How not to piss off a kernel subsystem maintainer - part 5