From owner-freebsd-hardware Tue Dec 31 19: 9:43 2002 Delivered-To: freebsd-hardware@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 236A437B401 for ; Tue, 31 Dec 2002 19:09:42 -0800 (PST) Received: from level.uwaterloo.ca (level.uwaterloo.ca [129.97.50.219]) by mx1.FreeBSD.org (Postfix) with ESMTP id 52C1243E4A for ; Tue, 31 Dec 2002 19:09:41 -0800 (PST) (envelope-from bruce@engmail.uwaterloo.ca) Received: from level.uwaterloo.ca (localhost [127.0.0.1]) by level.uwaterloo.ca (8.12.3/8.12.3) with ESMTP id h0138M5n050101; Tue, 31 Dec 2002 22:08:22 -0500 (EST) (envelope-from bruce@engmail.uwaterloo.ca) Received: (from www@localhost) by level.uwaterloo.ca (8.12.3/8.12.3/Submit) id h0138MZb050100; Tue, 31 Dec 2002 22:08:22 -0500 (EST) X-Authentication-Warning: level.uwaterloo.ca: www set sender to bruce@engmail.uwaterloo.ca using -f Received: from 65.93.103.72 ( [65.93.103.72]) as user bruce@engmail.uwaterloo.ca by www.nexusmail.uwaterloo.ca with HTTP; Tue, 31 Dec 2002 22:08:22 -0500 Message-ID: <1041390502.3e125ba66487e@www.nexusmail.uwaterloo.ca> Date: Tue, 31 Dec 2002 22:08:22 -0500 From: Bruce Campbell To: Mike Tancsa Cc: freebsd-hardware@freebsd.org Subject: Re: ata "fallback to PIO mode" on dual processor AMD systems References: <1041368236.3e1204ac45da5@www.nexusmail.uwaterloo.ca> <025701c2b112$ddfbf580$1200a8c0@gsicomp.on.ca> In-Reply-To: MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit User-Agent: Internet Messaging Program (IMP) 3.1 / FreeBSD-4.6.2 X-Originating-IP: 65.93.103.72 Sender: owner-freebsd-hardware@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Curiously, the 4 Intel systems that are working without problems are UDMA33: atapci0: port 0xb800-0xb80f at device 2.5 on pci0 while the problematic AMD systems are UDMA100: atapci0: port 0xb800-0xb80f at device 7.1 on pci0 The removeable drive bays use a short piece of 80 conductor cable to go from a large centronics type connector to the drive. Maybe it is a stub matching problem. The boxes that these enclosures came in had a sticker which said "UDMA33/66/100" but under the sticker it just said "UDMA33/66". Maybe I was ripped off. Quoting Mike Tancsa : > On Tue, 31 Dec 2002 16:49:57 -0500, in sentex.lists.freebsd.hardware you > wrote: > > > > All drives are installed in removeable drive bays. I don't have the > make/model > > on hand right now. They were $19 CAD. ($13USD). The low cost makes > > me suspicious now, but... > > > > possibly. If they are made by Startech, you will get the occasional bad > one that will show similar behaviour. We make heavy use of 3ware cards on > our FreeBSD servers and they are much more sensitive to bad trays in our > experience. We go through quite a few as we use pullouts on all our boxes > at the office and for customer servers we build. If possible, run the > problem drive directly on the cable, or move the drives around to see if > your problems follow the drives or the trays. > > ---Mike > Mike Tancsa (mike@sentex.net) > http://www.sentex.net/mike > -- Bruce Campbell Engineering Computing CPH-2374B University of Waterloo (519)888-4567 ext 5889 ---------------------------------------- This mail sent through www.mywaterloo.ca To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hardware" in the body of the message