<Thanks to Martin Diehl>
	o [CORRECT] Handle non-linear and shared skbs
	o [CORRECT] Tell kernel we can handle multithreaded receive
		<Of course, this has been tested extensively on SMP>