From: "Chris Wilkson" <ecwilkso@xxxxxxx>
> >double-height/single-height routine, but I ran into a situation where I
> >really DID need to waste just one cycle, and couldn't do it.
> Don't suppose you could reorganize, so that you branch to the lda
instruction?
> Not sure of the context here.
No, as I have code sections A,B and it goes like this
bcs B
A
B
So you see, one version runs just B, and the other runs A,B. They need to
be consecutive in memory. And the timing needs to be correct for both
versions. Also, reorganising so you branch to A instead of B simply shifts
the exact-same problem to B, which then has to find a cycle itself.
However, your next suggestion gives me a glimmer of hope on that front...
> Can you align the code so the the LDA instruction crosses a page boundary?
> That'll give you an extra cycle.
That's the most brilliant suggestion :) I'll try and mangle it into the
code somehow. As it exists, though, there's a branch back to the top of the
code, so that will take an additional cycle, too, and bugger up things. But
utilising the extra cycle when branching over a page boundary is certainly
something worth looking at. Thanks for that one!
> Question: Do the GRP timings depend on when the instruction starts, or
> when it finishes? If it's when it finishes, you can use the TIA mirror
> at GRP0+$100.
When the instruction completes, unfortunately. Good idea, and I'd spotted
that possibility after I sent my post. But also something to consider. You
don't need to use the mirror, just sta.w GRP0 will do the job.
> -Chris
--
_ _ _| _ _ _| _ * _ _ ,
(_|| )(_|( (/_\/\/ (_|(_|\/(_(/_ ,~' L_|\
,-' \
see my Museum of Soviet Calculators at ( \
http://www.taswegian.com/MOSCOW/soviet.html \ __ /
L,~' "\__/
@--> v
-
Archives (includes files) at http://www.biglist.com/lists/stella/archives/
Unsub & more at http://www.biglist.com/lists/stella/
