The "shifting" vs "multiplication" is just an optimisation since x * 2 == x << 1 for any positive x. As far as I'm aware shifting a negative number is undef

The "shifting" vs "multiplication" is just an optimisation since x * 2 == x << 1 for any positive x. As far as I'm aware shifting a negative number is undefined, which will be why they use multiplication for negatives.