tcp: TLP loss detection.

This is the second of the TLP patch series; it augments the basic TLP
algorithm with a loss detection scheme.

This patch implements a mechanism for loss detection when a Tail
loss probe retransmission plugs a hole thereby masking packet loss
from the sender. The loss detection algorithm relies on counting
TLP dupacks as outlined in Sec. 3 of:
http://tools.ietf.org/html/draft-dukkipati-tcpm-tcp-loss-probe-01

The basic idea is: Sender keeps track of TLP "episode" upon
retransmission of a TLP packet. An episode ends when the sender receives
an ACK above the SND.NXT (tracked by tlp_high_seq) at the time of the
episode. We want to make sure that before the episode ends the sender
receives a "TLP dupack", indicating that the TLP retransmission was
unnecessary, so there was no loss/hole that needed plugging. If the
sender gets no TLP dupack before the end of the episode, then it reduces
ssthresh and the congestion window, because the TLP packet arriving at
the receiver probably plugged a hole.

Signed-off-by: Nandita Dukkipati <nanditad@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

include/linux/tcp.h

@@ -204,6 +204,7 @@ struct tcp_sock {
 		syn_data:1,	/* SYN includes data */
 		syn_fastopen:1,	/* SYN includes Fast Open option */
 		syn_data_acked:1;/* data in SYN is acked by SYN-ACK */
+	u32	tlp_high_seq;	/* snd_nxt at the time of TLP retransmit. */
 
 /* RTT measurement */
 	u32	srtt;		/* smoothed round trip time << 3	*/

include/uapi/linux/snmp.h

@@ -203,6 +203,7 @@ enum
 	LINUX_MIB_TCPSLOWSTARTRETRANS,		/* TCPSlowStartRetrans */
 	LINUX_MIB_TCPTIMEOUTS,			/* TCPTimeouts */
 	LINUX_MIB_TCPLOSSPROBES,		/* TCPLossProbes */
+	LINUX_MIB_TCPLOSSPROBERECOVERY,		/* TCPLossProbeRecovery */
 	LINUX_MIB_TCPRENORECOVERYFAIL,		/* TCPRenoRecoveryFail */
 	LINUX_MIB_TCPSACKRECOVERYFAIL,		/* TCPSackRecoveryFail */
 	LINUX_MIB_TCPSCHEDULERFAILED,		/* TCPSchedulerFailed */

net/ipv4/proc.c

@@ -225,6 +225,7 @@ static const struct snmp_mib snmp4_net_list[] = {
 	SNMP_MIB_ITEM("TCPSlowStartRetrans", LINUX_MIB_TCPSLOWSTARTRETRANS),
 	SNMP_MIB_ITEM("TCPTimeouts", LINUX_MIB_TCPTIMEOUTS),
 	SNMP_MIB_ITEM("TCPLossProbes", LINUX_MIB_TCPLOSSPROBES),
+	SNMP_MIB_ITEM("TCPLossProbeRecovery", LINUX_MIB_TCPLOSSPROBERECOVERY),
 	SNMP_MIB_ITEM("TCPRenoRecoveryFail", LINUX_MIB_TCPRENORECOVERYFAIL),
 	SNMP_MIB_ITEM("TCPSackRecoveryFail", LINUX_MIB_TCPSACKRECOVERYFAIL),
 	SNMP_MIB_ITEM("TCPSchedulerFailed", LINUX_MIB_TCPSCHEDULERFAILED),

net/ipv4/tcp_input.c

@@ -2682,6 +2682,7 @@ static void tcp_init_cwnd_reduction(struct sock *sk, const bool set_ssthresh)
 	struct tcp_sock *tp = tcp_sk(sk);
 
 	tp->high_seq = tp->snd_nxt;
+	tp->tlp_high_seq = 0;
 	tp->snd_cwnd_cnt = 0;
 	tp->prior_cwnd = tp->snd_cwnd;
 	tp->prr_delivered = 0;
@@ -3569,6 +3570,38 @@ static void tcp_send_challenge_ack(struct sock *sk)
 	}
 }
 
+/* This routine deals with acks during a TLP episode.
+ * Ref: loss detection algorithm in draft-dukkipati-tcpm-tcp-loss-probe.
+ */
+static void tcp_process_tlp_ack(struct sock *sk, u32 ack, int flag)
+{
+	struct tcp_sock *tp = tcp_sk(sk);
+	bool is_tlp_dupack = (ack == tp->tlp_high_seq) &&
+			     !(flag & (FLAG_SND_UNA_ADVANCED |
+				       FLAG_NOT_DUP | FLAG_DATA_SACKED));
+
+	/* Mark the end of TLP episode on receiving TLP dupack or when
+	 * ack is after tlp_high_seq.
+	 */
+	if (is_tlp_dupack) {
+		tp->tlp_high_seq = 0;
+		return;
+	}
+
+	if (after(ack, tp->tlp_high_seq)) {
+		tp->tlp_high_seq = 0;
+		/* Don't reduce cwnd if DSACK arrives for TLP retrans. */
+		if (!(flag & FLAG_DSACKING_ACK)) {
+			tcp_init_cwnd_reduction(sk, true);
+			tcp_set_ca_state(sk, TCP_CA_CWR);
+			tcp_end_cwnd_reduction(sk);
+			tcp_set_ca_state(sk, TCP_CA_Open);
+			NET_INC_STATS_BH(sock_net(sk),
+					 LINUX_MIB_TCPLOSSPROBERECOVERY);
+		}
+	}
+}
+
 /* This routine deals with incoming acks, but not outgoing ones. */
 static int tcp_ack(struct sock *sk, const struct sk_buff *skb, int flag)
 {
@@ -3676,6 +3709,9 @@ static int tcp_ack(struct sock *sk, const struct sk_buff *skb, int flag)
 			tcp_cong_avoid(sk, ack, prior_in_flight);
 	}
 
+	if (tp->tlp_high_seq)
+		tcp_process_tlp_ack(sk, ack, flag);
+
 	if ((flag & FLAG_FORWARD_PROGRESS) || !(flag & FLAG_NOT_DUP)) {
 		struct dst_entry *dst = __sk_dst_get(sk);
 		if (dst)
@@ -3697,6 +3733,9 @@ no_queue:
 	 */
 	if (tcp_send_head(sk))
 		tcp_ack_probe(sk);
+
+	if (tp->tlp_high_seq)
+		tcp_process_tlp_ack(sk, ack, flag);
 	return 1;
 
 invalid_ack:

net/ipv4/tcp_minisocks.c

@@ -440,6 +440,7 @@ struct sock *tcp_create_openreq_child(struct sock *sk, struct request_sock *req,
 		newtp->fackets_out = 0;
 		newtp->snd_ssthresh = TCP_INFINITE_SSTHRESH;
 		tcp_enable_early_retrans(newtp);
+		newtp->tlp_high_seq = 0;
 
 		/* So many TCP implementations out there (incorrectly) count the
 		 * initial SYN frame in their delayed-ACK and congestion control

net/ipv4/tcp_output.c

@@ -2132,6 +2132,7 @@ bool tcp_schedule_loss_probe(struct sock *sk)
  */
 void tcp_send_loss_probe(struct sock *sk)
 {
+	struct tcp_sock *tp = tcp_sk(sk);
 	struct sk_buff *skb;
 	int pcount;
 	int mss = tcp_current_mss(sk);
@@ -2142,6 +2143,10 @@ void tcp_send_loss_probe(struct sock *sk)
 		goto rearm_timer;
 	}
 
+	/* At most one outstanding TLP retransmission. */
+	if (tp->tlp_high_seq)
+		goto rearm_timer;
+
 	/* Retransmit last segment. */
 	skb = tcp_write_queue_tail(sk);
 	if (WARN_ON(!skb))
@@ -2164,6 +2169,10 @@ void tcp_send_loss_probe(struct sock *sk)
 	if (skb->len > 0)
 		err = __tcp_retransmit_skb(sk, skb);
 
+	/* Record snd_nxt for loss detection. */
+	if (likely(!err))
+		tp->tlp_high_seq = tp->snd_nxt;
+
 rearm_timer:
 	inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS,
 				  inet_csk(sk)->icsk_rto,

net/ipv4/tcp_timer.c

@@ -356,6 +356,8 @@ void tcp_retransmit_timer(struct sock *sk)
 
 	WARN_ON(tcp_write_queue_empty(sk));
 
+	tp->tlp_high_seq = 0;
+
 	if (!tp->snd_wnd && !sock_flag(sk, SOCK_DEAD) &&
 	    !((1 << sk->sk_state) & (TCPF_SYN_SENT | TCPF_SYN_RECV))) {
 		/* Receiver dastardly shrinks window. Our retransmits