AVX2 permits working with 256-bit registers. Thus we can multiply 2 GF(2^127) elements in parallel. This commit adds 2 such functions for multiplication by 10 and 11).
- add Inverse operation to sl2 - fix a bug in xN()