[PATCH v2 1/2] ip6_gre: Fix MTU setting for ip6gretap

[Haishuang Yan]

When creat an ip6gretap interface with an unreachable route,
the MTU is about 14 bytes larger than what was needed.

If the remote address is reachable:
ping6 2001:0:130::1 -c 2
PING 2001:0:130::1(2001:0:130::1) 56 data bytes
64 bytes from 2001:0:130::1: icmp_seq=1 ttl=64 time=1.46 ms
64 bytes from 2001:0:130::1: icmp_seq=2 ttl=64 time=81.1 ms

--- 2001:0:130::1 ping statistics ---
2 packets transmitted, 2 received, 0% packet loss, time 1001ms
rtt min/avg/max/mdev = 1.465/41.316/81.167/39.851 ms

ip link add ip6gretap1 type ip6gretap\
 local 2001:0:130::2 remote 2001:0:130::1
ip link show ip6gretap1
11: ip6gretap1@NONE: <BROADCAST,MULTICAST> mtu 1434 ...
    link/ether c2:f3:f8:c1:2c:bf brd ff:ff:ff:ff:ff:ff

The MTU value 1434 is right. But if we delete the direct route:
ip -6 route del 2001:0:130::/64
ping6 2001:0:130::1 -c 2
connect: Network is unreachable
ip link add ip6gretap1 type ip6gretap\
 local 2001:0:130::2 remote 2001:0:130::1
ip link show ip6gretap1
12: ip6gretap1@NONE: <BROADCAST,MULTICAST> mtu 1448 ...
    link/ether 7e:e1:d2:c4:06:5e brd ff:ff:ff:ff:ff:ff

Now, the MTU value 1448 is larger than what was needed.

The reason is that if there is a reachable route, when
run following code in ip6gre_tnl_link_config:

	if (p->flags & IP6_TNL_F_CAP_XMIT) {
		int strict = (ipv6_addr_type(&p->raddr) &
			      (IPV6_ADDR_MULTICAST|IPV6_ADDR_LINKLOCAL));

		struct rt6_info *rt = rt6_lookup(t->net,
						 &p->raddr, &p->laddr,
						 p->link, strict);

		if (!rt)
			return;

		if (rt->dst.dev) {
			dev->hard_header_len = rt->dst.dev->hard_header_len +
					       t_hlen;

			if (set_mtu) {
				dev->mtu = rt->dst.dev->mtu - t_hlen;
				if (!(t->parms.flags & IP6_TNL_F_IGN_ENCAP_LIMIT))
					dev->mtu -= 8;
				if (dev->type == ARPHRD_ETHER)
					dev->mtu -= ETH_HLEN;

				if (dev->mtu < IPV6_MIN_MTU)
					dev->mtu = IPV6_MIN_MTU;
			}
		}
		ip6_rt_put(rt);
	}

Because rt is not NULL here, so dev->mtu will subtract the ethernet
header length later. But when rt is NULL, it just simply return, so
dev->mtu doesn't update correctly in this situation.

This patch first verify the dev->type is ARPHRD_ETHER for ip6gretap
interface, and then decrease the mtu as early as possible.

Signed-off-by: Haishuang Yan <yanhaishuang@xxxxxxxxxxxxxxxxxxxx>
---
Changes in v2:
  - Make the commit message more clearer.
---
 net/ipv6/ip6_gre.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/net/ipv6/ip6_gre.c b/net/ipv6/ip6_gre.c
index 4541fa5..8ea5a4d 100644
--- a/net/ipv6/ip6_gre.c
+++ b/net/ipv6/ip6_gre.c
@@ -1029,6 +1029,8 @@ static int ip6gre_tunnel_init_common(struct net_device *dev)
 
 	dev->hard_header_len = LL_MAX_HEADER + t_hlen;
 	dev->mtu = ETH_DATA_LEN - t_hlen;
+	if (dev->type == ARPHRD_ETHER)
+		dev->mtu -= ETH_HLEN;
 	if (!(tunnel->parms.flags & IP6_TNL_F_IGN_ENCAP_LIMIT))
 		dev->mtu -= 8;
 
-- 
1.8.3.1