If 'usequadmath' is enabled, then NV type is __float128. However Gconvert cannot convert 128bit floating point number, because it uses sprintf. So we use my_snprintf, it uses quadmath_snprintf when...