From owner-freebsd-current Tue Jul 29 01:42:04 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id BAA25021 for current-outgoing; Tue, 29 Jul 1997 01:42:04 -0700 (PDT) Received: from sos.freebsd.dk (sos.freebsd.dk [195.8.129.33]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id BAA25016 for ; Tue, 29 Jul 1997 01:42:00 -0700 (PDT) Received: (from sos@localhost) by sos.freebsd.dk (8.8.6/8.7.3) id KAA00348; Tue, 29 Jul 1997 10:42:05 +0200 (MEST) From: Søren Schmidt Message-Id: <199707290842.KAA00348@sos.freebsd.dk> Subject: Re: code talks: announcing EIDE bus master patches In-Reply-To: <199707290424.AAA23424@smoke.marlboro.vt.us> from john hood at "Jul 29, 97 00:24:09 am" To: cgull@smoke.marlboro.vt.us (john hood) Date: Tue, 29 Jul 1997 10:42:05 +0200 (MEST) Cc: freebsd-current@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL30 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk In reply to john hood who wrote: > Søren Schmidt writes: > > Well, I have two 4G EIDE Maxtors on my P6 :), when I get the bus master > > DMA going, it'll be a tough bunch to beat... > > It's going to go a little quicker :) > > I've been coding up bus-mastering support for EIDE drives, and I'm at > a point where I can release it to the world. I've got a patch against > 2.2.2 available and various notes for the enjoyment of the curious > kernel hacker. The wd driver is enough of a living fossil that the > patches should apply easily to any kernel since then, though. > > The code is designed to support the SFF-8038 standard for busmaster > controllers. It's generic enough to run on anything that nominally > conforms to this (which includes every current PCI chipset I've seen); > here's to hoping chipset quirks don't wreck this sublime quest. > > It's all very alpha yet, and it has been tested on exactly one system. > but if the surreal thrill of watching a system transfer 6MB/s off an > IDE disk with almost no cpu load mingled with the slight fear that > it's trashing your disk at 6MB/s appeals to you, go to: Well, doing a little surgery to it and sticking it into current, (the PCI subsystem has changed a little from 2.2.x to current) I'd say it works pretty well here too (440FX chipset Maxtor drives). I'l beat on it for a while, then commit the changes to -current. I think we should leave 2.2-STABLE out for the moment :) -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- Søren Schmidt (sos@FreeBSD.org) FreeBSD Core Team Even more code to hack -- will it ever end ..