Subject: RE: AOpen mb won't boot NetBSD
To: Gunnar Helliesen <gunnar@bitcon.no>
From: Manuel Bouyer <bouyer@chassiron.ensta.fr>
List: port-i386
Date: 02/02/1998 14:18:13
On Feb 2, Gunnar Helliesen wrote
> OK, that's the second report of problems with certain Seagate HD and
> motherboard combinations.
> 
> I've tried setting the geometry manually, I've tried "NORMAL", "LARGE"
> and "LBA" modes, all to no avail. If I turn off all probing, i.e.
> manually enter the geometry of the boot disk and set all other drives to
> "None" it still won't boot. That's the strange thing...

Mine was not happy with "large" I think. Both normal and LBA did work.

> 
> I see the same symptoms as you though, all IDE devices (two HDs and one
> Creative CD-ROM) all probe correctly if set to autoprobe. The BIOS just
> doesn't get beyond the probe. All this is a bit strange though, why
> should the BIOS even bother with the bootblocks at this early stage?

In my case it was not hanging because of the boot blocks. It was because
after probing and printing the drives, for each drive found it tries to
figure which mode to use. So at some point it tries to use the seagate
in 'large' mode -> hang.

> Does it read the bootblocks to help determine the correct geometry or
> mode? Why on earth does the BIOS hang at this stage instead of after the
> setup summary screen when it should start booting from the disk?

It may try to read the partition table, to use the geometry set up here
instead of the one reported by the disk. You can try to create a MBR
partition in which you'll put NetBSD (as when you setup your disk
for multi-os, but with only one partition). So you'll have the "standart" MBR
and partition table, and then the NetBSD disklabel and boot blocks into one of
these partitions.

--
Manuel Bouyer, LIP6, Universite Paris VI.
email: bouyer@rp.lip6.fr
       bouyer@ensta.fr
--