[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: GMP and the Intel compiler
From: |
Kevin Ryde |
Subject: |
Re: GMP and the Intel compiler |
Date: |
Thu, 23 Jan 2003 10:09:56 +1000 |
User-agent: |
Gnus/5.090013 (Oort Gnus v0.13) Emacs/21.1 (i386-debian-linux-gnu) |
Ronald Bruck <address@hidden> writes:
>
> Has anybody modified the source for GMP so that it will compile with
> the Intel C++ 7.0 compiler?
Not so far as I know.
> (Perhaps it's politically incorrect for
> Gnu software not to use the Gnu cc?)
Oh, well, we don't mind seeing vendor compilers work, but it's got to
be within the autoconf/automake/libtool framework. Anything else is
too hard.
> Intel claims, and my (few) tests
> seem to support, that it produces code 30% faster than gcc's.
GMP speed comes primarily from the assembler code. If any of the C
code is causing a big bottleneck we'd like to know and try to do
something about it.
> I've tried, of course, and failed miserably.
I think only the very latest libtool cvs has support for icc shared
libraries. A --disable-shared build will hopefully work, if the
compiler command line options are tolerably standard.
--
All followups to address@hidden please.