[TCP]: Re-place highest_sack check to a more robust position
authorIlpo Järvinen <ilpo.jarvinen@helsinki.fi>
Wed, 26 Sep 2007 05:44:14 +0000 (22:44 -0700)
committerDavid S. Miller <davem@sunset.davemloft.net>
Wed, 10 Oct 2007 23:52:38 +0000 (16:52 -0700)
I previously added checking to position that is rather poor as
state has already been adjusted quite a bit. Re-placing it above
all state changes should be more robust though the return should
never ever get executed regardless of its place :-).

Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>
net/ipv4/tcp_output.c

index cbb83acd830a89009ae88f522551340bcd267399..94c80113df7e06e3e3e92e63b5db6fde7c72f889 100644 (file)
@@ -1718,6 +1718,10 @@ static void tcp_retrans_try_collapse(struct sock *sk, struct sk_buff *skb, int m
                BUG_ON(tcp_skb_pcount(skb) != 1 ||
                       tcp_skb_pcount(next_skb) != 1);
 
+               if (WARN_ON(tp->sacked_out &&
+                   (TCP_SKB_CB(next_skb)->seq == tp->highest_sack)))
+                       return;
+
                /* Ok.  We will be able to collapse the packet. */
                tcp_unlink_write_queue(next_skb, sk);
 
@@ -1734,10 +1738,6 @@ static void tcp_retrans_try_collapse(struct sock *sk, struct sk_buff *skb, int m
                /* Update sequence range on original skb. */
                TCP_SKB_CB(skb)->end_seq = TCP_SKB_CB(next_skb)->end_seq;
 
-               if (WARN_ON(tp->sacked_out &&
-                   (TCP_SKB_CB(next_skb)->seq == tp->highest_sack)))
-                       return;
-
                /* Merge over control information. */
                flags |= TCP_SKB_CB(next_skb)->flags; /* This moves PSH/FIN etc. over */
                TCP_SKB_CB(skb)->flags = flags;