From: John Heffner <jheffner@psc.edu>
Date: Wed, 7 Feb 2007 05:57:34 +0000 (-0800)
Subject: Fix TCP FIN handling
X-Git-Tag: v2.6.20.2~86
X-Git-Url: http://git.ipfire.org/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=c1689102c93ed8754c2c1c09139e90c10f59683a;p=thirdparty%2Fkernel%2Fstable.git

Fix TCP FIN handling

We can accidently spit out a huge burst of packets with TSO
when the FIN back is piggybacked onto the final packet.

[TCP]: Don't apply FIN exception to full TSO segments.

Signed-off-by: John Heffner <jheffner@psc.edu>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
---

diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c
index 975f4472af29c..58b7111523f4a 100644
--- a/net/ipv4/tcp_output.c
+++ b/net/ipv4/tcp_output.c
@@ -965,7 +965,8 @@ static inline unsigned int tcp_cwnd_test(struct tcp_sock *tp, struct sk_buff *sk
 	u32 in_flight, cwnd;
 
 	/* Don't be strict about the congestion window for the final FIN.  */
-	if (TCP_SKB_CB(skb)->flags & TCPCB_FLAG_FIN)
+	if ((TCP_SKB_CB(skb)->flags & TCPCB_FLAG_FIN) &&
+	    tcp_skb_pcount(skb) == 1)
 		return 1;
 
 	in_flight = tcp_packets_in_flight(tp);