Obsoleting SBC-2 functions

Pat LaVarre p.lavarre at IEEE.org
Tue Nov 18 16:27:23 PST 2003

* From the T10 Reflector (t10 at t10.org), posted by:
* Pat LaVarre <p.lavarre at ieee.org>
>  > > following will be obsoleted from the SBC-2
>  > > standard unless someone comes forward with a
>  > > justification as to why they should not be
>  > > obsoleted:
>  > >
>  > > ... Flexible Disk mode page ...
> Slander or truth I do not know, but I hear wrong-headed PC Boot BIOS
> still shipping today fetch that page to help guess ...

Conversation continues in the semi-private space of usb dwg msc.  For me
I only see a clear right to forward to t10 my own contributions.  My
contributions include:


De facto people are shipping BIOS that refuse to talk PDT x00 HDD/Flash
unless those devices yield a mode page x05 H:S:C guess of the fictional
C:H:S geometry.

In this sense, mode page x05 is a "requirement".

Our only trouble now is that this requirement makes no sense.  The
device itself has no opinion over what fictional C:H:S geometry host
folk try to use to obscure lba's.  The device itself has only lba's.

BIOS folk who like seeing this requirement in place might wish to find
the time to make this requirement understood.  Making this requirement
understood increases the odds of people not ignoring [it], and correctly
implementing [it], and [not obsoleting it from SBC-2] ...


BIOS folk and device folk speak together!  History is made!  Hi! 


>From where do the BIOS folk suppose the drive folk get a C:H:S geometry?

The USB drive has only an LBA mode.  We cannot ask the USB drive to read
or write a C:H:S address.  The USB drive has no idea what fictional
C:H:S geometry breaks in a PC least often.


So the BIOS does not need mode page x05 except when the disk capacity is
below about 528MB?

In practice BIOS require mode page x05 more often?

What about if no disk is present at boot?  Do we need mode page x05 if
the drive can ever accept any small disk?


we care about more than the code of the C:H:S = 0:0:1 "partition" sector
and/or the code of the "boot" sector that starts the boot volume.  We
also cannot load MS-DOS without guessing a fictional C:H:S geometry.


Was there ever a standard for any range of capacity?

ATA HDD C:H:S came from ATA op xEC Identify?

>From where did floppy C:H:S come?

What's the place of the H:S = x40:20 SCSI boot standard for small
capacities that now is the H:S=xFF:3F SCSI boot standard for large


What happens if the disk is slow to spin up?


PC BIOS has a legacy of exporting a public interface, specifically the
INT13h interface, which encrypts lba's by factoring them using a
fictional C:H:S geometry that came from somewhere else.

It's the "somewhere else" that breaks the decryption when we no longer
have ATA op xEC Identify for hard drives, and something equally good for
floppy drives.


in concept, this is a legacy issue only.

In practice, some BIOS refuse to try to boot from devices that lack mode
page x05.  Then we cannot even try to discover if we're running a modern
O.S. rather than a legacy O.S.


* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to majordomo at t10.org

More information about the T10 mailing list