On 11/20/2017 05:09 AM, David Miller wrote: > From: Steffen Klassert <steffen.klass...@secunet.com> > Date: Mon, 20 Nov 2017 08:37:47 +0100 > >> This patchset implements asynchronous crypto handling >> in the layer 2 TX path. With this we can allow IPsec >> ESP GSO for software crypto. This also merges the IPsec >> GSO and non-GSO paths to both use validate_xmit_xfrm(). > ... > > Code looks generally fine to me. Only thing of note is that this > adds a new dev_requeue_skb() call site and that might intersect with > John Fastabend's RFC work to make qdiscs lockless.
Right, fortunately it doesn't appear that the conflicts are too hard to resolve. It looks like sch_redirect_xmit() will need to be resolved with an additional check to test if the qdisc lock is needed. Then assuming my reading is correct validate_xmit_xfrm(), xfrm_dev_resume(), and xfrm_dev_backlog() can run without qdisc lock already. Thanks, John > > Also, please adhere to the reverse christmas tree rule for the > ordering of your function local variables. > > Thank you. >