From mboxrd@z Thu Jan 1 00:00:00 1970 X-Msuck: nntp://news.gmane.org/gmane.linux.lib.musl.general/1515 Path: news.gmane.org!not-for-mail From: Solar Designer Newsgroups: gmane.linux.lib.musl.general Subject: Re: crypt_blowfish integration, optimization Date: Sat, 11 Aug 2012 02:52:44 +0400 Message-ID: <20120810225244.GA31063@openwall.com> References: <20120808214855.GL27715@brightrain.aerifal.cx> <20120809033613.GA24926@openwall.com> <20120809072940.GA26288@openwall.com> <20120809105348.GA27361@openwall.com> <20120809214654.GU27715@brightrain.aerifal.cx> <20120809222103.GA29365@openwall.com> <20120809223258.GW27715@brightrain.aerifal.cx> <20120810171803.GB29839@openwall.com> <20120810180814.GB27715@brightrain.aerifal.cx> Reply-To: musl@lists.openwall.com NNTP-Posting-Host: plane.gmane.org Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Trace: dough.gmane.org 1344639179 2304 80.91.229.3 (10 Aug 2012 22:52:59 GMT) X-Complaints-To: usenet@dough.gmane.org NNTP-Posting-Date: Fri, 10 Aug 2012 22:52:59 +0000 (UTC) To: musl@lists.openwall.com Original-X-From: musl-return-1516-gllmg-musl=m.gmane.org@lists.openwall.com Sat Aug 11 00:52:58 2012 Return-path: 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 ) id 1Szy4f-000464-CJ for gllmg-musl@plane.gmane.org; Sat, 11 Aug 2012 00:52:57 +0200 Original-Received: (qmail 3571 invoked by uid 550); 10 Aug 2012 22:52:46 -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 3561 invoked from network); 10 Aug 2012 22:52:46 -0000 Content-Disposition: inline In-Reply-To: <20120810180814.GB27715@brightrain.aerifal.cx> User-Agent: Mutt/1.4.2.3i Xref: news.gmane.org gmane.linux.lib.musl.general:1515 Archived-At: On Fri, Aug 10, 2012 at 02:08:14PM -0400, Rich Felker wrote: > Yes, -O3 includes -funroll-loops, That's good. BTW, for the version of code I posted last, going from -O2 to -O3 with gcc 3 actually slows things down by about 8% on 32-bit x86. > which is intelligent about choosing which loops to unroll. Yes, except that it can't know if the function is performance-critical or not. Alexander