VERY, very cool! Cant wait to see how your work on this progresses.
Sorry I broke in with that PCI CSII machine info, didn't track that you were specifically targeting the CS1 era PPC machines with 630 PDS from the NuBus Architecture generation. Should be very interesting info coming up indeed.
Do you have any plans for a future passthru connector to explore expanded use of the Networking capability of the CS1 Slot?
I figured I'd start with the easiest way of communicating between a microcontroller and the Mac -- using the slot's serial lines. If the card proves to be practical, then I can step up to memory transfers. That opens the door to various PDS slots and even NuBus.
That was very helpful, actually. I have a couple of Power Mac 6200s with CS1 slots that also won't show ROM address accesses. I would have spent days trying to debug that.
This brings up another concern --how will the 68040 cache affect bus snooping? In a tight loop, such as cold-boot memory tests, the bus will likely only show ram accesses. So, this whole thing is going to be a learning experience.