From mboxrd@z Thu Jan 1 00:00:00 1970 X-Msuck: nntp://news.gmane.org/gmane.linux.lib.musl.general/7074 Path: news.gmane.org!not-for-mail From: Denys Vlasenko Newsgroups: gmane.linux.lib.musl.general Subject: Re: [PATCH] x86_64/memset: use "small block" code for blocks up to 30 bytes long Date: Wed, 18 Feb 2015 10:05:31 +0100 Message-ID: References: <20150215040655.GM23507@brightrain.aerifal.cx> <20150215150313.GO23507@brightrain.aerifal.cx> <20150216173634.GA23507@brightrain.aerifal.cx> <20150217161222.GF23507@brightrain.aerifal.cx> <20150217174045.GH23507@brightrain.aerifal.cx> <20150217211247.GI23507@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 1424250371 3651 80.91.229.3 (18 Feb 2015 09:06:11 GMT) X-Complaints-To: usenet@ger.gmane.org NNTP-Posting-Date: Wed, 18 Feb 2015 09:06:11 +0000 (UTC) Cc: musl To: Rich Felker Original-X-From: musl-return-7087-gllmg-musl=m.gmane.org@lists.openwall.com Wed Feb 18 10:06:10 2015 Return-path: 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 ) id 1YO0aA-0002RZ-NQ for gllmg-musl@m.gmane.org; Wed, 18 Feb 2015 10:06:10 +0100 Original-Received: (qmail 1789 invoked by uid 550); 18 Feb 2015 09:06:08 -0000 Mailing-List: contact musl-help@lists.openwall.com; run by ezmlm Precedence: bulk List-Post: List-Help: List-Unsubscribe: List-Subscribe: Original-Received: (qmail 1737 invoked from network); 18 Feb 2015 09:06:03 -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 :cc:content-type; bh=EG/abwLfJaSf8ViPGzhLKb4Fz5j64T0uzhnCiawQ2eI=; b=DCXdXh6EAucI20AwfFpz5hIb9Rjr9iQ6KFwDDTM3IPeVZoNMBb705X2jijtDhH9sRK Ltg/1zb1H9k2qVowUsHMnApOlxc45ydLqJJjIoa84YJKWDoietIJLNDv3rh8fM1T+x+X 71gjT0E1Z4xFRdPsYEpMZXW5dRr5WWf+Lthd3NvrKl829W9SnqumawnqwFbvXSeraOJs Ua/A+DUs0eS5dOVu/HpnWcKw2ZTAQiKO9X/4GS9vZFFeoBHv8qcZjyYSqoFKcWaYKPsB 3CsdUZtovrKYRoX/zlgRG5v47/mJglFEwb1jyuC2QtZBfY/QDqEaflcsODNivAcSLHhO bvWQ== X-Received: by 10.140.23.199 with SMTP id 65mr1244832qgp.84.1424250351801; Wed, 18 Feb 2015 01:05:51 -0800 (PST) In-Reply-To: <20150217211247.GI23507@brightrain.aerifal.cx> Xref: news.gmane.org gmane.linux.lib.musl.general:7074 Archived-At: On Tue, Feb 17, 2015 at 10:12 PM, Rich Felker wrote: > For > me, I get ~1 cycle difference for a number of the small runs, but on > your test that measures rates in a way that's probably more applicable > to real-world use, there was no measureable difference one way or the > other. We had an enjoyable bike shed painting session, don't we? "Optimizing memset", a must for anyone who knows assembly :D