This commit is contained in:
Brooke Vibber 2024-12-31 05:11:26 -08:00
parent b56dc1e98b
commit 61eb1aaf21

View file

@ -1,5 +1,11 @@
things to try:
* skip add on the top-byte multiply in sqr8/mul8
* should save a few cycles, suggestion by jamey
* perform the zx += zx^s + cx in 32-bit space, before rounding
* should improve precision on max zoom, might cost a few cycles
* patch the entire expanded-ram imul8xe on top of imul8 to avoid the 3-cycle thunk penalty :D
* try 3.13 fixed point instead of 4.12 for more precision