1 files changed, 22 insertions, 117 deletions

diff --git a/net/ipv4/tcp_bic.c b/net/ipv4/tcp_bic.c
index ae35e060904..d5de69bc04f 100644
--- a/net/ipv4/tcp_bic.c
+++ b/net/ipv4/tcp_bic.c
@@ -1,18 +1,18 @@
 /*
  * Binary Increase Congestion control for TCP
- *
+ * Home page:
+ *      http://netsrv.csc.ncsu.edu/twiki/bin/view/Main/BIC
  * This is from the implementation of BICTCP in
  * Lison-Xu, Kahaled Harfoush, and Injong Rhee.
  *  "Binary Increase Congestion Control for Fast, Long Distance
  *  Networks" in InfoComm 2004
  * Available from:
- *  http://www.csc.ncsu.edu/faculty/rhee/export/bitcp.pdf
+ *  http://netsrv.csc.ncsu.edu/export/bitcp.pdf
  *
  * Unless BIC is enabled and congestion window is large
  * this behaves the same as the original Reno.
  */
 
-#include <linux/config.h>
 #include <linux/mm.h>
 #include <linux/module.h>
 #include <net/tcp.h>
@@ -30,9 +30,7 @@ static int fast_convergence = 1;
 static int max_increment = 16;
 static int low_window = 14;
 static int beta = 819;		/* = 819/1024 (BICTCP_BETA_SCALE) */
-static int low_utilization_threshold = 153;
-static int low_utilization_period = 2;
-static int initial_ssthresh = 100;
+static int initial_ssthresh;
 static int smooth_part = 20;
 
 module_param(fast_convergence, int, 0644);
@@ -43,10 +41,6 @@ module_param(low_window, int, 0644);
 MODULE_PARM_DESC(low_window, "lower bound on congestion window (for TCP friendliness)");
 module_param(beta, int, 0644);
 MODULE_PARM_DESC(beta, "beta for multiplicative increase");
-module_param(low_utilization_threshold, int, 0644);
-MODULE_PARM_DESC(low_utilization_threshold, "percent (scaled by 1024) for low utilization mode");
-module_param(low_utilization_period, int, 0644);
-MODULE_PARM_DESC(low_utilization_period, "if average delay exceeds then goto to low utilization mode (seconds)");
 module_param(initial_ssthresh, int, 0644);
 MODULE_PARM_DESC(initial_ssthresh, "initial value of slow start threshold");
 module_param(smooth_part, int, 0644);
@@ -60,11 +54,6 @@ struct bictcp {
 	u32	loss_cwnd;	/* congestion window at last loss */
 	u32	last_cwnd;	/* the last snd_cwnd */
 	u32	last_time;	/* time when updated last_cwnd */
-	u32	delay_min;	/* min delay */
-	u32	delay_max;	/* max delay */
-	u32	last_delay;
-	u8	low_utilization;/* 0: high; 1: low */
-	u32	low_utilization_start;	/* starting time of low utilization detection*/
 	u32	epoch_start;	/* beginning of an epoch */
 #define ACK_RATIO_SHIFT	4
 	u32	delayed_ack;	/* estimate the ratio of Packets/ACKs << 4 */
@@ -74,21 +63,19 @@ static inline void bictcp_reset(struct bictcp *ca)
 {
 	ca->cnt = 0;
 	ca->last_max_cwnd = 0;
-	ca->loss_cwnd = 0;
 	ca->last_cwnd = 0;
 	ca->last_time = 0;
-	ca->delay_min = 0;
-	ca->delay_max = 0;
-	ca->last_delay = 0;
-	ca->low_utilization = 0;
-	ca->low_utilization_start = 0;
 	ca->epoch_start = 0;
 	ca->delayed_ack = 2 << ACK_RATIO_SHIFT;
 }
 
 static void bictcp_init(struct sock *sk)
 {
-	bictcp_reset(inet_csk_ca(sk));
+	struct bictcp *ca = inet_csk_ca(sk);
+
+	bictcp_reset(ca);
+	ca->loss_cwnd = 0;
+
 	if (initial_ssthresh)
 		tcp_sk(sk)->snd_ssthresh = initial_ssthresh;
 }
@@ -143,8 +130,7 @@ static inline void bictcp_update(struct bictcp *ca, u32 cwnd)
 	}
 
 	/* if in slow start or link utilization is very low */
-	if ( ca->loss_cwnd == 0 ||
-	     (cwnd > ca->loss_cwnd && ca->low_utilization)) {
+	if (ca->last_max_cwnd == 0) {
 		if (ca->cnt > 20) /* increase cwnd 5% per RTT */
 			ca->cnt = 20;
 	}
@@ -154,88 +140,19 @@ static inline void bictcp_update(struct bictcp *ca, u32 cwnd)
 		ca->cnt = 1;
 }
 
-
-/* Detect low utilization in congestion avoidance */
-static inline void bictcp_low_utilization(struct sock *sk, int flag)
-{
-	const struct tcp_sock *tp = tcp_sk(sk);
-	struct bictcp *ca = inet_csk_ca(sk);
-	u32 dist, delay;
-
-	/* No time stamp */
-	if (!(tp->rx_opt.saw_tstamp && tp->rx_opt.rcv_tsecr) ||
-	     /* Discard delay samples right after fast recovery */
-	     tcp_time_stamp < ca->epoch_start + HZ ||
-	     /* this delay samples may not be accurate */
-	     flag == 0) {
-		ca->last_delay = 0;
-		goto notlow;
-	}
-
-	delay = ca->last_delay<<3;	/* use the same scale as tp->srtt*/
-	ca->last_delay = tcp_time_stamp - tp->rx_opt.rcv_tsecr;
-	if (delay == 0) 		/* no previous delay sample */
-		goto notlow;
-
-	/* first time call or link delay decreases */
-	if (ca->delay_min == 0 || ca->delay_min > delay) {
-		ca->delay_min = ca->delay_max = delay;
-		goto notlow;
-	}
-
-	if (ca->delay_max < delay)
-		ca->delay_max = delay;
-
-	/* utilization is low, if avg delay < dist*threshold
-	   for checking_period time */
-	dist = ca->delay_max - ca->delay_min;
-	if (dist <= ca->delay_min>>6 ||
-	    tp->srtt - ca->delay_min >=  (dist*low_utilization_threshold)>>10)
-		goto notlow;
-
-	if (ca->low_utilization_start == 0) {
-		ca->low_utilization = 0;
-		ca->low_utilization_start = tcp_time_stamp;
-	} else if ((s32)(tcp_time_stamp - ca->low_utilization_start)
-			> low_utilization_period*HZ) {
-		ca->low_utilization = 1;
-	}
-
-	return;
-
- notlow:
-	ca->low_utilization = 0;
-	ca->low_utilization_start = 0;
-
-}
-
-static void bictcp_cong_avoid(struct sock *sk, u32 ack,
-			      u32 seq_rtt, u32 in_flight, int data_acked)
+static void bictcp_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 	struct bictcp *ca = inet_csk_ca(sk);
 
-	bictcp_low_utilization(sk, data_acked);
-
-	if (in_flight < tp->snd_cwnd)
+	if (!tcp_is_cwnd_limited(sk))
 		return;
 
-	if (tp->snd_cwnd <= tp->snd_ssthresh) {
-		/* In "safe" area, increase. */
-		if (tp->snd_cwnd < tp->snd_cwnd_clamp)
-			tp->snd_cwnd++;
-	} else {
+	if (tp->snd_cwnd <= tp->snd_ssthresh)
+		tcp_slow_start(tp, acked);
+	else {
 		bictcp_update(ca, tp->snd_cwnd);
-
-                /* In dangerous area, increase slowly.
-		 * In theory this is tp->snd_cwnd += 1 / tp->snd_cwnd
-		 */
-		if (tp->snd_cwnd_cnt >= ca->cnt) {
-			if (tp->snd_cwnd < tp->snd_cwnd_clamp)
-				tp->snd_cwnd++;
-			tp->snd_cwnd_cnt = 0;
-		} else
-			tp->snd_cwnd_cnt++;
+		tcp_cong_avoid_ai(tp, ca->cnt);
 	}
 
 }
@@ -251,11 +168,6 @@ static u32 bictcp_recalc_ssthresh(struct sock *sk)
 
 	ca->epoch_start = 0;	/* end of epoch */
 
-	/* in case of wrong delay_max*/
-	if (ca->delay_min > 0 && ca->delay_max > ca->delay_min)
-		ca->delay_max = ca->delay_min
-			+ ((ca->delay_max - ca->delay_min)* 90) / 100;
-
 	/* Wmax and fast convergence */
 	if (tp->snd_cwnd < ca->last_max_cwnd && fast_convergence)
 		ca->last_max_cwnd = (tp->snd_cwnd * (BICTCP_BETA_SCALE + beta))
@@ -276,13 +188,7 @@ static u32 bictcp_undo_cwnd(struct sock *sk)
 {
 	const struct tcp_sock *tp = tcp_sk(sk);
 	const struct bictcp *ca = inet_csk_ca(sk);
-	return max(tp->snd_cwnd, ca->last_max_cwnd);
-}
-
-static u32 bictcp_min_cwnd(struct sock *sk)
-{
-	const struct tcp_sock *tp = tcp_sk(sk);
-	return tp->snd_ssthresh;
+	return max(tp->snd_cwnd, ca->loss_cwnd);
 }
 
 static void bictcp_state(struct sock *sk, u8 new_state)
@@ -291,14 +197,14 @@ static void bictcp_state(struct sock *sk, u8 new_state)
 		bictcp_reset(inet_csk_ca(sk));
 }
 
-/* Track delayed acknowledgement ratio using sliding window
+/* Track delayed acknowledgment ratio using sliding window
  * ratio = (15*ratio + sample) / 16
  */
-static void bictcp_acked(struct sock *sk, u32 cnt)
+static void bictcp_acked(struct sock *sk, u32 cnt, s32 rtt)
 {
 	const struct inet_connection_sock *icsk = inet_csk(sk);
 
-	if (cnt > 0 && 	icsk->icsk_ca_state == TCP_CA_Open) {
+	if (icsk->icsk_ca_state == TCP_CA_Open) {
 		struct bictcp *ca = inet_csk_ca(sk);
 		cnt -= ca->delayed_ack >> ACK_RATIO_SHIFT;
 		ca->delayed_ack += cnt;
@@ -306,13 +212,12 @@ static void bictcp_acked(struct sock *sk, u32 cnt)
 }
 
 
-static struct tcp_congestion_ops bictcp = {
+static struct tcp_congestion_ops bictcp __read_mostly = {
 	.init		= bictcp_init,
 	.ssthresh	= bictcp_recalc_ssthresh,
 	.cong_avoid	= bictcp_cong_avoid,
 	.set_state	= bictcp_state,
 	.undo_cwnd	= bictcp_undo_cwnd,
-	.min_cwnd	= bictcp_min_cwnd,
 	.pkts_acked     = bictcp_acked,
 	.owner		= THIS_MODULE,
 	.name		= "bic",
@@ -320,7 +225,7 @@ static struct tcp_congestion_ops bictcp = {
 
 static int __init bictcp_register(void)
 {
-	BUG_ON(sizeof(struct bictcp) > ICSK_CA_PRIV_SIZE);
+	BUILD_BUG_ON(sizeof(struct bictcp) > ICSK_CA_PRIV_SIZE);
 	return tcp_register_congestion_control(&bictcp);
 }