[PATCH mptcp-net v2 00/11] mptcp: avoid dup events + error + misc.

[PATCH mptcp-net v2 00/11] mptcp: avoid dup events + error + misc.

[Matthieu Baerts (NGI0)]

There are two fixes with their tests:

- a subflow closed event can be visible multiple times, see issue #603

- the subflow closed event didn't contain the error as expected.

The 4 first patches are for -net, the rest are small fixes and
improvements for net-next. Some tests should be quicker now.

Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
---
Changes in v2:
- Small updates on patches 1, 2, 3 and 5 (Geliang)
- Add Geliang RvB tags.
- Link to v1: https://patch.msgid.link/20251226-mptcp-issue-603-v1-0-bb30e331b839@kernel.org

---
Matthieu Baerts (NGI0) (11):
      mptcp: avoid dup SUB_CLOSED events after disconnect
      selftests: mptcp: check no dup close events after error
      mptcp: only reset subflow errors when propagated
      selftests: mptcp: check subflow errors in close events
      selftests: mptcp: join: wait for estab event instead of MPJ
      selftests: mptcp: join: fix wait_mpj helper
      selftests: mptcp: join: userspace: wait for new events
      selftests: mptcp: join chk_stale_nr: avoid dup stats
      selftests: mptcp: join: avoid declaring i if not used
      selftests: mptcp: connect: fix maybe-uninitialize warn
      selftests: mptcp: connect cleanup TFO setup

 net/mptcp/protocol.c                              |  13 +-
 tools/testing/selftests/net/mptcp/mptcp_connect.c |  19 ++-
 tools/testing/selftests/net/mptcp/mptcp_join.sh   | 149 +++++++++++++++-------
 3 files changed, 121 insertions(+), 60 deletions(-)
---
base-commit: b8bb952910af48167681524c5b217996c16382c1
change-id: 20251222-mptcp-issue-603-d85d967f7c0a

Best regards,
-- 
Matthieu Baerts (NGI0) <matttbe@kernel.org>

[PATCH mptcp-net v2 01/11] mptcp: avoid dup SUB_CLOSED events after disconnect

[Matthieu Baerts (NGI0)]

In case of subflow disconnect(), which can also happen with the first
subflow in case of errors like timeout or reset, mptcp_subflow_ctx_reset
will reset most fields from the mptcp_subflow_context structure,
including close_event_done. Then, when another subflow is closed, yet
another SUB_CLOSED event for the disconnected initial subflow is sent.
Because of the previous reset, there are no source address and
destination port.

A solution is then to also check the subflow's local id: it shouldn't be
negative anyway.

Another solution would be not to reset subflow->close_event_done at
disconnect time, but when reused. But then, probably the whole reset
could be done when being reused. Let's not change this logic, similar
to TCP with tcp_disconnect().

Fixes: d82809b6c5f2 ("mptcp: avoid duplicated SUB_CLOSED events")
Reported-by: Marco Angaroni <marco.angaroni@italtel.com>
Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/603
Reviewed-by: Geliang Tang <geliang@kernel.org>
Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
---
v2: rework comment (Geliang)
---
 net/mptcp/protocol.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c
index 900f26e21acd..335a1723c1dc 100644
--- a/net/mptcp/protocol.c
+++ b/net/mptcp/protocol.c
@@ -2622,8 +2622,8 @@ void mptcp_close_ssk(struct sock *sk, struct sock *ssk,
 	struct mptcp_sock *msk = mptcp_sk(sk);
 	struct sk_buff *skb;
 
-	/* The first subflow can already be closed and still in the list */
-	if (subflow->close_event_done)
+	/* The first subflow can already be closed or disconnected */
+	if (subflow->close_event_done || READ_ONCE(subflow->local_id) < 0)
 		return;
 
 	subflow->close_event_done = true;

-- 
2.51.0

[PATCH mptcp-net v2 02/11] selftests: mptcp: check no dup close events after error

[Matthieu Baerts (NGI0)]

This validates the previous commit: subflow closed events are re-sent
with less info when the initial subflow is disconnected after an error
and each time a subflow is closed after that.

In this new test, the userspace PM is involved because that's how it was
discovered, but it is not specific to it. The initial subflow is
terminated with a RESET, and that will cause the subflow disconnect.
Then, a new subflow is initiated, but also got rejected, which cause a
second subflow closed event, but not a third one.

While at it, in case of failure to get the expected amount of events,
the events are printed.

The 'Fixes' tag here below is the same as the one from the previous
commit: this patch here is not fixing anything wrong in the selftests,
but it validates the previous fix for an issue introduced by this commit
ID.

Fixes: d82809b6c5f2 ("mptcp: avoid duplicated SUB_CLOSED events")
Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
---
v2: avoid too long lines with iptables commands (Geliang)
---
 tools/testing/selftests/net/mptcp/mptcp_join.sh | 51 +++++++++++++++++++++++++
 1 file changed, 51 insertions(+)

diff --git a/tools/testing/selftests/net/mptcp/mptcp_join.sh b/tools/testing/selftests/net/mptcp/mptcp_join.sh
index f1a0d812e642..35ff769f440f 100755
--- a/tools/testing/selftests/net/mptcp/mptcp_join.sh
+++ b/tools/testing/selftests/net/mptcp/mptcp_join.sh
@@ -3871,11 +3871,32 @@ chk_evt_nr()
 	count=$(grep -cw "type:${evt}" "${evts}")
 	if [ "${count}" != "${exp}" ]; then
 		fail_test "got ${count} events, expected ${exp}"
+		cat "${evts}"
 	else
 		print_ok
 	fi
 }
 
+# $1: ns ; $2: event type ; $3: expected count
+wait_event()
+{
+	local ns="${1}"
+	local evt_name="${2}"
+	local exp="${3}"
+
+	local evt="${!evt_name}"
+	local evts="${evts_ns1}"
+	local count
+
+	[ "${ns}" == "ns2" ] && evts="${evts_ns2}"
+
+	for _ in $(seq 100); do
+		count=$(grep -cw "type:${evt}" "${evts}")
+		[ "${count}" -ge "${exp}" ] && break
+		sleep 0.1
+	done
+}
+
 userspace_tests()
 {
 	# userspace pm type prevents add_addr
@@ -4084,6 +4105,36 @@ userspace_tests()
 		kill_events_pids
 		mptcp_lib_kill_group_wait $tests_pid
 	fi
+
+	# userspace pm no duplicated spurious close events after an error
+	if reset_with_events "userspace pm no dup close events after error" &&
+	   continue_if mptcp_lib_has_file '/proc/sys/net/mptcp/pm_type'; then
+		set_userspace_pm $ns2
+		pm_nl_set_limits $ns1 0 2
+		{ timeout_test=120 test_linkfail=128 speed=slow \
+			run_tests $ns1 $ns2 10.0.1.1 & } 2>/dev/null
+		local tests_pid=$!
+		wait_event ns2 MPTCP_LIB_EVENT_ESTABLISHED 1
+		userspace_pm_add_sf $ns2 10.0.3.2 20
+		chk_mptcp_info subflows 1 subflows 1
+		chk_subflows_total 2 2
+
+		# force quick loss
+		ip netns exec $ns2 sysctl -q net.ipv4.tcp_syn_retries=1
+		if ip netns exec "${ns1}" ${iptables} -A INPUT -s "10.0.1.2" \
+		      -p tcp --tcp-option 30 -j REJECT --reject-with tcp-reset &&
+		   ip netns exec "${ns2}" ${iptables} -A INPUT -d "10.0.1.2" \
+		      -p tcp --tcp-option 30 -j REJECT --reject-with tcp-reset; then
+			wait_event ns2 MPTCP_LIB_EVENT_SUB_CLOSED 1
+			wait_event ns1 MPTCP_LIB_EVENT_SUB_CLOSED 1
+			chk_subflows_total 1 1
+			userspace_pm_add_sf $ns2 10.0.1.2 0
+			wait_event ns2 MPTCP_LIB_EVENT_SUB_CLOSED 2
+			chk_evt_nr ns2 MPTCP_LIB_EVENT_SUB_CLOSED 2
+		fi
+		kill_events_pids
+		mptcp_lib_kill_group_wait $tests_pid
+	fi
 }
 
 endpoint_tests()

-- 
2.51.0

[PATCH mptcp-net v2 03/11] mptcp: only reset subflow errors when propagated

[Matthieu Baerts (NGI0)]

Some subflow socket errors need to be reported to the MPTCP socket: the
initial subflow connect (MP_CAPABLE), and the ones from the fallback
sockets. The others are not propagated.

The issue is that sock_error() was used to retrieve the error, which was
also resetting the sk_err field. Because of that, when notifying the
userspace about subflow close events later on from the MPTCP worker, the
ssk->sk_err field was always 0.

Now, the error (sk_err) is only reset when propagating it to the msk.

Fixes: 15cc10453398 ("mptcp: deliver ssk errors to msk")
Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
---
Note: I guess we could also duplicate the error in mptcp_subflow_context
  struct before scheduling the worker, and use this field in
  mptcp_event_put_token_and_ssk(), but I don't see why we should always
  reset ssk->sk_err in __mptcp_subflow_error_report() in all cases.
v2: keep 'err' variable (Geliang)
---
 net/mptcp/protocol.c | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c
index 335a1723c1dc..1b87b3b0d089 100644
--- a/net/mptcp/protocol.c
+++ b/net/mptcp/protocol.c
@@ -817,11 +817,8 @@ static bool __mptcp_ofo_queue(struct mptcp_sock *msk)
 
 static bool __mptcp_subflow_error_report(struct sock *sk, struct sock *ssk)
 {
-	int err = sock_error(ssk);
 	int ssk_state;
-
-	if (!err)
-		return false;
+	int err;
 
 	/* only propagate errors on fallen-back sockets or
 	 * on MPC connect
@@ -829,6 +826,10 @@ static bool __mptcp_subflow_error_report(struct sock *sk, struct sock *ssk)
 	if (sk->sk_state != TCP_SYN_SENT && !__mptcp_check_fallback(mptcp_sk(sk)))
 		return false;
 
+	err = sock_error(ssk);
+	if (!err)
+		return false;
+
 	/* We need to propagate only transition to CLOSE state.
 	 * Orphaned socket will see such state change via
 	 * subflow_sched_work_if_closed() and that path will properly

-- 
2.51.0

[PATCH mptcp-net v2 04/11] selftests: mptcp: check subflow errors in close events

[Matthieu Baerts (NGI0)]

This validates the previous commit: subflow closed events should contain
an error field when a subflow got closed with an error, e.g. reset or
timeout.

For this test, the chk_evt_nr helper has been extended to check
attributes in the matched events.

In this test, the 2 subflow closed events should have an error.

The 'Fixes' tag here below is the same as the one from the previous
commit: this patch here is not fixing anything wrong in the selftests,
but it validates the previous fix for an issue introduced by this commit
ID.

Fixes: 15cc10453398 ("mptcp: deliver ssk errors to msk")
Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
---
 tools/testing/selftests/net/mptcp/mptcp_join.sh | 23 ++++++++++++++++++++---
 1 file changed, 20 insertions(+), 3 deletions(-)

diff --git a/tools/testing/selftests/net/mptcp/mptcp_join.sh b/tools/testing/selftests/net/mptcp/mptcp_join.sh
index 35ff769f440f..e70d3420954f 100755
--- a/tools/testing/selftests/net/mptcp/mptcp_join.sh
+++ b/tools/testing/selftests/net/mptcp/mptcp_join.sh
@@ -3846,21 +3846,28 @@ userspace_pm_chk_get_addr()
 	fi
 }
 
-# $1: ns ; $2: event type ; $3: count
+# $1: ns ; $2: event type ; $3: count ; [ $4: attr ; $5: attr count ]
 chk_evt_nr()
 {
 	local ns=${1}
 	local evt_name="${2}"
 	local exp="${3}"
+	local attr="${4}"
+	local attr_exp="${5}"
 
 	local evts="${evts_ns1}"
 	local evt="${!evt_name}"
+	local attr_name
 	local count
 
+	if [ -n "${attr}" ]; then
+		attr_name=", ${attr}: ${attr_exp}"
+	fi
+
 	evt_name="${evt_name:16}" # without MPTCP_LIB_EVENT_
 	[ "${ns}" == "ns2" ] && evts="${evts_ns2}"
 
-	print_check "event ${ns} ${evt_name} (${exp})"
+	print_check "event ${ns} ${evt_name} (${exp}${attr_name})"
 
 	if [[ "${evt_name}" = "LISTENER_"* ]] &&
 	   ! mptcp_lib_kallsyms_has "mptcp_event_pm_listener$"; then
@@ -3872,6 +3879,16 @@ chk_evt_nr()
 	if [ "${count}" != "${exp}" ]; then
 		fail_test "got ${count} events, expected ${exp}"
 		cat "${evts}"
+		return
+	elif [ -z "${attr}" ]; then
+		print_ok
+		return
+	fi
+
+	count=$(grep -w "type:${evt}" "${evts}" | grep -c ",${attr}:")
+	if [ "${count}" != "${attr_exp}" ]; then
+		fail_test "got ${count} event attributes, expected ${attr_exp}"
+		grep -w "type:${evt}" "${evts}"
 	else
 		print_ok
 	fi
@@ -4130,7 +4147,7 @@ userspace_tests()
 			chk_subflows_total 1 1
 			userspace_pm_add_sf $ns2 10.0.1.2 0
 			wait_event ns2 MPTCP_LIB_EVENT_SUB_CLOSED 2
-			chk_evt_nr ns2 MPTCP_LIB_EVENT_SUB_CLOSED 2
+			chk_evt_nr ns2 MPTCP_LIB_EVENT_SUB_CLOSED 2 error 2
 		fi
 		kill_events_pids
 		mptcp_lib_kill_group_wait $tests_pid

-- 
2.51.0

[PATCH mptcp-net v2 05/11] selftests: mptcp: join: wait for estab event instead of MPJ

[Matthieu Baerts (NGI0)]

'wait_mpj' was used just after having created a background connection,
but before creating new subflows. So no MPJ were sent. The intention was
to wait for the connection to be established, which was the same as
doing a simple sleep with a "random" value.

Instead, wait for an "established" event. With this, the tests can
finish quicker.

Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
---
v2: use reset_with_events instead of starting them manually (Geliang)
---
 tools/testing/selftests/net/mptcp/mptcp_join.sh | 19 ++++++++++---------
 1 file changed, 10 insertions(+), 9 deletions(-)

diff --git a/tools/testing/selftests/net/mptcp/mptcp_join.sh b/tools/testing/selftests/net/mptcp/mptcp_join.sh
index e70d3420954f..ff20d86ed399 100755
--- a/tools/testing/selftests/net/mptcp/mptcp_join.sh
+++ b/tools/testing/selftests/net/mptcp/mptcp_join.sh
@@ -3999,7 +3999,7 @@ userspace_tests()
 		{ timeout_test=120 test_linkfail=128 speed=5 \
 			run_tests $ns1 $ns2 10.0.1.1 & } 2>/dev/null
 		local tests_pid=$!
-		wait_mpj $ns1
+		wait_event ns1 MPTCP_LIB_EVENT_ESTABLISHED 1
 		userspace_pm_add_addr $ns1 10.0.2.1 10
 		userspace_pm_add_addr $ns1 10.0.3.1 20
 		chk_join_nr 2 2 2
@@ -4032,7 +4032,7 @@ userspace_tests()
 		{ timeout_test=120 test_linkfail=128 speed=5 \
 			run_tests $ns1 $ns2 10.0.1.1 & } 2>/dev/null
 		local tests_pid=$!
-		wait_mpj $ns2
+		wait_event ns2 MPTCP_LIB_EVENT_ESTABLISHED 1
 		userspace_pm_add_sf $ns2 10.0.3.2 20
 		chk_join_nr 1 1 1
 		chk_mptcp_info subflows 1 subflows 1
@@ -4060,7 +4060,7 @@ userspace_tests()
 		{ timeout_test=120 test_linkfail=128 speed=5 \
 			run_tests $ns1 $ns2 10.0.1.1 & } 2>/dev/null
 		local tests_pid=$!
-		wait_mpj $ns2
+		wait_event ns2 MPTCP_LIB_EVENT_ESTABLISHED 1
 		chk_mptcp_info subflows 0 subflows 0
 		chk_subflows_total 1 1
 		userspace_pm_add_sf $ns2 10.0.3.2 0
@@ -4081,7 +4081,7 @@ userspace_tests()
 		{ timeout_test=120 test_linkfail=128 speed=5 \
 			run_tests $ns1 $ns2 10.0.1.1 & } 2>/dev/null
 		local tests_pid=$!
-		wait_mpj $ns2
+		wait_event ns2 MPTCP_LIB_EVENT_ESTABLISHED 1
 		userspace_pm_add_sf $ns2 10.0.3.2 20
 		chk_join_nr 1 1 1
 		chk_mptcp_info subflows 1 subflows 1
@@ -4105,7 +4105,7 @@ userspace_tests()
 		{ timeout_test=120 test_linkfail=128 speed=5 \
 			run_tests $ns1 $ns2 10.0.1.1 & } 2>/dev/null
 		local tests_pid=$!
-		wait_mpj $ns1
+		wait_event ns1 MPTCP_LIB_EVENT_ESTABLISHED 1
 		userspace_pm_add_addr $ns1 10.0.2.1 10
 		chk_join_nr 1 1 1
 		chk_add_nr 1 1
@@ -4158,7 +4158,7 @@ endpoint_tests()
 {
 	# subflow_rebuild_header is needed to support the implicit flag
 	# userspace pm type prevents add_addr
-	if reset "implicit EP" &&
+	if reset_with_events "implicit EP" &&
 	   continue_if mptcp_lib_kallsyms_has "subflow_rebuild_header$"; then
 		pm_nl_set_limits $ns1 2 2
 		pm_nl_set_limits $ns2 2 2
@@ -4167,7 +4167,7 @@ endpoint_tests()
 			run_tests $ns1 $ns2 10.0.1.1 & } 2>/dev/null
 		local tests_pid=$!
 
-		wait_mpj $ns1
+		wait_event ns2 MPTCP_LIB_EVENT_ESTABLISHED 1
 		pm_nl_check_endpoint "creation" \
 			$ns2 10.0.2.2 id 1 flags implicit
 		chk_mptcp_info subflows 1 subflows 1
@@ -4181,6 +4181,7 @@ endpoint_tests()
 		pm_nl_check_endpoint "modif is allowed" \
 			$ns2 10.0.2.2 id 1 flags signal
 		mptcp_lib_kill_group_wait $tests_pid
+		kill_events_pids
 	fi
 
 	if reset_with_tcp_filter "delete and re-add" ns2 10.0.3.2 REJECT OUTPUT &&
@@ -4194,7 +4195,7 @@ endpoint_tests()
 			run_tests $ns1 $ns2 10.0.1.1 & } 2>/dev/null
 		local tests_pid=$!
 
-		wait_mpj $ns2
+		wait_event ns2 MPTCP_LIB_EVENT_ESTABLISHED 1
 		pm_nl_check_endpoint "creation" \
 			$ns2 10.0.2.2 id 2 flags subflow dev ns2eth2
 		chk_subflow_nr "before delete id 2" 2
@@ -4272,7 +4273,7 @@ endpoint_tests()
 			run_tests $ns1 $ns2 10.0.1.1 & } 2>/dev/null
 		local tests_pid=$!
 
-		wait_mpj $ns2
+		wait_event ns2 MPTCP_LIB_EVENT_ESTABLISHED 1
 		pm_nl_check_endpoint "creation" \
 			$ns1 10.0.2.1 id 1 flags signal
 		chk_subflow_nr "before delete" 2

-- 
2.51.0

[PATCH mptcp-net v2 06/11] selftests: mptcp: join: fix wait_mpj helper

[Matthieu Baerts (NGI0)]

It looks like most of the time, this helper was simply waiting a bit
more than one second: the previous MPJoin counter was often already at
the expected value. So at the end, it was just checking 10 times for
the MPJoin counter to change, but it was not happening. For the tests,
that was time, it was just waiting longer for nothing.

Instead, use 'wait_mpj' with the expected counter: in the tests, the MPJ
counter can easily be predicted. While at it, stop passing the netns as
argument: here the received MPJoin ACK is checked, which happens on the
server side. If later on, this needs to be checked on the client side,
the helper can be adapted for this case, but better avoid confusions now
if it is not needed.

While at it, stop using 'i' for the variable if it is not used.

With this, the tests can finish quicker.

Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
---
 tools/testing/selftests/net/mptcp/mptcp_join.sh | 30 ++++++++++++-------------
 1 file changed, 14 insertions(+), 16 deletions(-)

diff --git a/tools/testing/selftests/net/mptcp/mptcp_join.sh b/tools/testing/selftests/net/mptcp/mptcp_join.sh
index ff20d86ed399..6ab568d6b856 100755
--- a/tools/testing/selftests/net/mptcp/mptcp_join.sh
+++ b/tools/testing/selftests/net/mptcp/mptcp_join.sh
@@ -631,17 +631,15 @@ wait_rm_sf()
 	done
 }
 
+# $1: expected MPJ ACK Rx counter in $ns1
 wait_mpj()
 {
-	local ns="${1}"
-	local cnt old_cnt
+	local exp_cnt="${1}"
+	local cnt
 
-	old_cnt=$(mptcp_lib_get_counter ${ns} "MPTcpExtMPJoinAckRx")
-
-	local i
-	for i in $(seq 10); do
-		cnt=$(mptcp_lib_get_counter ${ns} "MPTcpExtMPJoinAckRx")
-		[ "$cnt" = "${old_cnt}" ] || break
+	for _ in $(seq 10); do
+		cnt=$(mptcp_lib_get_counter ${ns1} "MPTcpExtMPJoinAckRx")
+		[ "${cnt}" = "${exp_cnt}" ] && break
 		sleep 0.1
 	done
 }
@@ -4207,7 +4205,7 @@ endpoint_tests()
 		chk_mptcp_info subflows 0 subflows 0
 
 		pm_nl_add_endpoint $ns2 10.0.2.2 id 2 dev ns2eth2 flags subflow
-		wait_mpj $ns2
+		wait_mpj 2
 		chk_subflow_nr "after re-add id 2" 2
 		chk_mptcp_info subflows 1 subflows 1
 
@@ -4219,7 +4217,7 @@ endpoint_tests()
 		ip netns exec "${ns2}" ${iptables} -D OUTPUT -s "10.0.3.2" -p tcp -j REJECT
 		pm_nl_del_endpoint $ns2 3 10.0.3.2
 		pm_nl_add_endpoint $ns2 10.0.3.2 id 3 flags subflow
-		wait_mpj $ns2
+		wait_mpj 3
 		chk_subflow_nr "after no reject" 3
 		chk_mptcp_info subflows 2 subflows 2
 
@@ -4231,7 +4229,7 @@ endpoint_tests()
 			chk_mptcp_info subflows 2 subflows 2 # only decr for additional sf
 
 			pm_nl_add_endpoint $ns2 10.0.1.2 id 1 dev ns2eth1 flags subflow
-			wait_mpj $ns2
+			wait_mpj $((3 + i))
 			chk_subflow_nr "after re-add id 0 ($i)" 3
 			chk_mptcp_info subflows 3 subflows 3
 		done
@@ -4289,7 +4287,7 @@ endpoint_tests()
 
 		pm_nl_add_endpoint $ns1 10.0.2.1 id 1 flags signal
 		pm_nl_add_endpoint $ns1 10.0.3.1 id 2 flags signal
-		wait_mpj $ns2
+		wait_mpj 3
 		chk_subflow_nr "after re-add" 3
 		chk_mptcp_info subflows 2 subflows 2
 		chk_mptcp_info add_addr_signal 2 add_addr_accepted 2
@@ -4301,7 +4299,7 @@ endpoint_tests()
 		chk_mptcp_info add_addr_signal 2 add_addr_accepted 2
 
 		pm_nl_add_endpoint $ns1 10.0.1.1 id 99 flags signal
-		wait_mpj $ns2
+		wait_mpj 4
 		chk_subflow_nr "after re-add ID 0" 3
 		chk_mptcp_info subflows 3 subflows 3
 		chk_mptcp_info add_addr_signal 3 add_addr_accepted 2
@@ -4313,7 +4311,7 @@ endpoint_tests()
 		chk_mptcp_info add_addr_signal 2 add_addr_accepted 2
 
 		pm_nl_add_endpoint $ns1 10.0.1.1 id 88 flags signal
-		wait_mpj $ns2
+		wait_mpj 5
 		chk_subflow_nr "after re-re-add ID 0" 3
 		chk_mptcp_info subflows 3 subflows 3
 		chk_mptcp_info add_addr_signal 3 add_addr_accepted 2
@@ -4362,9 +4360,9 @@ endpoint_tests()
 		wait_rm_addr $ns2 0
 		ip netns exec "${ns2}" ${iptables} -D OUTPUT -s "10.0.3.2" -p tcp -j REJECT
 		pm_nl_add_endpoint $ns2 10.0.3.2 id 3 flags subflow
-		wait_mpj $ns2
+		wait_mpj 1
 		pm_nl_add_endpoint $ns1 10.0.3.1 id 2 flags signal
-		wait_mpj $ns2
+		wait_mpj 2
 		mptcp_lib_kill_group_wait $tests_pid
 
 		join_syn_tx=3 join_connect_err=1 \

-- 
2.51.0

[PATCH mptcp-net v2 07/11] selftests: mptcp: join: userspace: wait for new events

[Matthieu Baerts (NGI0)]

Instead of waiting for a random amount of time (1 second), wait for an
event to be received on the other side.

To do that, when an address is announced (userspace_pm_add_addr), the
ANNOUNCED is expected. When a new subflow is created
(userspace_pm_add_sf), the SUB_ESTABLISHED event is expected.

With this, the tests can finish quicker.

Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
---
 tools/testing/selftests/net/mptcp/mptcp_join.sh | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

diff --git a/tools/testing/selftests/net/mptcp/mptcp_join.sh b/tools/testing/selftests/net/mptcp/mptcp_join.sh
index 6ab568d6b856..4977e6ff17b4 100755
--- a/tools/testing/selftests/net/mptcp/mptcp_join.sh
+++ b/tools/testing/selftests/net/mptcp/mptcp_join.sh
@@ -3716,7 +3716,6 @@ userspace_pm_add_addr()
 	tk=$(mptcp_lib_evts_get_info token "$evts")
 
 	ip netns exec $1 ./pm_nl_ctl ann $2 token $tk id $3
-	sleep 1
 }
 
 # $1: ns ; $2: id
@@ -3747,7 +3746,6 @@ userspace_pm_add_sf()
 
 	ip netns exec $1 ./pm_nl_ctl csf lip $2 lid $3 \
 				rip $da rport $dp token $tk
-	sleep 1
 }
 
 # $1: ns ; $2: addr $3: event type
@@ -3999,7 +3997,9 @@ userspace_tests()
 		local tests_pid=$!
 		wait_event ns1 MPTCP_LIB_EVENT_ESTABLISHED 1
 		userspace_pm_add_addr $ns1 10.0.2.1 10
+		wait_event ns2 MPTCP_LIB_EVENT_ANNOUNCED 1
 		userspace_pm_add_addr $ns1 10.0.3.1 20
+		wait_event ns2 MPTCP_LIB_EVENT_ANNOUNCED 2
 		chk_join_nr 2 2 2
 		chk_add_nr 2 2
 		chk_mptcp_info subflows 2 subflows 2
@@ -4032,6 +4032,7 @@ userspace_tests()
 		local tests_pid=$!
 		wait_event ns2 MPTCP_LIB_EVENT_ESTABLISHED 1
 		userspace_pm_add_sf $ns2 10.0.3.2 20
+		wait_event ns2 MPTCP_LIB_EVENT_SUB_ESTABLISHED 1
 		chk_join_nr 1 1 1
 		chk_mptcp_info subflows 1 subflows 1
 		chk_subflows_total 2 2
@@ -4062,6 +4063,7 @@ userspace_tests()
 		chk_mptcp_info subflows 0 subflows 0
 		chk_subflows_total 1 1
 		userspace_pm_add_sf $ns2 10.0.3.2 0
+		wait_event ns2 MPTCP_LIB_EVENT_SUB_ESTABLISHED 1
 		userspace_pm_chk_dump_addr "${ns2}" \
 			"id 0 flags subflow 10.0.3.2" "id 0 subflow"
 		chk_join_nr 1 1 1
@@ -4081,6 +4083,7 @@ userspace_tests()
 		local tests_pid=$!
 		wait_event ns2 MPTCP_LIB_EVENT_ESTABLISHED 1
 		userspace_pm_add_sf $ns2 10.0.3.2 20
+		wait_event ns2 MPTCP_LIB_EVENT_SUB_ESTABLISHED 1
 		chk_join_nr 1 1 1
 		chk_mptcp_info subflows 1 subflows 1
 		chk_subflows_total 2 2
@@ -4105,6 +4108,7 @@ userspace_tests()
 		local tests_pid=$!
 		wait_event ns1 MPTCP_LIB_EVENT_ESTABLISHED 1
 		userspace_pm_add_addr $ns1 10.0.2.1 10
+		wait_event ns2 MPTCP_LIB_EVENT_ANNOUNCED 1
 		chk_join_nr 1 1 1
 		chk_add_nr 1 1
 		chk_mptcp_info subflows 1 subflows 1
@@ -4131,6 +4135,7 @@ userspace_tests()
 		local tests_pid=$!
 		wait_event ns2 MPTCP_LIB_EVENT_ESTABLISHED 1
 		userspace_pm_add_sf $ns2 10.0.3.2 20
+		wait_event ns2 MPTCP_LIB_EVENT_SUB_ESTABLISHED 1
 		chk_mptcp_info subflows 1 subflows 1
 		chk_subflows_total 2 2
 

-- 
2.51.0

[PATCH mptcp-net v2 08/11] selftests: mptcp: join chk_stale_nr: avoid dup stats

[Matthieu Baerts (NGI0)]

nstat outputs are already printed when calling 'fail_test', no need to
do it again.

While at it, no need to use the dump_stats variable, print the extra
stats directly. And use 'ip -n $ns' instead of 'ip netns exec $ns',
shorter and clearer.

Reviewed-by: Geliang Tang <geliang@kernel.org>
Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
---
 tools/testing/selftests/net/mptcp/mptcp_join.sh | 10 ++--------
 1 file changed, 2 insertions(+), 8 deletions(-)

diff --git a/tools/testing/selftests/net/mptcp/mptcp_join.sh b/tools/testing/selftests/net/mptcp/mptcp_join.sh
index 4977e6ff17b4..a8b9782a85df 100755
--- a/tools/testing/selftests/net/mptcp/mptcp_join.sh
+++ b/tools/testing/selftests/net/mptcp/mptcp_join.sh
@@ -1648,7 +1648,6 @@ chk_stale_nr()
 	local stale_min=$2
 	local stale_max=$3
 	local stale_delta=$4
-	local dump_stats
 	local stale_nr
 	local recover_nr
 
@@ -1664,16 +1663,11 @@ chk_stale_nr()
 		fail_test "got $stale_nr stale[s] $recover_nr recover[s], " \
 		     " expected stale in range [$stale_min..$stale_max]," \
 		     " stale-recover delta $stale_delta"
-		dump_stats=1
+		echo $ns stats
+		ip -n $ns -s link show
 	else
 		print_ok
 	fi
-
-	if [ "${dump_stats}" = 1 ]; then
-		echo $ns stats
-		ip netns exec $ns ip -s link show
-		ip netns exec $ns nstat -as | grep MPTcp
-	fi
 }
 
 chk_add_nr()

-- 
2.51.0

[PATCH mptcp-net v2 09/11] selftests: mptcp: join: avoid declaring i if not used

[Matthieu Baerts (NGI0)]

A few loops were declaring 'i', but this variable was not used.

To avoid confusions, use '_' instead: it is more explicit to mark that
this variable is not needed.

Reviewed-by: Geliang Tang <geliang@kernel.org>
Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
---
 tools/testing/selftests/net/mptcp/mptcp_join.sh | 9 +++------
 1 file changed, 3 insertions(+), 6 deletions(-)

diff --git a/tools/testing/selftests/net/mptcp/mptcp_join.sh b/tools/testing/selftests/net/mptcp/mptcp_join.sh
index a8b9782a85df..0f9253d607c3 100755
--- a/tools/testing/selftests/net/mptcp/mptcp_join.sh
+++ b/tools/testing/selftests/net/mptcp/mptcp_join.sh
@@ -603,8 +603,7 @@ wait_rm_addr()
 	local old_cnt="${2}"
 	local cnt
 
-	local i
-	for i in $(seq 10); do
+	for _ in $(seq 10); do
 		cnt=$(rm_addr_count ${ns})
 		[ "$cnt" = "${old_cnt}" ] || break
 		sleep 0.1
@@ -623,8 +622,7 @@ wait_rm_sf()
 	local old_cnt="${2}"
 	local cnt
 
-	local i
-	for i in $(seq 10); do
+	for _ in $(seq 10); do
 		cnt=$(rm_sf_count ${ns})
 		[ "$cnt" = "${old_cnt}" ] || break
 		sleep 0.1
@@ -648,8 +646,7 @@ wait_ll_ready()
 {
 	local ns="${1}"
 
-	local i
-	for i in $(seq 50); do
+	for _ in $(seq 50); do
 		ip -n "${ns}" -6 addr show scope link | grep "inet6 fe80" |
 			grep -qw "tentative" || break
 		sleep 0.1

-- 
2.51.0

[PATCH mptcp-net v2 10/11] selftests: mptcp: connect: fix maybe-uninitialize warn

[Matthieu Baerts (NGI0)]

This warning can be seen with GCC 15.2:

  mptcp_connect.c: In function ‘main_loop’:
  mptcp_connect.c:1422:37: warning: ‘peer’ may be used uninitialized [-Wmaybe-uninitialized]
   1422 |                 if (connect(fd, peer->ai_addr, peer->ai_addrlen))
        |                                 ~~~~^~~~~~~~~
  mptcp_connect.c:1377:26: note: ‘peer’ was declared here
   1377 |         struct addrinfo *peer;
        |                          ^~~~

This variable is set in sock_connect_mptcp() in some conditions. If not,
this helper returns an error, and the program stops. So this is a false
positive, but better remove it by initialising peer to NULL.

Reviewed-by: Geliang Tang <geliang@kernel.org>
Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
---
 tools/testing/selftests/net/mptcp/mptcp_connect.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tools/testing/selftests/net/mptcp/mptcp_connect.c b/tools/testing/selftests/net/mptcp/mptcp_connect.c
index 1c4fe60089a2..554886f23d6a 100644
--- a/tools/testing/selftests/net/mptcp/mptcp_connect.c
+++ b/tools/testing/selftests/net/mptcp/mptcp_connect.c
@@ -1354,8 +1354,8 @@ void xdisconnect(int fd)
 
 int main_loop(void)
 {
+	struct addrinfo *peer = NULL;
 	int fd = 0, ret, fd_in = 0;
-	struct addrinfo *peer;
 	struct wstate winfo;
 
 	if (cfg_input && cfg_sockopt_types.mptfo) {

-- 
2.51.0

[PATCH mptcp-net v2 11/11] selftests: mptcp: connect cleanup TFO setup

[Matthieu Baerts (NGI0)]

To the TFO, only the file descriptor is needed, the family is not.

Also, the error can be handled the same way when 'sendto()' or
'connect()' are used. Only the printed error message is different.

This avoids a bit of confusions.

Reviewed-by: Geliang Tang <geliang@kernel.org>
Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
---
 tools/testing/selftests/net/mptcp/mptcp_connect.c | 17 +++++++----------
 1 file changed, 7 insertions(+), 10 deletions(-)

diff --git a/tools/testing/selftests/net/mptcp/mptcp_connect.c b/tools/testing/selftests/net/mptcp/mptcp_connect.c
index 554886f23d6a..ff1a298d3469 100644
--- a/tools/testing/selftests/net/mptcp/mptcp_connect.c
+++ b/tools/testing/selftests/net/mptcp/mptcp_connect.c
@@ -259,7 +259,7 @@ static void set_transparent(int fd, int pf)
 	}
 }
 
-static void set_mptfo(int fd, int pf)
+static void set_mptfo(int fd)
 {
 	int qlen = 25;
 
@@ -336,7 +336,7 @@ static int sock_listen_mptcp(const char * const listenaddr,
 			set_transparent(sock, pf);
 
 		if (cfg_sockopt_types.mptfo)
-			set_mptfo(sock, pf);
+			set_mptfo(sock);
 
 		if (bind(sock, a->ai_addr, a->ai_addrlen) == 0)
 			break; /* success */
@@ -407,21 +407,18 @@ static int sock_connect_mptcp(const char * const remoteaddr,
 				*peer = a;
 				break; /* success */
 			}
+			perror("sendto()");
 		} else {
 			if (connect(sock, a->ai_addr, a->ai_addrlen) == 0) {
 				*peer = a;
 				break; /* success */
 			}
-		}
-		if (cfg_sockopt_types.mptfo) {
-			perror("sendto()");
-			close(sock);
-			sock = -1;
-		} else {
 			perror("connect()");
-			close(sock);
-			sock = -1;
 		}
+
+		/* error */
+		close(sock);
+		sock = -1;
 	}
 
 	freeaddrinfo(addr);

-- 
2.51.0

Re: [PATCH mptcp-net v2 00/11] mptcp: avoid dup events + error + misc.

[MPTCP CI]

Hi Matthieu,

Thank you for your modifications, that's great!

Our CI did some validations and here is its report:

- KVM Validation: normal (except selftest_mptcp_join): Success! ✅
- KVM Validation: normal (only selftest_mptcp_join): Success! ✅
- KVM Validation: debug (except selftest_mptcp_join): Success! ✅
- KVM Validation: debug (only selftest_mptcp_join): Success! ✅
- KVM Validation: btf-normal (only bpftest_all): Success! ✅
- KVM Validation: btf-debug (only bpftest_all): Success! ✅
- Task: https://github.com/multipath-tcp/mptcp_net-next/actions/runs/21371621189

Initiator: Patchew Applier
Commits: https://github.com/multipath-tcp/mptcp_net-next/commits/69110da63c84
Patchwork: https://patchwork.kernel.org/project/mptcp/list/?series=1047262


If there are some issues, you can reproduce them using the same environment as
the one used by the CI thanks to a docker image, e.g.:

    $ cd [kernel source code]
    $ docker run -v "${PWD}:${PWD}:rw" -w "${PWD}" --privileged --rm -it \
        --pull always mptcp/mptcp-upstream-virtme-docker:latest \
        auto-normal

For more details:

    https://github.com/multipath-tcp/mptcp-upstream-virtme-docker


Please note that despite all the efforts that have been already done to have a
stable tests suite when executed on a public CI like here, it is possible some
reported issues are not due to your modifications. Still, do not hesitate to
help us improve that ;-)

Cheers,
MPTCP GH Action bot
Bot operated by Matthieu Baerts (NGI0 Core)

Re: [PATCH mptcp-net v2 00/11] mptcp: avoid dup events + error + misc.

[Geliang Tang]

Hi Matt,

Thanks for this v2.

On Mon, 2026-01-26 at 20:18 +0100, Matthieu Baerts (NGI0) wrote:
> There are two fixes with their tests:
> 
> - a subflow closed event can be visible multiple times, see issue
> #603
> 
> - the subflow closed event didn't contain the error as expected.
> 
> The 4 first patches are for -net, the rest are small fixes and
> improvements for net-next. Some tests should be quicker now.
> 
> Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>

Looks good to me.

Reviewed-by: Geliang Tang <geliang@kernel.org>

-Geliang

> ---
> Changes in v2:
> - Small updates on patches 1, 2, 3 and 5 (Geliang)
> - Add Geliang RvB tags.
> - Link to v1:
> https://patch.msgid.link/20251226-mptcp-issue-603-v1-0-bb30e331b839@kernel.org
> 
> ---
> Matthieu Baerts (NGI0) (11):
>       mptcp: avoid dup SUB_CLOSED events after disconnect
>       selftests: mptcp: check no dup close events after error
>       mptcp: only reset subflow errors when propagated
>       selftests: mptcp: check subflow errors in close events
>       selftests: mptcp: join: wait for estab event instead of MPJ
>       selftests: mptcp: join: fix wait_mpj helper
>       selftests: mptcp: join: userspace: wait for new events
>       selftests: mptcp: join chk_stale_nr: avoid dup stats
>       selftests: mptcp: join: avoid declaring i if not used
>       selftests: mptcp: connect: fix maybe-uninitialize warn
>       selftests: mptcp: connect cleanup TFO setup
> 
>  net/mptcp/protocol.c                              |  13 +-
>  tools/testing/selftests/net/mptcp/mptcp_connect.c |  19 ++-
>  tools/testing/selftests/net/mptcp/mptcp_join.sh   | 149
> +++++++++++++++-------
>  3 files changed, 121 insertions(+), 60 deletions(-)
> ---
> base-commit: b8bb952910af48167681524c5b217996c16382c1
> change-id: 20251222-mptcp-issue-603-d85d967f7c0a
> 
> Best regards,

Re: [PATCH mptcp-net v2 00/11] mptcp: avoid dup events + error + misc.

[Matthieu Baerts]

Hi Geliang,

On 27/01/2026 07:49, Geliang Tang wrote:
> Hi Matt,
> 
> Thanks for this v2.
> 
> On Mon, 2026-01-26 at 20:18 +0100, Matthieu Baerts (NGI0) wrote:
>> There are two fixes with their tests:
>>
>> - a subflow closed event can be visible multiple times, see issue
>> #603
>>
>> - the subflow closed event didn't contain the error as expected.
>>
>> The 4 first patches are for -net, the rest are small fixes and
>> improvements for net-next. Some tests should be quicker now.
>>
>> Signed-off-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>
> 
> Looks good to me.
> 
> Reviewed-by: Geliang Tang <geliang@kernel.org>

Thank you for the review!

Now in our tree:

New patches for t/upstream-net and t/upstream:
- e00c1cde2f33: mptcp: avoid dup SUB_CLOSED events after disconnect
- 2690dee1d1c2: selftests: mptcp: check no dup close events after error
- b00ab6a1c0a3: mptcp: only reset subflow errors when propagated
- 16cfe3e8f86d: selftests: mptcp: check subflow errors in close events
- Results: e2aed0b16078..b9ebb632c0c5 (export-net)
- Results: f35d5bacc3d9..3a7897f294e0 (export)

New patches for t/upstream:
- 3a865473b8fc: selftests: mptcp: join: wait for estab event instead of MPJ
- 0e605a03b612: selftests: mptcp: join: fix wait_mpj helper
- 7ba1f3ed5a69: selftests: mptcp: join: userspace: wait for new events
- 9c01b3c47ce3: selftests: mptcp: join chk_stale_nr: avoid dup stats
- 29d0083334ab: selftests: mptcp: join: avoid declaring i if not used
- e6a4afd1527a: selftests: mptcp: connect: fix maybe-uninitialize warn
- 983eaf267fa4: selftests: mptcp: connect cleanup TFO setup
- Results: 3a7897f294e0..7db9a30f8664 (export)

Tests are now in progress:

- export-net:
https://github.com/multipath-tcp/mptcp_net-next/commit/f4278db6305ce32da9f5d541d3034ef2358f38e1/checks
- export:
https://github.com/multipath-tcp/mptcp_net-next/commit/954a6c81672b5ff22af3066565cbbc8631fd70f6/checks
- export:
https://github.com/multipath-tcp/mptcp_net-next/commit/8ad88e01fe8c2c5b0a6c897b0cf72bba60594b9a/checks

Cheers,
Matt
-- 
Sponsored by the NGI0 Core fund.