Series comparison

The msk socket uses to different timeout to track close related
events and retransmissions. The existing helpers do not indicate
clearly which timer they actually touch, making the related code
quite confusing.

Change the existing helpers name to avoid such confusion. No
functional change intended.

Signed-off-by: Paolo Abeni <pabeni@redhat.com>
---
 net/mptcp/protocol.c | 42 +++++++++++++++++++++---------------------
 net/mptcp/protocol.h |  2 +-
 net/mptcp/subflow.c  |  2 +-
 3 files changed, 23 insertions(+), 23 deletions(-)

diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c
index XXXXXXX..XXXXXXX 100644
--- a/net/mptcp/protocol.c
+++ b/net/mptcp/protocol.c
@@ -XXX,XX +XXX,XX @@ static bool __mptcp_move_skb(struct mptcp_sock *msk, struct sock *ssk,
 	return false;
 }
 
-static void mptcp_stop_timer(struct sock *sk)
+static void mptcp_stop_rtx_timer(struct sock *sk)
 {
 	struct inet_connection_sock *icsk = inet_csk(sk);
 
@@ -XXX,XX +XXX,XX @@ static void __mptcp_flush_join_list(struct sock *sk, struct list_head *join_list
 	}
 }
 
-static bool mptcp_timer_pending(struct sock *sk)
+static bool mptcp_rtx_timer_pending(struct sock *sk)
 {
 	return timer_pending(&inet_csk(sk)->icsk_retransmit_timer);
 }
 
-static void mptcp_reset_timer(struct sock *sk)
+static void mptcp_reset_rtx_timer(struct sock *sk)
 {
 	struct inet_connection_sock *icsk = inet_csk(sk);
 	unsigned long tout;
@@ -XXX,XX +XXX,XX @@ static void __mptcp_clean_una(struct sock *sk)
 out:
 	if (snd_una == READ_ONCE(msk->snd_nxt) &&
 	    snd_una == READ_ONCE(msk->write_seq)) {
-		if (mptcp_timer_pending(sk) && !mptcp_data_fin_enabled(msk))
-			mptcp_stop_timer(sk);
+		if (mptcp_rtx_timer_pending(sk) && !mptcp_data_fin_enabled(msk))
+			mptcp_stop_rtx_timer(sk);
 	} else {
-		mptcp_reset_timer(sk);
+		mptcp_reset_rtx_timer(sk);
 	}
 }
 
@@ -XXX,XX +XXX,XX @@ void __mptcp_push_pending(struct sock *sk, unsigned int flags)
 		mptcp_push_release(ssk, &info);
 
 	/* ensure the rtx timer is running */
-	if (!mptcp_timer_pending(sk))
-		mptcp_reset_timer(sk);
+	if (!mptcp_rtx_timer_pending(sk))
+		mptcp_reset_rtx_timer(sk);
 	if (do_check_data_fin)
 		mptcp_check_send_data_fin(sk);
 }
@@ -XXX,XX +XXX,XX @@ static void __mptcp_subflow_push_pending(struct sock *sk, struct sock *ssk, bool
 	if (copied) {
 		tcp_push(ssk, 0, info.mss_now, tcp_sk(ssk)->nonagle,
 			 info.size_goal);
-		if (!mptcp_timer_pending(sk))
-			mptcp_reset_timer(sk);
+		if (!mptcp_rtx_timer_pending(sk))
+			mptcp_reset_rtx_timer(sk);
 
 		if (msk->snd_data_fin_enable &&
 		    msk->snd_nxt + 1 == msk->write_seq)
@@ -XXX,XX +XXX,XX @@ static void mptcp_retransmit_timer(struct timer_list *t)
 	sock_put(sk);
 }
 
-static void mptcp_timeout_timer(struct timer_list *t)
+static void mptcp_tout_timer(struct timer_list *t)
 {
 	struct sock *sk = from_timer(sk, t, sk_timer);
 
@@ -XXX,XX +XXX,XX @@ static void __mptcp_retrans(struct sock *sk)
 reset_timer:
 	mptcp_check_and_set_pending(sk);
 
-	if (!mptcp_timer_pending(sk))
-		mptcp_reset_timer(sk);
+	if (!mptcp_rtx_timer_pending(sk))
+		mptcp_reset_rtx_timer(sk);
 }
 
 /* schedule the timeout timer for the relevant event: either close timeout
  * or mp_fail timeout. The close timeout takes precedence on the mp_fail one
  */
-void mptcp_reset_timeout(struct mptcp_sock *msk, unsigned long fail_tout)
+void mptcp_reset_tout_timer(struct mptcp_sock *msk, unsigned long fail_tout)
 {
 	struct sock *sk = (struct sock *)msk;
 	unsigned long timeout, close_timeout;
@@ -XXX,XX +XXX,XX @@ static void mptcp_mp_fail_no_response(struct mptcp_sock *msk)
 	WRITE_ONCE(mptcp_subflow_ctx(ssk)->fail_tout, 0);
 	unlock_sock_fast(ssk, slow);
 
-	mptcp_reset_timeout(msk, 0);
+	mptcp_reset_tout_timer(msk, 0);
 }
 
 static void mptcp_do_fastclose(struct sock *sk)
@@ -XXX,XX +XXX,XX @@ static void __mptcp_init_sock(struct sock *sk)
 
 	/* re-use the csk retrans timer for MPTCP-level retrans */
 	timer_setup(&msk->sk.icsk_retransmit_timer, mptcp_retransmit_timer, 0);
-	timer_setup(&sk->sk_timer, mptcp_timeout_timer, 0);
+	timer_setup(&sk->sk_timer, mptcp_tout_timer, 0);
 }
 
 static void mptcp_ca_reset(struct sock *sk)
@@ -XXX,XX +XXX,XX @@ void mptcp_subflow_shutdown(struct sock *sk, struct sock *ssk, int how)
 		} else {
 			pr_debug("Sending DATA_FIN on subflow %p", ssk);
 			tcp_send_ack(ssk);
-			if (!mptcp_timer_pending(sk))
-				mptcp_reset_timer(sk);
+			if (!mptcp_rtx_timer_pending(sk))
+				mptcp_reset_rtx_timer(sk);
 		}
 		break;
 	}
@@ -XXX,XX +XXX,XX @@ static void __mptcp_destroy_sock(struct sock *sk)
 
 	might_sleep();
 
-	mptcp_stop_timer(sk);
+	mptcp_stop_rtx_timer(sk);
 	sk_stop_timer(sk, &sk->sk_timer);
 	msk->pm.status = 0;
 	mptcp_release_sched(msk);
@@ -XXX,XX +XXX,XX @@ bool __mptcp_close(struct sock *sk, long timeout)
 		__mptcp_destroy_sock(sk);
 		do_cancel_work = true;
 	} else {
-		mptcp_reset_timeout(msk, 0);
+		mptcp_reset_tout_timer(msk, 0);
 	}
 
 	return do_cancel_work;
@@ -XXX,XX +XXX,XX @@ static int mptcp_disconnect(struct sock *sk, int flags)
 	mptcp_check_listen_stop(sk);
 	inet_sk_state_store(sk, TCP_CLOSE);
 
-	mptcp_stop_timer(sk);
+	mptcp_stop_rtx_timer(sk);
 	sk_stop_timer(sk, &sk->sk_timer);
 
 	if (msk->token)
diff --git a/net/mptcp/protocol.h b/net/mptcp/protocol.h
index XXXXXXX..XXXXXXX 100644
--- a/net/mptcp/protocol.h
+++ b/net/mptcp/protocol.h
@@ -XXX,XX +XXX,XX @@ void mptcp_get_options(const struct sk_buff *skb,
 
 void mptcp_finish_connect(struct sock *sk);
 void __mptcp_set_connected(struct sock *sk);
-void mptcp_reset_timeout(struct mptcp_sock *msk, unsigned long fail_tout);
+void mptcp_reset_tout_timer(struct mptcp_sock *msk, unsigned long fail_tout);
 static inline bool mptcp_is_fully_established(struct sock *sk)
 {
 	return inet_sk_state_load(sk) == TCP_ESTABLISHED &&
diff --git a/net/mptcp/subflow.c b/net/mptcp/subflow.c
index XXXXXXX..XXXXXXX 100644
--- a/net/mptcp/subflow.c
+++ b/net/mptcp/subflow.c
@@ -XXX,XX +XXX,XX @@ static void mptcp_subflow_fail(struct mptcp_sock *msk, struct sock *ssk)
 	WRITE_ONCE(subflow->fail_tout, fail_tout);
 	tcp_send_ack(ssk);
 
-	mptcp_reset_timeout(msk, subflow->fail_tout);
+	mptcp_reset_tout_timer(msk, subflow->fail_tout);
 }
 
 static bool subflow_check_data_avail(struct sock *ssk)
-- 
2.41.0

According to RFC 8684 section 3.3:

A connection is not closed unless [...] or an implementation-specific
  connection-level send timeout.

Currently the MPTCP protocol does not implement such timeout, and
connection timing-out at the TCP-level never move to close state.

Introduces a catch-up condition at subflow close time to move the
MPTCP socket to close, too.

That additionally allow removing similar existing inside the worker.

Finally, allow some additional timeout for plain ESTABLISHED mptcp
sockets, as the protocol allows creating new subflows even at that
point and making the connection functional again.

Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/430
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
---
v2 -> v3:
 - fix integer overflow in mptcp_close_tout_expired()
v1 -> v2:
 - really set the close timeout as need.
---
 net/mptcp/protocol.c | 84 +++++++++++++++++++++-----------------------
 net/mptcp/protocol.h | 20 +++++++++++
 net/mptcp/subflow.c  |  1 +
 3 files changed, 62 insertions(+), 43 deletions(-)

diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c
index XXXXXXX..XXXXXXX 100644
--- a/net/mptcp/protocol.c
+++ b/net/mptcp/protocol.c
@@ -XXX,XX +XXX,XX @@ static bool __mptcp_finish_join(struct mptcp_sock *msk, struct sock *ssk)
 	mptcp_subflow_ctx(ssk)->subflow_id = msk->subflow_id++;
 	mptcp_sockopt_sync_locked(msk, ssk);
 	mptcp_subflow_joined(msk, ssk);
+	mptcp_stop_tout_timer(sk);
 	return true;
 }
 
@@ -XXX,XX +XXX,XX @@ static void __mptcp_close_ssk(struct sock *sk, struct sock *ssk,
 	bool dispose_it, need_push = false;
 
 	/* If the first subflow moved to a close state before accept, e.g. due
-	 * to an incoming reset, mptcp either:
-	 * - if either the subflow or the msk are dead, destroy the context
-	 *   (the subflow socket is deleted by inet_child_forget) and the msk
-	 * - otherwise do nothing at the moment and take action at accept and/or
-	 *   listener shutdown - user-space must be able to accept() the closed
-	 *   socket.
+	 * to an incoming reset or listener shoutdown, the subflow socket is
+	 * already deleted by inet_child_forget() the mptcp socket can't survive
+	 * too.
 	 */
-	if (msk->in_accept_queue && msk->first == ssk) {
-		if (!sock_flag(sk, SOCK_DEAD) && !sock_flag(ssk, SOCK_DEAD))
-			return;
-
+	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 */
+		inet_csk(sk)->icsk_mtup.probe_timestamp = tcp_jiffies32 -
+							  TCP_TIMEWAIT_LEN;
 		sock_set_flag(sk, SOCK_DEAD);
 		lock_sock_nested(ssk, SINGLE_DEPTH_NESTING);
 		mptcp_subflow_drop_ctx(ssk);
@@ -XXX,XX +XXX,XX @@ static void __mptcp_close_ssk(struct sock *sk, struct sock *ssk,
 out:
 	if (need_push)
 		__mptcp_push_pending(sk, 0);
+
+	/* Catch every 'all subflows closed' scenario, including peers silently
+	 * closing them, e.g. due to timeout.
+	 * For established sockets, allow an additional timeout before closing,
+	 * as the protocol can still create more subflows.
+	 */
+	if (list_is_singular(&msk->conn_list) && msk->first &&
+	    inet_sk_state_load(msk->first) == TCP_CLOSE) {
+		if (sk->sk_state != TCP_ESTABLISHED ||
+		    msk->in_accept_queue || sock_flag(sk, SOCK_DEAD)) {
+			inet_sk_state_store(sk, TCP_CLOSE);
+			mptcp_close_wake_up(sk);
+		} else {
+			mptcp_start_tout_timer(sk);
+		}
+	}
 }
 
 void mptcp_close_ssk(struct sock *sk, struct sock *ssk,
@@ -XXX,XX +XXX,XX @@ static void __mptcp_close_subflow(struct sock *sk)
 
 }
 
-static bool mptcp_should_close(const struct sock *sk)
+static bool mptcp_close_tout_expired(const struct sock *sk)
 {
-	s32 delta = tcp_jiffies32 - inet_csk(sk)->icsk_mtup.probe_timestamp;
-	struct mptcp_subflow_context *subflow;
-
-	if (delta >= TCP_TIMEWAIT_LEN || mptcp_sk(sk)->in_accept_queue)
-		return true;
+	if (!inet_csk(sk)->icsk_mtup.probe_timestamp ||
+	    sk->sk_state == TCP_CLOSE)
+		return false;
 
-	/* if all subflows are in closed status don't bother with additional
-	 * timeout
-	 */
-	mptcp_for_each_subflow(mptcp_sk(sk), subflow) {
-		if (inet_sk_state_load(mptcp_subflow_tcp_sock(subflow)) !=
-		    TCP_CLOSE)
-			return false;
-	}
-	return true;
+	return !time_before32(tcp_jiffies32,
+		  inet_csk(sk)->icsk_mtup.probe_timestamp + TCP_TIMEWAIT_LEN);
 }
 
 static void mptcp_check_fastclose(struct mptcp_sock *msk)
@@ -XXX,XX +XXX,XX @@ void mptcp_reset_tout_timer(struct mptcp_sock *msk, unsigned long fail_tout)
 	struct sock *sk = (struct sock *)msk;
 	unsigned long timeout, close_timeout;
 
-	if (!fail_tout && !sock_flag(sk, SOCK_DEAD))
+	if (!fail_tout && !inet_csk(sk)->icsk_mtup.probe_timestamp)
 		return;
 
 	close_timeout = inet_csk(sk)->icsk_mtup.probe_timestamp - tcp_jiffies32 + jiffies + TCP_TIMEWAIT_LEN;
@@ -XXX,XX +XXX,XX @@ void mptcp_reset_tout_timer(struct mptcp_sock *msk, unsigned long fail_tout)
 	/* the close timeout takes precedence on the fail one, and here at least one of
 	 * them is active
 	 */
-	timeout = sock_flag(sk, SOCK_DEAD) ? close_timeout : fail_tout;
+	timeout = inet_csk(sk)->icsk_mtup.probe_timestamp ? close_timeout : fail_tout;
 
 	sk_reset_timer(sk, &sk->sk_timer, timeout);
 }
@@ -XXX,XX +XXX,XX @@ static void mptcp_mp_fail_no_response(struct mptcp_sock *msk)
 	mptcp_subflow_reset(ssk);
 	WRITE_ONCE(mptcp_subflow_ctx(ssk)->fail_tout, 0);
 	unlock_sock_fast(ssk, slow);
-
-	mptcp_reset_tout_timer(msk, 0);
 }
 
 static void mptcp_do_fastclose(struct sock *sk)
@@ -XXX,XX +XXX,XX @@ static void mptcp_worker(struct work_struct *work)
 	if (test_and_clear_bit(MPTCP_WORK_CLOSE_SUBFLOW, &msk->flags))
 		__mptcp_close_subflow(sk);
 
-	/* There is no point in keeping around an orphaned sk timedout or
-	 * closed, but we need the msk around to reply to incoming DATA_FIN,
-	 * even if it is orphaned and in FIN_WAIT2 state
-	 */
-	if (sock_flag(sk, SOCK_DEAD)) {
-		if (mptcp_should_close(sk))
-			mptcp_do_fastclose(sk);
+	if (mptcp_close_tout_expired(sk)) {
+		mptcp_do_fastclose(sk);
+		mptcp_close_wake_up(sk);
+	}
 
-		if (sk->sk_state == TCP_CLOSE) {
-			__mptcp_destroy_sock(sk);
-			goto unlock;
-		}
+	if (sock_flag(sk, SOCK_DEAD) && sk->sk_state == TCP_CLOSE) {
+		__mptcp_destroy_sock(sk);
+		goto unlock;
 	}
 
 	if (test_and_clear_bit(MPTCP_WORK_RTX, &msk->flags))
@@ -XXX,XX +XXX,XX @@ bool __mptcp_close(struct sock *sk, long timeout)
 
 cleanup:
 	/* orphan all the subflows */
-	inet_csk(sk)->icsk_mtup.probe_timestamp = tcp_jiffies32;
 	mptcp_for_each_subflow(msk, subflow) {
 		struct sock *ssk = mptcp_subflow_tcp_sock(subflow);
 		bool slow = lock_sock_fast_nested(ssk);
@@ -XXX,XX +XXX,XX @@ bool __mptcp_close(struct sock *sk, long timeout)
 		__mptcp_destroy_sock(sk);
 		do_cancel_work = true;
 	} else {
-		mptcp_reset_tout_timer(msk, 0);
+		mptcp_start_tout_timer(sk);
 	}
 
 	return do_cancel_work;
@@ -XXX,XX +XXX,XX @@ static int mptcp_disconnect(struct sock *sk, int flags)
 	inet_sk_state_store(sk, TCP_CLOSE);
 
 	mptcp_stop_rtx_timer(sk);
-	sk_stop_timer(sk, &sk->sk_timer);
+	mptcp_stop_tout_timer(sk);
 
 	if (msk->token)
 		mptcp_event(MPTCP_EVENT_CLOSED, msk, NULL, GFP_KERNEL);
diff --git a/net/mptcp/protocol.h b/net/mptcp/protocol.h
index XXXXXXX..XXXXXXX 100644
--- a/net/mptcp/protocol.h
+++ b/net/mptcp/protocol.h
@@ -XXX,XX +XXX,XX @@ void mptcp_get_options(const struct sk_buff *skb,
 void mptcp_finish_connect(struct sock *sk);
 void __mptcp_set_connected(struct sock *sk);
 void mptcp_reset_tout_timer(struct mptcp_sock *msk, unsigned long fail_tout);
+
+static inline void mptcp_stop_tout_timer(struct sock *sk)
+{
+	if (!inet_csk(sk)->icsk_mtup.probe_timestamp)
+		return;
+
+	sk_stop_timer(sk, &sk->sk_timer);
+	inet_csk(sk)->icsk_mtup.probe_timestamp = 0;
+}
+
+static inline void mptcp_start_tout_timer(struct sock *sk)
+{
+	/* avoid 0 timestamp, at that means no close timeout */
+	inet_csk(sk)->icsk_mtup.probe_timestamp = tcp_jiffies32;
+	if (!inet_csk(sk)->icsk_mtup.probe_timestamp)
+		inet_csk(sk)->icsk_mtup.probe_timestamp = 1;
+
+	mptcp_reset_tout_timer(mptcp_sk(sk), 0);
+}
+
 static inline bool mptcp_is_fully_established(struct sock *sk)
 {
 	return inet_sk_state_load(sk) == TCP_ESTABLISHED &&
diff --git a/net/mptcp/subflow.c b/net/mptcp/subflow.c
index XXXXXXX..XXXXXXX 100644
--- a/net/mptcp/subflow.c
+++ b/net/mptcp/subflow.c
@@ -XXX,XX +XXX,XX @@ int __mptcp_subflow_connect(struct sock *sk, const struct mptcp_addr_info *loc,
 	mptcp_sock_graft(ssk, sk->sk_socket);
 	iput(SOCK_INODE(sf));
 	WRITE_ONCE(msk->allow_infinite_fallback, false);
+	mptcp_stop_tout_timer(sk);
 	return 0;
 
 failed_unlink:
-- 
2.41.0

Change the existing helpers name to avoid such confusion. No
functional change intended.

According to RFC 8684 section 3.3:

A connection is not closed unless [...] or an implementation-specific
  connection-level send timeout.

Currently the MPTCP protocol does not implement such timeout, and
connection timing-out at the TCP-level never move to close state.

Introduces a catch-up condition at subflow close time to move the
MPTCP socket to close, too.

That additionally allow removing similar existing inside the worker.

Finally, allow some additional timeout for plain ESTABLISHED mptcp
sockets, as the protocol allows creating new subflows even at that
point and making the connection functional again.

The issues is actually present since the beginning, but it basically
impossible to solve without a long chain of functional pre-requisites
topped by the blamed commit below.

Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/430
Fixes: bbd49d114d57 ("mptcp: consolidate transition to TCP_CLOSE in mptcp_do_fastclose()")
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
---
v4 -> v5:
 - let the tout always expire for dead socks (Mat)
 - fix comment type (Mat)
 - targeting net, added fixes tag

v3 -> v4:
 - fix off-by-one in timer update/check (sigh, trival math is too
   hard:/)

v2 -> v3:
 - fix integer overflow in mptcp_close_tout_expired()
v1 -> v2:
 - really set the close timeout as need.
---
 net/mptcp/protocol.c | 86 ++++++++++++++++++++++----------------------
 net/mptcp/protocol.h | 24 +++++++++++++
 net/mptcp/subflow.c  |  1 +
 3 files changed, 67 insertions(+), 44 deletions(-)

diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c
index XXXXXXX..XXXXXXX 100644
--- a/net/mptcp/protocol.c
+++ b/net/mptcp/protocol.c
@@ -XXX,XX +XXX,XX @@ static bool __mptcp_finish_join(struct mptcp_sock *msk, struct sock *ssk)
 	mptcp_subflow_ctx(ssk)->subflow_id = msk->subflow_id++;
 	mptcp_sockopt_sync_locked(msk, ssk);
 	mptcp_subflow_joined(msk, ssk);
+	mptcp_stop_tout_timer(sk);
 	return true;
 }
 
@@ -XXX,XX +XXX,XX @@ static void __mptcp_close_ssk(struct sock *sk, struct sock *ssk,
 	bool dispose_it, need_push = false;
 
 	/* If the first subflow moved to a close state before accept, e.g. due
-	 * to an incoming reset, mptcp either:
-	 * - if either the subflow or the msk are dead, destroy the context
-	 *   (the subflow socket is deleted by inet_child_forget) and the msk
-	 * - otherwise do nothing at the moment and take action at accept and/or
-	 *   listener shutdown - user-space must be able to accept() the closed
-	 *   socket.
+	 * to an incoming reset or listener shutdown, the subflow socket is
+	 * already deleted by inet_child_forget() and the mptcp socket can't
+	 * survive too.
 	 */
-	if (msk->in_accept_queue && msk->first == ssk) {
-		if (!sock_flag(sk, SOCK_DEAD) && !sock_flag(ssk, SOCK_DEAD))
-			return;
-
+	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);
 		lock_sock_nested(ssk, SINGLE_DEPTH_NESTING);
 		mptcp_subflow_drop_ctx(ssk);
@@ -XXX,XX +XXX,XX @@ static void __mptcp_close_ssk(struct sock *sk, struct sock *ssk,
 out:
 	if (need_push)
 		__mptcp_push_pending(sk, 0);
+
+	/* Catch every 'all subflows closed' scenario, including peers silently
+	 * closing them, e.g. due to timeout.
+	 * For established sockets, allow an additional timeout before closing,
+	 * as the protocol can still create more subflows.
+	 */
+	if (list_is_singular(&msk->conn_list) && msk->first &&
+	    inet_sk_state_load(msk->first) == TCP_CLOSE) {
+		if (sk->sk_state != TCP_ESTABLISHED ||
+		    msk->in_accept_queue || sock_flag(sk, SOCK_DEAD)) {
+			inet_sk_state_store(sk, TCP_CLOSE);
+			mptcp_close_wake_up(sk);
+		} else {
+			mptcp_start_tout_timer(sk);
+		}
+	}
 }
 
 void mptcp_close_ssk(struct sock *sk, struct sock *ssk,
@@ -XXX,XX +XXX,XX @@ static void __mptcp_close_subflow(struct sock *sk)
 
 }
 
-static bool mptcp_should_close(const struct sock *sk)
+static bool mptcp_close_tout_expired(const struct sock *sk)
 {
-	s32 delta = tcp_jiffies32 - inet_csk(sk)->icsk_mtup.probe_timestamp;
-	struct mptcp_subflow_context *subflow;
-
-	if (delta >= TCP_TIMEWAIT_LEN || mptcp_sk(sk)->in_accept_queue)
-		return true;
+	if (!inet_csk(sk)->icsk_mtup.probe_timestamp ||
+	    sk->sk_state == TCP_CLOSE)
+		return false;
 
-	/* if all subflows are in closed status don't bother with additional
-	 * timeout
-	 */
-	mptcp_for_each_subflow(mptcp_sk(sk), subflow) {
-		if (inet_sk_state_load(mptcp_subflow_tcp_sock(subflow)) !=
-		    TCP_CLOSE)
-			return false;
-	}
-	return true;
+	return time_after32(tcp_jiffies32,
+		  inet_csk(sk)->icsk_mtup.probe_timestamp + TCP_TIMEWAIT_LEN);
 }
 
 static void mptcp_check_fastclose(struct mptcp_sock *msk)
@@ -XXX,XX +XXX,XX @@ void mptcp_reset_tout_timer(struct mptcp_sock *msk, unsigned long fail_tout)
 	struct sock *sk = (struct sock *)msk;
 	unsigned long timeout, close_timeout;
 
-	if (!fail_tout && !sock_flag(sk, SOCK_DEAD))
+	if (!fail_tout && !inet_csk(sk)->icsk_mtup.probe_timestamp)
 		return;
 
-	close_timeout = inet_csk(sk)->icsk_mtup.probe_timestamp - tcp_jiffies32 + jiffies + TCP_TIMEWAIT_LEN;
+	close_timeout = inet_csk(sk)->icsk_mtup.probe_timestamp - tcp_jiffies32 + jiffies +
+			TCP_TIMEWAIT_LEN;
 
 	/* the close timeout takes precedence on the fail one, and here at least one of
 	 * them is active
 	 */
-	timeout = sock_flag(sk, SOCK_DEAD) ? close_timeout : fail_tout;
+	timeout = inet_csk(sk)->icsk_mtup.probe_timestamp ? close_timeout : fail_tout;
 
 	sk_reset_timer(sk, &sk->sk_timer, timeout);
 }
@@ -XXX,XX +XXX,XX @@ static void mptcp_mp_fail_no_response(struct mptcp_sock *msk)
 	mptcp_subflow_reset(ssk);
 	WRITE_ONCE(mptcp_subflow_ctx(ssk)->fail_tout, 0);
 	unlock_sock_fast(ssk, slow);
-
-	mptcp_reset_tout_timer(msk, 0);
 }
 
 static void mptcp_do_fastclose(struct sock *sk)
@@ -XXX,XX +XXX,XX @@ static void mptcp_worker(struct work_struct *work)
 	if (test_and_clear_bit(MPTCP_WORK_CLOSE_SUBFLOW, &msk->flags))
 		__mptcp_close_subflow(sk);
 
-	/* There is no point in keeping around an orphaned sk timedout or
-	 * closed, but we need the msk around to reply to incoming DATA_FIN,
-	 * even if it is orphaned and in FIN_WAIT2 state
-	 */
-	if (sock_flag(sk, SOCK_DEAD)) {
-		if (mptcp_should_close(sk))
-			mptcp_do_fastclose(sk);
+	if (mptcp_close_tout_expired(sk)) {
+		mptcp_do_fastclose(sk);
+		mptcp_close_wake_up(sk);
+	}
 
-		if (sk->sk_state == TCP_CLOSE) {
-			__mptcp_destroy_sock(sk);
-			goto unlock;
-		}
+	if (sock_flag(sk, SOCK_DEAD) && sk->sk_state == TCP_CLOSE) {
+		__mptcp_destroy_sock(sk);
+		goto unlock;
 	}
 
 	if (test_and_clear_bit(MPTCP_WORK_RTX, &msk->flags))
@@ -XXX,XX +XXX,XX @@ bool __mptcp_close(struct sock *sk, long timeout)
 
 cleanup:
 	/* orphan all the subflows */
-	inet_csk(sk)->icsk_mtup.probe_timestamp = tcp_jiffies32;
 	mptcp_for_each_subflow(msk, subflow) {
 		struct sock *ssk = mptcp_subflow_tcp_sock(subflow);
 		bool slow = lock_sock_fast_nested(ssk);
@@ -XXX,XX +XXX,XX @@ bool __mptcp_close(struct sock *sk, long timeout)
 		__mptcp_destroy_sock(sk);
 		do_cancel_work = true;
 	} else {
-		mptcp_reset_tout_timer(msk, 0);
+		mptcp_start_tout_timer(sk);
 	}
 
 	return do_cancel_work;
@@ -XXX,XX +XXX,XX @@ static int mptcp_disconnect(struct sock *sk, int flags)
 	inet_sk_state_store(sk, TCP_CLOSE);
 
 	mptcp_stop_rtx_timer(sk);
-	sk_stop_timer(sk, &sk->sk_timer);
+	mptcp_stop_tout_timer(sk);
 
 	if (msk->token)
 		mptcp_event(MPTCP_EVENT_CLOSED, msk, NULL, GFP_KERNEL);
diff --git a/net/mptcp/protocol.h b/net/mptcp/protocol.h
index XXXXXXX..XXXXXXX 100644
--- a/net/mptcp/protocol.h
+++ b/net/mptcp/protocol.h
@@ -XXX,XX +XXX,XX @@ void mptcp_get_options(const struct sk_buff *skb,
 void mptcp_finish_connect(struct sock *sk);
 void __mptcp_set_connected(struct sock *sk);
 void mptcp_reset_tout_timer(struct mptcp_sock *msk, unsigned long fail_tout);
+
+static inline void mptcp_stop_tout_timer(struct sock *sk)
+{
+	if (!inet_csk(sk)->icsk_mtup.probe_timestamp)
+		return;
+
+	sk_stop_timer(sk, &sk->sk_timer);
+	inet_csk(sk)->icsk_mtup.probe_timestamp = 0;
+}
+
+static inline void mptcp_set_close_tout(struct sock *sk, unsigned long tout)
+{
+	/* avoid 0 timestamp, at that means no close timeout */
+	inet_csk(sk)->icsk_mtup.probe_timestamp = tout;
+	if (!inet_csk(sk)->icsk_mtup.probe_timestamp)
+		inet_csk(sk)->icsk_mtup.probe_timestamp = 1;
+}
+
+static inline void mptcp_start_tout_timer(struct sock *sk)
+{
+	mptcp_set_close_tout(sk, tcp_jiffies32);
+	mptcp_reset_tout_timer(mptcp_sk(sk), 0);
+}
+
 static inline bool mptcp_is_fully_established(struct sock *sk)
 {
 	return inet_sk_state_load(sk) == TCP_ESTABLISHED &&
diff --git a/net/mptcp/subflow.c b/net/mptcp/subflow.c
index XXXXXXX..XXXXXXX 100644
--- a/net/mptcp/subflow.c
+++ b/net/mptcp/subflow.c
@@ -XXX,XX +XXX,XX @@ int __mptcp_subflow_connect(struct sock *sk, const struct mptcp_addr_info *loc,
 	mptcp_sock_graft(ssk, sk->sk_socket);
 	iput(SOCK_INODE(sf));
 	WRITE_ONCE(msk->allow_infinite_fallback, false);
+	mptcp_stop_tout_timer(sk);
 	return 0;
 
 failed_unlink:
-- 
2.41.0