> Actually, looking at my 6502 docs again, doesn't JMP support absolute
> indirect addressing? ie, if I store #$10 in memory location $80, and #$F6
> in location $81, the instruction JMP ($80) will jump to $F610. Trying it
> in the PC Atari debugger, the addressing mode does exist... so it should be
> possible to get single-cycle-precision delay in ROM, without
> self-modifiying code.
I was using the BNE to save cycles over the JMP and a byte of space.
But you're absolutely right...I never even thought about absolute indirect
addressing. Guess I misunderstood your intent. :)
Heh...trying to save cycles in a time wasting routine!!!! :)
So how about...
LDA COUNT ; WAIT CYCLES = 16 - COUNT
LSR A ; SETUP CARRY BIT
STA $80 ; STORE OFFSET VALUE
DELAY STA WSYNC
JMP ($80) ; WASTE 5 CYCLES
TABLE NOP ; WASTE 2 CYCLES
NOP ; WASTE 2 CYCLES
NOP ; WASTE 2 CYCLES
NOP ; WASTE 2 CYCLES
NOP ; WASTE 2 CYCLES
BCC JUMP ; WASTE 2 OR 3 CYCLES
END ...
During initialization, the address of TABLE should be stored in $80. As is,
this one should give you a delay of 5 or between 7 and 18 cycles (can't get 6).
I think. I'm relatively new to cycle counting.
-Chris
--
Archives (includes files) at http://www.biglist.com/lists/stella/archives/
Unsub & more at http://www.biglist.com/lists/stella/stella.html
Don't post pirate BINs to Stellalist. Be a programmer, not a pirate.
Write the best game, win framed autographs of famous Atari alumni!!
