Similarly, how much time was spent writing "code" that targets the PowerPC processor versus the FPGA module?
We spent 6 weeks getting the game to work (~10-15 hrs/ week). This was all in C targeting the PowerPC and included effective double buffering for graphics as well as all the game logic.
We then spent another 2 weeks getting interrupt driven sound to work. This was also in C, but was valuable because in order to understand the nature of the generated interrupts we had to understand the underlying bus architecture.
We had 3 weeks to do a basic hardware module (a slave on the peripheral bus that powered some 7-segment displays and generated interrupts when switches were pressed) and then to do the creative portion (which was another slave).
Finally, we built the Audio DMA controller, which took 4 weeks. The Audio DMA controller was by far the most challenging because the rebuild cycle was 20 minutes (terrible EDK) and debugging was difficult.
So I'd say about 50-50.
It was a pretty cool course. No homework, no final, no text books. Just the project and all the documentation we could get our hands on.
Once we had it set up and had interrupt driven code to handle sends and receives it all worked great.
PS: I love your project, good work!
It'd be a nice introduction for other fpga and design logic students...