This is the mail archive of the
mailing list for the glibc project.
Re: [PATCH] New numbers in the benchtests.
On 12/18/2017 11:19 AM, Joseph Myers wrote:
On Sat, 16 Dec 2017, Patrick McGehearty wrote:
On 12/15/2017 11:31 PM, Siddhesh Poyarekar wrote:
On Friday 15 December 2017 09:34 PM, Joseph Myers wrote:
On Thu, 14 Dec 2017, Ben Woodard wrote:
Here are the results of that part of the modified tests:
What do you get with the exp patch (version 8) that has been approved and
now just needs to be committed?
I assume you'll be committing that as reviewer; I don't think Paul has
For clarity, the exp() submitter is Patrick McGehearty [not Paul McGehearty
and I do not (yet) have commit access.
If you'd like the patch committed for you, please provide a GNU ChangeLog
entry for it.
Thank you for the offer.
Proposed ChangeLog entry:
2017-12-19 Patrick McGehearty <firstname.lastname@example.org>
* sysdeps/ieee754/dbl-64/e_exp.c: 5x faster __ieee754_exp()
* sysdeps/ieee754/dbl-64/eexp.tbl: New file for e_exp.
* sysdeps/ieee754/dbl-64/slowexp.c: Removed.
* sysdeps/x86_64/fpu/multiarch/slowexp-avx.c: Removed.
* sysdeps/x86_64/fpu/multiarch/slowexp-fma.c: Removed.
* sysdeps/x86_64/fpu/multiarch/slowexp-fma4.c: Removed.
* sysdeps/generic/math_private.h: Remove slowexp.
* sysdeps/ieee754/dbl-64/e_pow.c: Likewise.
* sysdeps/powerpc/power4/fpu/Makefile: Likewise.
* sysdeps/x86_64/fpu/multiarch/Makefile: Likewise.
* sysdeps/x86_64/fpu/multiarch/e_exp-avx.c: Likewise.
* sysdeps/x86_64/fpu/multiarch/e_exp-fma.c: Likewise.
* sysdeps/x86_64/fpu/multiarch/e_exp-fma4.c: Likewise.
* math/Makefile: Likewise.
* manual/probes.texi: Remove slowexp documentation.