From owner-freebsd-hackers Fri May 7 15:25:17 1999 Delivered-To: freebsd-hackers@freebsd.org Received: from herring.nlsystems.com (nlsys.demon.co.uk [158.152.125.33]) by hub.freebsd.org (Postfix) with ESMTP id E654B14CA1 for ; Fri, 7 May 1999 15:25:09 -0700 (PDT) (envelope-from dfr@nlsystems.com) Received: from localhost (dfr@localhost) by herring.nlsystems.com (8.9.3/8.8.8) with ESMTP id XAA18811; Fri, 7 May 1999 23:25:15 +0100 (BST) (envelope-from dfr@nlsystems.com) Date: Fri, 7 May 1999 23:25:15 +0100 (BST) From: Doug Rabson To: Andrew Reilly Cc: Zach Brown , Matthew Dillon , freebsd-hackers@freebsd.org Subject: Re: Pentium-III and FreeBSD? In-Reply-To: <19990508080842.A11385@gurney.reilly.home> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG On Sat, 8 May 1999, Andrew Reilly wrote: > On Fri, May 07, 1999 at 11:15:53AM -0400, Zach Brown wrote: > > > In regards to FP: The best place for extreme FP optimization is in > > > a high level FP library, not in native compiler-produced code. The > > > > yes. > > That's pretty much all I really want to do, but the catch with > the P-III is that the new SIMD registers do _not_ map over the > fp registers, the way MMX did. They are new processor state. > > So we need OS support to save and restore them on context > switches. Intel supplies a "patch" for NT that does this, > appently. > > Should be an interesting introduction to the bottom levels of > the kernel. If you fix this, I can review it. I more or less promised this support to the XFree86 folks if they started using the extra state. You should also consider the AMD 3DNow! stuff which also has extra state I think. -- Doug Rabson Mail: dfr@nlsystems.com Nonlinear Systems Ltd. Phone: +44 181 442 9037 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message