This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: PATCH: Optimized memset for x86-64
- From: "Jagasia, Harsha" <harsha dot jagasia at amd dot com>
- To: hj at lucon dot org, drepper at redhat dot com, libc-alpha at sourceware dot org
- Date: Fri, 19 Oct 2007 23:41:52 -0500
- Subject: Re: PATCH: Optimized memset for x86-64
Hi HJ,
>This is a rewrite of x86-64 memset driven by jump tables. It uses SSE
>register move for Intel64 and 8byte register move for AMD64.
>
As has been requested of our submissions in the past, please provide
performance results for AMD and Intel platforms.
We are in the process of reviewing the routine for AMD processors and
this comes at a short notice for us to do detailed measurements across
our processors. We will evaluate the impact of this routine on AMD
hardware before it is checked in. As Ulrich has mentioned in the past,
we should aim towards a blended implementation.
We also ask you review our outstanding submissions.
Thanks,
Harsha Jagasia
AMD