tcp: add support for usec resolution in TCP TS values

Back in 2015, Van Jacobson suggested to use usec resolution in TCP TS values.
This has been implemented in our private kernels.

Goals were :

1) better observability of delays in networking stacks.
2) better disambiguation of events based on TSval/ecr values.
3) building block for congestion control modules needing usec resolution.

Back then we implemented a schem based on private SYN options
to negotiate the feature.

For upstream submission, we chose to use a route attribute,
because this feature is probably going to be used in private
networks [1] [2].

ip route add 10/8 ... features tcp_usec_ts

Note that RFC 7323 recommends a
  "timestamp clock frequency in the range 1 ms to 1 sec per tick.",
but also mentions
  "the maximum acceptable clock frequency is one tick every 59 ns."

[1] Unfortunately RFC 7323 5.5 (Outdated Timestamps) suggests
to invalidate TS.Recent values after a flow was idle for more
than 24 days. This is the part making usec_ts a problem
for peers following this recommendation for long living
idle flows.

[2] Attempts to standardize usec ts went nowhere:

https://www.ietf.org/proceedings/97/slides/slides-97-tcpm-tcp-options-for-low-latency-00.pdf
https://datatracker.ietf.org/doc/draft-wang-tcpm-low-latency-opt/

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

1 files changed, 26 insertions, 14 deletions

diff --git a/net/ipv4/tcp_timer.c b/net/ipv4/tcp_timer.c
index bfcf3fe44c72..1f9f6c1c196b 100644
--- a/net/ipv4/tcp_timer.c
+++ b/net/ipv4/tcp_timer.c
@@ -26,14 +26,18 @@
 static u32 tcp_clamp_rto_to_user_timeout(const struct sock *sk)
 {
 	struct inet_connection_sock *icsk = inet_csk(sk);
-	u32 elapsed, start_ts, user_timeout;
+	const struct tcp_sock *tp = tcp_sk(sk);
+	u32 elapsed, user_timeout;
 	s32 remaining;
 
-	start_ts = tcp_sk(sk)->retrans_stamp;
 	user_timeout = READ_ONCE(icsk->icsk_user_timeout);
 	if (!user_timeout)
 		return icsk->icsk_rto;
-	elapsed = tcp_time_stamp_ts(tcp_sk(sk)) - start_ts;
+
+	elapsed = tcp_time_stamp_ts(tp) - tp->retrans_stamp;
+	if (tp->tcp_usec_ts)
+		elapsed /= USEC_PER_MSEC;
+
 	remaining = user_timeout - elapsed;
 	if (remaining <= 0)
 		return 1; /* user timeout has passed; fire ASAP */
@@ -212,12 +216,13 @@ static bool retransmits_timed_out(struct sock *sk,
 				  unsigned int boundary,
 				  unsigned int timeout)
 {
-	unsigned int start_ts;
+	struct tcp_sock *tp = tcp_sk(sk);
+	unsigned int start_ts, delta;
 
 	if (!inet_csk(sk)->icsk_retransmits)
 		return false;
 
-	start_ts = tcp_sk(sk)->retrans_stamp;
+	start_ts = tp->retrans_stamp;
 	if (likely(timeout == 0)) {
 		unsigned int rto_base = TCP_RTO_MIN;
 
@@ -226,7 +231,12 @@ static bool retransmits_timed_out(struct sock *sk,
 		timeout = tcp_model_timeout(sk, boundary, rto_base);
 	}
 
-	return (s32)(tcp_time_stamp_ts(tcp_sk(sk)) - start_ts - timeout) >= 0;
+	if (tp->tcp_usec_ts) {
+		/* delta maybe off up to a jiffy due to timer granularity. */
+		delta = tp->tcp_mstamp - start_ts + jiffies_to_usecs(1);
+		return (s32)(delta - timeout * USEC_PER_MSEC) >= 0;
+	}
+	return (s32)(tcp_time_stamp_ts(tp) - start_ts - timeout) >= 0;
 }
 
 /* A write timeout has occurred. Process the after effects. */
@@ -468,20 +478,18 @@ static void tcp_fastopen_synack_timer(struct sock *sk, struct request_sock *req)
 }
 
 static bool tcp_rtx_probe0_timed_out(const struct sock *sk,
-				     const struct sk_buff *skb)
+				     const struct sk_buff *skb,
+				     u32 rtx_delta)
 {
 	const struct tcp_sock *tp = tcp_sk(sk);
 	const int timeout = TCP_RTO_MAX * 2;
-	u32 rcv_delta, rtx_delta;
+	u32 rcv_delta;
 
 	rcv_delta = inet_csk(sk)->icsk_timeout - tp->rcv_tstamp;
 	if (rcv_delta <= timeout)
 		return false;
 
-	rtx_delta = (u32)msecs_to_jiffies(tcp_time_stamp_ts(tp) -
-			(tp->retrans_stamp ?: tcp_skb_timestamp_ts(false, skb)));
-
-	return rtx_delta > timeout;
+	return msecs_to_jiffies(rtx_delta) > timeout;
 }
 
 /**
@@ -534,7 +542,11 @@ void tcp_retransmit_timer(struct sock *sk)
 		struct inet_sock *inet = inet_sk(sk);
 		u32 rtx_delta;
 
-		rtx_delta = tcp_time_stamp_ts(tp) - (tp->retrans_stamp ?: tcp_skb_timestamp_ts(false, skb));
+		rtx_delta = tcp_time_stamp_ts(tp) - (tp->retrans_stamp ?: 
+				tcp_skb_timestamp_ts(tp->tcp_usec_ts, skb));
+		if (tp->tcp_usec_ts)
+			rtx_delta /= USEC_PER_MSEC;
+
 		if (sk->sk_family == AF_INET) {
 			net_dbg_ratelimited("Probing zero-window on %pI4:%u/%u, seq=%u:%u, recv %ums ago, lasting %ums\n",
 				&inet->inet_daddr, ntohs(inet->inet_dport),
@@ -551,7 +563,7 @@ void tcp_retransmit_timer(struct sock *sk)
 				rtx_delta);
 		}
 #endif
-		if (tcp_rtx_probe0_timed_out(sk, skb)) {
+		if (tcp_rtx_probe0_timed_out(sk, skb, rtx_delta)) {
 			tcp_write_err(sk);
 			goto out;
 		}