Re: open64 versus gcc

Sid Touati <>
27 Nov 2006 17:45:12 -0500

          From comp.compilers

Related articles
open64 versus gcc (dz) (2006-11-22)
Re: open64 versus gcc (shrey) (2006-11-24)
Re: open64 versus gcc (touati) (2006-11-24)
Re: open64 versus gcc (Steven Bosscher) (2006-11-26)
Re: open64 versus gcc (Sid Touati) (2006-11-27)
Re: open64 versus gcc (dz) (2006-11-29)
Re: open64 versus gcc (Diego Novillo) (2006-11-29)
Re: open64 versus gcc (A.L.) (2006-12-01)
Re: open64 versus gcc (A.L.) (2006-12-01)
Re: open64 versus gcc (Jonathan Thornburg -- remove -animal to reply) (2006-12-03)
Re: open64 versus gcc (Diego Novillo) (2006-12-03)
[4 later articles]
| List of all articles for this month |

From: Sid Touati <>
Newsgroups: comp.compilers
Date: 27 Nov 2006 17:45:12 -0500
Organization: I.N.R.I.A Rocquencourt
References: 06-11-094 06-11-100 06-11-104
Keywords: GCC, code
Posted-Date: 27 Nov 2006 17:45:12 EST

Steven Bosscher a écrit :
> On 24 Nov 2006 08:21:12 -0500, touati wrote:
>> Actually, Open64 produces better code than gcc.
> Ah, generalizations... It depends on the target you want to look at.

I agree that my initial sentence looks like a generalization, but I am
aware that it isn't. It is difficult to make a fair comparison between
compilers, there are too parameters to explore...

Indeed, since all code optimization techniques (and their order of
execution) are based on ad hoc heuristics, we can never guarantee that
a compiler is better than another for any input program. Usually,
benchmarks are used, but benchmarks are rarely representative of
programs (they may represent workloads, but not
programs). Consequently, for any pair of distinct compilers C1 and C2,
you can always find programs better optimized with C1, and others
better optimized with C2.


Post a followup to this message

Return to the comp.compilers page.
Search the comp.compilers archives again.