>> >> This makes no sense. > > The variable is a uint64_t, the function returns an int. Some > compilers will tell you the implied cast may be wrong, we just say > it's correct. Correct fix is in checking and casting hashsize. Not losing precision. Francois