Help!!!!!!!!!!

rjo__

You'll notice a lot of English here, without much reference to anything that is actually informative:)

I'm so close I can taste it. But that is as close as my humble brain is going to get me.
Release of Propeller2Explorer is right around the corner... it is just missing one thing.

Propeller2Explorer can dump hub memory over serial...And it can use serial to dump into hub memory.

It can dump cog memory over serial... and then dump it back into the cog. It can transfer an image and have the P2 display it.


I think you can see where I am going... let's say that I dump a cog... and write a little loader.spin so that the those original contents can be dumped back into a cog from serial... what do I have to do to get that cog to act on that dump? Or... is that enough and I worry too much? Similarly... let's say that I dump the original hub memory back into hub ram that is currently from a different program ...

ughhh. So unclear. Use your intuitions here:) It's a serial loader...I have the serial part... do I just shake the board?

mindrobots

When all else fails, just jmp!

The cog that did the loading, loaded the image to it's own cog ram, as it was running in hubexec. Assuming you accounted for any cog ram variables it is using, you should just be able to jmp to a valid address between $0 and $1f0 and the cog will try and execute it.

You could load a portion of cog ram from code in cogram but you would need to not overwrite your own loader code and the variables it is using.

Creative use of Shadow registers may help here.

Remember, one cog can't load another cog''s cogram or lutram. You need to start the cog first so it can load its own ram spaces.

potatohead

And it can do that from Hubexec too. Only a small routine is needed.

rjo__

got it... I'm going to start out one little step at a time. Thanks guys.