fn bounds_max<'a, T: Ord>(b1: Bound<&'a T>, b2: Bound<&'a T>) -> Bound<&'a T>
Return the largest of b1 and b2.
b1
b2
If one of the bounds is Unbounded, the other will be returned.