Verification of Remco's full 256x256 bit multiplication

This repo attempts to verify Remco's widely used mulDiv function.

At the moment it only verifies that the computation of the upper 256 bits of the 512 bit product via sub(sub(x1, x0), lt(x1, x0)) is verified.

Furthermore it verifies a specialization of fullMulDiv where the denominator is fixed to 2^128.

Credit to Remco Bloemen under MIT license: 2π.com/17/chinese-remainder-theorem/