badtrack and *big* disks

[Stacy L. Millions]

In article <670 at ctisbv.UUCP>, pim at ctisbv.UUCP (Pim Zandbergen) writes:
> ...
> This is what I've got:
> 
>  Olivetti M380/XP5 (20 MHz '386)
>  Western Digital WD-1007A-WAH ESDI harddisk controller
>  Micropolis 1558 338 MB harddisk
>  SCO XENIX 386 2.3.1
> 
> Although the harddisk has 1224 cylinders, the machine's CMOS RAM
> says it is BIOS-type 31, which has 814 cylinders.
> This probarbly is due to the remapping capablities of the WD-1007A-WAH,
> in order to handle the 1024 cylinder limit of SCO XENIX...

The 1024 cylinder limit is a *feature* not a bug :-). It is there to stay
compatible with DOS and other systems that use the fdisk structure. I quote
from my SCO Release Notes (2.2.3)

	The limitations for attached disks are 1024 cylinders (10 bits of
	cylinder addressing) and 16 heads, due to the *fdisk* structure shared
	by all operating systems. You can attach larger disks, but you cannot
	use more than 1024 cylinders, even with multiple fdisk partitions; the
	excess starage space in unaddressable.

Thank you Microsoft :-). 

I have installed SCO Xenix on a Maxtor 4380E 320MB drive using the Adaptec
ACB-2322 ESDI controller. I have used the entire drive for Xenix (no DOS
or other OS partitions) and been able to use all 1222 cylinders (the drive
has 1224, the controller wants 2 for its own bad track remapping). According
to the FDISK(C) manual entry, using the entire disk for Xenix,

	*fdisk* creates one partition that includes all the tracks on the
	disk, except the first track and the last cylinder.

I assume that this is one way of overriding the 1024 cylinder limit.

-stacy

-- 
"He to whom the early bird runs best learns wisdom and patience!
                          ... I can never remember proverbs" - Charlie Brown
S. L. Millions                                            ..!tmsoft!mcl!stacy