LAPACK auxiliary routine to test whether a double-precision floating-point number is NaN. - View it on GitHub