From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <SRS0=UrM4=EP=1wt.eu=w@gnuweeb.org>
X-Spam-Checker-Version: SpamAssassin 3.4.6 (2021-04-09) on gnuweeb.org
X-Spam-Level: 
X-Spam-Status: No, score=1.0 required=5.0 tests=HEADER_FROM_DIFFERENT_DOMAINS,
	SPF_HELO_PASS,SPF_PASS autolearn=no autolearn_force=no version=3.4.6
Authentication-Results: gnuweeb.org; dmarc=none (p=none dis=none) header.from=1wt.eu
Received-SPF: Pass (mailfrom) identity=mailfrom; client-ip=163.172.96.212; helo=1wt.eu; envelope-from=w@1wt.eu; receiver=<UNKNOWN> 
Received: from 1wt.eu (ded1.1wt.eu [163.172.96.212])
	by gnuweeb.org (Postfix) with ESMTP id 9FB0A24B336
	for <gwml@vger.gnuweeb.org>; Wed, 30 Aug 2023 22:23:36 +0700 (WIB)
Received: (from willy@localhost)
	by mail.home.local (8.17.1/8.17.1/Submit) id 37UFNMms027896;
	Wed, 30 Aug 2023 17:23:22 +0200
Date: Wed, 30 Aug 2023 17:23:22 +0200
From: Willy Tarreau <w@1wt.eu>
To: Ammar Faizi <ammarfaizi2@gnuweeb.org>
Cc: Alviro Iskandar Setiawan <alviro.iskandar@gnuweeb.org>,
        Thomas =?iso-8859-1?Q?Wei=DFschuh?= <linux@weissschuh.net>,
        Nicholas Rosenberg <inori@vnlx.org>,
        Michael William Jonathan <moe@gnuweeb.org>,
        GNU/Weeb Mailing List <gwml@vger.gnuweeb.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
Subject: Re: [RFC PATCH v1 2/5] tools/nolibc: x86-64: Use `rep stosb` for
 `memset()`
Message-ID: <ZO9e6h2jjVIMpBJP@1wt.eu>
References: <20230830135726.1939997-1-ammarfaizi2@gnuweeb.org>
 <20230830135726.1939997-3-ammarfaizi2@gnuweeb.org>
 <CAOG64qNGdPjXHTvrpYxjizOYvMRaB9_2t2raw6DJ3kK852KUHg@mail.gmail.com>
 <ZO9bv+GvgpphtGEi@biznet-home.integral.gnuweeb.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <ZO9bv+GvgpphtGEi@biznet-home.integral.gnuweeb.org>
List-Id: <gwml.vger.gnuweeb.org>

On Wed, Aug 30, 2023 at 10:09:51PM +0700, Ammar Faizi wrote:
> On Wed, Aug 30, 2023 at 09:24:45PM +0700, Alviro Iskandar Setiawan wrote:
> > Just a small idea to shrink this more, "mov %rdi, %rdx" and "mov %rdx,
> > %rax" can be replaced with "push %rdi" and "pop %rax" (they are just a
> > byte). So we can save 4 bytes more.
> > 
> > 0000000000001500 <memset>:
> >     1500: 48 89 f0     mov    %rsi,%rax
> >     1503: 48 89 d1     mov    %rdx,%rcx
> >     1506: 57           push   %rdi
> >     1507: f3 aa        rep stos %al,%es:(%rdi)
> >     1509: 58           pop    %rax
> >     150a: c3           ret
> > 
> > But I know you don't like it because it costs extra memory access.
> 
> Yes, that's an extra memory access. But I believe it doesn't hurt
> someone targetting -Os. In many cases, the compilers use push/pop to
> align the stack before a 'call' instruction. If they want to avoid extra
> memory access, they could have used "subq $8, %rsp" and "addq $8, %rsp".

Then "xchg %esi, %eax" is just one byte with no memory access ;-)

Willy