How to efficiently convert a number n to an integer with that many bits set? eg 8 -> 0b11111111?

Question

How do you efficiently convert a number n to a number with that many low significant bits set? That is, for an 8-bit integer:

0 -> 00000000
1