linux-stable.git/tools/testing/selftests/net/forwarding, branch linux-6.16.y Linux kernel stable tree Merge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net 2025-05-01T22:11:38+00:00 Jakub Kicinski kuba@kernel.org 2025-05-01T22:11:17+00:00 337079d31fb160959f87d7aa22a85b123c500c48 Cross-merge networking fixes after downstream PR (net-6.15-rc5). No conflicts or adjacent changes. Signed-off-by: Jakub Kicinski <kuba@kernel.org> 
Cross-merge networking fixes after downstream PR (net-6.15-rc5).

No conflicts or adjacent changes.

Signed-off-by: Jakub Kicinski <kuba@kernel.org>
selftests: net: tc_taprio: new test 2025-04-29T21:44:34+00:00 Vladimir Oltean vladimir.oltean@nxp.com 2025-04-26T14:48:58+00:00 4eb9da050f005fbbb7d301e8e99cfdb6e4771a0d Add a forwarding path test for tc-taprio, based on isochron. This is specifically intended for NICs with an offloaded data path (switchdev/DSA) and requires taprio 'flags 2'. Also, $h1 and $h2 must support hardware timestamping, and $h1 tc-etf offload, for isochron to work. Packets received by a switch while the egress port has a taprio schedule with an open gate for the traffic class must be sent right away. Packets received by the switch while the traffic class gate must be delayed until it opens. Packets received by the switch must be dropped if the gate for the traffic class never opens. Packets should pass if the maximum SDU for the traffic class allows it, and should be dropped otherwise. The schedule should auto-update itself if clock jumps take place while taprio is installed. Repeat most of the above tests after forcing two clock jumps, one backwards (in Jan 1970) and one back into the present. Symlink it from tools/testing/selftests/drivers/net/dsa, because usually DSA ports have the same MAC address, and we need STABLE_MAC_ADDRS=yes from its forwarding.config for the test to run successfully. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Link: https://patch.msgid.link/20250426144859.3128352-5-vladimir.oltean@nxp.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> 
Add a forwarding path test for tc-taprio, based on isochron. This is
specifically intended for NICs with an offloaded data path (switchdev/DSA)
and requires taprio 'flags 2'. Also, $h1 and $h2 must support hardware
timestamping, and $h1 tc-etf offload, for isochron to work.

Packets received by a switch while the egress port has a taprio schedule
with an open gate for the traffic class must be sent right away.

Packets received by the switch while the traffic class gate must be
delayed until it opens.

Packets received by the switch must be dropped if the gate for the
traffic class never opens.

Packets should pass if the maximum SDU for the traffic class allows it,
and should be dropped otherwise.

The schedule should auto-update itself if clock jumps take place while
taprio is installed. Repeat most of the above tests after forcing two
clock jumps, one backwards (in Jan 1970) and one back into the present.

Symlink it from tools/testing/selftests/drivers/net/dsa, because usually
DSA ports have the same MAC address, and we need STABLE_MAC_ADDRS=yes
from its forwarding.config for the test to run successfully.

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Link: https://patch.msgid.link/20250426144859.3128352-5-vladimir.oltean@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
selftests: net: tsn_lib: add window_size argument to isochron_do() 2025-04-29T21:44:34+00:00 Vladimir Oltean vladimir.oltean@nxp.com 2025-04-26T14:48:57+00:00 f52fe6efd61f54c5cb0e19ef1fde96cf23048a70 Make out-of-band testing (send a packet when its traffic class gate is closed, expecting it to be delayed) more predictable by allowing the window size to be customized by isochron_do(). From man isochron-send, the window size alters the advance time (the delta between the transmission time of the packet, and its expected TX time when using SO_TXTIME or tc-taprio on the sender). In absence of the argument, isochron-send defaults to maximizing the advance time (making it equal to the cycle length). The default behavior is exactly what is problematic. An advance time that is too large will make packets intended to be out-of-band still be potentially in-band with an open gate from the schedule's previous cycle. We need to allow that advance time to be reduced. Perhaps a bit confusingly, isochron_do() has a shift_time argument currently, but that does not help here. The shift time shifts both the user space wakeup time and the expected TX time by equal amounts, it is unable of bringing them closer to one another. Set the window size properly for the Ocelot PSFP selftest as well. That used to work due to a very carefully chosen SHIFT_TIME_NS. I've re-tested that the test still works properly. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Link: https://patch.msgid.link/20250426144859.3128352-4-vladimir.oltean@nxp.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> 
Make out-of-band testing (send a packet when its traffic class gate is
closed, expecting it to be delayed) more predictable by allowing the
window size to be customized by isochron_do().

From man isochron-send, the window size alters the advance time (the
delta between the transmission time of the packet, and its expected TX
time when using SO_TXTIME or tc-taprio on the sender). In absence of the
argument, isochron-send defaults to maximizing the advance time (making
it equal to the cycle length).

The default behavior is exactly what is problematic. An advance time
that is too large will make packets intended to be out-of-band still be
potentially in-band with an open gate from the schedule's previous cycle.
We need to allow that advance time to be reduced.

Perhaps a bit confusingly, isochron_do() has a shift_time argument
currently, but that does not help here. The shift time shifts both the
user space wakeup time and the expected TX time by equal amounts, it is
unable of bringing them closer to one another.

Set the window size properly for the Ocelot PSFP selftest as well.
That used to work due to a very carefully chosen SHIFT_TIME_NS.
I've re-tested that the test still works properly.

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Link: https://patch.msgid.link/20250426144859.3128352-4-vladimir.oltean@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
selftests: net: tsn_lib: create common helper for counting received packets 2025-04-29T21:44:34+00:00 Vladimir Oltean vladimir.oltean@nxp.com 2025-04-26T14:48:56+00:00 efa6eb7d77aaf5b05eed25c0ecbf7754cc325c83 This snippet will be necessary for a future isochron-based test, so provide a simpler high-level interface for counting the received packets. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Link: https://patch.msgid.link/20250426144859.3128352-3-vladimir.oltean@nxp.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> 
This snippet will be necessary for a future isochron-based test, so
provide a simpler high-level interface for counting the received
packets.

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Link: https://patch.msgid.link/20250426144859.3128352-3-vladimir.oltean@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
selftests: net: bridge_vlan_aware: test untagged/8021p-tagged with and without PVID 2025-04-26T01:45:43+00:00 Vladimir Oltean vladimir.oltean@nxp.com 2025-04-24T22:37:34+00:00 bf9de1dcd0eecd16020a677c900a70ea9b0a9714 Recent discussions around commit ad1afb003939 ("vlan_dev: VLAN 0 should be treated as "no vlan tag" (802.1p packet)") have sparked the question what happens with the DSA (and possibly other switchdev) data path when the bridge says that ports should have no PVID VLAN, but the 8021q module, as the result of a NETDEV_UP event, decides it should add VID 0 to the RX filter of those bridge ports. Do those bridge ports receive packets tagged with VID 0 or not, now? We don't know, there is no test. In the veth realm, this passes trivially, because veth is not VLAN filtering and this, the 8021q module lacks the instinct to add VID 0 in the first place. In the realm of VLAN filtering NICs with no switchdev offload, this should also pass, because the VLAN groups of the software bridge are consulted, where it can clearly be seen that a PVID is missing, even though the packet was initially accepted by the NIC. The test only poses a challenge for switchdev drivers, which usually have to program to hardware both VLANs from RX filtering, as well as from switchdev. Especially when a switchdev port joins a VLAN-aware bridge, it is unavoidable that it gains the NETIF_F_HW_VLAN_CTAG_FILTER feature, i.e. any 8021q uppers that the bridge port may have must also be committed to the RX filtering table of the interface. When a VLAN-tagged packet is physically received by the port, it is initially indistinguishable whether it will reach the bridge data path or the 8021q upper data path. That is rather the final step of the new tests that we introduce. We need to build context up to that stage, which means the following: - we need to test that 802.1p (VID 0) tagged traffic is received in the first place (on bridge ports with a valid PVID). This is the "8021p" test. - we need to test that the usual paths of reaching a configuration with no PVID on a bridge port are all covered and they all reach the same state. Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ido Schimmel <idosch@nvidia.com> Tested-by: Ido Schimmel <idosch@nvidia.com> Link: https://patch.msgid.link/20250424223734.3096202-2-vladimir.oltean@nxp.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> 
Recent discussions around commit ad1afb003939 ("vlan_dev: VLAN 0 should
be treated as "no vlan tag" (802.1p packet)") have sparked the question
what happens with the DSA (and possibly other switchdev) data path when
the bridge says that ports should have no PVID VLAN, but the 8021q
module, as the result of a NETDEV_UP event, decides it should add VID 0
to the RX filter of those bridge ports. Do those bridge ports receive
packets tagged with VID 0 or not, now? We don't know, there is no test.

In the veth realm, this passes trivially, because veth is not VLAN
filtering and this, the 8021q module lacks the instinct to add VID 0 in
the first place.

In the realm of VLAN filtering NICs with no switchdev offload, this
should also pass, because the VLAN groups of the software bridge are
consulted, where it can clearly be seen that a PVID is missing, even
though the packet was initially accepted by the NIC.

The test only poses a challenge for switchdev drivers, which usually
have to program to hardware both VLANs from RX filtering, as well as
from switchdev. Especially when a switchdev port joins a VLAN-aware
bridge, it is unavoidable that it gains the NETIF_F_HW_VLAN_CTAG_FILTER
feature, i.e. any 8021q uppers that the bridge port may have must also
be committed to the RX filtering table of the interface. When a
VLAN-tagged packet is physically received by the port, it is initially
indistinguishable whether it will reach the bridge data path or the
8021q upper data path.

That is rather the final step of the new tests that we introduce.
We need to build context up to that stage, which means the following:

- we need to test that 802.1p (VID 0) tagged traffic is received in the
  first place (on bridge ports with a valid PVID). This is the "8021p"
  test.

- we need to test that the usual paths of reaching a configuration with
  no PVID on a bridge port are all covered and they all reach the same
  state.

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Tested-by: Ido Schimmel <idosch@nvidia.com>
Link: https://patch.msgid.link/20250424223734.3096202-2-vladimir.oltean@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
selftests: net/bridge : add tests for per vlan snooping with stp state changes 2025-04-23T12:02:20+00:00 Yong Wang yongwang@nvidia.com 2025-04-17T13:43:14+00:00 aea45363e29dd16050e6ce333ce0d3696ac3b5a9 Change ALL_TESTS definition to "test-per-line". Add the test case of per vlan snooping with port stp state change to forwarding and also vlan equivalent case in both bridge_igmp.sh and bridge_mld.sh. Signed-off-by: Yong Wang <yongwang@nvidia.com> Reviewed-by: Andy Roulin <aroulin@nvidia.com> Reviewed-by: Ido Schimmel <idosch@nvidia.com> Signed-off-by: Petr Machata <petrm@nvidia.com> Acked-by: Nikolay Aleksandrov <razor@blackwall.org> Signed-off-by: David S. Miller <davem@davemloft.net> 
Change ALL_TESTS definition to "test-per-line".

Add the test case of per vlan snooping with port stp state change to
forwarding and also vlan equivalent case in both bridge_igmp.sh and
bridge_mld.sh.

Signed-off-by: Yong Wang <yongwang@nvidia.com>
Reviewed-by: Andy Roulin <aroulin@nvidia.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Acked-by: Nikolay Aleksandrov <razor@blackwall.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
selftests: vxlan_bridge: Test flood with unresolved FDB entry 2025-03-24T22:09:31+00:00 Amit Cohen amcohen@nvidia.com 2025-03-17T17:37:31+00:00 36ed81bcade9b96e0c7be7cb6bdb5b6773a3ace0 Extend flood test to configure FDB entry with unresolved destination IP, check that packets are not sent twice. Without the previous patch which handles such scenario in mlxsw, the tests fail: $ TESTS='test_flood' ./vxlan_bridge_1d.sh Running tests with UDP port 4789 TEST: VXLAN: flood [ OK ] TEST: VXLAN: flood, unresolved FDB entry [FAIL] vx2 ns2: Expected to capture 10 packets, got 20. $ TESTS='test_flood' ./vxlan_bridge_1q.sh INFO: Running tests with UDP port 4789 TEST: VXLAN: flood vlan 10 [ OK ] TEST: VXLAN: flood vlan 20 [ OK ] TEST: VXLAN: flood vlan 10, unresolved FDB entry [FAIL] vx10 ns2: Expected to capture 10 packets, got 20. TEST: VXLAN: flood vlan 20, unresolved FDB entry [FAIL] vx20 ns2: Expected to capture 10 packets, got 20. With the previous patch, the tests pass. Signed-off-by: Amit Cohen <amcohen@nvidia.com> Reviewed-by: Petr Machata <petrm@nvidia.com> Reviewed-by: Ido Schimmel <idosch@nvidia.com> Signed-off-by: Petr Machata <petrm@nvidia.com> Reviewed-by: Simon Horman <horms@kernel.org> Link: https://patch.msgid.link/7bc96e317531f3bf06319fb2ea447bd8666f29fa.1742224300.git.petrm@nvidia.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> 
Extend flood test to configure FDB entry with unresolved destination IP,
check that packets are not sent twice.

Without the previous patch which handles such scenario in mlxsw, the
tests fail:

$ TESTS='test_flood' ./vxlan_bridge_1d.sh
Running tests with UDP port 4789
TEST: VXLAN: flood                                                  [ OK ]
TEST: VXLAN: flood, unresolved FDB entry                            [FAIL]
        vx2 ns2: Expected to capture 10 packets, got 20.

$ TESTS='test_flood' ./vxlan_bridge_1q.sh
INFO: Running tests with UDP port 4789
TEST: VXLAN: flood vlan 10                                          [ OK ]
TEST: VXLAN: flood vlan 20                                          [ OK ]
TEST: VXLAN: flood vlan 10, unresolved FDB entry                    [FAIL]
        vx10 ns2: Expected to capture 10 packets, got 20.
TEST: VXLAN: flood vlan 20, unresolved FDB entry                    [FAIL]
        vx20 ns2: Expected to capture 10 packets, got 20.

With the previous patch, the tests pass.

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/7bc96e317531f3bf06319fb2ea447bd8666f29fa.1742224300.git.petrm@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
net: rename netns_local to netns_immutable 2025-03-04T11:44:48+00:00 Nicolas Dichtel nicolas.dichtel@6wind.com 2025-02-28T10:20:56+00:00 0c493da86374dffff7505e67289ad75b21f5b301 The name 'netns_local' is confusing. A following commit will export it via netlink, so let's use a more explicit name. Reported-by: Eric Dumazet <edumazet@google.com> Suggested-by: Kuniyuki Iwashima <kuniyu@amazon.com> Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com> Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com> 
The name 'netns_local' is confusing. A following commit will export it via
netlink, so let's use a more explicit name.

Reported-by: Eric Dumazet <edumazet@google.com>
Suggested-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Merge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net 2025-02-20T18:37:30+00:00 Jakub Kicinski kuba@kernel.org 2025-02-20T18:36:34+00:00 5d6ba5ab8582aa35c1ee98e47af28e6f6772596c Cross-merge networking fixes after downstream PR (net-6.14-rc4). No conflicts or adjacent changes. Signed-off-by: Jakub Kicinski <kuba@kernel.org> 
Cross-merge networking fixes after downstream PR (net-6.14-rc4).

No conflicts or adjacent changes.

Signed-off-by: Jakub Kicinski <kuba@kernel.org>
selftests/net/forwarding: Add a test case for tc-flower of mixed port and port-range 2025-02-20T02:54:59+00:00 Cong Wang xiyou.wangcong@gmail.com 2025-02-18T04:32:08+00:00 dfc1580f960bf70bdaacda8f3d644e3e58160f9d After this patch: # ./tc_flower_port_range.sh TEST: Port range matching - IPv4 UDP [ OK ] TEST: Port range matching - IPv4 TCP [ OK ] TEST: Port range matching - IPv6 UDP [ OK ] TEST: Port range matching - IPv6 TCP [ OK ] TEST: Port range matching - IPv4 UDP Drop [ OK ] Cc: Qiang Zhang <dtzq01@gmail.com> Cc: Jamal Hadi Salim <jhs@mojatatu.com> Cc: Jiri Pirko <jiri@resnulli.us> Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com> Reviewed-by: Ido Schimmel <idosch@nvidia.com> Tested-by: Ido Schimmel <idosch@nvidia.com> Link: https://patch.msgid.link/20250218043210.732959-3-xiyou.wangcong@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> 
After this patch:

 # ./tc_flower_port_range.sh
 TEST: Port range matching - IPv4 UDP                                [ OK ]
 TEST: Port range matching - IPv4 TCP                                [ OK ]
 TEST: Port range matching - IPv6 UDP                                [ OK ]
 TEST: Port range matching - IPv6 TCP                                [ OK ]
 TEST: Port range matching - IPv4 UDP Drop                           [ OK ]

Cc: Qiang Zhang <dtzq01@gmail.com>
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Cc: Jiri Pirko <jiri@resnulli.us>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Tested-by: Ido Schimmel <idosch@nvidia.com>
Link: https://patch.msgid.link/20250218043210.732959-3-xiyou.wangcong@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>