[TCP] FRTO: RFC4138 allows Nagle override when new data must be sent

This is a corner case where less than MSS sized new data thingie
is awaiting in the send queue. For F-RTO to work correctly, a
new data segment must be sent at certain point or F-RTO cannot
be used at all. RFC4138 allows overriding of Nagle at that
point.

Implementation uses frto_counter states 2 and 3 to distinguish
when Nagle override is needed.

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

3 files changed, 17 insertions, 7 deletions

diff --git a/include/net/tcp.h b/include/net/tcp.h
index c6ecd455edab..ef8f9d4dae85 100644
--- a/include/net/tcp.h
+++ b/include/net/tcp.h
@@ -1199,9 +1199,14 @@ static inline struct sk_buff *tcp_send_head(struct sock *sk)
 
 static inline void tcp_advance_send_head(struct sock *sk, struct sk_buff *skb)
 {
+	struct tcp_sock *tp = tcp_sk(sk);
+
 	sk->sk_send_head = skb->next;
 	if (sk->sk_send_head == (struct sk_buff *)&sk->sk_write_queue)
 		sk->sk_send_head = NULL;
+	/* Don't override Nagle indefinately with F-RTO */
+	if (tp->frto_counter == 2)
+		tp->frto_counter = 3;
 }
 
 static inline void tcp_check_send_head(struct sock *sk, struct sk_buff *skb_unlinked)
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index 6b669898b197..7641b2761a14 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -2637,7 +2637,9 @@ static void tcp_undo_spur_to_response(struct sock *sk, int flag)
  *                  algorithm is not part of the F-RTO detection algorithm
  *                  given in RFC4138 but can be selected separately).
  * Otherwise (basically on duplicate ACK), RTO was (likely) caused by a loss
- * and TCP falls back to conventional RTO recovery.
+ * and TCP falls back to conventional RTO recovery. F-RTO allows overriding
+ * of Nagle, this is done using frto_counter states 2 and 3, when a new data
+ * segment of any size sent during F-RTO, state 2 is upgraded to 3.
  *
  * Rationale: if the RTO was spurious, new ACKs should arrive from the
  * original window even after we transmit two new data segments.
@@ -2666,7 +2668,7 @@ static int tcp_process_frto(struct sock *sk, u32 prior_snd_una, int flag)
 		inet_csk(sk)->icsk_retransmits = 0;
 
 	if (!before(tp->snd_una, tp->frto_highmark)) {
-		tcp_enter_frto_loss(sk, tp->frto_counter + 1, flag);
+		tcp_enter_frto_loss(sk, (tp->frto_counter == 1 ? 2 : 3), flag);
 		return 1;
 	}
 
@@ -2692,7 +2694,7 @@ static int tcp_process_frto(struct sock *sk, u32 prior_snd_una, int flag)
 			return 1;
 		}
 
-		if ((tp->frto_counter == 2) &&
+		if ((tp->frto_counter >= 2) &&
 		    (!(flag&FLAG_FORWARD_PROGRESS) ||
 		     ((flag&FLAG_DATA_SACKED) && !(flag&FLAG_ONLY_ORIG_SACKED)))) {
 			/* RFC4138 shortcoming (see comment above) */
@@ -2709,14 +2711,15 @@ static int tcp_process_frto(struct sock *sk, u32 prior_snd_una, int flag)
 		if (!tcp_send_head(sk) ||
 		    after(TCP_SKB_CB(tcp_send_head(sk))->end_seq,
 				     tp->snd_una + tp->snd_wnd)) {
-			tcp_enter_frto_loss(sk, tp->frto_counter + 1, flag);
+			tcp_enter_frto_loss(sk, (tp->frto_counter == 1 ? 2 : 3),
+					    flag);
 			return 1;
 		}
 
 		tp->snd_cwnd = tcp_packets_in_flight(tp) + 2;
 		tp->frto_counter = 2;
 		return 1;
-	} else /* frto_counter == 2 */ {
+	} else {
 		switch (sysctl_tcp_frto_response) {
 		case 2:
 			tcp_undo_spur_to_response(sk, flag);
diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c
index b5fa3c19afee..0faacf9c419d 100644
--- a/net/ipv4/tcp_output.c
+++ b/net/ipv4/tcp_output.c
@@ -1035,8 +1035,10 @@ static inline int tcp_nagle_test(struct tcp_sock *tp, struct sk_buff *skb,
 	if (nonagle & TCP_NAGLE_PUSH)
 		return 1;
 
-	/* Don't use the nagle rule for urgent data (or for the final FIN).  */
-	if (tp->urg_mode ||
+	/* Don't use the nagle rule for urgent data (or for the final FIN).
+	 * Nagle can be ignored during F-RTO too (see RFC4138).
+	 */
+	if (tp->urg_mode || (tp->frto_counter == 2) ||
 	    (TCP_SKB_CB(skb)->flags & TCPCB_FLAG_FIN))
 		return 1;