From owner-freebsd-current  Mon Nov 11 05:08:21 1996
Return-Path: owner-current
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.5/8.7.3) id FAA12708
          for current-outgoing; Mon, 11 Nov 1996 05:08:21 -0800 (PST)
Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19])
          by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id FAA12687;
          Mon, 11 Nov 1996 05:08:10 -0800 (PST)
Received: (from bde@localhost) by godzilla.zeta.org.au (8.7.6/8.6.9) id AAA13000; Tue, 12 Nov 1996 00:04:59 +1100
Date: Tue, 12 Nov 1996 00:04:59 +1100
From: Bruce Evans <bde@zeta.org.au>
Message-Id: <199611111304.AAA13000@godzilla.zeta.org.au>
To: 100120.3121@CompuServe.COM, bde@zeta.org.au, Tor.Egge@idt.ntnu.no
Subject: Re: Follow on Atlantis board
Cc: freebsd-current@freebsd.org, freebsd-stable@freebsd.org
Sender: owner-current@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

I wrote:
>Good work.  The problem is almost certainly in get_diskinfo().  It
>is documented to clobber ES:DI in some manuals:
>
>Phoenix "System BIOS ...":
>	ES:DI = Address of Fixed Disk Parameters table
>Van Gilluwe "The Undocumented PC":
>	This function does _not_ return a pointer to the drive type
>	table as is commonly indicated in many other technical
>	references.
>Ralf Brown's Interrupt List version 45:
>	(PC,XT286,CONV,PS,ESDI,SCSI):
>	ES:DI -> drive parameter table (floppies only)

The first 2 references have a separate section for floppies.  They
agree with the 3rd reference for floppies.  So the floppy case is
always broken, and the hard disk case is broken if BIOS implements
what the 1st reference says.

Bruce