tcp: switch TCP TS option (RFC 7323) to 1ms clock

TCP Timestamps option is defined in RFC 7323

Traditionally on linux, it has been tied to the internal
'jiffies' variable, because it had been a cheap and good enough
generator.

For TCP flows on the Internet, 1 ms resolution would be much better
than 4ms or 10ms (HZ=250 or HZ=100 respectively)

For TCP flows in the DC, Google has used usec resolution for more
than two years with great success [1]

Receive size autotuning (DRS) is indeed more precise and converges
faster to optimal window size.

This patch converts tp->tcp_mstamp to a plain u64 value storing
a 1 usec TCP clock.

This choice will allow us to upstream the 1 usec TS option as
discussed in IETF 97.

[1] https://www.ietf.org/proceedings/97/slides/slides-97-tcpm-tcp-options-for-low-latency-00.pdf

Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

11 files changed, 112 insertions, 114 deletions

diff --git a/net/ipv4/syncookies.c b/net/ipv4/syncookies.c
index 0257d965f111..6426250a58ea 100644
--- a/net/ipv4/syncookies.c
+++ b/net/ipv4/syncookies.c
@@ -66,10 +66,10 @@ static u32 cookie_hash(__be32 saddr, __be32 daddr, __be16 sport, __be16 dport,
  * Since subsequent timestamps use the normal tcp_time_stamp value, we
  * must make sure that the resulting initial timestamp is <= tcp_time_stamp.
  */
-__u32 cookie_init_timestamp(struct request_sock *req)
+u64 cookie_init_timestamp(struct request_sock *req)
 {
 	struct inet_request_sock *ireq;
-	u32 ts, ts_now = tcp_time_stamp;
+	u32 ts, ts_now = tcp_time_stamp_raw();
 	u32 options = 0;
 
 	ireq = inet_rsk(req);
@@ -88,7 +88,7 @@ __u32 cookie_init_timestamp(struct request_sock *req)
 		ts <<= TSBITS;
 		ts |= options;
 	}
-	return ts;
+	return (u64)ts * (USEC_PER_SEC / TCP_TS_HZ);
 }
 
 
@@ -343,7 +343,7 @@ struct sock *cookie_v4_check(struct sock *sk, struct sk_buff *skb)
 	ireq->wscale_ok		= tcp_opt.wscale_ok;
 	ireq->tstamp_ok		= tcp_opt.saw_tstamp;
 	req->ts_recent		= tcp_opt.saw_tstamp ? tcp_opt.rcv_tsval : 0;
-	treq->snt_synack.v64	= 0;
+	treq->snt_synack	= 0;
 	treq->tfo_listener	= false;
 
 	ireq->ir_iif = inet_request_bound_dev_if(sk, skb);
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 850054800526..b5d18484746d 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -2706,7 +2706,7 @@ static int do_tcp_setsockopt(struct sock *sk, int level,
 		if (!tp->repair)
 			err = -EPERM;
 		else
-			tp->tsoffset = val - tcp_time_stamp;
+			tp->tsoffset = val - tcp_time_stamp_raw();
 		break;
 	case TCP_REPAIR_WINDOW:
 		err = tcp_repair_set_window(tp, optval, optlen);
@@ -3072,7 +3072,7 @@ static int do_tcp_getsockopt(struct sock *sk, int level,
 		break;
 
 	case TCP_TIMESTAMP:
-		val = tcp_time_stamp + tp->tsoffset;
+		val = tcp_time_stamp_raw() + tp->tsoffset;
 		break;
 	case TCP_NOTSENT_LOWAT:
 		val = tp->notsent_lowat;
diff --git a/net/ipv4/tcp_bbr.c b/net/ipv4/tcp_bbr.c
index 40dc4fc5f6ac..dbcc9352a48f 100644
--- a/net/ipv4/tcp_bbr.c
+++ b/net/ipv4/tcp_bbr.c
@@ -91,7 +91,7 @@ struct bbr {
 	struct minmax bw;	/* Max recent delivery rate in pkts/uS << 24 */
 	u32	rtt_cnt;	    /* count of packet-timed rounds elapsed */
 	u32     next_rtt_delivered; /* scb->tx.delivered at end of round */
-	struct skb_mstamp cycle_mstamp;  /* time of this cycle phase start */
+	u64	cycle_mstamp;	     /* time of this cycle phase start */
 	u32     mode:3,		     /* current bbr_mode in state machine */
 		prev_ca_state:3,     /* CA state on previous ACK */
 		packet_conservation:1,  /* use packet conservation? */
@@ -411,7 +411,7 @@ static bool bbr_is_next_cycle_phase(struct sock *sk,
 	struct tcp_sock *tp = tcp_sk(sk);
 	struct bbr *bbr = inet_csk_ca(sk);
 	bool is_full_length =
-		skb_mstamp_us_delta(&tp->delivered_mstamp, &bbr->cycle_mstamp) >
+		tcp_stamp_us_delta(tp->delivered_mstamp, bbr->cycle_mstamp) >
 		bbr->min_rtt_us;
 	u32 inflight, bw;
 
@@ -497,7 +497,7 @@ static void bbr_reset_lt_bw_sampling_interval(struct sock *sk)
 	struct tcp_sock *tp = tcp_sk(sk);
 	struct bbr *bbr = inet_csk_ca(sk);
 
-	bbr->lt_last_stamp = tp->delivered_mstamp.stamp_jiffies;
+	bbr->lt_last_stamp = div_u64(tp->delivered_mstamp, USEC_PER_MSEC);
 	bbr->lt_last_delivered = tp->delivered;
 	bbr->lt_last_lost = tp->lost;
 	bbr->lt_rtt_cnt = 0;
@@ -551,7 +551,7 @@ static void bbr_lt_bw_sampling(struct sock *sk, const struct rate_sample *rs)
 	struct bbr *bbr = inet_csk_ca(sk);
 	u32 lost, delivered;
 	u64 bw;
-	s32 t;
+	u32 t;
 
 	if (bbr->lt_use_bw) {	/* already using long-term rate, lt_bw? */
 		if (bbr->mode == BBR_PROBE_BW && bbr->round_start &&
@@ -603,15 +603,15 @@ static void bbr_lt_bw_sampling(struct sock *sk, const struct rate_sample *rs)
 		return;
 
 	/* Find average delivery rate in this sampling interval. */
-	t = (s32)(tp->delivered_mstamp.stamp_jiffies - bbr->lt_last_stamp);
-	if (t < 1)
-		return;		/* interval is less than one jiffy, so wait */
-	t = jiffies_to_usecs(t);
-	/* Interval long enough for jiffies_to_usecs() to return a bogus 0? */
-	if (t < 1) {
+	t = div_u64(tp->delivered_mstamp, USEC_PER_MSEC) - bbr->lt_last_stamp;
+	if ((s32)t < 1)
+		return;		/* interval is less than one ms, so wait */
+	/* Check if can multiply without overflow */
+	if (t >= ~0U / USEC_PER_MSEC) {
 		bbr_reset_lt_bw_sampling(sk);  /* interval too long; reset */
 		return;
 	}
+	t *= USEC_PER_MSEC;
 	bw = (u64)delivered * BW_UNIT;
 	do_div(bw, t);
 	bbr_lt_bw_interval_done(sk, bw);
@@ -825,7 +825,7 @@ static void bbr_init(struct sock *sk)
 	bbr->idle_restart = 0;
 	bbr->full_bw = 0;
 	bbr->full_bw_cnt = 0;
-	bbr->cycle_mstamp.v64 = 0;
+	bbr->cycle_mstamp = 0;
 	bbr->cycle_idx = 0;
 	bbr_reset_lt_bw_sampling(sk);
 	bbr_reset_startup_mode(sk);
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index 10e6775464f6..9a5a9e8eda89 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -441,7 +441,7 @@ void tcp_init_buffer_space(struct sock *sk)
 		tcp_sndbuf_expand(sk);
 
 	tp->rcvq_space.space = tp->rcv_wnd;
-	skb_mstamp_get(&tp->tcp_mstamp);
+	tcp_mstamp_refresh(tp);
 	tp->rcvq_space.time = tp->tcp_mstamp;
 	tp->rcvq_space.seq = tp->copied_seq;
 
@@ -555,11 +555,11 @@ static inline void tcp_rcv_rtt_measure(struct tcp_sock *tp)
 {
 	u32 delta_us;
 
-	if (tp->rcv_rtt_est.time.v64 == 0)
+	if (tp->rcv_rtt_est.time == 0)
 		goto new_measure;
 	if (before(tp->rcv_nxt, tp->rcv_rtt_est.seq))
 		return;
-	delta_us = skb_mstamp_us_delta(&tp->tcp_mstamp, &tp->rcv_rtt_est.time);
+	delta_us = tcp_stamp_us_delta(tp->tcp_mstamp, tp->rcv_rtt_est.time);
 	tcp_rcv_rtt_update(tp, delta_us, 1);
 
 new_measure:
@@ -571,13 +571,15 @@ static inline void tcp_rcv_rtt_measure_ts(struct sock *sk,
 					  const struct sk_buff *skb)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
+
 	if (tp->rx_opt.rcv_tsecr &&
 	    (TCP_SKB_CB(skb)->end_seq -
-	     TCP_SKB_CB(skb)->seq >= inet_csk(sk)->icsk_ack.rcv_mss))
-		tcp_rcv_rtt_update(tp,
-				   jiffies_to_usecs(tcp_time_stamp -
-						    tp->rx_opt.rcv_tsecr),
-				   0);
+	     TCP_SKB_CB(skb)->seq >= inet_csk(sk)->icsk_ack.rcv_mss)) {
+		u32 delta = tcp_time_stamp(tp) - tp->rx_opt.rcv_tsecr;
+		u32 delta_us = delta * (USEC_PER_SEC / TCP_TS_HZ);
+
+		tcp_rcv_rtt_update(tp, delta_us, 0);
+	}
 }
 
 /*
@@ -590,7 +592,7 @@ void tcp_rcv_space_adjust(struct sock *sk)
 	int time;
 	int copied;
 
-	time = skb_mstamp_us_delta(&tp->tcp_mstamp, &tp->rcvq_space.time);
+	time = tcp_stamp_us_delta(tp->tcp_mstamp, tp->rcvq_space.time);
 	if (time < (tp->rcv_rtt_est.rtt_us >> 3) || tp->rcv_rtt_est.rtt_us == 0)
 		return;
 
@@ -1134,8 +1136,8 @@ struct tcp_sacktag_state {
 	 * that was SACKed. RTO needs the earliest RTT to stay conservative,
 	 * but congestion control should still get an accurate delay signal.
 	 */
-	struct skb_mstamp first_sackt;
-	struct skb_mstamp last_sackt;
+	u64	first_sackt;
+	u64	last_sackt;
 	struct rate_sample *rate;
 	int	flag;
 };
@@ -1200,7 +1202,7 @@ static u8 tcp_sacktag_one(struct sock *sk,
 			  struct tcp_sacktag_state *state, u8 sacked,
 			  u32 start_seq, u32 end_seq,
 			  int dup_sack, int pcount,
-			  const struct skb_mstamp *xmit_time)
+			  u64 xmit_time)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 	int fack_count = state->fack_count;
@@ -1242,9 +1244,9 @@ static u8 tcp_sacktag_one(struct sock *sk,
 							   state->reord);
 				if (!after(end_seq, tp->high_seq))
 					state->flag |= FLAG_ORIG_SACK_ACKED;
-				if (state->first_sackt.v64 == 0)
-					state->first_sackt = *xmit_time;
-				state->last_sackt = *xmit_time;
+				if (state->first_sackt == 0)
+					state->first_sackt = xmit_time;
+				state->last_sackt = xmit_time;
 			}
 
 			if (sacked & TCPCB_LOST) {
@@ -1304,7 +1306,7 @@ static bool tcp_shifted_skb(struct sock *sk, struct sk_buff *skb,
 	 */
 	tcp_sacktag_one(sk, state, TCP_SKB_CB(skb)->sacked,
 			start_seq, end_seq, dup_sack, pcount,
-			&skb->skb_mstamp);
+			skb->skb_mstamp);
 	tcp_rate_skb_delivered(sk, skb, state->rate);
 
 	if (skb == tp->lost_skb_hint)
@@ -1356,8 +1358,8 @@ static bool tcp_shifted_skb(struct sock *sk, struct sk_buff *skb,
 		tcp_advance_highest_sack(sk, skb);
 
 	tcp_skb_collapse_tstamp(prev, skb);
-	if (unlikely(TCP_SKB_CB(prev)->tx.delivered_mstamp.v64))
-		TCP_SKB_CB(prev)->tx.delivered_mstamp.v64 = 0;
+	if (unlikely(TCP_SKB_CB(prev)->tx.delivered_mstamp))
+		TCP_SKB_CB(prev)->tx.delivered_mstamp = 0;
 
 	tcp_unlink_write_queue(skb, sk);
 	sk_wmem_free_skb(sk, skb);
@@ -1587,7 +1589,7 @@ static struct sk_buff *tcp_sacktag_walk(struct sk_buff *skb, struct sock *sk,
 						TCP_SKB_CB(skb)->end_seq,
 						dup_sack,
 						tcp_skb_pcount(skb),
-						&skb->skb_mstamp);
+						skb->skb_mstamp);
 			tcp_rate_skb_delivered(sk, skb, state->rate);
 
 			if (!before(TCP_SKB_CB(skb)->seq,
@@ -2936,9 +2938,12 @@ static inline bool tcp_ack_update_rtt(struct sock *sk, const int flag,
 	 * See draft-ietf-tcplw-high-performance-00, section 3.3.
 	 */
 	if (seq_rtt_us < 0 && tp->rx_opt.saw_tstamp && tp->rx_opt.rcv_tsecr &&
-	    flag & FLAG_ACKED)
-		seq_rtt_us = ca_rtt_us = jiffies_to_usecs(tcp_time_stamp -
-							  tp->rx_opt.rcv_tsecr);
+	    flag & FLAG_ACKED) {
+		u32 delta = tcp_time_stamp(tp) - tp->rx_opt.rcv_tsecr;
+		u32 delta_us = delta * (USEC_PER_SEC / TCP_TS_HZ);
+
+		seq_rtt_us = ca_rtt_us = delta_us;
+	}
 	if (seq_rtt_us < 0)
 		return false;
 
@@ -2960,12 +2965,8 @@ void tcp_synack_rtt_meas(struct sock *sk, struct request_sock *req)
 {
 	long rtt_us = -1L;
 
-	if (req && !req->num_retrans && tcp_rsk(req)->snt_synack.v64) {
-		struct skb_mstamp now;
-
-		skb_mstamp_get(&now);
-		rtt_us = skb_mstamp_us_delta(&now, &tcp_rsk(req)->snt_synack);
-	}
+	if (req && !req->num_retrans && tcp_rsk(req)->snt_synack)
+		rtt_us = tcp_stamp_us_delta(tcp_clock_us(), tcp_rsk(req)->snt_synack);
 
 	tcp_ack_update_rtt(sk, FLAG_SYN_ACKED, rtt_us, -1L, rtt_us);
 }
@@ -3003,7 +3004,7 @@ void tcp_rearm_rto(struct sock *sk)
 			struct sk_buff *skb = tcp_write_queue_head(sk);
 			const u32 rto_time_stamp =
 				tcp_skb_timestamp(skb) + rto;
-			s32 delta = (s32)(rto_time_stamp - tcp_time_stamp);
+			s32 delta = (s32)(rto_time_stamp - tcp_jiffies32);
 			/* delta may not be positive if the socket is locked
 			 * when the retrans timer fires and is rescheduled.
 			 */
@@ -3060,9 +3061,8 @@ static int tcp_clean_rtx_queue(struct sock *sk, int prior_fackets,
 			       struct tcp_sacktag_state *sack)
 {
 	const struct inet_connection_sock *icsk = inet_csk(sk);
-	struct skb_mstamp first_ackt, last_ackt;
+	u64 first_ackt, last_ackt;
 	struct tcp_sock *tp = tcp_sk(sk);
-	struct skb_mstamp *now = &tp->tcp_mstamp;
 	u32 prior_sacked = tp->sacked_out;
 	u32 reord = tp->packets_out;
 	bool fully_acked = true;
@@ -3075,7 +3075,7 @@ static int tcp_clean_rtx_queue(struct sock *sk, int prior_fackets,
 	bool rtt_update;
 	int flag = 0;
 
-	first_ackt.v64 = 0;
+	first_ackt = 0;
 
 	while ((skb = tcp_write_queue_head(sk)) && skb != tcp_send_head(sk)) {
 		struct tcp_skb_cb *scb = TCP_SKB_CB(skb);
@@ -3106,8 +3106,8 @@ static int tcp_clean_rtx_queue(struct sock *sk, int prior_fackets,
 			flag |= FLAG_RETRANS_DATA_ACKED;
 		} else if (!(sacked & TCPCB_SACKED_ACKED)) {
 			last_ackt = skb->skb_mstamp;
-			WARN_ON_ONCE(last_ackt.v64 == 0);
-			if (!first_ackt.v64)
+			WARN_ON_ONCE(last_ackt == 0);
+			if (!first_ackt)
 				first_ackt = last_ackt;
 
 			last_in_flight = TCP_SKB_CB(skb)->tx.in_flight;
@@ -3122,7 +3122,7 @@ static int tcp_clean_rtx_queue(struct sock *sk, int prior_fackets,
 			tp->delivered += acked_pcount;
 			if (!tcp_skb_spurious_retrans(tp, skb))
 				tcp_rack_advance(tp, sacked, scb->end_seq,
-						 &skb->skb_mstamp);
+						 skb->skb_mstamp);
 		}
 		if (sacked & TCPCB_LOST)
 			tp->lost_out -= acked_pcount;
@@ -3165,13 +3165,13 @@ 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 (likely(first_ackt.v64) && !(flag & FLAG_RETRANS_DATA_ACKED)) {
-		seq_rtt_us = skb_mstamp_us_delta(now, &first_ackt);
-		ca_rtt_us = skb_mstamp_us_delta(now, &last_ackt);
+	if (likely(first_ackt) && !(flag & FLAG_RETRANS_DATA_ACKED)) {
+		seq_rtt_us = tcp_stamp_us_delta(tp->tcp_mstamp, first_ackt);
+		ca_rtt_us = tcp_stamp_us_delta(tp->tcp_mstamp, last_ackt);
 	}
-	if (sack->first_sackt.v64) {
-		sack_rtt_us = skb_mstamp_us_delta(now, &sack->first_sackt);
-		ca_rtt_us = skb_mstamp_us_delta(now, &sack->last_sackt);
+	if (sack->first_sackt) {
+		sack_rtt_us = tcp_stamp_us_delta(tp->tcp_mstamp, sack->first_sackt);
+		ca_rtt_us = tcp_stamp_us_delta(tp->tcp_mstamp, sack->last_sackt);
 	}
 	sack->rate->rtt_us = ca_rtt_us; /* RTT of last (S)ACKed packet, or -1 */
 	rtt_update = tcp_ack_update_rtt(sk, flag, seq_rtt_us, sack_rtt_us,
@@ -3201,7 +3201,7 @@ static int tcp_clean_rtx_queue(struct sock *sk, int prior_fackets,
 		tp->fackets_out -= min(pkts_acked, tp->fackets_out);
 
 	} else if (skb && rtt_update && sack_rtt_us >= 0 &&
-		   sack_rtt_us > skb_mstamp_us_delta(now, &skb->skb_mstamp)) {
+		   sack_rtt_us > tcp_stamp_us_delta(tp->tcp_mstamp, skb->skb_mstamp)) {
 		/* Do not re-arm RTO if the sack RTT is measured from data sent
 		 * after when the head was last (re)transmitted. Otherwise the
 		 * timeout may continue to extend in loss recovery.
@@ -3553,7 +3553,7 @@ static int tcp_ack(struct sock *sk, const struct sk_buff *skb, int flag)
 	int acked = 0; /* Number of packets newly acked */
 	int rexmit = REXMIT_NONE; /* Flag to (re)transmit to recover losses */
 
-	sack_state.first_sackt.v64 = 0;
+	sack_state.first_sackt = 0;
 	sack_state.rate = &rs;
 
 	/* We very likely will need to access write queue head. */
@@ -5356,7 +5356,7 @@ void tcp_rcv_established(struct sock *sk, struct sk_buff *skb,
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 
-	skb_mstamp_get(&tp->tcp_mstamp);
+	tcp_mstamp_refresh(tp);
 	if (unlikely(!sk->sk_rx_dst))
 		inet_csk(sk)->icsk_af_ops->sk_rx_dst_set(sk, skb);
 	/*
@@ -5672,7 +5672,7 @@ static int tcp_rcv_synsent_state_process(struct sock *sk, struct sk_buff *skb,
 
 		if (tp->rx_opt.saw_tstamp && tp->rx_opt.rcv_tsecr &&
 		    !between(tp->rx_opt.rcv_tsecr, tp->retrans_stamp,
-			     tcp_time_stamp)) {
+			     tcp_time_stamp(tp))) {
 			NET_INC_STATS(sock_net(sk),
 					LINUX_MIB_PAWSACTIVEREJECTED);
 			goto reset_and_undo;
@@ -5917,7 +5917,7 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb)
 
 	case TCP_SYN_SENT:
 		tp->rx_opt.saw_tstamp = 0;
-		skb_mstamp_get(&tp->tcp_mstamp);
+		tcp_mstamp_refresh(tp);
 		queued = tcp_rcv_synsent_state_process(sk, skb, th);
 		if (queued >= 0)
 			return queued;
@@ -5929,7 +5929,7 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb)
 		return 0;
 	}
 
-	skb_mstamp_get(&tp->tcp_mstamp);
+	tcp_mstamp_refresh(tp);
 	tp->rx_opt.saw_tstamp = 0;
 	req = tp->fastopen_rsk;
 	if (req) {
@@ -6202,7 +6202,7 @@ static void tcp_openreq_init(struct request_sock *req,
 	req->cookie_ts = 0;
 	tcp_rsk(req)->rcv_isn = TCP_SKB_CB(skb)->seq;
 	tcp_rsk(req)->rcv_nxt = TCP_SKB_CB(skb)->seq + 1;
-	skb_mstamp_get(&tcp_rsk(req)->snt_synack);
+	tcp_rsk(req)->snt_synack = tcp_clock_us();
 	tcp_rsk(req)->last_oow_ack_time = 0;
 	req->mss = rx_opt->mss_clamp;
 	req->ts_recent = rx_opt->saw_tstamp ? rx_opt->rcv_tsval : 0;
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index d8fe25db79f2..191b2f78b19d 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -376,8 +376,9 @@ void tcp_v4_err(struct sk_buff *icmp_skb, u32 info)
 	struct sock *sk;
 	struct sk_buff *skb;
 	struct request_sock *fastopen;
-	__u32 seq, snd_una;
-	__u32 remaining;
+	u32 seq, snd_una;
+	s32 remaining;
+	u32 delta_us;
 	int err;
 	struct net *net = dev_net(icmp_skb->dev);
 
@@ -483,12 +484,12 @@ void tcp_v4_err(struct sk_buff *icmp_skb, u32 info)
 		skb = tcp_write_queue_head(sk);
 		BUG_ON(!skb);
 
-		skb_mstamp_get(&tp->tcp_mstamp);
+		tcp_mstamp_refresh(tp);
+		delta_us = (u32)(tp->tcp_mstamp - skb->skb_mstamp);
 		remaining = icsk->icsk_rto -
-			    min(icsk->icsk_rto,
-				tcp_time_stamp - tcp_skb_timestamp(skb));
+			    usecs_to_jiffies(delta_us);
 
-		if (remaining) {
+		if (remaining > 0) {
 			inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS,
 						  remaining, TCP_RTO_MAX);
 		} else {
@@ -812,7 +813,7 @@ static void tcp_v4_timewait_ack(struct sock *sk, struct sk_buff *skb)
 	tcp_v4_send_ack(sk, skb,
 			tcptw->tw_snd_nxt, tcptw->tw_rcv_nxt,
 			tcptw->tw_rcv_wnd >> tw->tw_rcv_wscale,
-			tcp_time_stamp + tcptw->tw_ts_offset,
+			tcp_time_stamp_raw() + tcptw->tw_ts_offset,
 			tcptw->tw_ts_recent,
 			tw->tw_bound_dev_if,
 			tcp_twsk_md5_key(tcptw),
@@ -840,7 +841,7 @@ static void tcp_v4_reqsk_send_ack(const struct sock *sk, struct sk_buff *skb,
 	tcp_v4_send_ack(sk, skb, seq,
 			tcp_rsk(req)->rcv_nxt,
 			req->rsk_rcv_wnd >> inet_rsk(req)->rcv_wscale,
-			tcp_time_stamp + tcp_rsk(req)->ts_off,
+			tcp_time_stamp_raw() + tcp_rsk(req)->ts_off,
 			req->ts_recent,
 			0,
 			tcp_md5_do_lookup(sk, (union tcp_md5_addr *)&ip_hdr(skb)->daddr,
diff --git a/net/ipv4/tcp_lp.c b/net/ipv4/tcp_lp.c
index ef3122abb373..ae10ed64fe13 100644
--- a/net/ipv4/tcp_lp.c
+++ b/net/ipv4/tcp_lp.c
@@ -37,7 +37,7 @@
 #include <net/tcp.h>
 
 /* resolution of owd */
-#define LP_RESOL       1000
+#define LP_RESOL       TCP_TS_HZ
 
 /**
  * enum tcp_lp_state
@@ -147,9 +147,9 @@ static u32 tcp_lp_remote_hz_estimator(struct sock *sk)
 	    tp->rx_opt.rcv_tsecr == lp->local_ref_time)
 		goto out;
 
-	m = HZ * (tp->rx_opt.rcv_tsval -
-		  lp->remote_ref_time) / (tp->rx_opt.rcv_tsecr -
-					  lp->local_ref_time);
+	m = TCP_TS_HZ *
+	    (tp->rx_opt.rcv_tsval - lp->remote_ref_time) /
+	    (tp->rx_opt.rcv_tsecr - lp->local_ref_time);
 	if (m < 0)
 		m = -m;
 
@@ -194,7 +194,7 @@ static u32 tcp_lp_owd_calculator(struct sock *sk)
 	if (lp->flag & LP_VALID_RHZ) {
 		owd =
 		    tp->rx_opt.rcv_tsval * (LP_RESOL / lp->remote_hz) -
-		    tp->rx_opt.rcv_tsecr * (LP_RESOL / HZ);
+		    tp->rx_opt.rcv_tsecr * (LP_RESOL / TCP_TS_HZ);
 		if (owd < 0)
 			owd = -owd;
 	}
@@ -264,7 +264,7 @@ static void tcp_lp_pkts_acked(struct sock *sk, const struct ack_sample *sample)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 	struct lp *lp = inet_csk_ca(sk);
-	u32 now = tcp_time_stamp;
+	u32 now = tcp_time_stamp(tp);
 	u32 delta;
 
 	if (sample->rtt_us > 0)
diff --git a/net/ipv4/tcp_minisocks.c b/net/ipv4/tcp_minisocks.c
index 6504f1082bdf..d0642df73044 100644
--- a/net/ipv4/tcp_minisocks.c
+++ b/net/ipv4/tcp_minisocks.c
@@ -455,7 +455,7 @@ struct sock *tcp_create_openreq_child(const struct sock *sk,
 		newtp->fackets_out = 0;
 		newtp->snd_ssthresh = TCP_INFINITE_SSTHRESH;
 		newtp->tlp_high_seq = 0;
-		newtp->lsndtime = treq->snt_synack.stamp_jiffies;
+		newtp->lsndtime = tcp_jiffies32;
 		newsk->sk_txhash = treq->txhash;
 		newtp->last_oow_ack_time = 0;
 		newtp->total_retrans = req->num_retrans;
@@ -526,7 +526,7 @@ struct sock *tcp_create_openreq_child(const struct sock *sk,
 		newtp->fastopen_req = NULL;
 		newtp->fastopen_rsk = NULL;
 		newtp->syn_data_acked = 0;
-		newtp->rack.mstamp.v64 = 0;
+		newtp->rack.mstamp = 0;
 		newtp->rack.advanced = 0;
 
 		__TCP_INC_STATS(sock_net(sk), TCP_MIB_PASSIVEOPENS);
diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c
index 65472e931a0b..478f75baee31 100644
--- a/net/ipv4/tcp_output.c
+++ b/net/ipv4/tcp_output.c
@@ -1962,7 +1962,7 @@ static bool tcp_tso_should_defer(struct sock *sk, struct sk_buff *skb,
 
 	head = tcp_write_queue_head(sk);
 
-	age = skb_mstamp_us_delta(&tp->tcp_mstamp, &head->skb_mstamp);
+	age = tcp_stamp_us_delta(tp->tcp_mstamp, head->skb_mstamp);
 	/* If next ACK is likely to come too late (half srtt), do not defer */
 	if (age < (tp->srtt_us >> 4))
 		goto send_now;
@@ -2279,7 +2279,7 @@ static bool tcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle,
 	}
 
 	max_segs = tcp_tso_segs(sk, mss_now);
-	skb_mstamp_get(&tp->tcp_mstamp);
+	tcp_mstamp_refresh(tp);
 	while ((skb = tcp_send_head(sk))) {
 		unsigned int limit;
 
@@ -3095,7 +3095,7 @@ void tcp_send_active_reset(struct sock *sk, gfp_t priority)
 	skb_reserve(skb, MAX_TCP_HEADER);
 	tcp_init_nondata_skb(skb, tcp_acceptable_seq(sk),
 			     TCPHDR_ACK | TCPHDR_RST);
-	skb_mstamp_get(&tcp_sk(sk)->tcp_mstamp);
+	tcp_mstamp_refresh(tcp_sk(sk));
 	/* Send it off. */
 	if (tcp_transmit_skb(sk, skb, 0, priority))
 		NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPABORTFAILED);
@@ -3191,10 +3191,10 @@ struct sk_buff *tcp_make_synack(const struct sock *sk, struct dst_entry *dst,
 	memset(&opts, 0, sizeof(opts));
 #ifdef CONFIG_SYN_COOKIES
 	if (unlikely(req->cookie_ts))
-		skb->skb_mstamp.stamp_jiffies = cookie_init_timestamp(req);
+		skb->skb_mstamp = cookie_init_timestamp(req);
 	else
 #endif
-	skb_mstamp_get(&skb->skb_mstamp);
+		skb->skb_mstamp = tcp_clock_us();
 
 #ifdef CONFIG_TCP_MD5SIG
 	rcu_read_lock();
@@ -3453,8 +3453,8 @@ int tcp_connect(struct sock *sk)
 		return -ENOBUFS;
 
 	tcp_init_nondata_skb(buff, tp->write_seq++, TCPHDR_SYN);
-	skb_mstamp_get(&tp->tcp_mstamp);
-	tp->retrans_stamp = tp->tcp_mstamp.stamp_jiffies;
+	tcp_mstamp_refresh(tp);
+	tp->retrans_stamp = tcp_time_stamp(tp);
 	tcp_connect_queue_skb(sk, buff);
 	tcp_ecn_send_syn(sk, buff);
 
@@ -3615,7 +3615,7 @@ void tcp_send_window_probe(struct sock *sk)
 {
 	if (sk->sk_state == TCP_ESTABLISHED) {
 		tcp_sk(sk)->snd_wl1 = tcp_sk(sk)->rcv_nxt - 1;
-		skb_mstamp_get(&tcp_sk(sk)->tcp_mstamp);
+		tcp_mstamp_refresh(tcp_sk(sk));
 		tcp_xmit_probe_skb(sk, 0, LINUX_MIB_TCPWINPROBE);
 	}
 }
diff --git a/net/ipv4/tcp_rate.c b/net/ipv4/tcp_rate.c
index c6a9fa894646..ad99569d4c1e 100644
--- a/net/ipv4/tcp_rate.c
+++ b/net/ipv4/tcp_rate.c
@@ -78,7 +78,7 @@ void tcp_rate_skb_delivered(struct sock *sk, struct sk_buff *skb,
 	struct tcp_sock *tp = tcp_sk(sk);
 	struct tcp_skb_cb *scb = TCP_SKB_CB(skb);
 
-	if (!scb->tx.delivered_mstamp.v64)
+	if (!scb->tx.delivered_mstamp)
 		return;
 
 	if (!rs->prior_delivered ||
@@ -89,9 +89,9 @@ void tcp_rate_skb_delivered(struct sock *sk, struct sk_buff *skb,
 		rs->is_retrans	     = scb->sacked & TCPCB_RETRANS;
 
 		/* Find the duration of the "send phase" of this window: */
-		rs->interval_us      = skb_mstamp_us_delta(
-						&skb->skb_mstamp,
-						&scb->tx.first_tx_mstamp);
+		rs->interval_us      = tcp_stamp_us_delta(
+						skb->skb_mstamp,
+						scb->tx.first_tx_mstamp);
 
 		/* Record send time of most recently ACKed packet: */
 		tp->first_tx_mstamp  = skb->skb_mstamp;
@@ -101,7 +101,7 @@ void tcp_rate_skb_delivered(struct sock *sk, struct sk_buff *skb,
 	 * we don't need to reset since it'll be freed soon.
 	 */
 	if (scb->sacked & TCPCB_SACKED_ACKED)
-		scb->tx.delivered_mstamp.v64 = 0;
+		scb->tx.delivered_mstamp = 0;
 }
 
 /* Update the connection delivery information and generate a rate sample. */
@@ -125,7 +125,7 @@ void tcp_rate_gen(struct sock *sk, u32 delivered, u32 lost,
 	rs->acked_sacked = delivered;	/* freshly ACKed or SACKed */
 	rs->losses = lost;		/* freshly marked lost */
 	/* Return an invalid sample if no timing information is available. */
-	if (!rs->prior_mstamp.v64) {
+	if (!rs->prior_mstamp) {
 		rs->delivered = -1;
 		rs->interval_us = -1;
 		return;
@@ -138,8 +138,8 @@ void tcp_rate_gen(struct sock *sk, u32 delivered, u32 lost,
 	 * longer phase.
 	 */
 	snd_us = rs->interval_us;				/* send phase */
-	ack_us = skb_mstamp_us_delta(&tp->tcp_mstamp,
-				     &rs->prior_mstamp); /* ack phase */
+	ack_us = tcp_stamp_us_delta(tp->tcp_mstamp,
+				    rs->prior_mstamp); /* ack phase */
 	rs->interval_us = max(snd_us, ack_us);
 
 	/* Normally we expect interval_us >= min-rtt.
diff --git a/net/ipv4/tcp_recovery.c b/net/ipv4/tcp_recovery.c
index cd72b3d3879e..fe9a493d0208 100644
--- a/net/ipv4/tcp_recovery.c
+++ b/net/ipv4/tcp_recovery.c
@@ -17,12 +17,9 @@ static void tcp_rack_mark_skb_lost(struct sock *sk, struct sk_buff *skb)
 	}
 }
 
-static bool tcp_rack_sent_after(const struct skb_mstamp *t1,
-				const struct skb_mstamp *t2,
-				u32 seq1, u32 seq2)
+static bool tcp_rack_sent_after(u64 t1, u64 t2, u32 seq1, u32 seq2)
 {
-	return skb_mstamp_after(t1, t2) ||
-	       (t1->v64 == t2->v64 && after(seq1, seq2));
+	return t1 > t2 || (t1 == t2 && after(seq1, seq2));
 }
 
 /* RACK loss detection (IETF draft draft-ietf-tcpm-rack-01):
@@ -72,14 +69,14 @@ static void tcp_rack_detect_loss(struct sock *sk, u32 *reo_timeout)
 		    scb->sacked & TCPCB_SACKED_ACKED)
 			continue;
 
-		if (tcp_rack_sent_after(&tp->rack.mstamp, &skb->skb_mstamp,
+		if (tcp_rack_sent_after(tp->rack.mstamp, skb->skb_mstamp,
 					tp->rack.end_seq, scb->end_seq)) {
 			/* Step 3 in draft-cheng-tcpm-rack-00.txt:
 			 * A packet is lost if its elapsed time is beyond
 			 * the recent RTT plus the reordering window.
 			 */
-			u32 elapsed = skb_mstamp_us_delta(&tp->tcp_mstamp,
-							  &skb->skb_mstamp);
+			u32 elapsed = tcp_stamp_us_delta(tp->tcp_mstamp,
+							 skb->skb_mstamp);
 			s32 remaining = tp->rack.rtt_us + reo_wnd - elapsed;
 
 			if (remaining < 0) {
@@ -127,16 +124,16 @@ void tcp_rack_mark_lost(struct sock *sk)
  * draft-cheng-tcpm-rack-00.txt
  */
 void tcp_rack_advance(struct tcp_sock *tp, u8 sacked, u32 end_seq,
-		      const struct skb_mstamp *xmit_time)
+		      u64 xmit_time)
 {
 	u32 rtt_us;
 
-	if (tp->rack.mstamp.v64 &&
-	    !tcp_rack_sent_after(xmit_time, &tp->rack.mstamp,
+	if (tp->rack.mstamp &&
+	    !tcp_rack_sent_after(xmit_time, tp->rack.mstamp,
 				 end_seq, tp->rack.end_seq))
 		return;
 
-	rtt_us = skb_mstamp_us_delta(&tp->tcp_mstamp, xmit_time);
+	rtt_us = tcp_stamp_us_delta(tp->tcp_mstamp, xmit_time);
 	if (sacked & TCPCB_RETRANS) {
 		/* If the sacked packet was retransmitted, it's ambiguous
 		 * whether the retransmission or the original (or the prior
@@ -152,7 +149,7 @@ void tcp_rack_advance(struct tcp_sock *tp, u8 sacked, u32 end_seq,
 			return;
 	}
 	tp->rack.rtt_us = rtt_us;
-	tp->rack.mstamp = *xmit_time;
+	tp->rack.mstamp = xmit_time;
 	tp->rack.end_seq = end_seq;
 	tp->rack.advanced = 1;
 }
diff --git a/net/ipv4/tcp_timer.c b/net/ipv4/tcp_timer.c
index 6629f47aa7f0..27a667bce806 100644
--- a/net/ipv4/tcp_timer.c
+++ b/net/ipv4/tcp_timer.c
@@ -153,8 +153,8 @@ static bool retransmits_timed_out(struct sock *sk,
 				  unsigned int timeout,
 				  bool syn_set)
 {
-	unsigned int linear_backoff_thresh, start_ts;
 	unsigned int rto_base = syn_set ? TCP_TIMEOUT_INIT : TCP_RTO_MIN;
+	unsigned int linear_backoff_thresh, start_ts;
 
 	if (!inet_csk(sk)->icsk_retransmits)
 		return false;
@@ -172,7 +172,7 @@ static bool retransmits_timed_out(struct sock *sk,
 			timeout = ((2 << linear_backoff_thresh) - 1) * rto_base +
 				(boundary - linear_backoff_thresh) * TCP_RTO_MAX;
 	}
-	return (tcp_time_stamp - start_ts) >= timeout;
+	return (tcp_time_stamp(tcp_sk(sk)) - start_ts) >= jiffies_to_msecs(timeout);
 }
 
 /* A write timeout has occurred. Process the after effects. */
@@ -341,7 +341,7 @@ static void tcp_probe_timer(struct sock *sk)
 	if (!start_ts)
 		tcp_send_head(sk)->skb_mstamp = tp->tcp_mstamp;
 	else if (icsk->icsk_user_timeout &&
-		 (s32)(tcp_time_stamp - start_ts) > icsk->icsk_user_timeout)
+		 (s32)(tcp_time_stamp(tp) - start_ts) > icsk->icsk_user_timeout)
 		goto abort;
 
 	max_probes = sock_net(sk)->ipv4.sysctl_tcp_retries2;
@@ -561,7 +561,7 @@ void tcp_write_timer_handler(struct sock *sk)
 		goto out;
 	}
 
-	skb_mstamp_get(&tcp_sk(sk)->tcp_mstamp);
+	tcp_mstamp_refresh(tcp_sk(sk));
 	event = icsk->icsk_pending;
 
 	switch (event) {