[Coco] CoCoRX

[Frank Pittel]

On Thu, Jun 04, 2015 at 01:37:30PM +1000, Mark McDougall wrote:
> On 4/06/2015 1:19 PM, Frank Pittel wrote:
> 
> >>>On Sat, May 30, 2015 at 02:43:25PM -0500, Melanie and John Mark Mobley wrote:
> >>>>We need a diagnostic test that will prove out the cycle accuracy timing of
> >>>>the CPU so that it can be used to test the FPGA design.
> 
> As others have mentioned, not only would it be quite difficult (but
> ultimately possible) to have software determine the cycle accuracy
> of the core, the best you'll get is a pass/fail indication, which is
> of very little use when you're "test[ing] the FPGA design".
> 
> The way you'd actually test the FPGA design would be to use
> purpose-built hardware to run the core in lock-step with a real
> 6809, taking care to exercise not only every instruction, but also
> every pre-condition as well (eg. conditional branches).
> 
> I actually have a PCB that allows you to plug a 6809 onto the
> expansion headers on a TerASIC DE1/2. It was designed and built by
> my colleague when he started work on his cycle-accurate 6809 core
> (which was never completed). What I have done, however, is use it to
> run 6809-based FPGA designs - including a Coco 1/2 - with an actual
> 6809, so I know the board works.
> 
> But again, as Frank said, isn't this primarily for "real" Coco's anyway?

I notice that the reply was to me but nothing in either of my replies were included, here. :-)
(for the humor impared I'm kidding)

While I both my replies were typed after less than 5 min of thinking about it but it occurs to me
that every instruction and every possible permutation and addressing mode.

For FPGA code the key would be to wrap code around the core to test it. I think we agree that the
test ROM is targeted towards actual Cocos and not FPGA boards.

Frank