[IPV6] Always add a fragment header after receiving TOO BIG w/ pmtu < 1280.

According to RFC2460, PMTU is set to the IPv6 Minimum Link
MTU (1280) and a fragment header should always be included
after a node receiving Too Big message reporting PMTU is
less than the IPv6 Minimum Link MTU (1280).

Signed-off-by: Hideaki YOSHIFUJI <yoshfuji@linux-ipv6.org>

5 files changed, 39 insertions, 16 deletions

diff --git a/include/linux/ip.h b/include/linux/ip.h
index 487152a404f8..7d7633632d28 100644
--- a/include/linux/ip.h
+++ b/include/linux/ip.h
@@ -152,6 +152,7 @@ struct inet_sock {
 };
 
 #define IPCORK_OPT	1	/* ip-options has been held in ipcork.opt */
+#define IPCORK_ALLFRAG	2	/* always fragment (for ipv6 for now) */
 
 static inline struct inet_sock *inet_sk(const struct sock *sk)
 {
diff --git a/include/linux/rtnetlink.h b/include/linux/rtnetlink.h
index c3d5d043a5ca..5f6310f0d880 100644
--- a/include/linux/rtnetlink.h
+++ b/include/linux/rtnetlink.h
@@ -346,6 +346,7 @@ enum
 #define RTAX_FEATURE_ECN	0x00000001
 #define RTAX_FEATURE_SACK	0x00000002
 #define RTAX_FEATURE_TIMESTAMP	0x00000004
+#define RTAX_FEATURE_ALLFRAG	0x00000008
 
 struct rta_session
 {
diff --git a/include/net/dst.h b/include/net/dst.h
index b7e47a7ad105..785c4a185ff5 100644
--- a/include/net/dst.h
+++ b/include/net/dst.h
@@ -124,6 +124,15 @@ dst_pmtu(const struct dst_entry *dst)
 	return mtu;
 }
 
+static inline u32
+dst_allfrag(const struct dst_entry *dst)
+{
+	int ret = dst_path_metric(dst, RTAX_FEATURES) & RTAX_FEATURE_ALLFRAG;
+	/* Yes, _exactly_. This is paranoia. */
+	barrier();
+	return ret;
+}
+
 static inline int
 dst_metric_locked(struct dst_entry *dst, int metric)
 {
diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c
index fce2a87e9e4a..d19cafe5d1e2 100644
--- a/net/ipv6/ip6_output.c
+++ b/net/ipv6/ip6_output.c
@@ -147,7 +147,7 @@ static int ip6_output2(struct sk_buff *skb)
 
 int ip6_output(struct sk_buff *skb)
 {
-	if (skb->len > dst_pmtu(skb->dst))
+	if (skb->len > dst_pmtu(skb->dst) || dst_allfrag(skb->dst))
 		return ip6_fragment(skb, ip6_output2);
 	else
 		return ip6_output2(skb);
@@ -848,6 +848,8 @@ int ip6_append_data(struct sock *sk, int getfrag(void *from, char *to, int offse
 		inet->cork.fl = *fl;
 		np->cork.hop_limit = hlimit;
 		inet->cork.fragsize = mtu = dst_pmtu(&rt->u.dst);
+		if (dst_allfrag(&rt->u.dst))
+			inet->cork.flags |= IPCORK_ALLFRAG;
 		inet->cork.length = 0;
 		sk->sk_sndmsg_page = NULL;
 		sk->sk_sndmsg_off = 0;
@@ -899,7 +901,7 @@ int ip6_append_data(struct sock *sk, int getfrag(void *from, char *to, int offse
 
 	while (length > 0) {
 		/* Check if the remaining data fits into current packet. */
-		copy = mtu - skb->len;
+		copy = (inet->cork.length <= mtu && !(inet->cork.flags & IPCORK_ALLFRAG) ? mtu : maxfraglen) - skb->len;
 		if (copy < length)
 			copy = maxfraglen - skb->len;
 
@@ -924,7 +926,7 @@ alloc_new_skb:
 			 * we know we need more fragment(s).
 			 */
 			datalen = length + fraggap;
-			if (datalen > mtu - fragheaderlen)
+			if (datalen > (inet->cork.length <= mtu && !(inet->cork.flags & IPCORK_ALLFRAG) ? mtu : maxfraglen) - fragheaderlen)
 				datalen = maxfraglen - fragheaderlen;
 
 			fraglen = datalen + fragheaderlen;
@@ -1158,6 +1160,7 @@ out:
 	if (np->cork.rt) {
 		dst_release(&np->cork.rt->u.dst);
 		np->cork.rt = NULL;
+		inet->cork.flags &= ~IPCORK_ALLFRAG;
 	}
 	memset(&inet->cork.fl, 0, sizeof(inet->cork.fl));
 	return err;
@@ -1185,6 +1188,7 @@ void ip6_flush_pending_frames(struct sock *sk)
 	if (np->cork.rt) {
 		dst_release(&np->cork.rt->u.dst);
 		np->cork.rt = NULL;
+		inet->cork.flags &= ~IPCORK_ALLFRAG;
 	}
 	memset(&inet->cork.fl, 0, sizeof(inet->cork.fl));
 }
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index c696043a8e47..e5bc02ccc0ea 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -628,8 +628,10 @@ static void ip6_rt_update_pmtu(struct dst_entry *dst, u32 mtu)
 
 	if (mtu < dst_pmtu(dst) && rt6->rt6i_dst.plen == 128) {
 		rt6->rt6i_flags |= RTF_MODIFIED;
-		if (mtu < IPV6_MIN_MTU)
+		if (mtu < IPV6_MIN_MTU) {
 			mtu = IPV6_MIN_MTU;
+			dst->metrics[RTAX_FEATURES-1] |= RTAX_FEATURE_ALLFRAG;
+		}
 		dst->metrics[RTAX_MTU-1] = mtu;
 	}
 }
@@ -1164,26 +1166,26 @@ void rt6_pmtu_discovery(struct in6_addr *daddr, struct in6_addr *saddr,
 			struct net_device *dev, u32 pmtu)
 {
 	struct rt6_info *rt, *nrt;
-
-	if (pmtu < IPV6_MIN_MTU) {
-		if (net_ratelimit())
-			printk(KERN_DEBUG "rt6_pmtu_discovery: invalid MTU value %d\n",
-			       pmtu);
-		/* According to RFC1981, the PMTU is set to the IPv6 minimum
-		   link MTU if the node receives a Packet Too Big message
-		   reporting next-hop MTU that is less than the IPv6 minimum MTU.
-		   */
-		pmtu = IPV6_MIN_MTU;
-	}
+	int allfrag = 0;
 
 	rt = rt6_lookup(daddr, saddr, dev->ifindex, 0);
-
 	if (rt == NULL)
 		return;
 
 	if (pmtu >= dst_pmtu(&rt->u.dst))
 		goto out;
 
+	if (pmtu < IPV6_MIN_MTU) {
+		/*
+		 * According to RFC2460, PMTU is set to the IPv6 Minimum Link 
+		 * MTU (1280) and a fragment header should always be included
+		 * after a node receiving Too Big message reporting PMTU is
+		 * less than the IPv6 Minimum Link MTU.
+		 */
+		pmtu = IPV6_MIN_MTU;
+		allfrag = 1;
+	}
+
 	/* New mtu received -> path was valid.
 	   They are sent only in response to data packets,
 	   so that this nexthop apparently is reachable. --ANK
@@ -1197,6 +1199,8 @@ void rt6_pmtu_discovery(struct in6_addr *daddr, struct in6_addr *saddr,
 	 */
 	if (rt->rt6i_flags & RTF_CACHE) {
 		rt->u.dst.metrics[RTAX_MTU-1] = pmtu;
+		if (allfrag)
+			rt->u.dst.metrics[RTAX_FEATURES-1] |= RTAX_FEATURE_ALLFRAG;
 		dst_set_expires(&rt->u.dst, ip6_rt_mtu_expires);
 		rt->rt6i_flags |= RTF_MODIFIED|RTF_EXPIRES;
 		goto out;
@@ -1211,6 +1215,8 @@ void rt6_pmtu_discovery(struct in6_addr *daddr, struct in6_addr *saddr,
 		nrt = rt6_cow(rt, daddr, saddr);
 		if (!nrt->u.dst.error) {
 			nrt->u.dst.metrics[RTAX_MTU-1] = pmtu;
+			if (allfrag)
+				nrt->u.dst.metrics[RTAX_FEATURES-1] |= RTAX_FEATURE_ALLFRAG;
 			/* According to RFC 1981, detecting PMTU increase shouldn't be
 			   happened within 5 mins, the recommended timer is 10 mins.
 			   Here this route expiration time is set to ip6_rt_mtu_expires
@@ -1232,6 +1238,8 @@ void rt6_pmtu_discovery(struct in6_addr *daddr, struct in6_addr *saddr,
 		dst_set_expires(&nrt->u.dst, ip6_rt_mtu_expires);
 		nrt->rt6i_flags |= RTF_DYNAMIC|RTF_CACHE|RTF_EXPIRES;
 		nrt->u.dst.metrics[RTAX_MTU-1] = pmtu;
+		if (allfrag)
+			nrt->u.dst.metrics[RTAX_FEATURES-1] |= RTAX_FEATURE_ALLFRAG;
 		ip6_ins_rt(nrt, NULL, NULL);
 	}