From owner-freebsd-amd64@FreeBSD.ORG Sun Feb 13 19:25:34 2005 Return-Path: Delivered-To: freebsd-amd64@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 06E7616A4CE for ; Sun, 13 Feb 2005 19:25:34 +0000 (GMT) Received: from troutmask.apl.washington.edu (troutmask.apl.washington.edu [128.208.78.105]) by mx1.FreeBSD.org (Postfix) with ESMTP id C013043D5A for ; Sun, 13 Feb 2005 19:25:33 +0000 (GMT) (envelope-from sgk@troutmask.apl.washington.edu) Received: from troutmask.apl.washington.edu (localhost [127.0.0.1]) j1DJPOvi041877; Sun, 13 Feb 2005 11:25:24 -0800 (PST) (envelope-from sgk@troutmask.apl.washington.edu) Received: (from sgk@localhost)j1DJPOig041876; Sun, 13 Feb 2005 11:25:24 -0800 (PST) (envelope-from sgk) Date: Sun, 13 Feb 2005 11:25:24 -0800 From: Steve Kargl To: Thomas Krause -CI- Message-ID: <20050213192524.GA41820@troutmask.apl.washington.edu> References: <420F9CA1.4040207@chef-ingenieur.de> <20050213184051.GA41566@troutmask.apl.washington.edu> <420FA306.8010102@chef-ingenieur.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <420FA306.8010102@chef-ingenieur.de> User-Agent: Mutt/1.4.2.1i cc: freebsd-amd64@freebsd.org Subject: Re: -mno-sse -mno-sse2 -mno-mmx -mno-3dnow when compiling kernel X-BeenThere: freebsd-amd64@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Porting FreeBSD to the AMD64 platform List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 13 Feb 2005 19:25:34 -0000 On Sun, Feb 13, 2005 at 07:57:10PM +0100, Thomas Krause -CI- wrote: > Steve Kargl schrieb: > >On Sun, Feb 13, 2005 at 07:29:53PM +0100, Thomas Krause -CI- wrote: > > > >>Hello, > >>why are the Switches "-mno-sse -mno-sse2 -mno-mmx -mno-3dnow" are set > >>when compiling an amd64 kernel? I think, the CPU supports all thease > >>features. > >> > > > > > >AFAIK, floating point is not permitted in the kernel. Why would > >you need to use sse, sse2, mmx, or 3dnow? > > > I don't know, if thease features are useful in the kernel. But when > compiling a x86 kernel, they are not disabled. I suggest you read the gcc.info file, in particular see the description of -mfpmathi option. It seems that if you are building i386 (or equivalent 32bit processor), then these options are automatically disabled by the compiler. If you are building for amd64 (or Intel's copycat), then these options are automatically enabled by the compiler. These options are used to compile floating point code. There is no floating point code in the kernel, so the options are useless and should be turned off. > As I remember, you can compile libpng using mmx features. (I'm unsure, > if you need a mmx-optimized kernel for that.) libpng uses floating point code, so it can take advantage of mmx. The kernel's feature list for my CPU, Features=0x183f9ff shows that the CPU supports MMX, and thus libpng compile with mmx will work just fine. -- Steve