From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 22299 invoked by alias); 18 Jul 2002 20:00:12 -0000 Mailing-List: contact gcc-patches-help@gcc.gnu.org; run by ezmlm Precedence: bulk List-Archive: List-Post: List-Help: Sender: gcc-patches-owner@gcc.gnu.org Received: (qmail 22273 invoked from network); 18 Jul 2002 20:00:10 -0000 Received: from unknown (HELO atrey.karlin.mff.cuni.cz) (195.113.31.123) by sources.redhat.com with SMTP; 18 Jul 2002 20:00:10 -0000 Received: by atrey.karlin.mff.cuni.cz (Postfix, from userid 4018) id 40D714FE03; Thu, 18 Jul 2002 22:00:12 +0200 (CEST) Date: Thu, 18 Jul 2002 13:06:00 -0000 From: Jan Hubicka To: Toon Moene Cc: Jan Hubicka , Richard Henderson , gcc-patches@gcc.gnu.org Subject: Re: cprop "fix" Message-ID: <20020718200012.GC10794@atrey.karlin.mff.cuni.cz> References: <20020713184902.GD1909@atrey.karlin.mff.cuni.cz> <20020717091651.A14260@redhat.com> <20020717163352.GG9922@atrey.karlin.mff.cuni.cz> <20020717163535.GH9922@atrey.karlin.mff.cuni.cz> <20020717104415.B14279@redhat.com> <20020718073825.GI9922@atrey.karlin.mff.cuni.cz> <20020718082108.GL9922@atrey.karlin.mff.cuni.cz> <3D370F8A.2BEC99D5@moene.indiv.nluug.nl> <20020718195351.GA10794@atrey.karlin.mff.cuni.cz> <3D371DD3.B61EA34C@moene.indiv.nluug.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <3D371DD3.B61EA34C@moene.indiv.nluug.nl> User-Agent: Mutt/1.3.28i X-SW-Source: 2002-07/txt/msg00976.txt.bz2 > Jan Hubicka wrote: > > > It does basically same as 386 version - slightly more PRE improvements > > as PRE does not have to work around parallels as i386 compiler does. > > > > I can't benchmark, as the machine is busy by something else :) > > But the patch has passed full bootstrap and regtesting on that machine > > as well, so hope the try_replace_reg works as supposed to now. > > Hmmm, yes, thanks, but are you going to commit the stuff, or am I better > off just aplying it to the current trunk and see what it buys us ? I am waiting for Richard's "OK" if possible :) I have the SPEC results with and without patch as I noted earlier, I can dig out the full table. It brings about 1% improvement to SPECint and 3% to SPECfp in the second case all comming from single test, so I think it is quite effective. Honza > > Thanks in advance, > > -- > Toon Moene - mailto:toon@moene.indiv.nluug.nl - phoneto: +31 346 214290 > Saturnushof 14, 3738 XG Maartensdijk, The Netherlands > Maintainer, GNU Fortran 77: http://gcc.gnu.org/onlinedocs/g77_news.html > Join GNU Fortran 95: http://g95.sourceforge.net/ (under construction)