mptcp: add a new sysctl for make after break timeout

The MPTCP protocol allows sockets with no alive subflows to stay
in ESTABLISHED status for and user-defined timeout, to allow for
later subflows creation.

Currently such timeout is constant - TCP_TIMEWAIT_LEN. Let the
user-space configure them via a newly added sysctl, to better cope
with busy servers and simplify (make them faster) the relevant
pktdrill tests.

Note that the new know does not apply to orphaned MPTCP socket
waiting for the data_fin handshake completion: they always wait
TCP_TIMEWAIT_LEN.

Reviewed-by: Mat Martineau <[e-mail]>
Signed-off-by: Paolo Abeni <[e-mail]>
Signed-off-by: Mat Martineau <[e-mail]>
Link: https://lore.kernel.org/r/[e-mail]
Signed-off-by: Jakub Kicinski <[e-mail]>

Documentation/networking/mptcp-sysctl.rst

···
 
 	Default: 120
 
+close_timeout - INTEGER (seconds)
+	Set the make-after-break timeout: in absence of any close or
+	shutdown syscall, MPTCP sockets will maintain the status
+	unchanged for such time, after the last subflow removal, before
+	moving to TCP_CLOSE.
+
+	The default value matches TCP_TIMEWAIT_LEN. This is a per-namespace
+	sysctl.
+
+	Default: 60
+
 checksum_enabled - BOOLEAN
 	Control whether DSS checksum can be enabled.
 

net/mptcp/ctrl.c

···
 #endif
 
 	unsigned int add_addr_timeout;
+	unsigned int close_timeout;
 	unsigned int stale_loss_cnt;
 	u8 mptcp_enabled;
 	u8 checksum_enabled;
···
 	return mptcp_get_pernet(net)->stale_loss_cnt;
 }
 
+unsigned int mptcp_close_timeout(const struct sock *sk)
+{
+	if (sock_flag(sk, SOCK_DEAD))
+		return TCP_TIMEWAIT_LEN;
+	return mptcp_get_pernet(sock_net(sk))->close_timeout;
+}
+
 int mptcp_get_pm_type(const struct net *net)
 {
 	return mptcp_get_pernet(net)->pm_type;
···
 {
 	pernet->mptcp_enabled = 1;
 	pernet->add_addr_timeout = TCP_RTO_MAX;
+	pernet->close_timeout = TCP_TIMEWAIT_LEN;
 	pernet->checksum_enabled = 0;
 	pernet->allow_join_initial_addr_port = 1;
 	pernet->stale_loss_cnt = 4;
···
 		.mode = 0644,
 		.proc_handler = proc_dostring,
 	},
+	{
+		.procname = "close_timeout",
+		.maxlen = sizeof(unsigned int),
+		.mode = 0644,
+		.proc_handler = proc_dointvec_jiffies,
+	},
 	{}
 };
 
···
 	table[4].data = &pernet->stale_loss_cnt;
 	table[5].data = &pernet->pm_type;
 	table[6].data = &pernet->scheduler;
+	table[7].data = &pernet->close_timeout;
 
 	hdr = register_net_sysctl_sz(net, MPTCP_SYSCTL_PATH, table,
 				     ARRAY_SIZE(mptcp_sysctl_table));

net/mptcp/protocol.c

···
 	if (msk->in_accept_queue && msk->first == ssk &&
 	    (sock_flag(sk, SOCK_DEAD) || sock_flag(ssk, SOCK_DEAD))) {
 		/* ensure later check in mptcp_worker() will dispose the msk */
-		mptcp_set_close_tout(sk, tcp_jiffies32 - (TCP_TIMEWAIT_LEN + 1));
 		sock_set_flag(sk, SOCK_DEAD);
+		mptcp_set_close_tout(sk, tcp_jiffies32 - (mptcp_close_timeout(sk) + 1));
 		lock_sock_nested(ssk, SINGLE_DEPTH_NESTING);
 		mptcp_subflow_drop_ctx(ssk);
 		goto out_release;
···
 		return false;
 
 	return time_after32(tcp_jiffies32,
-		  inet_csk(sk)->icsk_mtup.probe_timestamp + TCP_TIMEWAIT_LEN);
+		  inet_csk(sk)->icsk_mtup.probe_timestamp + mptcp_close_timeout(sk));
 }
 
 static void mptcp_check_fastclose(struct mptcp_sock *msk)
···
 		return;
 
 	close_timeout = inet_csk(sk)->icsk_mtup.probe_timestamp - tcp_jiffies32 + jiffies +
-			TCP_TIMEWAIT_LEN;
+			mptcp_close_timeout(sk);
 
 	/* the close timeout takes precedence on the fail one, and here at least one of
 	 * them is active

net/mptcp/protocol.h

···
 int mptcp_is_checksum_enabled(const struct net *net);
 int mptcp_allow_join_id0(const struct net *net);
 unsigned int mptcp_stale_loss_cnt(const struct net *net);
+unsigned int mptcp_close_timeout(const struct sock *sk);
 int mptcp_get_pm_type(const struct net *net);
 const char *mptcp_get_scheduler(const struct net *net);
 void mptcp_subflow_fully_established(struct mptcp_subflow_context *subflow,