In the OpenCL spec it looks like 'isequal()' is only defined for half, double, and float. Why don't I have this operator for integer types? I would've expected it to be the opposite, as it is...