ip: take care of last fragment in ip_append_data

While investigating a bit, I found ip_fragment() slow path was taken
because ip_append_data() provides following layout for a send(MTU +
N*(MTU - 20)) syscall :

- one skb with 1500 (mtu) bytes
- N fragments of 1480 (mtu-20) bytes (before adding IP header)
last fragment gets 17 bytes of trail data because of following bit:

	if (datalen == length + fraggap)
		alloclen += rt->dst.trailer_len;

Then esp4 adds 16 bytes of data (while trailer_len is 17... hmm...
another bug ?)

In ip_fragment(), we notice last fragment is too big (1496 + 20) > mtu,
so we take slow path, building another skb chain.

In order to avoid taking slow path, we should correct ip_append_data()
to make sure last fragment has real trail space, under mtu...

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

1 files changed, 6 insertions, 3 deletions

diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c
index e42762023c2..3551b6dc741 100644
--- a/net/ipv4/ip_output.c
+++ b/net/ipv4/ip_output.c
@@ -926,16 +926,19 @@ alloc_new_skb:
 			    !(rt->dst.dev->features&NETIF_F_SG))
 				alloclen = mtu;
 			else
-				alloclen = datalen + fragheaderlen;
+				alloclen = fraglen;
 
 			/* The last fragment gets additional space at tail.
 			 * Note, with MSG_MORE we overallocate on fragments,
 			 * because we have no idea what fragment will be
 			 * the last.
 			 */
-			if (datalen == length + fraggap)
+			if (datalen == length + fraggap) {
 				alloclen += rt->dst.trailer_len;
-
+				/* make sure mtu is not reached */
+				if (datalen > mtu - fragheaderlen - rt->dst.trailer_len)
+					datalen -= ALIGN(rt->dst.trailer_len, 8);
+			}
 			if (transhdrlen) {
 				skb = sock_alloc_send_skb(sk,
 						alloclen + hh_len + 15,