From mboxrd@z Thu Jan  1 00:00:00 1970
X-Msuck: nntp://news.gmane.org/gmane.linux.lib.musl.general/6978
Path: news.gmane.org!not-for-mail
From: Denys Vlasenko <vda.linux@googlemail.com>
Newsgroups: gmane.linux.lib.musl.general
Subject: Re: [PATCH 1/2] x86_64/memset: simple optimizations
Date: Wed, 11 Feb 2015 02:07:23 +0100
Message-ID: <CAK1hOcM+0OHph8Misqrgt_QspoPox8ybqrEmsCDUiEn76CCORQ@mail.gmail.com>
References: <1423589457-8407-1-git-send-email-vda.linux@googlemail.com>
 <20150210205047.GK23507@brightrain.aerifal.cx> <CAK1hOcNPLxqvnDAsX3hYHkKm5hwLDLDjx10yysEqR8EZMdCFcA@mail.gmail.com>
 <20150210213756.GM23507@brightrain.aerifal.cx>
Reply-To: musl@lists.openwall.com
NNTP-Posting-Host: plane.gmane.org
Mime-Version: 1.0
Content-Type: text/plain; charset=UTF-8
X-Trace: ger.gmane.org 1423616879 6121 80.91.229.3 (11 Feb 2015 01:07:59 GMT)
X-Complaints-To: usenet@ger.gmane.org
NNTP-Posting-Date: Wed, 11 Feb 2015 01:07:59 +0000 (UTC)
To: musl <musl@lists.openwall.com>
Original-X-From: musl-return-6991-gllmg-musl=m.gmane.org@lists.openwall.com Wed Feb 11 02:07:59 2015
Return-path: <musl-return-6991-gllmg-musl=m.gmane.org@lists.openwall.com>
Envelope-to: gllmg-musl@m.gmane.org
Original-Received: from mother.openwall.net ([195.42.179.200])
	by plane.gmane.org with smtp (Exim 4.69)
	(envelope-from <musl-return-6991-gllmg-musl=m.gmane.org@lists.openwall.com>)
	id 1YLLmX-0002Dt-Ad
	for gllmg-musl@m.gmane.org; Wed, 11 Feb 2015 02:07:57 +0100
Original-Received: (qmail 30116 invoked by uid 550); 11 Feb 2015 01:07:55 -0000
Mailing-List: contact musl-help@lists.openwall.com; run by ezmlm
Precedence: bulk
List-Post: <mailto:musl@lists.openwall.com>
List-Help: <mailto:musl-help@lists.openwall.com>
List-Unsubscribe: <mailto:musl-unsubscribe@lists.openwall.com>
List-Subscribe: <mailto:musl-subscribe@lists.openwall.com>
Original-Received: (qmail 30105 invoked from network); 11 Feb 2015 01:07:55 -0000
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;
        d=googlemail.com; s=20120113;
        h=mime-version:in-reply-to:references:from:date:message-id:subject:to
         :content-type;
        bh=emQW+R6ZNOWNDTo2Go1I1MvOY8EkW5D76x8xfJxslEw=;
        b=oYnDdrA2ALhsa+30mzv0irFNb7vbYNpkA4oOOaEVi19phr3h7yOFt/4piu4TNS9Mkz
         lficA8IrP8+CIBZaN4bd2RCYiDLanYpzJDCX+LVoNo8NAp7oKrTLoJdWlxSzS3jGEMYa
         cnLTV1+wvVb4LjL/ABVXoGAqGMLnhfcnLePTWXjsZQRbqzvlfKS2Vfu1cKKIuEKP23w9
         ldBnRwQI6rLYAgOVsyFuBxJ2p/xKgCwndfANxFMnXPUdRj63YKFBfHYR+yKozrsxc5kL
         XyrcrCZ1xVcIajqUL+K750PmFQ+OSKqC/69aRzFkfLasfZZwdcqHpZ9HCpiY4oLyLc8l
         Lijw==
X-Received: by 10.140.33.201 with SMTP id j67mr22946448qgj.10.1423616863209;
 Tue, 10 Feb 2015 17:07:43 -0800 (PST)
In-Reply-To: <20150210213756.GM23507@brightrain.aerifal.cx>
Xref: news.gmane.org gmane.linux.lib.musl.general:6978
Archived-At: <http://permalink.gmane.org/gmane.linux.lib.musl.general/6978>

On Tue, Feb 10, 2015 at 10:37 PM, Rich Felker <dalias@libc.org> wrote:
> OK. Based on some casual testing on my Celeron 847:
>
> - For small sizes, your patches make significant improvement, 20-30%.
>
> - For rep stosq path, the improvement is minimal (roughly 1-2 cycles).
>
> - Using 32-bit imul instead of 64-bit makes no difference at all.

That's because Celeron 847 is a Sandy Bridge CPU. Only Intel's "big"
CPUs starting from Nehalem have fast (and large in transistor count)
integer multiplier capable of 3-cycle 64-bit multiply.

Many other CPUs are worse, even Intel ones: Atoms are 13-cycle (!),
Silvermont: 5 cycles. AMD's Bulldozers: 6 cycles, Bobcat: 6-7, Jaguar:
6, K10: 4 cycles.

32-bit imul is 3 or 4 cycles on all these CPUs (well, Atom has 5).