[PATCH mptcp-next v9 0/9] userspace pm remove id 0 subflow & address

[PATCH mptcp-next v9 0/9] userspace pm remove id 0 subflow & address

[Geliang Tang]

v9:
 - update __mptcp_check_initial_subflow().
 - split the code that relies on the 'ss' modification into patch 9.

v8:
 - merge two squash-to patches for v7.
 - address to #428 too.

v7:
 - move evts_get_info into mptcp_lib.sh, update userspace_pm.sh with
   this helper too.
 - drop the last two cleanup patches, they'll be sent in another series.

v6:
 - add a new helper get_info_value, update chk_mptcp_info(),
   verify_listener_events() and pedit_action_pkts() too.
 - only patch 1, 2, 4 are changed, other patches are the same as v5.

v5:
 - update userspace_pm_add_addr/sf helpers.
 - add userspace_pm_rm_addr/sf helpers.
 - remove id 0 subflow or address will send MP_RST, add chk_rst_nr for
   them.
 - add two selftests cleanups using for-loop.

v4:
 - add evts_get_info and chk_subflows helpers.
 - split the selftests patch into three.

v3:
 - address Matt's comments in v2.

v2:
 - fix CI errors.

This patchset addresses #379 and #391, add the abilities to remove id 0
subflow and address for userspace PM. And a selftest.

Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/379
Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/391
Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/428

Geliang Tang (9):
  selftests: mptcp: add evts_get_info helper
  selftests: mptcp: update userspace pm test helpers
  selftests: mptcp: userspace pm remove id 0 subflow
  mptcp: userspace pm allow creating id 0 subflow
  selftests: mptcp: userspace pm create id 0 subflow
  mptcp: add no_initial_subflow mptcpi_flags
  mptcp: userspace pm remove id 0 address
  selftests: mptcp: userspace pm remove id 0 address
  selftests: mptcp: check no_initial_subflow flag

 include/uapi/linux/mptcp.h                    |   1 +
 net/mptcp/pm_netlink.c                        |   3 +-
 net/mptcp/pm_userspace.c                      |  33 ++-
 net/mptcp/protocol.h                          |   5 +
 net/mptcp/sockopt.c                           |   2 +
 .../testing/selftests/net/mptcp/mptcp_join.sh | 221 +++++++++++++-----
 .../testing/selftests/net/mptcp/mptcp_lib.sh  |  12 +
 .../selftests/net/mptcp/userspace_pm.sh       |  87 ++++---
 8 files changed, 252 insertions(+), 112 deletions(-)

-- 
2.35.3

[PATCH mptcp-next v9 1/9] selftests: mptcp: add evts_get_info helper

[Geliang Tang]

This patch adds a new helper get_info_value(), using 'sed' command to
parse the value of the given item name in the line with the given keyword,
to make chk_mptcp_info() and pedit_action_pkts() more readable.

Also add another helper evts_get_info() to use get_info_value() to parse
the output of 'pm_nl_ctl events' command, to make all the userpsace pm
selftests more readable, both in mptcp_join.sh and userspace_pm.sh.

Signed-off-by: Geliang Tang <geliang.tang@suse.com>
---
 .../testing/selftests/net/mptcp/mptcp_join.sh | 19 ++--
 .../testing/selftests/net/mptcp/mptcp_lib.sh  | 10 +++
 .../selftests/net/mptcp/userspace_pm.sh       | 86 +++++++++----------
 3 files changed, 57 insertions(+), 58 deletions(-)

diff --git a/tools/testing/selftests/net/mptcp/mptcp_join.sh b/tools/testing/selftests/net/mptcp/mptcp_join.sh
index ee1f89a872b3..27517ba770bc 100755
--- a/tools/testing/selftests/net/mptcp/mptcp_join.sh
+++ b/tools/testing/selftests/net/mptcp/mptcp_join.sh
@@ -1862,10 +1862,8 @@ chk_mptcp_info()
 
 	print_check "mptcp_info ${info1:0:8}=$exp1:$exp2"
 
-	cnt1=$(ss -N $ns1 -inmHM | grep "$info1:" |
-	       sed -n 's/.*\('"$info1"':\)\([[:digit:]]*\).*$/\2/p;q')
-	cnt2=$(ss -N $ns2 -inmHM | grep "$info2:" |
-	       sed -n 's/.*\('"$info2"':\)\([[:digit:]]*\).*$/\2/p;q')
+	cnt1=$(ss -N $ns1 -inmHM | mptcp_lib_get_info_value "$info1" "$info1")
+	cnt2=$(ss -N $ns2 -inmHM | mptcp_lib_get_info_value "$info2" "$info2")
 	# 'ss' only display active connections and counters that are not 0.
 	[ -z "$cnt1" ] && cnt1=0
 	[ -z "$cnt2" ] && cnt2=0
@@ -2828,13 +2826,13 @@ verify_listener_events()
 		return
 	fi
 
-	type=$(grep "type:$e_type," $evt | sed -n 's/.*\(type:\)\([[:digit:]]*\).*$/\2/p;q')
-	family=$(grep "type:$e_type," $evt | sed -n 's/.*\(family:\)\([[:digit:]]*\).*$/\2/p;q')
-	sport=$(grep "type:$e_type," $evt | sed -n 's/.*\(sport:\)\([[:digit:]]*\).*$/\2/p;q')
+	type=$(mptcp_lib_evts_get_info type "$evt" "$e_type")
+	family=$(mptcp_lib_evts_get_info family "$evt" "$e_type")
+	sport=$(mptcp_lib_evts_get_info sport "$evt" "$e_type")
 	if [ $family ] && [ $family = $AF_INET6 ]; then
-		saddr=$(grep "type:$e_type," $evt | sed -n 's/.*\(saddr6:\)\([0-9a-f:.]*\).*$/\2/p;q')
+		saddr=$(mptcp_lib_evts_get_info saddr6 "$evt" "$e_type")
 	else
-		saddr=$(grep "type:$e_type," $evt | sed -n 's/.*\(saddr4:\)\([0-9.]*\).*$/\2/p;q')
+		saddr=$(mptcp_lib_evts_get_info saddr4 "$evt" "$e_type")
 	fi
 
 	if [ $type ] && [ $type = $e_type ] &&
@@ -3229,8 +3227,7 @@ fastclose_tests()
 pedit_action_pkts()
 {
 	tc -n $ns2 -j -s action show action pedit index 100 | \
-		grep "packets" | \
-		sed 's/.*"packets":\([0-9]\+\),.*/\1/'
+		mptcp_lib_get_info_value \"packets\" packets
 }
 
 fail_tests()
diff --git a/tools/testing/selftests/net/mptcp/mptcp_lib.sh b/tools/testing/selftests/net/mptcp/mptcp_lib.sh
index 92a5befe8039..def35395a254 100644
--- a/tools/testing/selftests/net/mptcp/mptcp_lib.sh
+++ b/tools/testing/selftests/net/mptcp/mptcp_lib.sh
@@ -207,3 +207,13 @@ mptcp_lib_result_print_all_tap() {
 		printf "%s\n" "${subtest}"
 	done
 }
+
+# get the value of keyword $1 in the line marked by keyword $2
+mptcp_lib_get_info_value() {
+	grep "${2}" | sed -n 's/.*\('${1}':\)\([0-9a-f:.]*\).*$/\2/p;q'
+}
+
+# $1: info name ; $2: evts_ns ; $3: event type
+mptcp_lib_evts_get_info() {
+	cat "${2}" | mptcp_lib_get_info_value "${1}" "^type:${3:-1},"
+}
diff --git a/tools/testing/selftests/net/mptcp/userspace_pm.sh b/tools/testing/selftests/net/mptcp/userspace_pm.sh
index b25a3e33eb25..2413059a42e5 100755
--- a/tools/testing/selftests/net/mptcp/userspace_pm.sh
+++ b/tools/testing/selftests/net/mptcp/userspace_pm.sh
@@ -247,14 +247,11 @@ make_connection()
 	local server_token
 	local server_serverside
 
-	client_token=$(sed --unbuffered -n 's/.*\(token:\)\([[:digit:]]*\).*$/\2/p;q' "$client_evts")
-	client_port=$(sed --unbuffered -n 's/.*\(sport:\)\([[:digit:]]*\).*$/\2/p;q' "$client_evts")
-	client_serverside=$(sed --unbuffered -n 's/.*\(server_side:\)\([[:digit:]]*\).*$/\2/p;q'\
-				      "$client_evts")
-	server_token=$(grep "type:1," "$server_evts" |
-		       sed --unbuffered -n 's/.*\(token:\)\([[:digit:]]*\).*$/\2/p;q')
-	server_serverside=$(grep "type:1," "$server_evts" |
-			    sed --unbuffered -n 's/.*\(server_side:\)\([[:digit:]]*\).*$/\2/p;q')
+	client_token=$(mptcp_lib_evts_get_info token "$client_evts")
+	client_port=$(mptcp_lib_evts_get_info sport "$client_evts")
+	client_serverside=$(mptcp_lib_evts_get_info server_side "$client_evts")
+	server_token=$(mptcp_lib_evts_get_info token "$server_evts")
+	server_serverside=$(mptcp_lib_evts_get_info server_side "$server_evts")
 
 	print_test "Established IP${is_v6} MPTCP Connection ns2 => ns1"
 	if [ "$client_token" != "" ] && [ "$server_token" != "" ] && [ "$client_serverside" = 0 ] &&
@@ -340,16 +337,16 @@ verify_announce_event()
 	local dport
 	local id
 
-	type=$(sed --unbuffered -n 's/.*\(type:\)\([[:digit:]]*\).*$/\2/p;q' "$evt")
-	token=$(sed --unbuffered -n 's/.*\(token:\)\([[:digit:]]*\).*$/\2/p;q' "$evt")
+	type=$(mptcp_lib_evts_get_info type "$evt" $e_type)
+	token=$(mptcp_lib_evts_get_info token "$evt" $e_type)
 	if [ "$e_af" = "v6" ]
 	then
-		addr=$(sed --unbuffered -n 's/.*\(daddr6:\)\([0-9a-f:.]*\).*$/\2/p;q' "$evt")
+		addr=$(mptcp_lib_evts_get_info daddr6 "$evt" $e_type)
 	else
-		addr=$(sed --unbuffered -n 's/.*\(daddr4:\)\([0-9.]*\).*$/\2/p;q' "$evt")
+		addr=$(mptcp_lib_evts_get_info daddr4 "$evt" $e_type)
 	fi
-	dport=$(sed --unbuffered -n 's/.*\(dport:\)\([[:digit:]]*\).*$/\2/p;q' "$evt")
-	id=$(sed --unbuffered -n 's/.*\(rem_id:\)\([[:digit:]]*\).*$/\2/p;q' "$evt")
+	dport=$(mptcp_lib_evts_get_info dport "$evt" $e_type)
+	id=$(mptcp_lib_evts_get_info rem_id "$evt" $e_type)
 
 	check_expected "type" "token" "addr" "dport" "id"
 }
@@ -367,7 +364,7 @@ test_announce()
 	   $client_addr_id dev ns2eth1 > /dev/null 2>&1
 
 	local type
-	type=$(sed --unbuffered -n 's/.*\(type:\)\([[:digit:]]*\).*$/\2/p;q' "$server_evts")
+	type=$(mptcp_lib_evts_get_info type "$server_evts")
 	print_test "ADD_ADDR 10.0.2.2 (ns2) => ns1, invalid token"
 	if [ "$type" = "" ]
 	then
@@ -446,9 +443,9 @@ verify_remove_event()
 	local token
 	local id
 
-	type=$(sed --unbuffered -n 's/.*\(type:\)\([[:digit:]]*\).*$/\2/p;q' "$evt")
-	token=$(sed --unbuffered -n 's/.*\(token:\)\([[:digit:]]*\).*$/\2/p;q' "$evt")
-	id=$(sed --unbuffered -n 's/.*\(rem_id:\)\([[:digit:]]*\).*$/\2/p;q' "$evt")
+	type=$(mptcp_lib_evts_get_info type "$evt" $e_type)
+	token=$(mptcp_lib_evts_get_info token "$evt" $e_type)
+	id=$(mptcp_lib_evts_get_info rem_id "$evt" $e_type)
 
 	check_expected "type" "token" "id"
 }
@@ -466,7 +463,7 @@ test_remove()
 	   $client_addr_id > /dev/null 2>&1
 	print_test "RM_ADDR id:${client_addr_id} ns2 => ns1, invalid token"
 	local type
-	type=$(sed --unbuffered -n 's/.*\(type:\)\([[:digit:]]*\).*$/\2/p;q' "$server_evts")
+	type=$(mptcp_lib_evts_get_info type "$server_evts")
 	if [ "$type" = "" ]
 	then
 		test_pass
@@ -479,7 +476,7 @@ test_remove()
 	ip netns exec "$ns2" ./pm_nl_ctl rem token "$client4_token" id\
 	   $invalid_id > /dev/null 2>&1
 	print_test "RM_ADDR id:${invalid_id} ns2 => ns1, invalid id"
-	type=$(sed --unbuffered -n 's/.*\(type:\)\([[:digit:]]*\).*$/\2/p;q' "$server_evts")
+	type=$(mptcp_lib_evts_get_info type "$server_evts")
 	if [ "$type" = "" ]
 	then
 		test_pass
@@ -583,19 +580,19 @@ verify_subflow_events()
 		fi
 	fi
 
-	type=$(sed --unbuffered -n 's/.*\(type:\)\([[:digit:]]*\).*$/\2/p;q' "$evt")
-	token=$(sed --unbuffered -n 's/.*\(token:\)\([[:digit:]]*\).*$/\2/p;q' "$evt")
-	family=$(sed --unbuffered -n 's/.*\(family:\)\([[:digit:]]*\).*$/\2/p;q' "$evt")
-	dport=$(sed --unbuffered -n 's/.*\(dport:\)\([[:digit:]]*\).*$/\2/p;q' "$evt")
-	locid=$(sed --unbuffered -n 's/.*\(loc_id:\)\([[:digit:]]*\).*$/\2/p;q' "$evt")
-	remid=$(sed --unbuffered -n 's/.*\(rem_id:\)\([[:digit:]]*\).*$/\2/p;q' "$evt")
+	type=$(mptcp_lib_evts_get_info type "$evt" $e_type)
+	token=$(mptcp_lib_evts_get_info token "$evt" $e_type)
+	family=$(mptcp_lib_evts_get_info family "$evt" $e_type)
+	dport=$(mptcp_lib_evts_get_info dport "$evt" $e_type)
+	locid=$(mptcp_lib_evts_get_info loc_id "$evt" $e_type)
+	remid=$(mptcp_lib_evts_get_info rem_id "$evt" $e_type)
 	if [ "$family" = "$AF_INET6" ]
 	then
-		saddr=$(sed --unbuffered -n 's/.*\(saddr6:\)\([0-9a-f:.]*\).*$/\2/p;q' "$evt")
-		daddr=$(sed --unbuffered -n 's/.*\(daddr6:\)\([0-9a-f:.]*\).*$/\2/p;q' "$evt")
+		saddr=$(mptcp_lib_evts_get_info saddr6 "$evt" $e_type)
+		daddr=$(mptcp_lib_evts_get_info daddr6 "$evt" $e_type)
 	else
-		saddr=$(sed --unbuffered -n 's/.*\(saddr4:\)\([0-9.]*\).*$/\2/p;q' "$evt")
-		daddr=$(sed --unbuffered -n 's/.*\(daddr4:\)\([0-9.]*\).*$/\2/p;q' "$evt")
+		saddr=$(mptcp_lib_evts_get_info saddr4 "$evt" $e_type)
+		daddr=$(mptcp_lib_evts_get_info daddr4 "$evt" $e_type)
 	fi
 
 	check_expected "type" "token" "daddr" "dport" "family" "saddr" "locid" "remid"
@@ -630,7 +627,7 @@ test_subflows()
 	kill_wait $listener_pid
 
 	local sport
-	sport=$(sed --unbuffered -n 's/.*\(sport:\)\([[:digit:]]*\).*$/\2/p;q' "$server_evts")
+	sport=$(mptcp_lib_evts_get_info sport "$server_evts" $SUB_ESTABLISHED)
 
 	# DESTROY_SUBFLOW from server to client machine
 	:>"$server_evts"
@@ -668,7 +665,7 @@ test_subflows()
 	# Delete the listener from the client ns, if one was created
 	kill_wait $listener_pid
 
-	sport=$(sed --unbuffered -n 's/.*\(sport:\)\([[:digit:]]*\).*$/\2/p;q' "$server_evts")
+	sport=$(mptcp_lib_evts_get_info sport "$server_evts" $SUB_ESTABLISHED)
 
 	# DESTROY_SUBFLOW6 from server to client machine
 	:>"$server_evts"
@@ -707,7 +704,7 @@ test_subflows()
 	# Delete the listener from the client ns, if one was created
 	kill_wait $listener_pid
 
-	sport=$(sed --unbuffered -n 's/.*\(sport:\)\([[:digit:]]*\).*$/\2/p;q' "$server_evts")
+	sport=$(mptcp_lib_evts_get_info sport "$server_evts" $SUB_ESTABLISHED)
 
 	# DESTROY_SUBFLOW from server to client machine
 	:>"$server_evts"
@@ -745,7 +742,7 @@ test_subflows()
 	# Delete the listener from the server ns, if one was created
 	kill_wait $listener_pid
 
-	sport=$(sed --unbuffered -n 's/.*\(sport:\)\([[:digit:]]*\).*$/\2/p;q' "$client_evts")
+	sport=$(mptcp_lib_evts_get_info sport "$client_evts" $SUB_ESTABLISHED)
 
 	# DESTROY_SUBFLOW from client to server machine
 	:>"$client_evts"
@@ -784,7 +781,7 @@ test_subflows()
 	# Delete the listener from the server ns, if one was created
 	kill_wait $listener_pid
 
-	sport=$(sed --unbuffered -n 's/.*\(sport:\)\([[:digit:]]*\).*$/\2/p;q' "$client_evts")
+	sport=$(mptcp_lib_evts_get_info sport "$client_evts" $SUB_ESTABLISHED)
 
 	# DESTROY_SUBFLOW6 from client to server machine
 	:>"$client_evts"
@@ -821,7 +818,7 @@ test_subflows()
 	# Delete the listener from the server ns, if one was created
 	kill_wait $listener_pid
 
-	sport=$(sed --unbuffered -n 's/.*\(sport:\)\([[:digit:]]*\).*$/\2/p;q' "$client_evts")
+	sport=$(mptcp_lib_evts_get_info sport "$client_evts" $SUB_ESTABLISHED)
 
 	# DESTROY_SUBFLOW from client to server machine
 	:>"$client_evts"
@@ -867,7 +864,7 @@ test_subflows_v4_v6_mix()
 	# Delete the listener from the server ns, if one was created
 	kill_wait $listener_pid
 
-	sport=$(sed --unbuffered -n 's/.*\(sport:\)\([[:digit:]]*\).*$/\2/p;q' "$client_evts")
+	sport=$(mptcp_lib_evts_get_info sport "$client_evts" $SUB_ESTABLISHED)
 
 	# DESTROY_SUBFLOW from client to server machine
 	:>"$client_evts"
@@ -933,18 +930,13 @@ verify_listener_events()
 		print_test "CLOSE_LISTENER $e_saddr:$e_sport"
 	fi
 
-	type=$(grep "type:$e_type," $evt |
-	       sed --unbuffered -n 's/.*\(type:\)\([[:digit:]]*\).*$/\2/p;q')
-	family=$(grep "type:$e_type," $evt |
-		 sed --unbuffered -n 's/.*\(family:\)\([[:digit:]]*\).*$/\2/p;q')
-	sport=$(grep "type:$e_type," $evt |
-		sed --unbuffered -n 's/.*\(sport:\)\([[:digit:]]*\).*$/\2/p;q')
+	type=$(mptcp_lib_evts_get_info type $evt $e_type)
+	family=$(mptcp_lib_evts_get_info family $evt $e_type)
+	sport=$(mptcp_lib_evts_get_info sport $evt $e_type)
 	if [ $family ] && [ $family = $AF_INET6 ]; then
-		saddr=$(grep "type:$e_type," $evt |
-			sed --unbuffered -n 's/.*\(saddr6:\)\([0-9a-f:.]*\).*$/\2/p;q')
+		saddr=$(mptcp_lib_evts_get_info saddr6 $evt $e_type)
 	else
-		saddr=$(grep "type:$e_type," $evt |
-			sed --unbuffered -n 's/.*\(saddr4:\)\([0-9.]*\).*$/\2/p;q')
+		saddr=$(mptcp_lib_evts_get_info saddr4 $evt $e_type)
 	fi
 
 	check_expected "type" "family" "saddr" "sport"
-- 
2.35.3

[PATCH mptcp-next v9 2/9] selftests: mptcp: update userspace pm test helpers

[Geliang Tang]

This patch adds a new argument namespace to userspace_pm_add_addr() and
userspace_pm_add_sf() to make these two helper more versatile.

Add two more versatile helpers for userspace pm remove subflow or address:
userspace_pm_rm_addr() and userspace_pm_rm_sf(). The original test helpers
userspace_pm_rm_sf_addr_ns1() and userspace_pm_rm_sf_addr_ns2() can be
replaced by these new helpers.

Signed-off-by: Geliang Tang <geliang.tang@suse.com>
---
 .../testing/selftests/net/mptcp/mptcp_join.sh | 89 +++++++++----------
 .../testing/selftests/net/mptcp/mptcp_lib.sh  |  2 +
 .../selftests/net/mptcp/userspace_pm.sh       |  1 -
 3 files changed, 46 insertions(+), 46 deletions(-)

diff --git a/tools/testing/selftests/net/mptcp/mptcp_join.sh b/tools/testing/selftests/net/mptcp/mptcp_join.sh
index 27517ba770bc..2ebe1692edeb 100755
--- a/tools/testing/selftests/net/mptcp/mptcp_join.sh
+++ b/tools/testing/selftests/net/mptcp/mptcp_join.sh
@@ -3252,69 +3252,66 @@ fail_tests()
 	fi
 }
 
+# $1: ns ; $2: addr ; $3: id
 userspace_pm_add_addr()
 {
-	local addr=$1
-	local id=$2
+	local evts=$evts_ns1
 	local tk
 
-	tk=$(grep "type:1," "$evts_ns1" |
-	     sed -n 's/.*\(token:\)\([[:digit:]]*\).*$/\2/p;q')
-	ip netns exec $ns1 ./pm_nl_ctl ann $addr token $tk id $id
+	[ "$1" == "$ns2" ] && evts=$evts_ns2
+	tk=$(mptcp_lib_evts_get_info token "$evts")
+
+	ip netns exec $1 ./pm_nl_ctl ann $2 token $tk id $3
 	sleep 1
 }
 
-userspace_pm_rm_sf_addr_ns1()
+# $1: ns ; $2: id
+userspace_pm_rm_addr()
 {
-	local addr=$1
-	local id=$2
-	local tk sp da dp
-
-	tk=$(grep "type:1," "$evts_ns1" |
-	     sed -n 's/.*\(token:\)\([[:digit:]]*\).*$/\2/p;q')
-	sp=$(grep "type:10" "$evts_ns1" |
-	     sed -n 's/.*\(sport:\)\([[:digit:]]*\).*$/\2/p;q')
-	da=$(grep "type:10" "$evts_ns1" |
-	     sed -n 's/.*\(daddr6:\)\([0-9a-f:.]*\).*$/\2/p;q')
-	dp=$(grep "type:10" "$evts_ns1" |
-	     sed -n 's/.*\(dport:\)\([[:digit:]]*\).*$/\2/p;q')
-	ip netns exec $ns1 ./pm_nl_ctl rem token $tk id $id
-	ip netns exec $ns1 ./pm_nl_ctl dsf lip "::ffff:$addr" \
-				lport $sp rip $da rport $dp token $tk
-	wait_rm_addr $ns1 1
-	wait_rm_sf $ns1 1
+	local evts=$evts_ns1
+	local tk
+
+	[ "$1" == "$ns2" ] && evts=$evts_ns2
+	tk=$(mptcp_lib_evts_get_info token "$evts")
+
+	ip netns exec $1 ./pm_nl_ctl rem token $tk id $2
+	wait_rm_addr $1 1
 }
 
+# $1: ns ; $2: addr ; $3: id
 userspace_pm_add_sf()
 {
-	local addr=$1
-	local id=$2
+	local evts=$evts_ns1
 	local tk da dp
 
-	tk=$(sed -n 's/.*\(token:\)\([[:digit:]]*\).*$/\2/p;q' "$evts_ns2")
-	da=$(sed -n 's/.*\(daddr4:\)\([0-9.]*\).*$/\2/p;q' "$evts_ns2")
-	dp=$(sed -n 's/.*\(dport:\)\([[:digit:]]*\).*$/\2/p;q' "$evts_ns2")
-	ip netns exec $ns2 ./pm_nl_ctl csf lip $addr lid $id \
+	[ "$1" == "$ns2" ] && evts=$evts_ns2
+	tk=$(mptcp_lib_evts_get_info token "$evts")
+	da=$(mptcp_lib_evts_get_info daddr4 "$evts")
+	dp=$(mptcp_lib_evts_get_info dport "$evts")
+
+	ip netns exec $1 ./pm_nl_ctl csf lip $2 lid $3 \
 				rip $da rport $dp token $tk
 	sleep 1
 }
 
-userspace_pm_rm_sf_addr_ns2()
+# $1: ns ; $2: addr $3: event type
+userspace_pm_rm_sf()
 {
-	local addr=$1
-	local id=$2
+	local evts=$evts_ns1
+	local t=${3:-1}
+	local ip=4
 	local tk da dp sp
 
-	tk=$(sed -n 's/.*\(token:\)\([[:digit:]]*\).*$/\2/p;q' "$evts_ns2")
-	da=$(sed -n 's/.*\(daddr4:\)\([0-9.]*\).*$/\2/p;q' "$evts_ns2")
-	dp=$(sed -n 's/.*\(dport:\)\([[:digit:]]*\).*$/\2/p;q' "$evts_ns2")
-	sp=$(grep "type:10" "$evts_ns2" |
-	     sed -n 's/.*\(sport:\)\([[:digit:]]*\).*$/\2/p;q')
-	ip netns exec $ns2 ./pm_nl_ctl rem token $tk id $id
-	ip netns exec $ns2 ./pm_nl_ctl dsf lip $addr lport $sp \
+	[ "$1" == "$ns2" ] && evts=$evts_ns2
+	if is_v6 $2; then ip=6; fi
+	tk=$(mptcp_lib_evts_get_info token "$evts")
+	da=$(mptcp_lib_evts_get_info "daddr$ip" "$evts" $t)
+	dp=$(mptcp_lib_evts_get_info dport "$evts" $t)
+	sp=$(mptcp_lib_evts_get_info sport "$evts" $t)
+
+	ip netns exec $1 ./pm_nl_ctl dsf lip $2 lport $sp \
 				rip $da rport $dp token $tk
-	wait_rm_addr $ns2 1
-	wait_rm_sf $ns2 1
+	wait_rm_sf $1 1
 }
 
 userspace_tests()
@@ -3401,12 +3398,13 @@ userspace_tests()
 			run_tests $ns1 $ns2 10.0.1.1 &
 		local tests_pid=$!
 		wait_mpj $ns1
-		userspace_pm_add_addr 10.0.2.1 10
+		userspace_pm_add_addr $ns1 10.0.2.1 10
 		chk_join_nr 1 1 1
 		chk_add_nr 1 1
 		chk_mptcp_info subflows 1 subflows 1
 		chk_mptcp_info add_addr_signal 1 add_addr_accepted 1
-		userspace_pm_rm_sf_addr_ns1 10.0.2.1 10
+		userspace_pm_rm_addr $ns1 10
+		userspace_pm_rm_sf $ns1 "::ffff:10.0.2.1" $SUB_ESTABLISHED
 		chk_rm_nr 1 1 invert
 		chk_mptcp_info subflows 0 subflows 0
 		kill_events_pids
@@ -3422,10 +3420,11 @@ userspace_tests()
 			run_tests $ns1 $ns2 10.0.1.1 &
 		local tests_pid=$!
 		wait_mpj $ns2
-		userspace_pm_add_sf 10.0.3.2 20
+		userspace_pm_add_sf $ns2 10.0.3.2 20
 		chk_join_nr 1 1 1
 		chk_mptcp_info subflows 1 subflows 1
-		userspace_pm_rm_sf_addr_ns2 10.0.3.2 20
+		userspace_pm_rm_addr $ns2 20
+		userspace_pm_rm_sf $ns2 10.0.3.2 $SUB_ESTABLISHED
 		chk_rm_nr 1 1
 		chk_mptcp_info subflows 0 subflows 0
 		kill_events_pids
diff --git a/tools/testing/selftests/net/mptcp/mptcp_lib.sh b/tools/testing/selftests/net/mptcp/mptcp_lib.sh
index def35395a254..bb95dd967eb3 100644
--- a/tools/testing/selftests/net/mptcp/mptcp_lib.sh
+++ b/tools/testing/selftests/net/mptcp/mptcp_lib.sh
@@ -8,6 +8,8 @@ readonly KSFT_SKIP=4
 # shellcheck disable=SC2155 # declare and assign separately
 readonly KSFT_TEST=$(basename "${0}" | sed 's/\.sh$//g')
 
+SUB_ESTABLISHED=10 # MPTCP_EVENT_SUB_ESTABLISHED
+
 MPTCP_LIB_SUBTESTS=()
 
 # only if supported (or forced) and not disabled, see no-color.org
diff --git a/tools/testing/selftests/net/mptcp/userspace_pm.sh b/tools/testing/selftests/net/mptcp/userspace_pm.sh
index 2413059a42e5..283c62deb628 100755
--- a/tools/testing/selftests/net/mptcp/userspace_pm.sh
+++ b/tools/testing/selftests/net/mptcp/userspace_pm.sh
@@ -25,7 +25,6 @@ fi
 
 ANNOUNCED=6        # MPTCP_EVENT_ANNOUNCED
 REMOVED=7          # MPTCP_EVENT_REMOVED
-SUB_ESTABLISHED=10 # MPTCP_EVENT_SUB_ESTABLISHED
 SUB_CLOSED=11      # MPTCP_EVENT_SUB_CLOSED
 LISTENER_CREATED=15 #MPTCP_EVENT_LISTENER_CREATED
 LISTENER_CLOSED=16  #MPTCP_EVENT_LISTENER_CLOSED
-- 
2.35.3

[PATCH mptcp-next v9 3/9] selftests: mptcp: userspace pm remove id 0 subflow

[Geliang Tang]

This patch adds a selftest for userpsace PM to remove id 0 subflow. Use
userspace_pm_add_sf() to add a subflow, and pass initial ip address to
userspace_pm_rm_sf() to remove id 0 subflow.

Add a new helper chk_all_subflows(), in it use 'ss' command

        ss -ti | grep -c tcp-ulp-mptcp

to get the "correct" amount of subflows, including the initial one.

When closing the initial subflow in __mptcp_close_ssk(), dispose_it is
false, then tcp_disconnect is invoked. This will send a MP_RST to close
a subflow on the peer too. So chk_rst_nr() is added in this test, and
chk_all_subflows after closing the initial subflow is '1 1', not '2 1'.

Signed-off-by: Geliang Tang <geliang.tang@suse.com>
---
 .../testing/selftests/net/mptcp/mptcp_join.sh | 47 +++++++++++++++++++
 1 file changed, 47 insertions(+)

diff --git a/tools/testing/selftests/net/mptcp/mptcp_join.sh b/tools/testing/selftests/net/mptcp/mptcp_join.sh
index 2ebe1692edeb..966977980d5c 100755
--- a/tools/testing/selftests/net/mptcp/mptcp_join.sh
+++ b/tools/testing/selftests/net/mptcp/mptcp_join.sh
@@ -1881,6 +1881,31 @@ chk_mptcp_info()
 	fi
 }
 
+# $1: subflows in ns1 ; $2: subflows in ns2
+# number of all subflows, including the initial subflow.
+chk_all_subflows()
+{
+	local cnt1
+	local cnt2
+
+	print_check "all subflows $1:$2"
+
+	cnt1=$(ss -N $ns1 -ti | grep -c tcp-ulp-mptcp)
+	cnt2=$(ss -N $ns2 -ti | grep -c tcp-ulp-mptcp)
+
+	if [ "$1" != "$cnt1" ] || [ "$2" != "$cnt2" ]; then
+		fail_test "got subflows $cnt1:$cnt2 expected $1:$2"
+		dump_stats=1
+	else
+		print_ok
+	fi
+
+	if [ "$dump_stats" = 1 ]; then
+		ss -N $ns1 -ti
+		ss -N $ns2 -ti
+	fi
+}
+
 chk_link_usage()
 {
 	local ns=$1
@@ -3430,6 +3455,28 @@ userspace_tests()
 		kill_events_pids
 		wait $tests_pid
 	fi
+
+	# userspace pm remove id 0 subflow
+	if reset_with_events "userspace pm remove id 0 subflow" &&
+	   continue_if mptcp_lib_has_file '/proc/sys/net/mptcp/pm_type'; then
+		set_userspace_pm $ns2
+		pm_nl_set_limits $ns1 0 1
+		speed=10 \
+			run_tests $ns1 $ns2 10.0.1.1 &
+		local tests_pid=$!
+		wait_mpj $ns2
+		userspace_pm_add_sf $ns2 10.0.3.2 20
+		chk_join_nr 1 1 1
+		chk_mptcp_info subflows 1 subflows 1
+		chk_all_subflows 2 2
+		userspace_pm_rm_sf $ns2 10.0.1.2
+		chk_rm_nr 0 1
+		chk_rst_nr 1 1 invert
+		chk_mptcp_info subflows 1 subflows 1
+		chk_all_subflows 1 1
+		kill_events_pids
+		wait $tests_pid
+	fi
 }
 
 endpoint_tests()
-- 
2.35.3

[PATCH mptcp-next v9 4/9] mptcp: userspace pm allow creating id 0 subflow

[Geliang Tang]

This patch drops id 0 limitation in mptcp_nl_cmd_sf_create() to allow
creating additional subflows with the local addr ID 0.

There is no reason not to allow additional subflows from this local
address: we should be able to create new subflows from the initial
endpoint. This limitation was breaking fullmesh support from userspace.

Fixes: 702c2f646d42 ("mptcp: netlink: allow userspace-driven subflow establishment")
Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/391
Suggested-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: Geliang Tang <geliang.tang@suse.com>
---
 net/mptcp/pm_userspace.c | 6 ------
 1 file changed, 6 deletions(-)

diff --git a/net/mptcp/pm_userspace.c b/net/mptcp/pm_userspace.c
index b5a8aa4c1ebd..d042d32beb4d 100644
--- a/net/mptcp/pm_userspace.c
+++ b/net/mptcp/pm_userspace.c
@@ -307,12 +307,6 @@ int mptcp_nl_cmd_sf_create(struct sk_buff *skb, struct genl_info *info)
 		goto create_err;
 	}
 
-	if (addr_l.id == 0) {
-		NL_SET_ERR_MSG_ATTR(info->extack, laddr, "missing local addr id");
-		err = -EINVAL;
-		goto create_err;
-	}
-
 	err = mptcp_pm_parse_addr(raddr, info, &addr_r);
 	if (err < 0) {
 		NL_SET_ERR_MSG_ATTR(info->extack, raddr, "error parsing remote addr");
-- 
2.35.3

[PATCH mptcp-next v9 5/9] selftests: mptcp: userspace pm create id 0 subflow

[Geliang Tang]

This patch adds a selftest to create id 0 subflow. Pass id 0 to the
helper userspace_pm_add_sf() to create id 0 subflow. chk_mptcp_info
shows one subflow but chk_all_subflows shows two subflows in each
namespace.

Signed-off-by: Geliang Tang <geliang.tang@suse.com>
---
 .../testing/selftests/net/mptcp/mptcp_join.sh | 19 +++++++++++++++++++
 1 file changed, 19 insertions(+)

diff --git a/tools/testing/selftests/net/mptcp/mptcp_join.sh b/tools/testing/selftests/net/mptcp/mptcp_join.sh
index 966977980d5c..f1a7e04815b1 100755
--- a/tools/testing/selftests/net/mptcp/mptcp_join.sh
+++ b/tools/testing/selftests/net/mptcp/mptcp_join.sh
@@ -3456,6 +3456,25 @@ userspace_tests()
 		wait $tests_pid
 	fi
 
+	# userspace pm create id 0 subflow
+	if reset_with_events "userspace pm create id 0 subflow" &&
+	   continue_if mptcp_lib_has_file '/proc/sys/net/mptcp/pm_type'; then
+		set_userspace_pm $ns2
+		pm_nl_set_limits $ns1 0 1
+		speed=10 \
+			run_tests $ns1 $ns2 10.0.1.1 &
+		local tests_pid=$!
+		wait_mpj $ns2
+		chk_mptcp_info subflows 0 subflows 0
+		chk_all_subflows 1 1
+		userspace_pm_add_sf $ns2 10.0.3.2 0
+		chk_join_nr 1 1 1
+		chk_mptcp_info subflows 1 subflows 1
+		chk_all_subflows 2 2
+		kill_events_pids
+		wait $tests_pid
+	fi
+
 	# userspace pm remove id 0 subflow
 	if reset_with_events "userspace pm remove id 0 subflow" &&
 	   continue_if mptcp_lib_has_file '/proc/sys/net/mptcp/pm_type'; then
-- 
2.35.3

[PATCH mptcp-next v9 6/9] mptcp: add no_initial_subflow mptcpi_flags

[Geliang Tang]

If the initial subflow has been removed, we cannot know without checking
other counters, e.g. ss -ti <filter> | grep -c tcp-ulp-mptcp or
getsockopt(SOL_MPTCP, MPTCP_FULL_INFO, ...) (or others except MPTCP_INFO
of course) and then check mptcp_subflow_data->num_subflows to get the
total amount of subflows.

This patch adds a new flag NO_INITIAL_SUBFLOW in mptcpi_flags to know if
the initial subflow has been removed.

Add a new helper __mptcp_check_initial_subflow() to check whether the
initial subflow is closing or closed.

Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/428
Signed-off-by: Geliang Tang <geliang.tang@suse.com>
---
 include/uapi/linux/mptcp.h | 1 +
 net/mptcp/pm_netlink.c     | 3 ++-
 net/mptcp/protocol.h       | 5 +++++
 net/mptcp/sockopt.c        | 2 ++
 4 files changed, 10 insertions(+), 1 deletion(-)

diff --git a/include/uapi/linux/mptcp.h b/include/uapi/linux/mptcp.h
index ee9c49f949a2..c92421866aaf 100644
--- a/include/uapi/linux/mptcp.h
+++ b/include/uapi/linux/mptcp.h
@@ -107,6 +107,7 @@ enum {
 
 #define MPTCP_INFO_FLAG_FALLBACK		_BITUL(0)
 #define MPTCP_INFO_FLAG_REMOTE_KEY_RECEIVED	_BITUL(1)
+#define MPTCP_INFO_FLAG_NO_INITIAL_SUBFLOW	_BITUL(2)
 
 struct mptcp_info {
 	__u8	mptcpi_subflows;
diff --git a/net/mptcp/pm_netlink.c b/net/mptcp/pm_netlink.c
index 9661f3812682..4b196235c058 100644
--- a/net/mptcp/pm_netlink.c
+++ b/net/mptcp/pm_netlink.c
@@ -1462,7 +1462,8 @@ static int mptcp_nl_remove_id_zero_address(struct net *net,
 		struct sock *sk = (struct sock *)msk;
 		struct mptcp_addr_info msk_local;
 
-		if (list_empty(&msk->conn_list) || mptcp_pm_is_userspace(msk))
+		if (list_empty(&msk->conn_list) || mptcp_pm_is_userspace(msk) ||
+		    !__mptcp_check_initial_subflow(msk))
 			goto next;
 
 		mptcp_local_address((struct sock_common *)msk, &msk_local);
diff --git a/net/mptcp/protocol.h b/net/mptcp/protocol.h
index 3c938e3560e4..96e3846b50fe 100644
--- a/net/mptcp/protocol.h
+++ b/net/mptcp/protocol.h
@@ -1031,6 +1031,11 @@ static inline void __mptcp_do_fallback(struct mptcp_sock *msk)
 	set_bit(MPTCP_FALLBACK_DONE, &msk->flags);
 }
 
+static inline bool __mptcp_check_initial_subflow(const struct mptcp_sock *msk)
+{
+	return msk->first && inet_sk_state_load(msk->first) != TCP_CLOSE;
+}
+
 static inline void mptcp_do_fallback(struct sock *ssk)
 {
 	struct mptcp_subflow_context *subflow = mptcp_subflow_ctx(ssk);
diff --git a/net/mptcp/sockopt.c b/net/mptcp/sockopt.c
index 8260202c0066..a2e1c3d159d6 100644
--- a/net/mptcp/sockopt.c
+++ b/net/mptcp/sockopt.c
@@ -919,6 +919,8 @@ void mptcp_diag_fill_info(struct mptcp_sock *msk, struct mptcp_info *info)
 		flags |= MPTCP_INFO_FLAG_FALLBACK;
 	if (READ_ONCE(msk->can_ack))
 		flags |= MPTCP_INFO_FLAG_REMOTE_KEY_RECEIVED;
+	if (!__mptcp_check_initial_subflow(msk))
+		flags |= MPTCP_INFO_FLAG_NO_INITIAL_SUBFLOW;
 	info->mptcpi_flags = flags;
 	mptcp_data_lock(sk);
 	info->mptcpi_snd_una = msk->snd_una;
-- 
2.35.3

[PATCH mptcp-next v9 7/9] mptcp: userspace pm remove id 0 address

[Geliang Tang]

This patch adds the ability to send RM_ADDR for local ID 0. Check
whether id 0 address is removed, if not, put id 0 into a removing
list, pass it to mptcp_pm_remove_addr() to remove id 0 address.

There is no reason not to allow the userspace to remove the initial
address (ID 0). This special case was not taken into account not
letting the userspace to delete all addresses as announced.

Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/379
Fixes: d9a4594edabf ("mptcp: netlink: Add MPTCP_PM_CMD_REMOVE")
Signed-off-by: Geliang Tang <geliang.tang@suse.com>
---
 net/mptcp/pm_userspace.c | 27 +++++++++++++++++++++++++++
 1 file changed, 27 insertions(+)

diff --git a/net/mptcp/pm_userspace.c b/net/mptcp/pm_userspace.c
index d042d32beb4d..7bbfef98b6c7 100644
--- a/net/mptcp/pm_userspace.c
+++ b/net/mptcp/pm_userspace.c
@@ -208,6 +208,30 @@ int mptcp_nl_cmd_announce(struct sk_buff *skb, struct genl_info *info)
 	return err;
 }
 
+static int mptcp_userspace_remove_id_zero_address(struct mptcp_sock *msk,
+						  struct genl_info *info)
+{
+	struct mptcp_rm_list list = { .nr = 0 };
+	struct sock *sk = (struct sock *)msk;
+	int err = -EINVAL;
+
+	lock_sock(sk);
+	spin_lock_bh(&msk->pm.lock);
+	if (!__mptcp_check_initial_subflow(msk)) {
+		GENL_SET_ERR_MSG(info, "address with id 0 not found");
+		goto out;
+	}
+
+	list.ids[list.nr++] = 0;
+	mptcp_pm_remove_addr(msk, &list);
+	err = 0;
+out:
+	spin_unlock_bh(&msk->pm.lock);
+	release_sock(sk);
+	sock_put(sk);
+	return err;
+}
+
 int mptcp_nl_cmd_remove(struct sk_buff *skb, struct genl_info *info)
 {
 	struct nlattr *token = info->attrs[MPTCP_PM_ATTR_TOKEN];
@@ -239,6 +263,9 @@ int mptcp_nl_cmd_remove(struct sk_buff *skb, struct genl_info *info)
 		goto remove_err;
 	}
 
+	if (id_val == 0)
+		return mptcp_userspace_remove_id_zero_address(msk, info);
+
 	lock_sock((struct sock *)msk);
 
 	list_for_each_entry(entry, &msk->pm.userspace_pm_local_addr_list, list) {
-- 
2.35.3

[PATCH mptcp-next v9 8/9] selftests: mptcp: userspace pm remove id 0 address

[Geliang Tang]

This patch adds a selftest for userpsace PM to remove id 0 address.
Use userspace_pm_add_addr() helper to add a id 10 address, then use
userspace_pm_rm_addr() helper to remove id 0 address.

Signed-off-by: Geliang Tang <geliang.tang@suse.com>
---
 .../testing/selftests/net/mptcp/mptcp_join.sh | 24 +++++++++++++++++++
 1 file changed, 24 insertions(+)

diff --git a/tools/testing/selftests/net/mptcp/mptcp_join.sh b/tools/testing/selftests/net/mptcp/mptcp_join.sh
index f1a7e04815b1..19a6859bd207 100755
--- a/tools/testing/selftests/net/mptcp/mptcp_join.sh
+++ b/tools/testing/selftests/net/mptcp/mptcp_join.sh
@@ -3496,6 +3496,30 @@ userspace_tests()
 		kill_events_pids
 		wait $tests_pid
 	fi
+
+	# userspace pm remove id 0 address
+	if reset_with_events "userspace pm remove id 0 address" &&
+	   continue_if mptcp_lib_has_file '/proc/sys/net/mptcp/pm_type'; then
+		set_userspace_pm $ns1
+		pm_nl_set_limits $ns2 1 1
+		speed=10 \
+			run_tests $ns1 $ns2 10.0.1.1 &
+		local tests_pid=$!
+		wait_mpj $ns1
+		userspace_pm_add_addr $ns1 10.0.2.1 10
+		chk_join_nr 1 1 1
+		chk_add_nr 1 1
+		chk_mptcp_info subflows 1 subflows 1
+		chk_all_subflows 2 2
+		chk_mptcp_info add_addr_signal 1 add_addr_accepted 1
+		userspace_pm_rm_addr $ns1 0
+		chk_rm_nr 1 0 invert
+		chk_rst_nr 1 1 invert
+		chk_mptcp_info subflows 1 subflows 1
+		chk_all_subflows 1 1
+		kill_events_pids
+		wait $tests_pid
+	fi
 }
 
 endpoint_tests()
-- 
2.35.3

[PATCH mptcp-next v9 9/9] selftests: mptcp: check no_initial_subflow flag

[Geliang Tang]

Add a new argument all for chk_mptcp_info() to get the "correct" amount
of subflows, including the initial one, if it's set as "all".

New version of ss will test the no_initial_subflow flag to count number
of all subflows.

With this new flag, we can then compute the total amount of subflows
from mptcp_info by doing something like:

  if (mptcpi_flags & MPTCP_INFO_FLAG_HAS_INITIAL_SUBFLOW)
          mptcpi_subflows += 1;

Signed-off-by: Geliang Tang <geliang.tang@suse.com>
---
 .../testing/selftests/net/mptcp/mptcp_join.sh | 23 ++++++++++++++++++-
 1 file changed, 22 insertions(+), 1 deletion(-)

diff --git a/tools/testing/selftests/net/mptcp/mptcp_join.sh b/tools/testing/selftests/net/mptcp/mptcp_join.sh
index 19a6859bd207..29250f5729ef 100755
--- a/tools/testing/selftests/net/mptcp/mptcp_join.sh
+++ b/tools/testing/selftests/net/mptcp/mptcp_join.sh
@@ -1856,14 +1856,23 @@ chk_mptcp_info()
 	local exp1=$2
 	local info2=$3
 	local exp2=$4
+	local all=${5:-""}
 	local cnt1
 	local cnt2
 	local dump_stats
 
-	print_check "mptcp_info ${info1:0:8}=$exp1:$exp2"
+	print_check "mptcp_info $all ${info1:0:8}=$exp1:$exp2"
 
 	cnt1=$(ss -N $ns1 -inmHM | mptcp_lib_get_info_value "$info1" "$info1")
 	cnt2=$(ss -N $ns2 -inmHM | mptcp_lib_get_info_value "$info2" "$info2")
+	if [ "$all" == "all" ]; then
+		local flag1=0 flag2=0
+
+		flag1=$(ss -N $ns1 -inmHM | grep -c "no_initial_subflow")
+		flag2=$(ss -N $ns2 -inmHM | grep -c "no_initial_subflow")
+		[ $info1 == "subflows" ] && [ $flag1 == 0 ] && cnt1=$((cnt1+1))
+		[ $info2 == "subflows" ] && [ $flag2 == 0 ] && cnt2=$((cnt2+1))
+	fi
 	# 'ss' only display active connections and counters that are not 0.
 	[ -z "$cnt1" ] && cnt1=0
 	[ -z "$cnt2" ] && cnt2=0
@@ -1881,13 +1890,25 @@ chk_mptcp_info()
 	fi
 }
 
+ver()
+{
+	printf "%02d%02d%02d%02d" ${1//./ }
+}
+
 # $1: subflows in ns1 ; $2: subflows in ns2
 # number of all subflows, including the initial subflow.
 chk_all_subflows()
 {
+	local ver_iproute2
 	local cnt1
 	local cnt2
 
+	ver_iproute2=$(ss -v |
+		 sed -n 's/.*\(iproute2-\)\([0-9].[0-9].[0-9]\).*$/\2/p')
+	(( $(ver $ver_iproute2) > $(ver 6.5.0) )) && \
+		chk_mptcp_info subflows $1 subflows $2 all && \
+		return
+
 	print_check "all subflows $1:$2"
 
 	cnt1=$(ss -N $ns1 -ti | grep -c tcp-ulp-mptcp)
-- 
2.35.3

Re: [PATCH mptcp-next v9 6/9] mptcp: add no_initial_subflow mptcpi_flags

[Matthieu Baerts]

Hi Geliang,

On 14/09/2023 16:50, Geliang Tang wrote:
> If the initial subflow has been removed, we cannot know without checking
> other counters, e.g. ss -ti <filter> | grep -c tcp-ulp-mptcp or
> getsockopt(SOL_MPTCP, MPTCP_FULL_INFO, ...) (or others except MPTCP_INFO
> of course) and then check mptcp_subflow_data->num_subflows to get the
> total amount of subflows.
> 
> This patch adds a new flag NO_INITIAL_SUBFLOW in mptcpi_flags to know if
> the initial subflow has been removed.
> 
> Add a new helper __mptcp_check_initial_subflow() to check whether the
> initial subflow is closing or closed.
> 
> Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/428
> Signed-off-by: Geliang Tang <geliang.tang@suse.com>
> ---
>  include/uapi/linux/mptcp.h | 1 +
>  net/mptcp/pm_netlink.c     | 3 ++-
>  net/mptcp/protocol.h       | 5 +++++
>  net/mptcp/sockopt.c        | 2 ++
>  4 files changed, 10 insertions(+), 1 deletion(-)
> 
> diff --git a/include/uapi/linux/mptcp.h b/include/uapi/linux/mptcp.h
> index ee9c49f949a2..c92421866aaf 100644
> --- a/include/uapi/linux/mptcp.h
> +++ b/include/uapi/linux/mptcp.h
> @@ -107,6 +107,7 @@ enum {
>  
>  #define MPTCP_INFO_FLAG_FALLBACK		_BITUL(0)
>  #define MPTCP_INFO_FLAG_REMOTE_KEY_RECEIVED	_BITUL(1)
> +#define MPTCP_INFO_FLAG_NO_INITIAL_SUBFLOW	_BITUL(2)

With only one flag, the userspace cannot know if the kernel supports
this feature or not: if you are on an older kernel, this flag will not
be set if the initial subflow has been removed. It means that the
userspace cannot be sure about the total number of subflows. We might
need to set another flag is the initial subflow is there.

But then, instead of adding 2 flags, I wonder if it would not be better
to add a new counter (e.g. 'mptcpi_total_subflows'): we would avoid
confusions with 'mptcpi_subflows' which doesn't take into account the
initial subflow.

>  
>  struct mptcp_info {
>  	__u8	mptcpi_subflows;
> diff --git a/net/mptcp/pm_netlink.c b/net/mptcp/pm_netlink.c
> index 9661f3812682..4b196235c058 100644
> --- a/net/mptcp/pm_netlink.c
> +++ b/net/mptcp/pm_netlink.c
> @@ -1462,7 +1462,8 @@ static int mptcp_nl_remove_id_zero_address(struct net *net,
>  		struct sock *sk = (struct sock *)msk;
>  		struct mptcp_addr_info msk_local;
>  
> -		if (list_empty(&msk->conn_list) || mptcp_pm_is_userspace(msk))
> +		if (list_empty(&msk->conn_list) || mptcp_pm_is_userspace(msk) ||
> +		    !__mptcp_check_initial_subflow(msk))

Again, sorry it is a quick review but why do you need to do this
modification here?

>  			goto next;
>  
>  		mptcp_local_address((struct sock_common *)msk, &msk_local);
> diff --git a/net/mptcp/protocol.h b/net/mptcp/protocol.h
> index 3c938e3560e4..96e3846b50fe 100644
> --- a/net/mptcp/protocol.h
> +++ b/net/mptcp/protocol.h
> @@ -1031,6 +1031,11 @@ static inline void __mptcp_do_fallback(struct mptcp_sock *msk)
>  	set_bit(MPTCP_FALLBACK_DONE, &msk->flags);
>  }
>  
> +static inline bool __mptcp_check_initial_subflow(const struct mptcp_sock *msk)

Maybe "__mptcp_has_initial_subflow()"?

> +{
> +	return msk->first && inet_sk_state_load(msk->first) != TCP_CLOSE;

No need to do a READ_ONCE() when called from mptcp_diag_fill_info()?

Did you check if it is normal to have 'msk->first' still set after
having removed it? Why do we keep it? (I don't remember)

Cheers,
Matt

> +}
> +
>  static inline void mptcp_do_fallback(struct sock *ssk)
>  {
>  	struct mptcp_subflow_context *subflow = mptcp_subflow_ctx(ssk);
> diff --git a/net/mptcp/sockopt.c b/net/mptcp/sockopt.c
> index 8260202c0066..a2e1c3d159d6 100644
> --- a/net/mptcp/sockopt.c
> +++ b/net/mptcp/sockopt.c
> @@ -919,6 +919,8 @@ void mptcp_diag_fill_info(struct mptcp_sock *msk, struct mptcp_info *info)
>  		flags |= MPTCP_INFO_FLAG_FALLBACK;
>  	if (READ_ONCE(msk->can_ack))
>  		flags |= MPTCP_INFO_FLAG_REMOTE_KEY_RECEIVED;
> +	if (!__mptcp_check_initial_subflow(msk))
> +		flags |= MPTCP_INFO_FLAG_NO_INITIAL_SUBFLOW;
>  	info->mptcpi_flags = flags;
>  	mptcp_data_lock(sk);
>  	info->mptcpi_snd_una = msk->snd_una;

-- 
Tessares | Belgium | Hybrid Access Solutions
www.tessares.net

Re: selftests: mptcp: check no_initial_subflow flag: Tests Results

[MPTCP CI]

Hi Geliang,

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! ✅:
  - Task: https://cirrus-ci.com/task/6408327622557696
  - Summary: https://api.cirrus-ci.com/v1/artifact/task/6408327622557696/summary/summary.txt

- KVM Validation: debug (except selftest_mptcp_join):
  - Unstable: 2 failed test(s): packetdrill_mp_join packetdrill_regressions 🔴:
  - Task: https://cirrus-ci.com/task/6126852645847040
  - Summary: https://api.cirrus-ci.com/v1/artifact/task/6126852645847040/summary/summary.txt

- KVM Validation: debug (only selftest_mptcp_join):
  - Success! ✅:
  - Task: https://cirrus-ci.com/task/5563902692425728
  - Summary: https://api.cirrus-ci.com/v1/artifact/task/5563902692425728/summary/summary.txt

- KVM Validation: normal (only selftest_mptcp_join):
  - Success! ✅:
  - Task: https://cirrus-ci.com/task/5000952739004416
  - Summary: https://api.cirrus-ci.com/v1/artifact/task/5000952739004416/summary/summary.txt

Initiator: Patchew Applier
Commits: https://github.com/multipath-tcp/mptcp_net-next/commits/644351872382


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-debug

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 (Tessares)