On Thu, 15 May 1997, Jonathan Stone wrote:

.. Could someone else who's had problems with NetBSD-1.2D/pmax please try
.. the netbsd.generic kernel on ftp.netbsd,org
..   (ftp://ftp.netbsd.org/pub/NetBSD/arch/pmax/netbsd.generic)
.. on a 3100, and let me know if it works better, worse, or the same?


All kernels that I've built myself since about 25 March won't boot --
they fail as indicated below.

This includes Jonathans recent netbsd.generic of yesterday (or the
day before (I think that I got the second version)).

But -- BUT -- today's netbsd.3100 will boot and to me this is big
progress. I notice that this kernel is 1.2D, i.e. not quite -current.
What made it boot on my machine? sufficiently old kernel sources?
or a sufficiently old compiler/assembler/linker ?

And it seems to be even better: I have a disk, that has caused the system
to hang whenever non-trivial (or maybe even trivial) writes were done.
This problem has persisted since about December I think. With netbsd.3100
I can copy files to this disk without hanging. 

But the trees don't grow into the skies as we say over here: I cp'd
xc-1.tar.gz from an NFS-mounted file system to the problematic disk
and the I started a "tar zxf xc-1.tar.gz", and now the system hang
again, but with previous kernels any copying of files to this disk
would hang, just leaving a zero-length file after rebooting.

To me this seems that we're getting back on track again and I'm excited
about getting my DS3100 back in a state where I can keep it running on a
system that I have built from current sources. 

=========================
Result of booting any of my custom kernels built after 25 March
as well as trying to boot several of Jonathans test kernels up
to, but not including today's netbsd.3100.

>> boot -f rz(0,2,0)netbsd.ny
Boot: rz(0,2,0)netbsd.ny
Size: 994336+0+57780
Starting at 0x80030000


Excptn:	<vtr=NRML>
Excpn pc:	0x801ff9ac
Creg:		0x30000028<CE=3,EXC=II>
Sreg:		0x0
Sp:		0x8002ffe8
exit(-1)

regards
Erik Bertelsen