@Inginsub wherenum3 = dworda*dwordc num2 = dworda*dwordd + dwordb*dwordc num1 = dwordb*dowrdd and num2 fits in 32bits without overflowing