1 files changed, 121 insertions, 83 deletions

diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index 3ca2139a130b..25a89eaa669d 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -688,6 +688,34 @@ static void tcp_rtt_estimator(struct sock *sk, const __u32 mrtt)
 	}
 }
 
+/* Set the sk_pacing_rate to allow proper sizing of TSO packets.
+ * Note: TCP stack does not yet implement pacing.
+ * FQ packet scheduler can be used to implement cheap but effective
+ * TCP pacing, to smooth the burst on large writes when packets
+ * in flight is significantly lower than cwnd (or rwin)
+ */
+static void tcp_update_pacing_rate(struct sock *sk)
+{
+	const struct tcp_sock *tp = tcp_sk(sk);
+	u64 rate;
+
+	/* set sk_pacing_rate to 200 % of current rate (mss * cwnd / srtt) */
+	rate = (u64)tp->mss_cache * 2 * (HZ << 3);
+
+	rate *= max(tp->snd_cwnd, tp->packets_out);
+
+	/* Correction for small srtt : minimum srtt being 8 (1 jiffy << 3),
+	 * be conservative and assume srtt = 1 (125 us instead of 1.25 ms)
+	 * We probably need usec resolution in the future.
+	 * Note: This also takes care of possible srtt=0 case,
+	 * when tcp_rtt_estimator() was not yet called.
+	 */
+	if (tp->srtt > 8 + 2)
+		do_div(rate, tp->srtt);
+
+	sk->sk_pacing_rate = min_t(u64, rate, ~0U);
+}
+
 /* Calculate rto without backoff.  This is the second half of Van Jacobson's
  * routine referred to above.
  */
@@ -1048,6 +1076,7 @@ struct tcp_sacktag_state {
 	int reord;
 	int fack_count;
 	int flag;
+	s32 rtt; /* RTT measured by SACKing never-retransmitted data */
 };
 
 /* Check if skb is fully within the SACK block. In presence of GSO skbs,
@@ -1108,7 +1137,7 @@ static int tcp_match_skb_to_sack(struct sock *sk, struct sk_buff *skb,
 static u8 tcp_sacktag_one(struct sock *sk,
 			  struct tcp_sacktag_state *state, u8 sacked,
 			  u32 start_seq, u32 end_seq,
-			  bool dup_sack, int pcount)
+			  int dup_sack, int pcount, u32 xmit_time)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 	int fack_count = state->fack_count;
@@ -1148,6 +1177,9 @@ static u8 tcp_sacktag_one(struct sock *sk,
 							   state->reord);
 				if (!after(end_seq, tp->high_seq))
 					state->flag |= FLAG_ORIG_SACK_ACKED;
+				/* Pick the earliest sequence sacked for RTT */
+				if (state->rtt < 0)
+					state->rtt = tcp_time_stamp - xmit_time;
 			}
 
 			if (sacked & TCPCB_LOST) {
@@ -1205,7 +1237,8 @@ static bool tcp_shifted_skb(struct sock *sk, struct sk_buff *skb,
 	 * tcp_highest_sack_seq() when skb is highest_sack.
 	 */
 	tcp_sacktag_one(sk, state, TCP_SKB_CB(skb)->sacked,
-			start_seq, end_seq, dup_sack, pcount);
+			start_seq, end_seq, dup_sack, pcount,
+			TCP_SKB_CB(skb)->when);
 
 	if (skb == tp->lost_skb_hint)
 		tp->lost_cnt_hint += pcount;
@@ -1479,7 +1512,8 @@ static struct sk_buff *tcp_sacktag_walk(struct sk_buff *skb, struct sock *sk,
 						TCP_SKB_CB(skb)->seq,
 						TCP_SKB_CB(skb)->end_seq,
 						dup_sack,
-						tcp_skb_pcount(skb));
+						tcp_skb_pcount(skb),
+						TCP_SKB_CB(skb)->when);
 
 			if (!before(TCP_SKB_CB(skb)->seq,
 				    tcp_highest_sack_seq(tp)))
@@ -1536,7 +1570,7 @@ static int tcp_sack_cache_ok(const struct tcp_sock *tp, const struct tcp_sack_bl
 
 static int
 tcp_sacktag_write_queue(struct sock *sk, const struct sk_buff *ack_skb,
-			u32 prior_snd_una)
+			u32 prior_snd_una, s32 *sack_rtt)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 	const unsigned char *ptr = (skb_transport_header(ack_skb) +
@@ -1554,6 +1588,7 @@ tcp_sacktag_write_queue(struct sock *sk, const struct sk_buff *ack_skb,
 
 	state.flag = 0;
 	state.reord = tp->packets_out;
+	state.rtt = -1;
 
 	if (!tp->sacked_out) {
 		if (WARN_ON(tp->fackets_out))
@@ -1737,6 +1772,7 @@ out:
 	WARN_ON((int)tp->retrans_out < 0);
 	WARN_ON((int)tcp_packets_in_flight(tp) < 0);
 #endif
+	*sack_rtt = state.rtt;
 	return state.flag;
 }
 
@@ -1869,8 +1905,13 @@ void tcp_enter_loss(struct sock *sk, int how)
 	}
 	tcp_verify_left_out(tp);
 
-	tp->reordering = min_t(unsigned int, tp->reordering,
-			       sysctl_tcp_reordering);
+	/* Timeout in disordered state after receiving substantial DUPACKs
+	 * suggests that the degree of reordering is over-estimated.
+	 */
+	if (icsk->icsk_ca_state <= TCP_CA_Disorder &&
+	    tp->sacked_out >= sysctl_tcp_reordering)
+		tp->reordering = min_t(unsigned int, tp->reordering,
+				       sysctl_tcp_reordering);
 	tcp_set_ca_state(sk, TCP_CA_Loss);
 	tp->high_seq = tp->snd_nxt;
 	TCP_ECN_queue_cwr(tp);
@@ -2472,8 +2513,6 @@ static void tcp_try_to_open(struct sock *sk, int flag, const int prior_unsacked)
 
 	if (inet_csk(sk)->icsk_ca_state != TCP_CA_CWR) {
 		tcp_try_keep_open(sk);
-		if (inet_csk(sk)->icsk_ca_state != TCP_CA_Open)
-			tcp_moderate_cwnd(tp);
 	} else {
 		tcp_cwnd_reduction(sk, prior_unsacked, 0);
 	}
@@ -2792,65 +2831,51 @@ static void tcp_fastretrans_alert(struct sock *sk, const int acked,
 	tcp_xmit_retransmit_queue(sk);
 }
 
-void tcp_valid_rtt_meas(struct sock *sk, u32 seq_rtt)
+static inline bool tcp_ack_update_rtt(struct sock *sk, const int flag,
+				      s32 seq_rtt, s32 sack_rtt)
 {
-	tcp_rtt_estimator(sk, seq_rtt);
-	tcp_set_rto(sk);
-	inet_csk(sk)->icsk_backoff = 0;
-}
-EXPORT_SYMBOL(tcp_valid_rtt_meas);
+	const struct tcp_sock *tp = tcp_sk(sk);
+
+	/* Prefer RTT measured from ACK's timing to TS-ECR. This is because
+	 * broken middle-boxes or peers may corrupt TS-ECR fields. But
+	 * Karn's algorithm forbids taking RTT if some retransmitted data
+	 * is acked (RFC6298).
+	 */
+	if (flag & FLAG_RETRANS_DATA_ACKED)
+		seq_rtt = -1;
+
+	if (seq_rtt < 0)
+		seq_rtt = sack_rtt;
 
-/* Read draft-ietf-tcplw-high-performance before mucking
- * with this code. (Supersedes RFC1323)
- */
-static void tcp_ack_saw_tstamp(struct sock *sk, int flag)
-{
 	/* RTTM Rule: A TSecr value received in a segment is used to
 	 * update the averaged RTT measurement only if the segment
 	 * acknowledges some new data, i.e., only if it advances the
 	 * left edge of the send window.
-	 *
 	 * See draft-ietf-tcplw-high-performance-00, section 3.3.
-	 * 1998/04/10 Andrey V. Savochkin <saw@msu.ru>
-	 *
-	 * Changed: reset backoff as soon as we see the first valid sample.
-	 * If we do not, we get strongly overestimated rto. With timestamps
-	 * samples are accepted even from very old segments: f.e., when rtt=1
-	 * increases to 8, we retransmit 5 times and after 8 seconds delayed
-	 * answer arrives rto becomes 120 seconds! If at least one of segments
-	 * in window is lost... Voila.	 			--ANK (010210)
 	 */
-	struct tcp_sock *tp = tcp_sk(sk);
-
-	tcp_valid_rtt_meas(sk, tcp_time_stamp - tp->rx_opt.rcv_tsecr);
-}
+	if (seq_rtt < 0 && tp->rx_opt.saw_tstamp && tp->rx_opt.rcv_tsecr)
+		seq_rtt = tcp_time_stamp - tp->rx_opt.rcv_tsecr;
 
-static void tcp_ack_no_tstamp(struct sock *sk, u32 seq_rtt, int flag)
-{
-	/* We don't have a timestamp. Can only use
-	 * packets that are not retransmitted to determine
-	 * rtt estimates. Also, we must not reset the
-	 * backoff for rto until we get a non-retransmitted
-	 * packet. This allows us to deal with a situation
-	 * where the network delay has increased suddenly.
-	 * I.e. Karn's algorithm. (SIGCOMM '87, p5.)
-	 */
+	if (seq_rtt < 0)
+		return false;
 
-	if (flag & FLAG_RETRANS_DATA_ACKED)
-		return;
+	tcp_rtt_estimator(sk, seq_rtt);
+	tcp_set_rto(sk);
 
-	tcp_valid_rtt_meas(sk, seq_rtt);
+	/* RFC6298: only reset backoff on valid RTT measurement. */
+	inet_csk(sk)->icsk_backoff = 0;
+	return true;
 }
 
-static inline void tcp_ack_update_rtt(struct sock *sk, const int flag,
-				      const s32 seq_rtt)
+/* Compute time elapsed between (last) SYNACK and the ACK completing 3WHS. */
+static void tcp_synack_rtt_meas(struct sock *sk, struct request_sock *req)
 {
-	const struct tcp_sock *tp = tcp_sk(sk);
-	/* Note that peer MAY send zero echo. In this case it is ignored. (rfc1323) */
-	if (tp->rx_opt.saw_tstamp && tp->rx_opt.rcv_tsecr)
-		tcp_ack_saw_tstamp(sk, flag);
-	else if (seq_rtt >= 0)
-		tcp_ack_no_tstamp(sk, seq_rtt, flag);
+	struct tcp_sock *tp = tcp_sk(sk);
+	s32 seq_rtt = -1;
+
+	if (tp->lsndtime && !tp->total_retrans)
+		seq_rtt = tcp_time_stamp - tp->lsndtime;
+	tcp_ack_update_rtt(sk, FLAG_SYN_ACKED, seq_rtt, -1);
 }
 
 static void tcp_cong_avoid(struct sock *sk, u32 ack, u32 in_flight)
@@ -2939,7 +2964,7 @@ static u32 tcp_tso_acked(struct sock *sk, struct sk_buff *skb)
  * arrived at the other end.
  */
 static int tcp_clean_rtx_queue(struct sock *sk, int prior_fackets,
-			       u32 prior_snd_una)
+			       u32 prior_snd_una, s32 sack_rtt)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 	const struct inet_connection_sock *icsk = inet_csk(sk);
@@ -2978,8 +3003,6 @@ static int tcp_clean_rtx_queue(struct sock *sk, int prior_fackets,
 			if (sacked & TCPCB_SACKED_RETRANS)
 				tp->retrans_out -= acked_pcount;
 			flag |= FLAG_RETRANS_DATA_ACKED;
-			ca_seq_rtt = -1;
-			seq_rtt = -1;
 		} else {
 			ca_seq_rtt = now - scb->when;
 			last_ackt = skb->tstamp;
@@ -3031,6 +3054,10 @@ static int tcp_clean_rtx_queue(struct sock *sk, int prior_fackets,
 	if (skb && (TCP_SKB_CB(skb)->sacked & TCPCB_SACKED_ACKED))
 		flag |= FLAG_SACK_RENEGING;
 
+	if (tcp_ack_update_rtt(sk, flag, seq_rtt, sack_rtt) ||
+	    (flag & FLAG_ACKED))
+		tcp_rearm_rto(sk);
+
 	if (flag & FLAG_ACKED) {
 		const struct tcp_congestion_ops *ca_ops
 			= inet_csk(sk)->icsk_ca_ops;
@@ -3040,9 +3067,6 @@ static int tcp_clean_rtx_queue(struct sock *sk, int prior_fackets,
 			tcp_mtup_probe_success(sk);
 		}
 
-		tcp_ack_update_rtt(sk, flag, seq_rtt);
-		tcp_rearm_rto(sk);
-
 		if (tcp_is_reno(tp)) {
 			tcp_remove_reno_sacks(sk, pkts_acked);
 		} else {
@@ -3130,11 +3154,22 @@ static inline bool tcp_ack_is_dubious(const struct sock *sk, const int flag)
 		inet_csk(sk)->icsk_ca_state != TCP_CA_Open;
 }
 
+/* Decide wheather to run the increase function of congestion control. */
 static inline bool tcp_may_raise_cwnd(const struct sock *sk, const int flag)
 {
-	const struct tcp_sock *tp = tcp_sk(sk);
-	return (!(flag & FLAG_ECE) || tp->snd_cwnd < tp->snd_ssthresh) &&
-		!tcp_in_cwnd_reduction(sk);
+	if (tcp_in_cwnd_reduction(sk))
+		return false;
+
+	/* If reordering is high then always grow cwnd whenever data is
+	 * delivered regardless of its ordering. Otherwise stay conservative
+	 * and only grow cwnd on in-order delivery (RFC5681). A stretched ACK w/
+	 * new SACK or ECE mark may first advance cwnd here and later reduce
+	 * cwnd in tcp_fastretrans_alert() based on more states.
+	 */
+	if (tcp_sk(sk)->reordering > sysctl_tcp_reordering)
+		return flag & FLAG_FORWARD_PROGRESS;
+
+	return flag & FLAG_DATA_ACKED;
 }
 
 /* Check that window update is acceptable.
@@ -3269,11 +3304,12 @@ static int tcp_ack(struct sock *sk, const struct sk_buff *skb, int flag)
 	u32 ack_seq = TCP_SKB_CB(skb)->seq;
 	u32 ack = TCP_SKB_CB(skb)->ack_seq;
 	bool is_dupack = false;
-	u32 prior_in_flight;
+	u32 prior_in_flight, prior_cwnd = tp->snd_cwnd, prior_rtt = tp->srtt;
 	u32 prior_fackets;
 	int prior_packets = tp->packets_out;
 	const int prior_unsacked = tp->packets_out - tp->sacked_out;
 	int acked = 0; /* Number of packets newly acked */
+	s32 sack_rtt = -1;
 
 	/* If the ack is older than previous acks
 	 * then we can probably ignore it.
@@ -3330,7 +3366,8 @@ static int tcp_ack(struct sock *sk, const struct sk_buff *skb, int flag)
 		flag |= tcp_ack_update_window(sk, skb, ack, ack_seq);
 
 		if (TCP_SKB_CB(skb)->sacked)
-			flag |= tcp_sacktag_write_queue(sk, skb, prior_snd_una);
+			flag |= tcp_sacktag_write_queue(sk, skb, prior_snd_una,
+							&sack_rtt);
 
 		if (TCP_ECN_rcv_ecn_echo(tp, tcp_hdr(skb)))
 			flag |= FLAG_ECE;
@@ -3349,21 +3386,18 @@ static int tcp_ack(struct sock *sk, const struct sk_buff *skb, int flag)
 
 	/* See if we can take anything off of the retransmit queue. */
 	acked = tp->packets_out;
-	flag |= tcp_clean_rtx_queue(sk, prior_fackets, prior_snd_una);
+	flag |= tcp_clean_rtx_queue(sk, prior_fackets, prior_snd_una, sack_rtt);
 	acked -= tp->packets_out;
 
+	/* Advance cwnd if state allows */
+	if (tcp_may_raise_cwnd(sk, flag))
+		tcp_cong_avoid(sk, ack, prior_in_flight);
+
 	if (tcp_ack_is_dubious(sk, flag)) {
-		/* Advance CWND, if state allows this. */
-		if ((flag & FLAG_DATA_ACKED) && tcp_may_raise_cwnd(sk, flag))
-			tcp_cong_avoid(sk, ack, prior_in_flight);
 		is_dupack = !(flag & (FLAG_SND_UNA_ADVANCED | FLAG_NOT_DUP));
 		tcp_fastretrans_alert(sk, acked, prior_unsacked,
 				      is_dupack, flag);
-	} else {
-		if (flag & FLAG_DATA_ACKED)
-			tcp_cong_avoid(sk, ack, prior_in_flight);
 	}
-
 	if (tp->tlp_high_seq)
 		tcp_process_tlp_ack(sk, ack, flag);
 
@@ -3375,6 +3409,8 @@ static int tcp_ack(struct sock *sk, const struct sk_buff *skb, int flag)
 
 	if (icsk->icsk_pending == ICSK_TIME_RETRANS)
 		tcp_schedule_loss_probe(sk);
+	if (tp->srtt != prior_rtt || tp->snd_cwnd != prior_cwnd)
+		tcp_update_pacing_rate(sk);
 	return 1;
 
 no_queue:
@@ -3402,7 +3438,8 @@ old_ack:
 	 * If data was DSACKed, see if we can undo a cwnd reduction.
 	 */
 	if (TCP_SKB_CB(skb)->sacked) {
-		flag |= tcp_sacktag_write_queue(sk, skb, prior_snd_una);
+		flag |= tcp_sacktag_write_queue(sk, skb, prior_snd_una,
+						&sack_rtt);
 		tcp_fastretrans_alert(sk, acked, prior_unsacked,
 				      is_dupack, flag);
 	}
@@ -4102,6 +4139,7 @@ static void tcp_data_queue_ofo(struct sock *sk, struct sk_buff *skb)
 		if (!tcp_try_coalesce(sk, skb1, skb, &fragstolen)) {
 			__skb_queue_after(&tp->out_of_order_queue, skb1, skb);
 		} else {
+			tcp_grow_window(sk, skb);
 			kfree_skb_partial(skb, fragstolen);
 			skb = NULL;
 		}
@@ -4177,8 +4215,10 @@ add_sack:
 	if (tcp_is_sack(tp))
 		tcp_sack_new_ofo_skb(sk, seq, end_seq);
 end:
-	if (skb)
+	if (skb) {
+		tcp_grow_window(sk, skb);
 		skb_set_owner_r(skb, sk);
+	}
 }
 
 static int __must_check tcp_queue_rcv(struct sock *sk, struct sk_buff *skb, int hdrlen,
@@ -5013,8 +5053,8 @@ discard:
  *	the rest is checked inline. Fast processing is turned on in
  *	tcp_data_queue when everything is OK.
  */
-int tcp_rcv_established(struct sock *sk, struct sk_buff *skb,
-			const struct tcphdr *th, unsigned int len)
+void tcp_rcv_established(struct sock *sk, struct sk_buff *skb,
+			 const struct tcphdr *th, unsigned int len)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 
@@ -5091,7 +5131,7 @@ int tcp_rcv_established(struct sock *sk, struct sk_buff *skb,
 				tcp_ack(sk, skb, 0);
 				__kfree_skb(skb);
 				tcp_data_snd_check(sk);
-				return 0;
+				return;
 			} else { /* Header too small */
 				TCP_INC_STATS_BH(sock_net(sk), TCP_MIB_INERRS);
 				goto discard;
@@ -5184,7 +5224,7 @@ no_ack:
 			if (eaten)
 				kfree_skb_partial(skb, fragstolen);
 			sk->sk_data_ready(sk, 0);
-			return 0;
+			return;
 		}
 	}
 
@@ -5200,7 +5240,7 @@ slow_path:
 	 */
 
 	if (!tcp_validate_incoming(sk, skb, th, 1))
-		return 0;
+		return;
 
 step5:
 	if (tcp_ack(sk, skb, FLAG_SLOWPATH | FLAG_UPDATE_TS_RECENT) < 0)
@@ -5216,7 +5256,7 @@ step5:
 
 	tcp_data_snd_check(sk);
 	tcp_ack_snd_check(sk);
-	return 0;
+	return;
 
 csum_error:
 	TCP_INC_STATS_BH(sock_net(sk), TCP_MIB_CSUMERRORS);
@@ -5224,7 +5264,6 @@ csum_error:
 
 discard:
 	__kfree_skb(skb);
-	return 0;
 }
 EXPORT_SYMBOL(tcp_rcv_established);
 
@@ -5627,9 +5666,7 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb,
 		 * so release it.
 		 */
 		if (req) {
-			tcp_synack_rtt_meas(sk, req);
 			tp->total_retrans = req->num_retrans;
-
 			reqsk_fastopen_remove(sk, req, false);
 		} else {
 			/* Make sure socket is routed, for correct metrics. */
@@ -5654,6 +5691,7 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb,
 		tp->snd_una = TCP_SKB_CB(skb)->ack_seq;
 		tp->snd_wnd = ntohs(th->window) << tp->rx_opt.snd_wscale;
 		tcp_init_wl(tp, TCP_SKB_CB(skb)->seq);
+		tcp_synack_rtt_meas(sk, req);
 
 		if (tp->rx_opt.tstamp_ok)
 			tp->advmss -= TCPOLEN_TSTAMP_ALIGNED;