At 20:10 Uhr +0100 11.12.2000, Colin Wood wrote:
>C.R. Carlson wrote:
>> >Colin asked:
>> >where in the boot did it panic, and what exactly was the panic message?
>>
>> ** Well first of all, I found out that Hauke Fath posted on an apparently
>> identical problem in the archives dated 3/13/00, though I don't think he
>> was using a powerbook 170. I get all the way through the device id's (which
>> appear to be normal) and just a line past the PRAM announcement (that the
>> PRAM time doesn't appear to have been read correctly) when I get
>>    panic: cn open: cn_tab->cndev==NODEV.
>>    then I get put in the debugger.
>> This was with both kernels, in single user mode, even booting from the mac
>> side one time. 1.4.3 is running famously, even as I speak.
>
>this one sounds like you need to build the 1.5 devices.

Not likely, as the panic occurs before teh kernel goes to disk. You then
would get a "friendlier" error message along the lines of "no console
device found", not a panic.

For my Quadra 700 at least, the panic was caused by the kernel not
attaching internal video (PR 11000). That again was caused by a bus.h
mapping problem: As Scott found out, a memory area was already mapped by
esp related code when the intvid probe happened. If the PB 170 has a
graphics controller that contains registers for other tasks, too, the
problem may be the same.

Clarence: Is the problem still present when you boot a kernel from a
-current snapshot? Can you send-pr the problem, giving as much information
as you have, especially the console output, from booting a 1.5 kernel?

	hauke

--
"It's never straight up and down"     (DEVO)