public inbox for libc-ports@sourceware.org
 help / color / mirror / Atom feed
From: "Carlos O'Donell" <carlos@redhat.com>
To: Steve Ellcey <sellcey@mips.com>,
	       "Carlos O'Donell" <carlos@systemhalted.org>
Cc: Andrew Pinski <pinskia@gmail.com>,
	       "Joseph S. Myers" <joseph@codesourcery.com>,
	       "libc-ports@sourceware.org" <libc-ports@sourceware.org>
Subject: Re: [patch, mips] Improved memset for MIPS
Date: Fri, 13 Dec 2013 04:40:00 -0000	[thread overview]
Message-ID: <52AA8FAF.5070606@redhat.com> (raw)
In-Reply-To: <1386893669.2764.30.camel@ubuntu-sellcey>

On 12/12/2013 07:14 PM, Steve Ellcey wrote:
> On Thu, 2013-12-12 at 19:01 -0500, Carlos O'Donell wrote:
> 
>>> I noticed this patch causes some performance regressions on Octeon due
>>> to having 128 byte cache lines.
>>> Changing PREFETCH_CHUNK/PREFETCH_FOR_STORE to assume 128 byte cache
>>> line gives us the performance back and improves over the original code
>>> at least 15%.
>>> That is:
>>> #  define PREFETCH_CHUNK 128
>>> #  define PREFETCH_FOR_STORE(chunk, reg) \
>>>     pref PREFETCH_STORE_HINT, (chunk)*128(reg);
>>
>> Submit a patch for that?
>>
>> We have microbenchmarks now, but the next hardest
>> part is going to be archiving data by device so that
>> the community can help track performance and point
>> out regressions like this.
>>
>> Cheers,
>> Carlos.
> 
> Unless the change is under some kind of ifdef for Octeon changing this
> will probably slow down other MIPS chips.  Most of them have 32 byte
> cache lines.

Absolutely. I don't suggest he just change it, but Andrew would have to add
enough framework for Octeon to be enabled with an optimal implementation.
For example you could compile an alternate version with 128 byte cache
line support and select it via IFUNC based on AT_HWCAP?

Cheers,
Carlos.

  parent reply	other threads:[~2013-12-13  4:40 UTC|newest]

Thread overview: 32+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-09-05 17:06 Steve Ellcey 
2013-09-06  0:40 ` Mike Frysinger
2013-09-06 15:42   ` Steve Ellcey
2013-09-06  4:18 ` Carlos O'Donell
2013-09-06 16:03   ` Steve Ellcey
2013-09-06 17:12     ` Carlos O'Donell
2013-09-06 23:33       ` Steve Ellcey
2013-09-07  2:38         ` Carlos O'Donell
2013-09-10 20:31           ` Steve Ellcey
2013-09-10 21:01             ` Carlos O'Donell
2013-09-10 21:14               ` Steve Ellcey
2013-09-10 22:35                 ` Carlos O'Donell
2013-09-10 22:38                   ` Carlos O'Donell
2013-09-07  5:46         ` Andreas Schwab
2013-09-06 14:31 ` Joseph S. Myers
2013-09-06 15:58   ` Steve Ellcey
2013-09-06 16:09     ` Joseph S. Myers
2013-09-06 16:50       ` Steve Ellcey
2013-09-06 16:59         ` Joseph S. Myers
2013-09-06 17:43           ` Steve Ellcey
2013-09-06 18:57             ` Brooks Moses
2013-09-18 17:41           ` Steve Ellcey
2013-09-19 15:25             ` Carlos O'Donell
2013-09-19 17:02               ` Steve Ellcey
2013-09-20 16:43             ` Joseph S. Myers
2013-09-20 17:32               ` Steve Ellcey
2013-12-12 22:19                 ` Andrew Pinski
2013-12-13  0:01                   ` Carlos O'Donell
2013-12-13  0:14                     ` Steve Ellcey
2013-12-13  0:22                       ` Andrew Pinski
2013-12-13  4:40                       ` Carlos O'Donell [this message]
2013-09-06 16:59       ` Steve Ellcey

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=52AA8FAF.5070606@redhat.com \
    --to=carlos@redhat.com \
    --cc=carlos@systemhalted.org \
    --cc=joseph@codesourcery.com \
    --cc=libc-ports@sourceware.org \
    --cc=pinskia@gmail.com \
    --cc=sellcey@mips.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).