.uiuc.edu
Organization: U of Illinois, Dept. of Computer Science, Systems Research Group
Lines: 33

J.Pearce@cs.ucl.ac.uk writes:
>Can anyone make any comments on the status/limitations/bugs of the
>default cc compiler that comes with A/UX.

The default cc is an OK compiler, but nothing special. Its biggest
deficiency is that it has fixed-size tables for most everything
(something it shares with the standard cpp, as, and ld, although
with ld you seldom if every notice it). This means that large
systems (gcc, X11, etc) fail with the default cc.

I've never seen cc put out bad code (some claim they have, but I
haven't); on the other hand, I've never seen it put it very good
code either. Its optimizer is pretty poor, at least compared to
Gnu cc (gcc)'s. But, aside from blowing up repeatedly when its
tables overflow, I've never seen it actually spit out bad code.

That said, I much prefer gcc. It produces tighter, faster, smaller,
and just plain better code. In general it has better error messages
and it doesn't suffer from the fixed size tables problems. But
A/UX's cc is fine for most tasks.

>And just out of interest is a "plain" int represented as 2 or 4
>bytes ?

4 bytes.

--John

--------------------------------------------------------------------------
John L. Coolidge     Internet:coolidge@cs.uiuc.edu   UUCP:uiucdcs!coolidge
Of course I don't speak for the U of I (or anyone else except myself)
Copyright 1990 John L. Coolidge. Copying allowed if (and only if) attributed.
You may redistribute this article if and only if your recipients may as well.