From owner-freebsd-bugs  Wed Jun 14 20:13:25 1995
Return-Path: bugs-owner
Received: (from majordom@localhost)
          by freefall.cdrom.com (8.6.10/8.6.6) id UAA17523
          for bugs-outgoing; Wed, 14 Jun 1995 20:13:25 -0700
Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.34])
          by freefall.cdrom.com (8.6.10/8.6.6) with ESMTP id UAA17517
          for <bugs@FreeBSD.org>; Wed, 14 Jun 1995 20:13:20 -0700
Received: (from bde@localhost) by godzilla.zeta.org.au (8.6.9/8.6.9) id NAA27850; Thu, 15 Jun 1995 13:09:11 +1000
Date: Thu, 15 Jun 1995 13:09:11 +1000
From: Bruce Evans <bde@zeta.org.au>
Message-Id: <199506150309.NAA27850@godzilla.zeta.org.au>
To: hlew@genome.Stanford.EDU, terry@cs.weber.edu
Subject: Re: 2.05R panics on boot
Cc: bugs@FreeBSD.org
Sender: bugs-owner@FreeBSD.org
Precedence: bulk

>The DOS MBR:

>,------.---------------------.-------------------.
>| JUMP | DOS PARTITION TABLE | DOS MBR BOOT CODE |
>`------'---------------------'-------------------'

>The DM MBR:

>.------.-------------------------.---------------.
>| JUMP | FAKE DOS PARTIOTN TABLE | DM MAGIC CODE |
>`------'-------------------------'---------------'

There is code between the JUMP and the PARTITION TABLE too.


>The PBR (partition boot record; second stage boot):

>> Do I subtract 64 from the sector number while in the FreeBSD install 
>> program under disk partitioning or are you referring to a DOS utility to 
>> do this?

A driver should subtract it.

>OK, DM has a boot record it writes on the drives that's 63 or so sectors
>long (depends on the version, etc.).

FreeBSD assumes the fixed value 63 (for DM partition type 84).

>Now what is "OS-BS", "BootEasy", "LILO", the"OS/2 Boot Manager", etc. going
>to do?
>...
>OK, now say you boot off a BSD floppy.

>You are still screwed, because even if it had loaded the DM MBR and patched
>INT 13, except for actually loading, the BSD code doesn't use the BIOS
>at all.  Not that you could load the DM MBR anyway, if you boot off a
>floppy.

>Now BSD's new slice code knows about the DM MBR.  But the installation
>of the MBR code does not.

Yes it does (modulo bugs).

Only one case can't possibly work: loading the bootstrap off a floppy
and then attempting to boot from a hard disk (e.g., by typing
sd(1,a)/kernel).  The boot code doesn't know about DM, so it won't
work if DM needs to be loaded.

Bruce