Avoiding long Berge cycles II, exact bounds for all $n$