What is the fastest way to divide two 256-bit integers? They are represented as an array of 26-bit words, each stored in 32-bit integers. I have access to a 64-bit type if that's useful.
It seems like fast division methods in the literature all assume we are dividing a 2n-word number by an n-word number. Is there some simple transformation or folklore that lets me generalize these methods to arbitrary numbers?