The _ _MULS_C intrinsic has the following declaration:
float __MULS_C(float, float);
This intrinsic function returns the result of multiplying the first argument with the second argument, with IEEE chopped rounding used. Both arguments and the result are in IEEE single-precision format.