On the possibility of p-type SnO2