net: Make skb->skb_iif always track skb->dev
authorDavid S. Miller <davem@davemloft.net>
Mon, 23 Jul 2012 23:27:54 +0000 (16:27 -0700)
committerDavid S. Miller <davem@davemloft.net>
Mon, 23 Jul 2012 23:36:27 +0000 (16:36 -0700)
Make it follow device decapsulation, from things such as VLAN and
bonding.

The stuff that actually cares about pre-demuxed device pointers, is
handled by the "orig_dev" variable in __netif_receive_skb().  And
the only consumer of that is the po->origdev feature of AF_PACKET
sockets.

Signed-off-by: David S. Miller <davem@davemloft.net>
net/core/dev.c

index cca02ae7a844ffa5d9bb3897dd1e2998a41ca653..0ebaea16632fc348f6a48789831b737c7c51b707 100644 (file)
@@ -3173,8 +3173,6 @@ static int __netif_receive_skb(struct sk_buff *skb)
        if (netpoll_receive_skb(skb))
                return NET_RX_DROP;
 
-       if (!skb->skb_iif)
-               skb->skb_iif = skb->dev->ifindex;
        orig_dev = skb->dev;
 
        skb_reset_network_header(skb);
@@ -3186,6 +3184,7 @@ static int __netif_receive_skb(struct sk_buff *skb)
        rcu_read_lock();
 
 another_round:
+       skb->skb_iif = skb->dev->ifindex;
 
        __this_cpu_inc(softnet_data.processed);