From owner-freebsd-hackers Thu Mar 21 02:01:40 1996 Return-Path: owner-hackers Received: (from root@localhost) by freefall.freebsd.org (8.7.3/8.7.3) id CAA14122 for hackers-outgoing; Thu, 21 Mar 1996 02:01:40 -0800 (PST) Received: from phaeton.artisoft.com (phaeton.Artisoft.COM [198.17.250.211]) by freefall.freebsd.org (8.7.3/8.7.3) with SMTP id CAA14110 for ; Thu, 21 Mar 1996 02:01:37 -0800 (PST) Received: (from terry@localhost) by phaeton.artisoft.com (8.6.11/8.6.9) id CAA29401; Thu, 21 Mar 1996 02:51:32 -0700 From: Terry Lambert Message-Id: <199603210951.CAA29401@phaeton.artisoft.com> Subject: Re: Adding a damn 2nd disk To: bde@zeta.org.au (Bruce Evans) Date: Thu, 21 Mar 1996 02:51:32 -0700 (MST) Cc: andreas@knobel.gun.de, terry@lambert.org, dave@kachina.jetcafe.org, freebsd-hackers@FreeBSD.org In-Reply-To: <199603210258.NAA15921@godzilla.zeta.org.au> from "Bruce Evans" at Mar 21, 96 01:58:23 pm X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-hackers@FreeBSD.org X-Loop: FreeBSD.org Precedence: bulk > >Stop right there. > > >the concept of "disktab" is fundamentally flawed. It assumes that > >it is not possible to determine the disk size from the controller, > >and it assumes uniform boundry recording for seek optimization. > > Well, it isn't always possible. It fails for old MFM drives and for > floppies (floppies may be unformatted, or formatted with a nonstandard > number of sectors/track ...). /etc/disktab is still useful for holding > the defaults for such mouldy drives. Yick. Formatting should take care of this... the floppy driver should auto-detect format. The format has to be specified and done seperately anyway. The driver for old MFM drives should read the CMOS drive table. > >The seek optimization is, in reality, useless because ZBR media > >makes it very difficult (without SCSI II extended queries) to > >determine the real cylinder boundries to do the optimization. > > The old drives don't use ZBR. The optimization is broken in another way > for floppies - since the number of sectors/cylinder is only one or wwo > times larger than the smallest possible (ufs) block size and not a > multiple of the block size, many blocks span cylinder boundaries. > > >This baically leaves default sector sparing settings, sector counts > >designed to be on (no longer applicable) cylinder boundries, and > >slice geometries that should be interactively determined instead > >of frozen in an obsolete data file. > > >The disktab should go. > > I see that you have sold your stock of ESDI drives :-). No, but I see that the slice code forces a translated world view (fake cylinder boundries) on me pretty much anyway, unless I go to an extrordinary amount of effort. Now that you mention it, I would like to see code written for ZBR drives, at least the ones that could be queried with SCSI II. 8-). Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers.