openvswitch: set correct protocol on route lookup
authorJiri Benc <jbenc@redhat.com>
Fri, 5 Dec 2014 16:24:28 +0000 (17:24 +0100)
committerDavid S. Miller <davem@davemloft.net>
Tue, 9 Dec 2014 21:01:21 +0000 (16:01 -0500)
Respect what the caller passed to ovs_tunnel_get_egress_info.

Fixes: 8f0aad6f35f7e ("openvswitch: Extend packet attribute for egress tunnel info")
Signed-off-by: Jiri Benc <jbenc@redhat.com>
Acked-by: Pravin B Shelar <pshelar@nicira.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
net/openvswitch/vport.c

index e771a46933e552fa05001e9354ff77725b1b8545..9584526c077804f21d22726d4bb5bd7cc4260d36 100644 (file)
@@ -600,7 +600,7 @@ int ovs_tunnel_get_egress_info(struct ovs_tunnel_info *egress_tun_info,
        fl.saddr = tun_key->ipv4_src;
        fl.flowi4_tos = RT_TOS(tun_key->ipv4_tos);
        fl.flowi4_mark = skb_mark;
-       fl.flowi4_proto = IPPROTO_GRE;
+       fl.flowi4_proto = ipproto;
 
        rt = ip_route_output_key(net, &fl);
        if (IS_ERR(rt))