From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 14653 invoked by alias); 10 Sep 2008 15:46:39 -0000 Received: (qmail 14530 invoked by uid 22791); 10 Sep 2008 15:46:37 -0000 X-Spam-Check-By: sourceware.org Received: from atrey.karlin.mff.cuni.cz (HELO atrey.karlin.mff.cuni.cz) (195.113.26.193) by sourceware.org (qpsmtpd/0.31) with ESMTP; Wed, 10 Sep 2008 15:45:42 +0000 Received: by atrey.karlin.mff.cuni.cz (Postfix, from userid 4018) id 1D37FF0137; Wed, 10 Sep 2008 17:45:39 +0200 (CEST) Date: Wed, 10 Sep 2008 15:57:00 -0000 From: Jan Hubicka To: "H.J. Lu" Cc: Jan Hubicka , Jan Hubicka , gcc-patches@gcc.gnu.org Subject: Re: Fix SPEC2000 GCC regression Message-ID: <20080910154538.GE19183@atrey.karlin.mff.cuni.cz> References: <20080905112900.GM3195@kam.mff.cuni.cz> <6dc9ffc80809100710ue96b527o3e553bedc93eb410@mail.gmail.com> <20080910144117.GD19183@atrey.karlin.mff.cuni.cz> <6dc9ffc80809100801o6db96483m83bfc9818cf83b3b@mail.gmail.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <6dc9ffc80809100801o6db96483m83bfc9818cf83b3b@mail.gmail.com> User-Agent: Mutt/1.5.13 (2006-08-11) X-IsSubscribed: yes Mailing-List: contact gcc-patches-help@gcc.gnu.org; run by ezmlm Precedence: bulk List-Id: List-Archive: List-Post: List-Help: Sender: gcc-patches-owner@gcc.gnu.org X-SW-Source: 2008-09/txt/msg00827.txt.bz2 > On Wed, Sep 10, 2008 at 7:41 AM, Jan Hubicka wrote: > >> On Fri, Sep 5, 2008 at 4:29 AM, Jan Hubicka wrote: > >> > Hi, > >> > we have regression in SPEC2000 GCC. The problem is schedule_loop that > >> > contains memset calls of large blocks. Those calls are identified by > >> > profile code as cold, that is not wrong, but since blocks are large, we > >> > still spend important amount of time. > >> > > >> > This patch makes hot/cold heuristics more conservative when it comes to > >> > string operations of large or unknown size. Now we optimize for size > >> > only memsets in cold functions, not based on BB profile. > >> > > >> > Bootstrapped/regtested i686-linux. > >> > > >> > * i386.c (decide_alg): Be more conservative about optimizing for size. > >> > > >> > * gcc.target/i386/cold-attribute-1.c: Update testcase. > >> > >> Hi Honza, > >> > >> This patch caused 10% performance drop on 176.gcc in SPEC CPU 20006 > >> at -O2 -ffast-math on Intel Core 2. > > > > This mostl likely mean that your glibc has slow memset implementation. > > I saw similar drop on SPEC2000 GCC and Debian machines used by > > compilation farm. I was told that FSF glibc was finally updated with > > new GCC so hopefully this problem is solved now. ^^^ strongops :) > > > > Our (Opteron based) SPEC2006 tester seems fine > > http://gcc.opensuse.org/SPEC/CINT/sb-balakirew-head-64-2006/403_gcc_big.png > > > > I will verify it on Fedora 9 which has glibc 2.8. Great, thanks! Honza