Kai, I agree with your advice; however I would test it in an emulator. The OP ask for simulation, which, in my opinion only does half the job. The pitfall with a simulator is that you can 'prove' that your software works and then, if you have a hrdware glitch, there is no other diagnostic than "it does not work". I have some misgivings about the simulator 'proof' since no simulator verifies that the timing is 'acceptable' to the hardware. When I have problems like this I 'grow' the software, rather than writing the whole kit and kaboodle in one go and then asking "why does it not work"

Erik

An example of 'growing' software: first write a routine that reads a register, make that work with the emulator, then write a register, read it back, make that work with the emulator, then ...