Merge branch 'selftests-updates-for-mlxsw-driver-test'

Jiri Pirko says:

====================
selftests: updates for mlxsw driver test

This patchset contains tweaks to the existing tests and is also adding
couple of new ones, namely tests for shared buffer and red offload.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

tools/testing/selftests/drivers/net/mlxsw/blackhole_routes.sh

@@ -45,6 +45,7 @@ ALL_TESTS="
 	blackhole_ipv6
 "
 NUM_NETIFS=4
+: ${TIMEOUT:=20000} # ms
 source $lib_dir/tc_common.sh
 source $lib_dir/lib.sh
 
@@ -123,7 +124,7 @@ blackhole_ipv4()
 		skip_hw dst_ip 198.51.100.1 src_ip 192.0.2.1 ip_proto icmp \
 		action pass
 
-	ip -4 route show 198.51.100.0/30 | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload ip -4 route show 198.51.100.0/30
 	check_err $? "route not marked as offloaded when should"
 
 	ping_do $h1 198.51.100.1
@@ -147,7 +148,7 @@ blackhole_ipv6()
 		skip_hw dst_ip 2001:db8:2::1 src_ip 2001:db8:1::1 \
 		ip_proto icmpv6 action pass
 
-	ip -6 route show 2001:db8:2::/120 | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload ip -6 route show 2001:db8:2::/120
 	check_err $? "route not marked as offloaded when should"
 
 	ping6_do $h1 2001:db8:2::1

tools/testing/selftests/drivers/net/mlxsw/devlink_trap_l3_drops.sh

@@ -641,13 +641,9 @@ erif_disabled_test()
 	mz_pid=$!
 
 	sleep 5
-	# In order to see this trap we need a route that points to disabled RIF.
-	# When ipv6 address is flushed, there is a delay and the routes are
-	# deleted before the RIF and we cannot get state that we have route
-	# to disabled RIF.
-	# Delete IPv6 address first and then check this trap with flushing IPv4.
-	ip -6 add flush dev br0
-	ip -4 add flush dev br0
+	# Unlinking the port from the bridge will disable the RIF associated
+	# with br0 as it is no longer an upper of any mlxsw port.
+	ip link set dev $rp1 nomaster
 
 	t1_packets=$(devlink_trap_rx_packets_get $trap_name)
 	t1_bytes=$(devlink_trap_rx_bytes_get $trap_name)
@@ -659,7 +655,6 @@ erif_disabled_test()
 	log_test "Egress RIF disabled"
 
 	kill $mz_pid && wait $mz_pid &> /dev/null
-	ip link set dev $rp1 nomaster
 	__addr_add_del $rp1 add 192.0.2.2/24 2001:db8:1::2/64
 	ip link del dev br0 type bridge
 	devlink_trap_action_set $trap_name "drop"

tools/testing/selftests/drivers/net/mlxsw/mlxsw_lib.sh

+#!/bin/bash
+# SPDX-License-Identifier: GPL-2.0
+
+##############################################################################
+# Defines
+
+if [[ ! -v MLXSW_CHIP ]]; then
+	MLXSW_CHIP=$(devlink -j dev info $DEVLINK_DEV | jq -r '.[][]["driver"]')
+	if [ -z "$MLXSW_CHIP" ]; then
+		echo "SKIP: Device $DEVLINK_DEV doesn't support devlink info command"
+		exit 1
+	fi
+fi

tools/testing/selftests/drivers/net/mlxsw/router_scale.sh

@@ -2,16 +2,15 @@
 # SPDX-License-Identifier: GPL-2.0
 
 ROUTER_NUM_NETIFS=4
+: ${TIMEOUT:=20000} # ms
 
 router_h1_create()
 {
 	simple_if_init $h1 192.0.1.1/24
-	ip route add 193.0.0.0/8 via 192.0.1.2 dev $h1
 }
 
 router_h1_destroy()
 {
-	ip route del 193.0.0.0/8 via 192.0.1.2 dev $h1
 	simple_if_fini $h1 192.0.1.1/24
 }
 
@@ -64,13 +63,15 @@ router_setup_prepare()
 	router_create
 }
 
-router_offload_validate()
+wait_for_routes()
 {
-	local route_count=$1
-	local offloaded_count
+	local t0=$1; shift
+	local route_count=$1; shift
 
-	offloaded_count=$(ip route | grep -o 'offload' | wc -l)
-	[[ $offloaded_count -ge $route_count ]]
+	local t1=$(ip route | grep -o 'offload' | wc -l)
+	local delta=$((t1 - t0))
+	echo $delta
+	[[ $delta -ge $route_count ]]
 }
 
 router_routes_create()
@@ -90,8 +91,8 @@ router_routes_create()
 					break 3
 				fi
 
-				echo route add 193.${i}.${j}.${k}/32 via \
-				       192.0.2.1 dev $rp2  >> $ROUTE_FILE
+				echo route add 193.${i}.${j}.${k}/32 dev $rp2 \
+					>> $ROUTE_FILE
 				((count++))
 			done
 		done
@@ -111,45 +112,19 @@ router_test()
 {
 	local route_count=$1
 	local should_fail=$2
-	local count=0
+	local delta
 
 	RET=0
 
+	local t0=$(ip route | grep -o 'offload' | wc -l)
 	router_routes_create $route_count
+	delta=$(busywait "$TIMEOUT" wait_for_routes $t0 $route_count)
 
-	router_offload_validate $route_count
-	check_err_fail $should_fail $? "Offload of $route_count routes"
+	check_err_fail $should_fail $? "Offload routes: Expected $route_count, got $delta."
 	if [[ $RET -ne 0 ]] || [[ $should_fail -eq 1 ]]; then
 		return
 	fi
 
-	tc filter add dev $h2 ingress protocol ip pref 1 flower \
-		skip_sw dst_ip 193.0.0.0/8 action drop
-
-	for i in {0..255}
-	do
-		for j in {0..255}
-		do
-			for k in {0..255}
-			do
-				if [[ $count -eq $route_count ]]; then
-					break 3
-				fi
-
-				$MZ $h1 -c 1 -p 64 -a $h1mac -b $rp1mac \
-					-A 192.0.1.1 -B 193.${i}.${j}.${k} \
-					-t ip -q
-				((count++))
-			done
-		done
-	done
-
-	tc_check_packets "dev $h2 ingress" 1 $route_count
-	check_err $? "Offload mismatch"
-
-	tc filter del dev $h2 ingress protocol ip pref 1 flower \
-		skip_sw dst_ip 193.0.0.0/8 action drop
-
 	router_routes_destroy
 }
 

tools/testing/selftests/drivers/net/mlxsw/rtnetlink.sh

@@ -32,6 +32,7 @@ ALL_TESTS="
 	devlink_reload_test
 "
 NUM_NETIFS=2
+: ${TIMEOUT:=20000} # ms
 source $lib_dir/lib.sh
 source $lib_dir/devlink_lib.sh
 
@@ -360,20 +361,24 @@ vlan_rif_refcount_test()
 	ip link add link br0 name br0.10 up type vlan id 10
 	ip -6 address add 2001:db8:1::1/64 dev br0.10
 
-	ip -6 route get fibmatch 2001:db8:1::2 dev br0.10 | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload \
+		ip -6 route get fibmatch 2001:db8:1::2 dev br0.10
 	check_err $? "vlan rif was not created before adding port to vlan"
 
 	bridge vlan add vid 10 dev $swp1
-	ip -6 route get fibmatch 2001:db8:1::2 dev br0.10 | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload \
+		ip -6 route get fibmatch 2001:db8:1::2 dev br0.10
 	check_err $? "vlan rif was destroyed after adding port to vlan"
 
 	bridge vlan del vid 10 dev $swp1
-	ip -6 route get fibmatch 2001:db8:1::2 dev br0.10 | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload \
+		ip -6 route get fibmatch 2001:db8:1::2 dev br0.10
 	check_err $? "vlan rif was destroyed after removing port from vlan"
 
 	ip link set dev $swp1 nomaster
-	ip -6 route get fibmatch 2001:db8:1::2 dev br0.10 | grep -q offload
-	check_fail $? "vlan rif was not destroyed after unlinking port from bridge"
+	busywait "$TIMEOUT" not wait_for_offload \
+		ip -6 route get fibmatch 2001:db8:1::2 dev br0.10
+	check_err $? "vlan rif was not destroyed after unlinking port from bridge"
 
 	log_test "vlan rif refcount"
 
@@ -401,22 +406,28 @@ subport_rif_refcount_test()
 	ip -6 address add 2001:db8:1::1/64 dev bond1
 	ip -6 address add 2001:db8:2::1/64 dev bond1.10
 
-	ip -6 route get fibmatch 2001:db8:1::2 dev bond1 | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload \
+		ip -6 route get fibmatch 2001:db8:1::2 dev bond1
 	check_err $? "subport rif was not created on lag device"
-	ip -6 route get fibmatch 2001:db8:2::2 dev bond1.10 | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload \
+		ip -6 route get fibmatch 2001:db8:2::2 dev bond1.10
 	check_err $? "subport rif was not created on vlan device"
 
 	ip link set dev $swp1 nomaster
-	ip -6 route get fibmatch 2001:db8:1::2 dev bond1 | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload \
+		ip -6 route get fibmatch 2001:db8:1::2 dev bond1
 	check_err $? "subport rif of lag device was destroyed when should not"
-	ip -6 route get fibmatch 2001:db8:2::2 dev bond1.10 | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload \
+		ip -6 route get fibmatch 2001:db8:2::2 dev bond1.10
 	check_err $? "subport rif of vlan device was destroyed when should not"
 
 	ip link set dev $swp2 nomaster
-	ip -6 route get fibmatch 2001:db8:1::2 dev bond1 | grep -q offload
-	check_fail $? "subport rif of lag device was not destroyed when should"
-	ip -6 route get fibmatch 2001:db8:2::2 dev bond1.10 | grep -q offload
-	check_fail $? "subport rif of vlan device was not destroyed when should"
+	busywait "$TIMEOUT" not wait_for_offload \
+		ip -6 route get fibmatch 2001:db8:1::2 dev bond1
+	check_err $? "subport rif of lag device was not destroyed when should"
+	busywait "$TIMEOUT" not wait_for_offload \
+		ip -6 route get fibmatch 2001:db8:2::2 dev bond1.10
+	check_err $? "subport rif of vlan device was not destroyed when should"
 
 	log_test "subport rif refcount"
 
@@ -575,7 +586,8 @@ bridge_extern_learn_test()
 
 	bridge fdb add de:ad:be:ef:13:37 dev $swp1 master extern_learn
 
-	bridge fdb show brport $swp1 | grep de:ad:be:ef:13:37 | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload \
+		bridge fdb show brport $swp1 de:ad:be:ef:13:37
 	check_err $? "fdb entry not marked as offloaded when should"
 
 	log_test "externally learned fdb entry"
@@ -595,9 +607,11 @@ neigh_offload_test()
 	ip -6 neigh add 2001:db8:1::2 lladdr de:ad:be:ef:13:37 nud perm \
 		dev $swp1
 
-	ip -4 neigh show dev $swp1 | grep 192.0.2.2 | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload \
+		ip -4 neigh show dev $swp1 192.0.2.2
 	check_err $? "ipv4 neigh entry not marked as offloaded when should"
-	ip -6 neigh show dev $swp1 | grep 2001:db8:1::2 | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload \
+		ip -6 neigh show dev $swp1 2001:db8:1::2
 	check_err $? "ipv6 neigh entry not marked as offloaded when should"
 
 	log_test "neighbour offload indication"
@@ -623,25 +637,31 @@ nexthop_offload_test()
 	ip -6 route add 2001:db8:2::/64 vrf v$swp1 \
 		nexthop via 2001:db8:1::2 dev $swp1
 
-	ip -4 route show 198.51.100.0/24 vrf v$swp1 | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload \
+		ip -4 route show 198.51.100.0/24 vrf v$swp1
 	check_err $? "ipv4 nexthop not marked as offloaded when should"
-	ip -6 route show 2001:db8:2::/64 vrf v$swp1 | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload \
+		ip -6 route show 2001:db8:2::/64 vrf v$swp1
 	check_err $? "ipv6 nexthop not marked as offloaded when should"
 
 	ip link set dev $swp2 down
 	sleep 1
 
-	ip -4 route show 198.51.100.0/24 vrf v$swp1 | grep -q offload
-	check_fail $? "ipv4 nexthop marked as offloaded when should not"
-	ip -6 route show 2001:db8:2::/64 vrf v$swp1 | grep -q offload
-	check_fail $? "ipv6 nexthop marked as offloaded when should not"
+	busywait "$TIMEOUT" not wait_for_offload \
+		ip -4 route show 198.51.100.0/24 vrf v$swp1
+	check_err $? "ipv4 nexthop marked as offloaded when should not"
+	busywait "$TIMEOUT" not wait_for_offload \
+		ip -6 route show 2001:db8:2::/64 vrf v$swp1
+	check_err $? "ipv6 nexthop marked as offloaded when should not"
 
 	ip link set dev $swp2 up
 	setup_wait
 
-	ip -4 route show 198.51.100.0/24 vrf v$swp1 | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload \
+		ip -4 route show 198.51.100.0/24 vrf v$swp1
 	check_err $? "ipv4 nexthop not marked as offloaded after neigh add"
-	ip -6 route show 2001:db8:2::/64 vrf v$swp1 | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload \
+		ip -6 route show 2001:db8:2::/64 vrf v$swp1
 	check_err $? "ipv6 nexthop not marked as offloaded after neigh add"
 
 	log_test "nexthop offload indication"

tools/testing/selftests/drivers/net/mlxsw/sch_red_core.sh

+# SPDX-License-Identifier: GPL-2.0
+
+# This test sends a >1Gbps stream of traffic from H1, to the switch, which
+# forwards it to a 1Gbps port. This 1Gbps stream is then looped back to the
+# switch and forwarded to the port under test $swp3, which is also 1Gbps.
+#
+# This way, $swp3 should be 100% filled with traffic without any of it spilling
+# to the backlog. Any extra packets sent should almost 1:1 go to backlog. That
+# is what H2 is used for--it sends the extra traffic to create backlog.
+#
+# A RED Qdisc is installed on $swp3. The configuration is such that the minimum
+# and maximum size are 1 byte apart, so there is a very clear border under which
+# no marking or dropping takes place, and above which everything is marked or
+# dropped.
+#
+# The test uses the buffer build-up behavior to test the installed RED.
+#
+# In order to test WRED, $swp3 actually contains RED under PRIO, with two
+# different configurations. Traffic is prioritized using 802.1p and relies on
+# the implicit mlxsw configuration, where packet priority is taken 1:1 from the
+# 802.1p marking.
+#
+# +--------------------------+                     +--------------------------+
+# | H1                       |                     | H2                       |
+# |     + $h1.10             |                     |     + $h2.10             |
+# |     | 192.0.2.1/28       |                     |     | 192.0.2.2/28       |
+# |     |                    |                     |     |                    |
+# |     |         $h1.11 +   |                     |     |         $h2.11 +   |
+# |     |  192.0.2.17/28 |   |                     |     |  192.0.2.18/28 |   |
+# |     |                |   |                     |     |                |   |
+# |     \______    ______/   |                     |     \______    ______/   |
+# |            \ /           |                     |            \ /           |
+# |             + $h1        |                     |             + $h2        |
+# +-------------|------------+                     +-------------|------------+
+#               | >1Gbps                                         |
+# +-------------|------------------------------------------------|------------+
+# | SW          + $swp1                                          + $swp2      |
+# |     _______/ \___________                        ___________/ \_______    |
+# |    /                     \                      /                     \   |
+# |  +-|-----------------+   |                    +-|-----------------+   |   |
+# |  | + $swp1.10        |   |                    | + $swp2.10        |   |   |
+# |  |                   |   |        .-------------+ $swp5.10        |   |   |
+# |  |     BR1_10        |   |        |           |                   |   |   |
+# |  |                   |   |        |           |     BR2_10        |   |   |
+# |  | + $swp2.10        |   |        |           |                   |   |   |
+# |  +-|-----------------+   |        |           | + $swp3.10        |   |   |
+# |    |                     |        |           +-|-----------------+   |   |
+# |    |   +-----------------|-+      |             |   +-----------------|-+ |
+# |    |   |        $swp1.11 + |      |             |   |        $swp2.11 + | |
+# |    |   |                   |      | .-----------------+ $swp5.11        | |
+# |    |   |      BR1_11       |      | |           |   |                   | |
+# |    |   |                   |      | |           |   |      BR2_11       | |
+# |    |   |        $swp2.11 + |      | |           |   |                   | |
+# |    |   +-----------------|-+      | |           |   |        $swp3.11 + | |
+# |    |                     |        | |           |   +-----------------|-+ |
+# |    \_______   ___________/        | |           \___________   _______/   |
+# |            \ /                    \ /                       \ /           |
+# |             + $swp4                + $swp5                   + $swp3      |
+# +-------------|----------------------|-------------------------|------------+
+#               |                      |                         | 1Gbps
+#               \________1Gbps_________/                         |
+#                                   +----------------------------|------------+
+#                                   | H3                         + $h3        |
+#                                   |      _____________________/ \_______    |
+#                                   |     /                               \   |
+#                                   |     |                               |   |
+#                                   |     + $h3.10                 $h3.11 +   |
+#                                   |       192.0.2.3/28    192.0.2.19/28     |
+#                                   +-----------------------------------------+
+
+NUM_NETIFS=8
+CHECK_TC="yes"
+lib_dir=$(dirname $0)/../../../net/forwarding
+source $lib_dir/lib.sh
+source $lib_dir/devlink_lib.sh
+source qos_lib.sh
+
+ipaddr()
+{
+	local host=$1; shift
+	local vlan=$1; shift
+
+	echo 192.0.2.$((16 * (vlan - 10) + host))
+}
+
+host_create()
+{
+	local dev=$1; shift
+	local host=$1; shift
+
+	simple_if_init $dev
+	mtu_set $dev 10000
+
+	vlan_create $dev 10 v$dev $(ipaddr $host 10)/28
+	ip link set dev $dev.10 type vlan egress 0:0
+
+	vlan_create $dev 11 v$dev $(ipaddr $host 11)/28
+	ip link set dev $dev.11 type vlan egress 0:1
+}
+
+host_destroy()
+{
+	local dev=$1; shift
+
+	vlan_destroy $dev 11
+	vlan_destroy $dev 10
+	mtu_restore $dev
+	simple_if_fini $dev
+}
+
+h1_create()
+{
+	host_create $h1 1
+}
+
+h1_destroy()
+{
+	host_destroy $h1
+}
+
+h2_create()
+{
+	host_create $h2 2
+
+	# Some of the tests in this suite use multicast traffic. As this traffic
+	# enters BR2_10 resp. BR2_11, it is flooded to all other ports. Thus
+	# e.g. traffic ingressing through $swp2 is flooded to $swp3 (the
+	# intended destination) and $swp5 (which is intended as ingress for
+	# another stream of traffic).
+	#
+	# This is generally not a problem, but if the $swp5 throughput is lower
+	# than $swp2 throughput, there will be a build-up at $swp5. That may
+	# cause packets to fail to queue up at $swp3 due to shared buffer
+	# quotas, and the test to spuriously fail.
+	#
+	# Prevent this by setting the speed of $h2 to 1Gbps.
+
+	ethtool -s $h2 speed 1000 autoneg off
+}
+
+h2_destroy()
+{
+	ethtool -s $h2 autoneg on
+	host_destroy $h2
+}
+
+h3_create()
+{
+	host_create $h3 3
+	ethtool -s $h3 speed 1000 autoneg off
+}
+
+h3_destroy()
+{
+	ethtool -s $h3 autoneg on
+	host_destroy $h3
+}
+
+switch_create()
+{
+	local intf
+	local vlan
+
+	ip link add dev br1_10 type bridge
+	ip link add dev br1_11 type bridge
+
+	ip link add dev br2_10 type bridge
+	ip link add dev br2_11 type bridge
+
+	for intf in $swp1 $swp2 $swp3 $swp4 $swp5; do
+		ip link set dev $intf up
+		mtu_set $intf 10000
+	done
+
+	for intf in $swp1 $swp4; do
+		for vlan in 10 11; do
+			vlan_create $intf $vlan
+			ip link set dev $intf.$vlan master br1_$vlan
+			ip link set dev $intf.$vlan up
+		done
+	done
+
+	for intf in $swp2 $swp3 $swp5; do
+		for vlan in 10 11; do
+			vlan_create $intf $vlan
+			ip link set dev $intf.$vlan master br2_$vlan
+			ip link set dev $intf.$vlan up
+		done
+	done
+
+	ip link set dev $swp4.10 type vlan egress 0:0
+	ip link set dev $swp4.11 type vlan egress 0:1
+	for intf in $swp1 $swp2 $swp5; do
+		for vlan in 10 11; do
+			ip link set dev $intf.$vlan type vlan ingress 0:0 1:1
+		done
+	done
+
+	for intf in $swp2 $swp3 $swp4 $swp5; do
+		ethtool -s $intf speed 1000 autoneg off
+	done
+
+	ip link set dev br1_10 up
+	ip link set dev br1_11 up
+	ip link set dev br2_10 up
+	ip link set dev br2_11 up
+
+	local size=$(devlink_pool_size_thtype 0 | cut -d' ' -f 1)
+	devlink_port_pool_th_set $swp3 8 $size
+}
+
+switch_destroy()
+{
+	local intf
+	local vlan
+
+	devlink_port_pool_th_restore $swp3 8
+
+	tc qdisc del dev $swp3 root 2>/dev/null
+
+	ip link set dev br2_11 down
+	ip link set dev br2_10 down
+	ip link set dev br1_11 down
+	ip link set dev br1_10 down
+
+	for intf in $swp5 $swp4 $swp3 $swp2; do
+		ethtool -s $intf autoneg on
+	done
+
+	for intf in $swp5 $swp3 $swp2 $swp4 $swp1; do
+		for vlan in 11 10; do
+			ip link set dev $intf.$vlan down
+			ip link set dev $intf.$vlan nomaster
+			vlan_destroy $intf $vlan
+		done
+
+		mtu_restore $intf
+		ip link set dev $intf down
+	done
+
+	ip link del dev br2_11
+	ip link del dev br2_10
+	ip link del dev br1_11
+	ip link del dev br1_10
+}
+
+setup_prepare()
+{
+	h1=${NETIFS[p1]}
+	swp1=${NETIFS[p2]}
+
+	swp2=${NETIFS[p3]}
+	h2=${NETIFS[p4]}
+
+	swp3=${NETIFS[p5]}
+	h3=${NETIFS[p6]}
+
+	swp4=${NETIFS[p7]}
+	swp5=${NETIFS[p8]}
+
+	h3_mac=$(mac_get $h3)
+
+	vrf_prepare
+
+	h1_create
+	h2_create
+	h3_create
+	switch_create
+}
+
+cleanup()
+{
+	pre_cleanup
+
+	switch_destroy
+	h3_destroy
+	h2_destroy
+	h1_destroy
+
+	vrf_cleanup
+}
+
+ping_ipv4()
+{
+	ping_test $h1.10 $(ipaddr 3 10) " from host 1, vlan 10"
+	ping_test $h1.11 $(ipaddr 3 11) " from host 1, vlan 11"
+	ping_test $h2.10 $(ipaddr 3 10) " from host 2, vlan 10"
+	ping_test $h2.11 $(ipaddr 3 11) " from host 2, vlan 11"
+}
+
+get_tc()
+{
+	local vlan=$1; shift
+
+	echo $((vlan - 10))
+}
+
+get_qdisc_handle()
+{
+	local vlan=$1; shift
+
+	local tc=$(get_tc $vlan)
+	local band=$((8 - tc))
+
+	# Handle is 107: for TC1, 108: for TC0.
+	echo "10$band:"
+}
+
+get_qdisc_backlog()
+{
+	local vlan=$1; shift
+
+	qdisc_stats_get $swp3 $(get_qdisc_handle $vlan) .backlog
+}
+
+get_mc_transmit_queue()
+{
+	local vlan=$1; shift
+
+	local tc=$(($(get_tc $vlan) + 8))
+	ethtool_stats_get $swp3 tc_transmit_queue_tc_$tc
+}
+
+get_nmarked()
+{
+	local vlan=$1; shift
+
+	ethtool_stats_get $swp3 ecn_marked
+}
+
+get_qdisc_npackets()
+{
+	local vlan=$1; shift
+
+	busywait_for_counter 1100 +1 \
+		qdisc_stats_get $swp3 $(get_qdisc_handle $vlan) .packets
+}
+
+# This sends traffic in an attempt to build a backlog of $size. Returns 0 on
+# success. After 10 failed attempts it bails out and returns 1. It dumps the
+# backlog size to stdout.
+build_backlog()
+{
+	local vlan=$1; shift
+	local size=$1; shift
+	local proto=$1; shift
+
+	local tc=$((vlan - 10))
+	local band=$((8 - tc))
+	local cur=-1
+	local i=0
+
+	while :; do
+		local cur=$(busywait 1100 until_counter_is $((cur + 1)) \
+					    get_qdisc_backlog $vlan)
+		local diff=$((size - cur))
+		local pkts=$(((diff + 7999) / 8000))
+
+		if ((cur >= size)); then
+			echo $cur
+			return 0
+		elif ((i++ > 10)); then
+			echo $cur
+			return 1
+		fi
+
+		$MZ $h2.$vlan -p 8000 -a own -b $h3_mac \
+		    -A $(ipaddr 2 $vlan) -B $(ipaddr 3 $vlan) \
+		    -t $proto -q -c $pkts "$@"
+	done
+}
+
+check_marking()
+{
+	local vlan=$1; shift
+	local cond=$1; shift
+
+	local npackets_0=$(get_qdisc_npackets $vlan)
+	local nmarked_0=$(get_nmarked $vlan)
+	sleep 5
+	local npackets_1=$(get_qdisc_npackets $vlan)
+	local nmarked_1=$(get_nmarked $vlan)
+
+	local nmarked_d=$((nmarked_1 - nmarked_0))
+	local npackets_d=$((npackets_1 - npackets_0))
+	local pct=$((100 * nmarked_d / npackets_d))
+
+	echo $pct
+	((pct $cond))
+}
+
+do_ecn_test()
+{
+	local vlan=$1; shift
+	local limit=$1; shift
+	local backlog
+	local pct
+
+	# Main stream.
+	start_tcp_traffic $h1.$vlan $(ipaddr 1 $vlan) $(ipaddr 3 $vlan) \
+			  $h3_mac tos=0x01
+
+	# Build the below-the-limit backlog using UDP. We could use TCP just
+	# fine, but this way we get a proof that UDP is accepted when queue
+	# length is below the limit. The main stream is using TCP, and if the
+	# limit is misconfigured, we would see this traffic being ECN marked.
+	RET=0
+	backlog=$(build_backlog $vlan $((2 * limit / 3)) udp)
+	check_err $? "Could not build the requested backlog"
+	pct=$(check_marking $vlan "== 0")
+	check_err $? "backlog $backlog / $limit Got $pct% marked packets, expected == 0."
+	log_test "TC $((vlan - 10)): ECN backlog < limit"
+
+	# Now push TCP, because non-TCP traffic would be early-dropped after the
+	# backlog crosses the limit, and we want to make sure that the backlog
+	# is above the limit.
+	RET=0
+	backlog=$(build_backlog $vlan $((3 * limit / 2)) tcp tos=0x01)
+	check_err $? "Could not build the requested backlog"
+	pct=$(check_marking $vlan ">= 95")
+	check_err $? "backlog $backlog / $limit Got $pct% marked packets, expected >= 95."
+	log_test "TC $((vlan - 10)): ECN backlog > limit"
+
+	# Up there we saw that UDP gets accepted when backlog is below the
+	# limit. Now that it is above, it should all get dropped, and backlog
+	# building should fail.
+	RET=0
+	build_backlog $vlan $((2 * limit)) udp >/dev/null
+	check_fail $? "UDP traffic went into backlog instead of being early-dropped"
+	log_test "TC $((vlan - 10)): ECN backlog > limit: UDP early-dropped"
+
+	stop_traffic
+	sleep 1
+}
+
+do_red_test()
+{
+	local vlan=$1; shift
+	local limit=$1; shift
+	local backlog
+	local pct
+
+	# Use ECN-capable TCP to verify there's no marking even though the queue
+	# is above limit.
+	start_tcp_traffic $h1.$vlan $(ipaddr 1 $vlan) $(ipaddr 3 $vlan) \
+			  $h3_mac tos=0x01
+
+	# Pushing below the queue limit should work.
+	RET=0
+	backlog=$(build_backlog $vlan $((2 * limit / 3)) tcp tos=0x01)
+	check_err $? "Could not build the requested backlog"
+	pct=$(check_marking $vlan "== 0")
+	check_err $? "backlog $backlog / $limit Got $pct% marked packets, expected == 0."
+	log_test "TC $((vlan - 10)): RED backlog < limit"
+
+	# Pushing above should not.
+	RET=0
+	backlog=$(build_backlog $vlan $((3 * limit / 2)) tcp tos=0x01)
+	check_fail $? "Traffic went into backlog instead of being early-dropped"
+	pct=$(check_marking $vlan "== 0")
+	check_err $? "backlog $backlog / $limit Got $pct% marked packets, expected == 0."
+	local diff=$((limit - backlog))
+	pct=$((100 * diff / limit))
+	((0 <= pct && pct <= 5))
+	check_err $? "backlog $backlog / $limit expected <= 5% distance"
+	log_test "TC $((vlan - 10)): RED backlog > limit"
+
+	stop_traffic
+	sleep 1
+}
+
+do_mc_backlog_test()
+{
+	local vlan=$1; shift
+	local limit=$1; shift
+	local backlog
+	local pct
+
+	RET=0
+
+	start_tcp_traffic $h1.$vlan $(ipaddr 1 $vlan) $(ipaddr 3 $vlan) bc
+	start_tcp_traffic $h2.$vlan $(ipaddr 2 $vlan) $(ipaddr 3 $vlan) bc
+
+	qbl=$(busywait 5000 until_counter_is 500000 \
+		       get_qdisc_backlog $vlan)
+	check_err $? "Could not build MC backlog"
+
+	# Verify that we actually see the backlog on BUM TC. Do a busywait as
+	# well, performance blips might cause false fail.
+	local ebl
+	ebl=$(busywait 5000 until_counter_is 500000 \
+		       get_mc_transmit_queue $vlan)
+	check_err $? "MC backlog reported by qdisc not visible in ethtool"
+
+	stop_traffic
+	stop_traffic
+
+	log_test "TC $((vlan - 10)): Qdisc reports MC backlog"
+}

tools/testing/selftests/drivers/net/mlxsw/sch_red_ets.sh

+#!/bin/bash
+# SPDX-License-Identifier: GPL-2.0
+
+ALL_TESTS="
+	ping_ipv4
+	ecn_test
+	red_test
+	mc_backlog_test
+"
+: ${QDISC:=ets}
+source sch_red_core.sh
+
+# do_ecn_test first build 2/3 of the requested backlog and expects no marking,
+# and then builds 3/2 of it and does expect marking. The values of $BACKLOG1 and
+# $BACKLOG2 are far enough not to overlap, so that we can assume that if we do
+# see (do not see) marking, it is actually due to the configuration of that one
+# TC, and not due to configuration of the other TC leaking over.
+BACKLOG1=200000
+BACKLOG2=500000
+
+install_qdisc()
+{
+	local -a args=("$@")
+
+	tc qdisc add dev $swp3 root handle 10: $QDISC \
+	   bands 8 priomap 7 6 5 4 3 2 1 0
+	tc qdisc add dev $swp3 parent 10:8 handle 108: red \
+	   limit 1000000 min $BACKLOG1 max $((BACKLOG1 + 1)) \
+	   probability 1.0 avpkt 8000 burst 38 "${args[@]}"
+	tc qdisc add dev $swp3 parent 10:7 handle 107: red \
+	   limit 1000000 min $BACKLOG2 max $((BACKLOG2 + 1)) \
+	   probability 1.0 avpkt 8000 burst 63 "${args[@]}"
+	sleep 1
+}
+
+uninstall_qdisc()
+{
+	tc qdisc del dev $swp3 parent 10:7
+	tc qdisc del dev $swp3 parent 10:8
+	tc qdisc del dev $swp3 root
+}
+
+ecn_test()
+{
+	install_qdisc ecn
+
+	do_ecn_test 10 $BACKLOG1
+	do_ecn_test 11 $BACKLOG2
+
+	uninstall_qdisc
+}
+
+red_test()
+{
+	install_qdisc
+
+	do_red_test 10 $BACKLOG1
+	do_red_test 11 $BACKLOG2
+
+	uninstall_qdisc
+}
+
+mc_backlog_test()
+{
+	install_qdisc
+
+	# Note that the backlog numbers here do not correspond to RED
+	# configuration, but are arbitrary.
+	do_mc_backlog_test 10 $BACKLOG1
+	do_mc_backlog_test 11 $BACKLOG2
+
+	uninstall_qdisc
+}
+
+trap cleanup EXIT
+
+setup_prepare
+setup_wait
+
+bail_on_lldpad
+tests_run
+
+exit $EXIT_STATUS

tools/testing/selftests/drivers/net/mlxsw/sch_red_prio.sh

+#!/bin/bash
+# SPDX-License-Identifier: GPL-2.0
+
+QDISC=prio
+source sch_red_ets.sh

tools/testing/selftests/drivers/net/mlxsw/sch_red_root.sh

+#!/bin/bash
+# SPDX-License-Identifier: GPL-2.0
+
+ALL_TESTS="
+	ping_ipv4
+	ecn_test
+	red_test
+	mc_backlog_test
+"
+source sch_red_core.sh
+
+BACKLOG=300000
+
+install_qdisc()
+{
+	local -a args=("$@")
+
+	tc qdisc add dev $swp3 root handle 108: red \
+	   limit 1000000 min $BACKLOG max $((BACKLOG + 1)) \
+	   probability 1.0 avpkt 8000 burst 38 "${args[@]}"
+	sleep 1
+}
+
+uninstall_qdisc()
+{
+	tc qdisc del dev $swp3 root
+}
+
+ecn_test()
+{
+	install_qdisc ecn
+	do_ecn_test 10 $BACKLOG
+	uninstall_qdisc
+}
+
+red_test()
+{
+	install_qdisc
+	do_red_test 10 $BACKLOG
+	uninstall_qdisc
+}
+
+mc_backlog_test()
+{
+	install_qdisc
+	# Note that the backlog value here does not correspond to RED
+	# configuration, but is arbitrary.
+	do_mc_backlog_test 10 $BACKLOG
+	uninstall_qdisc
+}
+
+trap cleanup EXIT
+
+setup_prepare
+setup_wait
+
+bail_on_lldpad
+tests_run
+
+exit $EXIT_STATUS

tools/testing/selftests/drivers/net/mlxsw/sharedbuffer.sh

+#!/bin/bash
+# SPDX-License-Identifier: GPL-2.0
+
+ALL_TESTS="
+	port_pool_test
+	port_tc_ip_test
+	port_tc_arp_test
+"
+
+NUM_NETIFS=2
+source ../../../net/forwarding/lib.sh
+source ../../../net/forwarding/devlink_lib.sh
+source mlxsw_lib.sh
+
+SB_POOL_ING=0
+SB_POOL_EGR_CPU=10
+
+SB_ITC_CPU_IP=3
+SB_ITC_CPU_ARP=2
+SB_ITC=0
+
+h1_create()
+{
+	simple_if_init $h1 192.0.1.1/24
+}
+
+h1_destroy()
+{
+	simple_if_fini $h1 192.0.1.1/24
+}
+
+h2_create()
+{
+	simple_if_init $h2 192.0.1.2/24
+}
+
+h2_destroy()
+{
+	simple_if_fini $h2 192.0.1.2/24
+}
+
+sb_occ_pool_check()
+{
+	local dl_port=$1; shift
+	local pool=$1; shift
+	local exp_max_occ=$1
+	local max_occ
+	local err=0
+
+	max_occ=$(devlink sb -j occupancy show $dl_port \
+		  | jq -e ".[][][\"pool\"][\"$pool\"][\"max\"]")
+
+	if [[ "$max_occ" -ne "$exp_max_occ" ]]; then
+		err=1
+	fi
+
+	echo $max_occ
+	return $err
+}
+
+sb_occ_itc_check()
+{
+	local dl_port=$1; shift
+	local itc=$1; shift
+	local exp_max_occ=$1
+	local max_occ
+	local err=0
+
+	max_occ=$(devlink sb -j occupancy show $dl_port \
+		  | jq -e ".[][][\"itc\"][\"$itc\"][\"max\"]")
+
+	if [[ "$max_occ" -ne "$exp_max_occ" ]]; then
+		err=1
+	fi
+
+	echo $max_occ
+	return $err
+}
+
+sb_occ_etc_check()
+{
+	local dl_port=$1; shift
+	local etc=$1; shift
+	local exp_max_occ=$1; shift
+	local max_occ
+	local err=0
+
+	max_occ=$(devlink sb -j occupancy show $dl_port \
+		  | jq -e ".[][][\"etc\"][\"$etc\"][\"max\"]")
+
+	if [[ "$max_occ" -ne "$exp_max_occ" ]]; then
+		err=1
+	fi
+
+	echo $max_occ
+	return $err
+}
+
+port_pool_test()
+{
+	local exp_max_occ=288
+	local max_occ
+
+	devlink sb occupancy clearmax $DEVLINK_DEV
+
+	$MZ $h1 -c 1 -p 160 -a $h1mac -b $h2mac -A 192.0.1.1 -B 192.0.1.2 \
+		-t ip -q
+
+	devlink sb occupancy snapshot $DEVLINK_DEV
+
+	RET=0
+	max_occ=$(sb_occ_pool_check $dl_port1 $SB_POOL_ING $exp_max_occ)
+	check_err $? "Expected iPool($SB_POOL_ING) max occupancy to be $exp_max_occ, but got $max_occ"
+	log_test "physical port's($h1) ingress pool"
+
+	RET=0
+	max_occ=$(sb_occ_pool_check $dl_port2 $SB_POOL_ING $exp_max_occ)
+	check_err $? "Expected iPool($SB_POOL_ING) max occupancy to be $exp_max_occ, but got $max_occ"
+	log_test "physical port's($h2) ingress pool"
+
+	RET=0
+	max_occ=$(sb_occ_pool_check $cpu_dl_port $SB_POOL_EGR_CPU $exp_max_occ)
+	check_err $? "Expected ePool($SB_POOL_EGR_CPU) max occupancy to be $exp_max_occ, but got $max_occ"
+	log_test "CPU port's egress pool"
+}
+
+port_tc_ip_test()
+{
+	local exp_max_occ=288
+	local max_occ
+
+	devlink sb occupancy clearmax $DEVLINK_DEV
+
+	$MZ $h1 -c 1 -p 160 -a $h1mac -b $h2mac -A 192.0.1.1 -B 192.0.1.2 \
+		-t ip -q
+
+	devlink sb occupancy snapshot $DEVLINK_DEV
+
+	RET=0
+	max_occ=$(sb_occ_itc_check $dl_port2 $SB_ITC $exp_max_occ)
+	check_err $? "Expected ingress TC($SB_ITC) max occupancy to be $exp_max_occ, but got $max_occ"
+	log_test "physical port's($h1) ingress TC - IP packet"
+
+	RET=0
+	max_occ=$(sb_occ_itc_check $dl_port2 $SB_ITC $exp_max_occ)
+	check_err $? "Expected ingress TC($SB_ITC) max occupancy to be $exp_max_occ, but got $max_occ"
+	log_test "physical port's($h2) ingress TC - IP packet"
+
+	RET=0
+	max_occ=$(sb_occ_etc_check $cpu_dl_port $SB_ITC_CPU_IP $exp_max_occ)
+	check_err $? "Expected egress TC($SB_ITC_CPU_IP) max occupancy to be $exp_max_occ, but got $max_occ"
+	log_test "CPU port's egress TC - IP packet"
+}
+
+port_tc_arp_test()
+{
+	local exp_max_occ=96
+	local max_occ
+
+	if [[ $MLXSW_CHIP != "mlxsw_spectrum" ]]; then
+		exp_max_occ=144
+	fi
+
+	devlink sb occupancy clearmax $DEVLINK_DEV
+
+	$MZ $h1 -c 1 -p 160 -a $h1mac -A 192.0.1.1 -t arp -q
+
+	devlink sb occupancy snapshot $DEVLINK_DEV
+
+	RET=0
+	max_occ=$(sb_occ_itc_check $dl_port2 $SB_ITC $exp_max_occ)
+	check_err $? "Expected ingress TC($SB_ITC) max occupancy to be $exp_max_occ, but got $max_occ"
+	log_test "physical port's($h1) ingress TC - ARP packet"
+
+	RET=0
+	max_occ=$(sb_occ_itc_check $dl_port2 $SB_ITC $exp_max_occ)
+	check_err $? "Expected ingress TC($SB_ITC) max occupancy to be $exp_max_occ, but got $max_occ"
+	log_test "physical port's($h2) ingress TC - ARP packet"
+
+	RET=0
+	max_occ=$(sb_occ_etc_check $cpu_dl_port $SB_ITC_CPU_ARP $exp_max_occ)
+	check_err $? "Expected egress TC($SB_ITC_IP2ME) max occupancy to be $exp_max_occ, but got $max_occ"
+	log_test "CPU port's egress TC - ARP packet"
+}
+
+setup_prepare()
+{
+	h1=${NETIFS[p1]}
+	h2=${NETIFS[p2]}
+
+	h1mac=$(mac_get $h1)
+	h2mac=$(mac_get $h2)
+
+	dl_port1=$(devlink_port_by_netdev $h1)
+	dl_port2=$(devlink_port_by_netdev $h2)
+
+	cpu_dl_port=$(devlink_cpu_port_get)
+
+	vrf_prepare
+
+	h1_create
+	h2_create
+}
+
+cleanup()
+{
+	pre_cleanup
+
+	h2_destroy
+	h1_destroy
+
+	vrf_cleanup
+}
+
+trap cleanup EXIT
+
+setup_prepare
+setup_wait
+
+tests_run
+
+exit $EXIT_STATUS

tools/testing/selftests/drivers/net/mlxsw/sharedbuffer_configuration.py

+#!/usr/bin/python
+# SPDX-License-Identifier: GPL-2.0
+
+import subprocess
+import json as j
+import random
+
+
+class SkipTest(Exception):
+    pass
+
+
+class RandomValuePicker:
+    """
+    Class for storing shared buffer configuration. Can handle 3 different
+    objects, pool, tcbind and portpool. Provide an interface to get random
+    values for a specific object type as the follow:
+      1. Pool:
+         - random size
+
+      2. TcBind:
+         - random pool number
+         - random threshold
+
+      3. PortPool:
+         - random threshold
+    """
+    def __init__(self, pools):
+        self._pools = []
+        for pool in pools:
+            self._pools.append(pool)
+
+    def _cell_size(self):
+        return self._pools[0]["cell_size"]
+
+    def _get_static_size(self, th):
+        # For threshold of 16, this works out to be about 12MB on Spectrum-1,
+        # and about 17MB on Spectrum-2.
+        return th * 8000 * self._cell_size()
+
+    def _get_size(self):
+        return self._get_static_size(16)
+
+    def _get_thtype(self):
+        return "static"
+
+    def _get_th(self, pool):
+        # Threshold value could be any integer between 3 to 16
+        th = random.randint(3, 16)
+        if pool["thtype"] == "dynamic":
+            return th
+        else:
+            return self._get_static_size(th)
+
+    def _get_pool(self, direction):
+        ing_pools = []
+        egr_pools = []
+        for pool in self._pools:
+            if pool["type"] == "ingress":
+                ing_pools.append(pool)
+            else:
+                egr_pools.append(pool)
+        if direction == "ingress":
+            arr = ing_pools
+        else:
+            arr = egr_pools
+        return arr[random.randint(0, len(arr) - 1)]
+
+    def get_value(self, objid):
+        if isinstance(objid, Pool):
+            if objid["pool"] in [4, 8, 9, 10]:
+                # The threshold type of pools 4, 8, 9 and 10 cannot be changed
+                raise SkipTest()
+            else:
+                return (self._get_size(), self._get_thtype())
+        if isinstance(objid, TcBind):
+            if objid["tc"] >= 8:
+                # Multicast TCs cannot be changed
+                raise SkipTest()
+            else:
+                pool = self._get_pool(objid["type"])
+                th = self._get_th(pool)
+                pool_n = pool["pool"]
+                return (pool_n, th)
+        if isinstance(objid, PortPool):
+            pool_n = objid["pool"]
+            pool = self._pools[pool_n]
+            assert pool["pool"] == pool_n
+            th = self._get_th(pool)
+            return (th,)
+
+
+class RecordValuePickerException(Exception):
+    pass
+
+
+class RecordValuePicker:
+    """
+    Class for storing shared buffer configuration. Can handle 2 different
+    objects, pool and tcbind. Provide an interface to get the stored values per
+    object type.
+    """
+    def __init__(self, objlist):
+        self._recs = []
+        for item in objlist:
+            self._recs.append({"objid": item, "value": item.var_tuple()})
+
+    def get_value(self, objid):
+        if isinstance(objid, Pool) and objid["pool"] in [4, 8, 9, 10]:
+            # The threshold type of pools 4, 8, 9 and 10 cannot be changed
+            raise SkipTest()
+        if isinstance(objid, TcBind) and objid["tc"] >= 8:
+            # Multicast TCs cannot be changed
+            raise SkipTest()
+        for rec in self._recs:
+            if rec["objid"].weak_eq(objid):
+                return rec["value"]
+        raise RecordValuePickerException()
+
+
+def run_cmd(cmd, json=False):
+    out = subprocess.check_output(cmd, shell=True)
+    if json:
+        return j.loads(out)
+    return out
+
+
+def run_json_cmd(cmd):
+    return run_cmd(cmd, json=True)
+
+
+def log_test(test_name, err_msg=None):
+    if err_msg:
+        print("\t%s" % err_msg)
+        print("TEST: %-80s  [FAIL]" % test_name)
+    else:
+        print("TEST: %-80s  [ OK ]" % test_name)
+
+
+class CommonItem(dict):
+    varitems = []
+
+    def var_tuple(self):
+        ret = []
+        self.varitems.sort()
+        for key in self.varitems:
+            ret.append(self[key])
+        return tuple(ret)
+
+    def weak_eq(self, other):
+        for key in self:
+            if key in self.varitems:
+                continue
+            if self[key] != other[key]:
+                return False
+        return True
+
+
+class CommonList(list):
+    def get_by(self, by_obj):
+        for item in self:
+            if item.weak_eq(by_obj):
+                return item
+        return None
+
+    def del_by(self, by_obj):
+        for item in self:
+            if item.weak_eq(by_obj):
+                self.remove(item)
+
+
+class Pool(CommonItem):
+    varitems = ["size", "thtype"]
+
+    def dl_set(self, dlname, size, thtype):
+        run_cmd("devlink sb pool set {} sb {} pool {} size {} thtype {}".format(dlname, self["sb"],
+                                                                                self["pool"],
+                                                                                size, thtype))
+
+
+class PoolList(CommonList):
+    pass
+
+
+def get_pools(dlname, direction=None):
+    d = run_json_cmd("devlink sb pool show -j")
+    pools = PoolList()
+    for pooldict in d["pool"][dlname]:
+        if not direction or direction == pooldict["type"]:
+            pools.append(Pool(pooldict))
+    return pools
+
+
+def do_check_pools(dlname, pools, vp):
+    for pool in pools:
+        pre_pools = get_pools(dlname)
+        try:
+            (size, thtype) = vp.get_value(pool)
+        except SkipTest:
+            continue
+        pool.dl_set(dlname, size, thtype)
+        post_pools = get_pools(dlname)
+        pool = post_pools.get_by(pool)
+
+        err_msg = None
+        if pool["size"] != size:
+            err_msg = "Incorrect pool size (got {}, expected {})".format(pool["size"], size)
+        if pool["thtype"] != thtype:
+            err_msg = "Incorrect pool threshold type (got {}, expected {})".format(pool["thtype"], thtype)
+
+        pre_pools.del_by(pool)
+        post_pools.del_by(pool)
+        if pre_pools != post_pools:
+            err_msg = "Other pool setup changed as well"
+        log_test("pool {} of sb {} set verification".format(pool["pool"],
+                                                            pool["sb"]), err_msg)
+
+
+def check_pools(dlname, pools):
+    # Save defaults
+    record_vp = RecordValuePicker(pools)
+
+    # For each pool, set random size and static threshold type
+    do_check_pools(dlname, pools, RandomValuePicker(pools))
+
+    # Restore defaults
+    do_check_pools(dlname, pools, record_vp)
+
+
+class TcBind(CommonItem):
+    varitems = ["pool", "threshold"]
+
+    def __init__(self, port, d):
+        super(TcBind, self).__init__(d)
+        self["dlportname"] = port.name
+
+    def dl_set(self, pool, th):
+        run_cmd("devlink sb tc bind set {} sb {} tc {} type {} pool {} th {}".format(self["dlportname"],
+                                                                                     self["sb"],
+                                                                                     self["tc"],
+                                                                                     self["type"],
+                                                                                     pool, th))
+
+
+class TcBindList(CommonList):
+    pass
+
+
+def get_tcbinds(ports, verify_existence=False):
+    d = run_json_cmd("devlink sb tc bind show -j -n")
+    tcbinds = TcBindList()
+    for port in ports:
+        err_msg = None
+        if port.name not in d["tc_bind"] or len(d["tc_bind"][port.name]) == 0:
+            err_msg = "No tc bind for port"
+        else:
+            for tcbinddict in d["tc_bind"][port.name]:
+                tcbinds.append(TcBind(port, tcbinddict))
+        if verify_existence:
+            log_test("tc bind existence for port {} verification".format(port.name), err_msg)
+    return tcbinds
+
+
+def do_check_tcbind(ports, tcbinds, vp):
+    for tcbind in tcbinds:
+        pre_tcbinds = get_tcbinds(ports)
+        try:
+            (pool, th) = vp.get_value(tcbind)
+        except SkipTest:
+            continue
+        tcbind.dl_set(pool, th)
+        post_tcbinds = get_tcbinds(ports)
+        tcbind = post_tcbinds.get_by(tcbind)
+
+        err_msg = None
+        if tcbind["pool"] != pool:
+            err_msg = "Incorrect pool (got {}, expected {})".format(tcbind["pool"], pool)
+        if tcbind["threshold"] != th:
+            err_msg = "Incorrect threshold (got {}, expected {})".format(tcbind["threshold"], th)
+
+        pre_tcbinds.del_by(tcbind)
+        post_tcbinds.del_by(tcbind)
+        if pre_tcbinds != post_tcbinds:
+            err_msg = "Other tc bind setup changed as well"
+        log_test("tc bind {}-{} of sb {} set verification".format(tcbind["dlportname"],
+                                                                  tcbind["tc"],
+                                                                  tcbind["sb"]), err_msg)
+
+
+def check_tcbind(dlname, ports, pools):
+    tcbinds = get_tcbinds(ports, verify_existence=True)
+
+    # Save defaults
+    record_vp = RecordValuePicker(tcbinds)
+
+    # Bind each port and unicast TC (TCs < 8) to a random pool and a random
+    # threshold
+    do_check_tcbind(ports, tcbinds, RandomValuePicker(pools))
+
+    # Restore defaults
+    do_check_tcbind(ports, tcbinds, record_vp)
+
+
+class PortPool(CommonItem):
+    varitems = ["threshold"]
+
+    def __init__(self, port, d):
+        super(PortPool, self).__init__(d)
+        self["dlportname"] = port.name
+
+    def dl_set(self, th):
+        run_cmd("devlink sb port pool set {} sb {} pool {} th {}".format(self["dlportname"],
+                                                                         self["sb"],
+                                                                         self["pool"], th))
+
+
+class PortPoolList(CommonList):
+    pass
+
+
+def get_portpools(ports, verify_existence=False):
+    d = run_json_cmd("devlink sb port pool -j -n")
+    portpools = PortPoolList()
+    for port in ports:
+        err_msg = None
+        if port.name not in d["port_pool"] or len(d["port_pool"][port.name]) == 0:
+            err_msg = "No port pool for port"
+        else:
+            for portpooldict in d["port_pool"][port.name]:
+                portpools.append(PortPool(port, portpooldict))
+        if verify_existence:
+            log_test("port pool existence for port {} verification".format(port.name), err_msg)
+    return portpools
+
+
+def do_check_portpool(ports, portpools, vp):
+    for portpool in portpools:
+        pre_portpools = get_portpools(ports)
+        (th,) = vp.get_value(portpool)
+        portpool.dl_set(th)
+        post_portpools = get_portpools(ports)
+        portpool = post_portpools.get_by(portpool)
+
+        err_msg = None
+        if portpool["threshold"] != th:
+            err_msg = "Incorrect threshold (got {}, expected {})".format(portpool["threshold"], th)
+
+        pre_portpools.del_by(portpool)
+        post_portpools.del_by(portpool)
+        if pre_portpools != post_portpools:
+            err_msg = "Other port pool setup changed as well"
+        log_test("port pool {}-{} of sb {} set verification".format(portpool["dlportname"],
+                                                                    portpool["pool"],
+                                                                    portpool["sb"]), err_msg)
+
+
+def check_portpool(dlname, ports, pools):
+    portpools = get_portpools(ports, verify_existence=True)
+
+    # Save defaults
+    record_vp = RecordValuePicker(portpools)
+
+    # For each port pool, set a random threshold
+    do_check_portpool(ports, portpools, RandomValuePicker(pools))
+
+    # Restore defaults
+    do_check_portpool(ports, portpools, record_vp)
+
+
+class Port:
+    def __init__(self, name):
+        self.name = name
+
+
+class PortList(list):
+    pass
+
+
+def get_ports(dlname):
+    d = run_json_cmd("devlink port show -j")
+    ports = PortList()
+    for name in d["port"]:
+        if name.find(dlname) == 0 and d["port"][name]["flavour"] == "physical":
+            ports.append(Port(name))
+    return ports
+
+
+def get_device():
+    devices_info = run_json_cmd("devlink -j dev info")["info"]
+    for d in devices_info:
+        if "mlxsw_spectrum" in devices_info[d]["driver"]:
+            return d
+    return None
+
+
+class UnavailableDevlinkNameException(Exception):
+    pass
+
+
+def test_sb_configuration():
+    # Use static seed
+    random.seed(0)
+
+    dlname = get_device()
+    if not dlname:
+        raise UnavailableDevlinkNameException()
+
+    ports = get_ports(dlname)
+    pools = get_pools(dlname)
+
+    check_pools(dlname, pools)
+    check_tcbind(dlname, ports, pools)
+    check_portpool(dlname, ports, pools)
+
+
+test_sb_configuration()

tools/testing/selftests/drivers/net/mlxsw/spectrum-2/resource_scale.sh

@@ -8,8 +8,9 @@ source $lib_dir/lib.sh
 source $lib_dir/tc_common.sh
 source $lib_dir/devlink_lib.sh
 
-if [ "$DEVLINK_VIDDID" != "15b3:cf6c" ]; then
-	echo "SKIP: test is tailored for Mellanox Spectrum-2"
+if [[ "$DEVLINK_VIDDID" != "15b3:cf6c" && \
+	"$DEVLINK_VIDDID" != "15b3:cf70" ]]; then
+	echo "SKIP: test is tailored for Mellanox Spectrum-2 and Spectrum-3"
 	exit 1
 fi
 

tools/testing/selftests/drivers/net/mlxsw/tc_flower_restrictions.sh

@@ -3,7 +3,11 @@
 
 lib_dir=$(dirname $0)/../../../net/forwarding
 
-ALL_TESTS="shared_block_drop_test"
+ALL_TESTS="
+	shared_block_drop_test
+	egress_redirect_test
+	multi_mirror_test
+"
 NUM_NETIFS=2
 
 source $lib_dir/tc_common.sh
@@ -69,6 +73,88 @@ shared_block_drop_test()
 	log_test "shared block drop"
 }
 
+egress_redirect_test()
+{
+	RET=0
+
+	# It is forbidden in mlxsw driver to have mirred redirect on
+	# egress-bound block.
+
+	tc qdisc add dev $swp1 ingress_block 22 clsact
+	check_err $? "Failed to create clsact with ingress block"
+
+	tc filter add block 22 protocol ip pref 1 handle 101 flower \
+		skip_sw dst_ip 192.0.2.2 \
+		action mirred egress redirect dev $swp2
+	check_err $? "Failed to add redirect rule to ingress bound block"
+
+	tc qdisc add dev $swp2 ingress_block 22 clsact
+	check_err $? "Failed to create another clsact with ingress shared block"
+
+	tc qdisc del dev $swp2 clsact
+
+	tc qdisc add dev $swp2 egress_block 22 clsact
+	check_fail $? "Incorrect success to create another clsact with egress shared block"
+
+	tc filter del block 22 protocol ip pref 1 handle 101 flower
+
+	tc qdisc add dev $swp2 egress_block 22 clsact
+	check_err $? "Failed to create another clsact with egress shared block after blocker redirect rule removed"
+
+	tc filter add block 22 protocol ip pref 1 handle 101 flower \
+		skip_sw dst_ip 192.0.2.2 \
+		action mirred egress redirect dev $swp2
+	check_fail $? "Incorrect success to add redirect rule to mixed bound block"
+
+	tc qdisc del dev $swp1 clsact
+
+	tc qdisc add dev $swp1 egress_block 22 clsact
+	check_err $? "Failed to create another clsact with egress shared block"
+
+	tc filter add block 22 protocol ip pref 1 handle 101 flower \
+		skip_sw dst_ip 192.0.2.2 \
+		action mirred egress redirect dev $swp2
+	check_fail $? "Incorrect success to add redirect rule to egress bound shared block"
+
+	tc qdisc del dev $swp2 clsact
+
+	tc filter add block 22 protocol ip pref 1 handle 101 flower \
+		skip_sw dst_ip 192.0.2.2 \
+		action mirred egress redirect dev $swp2
+	check_fail $? "Incorrect success to add redirect rule to egress bound block"
+
+	tc qdisc del dev $swp1 clsact
+
+	log_test "shared block drop"
+}
+
+multi_mirror_test()
+{
+	RET=0
+
+	# It is forbidden in mlxsw driver to have multiple mirror
+	# actions in a single rule.
+
+	tc qdisc add dev $swp1 clsact
+
+	tc filter add dev $swp1 ingress protocol ip pref 1 handle 101 flower \
+		skip_sw dst_ip 192.0.2.2 \
+		action mirred egress mirror dev $swp2
+	check_err $? "Failed to add rule with single mirror action"
+
+	tc filter del dev $swp1 ingress protocol ip pref 1 handle 101 flower
+
+	tc filter add dev $swp1 ingress protocol ip pref 1 handle 101 flower \
+		skip_sw dst_ip 192.0.2.2 \
+		action mirred egress mirror dev $swp2 \
+		action mirred egress mirror dev $swp1
+	check_fail $? "Incorrect success to add rule with two mirror actions"
+
+	tc qdisc del dev $swp1 clsact
+
+	log_test "multi mirror"
+}
+
 setup_prepare()
 {
 	swp1=${NETIFS[p1]}

tools/testing/selftests/drivers/net/mlxsw/tc_flower_scale.sh

@@ -2,9 +2,9 @@
 # SPDX-License-Identifier: GPL-2.0
 
 # Test for resource limit of offloaded flower rules. The test adds a given
-# number of flower matches for different IPv6 addresses, then generates traffic,
-# and ensures each was hit exactly once. This file contains functions to set up
-# a testing topology and run the test, and is meant to be sourced from a test
+# number of flower matches for different IPv6 addresses, then check the offload
+# indication for all of the tc flower rules. This file contains functions to set
+# up a testing topology and run the test, and is meant to be sourced from a test
 # script that calls the testing routine with a given number of rules.
 
 TC_FLOWER_NUM_NETIFS=2
@@ -94,22 +94,15 @@ __tc_flower_test()
 
 	tc_flower_rules_create $count $should_fail
 
-	for ((i = 0; i < count; ++i)); do
-		$MZ $h1 -q -c 1 -t ip -p 20 -b bc -6 \
-			-A 2001:db8:2::1 \
-			-B $(tc_flower_addr $i)
-	done
-
-	MISMATCHES=$(
-		tc -j -s filter show dev $h2 ingress |
-		jq -r '[ .[] | select(.kind == "flower") | .options |
-		         values as $rule | .actions[].stats.packets |
-		         select(. != 1) | "\(.) on \($rule.keys.dst_ip)" ] |
-		       join(", ")'
-	)
-
-	test -z "$MISMATCHES"
-	check_err $? "Expected to capture 1 packet for each IP, but got $MISMATCHES"
+	offload_count=$(tc -j -s filter show dev $h2 ingress    |
+			jq -r '[ .[] | select(.kind == "flower") |
+			.options | .in_hw ]' | jq .[] | wc -l)
+	[[ $((offload_count - 1)) -eq $count ]]
+	if [[ $should_fail -eq 0 ]]; then
+		check_err $? "Offload mismatch"
+	else
+		check_err_fail $should_fail $? "Offload more than expacted"
+	fi
 }
 
 tc_flower_test()

tools/testing/selftests/drivers/net/mlxsw/vxlan.sh

@@ -9,6 +9,7 @@ lib_dir=$(dirname $0)/../../../net/forwarding
 ALL_TESTS="sanitization_test offload_indication_test \
 	sanitization_vlan_aware_test offload_indication_vlan_aware_test"
 NUM_NETIFS=2
+: ${TIMEOUT:=20000} # ms
 source $lib_dir/lib.sh
 
 setup_prepare()
@@ -470,8 +471,8 @@ offload_indication_fdb_flood_test()
 
 	bridge fdb append 00:00:00:00:00:00 dev vxlan0 self dst 198.51.100.2
 
-	bridge fdb show brport vxlan0 | grep 00:00:00:00:00:00 \
-		| grep -q offload
+	busywait "$TIMEOUT" wait_for_offload grep_bridge_fdb 00:00:00:00:00:00 \
+		bridge fdb show brport vxlan0
 	check_err $?
 
 	bridge fdb del 00:00:00:00:00:00 dev vxlan0 self
@@ -486,11 +487,11 @@ offload_indication_fdb_bridge_test()
 	bridge fdb add de:ad:be:ef:13:37 dev vxlan0 self master static \
 		dst 198.51.100.2
 
-	bridge fdb show brport vxlan0 | grep de:ad:be:ef:13:37 | grep self \
-		| grep -q offload
+	busywait "$TIMEOUT" wait_for_offload grep_bridge_fdb \
+		de:ad:be:ef:13:37 self bridge fdb show brport vxlan0
 	check_err $?
-	bridge fdb show brport vxlan0 | grep de:ad:be:ef:13:37 | grep -v self \
-		| grep -q offload
+	busywait "$TIMEOUT" wait_for_offload grep_bridge_fdb \
+		de:ad:be:ef:13:37 self -v bridge fdb show brport vxlan0
 	check_err $?
 
 	log_test "vxlan entry offload indication - initial state"
@@ -500,9 +501,9 @@ offload_indication_fdb_bridge_test()
 	RET=0
 
 	bridge fdb del de:ad:be:ef:13:37 dev vxlan0 master
-	bridge fdb show brport vxlan0 | grep de:ad:be:ef:13:37 | grep self \
-		| grep -q offload
-	check_fail $?
+	busywait "$TIMEOUT" not wait_for_offload grep_bridge_fdb \
+		de:ad:be:ef:13:37 self bridge fdb show brport vxlan0
+	check_err $?
 
 	log_test "vxlan entry offload indication - after removal from bridge"
 
@@ -511,11 +512,11 @@ offload_indication_fdb_bridge_test()
 	RET=0
 
 	bridge fdb add de:ad:be:ef:13:37 dev vxlan0 master static
-	bridge fdb show brport vxlan0 | grep de:ad:be:ef:13:37 | grep self \
-		| grep -q offload
+	busywait "$TIMEOUT" wait_for_offload grep_bridge_fdb \
+		de:ad:be:ef:13:37 self bridge fdb show brport vxlan0
 	check_err $?
-	bridge fdb show brport vxlan0 | grep de:ad:be:ef:13:37 | grep -v self \
-		| grep -q offload
+	busywait "$TIMEOUT" wait_for_offload grep_bridge_fdb \
+		de:ad:be:ef:13:37 self -v bridge fdb show brport vxlan0
 	check_err $?
 
 	log_test "vxlan entry offload indication - after re-add to bridge"
@@ -525,9 +526,9 @@ offload_indication_fdb_bridge_test()
 	RET=0
 
 	bridge fdb del de:ad:be:ef:13:37 dev vxlan0 self
-	bridge fdb show brport vxlan0 | grep de:ad:be:ef:13:37 | grep -v self \
-		| grep -q offload
-	check_fail $?
+	busywait "$TIMEOUT" not wait_for_offload grep_bridge_fdb \
+		de:ad:be:ef:13:37 self -v bridge fdb show brport vxlan0
+	check_err $?
 
 	log_test "vxlan entry offload indication - after removal from vxlan"
 
@@ -536,11 +537,11 @@ offload_indication_fdb_bridge_test()
 	RET=0
 
 	bridge fdb add de:ad:be:ef:13:37 dev vxlan0 self dst 198.51.100.2
-	bridge fdb show brport vxlan0 | grep de:ad:be:ef:13:37 | grep self \
-		| grep -q offload
+	busywait "$TIMEOUT" wait_for_offload grep_bridge_fdb \
+		de:ad:be:ef:13:37 self bridge fdb show brport vxlan0
 	check_err $?
-	bridge fdb show brport vxlan0 | grep de:ad:be:ef:13:37 | grep -v self \
-		| grep -q offload
+	busywait "$TIMEOUT" wait_for_offload grep_bridge_fdb \
+		de:ad:be:ef:13:37 self -v bridge fdb show brport vxlan0
 	check_err $?
 
 	log_test "vxlan entry offload indication - after re-add to vxlan"
@@ -558,27 +559,32 @@ offload_indication_decap_route_test()
 {
 	RET=0
 
-	ip route show table local | grep 198.51.100.1 | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload \
+		ip route show table local 198.51.100.1
 	check_err $?
 
 	ip link set dev vxlan0 down
-	ip route show table local | grep 198.51.100.1 | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload \
+		ip route show table local 198.51.100.1
 	check_err $?
 
 	ip link set dev vxlan1 down
-	ip route show table local | grep 198.51.100.1 | grep -q offload
-	check_fail $?
+	busywait "$TIMEOUT" not wait_for_offload \
+		ip route show table local 198.51.100.1
+	check_err $?
 
 	log_test "vxlan decap route - vxlan device down"
 
 	RET=0
 
 	ip link set dev vxlan1 up
-	ip route show table local | grep 198.51.100.1 | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload \
+		ip route show table local 198.51.100.1
 	check_err $?
 
 	ip link set dev vxlan0 up
-	ip route show table local | grep 198.51.100.1 | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload \
+		ip route show table local 198.51.100.1
 	check_err $?
 
 	log_test "vxlan decap route - vxlan device up"
@@ -586,11 +592,13 @@ offload_indication_decap_route_test()
 	RET=0
 
 	ip address delete 198.51.100.1/32 dev lo
-	ip route show table local | grep 198.51.100.1 | grep -q offload
-	check_fail $?
+	busywait "$TIMEOUT" not wait_for_offload \
+		ip route show table local 198.51.100.1
+	check_err $?
 
 	ip address add 198.51.100.1/32 dev lo
-	ip route show table local | grep 198.51.100.1 | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload \
+		ip route show table local 198.51.100.1
 	check_err $?
 
 	log_test "vxlan decap route - add local route"
@@ -598,16 +606,19 @@ offload_indication_decap_route_test()
 	RET=0
 
 	ip link set dev $swp1 nomaster
-	ip route show table local | grep 198.51.100.1 | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload \
+		ip route show table local 198.51.100.1
 	check_err $?
 
 	ip link set dev $swp2 nomaster
-	ip route show table local | grep 198.51.100.1 | grep -q offload
-	check_fail $?
+	busywait "$TIMEOUT" not wait_for_offload \
+		ip route show table local 198.51.100.1
+	check_err $?
 
 	ip link set dev $swp1 master br0
 	ip link set dev $swp2 master br1
-	ip route show table local | grep 198.51.100.1 | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload \
+		ip route show table local 198.51.100.1
 	check_err $?
 
 	log_test "vxlan decap route - local ports enslavement"
@@ -615,12 +626,14 @@ offload_indication_decap_route_test()
 	RET=0
 
 	ip link del dev br0
-	ip route show table local | grep 198.51.100.1 | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload \
+		ip route show table local 198.51.100.1
 	check_err $?
 
 	ip link del dev br1
-	ip route show table local | grep 198.51.100.1 | grep -q offload
-	check_fail $?
+	busywait "$TIMEOUT" not wait_for_offload \
+		ip route show table local 198.51.100.1
+	check_err $?
 
 	log_test "vxlan decap route - bridge device deletion"
 
@@ -632,16 +645,19 @@ offload_indication_decap_route_test()
 	ip link set dev $swp2 master br1
 	ip link set dev vxlan0 master br0
 	ip link set dev vxlan1 master br1
-	ip route show table local | grep 198.51.100.1 | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload \
+		ip route show table local 198.51.100.1
 	check_err $?
 
 	ip link del dev vxlan0
-	ip route show table local | grep 198.51.100.1 | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload \
+		ip route show table local 198.51.100.1
 	check_err $?
 
 	ip link del dev vxlan1
-	ip route show table local | grep 198.51.100.1 | grep -q offload
-	check_fail $?
+	busywait "$TIMEOUT" not wait_for_offload \
+		ip route show table local 198.51.100.1
+	check_err $?
 
 	log_test "vxlan decap route - vxlan device deletion"
 
@@ -656,12 +672,15 @@ check_fdb_offloaded()
 	local mac=00:11:22:33:44:55
 	local zmac=00:00:00:00:00:00
 
-	bridge fdb show dev vxlan0 | grep $mac | grep self | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload grep_bridge_fdb $mac self \
+		bridge fdb show dev vxlan0
 	check_err $?
-	bridge fdb show dev vxlan0 | grep $mac | grep master | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload grep_bridge_fdb $mac master \
+		bridge fdb show dev vxlan0
 	check_err $?
 
-	bridge fdb show dev vxlan0 | grep $zmac | grep self | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload grep_bridge_fdb $zmac self \
+		bridge fdb show dev vxlan0
 	check_err $?
 }
 
@@ -672,13 +691,15 @@ check_vxlan_fdb_not_offloaded()
 
 	bridge fdb show dev vxlan0 | grep $mac | grep -q self
 	check_err $?
-	bridge fdb show dev vxlan0 | grep $mac | grep self | grep -q offload
-	check_fail $?
+	busywait "$TIMEOUT" not wait_for_offload grep_bridge_fdb $mac self \
+		bridge fdb show dev vxlan0
+	check_err $?
 
 	bridge fdb show dev vxlan0 | grep $zmac | grep -q self
 	check_err $?
-	bridge fdb show dev vxlan0 | grep $zmac | grep self | grep -q offload
-	check_fail $?
+	busywait "$TIMEOUT" not wait_for_offload grep_bridge_fdb $zmac self \
+		bridge fdb show dev vxlan0
+	check_err $?
 }
 
 check_bridge_fdb_not_offloaded()
@@ -688,8 +709,9 @@ check_bridge_fdb_not_offloaded()
 
 	bridge fdb show dev vxlan0 | grep $mac | grep -q master
 	check_err $?
-	bridge fdb show dev vxlan0 | grep $mac | grep master | grep -q offload
-	check_fail $?
+	busywait "$TIMEOUT" not wait_for_offload grep_bridge_fdb $mac master \
+		bridge fdb show dev vxlan0
+	check_err $?
 }
 
 __offload_indication_join_vxlan_first()
@@ -771,12 +793,14 @@ __offload_indication_join_vxlan_last()
 
 	ip link set dev $swp1 master br0
 
-	bridge fdb show dev vxlan0 | grep $zmac | grep self | grep -q offload
-	check_fail $?
+	busywait "$TIMEOUT" not wait_for_offload grep_bridge_fdb $zmac self \
+		bridge fdb show dev vxlan0
+	check_err $?
 
 	ip link set dev vxlan0 master br0
 
-	bridge fdb show dev vxlan0 | grep $zmac | grep self | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload grep_bridge_fdb $zmac self \
+		bridge fdb show dev vxlan0
 	check_err $?
 
 	log_test "offload indication - attach vxlan last"
@@ -866,8 +890,9 @@ sanitization_vlan_aware_test()
 	ip link set dev $swp1 master br0 &> /dev/null
 	check_fail $?
 
-	ip route show table local | grep 198.51.100.1 | grep -q offload
-	check_fail $?
+	busywait "$TIMEOUT" not wait_for_offload \
+		ip route show table local 198.51.100.1
+	check_err $?
 
 	log_test "vlan-aware - failed enslavement to bridge due to conflict"
 
@@ -929,11 +954,11 @@ offload_indication_vlan_aware_fdb_test()
 	bridge fdb add de:ad:be:ef:13:37 dev vxlan10 self master static \
 		dst 198.51.100.2 vlan 10
 
-	bridge fdb show brport vxlan10 | grep de:ad:be:ef:13:37 | grep self \
-		| grep -q offload
+	busywait "$TIMEOUT" wait_for_offload grep_bridge_fdb \
+		de:ad:be:ef:13:37 self bridge fdb show brport vxlan10
 	check_err $?
-	bridge fdb show brport vxlan10 | grep de:ad:be:ef:13:37 | grep -v self \
-		| grep -q offload
+	busywait "$TIMEOUT" wait_for_offload grep_bridge_fdb \
+		de:ad:be:ef:13:37 self -v bridge fdb show brport vxlan10
 	check_err $?
 
 	log_test "vxlan entry offload indication - initial state"
@@ -943,9 +968,9 @@ offload_indication_vlan_aware_fdb_test()
 	RET=0
 
 	bridge fdb del de:ad:be:ef:13:37 dev vxlan10 master vlan 10
-	bridge fdb show brport vxlan10 | grep de:ad:be:ef:13:37 | grep self \
-		| grep -q offload
-	check_fail $?
+	busywait "$TIMEOUT" not wait_for_offload grep_bridge_fdb \
+		de:ad:be:ef:13:37 self bridge fdb show brport vxlan10
+	check_err $?
 
 	log_test "vxlan entry offload indication - after removal from bridge"
 
@@ -954,11 +979,11 @@ offload_indication_vlan_aware_fdb_test()
 	RET=0
 
 	bridge fdb add de:ad:be:ef:13:37 dev vxlan10 master static vlan 10
-	bridge fdb show brport vxlan10 | grep de:ad:be:ef:13:37 | grep self \
-		| grep -q offload
+	busywait "$TIMEOUT" wait_for_offload grep_bridge_fdb \
+		de:ad:be:ef:13:37 self bridge fdb show brport vxlan10
 	check_err $?
-	bridge fdb show brport vxlan10 | grep de:ad:be:ef:13:37 | grep -v self \
-		| grep -q offload
+	busywait "$TIMEOUT" wait_for_offload grep_bridge_fdb \
+		de:ad:be:ef:13:37 self -v bridge fdb show brport vxlan10
 	check_err $?
 
 	log_test "vxlan entry offload indication - after re-add to bridge"
@@ -968,9 +993,9 @@ offload_indication_vlan_aware_fdb_test()
 	RET=0
 
 	bridge fdb del de:ad:be:ef:13:37 dev vxlan10 self
-	bridge fdb show brport vxlan10 | grep de:ad:be:ef:13:37 | grep -v self \
-		| grep -q offload
-	check_fail $?
+	busywait "$TIMEOUT" not wait_for_offload grep_bridge_fdb \
+		de:ad:be:ef:13:37 self -v bridge fdb show brport vxlan10
+	check_err $?
 
 	log_test "vxlan entry offload indication - after removal from vxlan"
 
@@ -979,11 +1004,11 @@ offload_indication_vlan_aware_fdb_test()
 	RET=0
 
 	bridge fdb add de:ad:be:ef:13:37 dev vxlan10 self dst 198.51.100.2
-	bridge fdb show brport vxlan10 | grep de:ad:be:ef:13:37 | grep self \
-		| grep -q offload
+	busywait "$TIMEOUT" wait_for_offload grep_bridge_fdb \
+		de:ad:be:ef:13:37 self bridge fdb show brport vxlan10
 	check_err $?
-	bridge fdb show brport vxlan10 | grep de:ad:be:ef:13:37 | grep -v self \
-		| grep -q offload
+	busywait "$TIMEOUT" wait_for_offload grep_bridge_fdb \
+		de:ad:be:ef:13:37 self -v bridge fdb show brport vxlan10
 	check_err $?
 
 	log_test "vxlan entry offload indication - after re-add to vxlan"
@@ -995,28 +1020,31 @@ offload_indication_vlan_aware_decap_route_test()
 {
 	RET=0
 
-	ip route show table local | grep 198.51.100.1 | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload \
+		ip route show table local 198.51.100.1
 	check_err $?
 
 	# Toggle PVID flag on one VxLAN device and make sure route is still
 	# marked as offloaded
 	bridge vlan add vid 10 dev vxlan10 untagged
 
-	ip route show table local | grep 198.51.100.1 | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload \
+		ip route show table local 198.51.100.1
 	check_err $?
 
 	# Toggle PVID flag on second VxLAN device and make sure route is no
 	# longer marked as offloaded
 	bridge vlan add vid 20 dev vxlan20 untagged
 
-	ip route show table local | grep 198.51.100.1 | grep -q offload
-	check_fail $?
+	busywait "$TIMEOUT" not wait_for_offload \
+		ip route show table local 198.51.100.1
+	check_err $?
 
 	# Toggle PVID flag back and make sure route is marked as offloaded
 	bridge vlan add vid 10 dev vxlan10 pvid untagged
 	bridge vlan add vid 20 dev vxlan20 pvid untagged
 
-	ip route show table local | grep 198.51.100.1 | grep -q offload
+	busywait "$TIMEOUT" wait_for_offload ip route show table local 198.51.100.1
 	check_err $?
 
 	log_test "vxlan decap route - vni map/unmap"
@@ -1069,33 +1097,33 @@ offload_indication_vlan_aware_l3vni_test()
 	ip link set dev vxlan0 master br0
 	bridge vlan add dev vxlan0 vid 10 pvid untagged
 
-	bridge fdb show brport vxlan0 | grep $zmac | grep self \
-		| grep -q offload
+	busywait "$TIMEOUT" wait_for_offload grep_bridge_fdb $zmac self \
+		bridge fdb show brport vxlan0
 	check_err $? "vxlan tunnel not offloaded when should"
 
 	# Configure a VLAN interface and make sure tunnel is offloaded
 	ip link add link br0 name br10 up type vlan id 10
 	sysctl_set net.ipv6.conf.br10.disable_ipv6 0
 	ip -6 address add 2001:db8:1::1/64 dev br10
-	bridge fdb show brport vxlan0 | grep $zmac | grep self \
-		| grep -q offload
+	busywait "$TIMEOUT" wait_for_offload grep_bridge_fdb $zmac self \
+		bridge fdb show brport vxlan0
 	check_err $? "vxlan tunnel not offloaded when should"
 
 	# Unlink the VXLAN device, make sure tunnel is no longer offloaded,
 	# then add it back to the bridge and make sure it is offloaded
 	ip link set dev vxlan0 nomaster
-	bridge fdb show brport vxlan0 | grep $zmac | grep self \
-		| grep -q offload
-	check_fail $? "vxlan tunnel offloaded after unlinked from bridge"
+	busywait "$TIMEOUT" not wait_for_offload grep_bridge_fdb $zmac self \
+		bridge fdb show brport vxlan0
+	check_err $? "vxlan tunnel offloaded after unlinked from bridge"
 
 	ip link set dev vxlan0 master br0
-	bridge fdb show brport vxlan0 | grep $zmac | grep self \
-		| grep -q offload
-	check_fail $? "vxlan tunnel offloaded despite no matching vid"
+	busywait "$TIMEOUT" not wait_for_offload grep_bridge_fdb $zmac self \
+		bridge fdb show brport vxlan0
+	check_err $? "vxlan tunnel offloaded despite no matching vid"
 
 	bridge vlan add dev vxlan0 vid 10 pvid untagged
-	bridge fdb show brport vxlan0 | grep $zmac | grep self \
-		| grep -q offload
+	busywait "$TIMEOUT" wait_for_offload grep_bridge_fdb $zmac self \
+		bridge fdb show brport vxlan0
 	check_err $? "vxlan tunnel not offloaded after adding vid"
 
 	log_test "vxlan - l3 vni"

tools/testing/selftests/net/forwarding/devlink_lib.sh

@@ -35,6 +35,12 @@ if [ $? -ne 0 ]; then
 	exit 1
 fi
 
+devlink dev help 2>&1 | grep info &> /dev/null
+if [ $? -ne 0 ]; then
+	echo "SKIP: iproute2 too old, missing devlink dev info support"
+	exit 1
+fi
+
 ##############################################################################
 # Devlink helpers
 
@@ -413,3 +419,19 @@ devlink_trap_drop_cleanup()
 	kill $mz_pid && wait $mz_pid &> /dev/null
 	tc filter del dev $dev egress protocol $proto pref $pref handle $handle flower
 }
+
+devlink_port_by_netdev()
+{
+	local if_name=$1
+
+	devlink -j port show $if_name | jq -e '.[] | keys' | jq -r '.[]'
+}
+
+devlink_cpu_port_get()
+{
+	local cpu_dl_port_num=$(devlink port list | grep "$DEVLINK_DEV" |
+				grep cpu | cut -d/ -f3 | cut -d: -f1 |
+				sed -n '1p')
+
+	echo "$DEVLINK_DEV/$cpu_dl_port_num"
+}

tools/testing/selftests/net/forwarding/lib.sh

@@ -248,6 +248,33 @@ busywait()
 	done
 }
 
+not()
+{
+	"$@"
+	[[ $? != 0 ]]
+}
+
+grep_bridge_fdb()
+{
+	local addr=$1; shift
+	local word
+	local flag
+
+	if [ "$1" == "self" ] || [ "$1" == "master" ]; then
+		word=$1; shift
+		if [ "$1" == "-v" ]; then
+			flag=$1; shift
+		fi
+	fi
+
+	$@ | grep $addr | grep $flag "$word"
+}
+
+wait_for_offload()
+{
+	"$@" | grep -q offload
+}
+
 until_counter_is()
 {
 	local value=$1; shift
@@ -607,6 +634,16 @@ ethtool_stats_get()
 	ethtool -S $dev | grep "^ *$stat:" | head -n 1 | cut -d: -f2
 }
 
+qdisc_stats_get()
+{
+	local dev=$1; shift
+	local handle=$1; shift
+	local selector=$1; shift
+
+	tc -j -s qdisc show dev "$dev" \
+	    | jq '.[] | select(.handle == "'"$handle"'") | '"$selector"
+}
+
 humanize()
 {
 	local speed=$1; shift
@@ -1132,18 +1169,29 @@ flood_test()
 	flood_multicast_test $br_port $host1_if $host2_if
 }
 
-start_traffic()
+__start_traffic()
 {
+	local proto=$1; shift
 	local h_in=$1; shift    # Where the traffic egresses the host
 	local sip=$1; shift
 	local dip=$1; shift
 	local dmac=$1; shift
 
 	$MZ $h_in -p 8000 -A $sip -B $dip -c 0 \
-		-a own -b $dmac -t udp -q &
+		-a own -b $dmac -t "$proto" -q "$@" &
 	sleep 1
 }
 
+start_traffic()
+{
+	__start_traffic udp "$@"
+}
+
+start_tcp_traffic()
+{
+	__start_traffic tcp "$@"
+}
+
 stop_traffic()
 {
 	# Suppress noise from killing mausezahn.