Skip to content

Commit

Permalink
tcp: fix a stale ooo_last_skb after a replace
Browse files Browse the repository at this point in the history
When skb replaces another one in ooo queue, I forgot to also
update tp->ooo_last_skb as well, if the replaced skb was the last one
in the queue.

To fix this, we simply can re-use the code that runs after an insertion,
trying to merge skbs at the right of current skb.

This not only fixes the bug, but also remove all small skbs that might
be a subset of the new one.

Example:

We receive segments 2001:3001,  4001:5001

Then we receive 2001:8001 : We should replace 2001:3001 with the big
skb, but also remove 4001:50001 from the queue to save space.

packetdrill test demonstrating the bug

0.000 socket(..., SOCK_STREAM, IPPROTO_TCP) = 3
+0 setsockopt(3, SOL_SOCKET, SO_REUSEADDR, [1], 4) = 0
+0 bind(3, ..., ...) = 0
+0 listen(3, 1) = 0

+0 < S 0:0(0) win 32792 <mss 1000,sackOK,nop,nop,nop,wscale 7>
+0 > S. 0:0(0) ack 1 <mss 1460,nop,nop,sackOK,nop,wscale 7>
+0.100 < . 1:1(0) ack 1 win 1024
+0 accept(3, ..., ...) = 4

+0.01 < . 1001:2001(1000) ack 1 win 1024
+0    > . 1:1(0) ack 1 <nop,nop, sack 1001:2001>

+0.01 < . 1001:3001(2000) ack 1 win 1024
+0    > . 1:1(0) ack 1 <nop,nop, sack 1001:2001 1001:3001>

Fixes: 9f5afea ("tcp: use an RB tree for ooo receive queue")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Yuchung Cheng <ycheng@google.com>
Cc: Yaogong Wang <wygivan@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
  • Loading branch information
Eric Dumazet authored and David S. Miller committed Sep 16, 2016
1 parent bf0a6da commit 76f0dcb
Showing 1 changed file with 2 additions and 1 deletion.
3 changes: 2 additions & 1 deletion net/ipv4/tcp_input.c
Original file line number Diff line number Diff line change
Expand Up @@ -4502,7 +4502,7 @@ static void tcp_data_queue_ofo(struct sock *sk, struct sk_buff *skb)
NET_INC_STATS(sock_net(sk),
LINUX_MIB_TCPOFOMERGE);
__kfree_skb(skb1);
goto add_sack;
goto merge_right;
}
} else if (tcp_try_coalesce(sk, skb1, skb, &fragstolen)) {
goto coalesce_done;
Expand All @@ -4514,6 +4514,7 @@ static void tcp_data_queue_ofo(struct sock *sk, struct sk_buff *skb)
rb_link_node(&skb->rbnode, parent, p);
rb_insert_color(&skb->rbnode, &tp->out_of_order_queue);

merge_right:
/* Remove other segments covered by skb. */
while ((q = rb_next(&skb->rbnode)) != NULL) {
skb1 = rb_entry(q, struct sk_buff, rbnode);
Expand Down

0 comments on commit 76f0dcb

Please sign in to comment.