Subject: Re: Exactly what does the 1024 cyl limit break?
To: Joel Reicher <void@yoyo.cc.monash.edu.au>
From: Frank van der Linden <frank@wins.uva.nl>
List: port-i386
Date: 11/27/1997 12:28:36
On Thu, Nov 27, 1997 at 09:35:01PM +1100, Joel Reicher wrote:
> I can't even be sure I'm asking the right question. I know there is some 
> sort of limit on hard drive size for _something_, and I vaguely remember 
> it being 1024 cylinders, LBA or not. I also seem to remember it is only a 
> BIOS limitation and affects booting, but nothing else.


The limit is in the interface to the BIOS. You can only specify 10 bits
for the cylinder number, so that means you can have cylinders 0-1023,
no matter what mode you're in.

Most modern BIOSs work around this by enabling a fake geometry, in
which the number of cylinders is trimmed down to a value <= 1024,
and the number of heads multiplied by the same factor, giving
the same total size.

> If I have all of 
> this right, then which bit of the boot process does it affect? Would I be 
> right in saying that the only restriction it places on NetBSD is that the 
> kernel being booted must be below cyl 1024?

Yes, this is correct, with the addition that 'cylinder 1024' means 'cylinder
1024 as the BIOS knows it', which may be different from the NetBSD view
of things, as NetBSD always uses the 'real' geometry as reported by the disk.
I am modifiying the install program so that it will check if your root
partition lies within the first 1024 BIOS cylinders.

- Frank