This is the mail archive of the
mailing list for the GCC project.
Re: [BENCHMARK]-mfpmath=sse should disable x387 intrinsics
- From: Roger Sayle <roger at eyesopen dot com>
- To: Richard Guenther <richard dot guenther at gmail dot com>
- Cc: Uros Bizjak <uros at kss-loka dot si>, <gcc-patches at gcc dot gnu dot org>
- Date: Thu, 25 Nov 2004 10:45:09 -0700 (MST)
- Subject: Re: [BENCHMARK]-mfpmath=sse should disable x387 intrinsics
On Thu, 25 Nov 2004, Richard Guenther wrote:
> Now I actually read the patch and looked at the patch context.
> Simulating the effect of your patch results in
> -mfpmath=sse -mno-fancy-math-387 -D__NO_MATH_INLINES: 55.8s
> -mfpmath=sse -mno-fancy-math-387: 55.8s
> Oh - and your patch misses updating of the documentation of -mfpmath (and
> possibly -mno-fancy-math-387).
Now you've misread the documentation for -mno-fancy-math-387 which
currently has absolutely no effect on recent processors:
const int x86_arch_always_fancy_math_387 = m_PENT | m_PPRO | m_ATHLON_K8
| m_PENT4 | m_NOCONA;
Before complaining much more, might I strongly suggest that you actually
try running some benchmarks with current CVS. Clearly, you're not exactly
sure what the performance impact of my patch is/will be. I'd not rule out
that there could be a performance improvement in which case much of the
previous discussions would be moot.