From owner-freebsd-current@FreeBSD.ORG Fri Dec 12 13:08:47 2003 Return-Path: Delivered-To: freebsd-current@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id ECAA116A4CE for ; Fri, 12 Dec 2003 13:08:47 -0800 (PST) Received: from paladin.fortunaty.net (fortunaty.net [217.160.129.175]) by mx1.FreeBSD.org (Postfix) with SMTP id 2293643D31 for ; Fri, 12 Dec 2003 13:08:46 -0800 (PST) (envelope-from ah@paladin.fortunaty.net) Received: (qmail 17777 invoked by uid 501); 12 Dec 2003 21:08:44 -0000 Date: 12 Dec 2003 21:08:44 -0000 Message-ID: <20031212210844.17776.qmail@paladin.fortunaty.net> User-Agent: Emai/0.0.2p1 Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Disposition: inline References: <1071241208.970.49.camel@localhost>,<200312121823.22352.postfix@sendmail.ru> In-Reply-To: <200312121823.22352.postfix@sendmail.ru> From: Andreas Hauser To: toxa cc: Ryan Sommers cc: freebsd-current@freebsd.org Subject: Re: CURRENT state of acpi and -O'ptimisations in buildworld X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Discussions about the use of FreeBSD-current List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 12 Dec 2003 21:08:48 -0000 postfix wrote @ Fri, 12 Dec 2003 18:23:22 +0300: > On Friday 12 December 2003 18:00, Ryan Sommers wrote: > > > > > > The second is about buildworld. Usually I use this opimisaton flags to > > > build world and userland (putting them to /etc/make.conf): > > > > > > CFLAGS= -O3 -pipe -ffast-math -funroll-loops -mno-sse2 -march=pentium4 > > > COPTFLAGS= -O2 -pipe -march=pentium4 -ffast-math -funroll-loops > > > CPUTYPE= p4 > > > > Building with any optimization higher then -O is not supported. > > Does it mean that -O3 really translated into -O then making buildworld OR that > with any optimisation higher than -O you may have _success_ with buildworld > BUT binaries may be _broken_? I have used -O2 and -O3 with 5-CURRENT for many > months (cvsuping to -current since 5.1-release) without any problem until > tonight. There was a time with gcc when optimizations had a lot of problems. These seem over now and i like to work on getting at least -Os (which kinda implies -O2) to work. But still most testing probably is done with -O only, so this therefore will be more stable. Also when you try to bring attention to problems with optimizations try to see wether the source code or the compiler is the problem. If it's the code your chances of getting help will be higher. DES committed some patches for libpam recently, these should get you a bit farther with buildworld. Not through though. Currently i am working on fixing /usr.sbin/kldxref. Maybe you like to offer to test the patches as soon as they make a full buildworld go? Andy