zielaj
Results
2
comments of
zielaj
One option is to put the burden of choosing `ON` vs `OFF` for `WITH_DIV32` on the compiler instead. For example, clang today includes a dynamic check for actual operand sizes...
This comment is perhaph against the spirit of this thread, but since we've been discussing integer division optimizations, here's yet another possibility. I've noticed that a lot of CPU in...