Re: better late than never: amigappc progress

[Ignatios Souvatzis <is%netbsd.org@localhost>]

On Sat, Feb 13, 2010 at 01:16:27PM +0200, Jukka Andberg wrote:
> On Thu, 11 Feb 2010 21:16:48 +0100 Frank Wille wrote:
> > No surprise. I just asked if somebody owns a BPPC, not that amigappc is
> > already working on it... ;)
> 
> One can hope...
> 
> > > Enforcer shows hits to 0xfff03100+.
> > 
> > Hits while accessing this address, or from a program running there?
> > 
> > The 68k-part of the kernel startup routine is located at 0xfff03000,
> > which will copy the kernel to 0x08000000 (probably this is causing the
> > hits on an A1200).
> 
> First from writing to that address and then from trying to execute from 
> there, 
> if I read it right. What memory should be at 0xfff03000?

Listen, young padawan:

PPC chips find their vector table (which is, btw., not an address
table like on m68k, but a list of short code snippets, normally
branch instructions, like on intel 8080/8085 and Z80) either at
0x00000000 or at 0xFFF00000 (IIRC), depending on an configuration
input pin to the CPU, with the MMU disabled. As Chipmem is... not
exactly fast enough for interupt and trap duty, sane PPC boards
for the Amiga map the last 512k or so of their memory there and
run in high-vector-table mode.

Regards,
        -is