[NET]: rtnl_unlock cleanups
authorDenis V. Lunev <den@openvz.org>
Thu, 11 Oct 2007 04:12:58 +0000 (21:12 -0700)
committerDavid S. Miller <davem@davemloft.net>
Thu, 11 Oct 2007 04:12:58 +0000 (21:12 -0700)
commit1536cc0d55a2820b71daf912060fe43ec15630c2
tree0561a25ba3623e987fb801da529ea7c2270f0d11
parentd1ec3b772233826bf156284170632563790dbabf
[NET]: rtnl_unlock cleanups

There is no need to process outstanding netlink user->kernel packets
during rtnl_unlock now. There is no rtnl_trylock in the rtnetlink_rcv
anymore.

Normal code path is the following:
netlink_sendmsg
   netlink_unicast
       netlink_sendskb
           skb_queue_tail
           netlink_data_ready
               rtnetlink_rcv
                   mutex_lock(&rtnl_mutex);
                   netlink_run_queue(sk, qlen, &rtnetlink_rcv_msg);
                   mutex_unlock(&rtnl_mutex);

So, it is possible, that packets can be present in the rtnl->sk_receive_queue
during rtnl_unlock, but there is no need to process them at that moment as
rtnetlink_rcv for that packet is pending.

Signed-off-by: Denis V. Lunev <den@openvz.org>
Acked-by: Alexey Kuznetsov <kuznet@ms2.inr.ac.ru>
Signed-off-by: David S. Miller <davem@davemloft.net>
net/core/rtnetlink.c