public inbox for git@vger.kernel.org 
 help / color / mirror / Atom feed
From: Artur Skawina <art.08.09@gmail•com>
To: Sebastian Schuberth <sschuberth@gmail•com>
Cc: Johannes Sixt <j.sixt@viscovery•net>,
	msysGit <msysgit@googlegroups•com>,
	Junio C Hamano <gitster@pobox•com>,
	Git Mailing List <git@vger•kernel.org>
Subject: Re: [PATCH] block-sha1: Windows declares ntohl() in winsock2.h
Date: Tue, 18 Aug 2009 14:56:49 +0200	[thread overview]
Message-ID: <4A8AA511.1060205@gmail.com> (raw)
In-Reply-To: <4A8A8661.5060908@gmail.com>

Sebastian Schuberth wrote:
> As ntohl()/htonl() are function calls (that internally do shifts), I
> doubt they're faster than the shift macros, though I haven't measured
> it. However, I do not suggest to go for the macros on Windows/Intel, but
> to apply the following patch on top of your patch:

> On Windows/Intel, ntohl()/htonl() are function calls that do shifts to
> swap the
> byte order. Using the native bswap instruction boths gets rid of the
> shifts and
> the function call overhead to gain some performance.

Umm, nothing like this should be needed on linux; the compiler/glibc
will choose bswap itself. (see endian.h and bits/byteswap.h).
I did try using __builtin_bswap32 directly and the result was a few
(3 or 4, iirc) differently scheduled instructions, that's all, no
performance difference.

>   * Performance might be improved if the CPU architecture is OK with
> - * unaligned 32-bit loads and a fast ntohl() is available.
> + * unaligned 32-bit loads and a fast ntohl() is available. On Intel,
> + * use the bswap built-in to get rid of the function call overhead.
>   * Otherwise fall back to byte loads and shifts which is portable,
>   * and is faster on architectures with memory alignment issues.
>   */
> 
> -#if defined(__i386__) || defined(__x86_64__) || \
> -    defined(__ppc__) || defined(__ppc64__) || \
> -    defined(__powerpc__) || defined(__powerpc64__) || \
> -    defined(__s390__) || defined(__s390x__)
> +#if defined(__i386__) || defined(__x86_64__)
>
> +#define get_be32(p)    __builtin_bswap32(*(unsigned int *)(p))
> +#define put_be32(p, v)    do { *(unsigned int *)(p) = __builtin_bswap32(v); } while (0)
> +
> +#elif defined(__ppc__) || defined(__ppc64__) || \
> +      defined(__powerpc__) || defined(__powerpc64__) || \
> +      defined(__s390__) || defined(__s390x__)

I'd limit it to windows and any other ia32 platform that doesn't pick the
bswaps itself; as is, it just adds an unnecessary hidden gcc dependency.

Hmm, it's actually a gcc-4.3+ dependency, so it won't even build w/ gcc 4.2;
something like this would be required: "(__GNUC__>=4 && __GNUC_MINOR__>=3)" .

artur

  parent reply	other threads:[~2009-08-18 12:57 UTC|newest]

Thread overview: 34+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-08-18  7:15 [PATCH] block-sha1: Windows declares ntohl() in winsock2.h Johannes Sixt
2009-08-18 10:45 ` Sebastian Schuberth
2009-08-18 11:07   ` Junio C Hamano
2009-08-18 11:28     ` Sebastian Schuberth
2009-08-18 12:56   ` Artur Skawina [this message]
2009-08-18 13:17     ` Sebastian Schuberth
2009-08-18 13:39       ` Junio C Hamano
2009-08-18 15:40         ` Linus Torvalds
2009-08-18 16:08           ` Linus Torvalds
2009-08-18 16:23             ` Sebastian Schuberth
2009-08-18 16:43               ` Junio C Hamano
2009-08-18 16:30         ` Nicolas Pitre
2009-08-18 16:43           ` Nicolas Pitre
2009-08-18 16:50             ` Junio C Hamano
2009-08-18 18:01               ` Nicolas Pitre
2009-08-18 19:00                 ` Junio C Hamano
2009-08-18 19:22                   ` Nicolas Pitre
2009-08-18 19:26                     ` [PATCH] make sure byte swapping is optimal for git Nicolas Pitre
2009-08-18 19:37                     ` [PATCH] block-sha1: guard gcc extensions with __GNUC__ Nicolas Pitre
2009-08-18 19:40                   ` [PATCH] block-sha1: Windows declares ntohl() in winsock2.h Junio C Hamano
2009-08-18 19:56                   ` Brandon Casey
2009-08-18 20:10                     ` Nicolas Pitre
2009-08-18 20:25                       ` Linus Torvalds
2009-08-18 20:29                       ` Brandon Casey
2009-08-20  2:26                     ` Nicolas Pitre
2009-08-20  2:30                       ` Linus Torvalds
2009-08-20  2:45                       ` Brandon Casey
2009-08-18 16:59             ` Sebastian Schuberth
2009-08-18 17:05               ` Junio C Hamano
2009-08-18 18:16                 ` Nicolas Pitre
2009-08-18 20:17                   ` Junio C Hamano
2009-08-18 20:30                     ` Junio C Hamano
2009-08-18 20:49                       ` Nicolas Pitre
2009-08-18 18:10               ` Nicolas Pitre

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=4A8AA511.1060205@gmail.com \
    --to=art.08.09@gmail$(echo .)com \
    --cc=git@vger$(echo .)kernel.org \
    --cc=gitster@pobox$(echo .)com \
    --cc=j.sixt@viscovery$(echo .)net \
    --cc=msysgit@googlegroups$(echo .)com \
    --cc=sschuberth@gmail$(echo .)com \
    /path/to/YOUR_REPLY

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