/*
- * TCP CUBIC: Binary Increase Congestion control for TCP v2.0
- *
+ * TCP CUBIC: Binary Increase Congestion control for TCP v2.3
+ * Home page:
+ * http://netsrv.csc.ncsu.edu/twiki/bin/view/Main/BIC
* This is from the implementation of CUBIC TCP in
- * Injong Rhee, Lisong Xu.
- * "CUBIC: A New TCP-Friendly High-Speed TCP Variant
- * in PFLDnet 2005
+ * Sangtae Ha, Injong Rhee and Lisong Xu,
+ * "CUBIC: A New TCP-Friendly High-Speed TCP Variant"
+ * in ACM SIGOPS Operating System Review, July 2008.
+ * Available from:
+ * http://netsrv.csc.ncsu.edu/export/cubic_a_new_tcp_2008.pdf
+ *
+ * CUBIC integrates a new slow start algorithm, called HyStart.
+ * The details of HyStart are presented in
+ * Sangtae Ha and Injong Rhee,
+ * "Taming the Elephants: New TCP Slow Start", NCSU TechReport 2008.
* Available from:
- * http://www.csc.ncsu.edu/faculty/rhee/export/bitcp/cubic-paper.pdf
+ * http://netsrv.csc.ncsu.edu/export/hystart_techreport_2008.pdf
+ *
+ * All testing results are available from:
+ * http://netsrv.csc.ncsu.edu/wiki/index.php/TCP_Testing
*
* Unless CUBIC 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 <linux/math64.h>
#include <net/tcp.h>
-
#define BICTCP_BETA_SCALE 1024 /* Scale factor beta calculation
* max_cwnd = snd_cwnd * beta
*/
-#define BICTCP_B 4 /*
- * In binary search,
- * go to point (max+min)/N
- */
#define BICTCP_HZ 10 /* BIC HZ 2^10 = 1024 */
-static int fast_convergence = 1;
-static int max_increment = 16;
-static int beta = 819; /* = 819/1024 (BICTCP_BETA_SCALE) */
-static int initial_ssthresh = 100;
-static int bic_scale = 41;
-static int tcp_friendliness = 1;
+/* Two methods of hybrid slow start */
+#define HYSTART_ACK_TRAIN 0x1
+#define HYSTART_DELAY 0x2
+
+/* Number of delay samples for detecting the increase of delay */
+#define HYSTART_MIN_SAMPLES 8
+#define HYSTART_DELAY_MIN (2U<<3)
+#define HYSTART_DELAY_MAX (16U<<3)
+#define HYSTART_DELAY_THRESH(x) clamp(x, HYSTART_DELAY_MIN, HYSTART_DELAY_MAX)
+static int fast_convergence __read_mostly = 1;
+static int beta __read_mostly = 717; /* = 717/1024 (BICTCP_BETA_SCALE) */
+static int initial_ssthresh __read_mostly;
+static int bic_scale __read_mostly = 41;
+static int tcp_friendliness __read_mostly = 1;
+
+static int hystart __read_mostly = 1;
+static int hystart_detect __read_mostly = HYSTART_ACK_TRAIN | HYSTART_DELAY;
+static int hystart_low_window __read_mostly = 16;
+
+static u32 cube_rtt_scale __read_mostly;
+static u32 beta_scale __read_mostly;
+static u64 cube_factor __read_mostly;
+
+/* Note parameters that are used for precomputing scale factors are read-only */
module_param(fast_convergence, int, 0644);
MODULE_PARM_DESC(fast_convergence, "turn on/off fast convergence");
-module_param(max_increment, int, 0644);
-MODULE_PARM_DESC(max_increment, "Limit on increment allowed during binary search");
module_param(beta, int, 0644);
MODULE_PARM_DESC(beta, "beta for multiplicative increase");
module_param(initial_ssthresh, int, 0644);
MODULE_PARM_DESC(initial_ssthresh, "initial value of slow start threshold");
-module_param(bic_scale, int, 0644);
+module_param(bic_scale, int, 0444);
MODULE_PARM_DESC(bic_scale, "scale (scaled by 1024) value for bic function (bic_scale/1024)");
module_param(tcp_friendliness, int, 0644);
MODULE_PARM_DESC(tcp_friendliness, "turn on/off tcp friendliness");
-
+module_param(hystart, int, 0644);
+MODULE_PARM_DESC(hystart, "turn on/off hybrid slow start algorithm");
+module_param(hystart_detect, int, 0644);
+MODULE_PARM_DESC(hystart_detect, "hyrbrid slow start detection mechanisms"
+ " 1: packet-train 2: delay 3: both packet-train and delay");
+module_param(hystart_low_window, int, 0644);
+MODULE_PARM_DESC(hystart_low_window, "lower bound cwnd for hybrid slow start");
/* BIC TCP Parameters */
struct bictcp {
u32 ack_cnt; /* number of acks */
u32 tcp_cwnd; /* estimated tcp cwnd */
#define ACK_RATIO_SHIFT 4
- u32 delayed_ack; /* estimate the ratio of Packets/ACKs << 4 */
+ u16 delayed_ack; /* estimate the ratio of Packets/ACKs << 4 */
+ u8 sample_cnt; /* number of samples to decide curr_rtt */
+ u8 found; /* the exit point is found? */
+ u32 round_start; /* beginning of each round */
+ u32 end_seq; /* end_seq of the round */
+ u32 last_jiffies; /* last time when the ACK spacing is close */
+ u32 curr_rtt; /* the minimum rtt of current round */
};
static inline void bictcp_reset(struct bictcp *ca)
ca->delayed_ack = 2 << ACK_RATIO_SHIFT;
ca->ack_cnt = 0;
ca->tcp_cwnd = 0;
+ ca->found = 0;
+}
+
+static inline void bictcp_hystart_reset(struct sock *sk)
+{
+ struct tcp_sock *tp = tcp_sk(sk);
+ struct bictcp *ca = inet_csk_ca(sk);
+
+ ca->round_start = ca->last_jiffies = jiffies;
+ ca->end_seq = tp->snd_nxt;
+ ca->curr_rtt = 0;
+ ca->sample_cnt = 0;
}
static void bictcp_init(struct sock *sk)
{
bictcp_reset(inet_csk_ca(sk));
- if (initial_ssthresh)
+
+ if (hystart)
+ bictcp_hystart_reset(sk);
+
+ if (!hystart && initial_ssthresh)
tcp_sk(sk)->snd_ssthresh = initial_ssthresh;
}
-/* 65536 times the cubic root */
-static const u64 cubic_table[8]
- = {0, 65536, 82570, 94519, 104030, 112063, 119087, 125367};
-
-/*
- * calculate the cubic root of x
- * the basic idea is that x can be expressed as i*8^j
- * so cubic_root(x) = cubic_root(i)*2^j
- * in the following code, x is i, and y is 2^j
- * because of integer calculation, there are errors in calculation
- * so finally use binary search to find out the exact solution
+/* calculate the cubic root of x using a table lookup followed by one
+ * Newton-Raphson iteration.
+ * Avg err ~= 0.195%
*/
-static u32 cubic_root(u64 x)
+static u32 cubic_root(u64 a)
{
- u64 y, app, target, start, end, mid, start_diff, end_diff;
-
- if (x == 0)
- return 0;
-
- target = x;
-
- /* first estimate lower and upper bound */
- y = 1;
- while (x >= 8){
- x = (x >> 3);
- y = (y << 1);
- }
- start = (y*cubic_table[x])>>16;
- if (x==7)
- end = (y<<1);
- else
- end = (y*cubic_table[x+1]+65535)>>16;
-
- /* binary search for more accurate one */
- while (start < end-1) {
- mid = (start+end) >> 1;
- app = mid*mid*mid;
- if (app < target)
- start = mid;
- else if (app > target)
- end = mid;
- else
- return mid;
- }
-
- /* find the most accurate one from start and end */
- app = start*start*start;
- if (app < target)
- start_diff = target - app;
- else
- start_diff = app - target;
- app = end*end*end;
- if (app < target)
- end_diff = target - app;
- else
- end_diff = app - target;
-
- if (start_diff < end_diff)
- return (u32)start;
- else
- return (u32)end;
-}
+ u32 x, b, shift;
+ /*
+ * cbrt(x) MSB values for x MSB values in [0..63].
+ * Precomputed then refined by hand - Willy Tarreau
+ *
+ * For x in [0..63],
+ * v = cbrt(x << 18) - 1
+ * cbrt(x) = (v[x] + 10) >> 6
+ */
+ static const u8 v[] = {
+ /* 0x00 */ 0, 54, 54, 54, 118, 118, 118, 118,
+ /* 0x08 */ 123, 129, 134, 138, 143, 147, 151, 156,
+ /* 0x10 */ 157, 161, 164, 168, 170, 173, 176, 179,
+ /* 0x18 */ 181, 185, 187, 190, 192, 194, 197, 199,
+ /* 0x20 */ 200, 202, 204, 206, 209, 211, 213, 215,
+ /* 0x28 */ 217, 219, 221, 222, 224, 225, 227, 229,
+ /* 0x30 */ 231, 232, 234, 236, 237, 239, 240, 242,
+ /* 0x38 */ 244, 245, 246, 248, 250, 251, 252, 254,
+ };
+
+ b = fls64(a);
+ if (b < 7) {
+ /* a in [0..63] */
+ return ((u32)v[(u32)a] + 35) >> 6;
+ }
-static inline u32 bictcp_K(u32 dist, u32 srtt)
-{
- u64 d64;
- u32 d32;
- u32 count;
- u32 result;
-
- /* calculate the "K" for (wmax-cwnd) = c/rtt * K^3
- so K = cubic_root( (wmax-cwnd)*rtt/c )
- the unit of K is bictcp_HZ=2^10, not HZ
-
- c = bic_scale >> 10
- rtt = (tp->srtt >> 3 ) / HZ
-
- the following code has been designed and tested for
- cwnd < 1 million packets
- RTT < 100 seconds
- HZ < 1,000,00 (corresponding to 10 nano-second)
-
- */
-
- /* 1/c * 2^2*bictcp_HZ */
- d32 = (1 << (10+2*BICTCP_HZ)) / bic_scale;
- d64 = (__u64)d32;
-
- /* srtt * 2^count / HZ
- 1) to get a better accuracy of the following d32,
- the larger the "count", the better the accuracy
- 2) and avoid overflow of the following d64
- the larger the "count", the high possibility of overflow
- 3) so find a "count" between bictcp_hz-3 and bictcp_hz
- "count" may be less than bictcp_HZ,
- then d64 becomes 0. that is OK
- */
- d32 = srtt;
- count = 0;
- while (((d32 & 0x80000000)==0) && (count < BICTCP_HZ)){
- d32 = d32 << 1;
- count++;
- }
- d32 = d32 / HZ;
-
- /* (wmax-cwnd) * (srtt>>3 / HZ) / c * 2^(3*bictcp_HZ) */
- d64 = (d64 * dist * d32) >> (count+3-BICTCP_HZ);
-
- /* cubic root */
- d64 = cubic_root(d64);
-
- result = (u32)d64;
- return result;
+ b = ((b * 84) >> 8) - 1;
+ shift = (a >> (b * 3));
+
+ x = ((u32)(((u32)v[shift] + 10) << b)) >> 6;
+
+ /*
+ * Newton-Raphson iteration
+ * 2
+ * x = ( 2 * x + a / x ) / 3
+ * k+1 k k
+ */
+ x = (2 * x + (u32)div64_u64(a, (u64)x * (u64)(x - 1)));
+ x = ((x * 341) >> 10);
+ return x;
}
/*
*/
static inline void bictcp_update(struct bictcp *ca, u32 cwnd)
{
- u64 d64;
- u32 d32, t, srtt, bic_target, min_cnt, max_cnt;
+ u64 offs;
+ u32 delta, t, bic_target, max_cnt;
ca->ack_cnt++; /* count the number of ACKs */
ca->last_cwnd = cwnd;
ca->last_time = tcp_time_stamp;
- srtt = (HZ << 3)/10; /* use real time-based growth function */
-
if (ca->epoch_start == 0) {
ca->epoch_start = tcp_time_stamp; /* record the beginning of an epoch */
ca->ack_cnt = 1; /* start counting */
ca->bic_K = 0;
ca->bic_origin_point = cwnd;
} else {
- ca->bic_K = bictcp_K(ca->last_max_cwnd-cwnd, srtt);
+ /* Compute new K based on
+ * (wmax-cwnd) * (srtt>>3 / HZ) / c * 2^(3*bictcp_HZ)
+ */
+ ca->bic_K = cubic_root(cube_factor
+ * (ca->last_max_cwnd - cwnd));
ca->bic_origin_point = ca->last_max_cwnd;
}
}
- /* cubic function - calc*/
- /* calculate c * time^3 / rtt,
- * while considering overflow in calculation of time^3
- * (so time^3 is done by using d64)
+ /* cubic function - calc*/
+ /* calculate c * time^3 / rtt,
+ * while considering overflow in calculation of time^3
+ * (so time^3 is done by using 64 bit)
* and without the support of division of 64bit numbers
- * (so all divisions are done by using d32)
- * also NOTE the unit of those veriables
- * time = (t - K) / 2^bictcp_HZ
- * c = bic_scale >> 10
+ * (so all divisions are done by using 32 bit)
+ * also NOTE the unit of those veriables
+ * time = (t - K) / 2^bictcp_HZ
+ * c = bic_scale >> 10
* rtt = (srtt >> 3) / HZ
* !!! The following code does not have overflow problems,
* if the cwnd < 1 million packets !!!
- */
+ */
/* change the unit from HZ to bictcp_HZ */
- t = ((tcp_time_stamp + ca->delay_min - ca->epoch_start)
+ t = ((tcp_time_stamp + (ca->delay_min>>3) - ca->epoch_start)
<< BICTCP_HZ) / HZ;
- if (t < ca->bic_K) /* t - K */
- d32 = ca->bic_K - t;
- else
- d32 = t - ca->bic_K;
-
- d64 = (u64)d32;
- d32 = (bic_scale << 3) * HZ / srtt; /* 1024*c/rtt */
- d64 = (d32 * d64 * d64 * d64) >> (10+3*BICTCP_HZ); /* c/rtt * (t-K)^3 */
- d32 = (u32)d64;
- if (t < ca->bic_K) /* below origin*/
- bic_target = ca->bic_origin_point - d32;
- else /* above origin*/
- bic_target = ca->bic_origin_point + d32;
-
- /* cubic function - calc bictcp_cnt*/
- if (bic_target > cwnd) {
+ if (t < ca->bic_K) /* t - K */
+ offs = ca->bic_K - t;
+ else
+ offs = t - ca->bic_K;
+
+ /* c/rtt * (t-K)^3 */
+ delta = (cube_rtt_scale * offs * offs * offs) >> (10+3*BICTCP_HZ);
+ if (t < ca->bic_K) /* below origin*/
+ bic_target = ca->bic_origin_point - delta;
+ else /* above origin*/
+ bic_target = ca->bic_origin_point + delta;
+
+ /* cubic function - calc bictcp_cnt*/
+ if (bic_target > cwnd) {
ca->cnt = cwnd / (bic_target - cwnd);
- } else {
- ca->cnt = 100 * cwnd; /* very small increment*/
- }
-
- if (ca->delay_min > 0) {
- /* max increment = Smax * rtt / 0.1 */
- min_cnt = (cwnd * HZ * 8)/(10 * max_increment * ca->delay_min);
- if (ca->cnt < min_cnt)
- ca->cnt = min_cnt;
+ } else {
+ ca->cnt = 100 * cwnd; /* very small increment*/
}
- /* slow start and low utilization */
- if (ca->loss_cwnd == 0) /* could be aggressive in slow start */
- ca->cnt = 50;
-
/* TCP Friendly */
if (tcp_friendliness) {
- u32 scale = 8*(BICTCP_BETA_SCALE+beta)/3/(BICTCP_BETA_SCALE-beta);
- d32 = (cwnd * scale) >> 3;
- while (ca->ack_cnt > d32) { /* update tcp cwnd */
- ca->ack_cnt -= d32;
- ca->tcp_cwnd++;
+ u32 scale = beta_scale;
+ delta = (cwnd * scale) >> 3;
+ while (ca->ack_cnt > delta) { /* update tcp cwnd */
+ ca->ack_cnt -= delta;
+ ca->tcp_cwnd++;
}
if (ca->tcp_cwnd > cwnd){ /* if bic is slower than tcp */
- d32 = ca->tcp_cwnd - cwnd;
- max_cnt = cwnd / d32;
+ delta = ca->tcp_cwnd - cwnd;
+ max_cnt = cwnd / delta;
if (ca->cnt > max_cnt)
ca->cnt = max_cnt;
}
- }
+ }
ca->cnt = (ca->cnt << ACK_RATIO_SHIFT) / ca->delayed_ack;
if (ca->cnt == 0) /* cannot be zero */
ca->cnt = 1;
}
-
-/* Keep track of minimum rtt */
-static inline void measure_delay(struct sock *sk)
-{
- const struct tcp_sock *tp = tcp_sk(sk);
- struct bictcp *ca = inet_csk_ca(sk);
- u32 delay;
-
- /* No time stamp */
- if (!(tp->rx_opt.saw_tstamp && tp->rx_opt.rcv_tsecr) ||
- /* Discard delay samples right after fast recovery */
- (s32)(tcp_time_stamp - ca->epoch_start) < HZ)
- return;
-
- delay = tcp_time_stamp - tp->rx_opt.rcv_tsecr;
- if (delay == 0)
- delay = 1;
-
- /* first time call or link delay decreases */
- if (ca->delay_min == 0 || ca->delay_min > delay)
- ca->delay_min = delay;
-}
-
-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 in_flight)
{
struct tcp_sock *tp = tcp_sk(sk);
struct bictcp *ca = inet_csk_ca(sk);
- if (data_acked)
- measure_delay(sk);
-
if (!tcp_is_cwnd_limited(sk, in_flight))
return;
- if (tp->snd_cwnd <= tp->snd_ssthresh)
+ if (tp->snd_cwnd <= tp->snd_ssthresh) {
+ if (hystart && after(ack, ca->end_seq))
+ bictcp_hystart_reset(sk);
tcp_slow_start(tp);
- else {
+ } 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);
}
}
return max(tcp_sk(sk)->snd_cwnd, ca->last_max_cwnd);
}
-static u32 bictcp_min_cwnd(struct sock *sk)
+static void bictcp_state(struct sock *sk, u8 new_state)
{
- return tcp_sk(sk)->snd_ssthresh;
+ if (new_state == TCP_CA_Loss) {
+ bictcp_reset(inet_csk_ca(sk));
+ bictcp_hystart_reset(sk);
+ }
}
-static void bictcp_state(struct sock *sk, u8 new_state)
+static void hystart_update(struct sock *sk, u32 delay)
{
- if (new_state == TCP_CA_Loss)
- bictcp_reset(inet_csk_ca(sk));
+ struct tcp_sock *tp = tcp_sk(sk);
+ struct bictcp *ca = inet_csk_ca(sk);
+
+ if (!(ca->found & hystart_detect)) {
+ u32 curr_jiffies = jiffies;
+
+ /* first detection parameter - ack-train detection */
+ if (curr_jiffies - ca->last_jiffies <= msecs_to_jiffies(2)) {
+ ca->last_jiffies = curr_jiffies;
+ if (curr_jiffies - ca->round_start >= ca->delay_min>>4)
+ ca->found |= HYSTART_ACK_TRAIN;
+ }
+
+ /* obtain the minimum delay of more than sampling packets */
+ if (ca->sample_cnt < HYSTART_MIN_SAMPLES) {
+ if (ca->curr_rtt == 0 || ca->curr_rtt > delay)
+ ca->curr_rtt = delay;
+
+ ca->sample_cnt++;
+ } else {
+ if (ca->curr_rtt > ca->delay_min +
+ HYSTART_DELAY_THRESH(ca->delay_min>>4))
+ ca->found |= HYSTART_DELAY;
+ }
+ /*
+ * Either one of two conditions are met,
+ * we exit from slow start immediately.
+ */
+ if (ca->found & hystart_detect)
+ tp->snd_ssthresh = tp->snd_cwnd;
+ }
}
/* 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_us)
{
const struct inet_connection_sock *icsk = inet_csk(sk);
+ const struct tcp_sock *tp = tcp_sk(sk);
+ struct bictcp *ca = inet_csk_ca(sk);
+ u32 delay;
- if (cnt > 0 && icsk->icsk_ca_state == TCP_CA_Open) {
- struct bictcp *ca = inet_csk_ca(sk);
+ if (icsk->icsk_ca_state == TCP_CA_Open) {
cnt -= ca->delayed_ack >> ACK_RATIO_SHIFT;
ca->delayed_ack += cnt;
}
-}
+ /* Some calls are for duplicates without timetamps */
+ if (rtt_us < 0)
+ return;
+
+ /* Discard delay samples right after fast recovery */
+ if ((s32)(tcp_time_stamp - ca->epoch_start) < HZ)
+ return;
+
+ delay = usecs_to_jiffies(rtt_us) << 3;
+ if (delay == 0)
+ delay = 1;
+
+ /* first time call or link delay decreases */
+ if (ca->delay_min == 0 || ca->delay_min > delay)
+ ca->delay_min = delay;
+
+ /* hystart triggers when cwnd is larger than some threshold */
+ if (hystart && tp->snd_cwnd <= tp->snd_ssthresh &&
+ tp->snd_cwnd >= hystart_low_window)
+ hystart_update(sk, delay);
+}
static struct tcp_congestion_ops cubictcp = {
.init = bictcp_init,
.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 = "cubic",
static int __init cubictcp_register(void)
{
- BUG_ON(sizeof(struct bictcp) > ICSK_CA_PRIV_SIZE);
+ BUILD_BUG_ON(sizeof(struct bictcp) > ICSK_CA_PRIV_SIZE);
+
+ /* Precompute a bunch of the scaling factors that are used per-packet
+ * based on SRTT of 100ms
+ */
+
+ beta_scale = 8*(BICTCP_BETA_SCALE+beta)/ 3 / (BICTCP_BETA_SCALE - beta);
+
+ cube_rtt_scale = (bic_scale * 10); /* 1024*c/rtt */
+
+ /* calculate the "K" for (wmax-cwnd) = c/rtt * K^3
+ * so K = cubic_root( (wmax-cwnd)*rtt/c )
+ * the unit of K is bictcp_HZ=2^10, not HZ
+ *
+ * c = bic_scale >> 10
+ * rtt = 100ms
+ *
+ * the following code has been designed and tested for
+ * cwnd < 1 million packets
+ * RTT < 100 seconds
+ * HZ < 1,000,00 (corresponding to 10 nano-second)
+ */
+
+ /* 1/c * 2^2*bictcp_HZ * srtt */
+ cube_factor = 1ull << (10+3*BICTCP_HZ); /* 2^40 */
+
+ /* divide by bic_scale and by constant Srtt (100ms) */
+ do_div(cube_factor, bic_scale * 10);
+
return tcp_register_congestion_control(&cubictcp);
}
MODULE_AUTHOR("Sangtae Ha, Stephen Hemminger");
MODULE_LICENSE("GPL");
MODULE_DESCRIPTION("CUBIC TCP");
-MODULE_VERSION("2.0");
+MODULE_VERSION("2.3");
git://ftp.safe.ca / safe / jmp / linux-2.6 / blobdiff