git.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Janos Laube <janos.dev@gmail.com>
To: Johannes.Schindelin@gmx.de
Cc: Marius Storm-Olsen <marius@trolltech.com>,
	git@vger.kernel.org, msysgit@googlegroups.com
Subject: Re: [PATCH] Add custom memory allocator to MinGW and  MacOS builds
Date: Fri, 3 Apr 2009 15:50:08 +0000	[thread overview]
Message-ID: <9d6091530904030850x4eff9e54ga5fb6b13cd036335@mail.gmail.com> (raw)
In-Reply-To: <alpine.DEB.1.00.0904031659290.10279@pacific.mpi-cbg.de>


>
> Hi,
>
> On Fri, 3 Apr 2009, Marius Storm-Olsen wrote:
>
>> The standard allocator on Windows is pretty bad prior to Windows Vista,
>> and nedmalloc is better than the modified dlmalloc provided with newer
>> versions of the MinGW libc.
>>
>> Note that both Windows and MacOSX need to have an implementation of
>> strdup, which I added at the end of nedmalloc.c, to ensure that strdup
>> use the same heap. So, setting USE_NED_ALLOCATION = YesPlease will also
>> add this function reimplementation, unless you define
>> NEDMALLOC_DONT_OVERRIDE_STRDUP.
>>
>> NedMalloc stats in Git
>> ----------------------
>>   All results are the best result out of 3 runs. The
>>   benchmarks have been done on different hardware, so
>>   the repack times are not comparable.
>>
>>   These benchmarks are all based on 'git repack -adf'
>>   on the Linux kernel.
>>
>>   XP
>>   -----------------------------------------------
>>   MinGW               Threads  Total Time   Speed
>>   -----------------------------------------------
>>   3.4.2                  (1T)  00:12:28.422
>>   3.4.2     + nedmalloc  (1T)  00:07:25.437 1.68x
>>
>>   3.4.5                  (1T)  00:12:20.718
>>   3.4.5     + nedmalloc  (1T)  00:07:24.809 1.67x
>>
>>   4.3.3-tdm              (1T)  00:12:01.843
>>   4.3.3-tdm + nedmalloc  (1T)  00:07:16.468 1.65x
>>
>>   4.3.3-tdm              (2T)  00:07:35.062
>>   4.3.3-tdm + nedmalloc  (2T)  00:04:57.874 1.54x
>
> Wow!  That is impressive.

indeed, haven't time to test it myself currently tho :-( i hope
someone can confirm this massive performance boost and that this patch
makes it into msysgit very soon. thanks for the work :-)

      reply	other threads:[~2009-04-03 15:51 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-04-03 13:52 [PATCH] Add custom memory allocator to MinGW and MacOS builds Marius Storm-Olsen
2009-04-03 14:20 ` Marius Storm-Olsen
2009-04-03 21:12   ` [msysGit] " Pat Thoyts
2009-04-03 23:16     ` Johannes Schindelin
2009-04-03 23:42       ` [msysGit] " Pat Thoyts
2009-04-04  5:16     ` Marius Storm-Olsen
2009-04-04  6:35       ` [msysGit] " Johannes Sixt
2009-04-05 18:43         ` Marius Storm-Olsen
2009-04-03 15:01 ` Johannes Schindelin
2009-04-03 15:50   ` Janos Laube [this message]

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=9d6091530904030850x4eff9e54ga5fb6b13cd036335@mail.gmail.com \
    --to=janos.dev@gmail.com \
    --cc=Johannes.Schindelin@gmx.de \
    --cc=git@vger.kernel.org \
    --cc=marius@trolltech.com \
    --cc=msysgit@googlegroups.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 NNTP newsgroup(s).