From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 29069 invoked by alias); 4 Jun 2008 20:02:06 -0000 Received: (qmail 29055 invoked by uid 22791); 4 Jun 2008 20:02:05 -0000 X-Spam-Check-By: sourceware.org Received: from smtp-out.google.com (HELO smtp-out.google.com) (216.239.33.17) by sourceware.org (qpsmtpd/0.31) with ESMTP; Wed, 04 Jun 2008 20:01:43 +0000 Received: from spaceape23.eur.corp.google.com (spaceape23.eur.corp.google.com [172.28.16.75]) by smtp-out.google.com with ESMTP id m54K1aQo028434 for ; Wed, 4 Jun 2008 21:01:36 +0100 Received: from yw-out-1718.google.com (ywa6.prod.google.com [10.192.1.6]) by spaceape23.eur.corp.google.com with ESMTP id m54K1YAM005075 for ; Wed, 4 Jun 2008 21:01:36 +0100 Received: by yw-out-1718.google.com with SMTP id 6so144235ywa.84 for ; Wed, 04 Jun 2008 13:01:33 -0700 (PDT) Received: by 10.150.191.10 with SMTP id o10mr682748ybf.56.1212609693836; Wed, 04 Jun 2008 13:01:33 -0700 (PDT) Received: by 10.150.197.17 with HTTP; Wed, 4 Jun 2008 13:01:33 -0700 (PDT) Message-ID: <65dd6fd50806041301q63536cb4u9c541bfa343c5cb0@mail.gmail.com> Date: Wed, 04 Jun 2008 20:02:00 -0000 From: "Ollie Wild" To: "Chris Lattner" Subject: Re: [whopr] Design/implementation alternatives for the driver and WPA Cc: "Kenneth Zadeck" , "Diego Novillo" , gcc@gcc.gnu.org, "Jan Hubicka" , "Rafael Espindola" , "Robert Hundt" In-Reply-To: <89069638-6D2B-4AE6-ACB3-99A2B09091BA@apple.com> MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline References: <96A87E23-CD76-41E3-A4BD-DD399C847F17@apple.com> <4846B475.8030907@naturalbridge.com> <8AA7CF45-8855-49EE-A4FD-5C4441639203@apple.com> <65dd6fd50806041223l1871ecfbh384aa175c3da0645@mail.gmail.com> <89069638-6D2B-4AE6-ACB3-99A2B09091BA@apple.com> X-IsSubscribed: yes Mailing-List: contact gcc-help@gcc.gnu.org; run by ezmlm Precedence: bulk List-Id: List-Archive: List-Post: List-Help: Sender: gcc-owner@gcc.gnu.org X-SW-Source: 2008-06/txt/msg00079.txt.bz2 On Wed, Jun 4, 2008 at 12:33 PM, Chris Lattner wrote: > > Right, I understand that you design "stacks" on LTO. It just seems strange > to work on the advanced stuff before the basic GCC LTO stuff is close to > being useful. To some degree, we're scratching our own itch here. Basic LTO doesn't help us much. Obviously, though, we want to implement this in a way which is generally useful to the external community. The scalability techniques required to work with distcc are different from the techniques which are useful on a single machine. > I don't know anything about the implementation of the HP or Intel LTO > implementation, but it sounds like there is much room for improvement there. > With LLVM LTO, we see a compile-time slowdown on the order of 30-50% switch > from O3 to O4, not an order of magnitude. There is also still much room for > improvement in the LLVM implementation of course. I think we're working from different baselines. We use distributed techniques for compiling individual .o files. With a tool like distcc, you can get something on the order of 20x speedup. Linking becomes 20% or more of total execution time. LTO *is* an order of magnitude increase compared to a basic link operation. Ollie