This is the mail archive of the libc-alpha@sourceware.org mailing list for the glibc project.
| Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
|---|---|---|
| Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |
| Other format: | [Raw text] | |
Hello,
Code was replaced to sysdeps/i386/i686/multiarch. New patch is attached.
My performance data was about { expf (which uses __ieee754_expf),
__ieee754_expf } not about { __expf_finite, __ieee754_expf }. Sorry
for misprint in names at my previous letter.
We didn't check __expf_finite but believe it becomes faster as well.
2012-04-05 Liubov Dmitrieva <liubov.dmitrieva@gmail.com>
* sysdeps/i386/i686/multiarch/Makefile: Add
(libm-sysdep_routines): e_expf-sse2 e_expf_finite.sse2 init-arch
* sysdeps/i386/i686/multiarch/e_expf.S: New file.
* sysdeps/i386/i686/multiarch/e_expf-sse2.S: New file.
* sysdeps/i386/i686/multiarch/e_expf_finite-sse2.S: New file.
--
Liubov Dmitrieva
Intel Corporation
2012/4/3 Ulrich Drepper <drepper@gmail.com>:
> On Mon, Apr 2, 2012 at 10:43, H.J. Lu <hjl.tools@gmail.com> wrote:
>> I think all i386 multiarch codes should be under
>> sysdeps/i386/i686.
>
> Indeed. ?None of this applies to pre-i686 hardware.
Attachment:
expf_x86_32.patch
Description: Binary data
| Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
|---|---|---|
| Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |