This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Little C-benchmarks of gcc-XXX: Compilation timings of make foreach GCC's version.
- From: Scott Robert Ladd <coyote at coyotegulch dot com>
- To: jc-nospam at jr-pizarro dot jazztel dot es
- Cc: gcc at gcc dot gnu dot org
- Date: Wed, 02 Jun 2004 11:58:45 -0400
- Subject: Re: Little C-benchmarks of gcc-XXX: Compilation timings of make foreach GCC's version.
- References: <1086181666.40bdd1220b9a7@webmail.jazznet.es>
jc-nospam@jr-pizarro.jazztel.es wrote:
I did continue since yesterday with the same options/flags & architecture ...
http://gcc.gnu.org/ml/gcc/2004-06/msg00045.html
and i'd added the -O3 option (even if it was not very important).
I've done quite a few GCC benchmarks over the years, and debate the
value of comparing against GCC 2.95, especially in terms of generated
code performance.
GCC 2.95 simply isn't an option for anyone writing C99 or C++ or (soon)
Fortran 95. The 3.x series compilers offer features not found in earlier
compilers; while I won't dispute that GCC 2.95 may have compiled faster
or produced speedier code for traditional C, we also need to recognize
that the current compilers offer many advantages in terms of language
support and error reporting.
--
Scott Robert Ladd
Coyote Gulch Productions (http://www.coyotegulch.com)
Software Invention for High-Performance Computing