tcp: use tp->tcp_mstamp in output path

Idea is to later convert tp->tcp_mstamp to a full u64 counter
using usec resolution, so that we can later have fine
grained TCP TS clock (RFC 7323), regardless of HZ value.

We try to refresh tp->tcp_mstamp only when necessary.

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>

4 files changed, 14 insertions, 12 deletions

diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 5ab2aac5ca19..d8fe25db79f2 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -483,6 +483,7 @@ 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);
 		remaining = icsk->icsk_rto -
 			    min(icsk->icsk_rto,
 				tcp_time_stamp - tcp_skb_timestamp(skb));
diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c
index a32172d69a03..4c8a6eaba6b3 100644
--- a/net/ipv4/tcp_output.c
+++ b/net/ipv4/tcp_output.c
@@ -997,8 +997,8 @@ static int tcp_transmit_skb(struct sock *sk, struct sk_buff *skb, int clone_it,
 	BUG_ON(!skb || !tcp_skb_pcount(skb));
 	tp = tcp_sk(sk);
 
+	skb->skb_mstamp = tp->tcp_mstamp;
 	if (clone_it) {
-		skb_mstamp_get(&skb->skb_mstamp);
 		TCP_SKB_CB(skb)->tx.in_flight = TCP_SKB_CB(skb)->end_seq
 			- tp->snd_una;
 		tcp_rate_skb_sent(sk, skb);
@@ -1906,7 +1906,6 @@ static bool tcp_tso_should_defer(struct sock *sk, struct sk_buff *skb,
 	const struct inet_connection_sock *icsk = inet_csk(sk);
 	u32 age, send_win, cong_win, limit, in_flight;
 	struct tcp_sock *tp = tcp_sk(sk);
-	struct skb_mstamp now;
 	struct sk_buff *head;
 	int win_divisor;
 
@@ -1962,8 +1961,8 @@ static bool tcp_tso_should_defer(struct sock *sk, struct sk_buff *skb,
 	}
 
 	head = tcp_write_queue_head(sk);
-	skb_mstamp_get(&now);
-	age = skb_mstamp_us_delta(&now, &head->skb_mstamp);
+
+	age = skb_mstamp_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;
@@ -2280,6 +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);
 	while ((skb = tcp_send_head(sk))) {
 		unsigned int limit;
 
@@ -2291,7 +2291,7 @@ static bool tcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle,
 
 		if (unlikely(tp->repair) && tp->repair_queue == TCP_SEND_QUEUE) {
 			/* "skb_mstamp" is used as a start point for the retransmit timer */
-			skb_mstamp_get(&skb->skb_mstamp);
+			skb->skb_mstamp = tp->tcp_mstamp;
 			goto repair; /* Skip network transmission */
 		}
 
@@ -2879,7 +2879,7 @@ int __tcp_retransmit_skb(struct sock *sk, struct sk_buff *skb, int segs)
 		     skb_headroom(skb) >= 0xFFFF)) {
 		struct sk_buff *nskb;
 
-		skb_mstamp_get(&skb->skb_mstamp);
+		skb->skb_mstamp = tp->tcp_mstamp;
 		nskb = __pskb_copy(skb, MAX_TCP_HEADER, GFP_ATOMIC);
 		err = nskb ? tcp_transmit_skb(sk, nskb, 0, GFP_ATOMIC) :
 			     -ENOBUFS;
@@ -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(&skb->skb_mstamp);
+	skb_mstamp_get(&tcp_sk(sk)->tcp_mstamp);
 	/* Send it off. */
 	if (tcp_transmit_skb(sk, skb, 0, priority))
 		NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPABORTFAILED);
@@ -3453,7 +3453,8 @@ int tcp_connect(struct sock *sk)
 		return -ENOBUFS;
 
 	tcp_init_nondata_skb(buff, tp->write_seq++, TCPHDR_SYN);
-	tp->retrans_stamp = tcp_time_stamp;
+	skb_mstamp_get(&tp->tcp_mstamp);
+	tp->retrans_stamp = tp->tcp_mstamp.stamp_jiffies;
 	tcp_connect_queue_skb(sk, buff);
 	tcp_ecn_send_syn(sk, buff);
 
@@ -3572,7 +3573,6 @@ void tcp_send_ack(struct sock *sk)
 	skb_set_tcp_pure_ack(buff);
 
 	/* Send it off, this clears delayed acks for us. */
-	skb_mstamp_get(&buff->skb_mstamp);
 	tcp_transmit_skb(sk, buff, 0, (__force gfp_t)0);
 }
 EXPORT_SYMBOL_GPL(tcp_send_ack);
@@ -3606,15 +3606,16 @@ static int tcp_xmit_probe_skb(struct sock *sk, int urgent, int mib)
 	 * send it.
 	 */
 	tcp_init_nondata_skb(skb, tp->snd_una - !urgent, TCPHDR_ACK);
-	skb_mstamp_get(&skb->skb_mstamp);
 	NET_INC_STATS(sock_net(sk), mib);
 	return tcp_transmit_skb(sk, skb, 0, (__force gfp_t)0);
 }
 
+/* Called from setsockopt( ... TCP_REPAIR ) */
 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_xmit_probe_skb(sk, 0, LINUX_MIB_TCPWINPROBE);
 	}
 }
diff --git a/net/ipv4/tcp_recovery.c b/net/ipv4/tcp_recovery.c
index 362b8c75bfab..cd72b3d3879e 100644
--- a/net/ipv4/tcp_recovery.c
+++ b/net/ipv4/tcp_recovery.c
@@ -166,7 +166,6 @@ void tcp_rack_reo_timeout(struct sock *sk)
 	u32 timeout, prior_inflight;
 
 	prior_inflight = tcp_packets_in_flight(tp);
-	skb_mstamp_get(&tp->tcp_mstamp);
 	tcp_rack_detect_loss(sk, &timeout);
 	if (prior_inflight != tcp_packets_in_flight(tp)) {
 		if (inet_csk(sk)->icsk_ca_state != TCP_CA_Recovery) {
diff --git a/net/ipv4/tcp_timer.c b/net/ipv4/tcp_timer.c
index 86934bcf685a..ec7c5473c788 100644
--- a/net/ipv4/tcp_timer.c
+++ b/net/ipv4/tcp_timer.c
@@ -339,7 +339,7 @@ static void tcp_probe_timer(struct sock *sk)
 	 */
 	start_ts = tcp_skb_timestamp(tcp_send_head(sk));
 	if (!start_ts)
-		skb_mstamp_get(&tcp_send_head(sk)->skb_mstamp);
+		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)
 		goto abort;
@@ -561,6 +561,7 @@ void tcp_write_timer_handler(struct sock *sk)
 		goto out;
 	}
 
+	skb_mstamp_get(&tcp_sk(sk)->tcp_mstamp);
 	event = icsk->icsk_pending;
 
 	switch (event) {