geneve: Relax MTU constraints

BugLink: http://bugs.launchpad.net/bugs/1596575

[ Upstream commit 55e5bfb5 ]

Allow the MTU of geneve devices to be set to large values, in order to
exploit underlying networks with larger frame sizes.

GENEVE does not have a fixed encapsulation overhead (an openvswitch
rule can add variable length options), so there is no relevant maximum
MTU to enforce.  A maximum of IP_MAX_MTU is used instead.
Encapsulated packets that are too big for the underlying network will
get dropped on the floor.
Signed-off-by: David Wragg <david@weave.works>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Tim Gardner <tim.gardner@canonical.com>
Signed-off-by: Kamal Mostafa <kamal@canonical.com>

drivers/net/geneve.c

@@ -1038,6 +1038,17 @@ static netdev_tx_t geneve_xmit(struct sk_buff *skb, struct net_device *dev)
 	return geneve_xmit_skb(skb, dev, info);
 }
 
+static int geneve_change_mtu(struct net_device *dev, int new_mtu)
+{
+	/* GENEVE overhead is not fixed, so we can't enforce a more
+	 * precise max MTU.
+	 */
+	if (new_mtu < 68 || new_mtu > IP_MAX_MTU)
+		return -EINVAL;
+	dev->mtu = new_mtu;
+	return 0;
+}
+
 static int geneve_fill_metadata_dst(struct net_device *dev, struct sk_buff *skb)
 {
 	struct ip_tunnel_info *info = skb_tunnel_info(skb);
@@ -1082,7 +1093,7 @@ static const struct net_device_ops geneve_netdev_ops = {
 	.ndo_stop		= geneve_stop,
 	.ndo_start_xmit		= geneve_xmit,
 	.ndo_get_stats64	= ip_tunnel_get_stats64,
-	.ndo_change_mtu		= eth_change_mtu,
+	.ndo_change_mtu		= geneve_change_mtu,
 	.ndo_validate_addr	= eth_validate_addr,
 	.ndo_set_mac_address	= eth_mac_addr,
 	.ndo_fill_metadata_dst	= geneve_fill_metadata_dst,