From mboxrd@z Thu Jan  1 00:00:00 1970
X-Msuck: nntp://news.gmane.org/gmane.linux.lib.musl.general/3859
Path: news.gmane.org!not-for-mail
From: Luca Barbato <lu_zero@gentoo.org>
Newsgroups: gmane.linux.lib.musl.general
Subject: Re: Optimized C memcpy
Date: Thu, 08 Aug 2013 15:17:18 +0200
Message-ID: <52039A5E.9050503@gentoo.org>
References: <20130807182123.GA17670@brightrain.aerifal.cx> <1375966759.14128.7503991.4EB49755@webmail.messagingengine.com> <1375967031.16501.7506555.32466BB3@webmail.messagingengine.com>
Reply-To: musl@lists.openwall.com
NNTP-Posting-Host: plane.gmane.org
Mime-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit
X-Trace: ger.gmane.org 1375967768 27626 80.91.229.3 (8 Aug 2013 13:16:08 GMT)
X-Complaints-To: usenet@ger.gmane.org
NNTP-Posting-Date: Thu, 8 Aug 2013 13:16:08 +0000 (UTC)
To: musl@lists.openwall.com
Original-X-From: musl-return-3863-gllmg-musl=m.gmane.org@lists.openwall.com Thu Aug 08 15:16:10 2013
Return-path: <musl-return-3863-gllmg-musl=m.gmane.org@lists.openwall.com>
Envelope-to: gllmg-musl@plane.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-3863-gllmg-musl=m.gmane.org@lists.openwall.com>)
	id 1V7Q4X-0005oD-Cv
	for gllmg-musl@plane.gmane.org; Thu, 08 Aug 2013 15:16:09 +0200
Original-Received: (qmail 25619 invoked by uid 550); 8 Aug 2013 13:16:09 -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 25608 invoked from network); 8 Aug 2013 13:16:08 -0000
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:17.0) Gecko/20130411 Thunderbird/17.0.5
In-Reply-To: <1375967031.16501.7506555.32466BB3@webmail.messagingengine.com>
Xref: news.gmane.org gmane.linux.lib.musl.general:3859
Archived-At: <http://permalink.gmane.org/gmane.linux.lib.musl.general/3859>

On 08/08/13 15:03, Andrew Bradford wrote:
> Sorry, I now see why 4 byte blocks won't work due to the misalignment,
> but 8 or 16 seem like they should be possible.
> Is it just the evaluation of the for loop being expensive that's trying
> to be avoided?

Probably 32bytes uses cachelines better on average.

lu