I don't understand how to determine which banks to use. I've seen gbaguy's code use bank0 for code, bank1 for init info (where prog starts, vblank, etc), and bank2 as CHR-ROM.
Then he .org's bank0 at $8000 and bank1 at $FFFA, which seems reasonable CPU memory locations. But then he .org's bank2 at $0000 which would be at zero page in the CPU memory map (but at the pattern table in the PPU memory map). So how it is done to make the mapper "know" where to put the stuff, CPU mem or PPU mem?
Is it a matter of a certain bank number, or which addresses to use?
I'm using nesasm.
Then he .org's bank0 at $8000 and bank1 at $FFFA, which seems reasonable CPU memory locations. But then he .org's bank2 at $0000 which would be at zero page in the CPU memory map (but at the pattern table in the PPU memory map). So how it is done to make the mapper "know" where to put the stuff, CPU mem or PPU mem?
Is it a matter of a certain bank number, or which addresses to use?
I'm using nesasm.