Pointless tainting?: evaluating the practicality of pointer tainting