pub const USE_LIBM: bool = false;
True if using libm for floating-point math, false if using internal approximation functions