>Dan Winship wrote:
>
>> Well, once you manage to get OF to boot a kernel, it's not likely to
>> actually work...
>>
>> -- Dan
>
>Well, it booted, but it didn't get far...
>
>(using netbsd-GENERIC_MD, 8/11/99)
>
>cpu version c (Revision 207)
>real mem = 402653184
>avail mem = 380219392
>using 800 buffers containing 3276800 bytes of memory
>mainbus0 (root)
>cpu0 at mainbus0
>trap type 200 at 272764
>stopped in  at 0x234a58: lwz r0, r1, 0x14
>db> _
>
>To my (admittedly untrained) eyes, it looks like it's breaking when it
>tries to talk to the PCI controller. Could this be a bus timing issue,
>perhaps?
>
>Also, does anyone know the boot sequence step-by-step (that is, what the
>kernel looks for when)? This might help me figure out what is happening.

Yes, that would be cool if people in the know enlightened the rest of us
about the exact boot sequence
(preferrably with the routine names involved) I'm particularly interested
in the early stages ( before kernel  main is invoked). Or maybe a pointer
to the source of info?
The "pink book" talks about some "machine dependent setup"  and briefly
mentions setting up the stack, address translation, etc - that's what I'm
particularly  interested to know


Andy