> On Wed, 23 Feb 2000, NetBSD Mailing list wrote:
> 
> > FYI, booting is exhibited as: boot -f rz(0,0)/netbsd
> 
>   Whoops - now I think I see what is going on:
> 
>   The bug in the code was extracting the boot device information if it
> started with "rz(", but did *not* have the correct format for the rest of
> the string.  It was taking the '/' as the partition, which resulted in -1
> as the partition value.  That would be masked to 7, which corresponds to
> 'h'.  The old parsing code looks like it would have accepted rz(0,0), and
> defaulted the partition to 0, which is partition 'a'.
> 
>   Try using rz(0,0,0) in your bootpatch and see if that works any better.

Yup, that works.   Thanks.  And, the kernel looks good as before.

Now, with regards to the console setting of bootpath, what is the resolution
for having NetBSD properly recognise the kernel booted for kvm_mkdb?

	-471 auto: rz(0,0,0)netbsd
	...
	kvm_mkdb: rz(0,0,0)netbsd: No such file or directory 

The relevant /etc/rc code is:

/etc/rc:        kvm_mkdb `sysctl -n machdep.booted_kernel`

This is what pops up upon boot when bootpath is changed.  It doesn't seem to
the second time around when an autoboot kicks in, so I don't know if this is
because it created the DB properly, or just filled it in with placeholder
information the first time.  You can repeat the above error by simply rm'ing
/var/db/kvm.db.

Shall I simply 'ln -s /netbsd "/rz(0,0,0)/netbsd"'?   It works ;)

Thanks!
-skots
--
Scott G. Akmentins-Taylor     InterNet: staylor@mrynet.com
MRY Systems			        staylor@mrynet.lv
    (Skots Gregorijs Akmentins-Teilors -- just call me "Skots")
		----- Labak miris neka sarkans -----