This is the mail archive of the
glibc-bugs@sourceware.org
mailing list for the glibc project.
[Bug string/25131] memcpy perfomance problem with ARM 32 A9be due to high cache-misses
- From: "helugang at huawei dot com" <sourceware-bugzilla at sourceware dot org>
- To: glibc-bugs at sourceware dot org
- Date: Tue, 26 Nov 2019 03:12:26 +0000
- Subject: [Bug string/25131] memcpy perfomance problem with ARM 32 A9be due to high cache-misses
- Auto-submitted: auto-generated
- References: <bug-25131-131@http.sourceware.org/bugzilla/>
https://sourceware.org/bugzilla/show_bug.cgi?id=25131
--- Comment #14 from helugang <helugang at huawei dot com> ---
(In reply to Adhemerval Zanella from comment #13)
> I forgot to mention that if compiler also targets armv7 with -mfpu=neon as
> default the ifunc selector won't be used and instead
> sysdeps/arm/armv7/multiarch/memcpy_neon.S will set as the default
> implementation.
Thank you very much for the detailed explanation.
Currently in our platform -march=armv7-a is set ,so __memcpy_vfp is working
now.
According to your suggestion, we can use __memcpy_neon with -mfpu=neon set and
if we build glibc with --disable-multi-arch configured,the default
implementation __memcpy_arm will work.
Is my understanding correct?
We will try to re-evalute the three ways.
--
You are receiving this mail because:
You are on the CC list for the bug.