[PATCH 4.4 26/28] ipv4: Set skb->protocol properly for local output

From: Greg Kroah-Hartman
Date: Fri Dec 09 2016 - 11:19:10 EST

4.4-stable review patch. If anyone has any objections, please let me know.


From: Eli Cooper <elicooper@xxxxxxx>

commit f4180439109aa720774baafdd798b3234ab1a0d2 upstream.

When xfrm is applied to TSO/GSO packets, it follows this path:

xfrm_output() -> xfrm_output_gso() -> skb_gso_segment()

where skb_gso_segment() relies on skb->protocol to function properly.

This patch sets skb->protocol to ETH_P_IP before dst_output() is called,
fixing a bug where GSO packets sent through a sit tunnel are dropped
when xfrm is involved.

Signed-off-by: Eli Cooper <elicooper@xxxxxxx>
Signed-off-by: David S. Miller <davem@xxxxxxxxxxxxx>
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>

net/ipv4/ip_output.c | 3 +++
1 file changed, 3 insertions(+)

--- a/net/ipv4/ip_output.c
+++ b/net/ipv4/ip_output.c
@@ -102,6 +102,9 @@ int __ip_local_out(struct net *net, stru

iph->tot_len = htons(skb->len);
+ skb->protocol = htons(ETH_P_IP);
net, sk, skb, NULL, skb_dst(skb)->dev,