netdev - [PATCH net 3/3] selftests: bonding: add mux and churn state testing

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20251124043310.34073-4-liuhangbin@gmail.com>
Date: Mon, 24 Nov 2025 04:33:10 +0000
From: Hangbin Liu <liuhangbin@...il.com>
To: netdev@...r.kernel.org
Cc: Jay Vosburgh <jv@...sburgh.net>,
	Andrew Lunn <andrew+netdev@...n.ch>,
	"David S. Miller" <davem@...emloft.net>,
	Eric Dumazet <edumazet@...gle.com>,
	Jakub Kicinski <kuba@...nel.org>,
	Paolo Abeni <pabeni@...hat.com>,
	Simon Horman <horms@...nel.org>,
	Mahesh Bandewar <maheshb@...gle.com>,
	Shuah Khan <shuah@...nel.org>,
	linux-kselftest@...r.kernel.org,
	Hangbin Liu <liuhangbin@...il.com>
Subject: [PATCH net 3/3] selftests: bonding: add mux and churn state testing

Rename the current LACP priority test to LACP ad_select testing, and
extend it to include validation of the actor state machine and churn
state logic. The updated tests verify that both the mux state machine and
the churn state machine behave correctly under aggregator selection and
failover scenarios.

Signed-off-by: Hangbin Liu <liuhangbin@...il.com>
---
 .../selftests/drivers/net/bonding/Makefile    |  2 +-
 ...nd_lacp_prio.sh => bond_lacp_ad_select.sh} | 73 +++++++++++++++++++
 2 files changed, 74 insertions(+), 1 deletion(-)
 rename tools/testing/selftests/drivers/net/bonding/{bond_lacp_prio.sh => bond_lacp_ad_select.sh} (64%)

diff --git a/tools/testing/selftests/drivers/net/bonding/Makefile b/tools/testing/selftests/drivers/net/bonding/Makefile
index 6c5c60adb5e8..e7bddfbf0f7a 100644
--- a/tools/testing/selftests/drivers/net/bonding/Makefile
+++ b/tools/testing/selftests/drivers/net/bonding/Makefile
@@ -7,7 +7,7 @@ TEST_PROGS := \
 	bond-eth-type-change.sh \
 	bond-lladdr-target.sh \
 	bond_ipsec_offload.sh \
-	bond_lacp_prio.sh \
+	bond_lacp_ad_select.sh \
 	bond_macvlan_ipvlan.sh \
 	bond_options.sh \
 	bond_passive_lacp.sh \
diff --git a/tools/testing/selftests/drivers/net/bonding/bond_lacp_prio.sh b/tools/testing/selftests/drivers/net/bonding/bond_lacp_ad_select.sh
similarity index 64%
rename from tools/testing/selftests/drivers/net/bonding/bond_lacp_prio.sh
rename to tools/testing/selftests/drivers/net/bonding/bond_lacp_ad_select.sh
index a483d505c6a8..9f0b3de4f55c 100755
--- a/tools/testing/selftests/drivers/net/bonding/bond_lacp_prio.sh
+++ b/tools/testing/selftests/drivers/net/bonding/bond_lacp_ad_select.sh
@@ -89,6 +89,65 @@ test_agg_reselect()
 		RET=1
 }
 
+is_distributing()
+{
+	ip -j -n "$c_ns" -d link show "$1" \
+		| jq -e '.[].linkinfo.info_slave_data.ad_actor_oper_port_state_str | index("distributing")' > /dev/null
+}
+
+get_churn_state()
+{
+	local slave=$1
+	# shellcheck disable=SC2016
+	ip netns exec "$c_ns" awk -v s="$slave" '
+	$0 ~ "Slave Interface: " s {found=1}
+	found && /Actor Churn State:/ { print $4; exit }
+	' /proc/net/bonding/bond0
+}
+
+check_slave_state()
+{
+	local state=$1
+	local slave_0=$2
+	local slave_1=$3
+	local churn_state
+	RET=0
+
+	s0_agg_id=$(cmd_jq "ip -n ${c_ns} -d -j link show $slave_0" \
+		".[].linkinfo.info_slave_data.ad_aggregator_id")
+	s1_agg_id=$(cmd_jq "ip -n ${c_ns} -d -j link show $slave_1" \
+		".[].linkinfo.info_slave_data.ad_aggregator_id")
+	if [ "${s0_agg_id}" -ne "${s1_agg_id}" ]; then
+		log_info "$state: $slave_0 $slave_1 agg ids are different"
+		RET=1
+	fi
+
+	for s in "$slave_0" "$slave_1"; do
+		churn_state=$(get_churn_state "$s")
+		if [ "$state" = "active" ]; then
+			if ! is_distributing "$s"; then
+				log_info "$state: $s is not in distributing state"
+				RET=1
+			fi
+			if [ "$churn_state" != "none" ]; then
+				log_info "$state: $s churn state $churn_state"
+				RET=1
+			fi
+		else
+			# Backup state, should be in churned and not distributing
+			if is_distributing "$s"; then
+				log_info "$state: $s is in distributing state"
+				RET=1
+			fi
+			if [ "$churn_state" != "churned" ]; then
+				log_info "$state: $s churn state $churn_state"
+				# shellcheck disable=SC2034
+				RET=1
+			fi
+		fi
+	done
+}
+
 trap cleanup_all_ns EXIT
 setup_ns c_ns s_ns b_ns
 setup_links
@@ -98,11 +157,25 @@ log_test "bond 802.3ad" "actor_port_prio setting"
 
 test_agg_reselect eth0
 log_test "bond 802.3ad" "actor_port_prio select"
+# sleep for a while to make sure the mux state machine has completed.
+sleep 10
+check_slave_state active eth0 eth1
+log_test "bond 802.3ad" "active state/churn checking"
+# wait for churn timer expired, need a bit longer as we restart eth1
+sleep 55
+check_slave_state backup eth2 eth3
+log_test "bond 802.3ad" "backup state/churn checking"
 
 # Change the actor port prio and re-test
 ip -n "${c_ns}" link set eth0 type bond_slave actor_port_prio 10
 ip -n "${c_ns}" link set eth2 type bond_slave actor_port_prio 1000
 test_agg_reselect eth2
 log_test "bond 802.3ad" "actor_port_prio switch"
+sleep 10
+check_slave_state active eth2 eth3
+log_test "bond 802.3ad" "active state/churn checking"
+sleep 55
+check_slave_state backup eth0 eth1
+log_test "bond 802.3ad" "backup state/churn checking"
 
 exit "${EXIT_STATUS}"
-- 
2.50.1