This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: [PATCH] x86: Use AVX2 memcpy/memset on Skylake server [BZ #21396]
- From: "H.J. Lu" <hjl dot tools at gmail dot com>
- To: Florian Weimer <fweimer at redhat dot com>
- Cc: GNU C Library <libc-alpha at sourceware dot org>
- Date: Tue, 18 Apr 2017 12:21:39 -0700
- Subject: Re: [PATCH] x86: Use AVX2 memcpy/memset on Skylake server [BZ #21396]
- Authentication-results: sourceware.org; auth=none
- References: <20170418183712.GA22211@intel.com> <a91779cf-519b-16c6-f1bb-af4cbadec23a@redhat.com>
On Tue, Apr 18, 2017 at 12:08 PM, Florian Weimer <fweimer@redhat.com> wrote:
> On 04/18/2017 08:37 PM, H.J. Lu wrote:
>>
>> On Skylake server, AVX512 load/store instructions in memcpy/memset may
>> lead to lower CPU turbo frequency in certain situations. Use of AVX2
>> in memcpy/memset has been observed to have improved overall performance
>> in many workloads due to the higher frequency.
>
>
> Since you bring this up: I think I have read reports that this happens with
> AVX2 and some parts, too.
That is true and AVX512 has higher impact on CPU frequency than AVX2.
> The change itself looks reasonable.
>
> Thanks,
> Florian
--
H.J.