fp_floor_byte

Been hoping to port this code to my own CPU and so reading the repo.
But I've gotten stuck on this shared function.
Can't for the life of me see how it works.

Is it used or is it junk and buggy?
Seems for positive numbers it return zero, and for negative numbers other weird results, non zero fraction value gets returned and so on.

Any clues?

fp_floor looks sane tho.