From owner-freebsd-hackers Tue Jul 4 02:31:08 1995 Return-Path: hackers-owner Received: (from majordom@localhost) by freefall.cdrom.com (8.6.10/8.6.6) id CAA20254 for hackers-outgoing; Tue, 4 Jul 1995 02:31:08 -0700 Received: from who.cdrom.com (who.cdrom.com [192.216.222.3]) by freefall.cdrom.com (8.6.10/8.6.6) with ESMTP id CAA20248 for ; Tue, 4 Jul 1995 02:31:07 -0700 Received: from hq.icb.chel.su (icb-rich-gw.icb.chel.su [193.125.10.34]) by who.cdrom.com (8.6.11/8.6.11) with ESMTP id CAA04423 for ; Tue, 4 Jul 1995 02:24:03 -0700 Received: from localhost (babkin@localhost) by hq.icb.chel.su (8.6.5/8.6.5) id PAA03710; Tue, 4 Jul 1995 15:15:55 +0500 From: "Serge A. Babkin" Message-Id: <199507041015.PAA03710@hq.icb.chel.su> Subject: Re: FBSD & old HDs To: bde@zeta.org.au (Bruce Evans) Date: Tue, 4 Jul 1995 15:15:54 +0500 (GMT+0500) Cc: bde@zeta.org.au, hackers@freebsd.org In-Reply-To: <199507040725.RAA14172@godzilla.zeta.org.au> from "Bruce Evans" at Jul 4, 95 05:25:27 pm X-Mailer: ELM [version 2.4 PL23] Content-Type: text Content-Length: 2160 Sender: hackers-owner@freebsd.org Precedence: bulk > >But MS-DOS worked absolutely OK on them and FBSD absolutely didn't worked > >with XEBEC drive. I think the first thing FBSD should try to read is > > DOS FORMAT marks bad sectors and doesn't use them for the file system. > FreeBSD would mark bad sectors in a similar way if you ran badsect(8). > Both methods fail a file system metadata block is bad. bad144 solves ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > this problem. It looks like you said the same thing I said: "If DOS can use this disk then FreeBSD can at least try to use this disk too". BTW as I know bad sectors do not return absolutely no data, they return data with errors and a flag that says about wrong checksum. But the message I got said that driver gets absolutely no response from drive (if I have understood it right). > >MBR, and then if there is no FBSD slice it should not try to read anything > >else (at least for the install diskette). But it got problems at the first > > FreeBSD slice'ness is determined by attempting to read the label. You can > have a label and FreeBSD file systems on slices of any type. I didn't know that. And what it does with unused slice records in MBR ? Theoretically it's possible that these record will contain garbage saying that this slice is located out of the disk and when slice code will try to determine the type of that slice it will get out of geometry and will try to read nonexisting sectors. Obviously that this scenery is possible only if there were no chechs. But anyway try to read out of disk's geometry should give an immediate error report. > >Yes, maybe. But the `subject' is slightly misleading, really in the 2nd case > >only computer was old, not the drive. IDE controllers are known to be simply > >buffer logic with absolutely no intelligence, so I think there should be no > >interference between old IDE controller and new drive. > > `Old' means 1985 MFM to me :-). When did IDE drives first come out? I > didn't own one until late 1994. XEBEC drive was produced in 1991. Serge Babkin ! (babkin@hq.icb.chel.su) ! Headquarter of Joint Stock Commercial Bank "Chelindbank" ! Chelyabinsk, Russia