Subject: Re: Tracking -current while not trashing stable installation
To: <>
From: David Laight <david@l8s.co.uk>
List: netbsd-help
Date: 01/25/2002 10:23:10
Thomas Mueller wrote:

> 
> I installed Slackware 8.0, ran lilo a second time, to include the NetBSD primary
> partition.  Using Linux cfdisk, prior to installing Slackware 8.0, I deleted
> both the NetBSD primary partition and the adjacent primary partition originally
> intended for OpenBSD 2.9, and made the combined space into a new primary
> partition type a9 (NetBSD, but Unknown to Linux).  lilo.conf needed an "lba32"
> line (without the quotes) to be able to see above 8 GB, ran seemingly OK.  I
> think that partition was bootable from the previous install attempt that only
> installed minimally, and it might possibly still be, though I think I would
> redisklabel and reinstall 1.5.2 from the beginning.

Looks like lilo is loading and executing the pbr code....  I'm surprised
it
needs to be told to do LBA, netBSD bootstrap susses it dynamically.
> 
> I saw messages to the effect that OpenBSD can't boot from above 8 GB, and I
> found OpenBSD fdisk can't see above 8 GB, and OpenBSD disklabel showed those two
> primary partitions, now combined into one, as overlapping, though it was OK to
> Linux fdisk and cfdisk.  So now I figure I have enough without OpenBSD.

Looks like openBSD has a broken fdisk!  and a useless bootstrap ;-)
> 

> I guess changing the partition type, as Linux fdisk or cfdisk (and NetBSD
> fdisk?) would do, only changes one byte in the partition table but nothing in
> the actual partition, thus making it easy to change back?

Yep...

> 
> I will have some reading to do, on accessing partitions on other disks.  I have
> a variable configuration regarding other disks because of a mobile rack, which
> now holds the larger of two hard drives (1270 MB) from the old computer.

Eack disk needs its own disklabel......

>  New computer hard drive is 40 GB.

A small one :-)

	David