> 
> Yup, tried sdNg for all in N (including N) - You never know, could have
> been behaving like a Dec Alpha !
> 
> Getting the JAZ drive to work is one of the steps I'm taking to help me
> plunge into MacBSD development (in a tentative kinda way, if you see what
> I mean. %)

Probably a good place to start. :-)

> >> the drives, yet the kernel can't. Looking at the bootstrap output
> >> I noticed the following...
> >>
> >>    obio0 at mainbus0
> >>           :
> >>    ncrscsi0 at obio0
> >>    scsibus0 at ncrscsi0
> >>    sd0 at scsibus0 targ 0 lun 0: ... (which is my internal drive :)
> >>    sd0: 244MB ...
> >
> >What are the next two lines? There should be an sd1 entry, and an sd2 entry.
> >
> 
> Nope, I got...
> 
>    zsc0 at obio0 chip type 0
>    zstty0 at zsc0 channel 0
>    zstty1 at zsc0 channel 1  sheilding from LocalTalk
>    fpu0 at obio0 (emulator)

That's the problem. All the hard disks show up during autoconfig at boot (we
don't yet support post-boot configuration).

The easy way to tell, I think, is to get info on the two drives, under MacOS.
When you Get Info on a drive (select the drive icon on the desktop, then
Command-I), MacOS tells you where the drive lives. On the 8500 at work, it
tells you which SCSI bus. If the internal drive and the external/Jaz are on
different busses, then the dock has its own SCSI port.

It would take a bit of work to get NetBSD to be happy w/ multiple SCSI
ports (well, to get port-mac68k happy; M.I. NetBSD is quite happy now).
The mac68k drivers kinda assume there's only one chip AFAIK. To support
multiple chips, two things need to be done. 1) the mac68k side of the
5380 driver needs to deal with multiple chips, and 2) someone needs
to write an attach routine for the doc's chip.

>    graf1 at macvid0
>    changing root device to sd0a.
>    PRAM time does not appear to have been read corrctly  (at least im getting
>    PRAM 0x83daf80, macos_boottime; 0x32fdb2e5.            some 'normal'
> errors ;)

That "normal" error means you either don't have your timezone set correctly in
the Map Control panel, or you don't have the GMT offset set right in the booter.

> There are no other diagnostics earlier on about nnubus. Are there any sniffer
> utilities in existance that do a complete crawl of the machine's innards to see
> what it's got under the bonnet [hood :] (that run under BSD) ?

There is the Slots program from Apple. I think there's a copy on puma, somewhere.
It will tell you what's in each NuBus card. Probably what would be needed would be
a NuBus card driver which recognizes all the things on a Duo Dock, and sets them
up.

> I've downloaded the entire source tree last night to my PM6400, so I might
> have a
> look at the kernel's crawler code - but any further ideas would be good. Inci-
> dentally, is there a port which works on a PowerMac 6400/200 yet ? It's got that
> darn IDE, don'tcherknow.

Not yet, but hopefully soon. Since it's a PowerPC Mac, it would need port-powerpc.
A number of folks are getting together on it, and good things should happen
soon. port-powerpc runs now on OpenFirmware machines. Unfortunately gcc can't
build the right output format to get the PowerMacs to load a kernel. Also,
some folks have ideas on how to make the same kernel work on an OpenFirmware
machine and on a non-OpenFirmware machine (so the same kernel could run a
Motorola non-mac clone PPC computer, a MacOS clone, a PCI PowerMac, and a
NuBus PowerMac.

Take care,

Bill