From owner-freebsd-hackers Mon May 4 23:19:13 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id XAA02896 for freebsd-hackers-outgoing; Mon, 4 May 1998 23:19:13 -0700 (PDT) (envelope-from owner-freebsd-hackers@FreeBSD.ORG) Received: from implode.root.com (implode.root.com [198.145.90.17]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id XAA02880 for ; Mon, 4 May 1998 23:19:07 -0700 (PDT) (envelope-from root@implode.root.com) Received: from implode.root.com (localhost [127.0.0.1]) by implode.root.com (8.8.5/8.8.5) with ESMTP id XAA23684; Mon, 4 May 1998 23:17:58 -0700 (PDT) Message-Id: <199805050617.XAA23684@implode.root.com> To: Bill Paul cc: hackers@FreeBSD.ORG Subject: Re: Call for testers for ThunderLAN ethernet driver In-reply-to: Your message of "Tue, 05 May 1998 01:05:51 EDT." <199805050505.BAA28835@skynet.ctr.columbia.edu> From: David Greenman Reply-To: dg@root.com Date: Mon, 04 May 1998 23:17:58 -0700 Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG >I set up the driver to use PCI memory mapping to access the >ThunderLAN's registers as opposed to using programmed I/O, which is >what both the Linux and NetBSD drivers do. There probably isn't any >particular advantage to doing it this way, but it seemed niftier somehow. Memory mapped access to the registers will be faster than PIO on P6 class processors. I don't recall all of the issues, but I believe that doing PIO causes the instruction pipeline to get flushed. Anyway, I use mapped register access in the fxp driver and I've never had a complaint. -DG David Greenman Co-founder/Principal Architect, The FreeBSD Project To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message