diff options
| author | Yuchung Cheng <ycheng@google.com> | 2013-09-03 14:14:35 -0700 |
|---|---|---|
| committer | David S. Miller <davem@davemloft.net> | 2013-09-04 14:41:55 -0400 |
| commit | 52f20e655d9f6f7f937a1cdacf219d9df3ab6166 (patch) | |
| tree | 234a887f331af07ebb2ff5a1ae80823f8e1456b5 /net | |
| parent | 430eda6d6d568eded71dfd1be5a16c0c1379e201 (diff) | |
tcp: better comments for RTO initiallization
Commit 1b7fdd2ab585("tcp: do not use cached RTT for RTT estimation")
removes important comments on how RTO is initialized and updated.
Hopefully this patch puts those information back.
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Diffstat (limited to 'net')
| -rw-r--r-- | net/ipv4/tcp_metrics.c | 26 |
1 files changed, 20 insertions, 6 deletions
diff --git a/net/ipv4/tcp_metrics.c b/net/ipv4/tcp_metrics.c index 273ed735cca2..4a22f3e715df 100644 --- a/net/ipv4/tcp_metrics.c +++ b/net/ipv4/tcp_metrics.c @@ -481,13 +481,27 @@ void tcp_init_metrics(struct sock *sk) crtt = tcp_metric_get_jiffies(tm, TCP_METRIC_RTT); rcu_read_unlock(); reset: + /* The initial RTT measurement from the SYN/SYN-ACK is not ideal + * to seed the RTO for later data packets because SYN packets are + * small. Use the per-dst cached values to seed the RTO but keep + * the RTT estimator variables intact (e.g., srtt, mdev, rttvar). + * Later the RTO will be updated immediately upon obtaining the first + * data RTT sample (tcp_rtt_estimator()). Hence the cached RTT only + * influences the first RTO but not later RTT estimation. + * + * But if RTT is not available from the SYN (due to retransmits or + * syn cookies) or the cache, force a conservative 3secs timeout. + * + * A bit of theory. RTT is time passed after "normal" sized packet + * is sent until it is ACKed. In normal circumstances sending small + * packets force peer to delay ACKs and calculation is correct too. + * The algorithm is adaptive and, provided we follow specs, it + * NEVER underestimate RTT. BUT! If peer tries to make some clever + * tricks sort of "quick acks" for time long enough to decrease RTT + * to low value, and then abruptly stops to do it and starts to delay + * ACKs, wait for troubles. + */ if (crtt > tp->srtt) { - /* Initial RTT (tp->srtt) from SYN usually don't measure - * serialization delay on low BW links well so RTO may be - * under-estimated. Stay conservative and seed RTO with - * the RTTs from past data exchanges, using the same seeding - * formula in tcp_rtt_estimator(). - */ inet_csk(sk)->icsk_rto = crtt + max(crtt >> 2, tcp_rto_min(sk)); } else if (tp->srtt == 0) { /* RFC6298: 5.7 We've failed to get a valid RTT sample from |