This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: [PATCH] Fix POWER4/POWER7 optimized strncmp to not read pastdiffering bytes
- From: Ryan Arnold <ryan dot arnold at gmail dot com>
- To: Ulrich Drepper <drepper at gmail dot com>
- Cc: Andreas Schwab <schwab at linux-m68k dot org>, libc-alpha at sourceware dot org, Steven Munroe <sjmunroe at us dot ibm dot com>
- Date: Sun, 17 Apr 2011 20:02:12 -0500
- Subject: Re: [PATCH] Fix POWER4/POWER7 optimized strncmp to not read pastdiffering bytes
- References: <m2vcz3rsio.fsf@igel.home><BANLkTikj+aX348gTYhEWmhUt5nZ0F4ihLQ@mail.gmail.com>
It corrects the behavior exposed in the bug.
I think you should apply it and we'll submit a patch later after we do
some performance analysis if we find any significant performance
regression. I have someone looking at it right now.
What kind of runway do we have for making 2.14? If nothing else we
can get the 2.14 maintainer to backport a perf fix after the branch is
tagged.
Ryan
On Sun, Apr 17, 2011 at 7:51 PM, Ulrich Drepper <drepper@gmail.com> wrote:
> On Mon, Mar 28, 2011 at 14:47, Andreas Schwab <schwab@linux-m68k.org> wrote:
>> 2011-03-28 ÂAndreas Schwab Â<schwab@linux-m68k.org>
>>
>> Â Â Â Â* sysdeps/powerpc/powerpc32/power4/strncmp.S: Don't read past
>> Â Â Â Âdiffering bytes.
>> Â Â Â Â* sysdeps/powerpc/powerpc64/power4/strncmp.S: Likewise.
>> Â Â Â Â* sysdeps/powerpc/powerpc32/power7/strncmp.S: Likewise.
>> Â Â Â Â* sysdeps/powerpc/powerpc64/power7/strncmp.S: Likewise.
>
> Is this patch correct? ÂI haven't seen anything from the ppc people.
>