- All internal memory handlers are now private members
- DDRs zeroed when device reset (as per datasheet)
- Port writes no longer transmit bits not set in DDR
- DDR updates may cause port bits to be written out (this is important)
- Add external read/write handlers for ports
* Identify cabinet type DIP switch
* Run the MCUs faster to get it back to same state as with simulation
* Remove now-unnecessary MCU simulation
(nw) I still don't want to drop the NOT_WORKING flag as the game's
clearly still not quite right, but it's as playable as it was previously
now, but using original MCU code for CPU comms. The clocks are still
guesses, just better guesses.
* Split out gsword and josvolly functionalit from base driver class
* Hooked up josvolly MCUs for communication between CPUs
* Added DIP locations to josvolly based on test mode
* Named unknown DIP settings after the location they control in main RAM
* Hooked up josvolly MCUs for communication, still preliminary
* Marked josvolly NOT_WORKING, simulation still present for reference
(nw) The MCU code only makes sense if there are a pair of them hooked up
together. There are effectively two programs, selected by whether P27
is tied high (master) or low (slave) at reset. Master and slave have
all corresponding P1 lines connected, master P20 connects to slave
TEST0, and master P21 connects to slave TEST1. You can see commands
and responses pass between the master and slave, it's very chatty. Test
mode seems to work, but game mode fails a check after a couple of
seconds and dumps status codes on the screen. There's still something
missing in the MCU hookup, and I'd guess there's something attached to
the AY output ports, too.
Mario driver now uses netlist audio implementation instead of discrete
implementation. The previous discrete sound emulation has not been
removed yet because it still contains a lot of documentation.
[Couriersud]
