[many sad error problems deleted]

> sd0(ncr0:0:0): non-media hardware failure, data = 6d 00 0f bf eb 83 fd ff
> 74 14 80 0e 02 8b ac 24 ac 01 ... etc (it was about 6 lines of hex dump,
> and I didn't feel like writing it all down :) Is there any way to get
> these kernel messages sent to a serial port, btw?) It repeated that
> message about 6 or 7 times before the system panicked (unable to mount 
> root, I believe).

The mac68k port has this feature. It's really easy to add. At one point,
the kernel goes off and looks for consoles. The serial driver will,
if it's not the console-of-choice, look at a boot environment variable
(which was set in the MacOS booting program) to see if we're doing
serial echo. If so, the serial port gets initialized w/ interrupts off.

The itecnputc routine (which draws characters on the console screen)
looks at this boot variable also, and calls sercnputc to echo characters 
out the serial port too.

I'm not sure how to go about setting such a flag during boot on other 
ports. If the machine comes up directly into netbsd, it might be a bit 
tricky to convey this info. But other than that, serial echo isn't
too big a deal.

Take care,

Bill