tcp: add sysctl_tcp_rto_min_us
[ Upstream commit f086edef71be7174a16c1ed67ac65a085cda28b1 ]
Adding a sysctl knob to allow user to specify a default
rto_min at socket init time, other than using the hard
coded 200ms default rto_min.
Note that the rto_min route option has the highest precedence
for configuring this setting, followed by the TCP_BPF_RTO_MIN
socket option, followed by the tcp_rto_min_us sysctl.
BUG=b/345799473
TEST=presubmit
RELEASE_NOTE=Added tcp_rto_min_us sysctl.
cos-patch: bug
Signed-off-by: Kevin Yang <yyd@google.com>
Reviewed-by: Neal Cardwell <ncardwell@google.com>
Reviewed-by: Yuchung Cheng <ycheng@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Tony Lu <tonylu@linux.alibaba.com>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Change-Id: I5f1ef617eb30f9f4498a35035e6edc872b9a2e89
Reviewed-on: https://cos-review.googlesource.com/c/third_party/kernel/+/74682
Reviewed-by: Arnav Kansal <rnv@google.com>
Reviewed-by: Kevin Berry <kpberry@google.com>
Tested-by: Cusky Presubmit Bot <presubmit@cos-infra-prod.iam.gserviceaccount.com>
Main-Branch-Verified: Cusky Presubmit Bot <presubmit@cos-infra-prod.iam.gserviceaccount.com>
diff --git a/Documentation/networking/ip-sysctl.rst b/Documentation/networking/ip-sysctl.rst
index f539f17..498147b 100644
--- a/Documentation/networking/ip-sysctl.rst
+++ b/Documentation/networking/ip-sysctl.rst
@@ -1094,6 +1094,19 @@
Default: 0
+tcp_rto_min_us - INTEGER
+ Minimal TCP retransmission timeout (in microseconds). Note that the
+ rto_min route option has the highest precedence for configuring this
+ setting, followed by the TCP_BPF_RTO_MIN socket option, followed by
+ this tcp_rto_min_us sysctl.
+
+ The recommended practice is to use a value less or equal to 200000
+ microseconds.
+
+ Possible Values: 1 - INT_MAX
+
+ Default: 200000
+
UDP variables
=============
diff --git a/include/net/netns/ipv4.h b/include/net/netns/ipv4.h
index ad6b02f..2b913e0 100644
--- a/include/net/netns/ipv4.h
+++ b/include/net/netns/ipv4.h
@@ -141,6 +141,7 @@ struct netns_ipv4 {
u8 sysctl_tcp_sack;
u8 sysctl_tcp_window_scaling;
u8 sysctl_tcp_timestamps;
+ int sysctl_tcp_rto_min_us;
u8 sysctl_tcp_early_retrans;
u8 sysctl_tcp_recovery;
u8 sysctl_tcp_thin_linear_timeouts;
diff --git a/net/ipv4/sysctl_net_ipv4.c b/net/ipv4/sysctl_net_ipv4.c
index f4d8396..79a38ca 100644
--- a/net/ipv4/sysctl_net_ipv4.c
+++ b/net/ipv4/sysctl_net_ipv4.c
@@ -1405,6 +1405,14 @@ static struct ctl_table ipv4_net_table[] = {
.extra1 = SYSCTL_ZERO,
.extra2 = SYSCTL_ONE,
},
+ {
+ .procname = "tcp_rto_min_us",
+ .data = &init_net.ipv4.sysctl_tcp_rto_min_us,
+ .maxlen = sizeof(int),
+ .mode = 0644,
+ .proc_handler = proc_dointvec_minmax,
+ .extra1 = SYSCTL_ONE,
+ },
{ }
};
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 2aa92657..2f91a5f 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -414,6 +414,7 @@ void tcp_init_sock(struct sock *sk)
{
struct inet_connection_sock *icsk = inet_csk(sk);
struct tcp_sock *tp = tcp_sk(sk);
+ int rto_min_us;
tp->out_of_order_queue = RB_ROOT;
sk->tcp_rtx_queue = RB_ROOT;
@@ -422,7 +423,8 @@ void tcp_init_sock(struct sock *sk)
INIT_LIST_HEAD(&tp->tsorted_sent_queue);
icsk->icsk_rto = TCP_TIMEOUT_INIT;
- icsk->icsk_rto_min = TCP_RTO_MIN;
+ rto_min_us = READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_rto_min_us);
+ icsk->icsk_rto_min = usecs_to_jiffies(rto_min_us);
icsk->icsk_delack_max = TCP_DELACK_MAX;
tp->mdev_us = jiffies_to_usecs(TCP_TIMEOUT_INIT);
minmax_reset(&tp->rtt_min, tcp_jiffies32, ~0U);
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 8d40e06..ca2e3f36 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -3234,6 +3234,8 @@ static int __net_init tcp_sk_init(struct net *net)
net->ipv4.sysctl_tcp_shrink_window = 0;
+ net->ipv4.sysctl_tcp_rto_min_us = jiffies_to_usecs(TCP_RTO_MIN);
+
return 0;
}
