1 .. SPDX-License-Identifier: BSD-3-Clause
2 Copyright(c) 2010-2016 Intel Corporation.
6 Testpmd Runtime Functions
7 =========================
9 Where the testpmd application is started in interactive mode, (``-i|--interactive``),
10 it displays a prompt that can be used to start and stop forwarding,
11 configure the application, display statistics (including the extended NIC
12 statistics aka xstats) , set the Flow Director and other tasks::
16 The testpmd prompt has some, limited, readline support.
17 Common bash command-line functions such as ``Ctrl+a`` and ``Ctrl+e`` to go to the start and end of the prompt line are supported
18 as well as access to the command history via the up-arrow.
20 There is also support for tab completion.
21 If you type a partial command and hit ``<TAB>`` you get a list of the available completions:
23 .. code-block:: console
25 testpmd> show port <TAB>
27 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
28 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
29 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
30 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
36 Some examples in this document are too long to fit on one line are are shown wrapped at `"\\"` for display purposes::
38 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
39 (pause_time) (send_xon) (port_id)
41 In the real ``testpmd>`` prompt these commands should be on a single line.
46 The testpmd has on-line help for the functions that are available at runtime.
47 These are divided into sections and can be accessed using help, help section or help all:
49 .. code-block:: console
53 help control : Start and stop forwarding.
54 help display : Displaying port, stats and config information.
55 help config : Configuration information.
56 help ports : Configuring ports.
57 help registers : Reading and setting port registers.
58 help filters : Filters configuration help.
59 help all : All of the above sections.
62 Command File Functions
63 ----------------------
65 To facilitate loading large number of commands or to avoid cutting and pasting where not
66 practical or possible testpmd supports alternative methods for executing commands.
68 * If started with the ``--cmdline-file=FILENAME`` command line argument testpmd
69 will execute all CLI commands contained within the file immediately before
70 starting packet forwarding or entering interactive mode.
72 .. code-block:: console
74 ./testpmd -n4 -r2 ... -- -i --cmdline-file=/home/ubuntu/flow-create-commands.txt
75 Interactive-mode selected
76 CLI commands to be read from /home/ubuntu/flow-create-commands.txt
77 Configuring Port 0 (socket 0)
78 Port 0: 7C:FE:90:CB:74:CE
79 Configuring Port 1 (socket 0)
80 Port 1: 7C:FE:90:CB:74:CA
81 Checking link statuses...
82 Port 0 Link Up - speed 10000 Mbps - full-duplex
83 Port 1 Link Up - speed 10000 Mbps - full-duplex
89 Flow rule #498 created
90 Flow rule #499 created
91 Read all CLI commands from /home/ubuntu/flow-create-commands.txt
95 * At run-time additional commands can be loaded in bulk by invoking the ``load FILENAME``
98 .. code-block:: console
100 testpmd> load /home/ubuntu/flow-create-commands.txt
105 Flow rule #498 created
106 Flow rule #499 created
107 Read all CLI commands from /home/ubuntu/flow-create-commands.txt
111 In all cases output from any included command will be displayed as standard output.
112 Execution will continue until the end of the file is reached regardless of
113 whether any errors occur. The end user must examine the output to determine if
114 any failures occurred.
123 Start packet forwarding with current configuration::
130 Start packet forwarding with current configuration after sending specified number of bursts of packets::
132 testpmd> start tx_first (""|burst_num)
134 The default burst number is 1 when ``burst_num`` not presented.
139 Stop packet forwarding, and display accumulated statistics::
154 The functions in the following sections are used to display information about the
155 testpmd configuration or the NIC status.
160 Display information for a given port or all ports::
162 testpmd> show port (info|stats|xstats|fdir|stat_qmap|dcb_tc|cap) (port_id|all)
164 The available information categories are:
166 * ``info``: General port information such as MAC address.
168 * ``stats``: RX/TX statistics.
170 * ``xstats``: RX/TX extended NIC statistics.
172 * ``fdir``: Flow Director information and statistics.
174 * ``stat_qmap``: Queue statistics mapping.
176 * ``dcb_tc``: DCB information such as TC mapping.
178 * ``cap``: Supported offload capabilities.
182 .. code-block:: console
184 testpmd> show port info 0
186 ********************* Infos for port 0 *********************
188 MAC address: XX:XX:XX:XX:XX:XX
190 memory allocation on the socket: 0
192 Link speed: 40000 Mbps
193 Link duplex: full-duplex
194 Promiscuous mode: enabled
195 Allmulticast mode: disabled
196 Maximum number of MAC addresses: 64
197 Maximum number of MAC addresses of hash filtering: 0
202 Redirection table size: 512
203 Supported flow types:
223 Display the rss redirection table entry indicated by masks on port X::
225 testpmd> show port (port_id) rss reta (size) (mask0, mask1...)
227 size is used to indicate the hardware supported reta size
232 Display the RSS hash functions and RSS hash key of a port::
234 testpmd> show port (port_id) rss-hash ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2-payload|ipv6-ex|ipv6-tcp-ex|ipv6-udp-ex [key]
239 Clear the port statistics for a given port or for all ports::
241 testpmd> clear port (info|stats|xstats|fdir|stat_qmap) (port_id|all)
245 testpmd> clear port stats all
250 Display information for a given port's RX/TX queue::
252 testpmd> show (rxq|txq) info (port_id) (queue_id)
257 Displays the configuration of the application.
258 The configuration comes from the command-line, the runtime or the application defaults::
260 testpmd> show config (rxtx|cores|fwd|txpkts)
262 The available information categories are:
264 * ``rxtx``: RX/TX configuration items.
266 * ``cores``: List of forwarding cores.
268 * ``fwd``: Packet forwarding configuration.
270 * ``txpkts``: Packets to TX configuration.
274 .. code-block:: console
276 testpmd> show config rxtx
278 io packet forwarding - CRC stripping disabled - packets/burst=16
279 nb forwarding cores=2 - nb forwarding ports=1
280 RX queues=1 - RX desc=128 - RX free threshold=0
281 RX threshold registers: pthresh=8 hthresh=8 wthresh=4
282 TX queues=1 - TX desc=512 - TX free threshold=0
283 TX threshold registers: pthresh=36 hthresh=0 wthresh=0
284 TX RS bit threshold=0 - TXQ flags=0x0
289 Set the packet forwarding mode::
291 testpmd> set fwd (io|mac|macswap|flowgen| \
292 rxonly|txonly|csum|icmpecho) (""|retry)
294 ``retry`` can be specified for forwarding engines except ``rx_only``.
296 The available information categories are:
298 * ``io``: Forwards packets "as-is" in I/O mode.
299 This is the fastest possible forwarding operation as it does not access packets data.
300 This is the default mode.
302 * ``mac``: Changes the source and the destination Ethernet addresses of packets before forwarding them.
303 Default application behaviour is to set source Ethernet address to that of the transmitting interface, and destination
304 address to a dummy value (set during init). The user may specify a target destination Ethernet address via the 'eth-peer' or
305 'eth-peer-configfile' command-line options. It is not currently possible to specify a specific source Ethernet address.
307 * ``macswap``: MAC swap forwarding mode.
308 Swaps the source and the destination Ethernet addresses of packets before forwarding them.
310 * ``flowgen``: Multi-flow generation mode.
311 Originates a number of flows (with varying destination IP addresses), and terminate receive traffic.
313 * ``rxonly``: Receives packets but doesn't transmit them.
315 * ``txonly``: Generates and transmits packets without receiving any.
317 * ``csum``: Changes the checksum field with hardware or software methods depending on the offload flags on the packet.
319 * ``icmpecho``: Receives a burst of packets, lookup for IMCP echo requests and, if any, send back ICMP echo replies.
321 * ``ieee1588``: Demonstrate L2 IEEE1588 V2 PTP timestamping for RX and TX. Requires ``CONFIG_RTE_LIBRTE_IEEE1588=y``.
323 * ``tm``: Traffic Management forwarding mode
324 Demonstrates the use of ethdev traffic management APIs and softnic PMD for
325 QoS traffic management. In this mode, 5-level hierarchical QoS scheduler is
326 available as an default option that can be enabled through CLI. The user can
327 also modify the default hierarchy or specify the new hierarchy through CLI for
328 implementing QoS scheduler. Requires ``CONFIG_RTE_LIBRTE_PMD_SOFTNIC=y`` ``CONFIG_RTE_LIBRTE_SCHED=y``.
332 testpmd> set fwd rxonly
334 Set rxonly packet forwarding mode
340 Display an RX descriptor for a port RX queue::
342 testpmd> read rxd (port_id) (queue_id) (rxd_id)
346 testpmd> read rxd 0 0 4
347 0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180
352 Display a TX descriptor for a port TX queue::
354 testpmd> read txd (port_id) (queue_id) (txd_id)
358 testpmd> read txd 0 0 4
359 0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C
364 Get loaded dynamic device personalization (DDP) package info list::
366 testpmd> ddp get list (port_id)
371 Display information about dynamic device personalization (DDP) profile::
373 testpmd> ddp get info (profile_path)
378 Display VF statistics::
380 testpmd> show vf stats (port_id) (vf_id)
385 Reset VF statistics::
387 testpmd> clear vf stats (port_id) (vf_id)
389 show port pctype mapping
390 ~~~~~~~~~~~~~~~~~~~~~~~~
392 List all items from the pctype mapping table::
394 testpmd> show port (port_id) pctype mapping
396 show rx offloading capabilities
397 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
399 List all per queue and per port Rx offloading capabilities of a port::
401 testpmd> show port (port_id) rx_offload capabilities
403 show rx offloading configuration
404 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
406 List port level and all queue level Rx offloading configuration::
408 testpmd> show port (port_id) rx_offload configuration
410 show tx offloading capabilities
411 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
413 List all per queue and per port Tx offloading capabilities of a port::
415 testpmd> show port (port_id) tx_offload capabilities
417 show tx offloading configuration
418 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
420 List port level and all queue level Tx offloading configuration::
422 testpmd> show port (port_id) tx_offload configuration
425 Configuration Functions
426 -----------------------
428 The testpmd application can be configured from the runtime as well as from the command-line.
430 This section details the available configuration functions that are available.
434 Configuration changes only become active when forwarding is started/restarted.
439 Reset forwarding to the default configuration::
446 Set the debug verbosity level::
448 testpmd> set verbose (level)
450 Currently the only available levels are 0 (silent except for error) and 1 (fully verbose).
455 Set the log level for a log type::
457 testpmd> set log global|(type) (level)
461 * ``type`` is the log name.
463 * ``level`` is the log level.
465 For example, to change the global log level::
466 testpmd> set log global (level)
468 Regexes can also be used for type. To change log level of user1, user2 and user3::
469 testpmd> set log user[1-3] (level)
474 Set the number of ports used by the application:
478 This is equivalent to the ``--nb-ports`` command-line option.
483 Set the number of cores used by the application::
485 testpmd> set nbcore (num)
487 This is equivalent to the ``--nb-cores`` command-line option.
491 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port.
496 Set the forwarding cores hexadecimal mask::
498 testpmd> set coremask (mask)
500 This is equivalent to the ``--coremask`` command-line option.
504 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding.
509 Set the forwarding ports hexadecimal mask::
511 testpmd> set portmask (mask)
513 This is equivalent to the ``--portmask`` command-line option.
518 Set number of packets per burst::
520 testpmd> set burst (num)
522 This is equivalent to the ``--burst command-line`` option.
524 When retry is enabled, the transmit delay time and number of retries can also be set::
526 testpmd> set burst tx delay (microseconds) retry (num)
531 Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode::
533 testpmd> set txpkts (x[,y]*)
535 Where x[,y]* represents a CSV list of values, without white space.
540 Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes::
542 testpmd> set txsplit (off|on|rand)
546 * ``off`` disable packet copy & split for CSUM mode.
548 * ``on`` split outgoing packet into multiple segments. Size of each segment
549 and number of segments per packet is determined by ``set txpkts`` command
552 * ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments.
557 Set the list of forwarding cores::
559 testpmd> set corelist (x[,y]*)
561 For example, to change the forwarding cores:
563 .. code-block:: console
565 testpmd> set corelist 3,1
566 testpmd> show config fwd
568 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled
569 Logical Core 3 (socket 0) forwards packets on 1 streams:
570 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01
571 Logical Core 1 (socket 0) forwards packets on 1 streams:
572 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
576 The cores are used in the same order as specified on the command line.
581 Set the list of forwarding ports::
583 testpmd> set portlist (x[,y]*)
585 For example, to change the port forwarding:
587 .. code-block:: console
589 testpmd> set portlist 0,2,1,3
590 testpmd> show config fwd
592 io packet forwarding - ports=4 - cores=1 - streams=4
593 Logical Core 3 (socket 0) forwards packets on 4 streams:
594 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01
595 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
596 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03
597 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02
602 Enable/disable tx loopback::
604 testpmd> set tx loopback (port_id) (on|off)
609 set drop enable bit for all queues::
611 testpmd> set all queues drop (port_id) (on|off)
613 set split drop enable (for VF)
614 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
616 set split drop enable bit for VF from PF::
618 testpmd> set vf split drop (port_id) (vf_id) (on|off)
620 set mac antispoof (for VF)
621 ~~~~~~~~~~~~~~~~~~~~~~~~~~
623 Set mac antispoof for a VF from the PF::
625 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off)
630 Enable/disable MACsec offload::
632 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off)
633 testpmd> set macsec offload (port_id) off
638 Configure MACsec secure connection (SC)::
640 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi)
644 The pi argument is ignored for tx.
645 Check the NIC Datasheet for hardware limits.
650 Configure MACsec secure association (SA)::
652 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key)
656 The IDX value must be 0 or 1.
657 Check the NIC Datasheet for hardware limits.
659 set broadcast mode (for VF)
660 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
662 Set broadcast mode for a VF from the PF::
664 testpmd> set vf broadcast (port_id) (vf_id) (on|off)
669 Set the VLAN strip on a port::
671 testpmd> vlan set strip (on|off) (port_id)
676 Set the VLAN strip for a queue on a port::
678 testpmd> vlan set stripq (on|off) (port_id,queue_id)
680 vlan set stripq (for VF)
681 ~~~~~~~~~~~~~~~~~~~~~~~~
683 Set VLAN strip for all queues in a pool for a VF from the PF::
685 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off)
687 vlan set insert (for VF)
688 ~~~~~~~~~~~~~~~~~~~~~~~~
690 Set VLAN insert for a VF from the PF::
692 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id)
694 vlan set tag (for VF)
695 ~~~~~~~~~~~~~~~~~~~~~
697 Set VLAN tag for a VF from the PF::
699 testpmd> set vf vlan tag (port_id) (vf_id) (on|off)
701 vlan set antispoof (for VF)
702 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
704 Set VLAN antispoof for a VF from the PF::
706 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off)
711 Set the VLAN filter on a port::
713 testpmd> vlan set filter (on|off) (port_id)
718 Set the VLAN QinQ (extended queue in queue) on for a port::
720 testpmd> vlan set qinq (on|off) (port_id)
725 Set the inner or outer VLAN TPID for packet filtering on a port::
727 testpmd> vlan set (inner|outer) tpid (value) (port_id)
731 TPID value must be a 16-bit number (value <= 65536).
736 Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID::
738 testpmd> rx_vlan add (vlan_id|all) (port_id)
742 VLAN filter must be set on that port. VLAN ID < 4096.
743 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries
744 in VFTA table. This is important if enabling all vlan_ids.
749 Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID::
751 testpmd> rx_vlan rm (vlan_id|all) (port_id)
756 Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID::
758 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask)
763 Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID::
765 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask)
770 Add a tunnel filter on a port::
772 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \
773 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
774 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
776 The available information categories are:
778 * ``vxlan``: Set tunnel type as VXLAN.
780 * ``nvgre``: Set tunnel type as NVGRE.
782 * ``ipingre``: Set tunnel type as IP-in-GRE.
784 * ``imac-ivlan``: Set filter type as Inner MAC and VLAN.
786 * ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID.
788 * ``imac-tenid``: Set filter type as Inner MAC and tenant ID.
790 * ``imac``: Set filter type as Inner MAC.
792 * ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID.
794 * ``oip``: Set filter type as Outer IP.
796 * ``iip``: Set filter type as Inner IP.
800 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \
801 192.168.2.2 0 ipingre oip 1 1
803 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP.
808 Remove a tunnel filter on a port::
810 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \
811 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
812 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
817 Add an UDP port for VXLAN packet filter on a port::
819 testpmd> rx_vxlan_port add (udp_port) (port_id)
824 Remove an UDP port for VXLAN packet filter on a port::
826 testpmd> rx_vxlan_port rm (udp_port) (port_id)
831 Set hardware insertion of VLAN IDs in packets sent on a port::
833 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer]
835 For example, set a single VLAN ID (5) insertion on port 0::
839 Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1::
847 Set port based hardware insertion of VLAN ID in packets sent on a port::
849 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off)
854 Disable hardware insertion of a VLAN header in packets sent on a port::
856 testpmd> tx_vlan reset (port_id)
861 Select hardware or software calculation of the checksum when
862 transmitting a packet using the ``csum`` forwarding engine::
864 testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id)
868 * ``ip|udp|tcp|sctp`` always relate to the inner layer.
870 * ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized
871 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are
872 supported). See also the ``csum parse-tunnel`` command.
876 Check the NIC Datasheet for hardware limits.
881 Set RSS queue region span on a port::
883 testpmd> set port (port_id) queue-region region_id (value) \
884 queue_start_index (value) queue_num (value)
886 Set flowtype mapping on a RSS queue region on a port::
888 testpmd> set port (port_id) queue-region region_id (value) flowtype (value)
892 * For the flowtype(pctype) of packet,the specific index for each type has
893 been defined in file i40e_type.h as enum i40e_filter_pctype.
895 Set user priority mapping on a RSS queue region on a port::
897 testpmd> set port (port_id) queue-region UP (value) region_id (value)
899 Flush all queue region related configuration on a port::
901 testpmd> set port (port_id) queue-region flush (on|off)
905 * "on"is just an enable function which server for other configuration,
906 it is for all configuration about queue region from up layer,
907 at first will only keep in DPDK softwarestored in driver,
908 only after "flush on", it commit all configuration to HW.
909 "off" is just clean all configuration about queue region just now,
910 and restore all to DPDK i40e driver default config when start up.
912 Show all queue region related configuration info on a port::
914 testpmd> show port (port_id) queue-region
918 Queue region only support on PF by now, so these command is
919 only for configuration of queue region on PF port.
924 Define how tunneled packets should be handled by the csum forward
927 testpmd> csum parse-tunnel (on|off) (tx_port_id)
929 If enabled, the csum forward engine will try to recognize supported
930 tunnel headers (vxlan, gre, ipip).
932 If disabled, treat tunnel packets as non-tunneled packets (a inner
933 header is handled as a packet payload).
937 The port argument is the TX port like in the ``csum set`` command.
941 Consider a packet in packet like the following::
943 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in
945 * If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
946 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the
947 ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``).
949 * If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
950 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``.
955 Display tx checksum offload configuration::
957 testpmd> csum show (port_id)
962 Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine::
964 testpmd> tso set (segsize) (port_id)
968 Check the NIC datasheet for hardware limits.
973 Display the status of TCP Segmentation Offload::
975 testpmd> tso show (port_id)
980 Enable or disable GRO in ``csum`` forwarding engine::
982 testpmd> set port <port_id> gro on|off
984 If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4
985 packets received from the given port.
987 If disabled, packets received from the given port won't be performed
988 GRO. By default, GRO is disabled for all ports.
992 When enable GRO for a port, TCP/IPv4 packets received from the port
993 will be performed GRO. After GRO, all merged packets have bad
994 checksums, since the GRO library doesn't re-calculate checksums for
995 the merged packets. Therefore, if users want the merged packets to
996 have correct checksums, please select HW IP checksum calculation and
997 HW TCP checksum calculation for the port which the merged packets are
1003 Display GRO configuration for a given port::
1005 testpmd> show port <port_id> gro
1010 Set the cycle to flush the GROed packets from reassembly tables::
1012 testpmd> set gro flush <cycles>
1014 When enable GRO, the csum forwarding engine performs GRO on received
1015 packets, and the GROed packets are stored in reassembly tables. Users
1016 can use this command to determine when the GROed packets are flushed
1017 from the reassembly tables.
1019 The ``cycles`` is measured in GRO operation times. The csum forwarding
1020 engine flushes the GROed packets from the tables every ``cycles`` GRO
1023 By default, the value of ``cycles`` is 1, which means flush GROed packets
1024 from the reassembly tables as soon as one GRO operation finishes. The value
1025 of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``.
1027 Please note that the large value of ``cycles`` may cause the poor TCP/IP
1028 stack performance. Because the GROed packets are delayed to arrive the
1029 stack, thus causing more duplicated ACKs and TCP retransmissions.
1034 Toggle per-port GSO support in ``csum`` forwarding engine::
1036 testpmd> set port <port_id> gso on|off
1038 If enabled, the csum forwarding engine will perform GSO on supported IPv4
1039 packets, transmitted on the given port.
1041 If disabled, packets transmitted on the given port will not undergo GSO.
1042 By default, GSO is disabled for all ports.
1046 When GSO is enabled on a port, supported IPv4 packets transmitted on that
1047 port undergo GSO. Afterwards, the segmented packets are represented by
1048 multi-segment mbufs; however, the csum forwarding engine doesn't calculation
1049 of checksums for GSO'd segments in SW. As a result, if users want correct
1050 checksums in GSO segments, they should enable HW checksum calculation for
1053 For example, HW checksum calculation for VxLAN GSO'd packets may be enabled
1054 by setting the following options in the csum forwarding engine:
1056 testpmd> csum set outer_ip hw <port_id>
1058 testpmd> csum set ip hw <port_id>
1060 testpmd> csum set tcp hw <port_id>
1062 UDP GSO is the same as IP fragmentation, which treats the UDP header
1063 as the payload and does not modify it during segmentation. That is,
1064 after UDP GSO, only the first output fragment has the original UDP
1065 header. Therefore, users need to enable HW IP checksum calculation
1066 and SW UDP checksum calculation for GSO-enabled ports, if they want
1067 correct checksums for UDP/IPv4 packets.
1072 Set the maximum GSO segment size (measured in bytes), which includes the
1073 packet header and the packet payload for GSO-enabled ports (global)::
1075 testpmd> set gso segsz <length>
1080 Display the status of Generic Segmentation Offload for a given port::
1082 testpmd> show port <port_id> gso
1087 Add an alternative MAC address to a port::
1089 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX)
1094 Remove a MAC address from a port::
1096 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX)
1098 mac_addr add (for VF)
1099 ~~~~~~~~~~~~~~~~~~~~~
1101 Add an alternative MAC address for a VF to a port::
1103 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX)
1108 Set the default MAC address for a port::
1110 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX)
1112 mac_addr set (for VF)
1113 ~~~~~~~~~~~~~~~~~~~~~
1115 Set the MAC address for a VF from the PF::
1117 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX)
1122 Set the forwarding peer address for certain port::
1124 testpmd> set eth-peer (port_id) (perr_addr)
1126 This is equivalent to the ``--eth-peer`` command-line option.
1131 Set the unicast hash filter(s) on/off for a port::
1133 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
1138 Set the promiscuous mode on for a port or for all ports.
1139 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1141 testpmd> set promisc (port_id|all) (on|off)
1146 Set the allmulti mode for a port or for all ports::
1148 testpmd> set allmulti (port_id|all) (on|off)
1150 Same as the ifconfig (8) option. Controls how multicast packets are handled.
1152 set promisc (for VF)
1153 ~~~~~~~~~~~~~~~~~~~~
1155 Set the unicast promiscuous mode for a VF from PF.
1156 It's supported by Intel i40e NICs now.
1157 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1159 testpmd> set vf promisc (port_id) (vf_id) (on|off)
1161 set allmulticast (for VF)
1162 ~~~~~~~~~~~~~~~~~~~~~~~~~
1164 Set the multicast promiscuous mode for a VF from PF.
1165 It's supported by Intel i40e NICs now.
1166 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1168 testpmd> set vf allmulti (port_id) (vf_id) (on|off)
1170 set tx max bandwidth (for VF)
1171 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1173 Set TX max absolute bandwidth (Mbps) for a VF from PF::
1175 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth)
1177 set tc tx min bandwidth (for VF)
1178 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1180 Set all TCs' TX min relative bandwidth (%) for a VF from PF::
1182 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...)
1184 set tc tx max bandwidth (for VF)
1185 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1187 Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF::
1189 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth)
1191 set tc strict link priority mode
1192 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1194 Set some TCs' strict link priority mode on a physical port::
1196 testpmd> set tx strict-link-priority (port_id) (tc_bitmap)
1198 set tc tx min bandwidth
1199 ~~~~~~~~~~~~~~~~~~~~~~~
1201 Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs::
1203 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...)
1208 Set the link flow control parameter on a port::
1210 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1211 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
1212 autoneg (on|off) (port_id)
1216 * ``high_water`` (integer): High threshold value to trigger XOFF.
1218 * ``low_water`` (integer): Low threshold value to trigger XON.
1220 * ``pause_time`` (integer): Pause quota in the Pause frame.
1222 * ``send_xon`` (0/1): Send XON frame.
1224 * ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
1226 * ``autoneg``: Change the auto-negotiation parameter.
1231 Set the priority flow control parameter on a port::
1233 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1234 (pause_time) (priority) (port_id)
1238 * ``high_water`` (integer): High threshold value.
1240 * ``low_water`` (integer): Low threshold value.
1242 * ``pause_time`` (integer): Pause quota in the Pause frame.
1244 * ``priority`` (0-7): VLAN User Priority.
1249 Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
1251 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
1253 For example, to set rx queue 2 on port 0 to mapping 5::
1255 testpmd>set stat_qmap rx 0 2 5
1257 set xstats-hide-zero
1258 ~~~~~~~~~~~~~~~~~~~~
1260 Set the option to hide zero values for xstats display::
1262 testpmd> set xstats-hide-zero on|off
1266 By default, the zero values are displayed for xstats.
1268 set port - rx/tx (for VF)
1269 ~~~~~~~~~~~~~~~~~~~~~~~~~
1271 Set VF receive/transmit from a port::
1273 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
1275 set port - mac address filter (for VF)
1276 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1278 Add/Remove unicast or multicast MAC addr filter for a VF::
1280 testpmd> set port (port_id) vf (vf_id) (mac_addr) \
1281 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
1283 set port - rx mode(for VF)
1284 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1286 Set the VF receive mode of a port::
1288 testpmd> set port (port_id) vf (vf_id) \
1289 rxmode (AUPE|ROPE|BAM|MPE) (on|off)
1291 The available receive modes are:
1293 * ``AUPE``: Accepts untagged VLAN.
1295 * ``ROPE``: Accepts unicast hash.
1297 * ``BAM``: Accepts broadcast packets.
1299 * ``MPE``: Accepts all multicast packets.
1301 set port - tx_rate (for Queue)
1302 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1304 Set TX rate limitation for a queue on a port::
1306 testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1308 set port - tx_rate (for VF)
1309 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
1311 Set TX rate limitation for queues in VF on a port::
1313 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1315 set port - mirror rule
1316 ~~~~~~~~~~~~~~~~~~~~~~
1318 Set pool or vlan type mirror rule for a port::
1320 testpmd> set port (port_id) mirror-rule (rule_id) \
1321 (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1322 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1324 Set link mirror rule for a port::
1326 testpmd> set port (port_id) mirror-rule (rule_id) \
1327 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1329 For example to enable mirror traffic with vlan 0,1 to pool 0::
1331 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1333 reset port - mirror rule
1334 ~~~~~~~~~~~~~~~~~~~~~~~~
1336 Reset a mirror rule for a port::
1338 testpmd> reset port (port_id) mirror-rule (rule_id)
1343 Set the flush on RX streams before forwarding.
1344 The default is flush ``on``.
1345 Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1347 testpmd> set flush_rx off
1352 Set the bypass mode for the lowest port on bypass enabled NIC::
1354 testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1359 Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1361 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1362 mode (normal|bypass|isolate) (port_id)
1366 * ``timeout``: Enable bypass after watchdog timeout.
1368 * ``os_on``: Enable bypass when OS/board is powered on.
1370 * ``os_off``: Enable bypass when OS/board is powered off.
1372 * ``power_on``: Enable bypass when power supply is turned on.
1374 * ``power_off``: Enable bypass when power supply is turned off.
1380 Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1382 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1387 Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1389 testpmd> show bypass config (port_id)
1394 Set link up for a port::
1396 testpmd> set link-up port (port id)
1401 Set link down for a port::
1403 testpmd> set link-down port (port id)
1408 Enable E-tag insertion for a VF on a port::
1410 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1412 Disable E-tag insertion for a VF on a port::
1414 testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1416 Enable/disable E-tag stripping on a port::
1418 testpmd> E-tag set stripping (on|off) port (port_id)
1420 Enable/disable E-tag based forwarding on a port::
1422 testpmd> E-tag set forwarding (on|off) port (port_id)
1424 Add an E-tag forwarding filter on a port::
1426 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1428 Delete an E-tag forwarding filter on a port::
1429 testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1434 Load a dynamic device personalization (DDP) profile and store backup profile::
1436 testpmd> ddp add (port_id) (profile_path[,backup_profile_path])
1441 Delete a dynamic device personalization profile and restore backup profile::
1443 testpmd> ddp del (port_id) (backup_profile_path)
1448 List all items from the ptype mapping table::
1450 testpmd> ptype mapping get (port_id) (valid_only)
1454 * ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0).
1456 Replace a specific or a group of software defined ptype with a new one::
1458 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type)
1462 * ``target``: A specific software ptype or a mask to represent a group of software ptypes.
1464 * ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1).
1466 * ``pkt_type``: The new software ptype to replace the old ones.
1468 Update hardware defined ptype to software defined packet type mapping table::
1470 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype)
1474 * ``hw_ptype``: hardware ptype as the index of the ptype mapping table.
1476 * ``sw_ptype``: software ptype as the value of the ptype mapping table.
1478 Reset ptype mapping table::
1480 testpmd> ptype mapping reset (port_id)
1482 config per port Rx offloading
1483 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1485 Enable or disable a per port Rx offloading on all Rx queues of a port::
1487 testpmd> port config (port_id) rx_offload (offloading) on|off
1489 * ``offloading``: can be any of these offloading capability:
1490 vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro,
1491 qinq_strip, outer_ipv4_cksum, macsec_strip,
1492 header_split, vlan_filter, vlan_extend, jumbo_frame,
1493 crc_strip, scatter, timestamp, security
1495 This command should be run when the port is stopped, or else it will fail.
1497 config per queue Rx offloading
1498 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1500 Enable or disable a per queue Rx offloading only on a specific Rx queue::
1502 testpmd> port (port_id) rxq (queue_id) rx_offload (offloading) on|off
1504 * ``offloading``: can be any of these offloading capability:
1505 vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro,
1506 qinq_strip, outer_ipv4_cksum, macsec_strip,
1507 header_split, vlan_filter, vlan_extend, jumbo_frame,
1508 crc_strip, scatter, timestamp, security
1510 This command should be run when the port is stopped, or else it will fail.
1512 config per port Tx offloading
1513 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1515 Enable or disable a per port Tx offloading on all Tx queues of a port::
1517 testpmd> port config (port_id) tx_offload (offloading) on|off
1519 * ``offloading``: can be any of these offloading capability:
1520 vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum,
1521 sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum,
1522 qinq_insert, vxlan_tnl_tso, gre_tnl_tso,
1523 ipip_tnl_tso, geneve_tnl_tso, macsec_insert,
1524 mt_lockfree, multi_segs, fast_free, security
1526 This command should be run when the port is stopped, or else it will fail.
1528 config per queue Tx offloading
1529 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1531 Enable or disable a per queue Tx offloading only on a specific Tx queue::
1533 testpmd> port (port_id) txq (queue_id) tx_offload (offloading) on|off
1535 * ``offloading``: can be any of these offloading capability:
1536 vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum,
1537 sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum,
1538 qinq_insert, vxlan_tnl_tso, gre_tnl_tso,
1539 ipip_tnl_tso, geneve_tnl_tso, macsec_insert,
1540 mt_lockfree, multi_segs, fast_free, security
1542 This command should be run when the port is stopped, or else it will fail.
1544 Config VXLAN Encap outer layers
1545 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1547 Configure the outer layer to encapsulate a packet inside a VXLAN tunnel::
1549 set vxlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \
1550 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) eth-src (eth-src) \
1553 set vxlan-with-vlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \
1554 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) vlan-tci (vlan-tci) \
1555 eth-src (eth-src) eth-dst (eth-dst)
1557 Those command will set an internal configuration inside testpmd, any following
1558 flow rule using the action vxlan_encap will use the last configuration set.
1559 To have a different encapsulation header, one of those commands must be called
1560 before the flow rule creation.
1565 The following sections show functions for configuring ports.
1569 Port configuration changes only become active when forwarding is started/restarted.
1574 Attach a port specified by pci address or virtual device args::
1576 testpmd> port attach (identifier)
1578 To attach a new pci device, the device should be recognized by kernel first.
1579 Then it should be moved under DPDK management.
1580 Finally the port can be attached to testpmd.
1582 For example, to move a pci device using ixgbe under DPDK management:
1584 .. code-block:: console
1586 # Check the status of the available devices.
1587 ./usertools/dpdk-devbind.py --status
1589 Network devices using DPDK-compatible driver
1590 ============================================
1593 Network devices using kernel driver
1594 ===================================
1595 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1598 # Bind the device to igb_uio.
1599 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1602 # Recheck the status of the devices.
1603 ./usertools/dpdk-devbind.py --status
1604 Network devices using DPDK-compatible driver
1605 ============================================
1606 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1608 To attach a port created by virtual device, above steps are not needed.
1610 For example, to attach a port whose pci address is 0000:0a:00.0.
1612 .. code-block:: console
1614 testpmd> port attach 0000:0a:00.0
1615 Attaching a new port...
1616 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1617 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1618 EAL: PCI memory mapped at 0x7f83bfa00000
1619 EAL: PCI memory mapped at 0x7f83bfa80000
1620 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1621 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1622 Port 0 is attached. Now total ports is 1
1625 For example, to attach a port created by pcap PMD.
1627 .. code-block:: console
1629 testpmd> port attach net_pcap0
1630 Attaching a new port...
1631 PMD: Initializing pmd_pcap for net_pcap0
1632 PMD: Creating pcap-backed ethdev on numa socket 0
1633 Port 0 is attached. Now total ports is 1
1636 In this case, identifier is ``net_pcap0``.
1637 This identifier format is the same as ``--vdev`` format of DPDK applications.
1639 For example, to re-attach a bonded port which has been previously detached,
1640 the mode and slave parameters must be given.
1642 .. code-block:: console
1644 testpmd> port attach net_bond_0,mode=0,slave=1
1645 Attaching a new port...
1646 EAL: Initializing pmd_bond for net_bond_0
1647 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1648 Port 0 is attached. Now total ports is 1
1655 Detach a specific port::
1657 testpmd> port detach (port_id)
1659 Before detaching a port, the port should be stopped and closed.
1661 For example, to detach a pci device port 0.
1663 .. code-block:: console
1665 testpmd> port stop 0
1668 testpmd> port close 0
1672 testpmd> port detach 0
1674 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1675 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1676 EAL: PCI memory unmapped at 0x7f83bfa00000
1677 EAL: PCI memory unmapped at 0x7f83bfa80000
1681 For example, to detach a virtual device port 0.
1683 .. code-block:: console
1685 testpmd> port stop 0
1688 testpmd> port close 0
1692 testpmd> port detach 0
1694 PMD: Closing pcap ethdev on numa socket 0
1695 Port 'net_pcap0' is detached. Now total ports is 0
1698 To remove a pci device completely from the system, first detach the port from testpmd.
1699 Then the device should be moved under kernel management.
1700 Finally the device can be removed using kernel pci hotplug functionality.
1702 For example, to move a pci device under kernel management:
1704 .. code-block:: console
1706 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1708 ./usertools/dpdk-devbind.py --status
1710 Network devices using DPDK-compatible driver
1711 ============================================
1714 Network devices using kernel driver
1715 ===================================
1716 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1718 To remove a port created by a virtual device, above steps are not needed.
1723 Start all ports or a specific port::
1725 testpmd> port start (port_id|all)
1730 Stop all ports or a specific port::
1732 testpmd> port stop (port_id|all)
1737 Close all ports or a specific port::
1739 testpmd> port close (port_id|all)
1741 port config - queue ring size
1742 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1744 Configure a rx/tx queue ring size::
1746 testpmd> port (port_id) (rxq|txq) (queue_id) ring_size (value)
1748 Only take effect after command that (re-)start the port or command that setup specific queue.
1750 port start/stop queue
1751 ~~~~~~~~~~~~~~~~~~~~~
1753 Start/stop a rx/tx queue on a specific port::
1755 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1758 ~~~~~~~~~~~~~~~~~~~~~
1760 Setup a rx/tx queue on a specific port::
1762 testpmd> port (port_id) (rxq|txq) (queue_id) setup
1764 Only take effect when port is started.
1769 Set the speed and duplex mode for all ports or a specific port::
1771 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1772 duplex (half|full|auto)
1774 port config - queues/descriptors
1775 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1777 Set number of queues/descriptors for rxq, txq, rxd and txd::
1779 testpmd> port config all (rxq|txq|rxd|txd) (value)
1781 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1783 port config - max-pkt-len
1784 ~~~~~~~~~~~~~~~~~~~~~~~~~
1786 Set the maximum packet length::
1788 testpmd> port config all max-pkt-len (value)
1790 This is equivalent to the ``--max-pkt-len`` command-line option.
1792 port config - CRC Strip
1793 ~~~~~~~~~~~~~~~~~~~~~~~
1795 Set hardware CRC stripping on or off for all ports::
1797 testpmd> port config all crc-strip (on|off)
1799 CRC stripping is on by default.
1801 The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option.
1803 port config - scatter
1804 ~~~~~~~~~~~~~~~~~~~~~~~
1806 Set RX scatter mode on or off for all ports::
1808 testpmd> port config all scatter (on|off)
1810 RX scatter mode is off by default.
1812 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1814 port config - RX Checksum
1815 ~~~~~~~~~~~~~~~~~~~~~~~~~
1817 Set hardware RX checksum offload to on or off for all ports::
1819 testpmd> port config all rx-cksum (on|off)
1821 Checksum offload is off by default.
1823 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1828 Set hardware VLAN on or off for all ports::
1830 testpmd> port config all hw-vlan (on|off)
1832 Hardware VLAN is off by default.
1834 The ``on`` option is equivalent to the ``--enable-hw-vlan`` command-line option.
1836 port config - VLAN filter
1837 ~~~~~~~~~~~~~~~~~~~~~~~~~
1839 Set hardware VLAN filter on or off for all ports::
1841 testpmd> port config all hw-vlan-filter (on|off)
1843 Hardware VLAN filter is off by default.
1845 The ``on`` option is equivalent to the ``--enable-hw-vlan-filter`` command-line option.
1847 port config - VLAN strip
1848 ~~~~~~~~~~~~~~~~~~~~~~~~
1850 Set hardware VLAN strip on or off for all ports::
1852 testpmd> port config all hw-vlan-strip (on|off)
1854 Hardware VLAN strip is off by default.
1856 The ``on`` option is equivalent to the ``--enable-hw-vlan-strip`` command-line option.
1858 port config - VLAN extend
1859 ~~~~~~~~~~~~~~~~~~~~~~~~~
1861 Set hardware VLAN extend on or off for all ports::
1863 testpmd> port config all hw-vlan-extend (on|off)
1865 Hardware VLAN extend is off by default.
1867 The ``on`` option is equivalent to the ``--enable-hw-vlan-extend`` command-line option.
1869 port config - Drop Packets
1870 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1872 Set packet drop for packets with no descriptors on or off for all ports::
1874 testpmd> port config all drop-en (on|off)
1876 Packet dropping for packets with no descriptors is off by default.
1878 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1883 Set the RSS (Receive Side Scaling) mode on or off::
1885 testpmd> port config all rss (all|default|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1887 RSS is on by default.
1889 The ``all`` option is equivalent to ip|tcp|udp|sctp|ether.
1890 The ``default`` option enables all supported RSS types reported by device info.
1891 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1893 port config - RSS Reta
1894 ~~~~~~~~~~~~~~~~~~~~~~
1896 Set the RSS (Receive Side Scaling) redirection table::
1898 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1903 Set the DCB mode for an individual port::
1905 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1907 The traffic class should be 4 or 8.
1912 Set the number of packets per burst::
1914 testpmd> port config all burst (value)
1916 This is equivalent to the ``--burst`` command-line option.
1918 port config - Threshold
1919 ~~~~~~~~~~~~~~~~~~~~~~~
1921 Set thresholds for TX/RX queues::
1923 testpmd> port config all (threshold) (value)
1925 Where the threshold type can be:
1927 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1929 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1931 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1933 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1935 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1937 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1939 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1941 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1943 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1945 These threshold options are also available from the command-line.
1950 Set the value of ether-type for E-tag::
1952 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1954 Enable/disable the E-tag support::
1956 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1958 port config pctype mapping
1959 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1961 Reset pctype mapping table::
1963 testpmd> port config (port_id) pctype mapping reset
1965 Update hardware defined pctype to software defined flow type mapping table::
1967 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id)
1971 * ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table.
1973 * ``flow_type_id``: software flow type id as the index of the pctype mapping table.
1975 port config input set
1976 ~~~~~~~~~~~~~~~~~~~~~
1978 Config RSS/FDIR/FDIR flexible payload input set for some pctype::
1979 testpmd> port config (port_id) pctype (pctype_id) \
1980 (hash_inset|fdir_inset|fdir_flx_inset) \
1981 (get|set|clear) field (field_idx)
1983 Clear RSS/FDIR/FDIR flexible payload input set for some pctype::
1984 testpmd> port config (port_id) pctype (pctype_id) \
1985 (hash_inset|fdir_inset|fdir_flx_inset) clear all
1989 * ``pctype_id``: hardware packet classification types.
1990 * ``field_idx``: hardware field index.
1992 port config udp_tunnel_port
1993 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
1995 Add/remove UDP tunnel port for VXLAN/GENEVE tunneling protocols::
1996 testpmd> port config (port_id) udp_tunnel_port add|rm vxlan|geneve (udp_port)
1998 Link Bonding Functions
1999 ----------------------
2001 The Link Bonding functions make it possible to dynamically create and
2002 manage link bonding devices from within testpmd interactive prompt.
2004 create bonded device
2005 ~~~~~~~~~~~~~~~~~~~~
2007 Create a new bonding device::
2009 testpmd> create bonded device (mode) (socket)
2011 For example, to create a bonded device in mode 1 on socket 0::
2013 testpmd> create bonded 1 0
2014 created new bonded device (port X)
2019 Adds Ethernet device to a Link Bonding device::
2021 testpmd> add bonding slave (slave id) (port id)
2023 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
2025 testpmd> add bonding slave 6 10
2028 remove bonding slave
2029 ~~~~~~~~~~~~~~~~~~~~
2031 Removes an Ethernet slave device from a Link Bonding device::
2033 testpmd> remove bonding slave (slave id) (port id)
2035 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
2037 testpmd> remove bonding slave 6 10
2042 Set the Link Bonding mode of a Link Bonding device::
2044 testpmd> set bonding mode (value) (port id)
2046 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
2048 testpmd> set bonding mode 3 10
2053 Set an Ethernet slave device as the primary device on a Link Bonding device::
2055 testpmd> set bonding primary (slave id) (port id)
2057 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
2059 testpmd> set bonding primary 6 10
2064 Set the MAC address of a Link Bonding device::
2066 testpmd> set bonding mac (port id) (mac)
2068 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
2070 testpmd> set bonding mac 10 00:00:00:00:00:01
2072 set bonding xmit_balance_policy
2073 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2075 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
2077 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
2079 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
2081 testpmd> set bonding xmit_balance_policy 10 l34
2084 set bonding mon_period
2085 ~~~~~~~~~~~~~~~~~~~~~~
2087 Set the link status monitoring polling period in milliseconds for a bonding device.
2089 This adds support for PMD slave devices which do not support link status interrupts.
2090 When the mon_period is set to a value greater than 0 then all PMD's which do not support
2091 link status ISR will be queried every polling interval to check if their link status has changed::
2093 testpmd> set bonding mon_period (port_id) (value)
2095 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
2097 testpmd> set bonding mon_period 5 150
2100 set bonding lacp dedicated_queue
2101 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2103 Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic
2104 when in mode 4 (link-aggregration-802.3ad)::
2106 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable)
2109 set bonding agg_mode
2110 ~~~~~~~~~~~~~~~~~~~~
2112 Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad)::
2114 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable)
2120 Show the current configuration of a Link Bonding device::
2122 testpmd> show bonding config (port id)
2125 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
2126 in balance mode with a transmission policy of layer 2+3::
2128 testpmd> show bonding config 9
2130 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
2132 Active Slaves (3): [1 3 4]
2139 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
2140 This is mainly useful for debugging purposes.
2141 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
2142 and fields that can be accessed.
2147 Display the value of a port register::
2149 testpmd> read reg (port_id) (address)
2151 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
2153 testpmd> read reg 0 0xEE00
2154 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
2159 Display a port register bit field::
2161 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
2163 For example, reading the lowest two bits from the register in the example above::
2165 testpmd> read regfield 0 0xEE00 0 1
2166 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
2171 Display a single port register bit::
2173 testpmd> read regbit (port_id) (address) (bit_x)
2175 For example, reading the lowest bit from the register in the example above::
2177 testpmd> read regbit 0 0xEE00 0
2178 port 0 PCI register at offset 0xEE00: bit 0=1
2183 Set the value of a port register::
2185 testpmd> write reg (port_id) (address) (value)
2187 For example, to clear a register::
2189 testpmd> write reg 0 0xEE00 0x0
2190 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
2195 Set bit field of a port register::
2197 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
2199 For example, writing to the register cleared in the example above::
2201 testpmd> write regfield 0 0xEE00 0 1 2
2202 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
2207 Set single bit value of a port register::
2209 testpmd> write regbit (port_id) (address) (bit_x) (value)
2211 For example, to set the high bit in the register from the example above::
2213 testpmd> write regbit 0 0xEE00 31 1
2214 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
2216 Traffic Metering and Policing
2217 -----------------------------
2219 The following section shows functions for configuring traffic metering and
2220 policing on the ethernet device through the use of generic ethdev API.
2222 show port traffic management capability
2223 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2225 Show traffic metering and policing capability of the port::
2227 testpmd> show port meter cap (port_id)
2229 add port meter profile (srTCM rfc2967)
2230 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2232 Add meter profile (srTCM rfc2697) to the ethernet device::
2234 testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \
2239 * ``profile_id``: ID for the meter profile.
2240 * ``cir``: Committed Information Rate (CIR) (bytes/second).
2241 * ``cbs``: Committed Burst Size (CBS) (bytes).
2242 * ``ebs``: Excess Burst Size (EBS) (bytes).
2244 add port meter profile (trTCM rfc2968)
2245 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2247 Add meter profile (srTCM rfc2698) to the ethernet device::
2249 testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \
2250 (cir) (pir) (cbs) (pbs)
2254 * ``profile_id``: ID for the meter profile.
2255 * ``cir``: Committed information rate (bytes/second).
2256 * ``pir``: Peak information rate (bytes/second).
2257 * ``cbs``: Committed burst size (bytes).
2258 * ``pbs``: Peak burst size (bytes).
2260 add port meter profile (trTCM rfc4115)
2261 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2263 Add meter profile (trTCM rfc4115) to the ethernet device::
2265 testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \
2266 (cir) (eir) (cbs) (ebs)
2270 * ``profile_id``: ID for the meter profile.
2271 * ``cir``: Committed information rate (bytes/second).
2272 * ``eir``: Excess information rate (bytes/second).
2273 * ``cbs``: Committed burst size (bytes).
2274 * ``ebs``: Excess burst size (bytes).
2276 delete port meter profile
2277 ~~~~~~~~~~~~~~~~~~~~~~~~~
2279 Delete meter profile from the ethernet device::
2281 testpmd> del port meter profile (port_id) (profile_id)
2286 Create new meter object for the ethernet device::
2288 testpmd> create port meter (port_id) (mtr_id) (profile_id) \
2289 (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \
2290 (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\
2295 * ``mtr_id``: meter object ID.
2296 * ``profile_id``: ID for the meter profile.
2297 * ``meter_enable``: When this parameter has a non-zero value, the meter object
2298 gets enabled at the time of creation, otherwise remains disabled.
2299 * ``g_action``: Policer action for the packet with green color.
2300 * ``y_action``: Policer action for the packet with yellow color.
2301 * ``r_action``: Policer action for the packet with red color.
2302 * ``stats_mask``: Mask of statistics counter types to be enabled for the
2304 * ``shared``: When this parameter has a non-zero value, the meter object is
2305 shared by multiple flows. Otherwise, meter object is used by single flow.
2306 * ``use_pre_meter_color``: When this parameter has a non-zero value, the
2307 input color for the current meter object is determined by the latest meter
2308 object in the same flow. Otherwise, the current meter object uses the
2309 *dscp_table* to determine the input color.
2310 * ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input
2311 color, 0 <= x <= 63.
2316 Enable meter for the ethernet device::
2318 testpmd> enable port meter (port_id) (mtr_id)
2323 Disable meter for the ethernet device::
2325 testpmd> disable port meter (port_id) (mtr_id)
2330 Delete meter for the ethernet device::
2332 testpmd> del port meter (port_id) (mtr_id)
2334 Set port meter profile
2335 ~~~~~~~~~~~~~~~~~~~~~~
2337 Set meter profile for the ethernet device::
2339 testpmd> set port meter profile (port_id) (mtr_id) (profile_id)
2341 set port meter dscp table
2342 ~~~~~~~~~~~~~~~~~~~~~~~~~
2344 Set meter dscp table for the ethernet device::
2346 testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \
2347 (dscp_tbl_entry1)...(dscp_tbl_entry63)]
2349 set port meter policer action
2350 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2352 Set meter policer action for the ethernet device::
2354 testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \
2355 (action0) [(action1) (action1)]
2359 * ``action_mask``: Bit mask indicating which policer actions need to be
2360 updated. One or more policer actions can be updated in a single function
2361 invocation. To update the policer action associated with color C, bit
2362 (1 << C) needs to be set in *action_mask* and element at position C
2363 in the *actions* array needs to be valid.
2364 * ``actionx``: Policer action for the color x,
2365 RTE_MTR_GREEN <= x < RTE_MTR_COLORS
2367 set port meter stats mask
2368 ~~~~~~~~~~~~~~~~~~~~~~~~~
2370 Set meter stats mask for the ethernet device::
2372 testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask)
2376 * ``stats_mask``: Bit mask indicating statistics counter types to be enabled.
2378 show port meter stats
2379 ~~~~~~~~~~~~~~~~~~~~~
2381 Show meter stats of the ethernet device::
2383 testpmd> show port meter stats (port_id) (mtr_id) (clear)
2387 * ``clear``: Flag that indicates whether the statistics counters should
2388 be cleared (i.e. set to zero) immediately after they have been read or not.
2393 The following section shows functions for configuring traffic management on
2394 on the ethernet device through the use of generic TM API.
2396 show port traffic management capability
2397 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2399 Show traffic management capability of the port::
2401 testpmd> show port tm cap (port_id)
2403 show port traffic management capability (hierarchy level)
2404 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2406 Show traffic management hierarchy level capability of the port::
2408 testpmd> show port tm level cap (port_id) (level_id)
2410 show port traffic management capability (hierarchy node level)
2411 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2413 Show the traffic management hierarchy node capability of the port::
2415 testpmd> show port tm node cap (port_id) (node_id)
2417 show port traffic management hierarchy node type
2418 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2420 Show the port traffic management hierarchy node type::
2422 testpmd> show port tm node type (port_id) (node_id)
2424 show port traffic management hierarchy node stats
2425 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2427 Show the port traffic management hierarchy node statistics::
2429 testpmd> show port tm node stats (port_id) (node_id) (clear)
2433 * ``clear``: When this parameter has a non-zero value, the statistics counters
2434 are cleared (i.e. set to zero) immediately after they have been read,
2435 otherwise the statistics counters are left untouched.
2437 Add port traffic management private shaper profile
2438 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2440 Add the port traffic management private shaper profile::
2442 testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \
2443 (tb_rate) (tb_size) (packet_length_adjust)
2447 * ``shaper_profile id``: Shaper profile ID for the new profile.
2448 * ``tb_rate``: Token bucket rate (bytes per second).
2449 * ``tb_size``: Token bucket size (bytes).
2450 * ``packet_length_adjust``: The value (bytes) to be added to the length of
2451 each packet for the purpose of shaping. This parameter value can be used to
2452 correct the packet length with the framing overhead bytes that are consumed
2455 Delete port traffic management private shaper profile
2456 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2458 Delete the port traffic management private shaper::
2460 testpmd> del port tm node shaper profile (port_id) (shaper_profile_id)
2464 * ``shaper_profile id``: Shaper profile ID that needs to be deleted.
2466 Add port traffic management shared shaper
2467 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2469 Create the port traffic management shared shaper::
2471 testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \
2476 * ``shared_shaper_id``: Shared shaper ID to be created.
2477 * ``shaper_profile id``: Shaper profile ID for shared shaper.
2479 Set port traffic management shared shaper
2480 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2482 Update the port traffic management shared shaper::
2484 testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \
2489 * ``shared_shaper_id``: Shared shaper ID to be update.
2490 * ``shaper_profile id``: Shaper profile ID for shared shaper.
2492 Delete port traffic management shared shaper
2493 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2495 Delete the port traffic management shared shaper::
2497 testpmd> del port tm node shared shaper (port_id) (shared_shaper_id)
2501 * ``shared_shaper_id``: Shared shaper ID to be deleted.
2503 Set port traffic management hiearchy node private shaper
2504 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2506 set the port traffic management hierarchy node private shaper::
2508 testpmd> set port tm node shaper profile (port_id) (node_id) \
2513 * ``shaper_profile id``: Private shaper profile ID to be enabled on the
2516 Add port traffic management WRED profile
2517 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2519 Create a new WRED profile::
2521 testpmd> add port tm node wred profile (port_id) (wred_profile_id) \
2522 (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \
2523 (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \
2524 (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r)
2528 * ``wred_profile id``: Identifier for the newly create WRED profile
2529 * ``color_g``: Packet color (green)
2530 * ``min_th_g``: Minimum queue threshold for packet with green color
2531 * ``max_th_g``: Minimum queue threshold for packet with green color
2532 * ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp)
2533 * ``wq_log2_g``: Negated log2 of queue weight (wq)
2534 * ``color_y``: Packet color (yellow)
2535 * ``min_th_y``: Minimum queue threshold for packet with yellow color
2536 * ``max_th_y``: Minimum queue threshold for packet with yellow color
2537 * ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp)
2538 * ``wq_log2_y``: Negated log2 of queue weight (wq)
2539 * ``color_r``: Packet color (red)
2540 * ``min_th_r``: Minimum queue threshold for packet with yellow color
2541 * ``max_th_r``: Minimum queue threshold for packet with yellow color
2542 * ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp)
2543 * ``wq_log2_r``: Negated log2 of queue weight (wq)
2545 Delete port traffic management WRED profile
2546 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2548 Delete the WRED profile::
2550 testpmd> del port tm node wred profile (port_id) (wred_profile_id)
2552 Add port traffic management hierarchy nonleaf node
2553 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2555 Add nonleaf node to port traffic management hiearchy::
2557 testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \
2558 (priority) (weight) (level_id) (shaper_profile_id) \
2559 (n_sp_priorities) (stats_mask) (n_shared_shapers) \
2560 [(shared_shaper_0) (shared_shaper_1) ...] \
2564 * ``parent_node_id``: Node ID of the parent.
2565 * ``priority``: Node priority (highest node priority is zero). This is used by
2566 the SP algorithm running on the parent node for scheduling this node.
2567 * ``weight``: Node weight (lowest weight is one). The node weight is relative
2568 to the weight sum of all siblings that have the same priority. It is used by
2569 the WFQ algorithm running on the parent node for scheduling this node.
2570 * ``level_id``: Hiearchy level of the node.
2571 * ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2573 * ``n_sp_priorities``: Number of strict priorities.
2574 * ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2575 * ``n_shared_shapers``: Number of shared shapers.
2576 * ``shared_shaper_id``: Shared shaper id.
2578 Add port traffic management hierarchy leaf node
2579 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2581 Add leaf node to port traffic management hiearchy::
2583 testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \
2584 (priority) (weight) (level_id) (shaper_profile_id) \
2585 (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \
2586 [(shared_shaper_id) (shared_shaper_id) ...] \
2590 * ``parent_node_id``: Node ID of the parent.
2591 * ``priority``: Node priority (highest node priority is zero). This is used by
2592 the SP algorithm running on the parent node for scheduling this node.
2593 * ``weight``: Node weight (lowest weight is one). The node weight is relative
2594 to the weight sum of all siblings that have the same priority. It is used by
2595 the WFQ algorithm running on the parent node for scheduling this node.
2596 * ``level_id``: Hiearchy level of the node.
2597 * ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2599 * ``cman_mode``: Congestion management mode to be enabled for this node.
2600 * ``wred_profile_id``: WRED profile id to be enabled for this node.
2601 * ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2602 * ``n_shared_shapers``: Number of shared shapers.
2603 * ``shared_shaper_id``: Shared shaper id.
2605 Delete port traffic management hierarchy node
2606 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2608 Delete node from port traffic management hiearchy::
2610 testpmd> del port tm node (port_id) (node_id)
2612 Update port traffic management hierarchy parent node
2613 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2615 Update port traffic management hierarchy parent node::
2617 testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \
2620 This function can only be called after the hierarchy commit invocation. Its
2621 success depends on the port support for this operation, as advertised through
2622 the port capability set. This function is valid for all nodes of the traffic
2623 management hierarchy except root node.
2625 Suspend port traffic management hierarchy node
2626 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2628 testpmd> suspend port tm node (port_id) (node_id)
2630 Resume port traffic management hierarchy node
2631 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2633 testpmd> resume port tm node (port_id) (node_id)
2635 Commit port traffic management hierarchy
2636 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2638 Commit the traffic management hierarchy on the port::
2640 testpmd> port tm hierarchy commit (port_id) (clean_on_fail)
2644 * ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function
2645 call failure. On the other hand, hierarchy is preserved when this parameter
2648 Set port traffic management default hierarchy (tm forwarding mode)
2649 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2651 set the traffic management default hierarchy on the port::
2653 testpmd> set port tm hierarchy default (port_id)
2658 This section details the available filter functions that are available.
2660 Note these functions interface the deprecated legacy filtering framework,
2661 superseded by *rte_flow*. See `Flow rules management`_.
2664 ~~~~~~~~~~~~~~~~~~~~
2666 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
2668 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
2669 ethertype (ether_type) (drop|fwd) queue (queue_id)
2671 The available information parameters are:
2673 * ``port_id``: The port which the Ethertype filter assigned on.
2675 * ``mac_addr``: Compare destination mac address.
2677 * ``mac_ignr``: Ignore destination mac address match.
2679 * ``mac_address``: Destination mac address to match.
2681 * ``ether_type``: The EtherType value want to match,
2682 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
2684 * ``queue_id``: The receive queue associated with this EtherType filter.
2685 It is meaningless when deleting or dropping.
2687 Example, to add/remove an ethertype filter rule::
2689 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
2690 ethertype 0x0806 fwd queue 3
2692 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
2693 ethertype 0x0806 fwd queue 3
2698 Add or delete a 2-tuple filter,
2699 which identifies packets by specific protocol and destination TCP/UDP port
2700 and forwards packets into one of the receive queues::
2702 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
2703 protocol (protocol_value) mask (mask_value) \
2704 tcp_flags (tcp_flags_value) priority (prio_value) \
2707 The available information parameters are:
2709 * ``port_id``: The port which the 2-tuple filter assigned on.
2711 * ``dst_port_value``: Destination port in L4.
2713 * ``protocol_value``: IP L4 protocol.
2715 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
2717 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
2719 * ``prio_value``: Priority of this filter.
2721 * ``queue_id``: The receive queue associated with this 2-tuple filter.
2723 Example, to add/remove an 2tuple filter rule::
2725 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
2726 tcp_flags 0x02 priority 3 queue 3
2728 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
2729 tcp_flags 0x02 priority 3 queue 3
2734 Add or delete a 5-tuple filter,
2735 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
2736 and routes packets into one of the receive queues::
2738 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
2739 (src_address) dst_port (dst_port_value) \
2740 src_port (src_port_value) protocol (protocol_value) \
2741 mask (mask_value) tcp_flags (tcp_flags_value) \
2742 priority (prio_value) queue (queue_id)
2744 The available information parameters are:
2746 * ``port_id``: The port which the 5-tuple filter assigned on.
2748 * ``dst_address``: Destination IP address.
2750 * ``src_address``: Source IP address.
2752 * ``dst_port_value``: TCP/UDP destination port.
2754 * ``src_port_value``: TCP/UDP source port.
2756 * ``protocol_value``: L4 protocol.
2758 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
2760 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
2762 * ``prio_value``: The priority of this filter.
2764 * ``queue_id``: The receive queue associated with this 5-tuple filter.
2766 Example, to add/remove an 5tuple filter rule::
2768 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2769 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2770 flags 0x0 priority 3 queue 3
2772 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2773 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2774 flags 0x0 priority 3 queue 3
2779 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
2781 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
2783 The available information parameters are:
2785 * ``port_id``: The port which the SYN filter assigned on.
2787 * ``high``: This SYN filter has higher priority than other filters.
2789 * ``low``: This SYN filter has lower priority than other filters.
2791 * ``queue_id``: The receive queue associated with this SYN filter
2795 testpmd> syn_filter 0 add priority high queue 3
2800 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
2801 and routed into one of the receive queues::
2803 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
2804 mask (mask_value) priority (prio_value) queue (queue_id)
2806 The available information parameters are:
2808 * ``port_id``: The port which the Flex filter is assigned on.
2810 * ``len_value``: Filter length in bytes, no greater than 128.
2812 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
2814 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
2816 * ``prio_value``: The priority of this filter.
2818 * ``queue_id``: The receive queue associated with this Flex filter.
2822 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
2823 mask 000C priority 3 queue 3
2825 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
2826 mask 000C priority 3 queue 3
2829 .. _testpmd_flow_director:
2831 flow_director_filter
2832 ~~~~~~~~~~~~~~~~~~~~
2834 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
2836 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
2837 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
2839 * Perfect match filters.
2840 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2841 The masked fields are for IP flow.
2843 * Signature filters.
2844 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
2846 * Perfect-mac-vlan match filters.
2847 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2848 The masked fields are for MAC VLAN flow.
2850 * Perfect-tunnel match filters.
2851 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2852 The masked fields are for tunnel flow.
2854 * Perfect-raw-flow-type match filters.
2855 The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet.
2856 The masked fields are specified by input sets.
2858 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
2859 per flow type and the flexible payload.
2861 The Flow Director can also mask out parts of all of these fields so that filters
2862 are only applied to certain fields or parts of the fields.
2864 Note that for raw flow type mode the source and destination fields in the
2865 raw packet buffer need to be presented in a reversed order with respect
2866 to the expected received packets.
2867 For example: IP source and destination addresses or TCP/UDP/SCTP
2868 source and destination ports
2870 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
2872 # Commands to add flow director filters of different flow types::
2874 flow_director_filter (port_id) mode IP (add|del|update) \
2875 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
2876 src (src_ip_address) dst (dst_ip_address) \
2877 tos (tos_value) proto (proto_value) ttl (ttl_value) \
2878 vlan (vlan_value) flexbytes (flexbytes_value) \
2879 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
2882 flow_director_filter (port_id) mode IP (add|del|update) \
2883 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
2884 src (src_ip_address) (src_port) \
2885 dst (dst_ip_address) (dst_port) \
2886 tos (tos_value) ttl (ttl_value) \
2887 vlan (vlan_value) flexbytes (flexbytes_value) \
2888 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
2891 flow_director_filter (port_id) mode IP (add|del|update) \
2892 flow (ipv4-sctp|ipv6-sctp) \
2893 src (src_ip_address) (src_port) \
2894 dst (dst_ip_address) (dst_port) \
2895 tos (tos_value) ttl (ttl_value) \
2896 tag (verification_tag) vlan (vlan_value) \
2897 flexbytes (flexbytes_value) (drop|fwd) \
2898 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
2900 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
2901 ether (ethertype) flexbytes (flexbytes_value) \
2902 (drop|fwd) pf|vf(vf_id) queue (queue_id)
2905 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
2906 mac (mac_address) vlan (vlan_value) \
2907 flexbytes (flexbytes_value) (drop|fwd) \
2908 queue (queue_id) fd_id (fd_id_value)
2910 flow_director_filter (port_id) mode Tunnel (add|del|update) \
2911 mac (mac_address) vlan (vlan_value) \
2912 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
2913 flexbytes (flexbytes_value) (drop|fwd) \
2914 queue (queue_id) fd_id (fd_id_value)
2916 flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \
2917 (drop|fwd) queue (queue_id) fd_id (fd_id_value) \
2918 packet (packet file name)
2920 For example, to add an ipv4-udp flow type filter::
2922 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
2923 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
2924 fwd pf queue 1 fd_id 1
2926 For example, add an ipv4-other flow type filter::
2928 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
2929 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
2930 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
2935 Flush all flow director filters on a device::
2937 testpmd> flush_flow_director (port_id)
2939 Example, to flush all flow director filter on port 0::
2941 testpmd> flush_flow_director 0
2946 Set flow director's input masks::
2948 flow_director_mask (port_id) mode IP vlan (vlan_value) \
2949 src_mask (ipv4_src) (ipv6_src) (src_port) \
2950 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2952 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2954 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2955 mac (mac_value) tunnel-type (tunnel_type_value) \
2956 tunnel-id (tunnel_id_value)
2958 Example, to set flow director mask on port 0::
2960 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2961 src_mask 255.255.255.255 \
2962 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2963 dst_mask 255.255.255.255 \
2964 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2966 flow_director_flex_mask
2967 ~~~~~~~~~~~~~~~~~~~~~~~
2969 set masks of flow director's flexible payload based on certain flow type::
2971 testpmd> flow_director_flex_mask (port_id) \
2972 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2973 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2974 l2_payload|all) (mask)
2976 Example, to set flow director's flex mask for all flow type on port 0::
2978 testpmd> flow_director_flex_mask 0 flow all \
2979 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2982 flow_director_flex_payload
2983 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2985 Configure flexible payload selection::
2987 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2989 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2991 testpmd> flow_director_flex_payload 0 l4 \
2992 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2994 get_sym_hash_ena_per_port
2995 ~~~~~~~~~~~~~~~~~~~~~~~~~
2997 Get symmetric hash enable configuration per port::
2999 get_sym_hash_ena_per_port (port_id)
3001 For example, to get symmetric hash enable configuration of port 1::
3003 testpmd> get_sym_hash_ena_per_port 1
3005 set_sym_hash_ena_per_port
3006 ~~~~~~~~~~~~~~~~~~~~~~~~~
3008 Set symmetric hash enable configuration per port to enable or disable::
3010 set_sym_hash_ena_per_port (port_id) (enable|disable)
3012 For example, to set symmetric hash enable configuration of port 1 to enable::
3014 testpmd> set_sym_hash_ena_per_port 1 enable
3016 get_hash_global_config
3017 ~~~~~~~~~~~~~~~~~~~~~~
3019 Get the global configurations of hash filters::
3021 get_hash_global_config (port_id)
3023 For example, to get the global configurations of hash filters of port 1::
3025 testpmd> get_hash_global_config 1
3027 set_hash_global_config
3028 ~~~~~~~~~~~~~~~~~~~~~~
3030 Set the global configurations of hash filters::
3032 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
3033 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
3034 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \
3037 For example, to enable simple_xor for flow type of ipv6 on port 2::
3039 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
3044 Set the input set for hash::
3046 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
3047 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
3048 l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
3049 ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
3050 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
3051 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
3052 fld-8th|none) (select|add)
3054 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
3056 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
3061 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
3062 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
3064 Set the input set for flow director::
3066 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
3067 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
3068 l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
3069 ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
3070 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
3071 sctp-dst-port|sctp-veri-tag|none) (select|add)
3073 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
3075 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
3080 Set different GRE key length for input set::
3082 global_config (port_id) gre-key-len (number in bytes)
3084 For example to set GRE key length for input set to 4 bytes on port 0::
3086 testpmd> global_config 0 gre-key-len 4
3089 .. _testpmd_rte_flow:
3091 Flow rules management
3092 ---------------------
3094 Control of the generic flow API (*rte_flow*) is fully exposed through the
3095 ``flow`` command (validation, creation, destruction, queries and operation
3098 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
3099 features simultaneously may cause undefined side-effects and is therefore
3105 Because the ``flow`` command uses dynamic tokens to handle the large number
3106 of possible flow rules combinations, its behavior differs slightly from
3107 other commands, in particular:
3109 - Pressing *?* or the *<tab>* key displays contextual help for the current
3110 token, not that of the entire command.
3112 - Optional and repeated parameters are supported (provided they are listed
3113 in the contextual help).
3115 The first parameter stands for the operation mode. Possible operations and
3116 their general syntax are described below. They are covered in detail in the
3119 - Check whether a flow rule can be created::
3121 flow validate {port_id}
3122 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3123 pattern {item} [/ {item} [...]] / end
3124 actions {action} [/ {action} [...]] / end
3126 - Create a flow rule::
3128 flow create {port_id}
3129 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3130 pattern {item} [/ {item} [...]] / end
3131 actions {action} [/ {action} [...]] / end
3133 - Destroy specific flow rules::
3135 flow destroy {port_id} rule {rule_id} [...]
3137 - Destroy all flow rules::
3139 flow flush {port_id}
3141 - Query an existing flow rule::
3143 flow query {port_id} {rule_id} {action}
3145 - List existing flow rules sorted by priority, filtered by group
3148 flow list {port_id} [group {group_id}] [...]
3150 - Restrict ingress traffic to the defined flow rules::
3152 flow isolate {port_id} {boolean}
3154 Validating flow rules
3155 ~~~~~~~~~~~~~~~~~~~~~
3157 ``flow validate`` reports whether a flow rule would be accepted by the
3158 underlying device in its current state but stops short of creating it. It is
3159 bound to ``rte_flow_validate()``::
3161 flow validate {port_id}
3162 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3163 pattern {item} [/ {item} [...]] / end
3164 actions {action} [/ {action} [...]] / end
3166 If successful, it will show::
3170 Otherwise it will show an error message of the form::
3172 Caught error type [...] ([...]): [...]
3174 This command uses the same parameters as ``flow create``, their format is
3175 described in `Creating flow rules`_.
3177 Check whether redirecting any Ethernet packet received on port 0 to RX queue
3178 index 6 is supported::
3180 testpmd> flow validate 0 ingress pattern eth / end
3181 actions queue index 6 / end
3185 Port 0 does not support TCPv6 rules::
3187 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
3189 Caught error type 9 (specific pattern item): Invalid argument
3195 ``flow create`` validates and creates the specified flow rule. It is bound
3196 to ``rte_flow_create()``::
3198 flow create {port_id}
3199 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3200 pattern {item} [/ {item} [...]] / end
3201 actions {action} [/ {action} [...]] / end
3203 If successful, it will return a flow rule ID usable with other commands::
3205 Flow rule #[...] created
3207 Otherwise it will show an error message of the form::
3209 Caught error type [...] ([...]): [...]
3211 Parameters describe in the following order:
3213 - Attributes (*group*, *priority*, *ingress*, *egress*, *transfer* tokens).
3214 - A matching pattern, starting with the *pattern* token and terminated by an
3216 - Actions, starting with the *actions* token and terminated by an *end*
3219 These translate directly to *rte_flow* objects provided as-is to the
3220 underlying functions.
3222 The shortest valid definition only comprises mandatory tokens::
3224 testpmd> flow create 0 pattern end actions end
3226 Note that PMDs may refuse rules that essentially do nothing such as this
3229 **All unspecified object values are automatically initialized to 0.**
3234 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
3235 specified before the ``pattern`` token.
3237 - ``group {group id}``: priority group.
3238 - ``priority {level}``: priority level within group.
3239 - ``ingress``: rule applies to ingress traffic.
3240 - ``egress``: rule applies to egress traffic.
3241 - ``transfer``: apply rule directly to endpoints found in pattern.
3243 Each instance of an attribute specified several times overrides the previous
3244 value as shown below (group 4 is used)::
3246 testpmd> flow create 0 group 42 group 24 group 4 [...]
3248 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
3250 While not specifying a direction is an error, some rules may allow both
3253 Most rules affect RX therefore contain the ``ingress`` token::
3255 testpmd> flow create 0 ingress pattern [...]
3260 A matching pattern starts after the ``pattern`` token. It is made of pattern
3261 items and is terminated by a mandatory ``end`` item.
3263 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
3264 rte_flow_item_type``).
3266 The ``/`` token is used as a separator between pattern items as shown
3269 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
3271 Note that protocol items like these must be stacked from lowest to highest
3272 layer to make sense. For instance, the following rule is either invalid or
3273 unlikely to match any packet::
3275 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
3277 More information on these restrictions can be found in the *rte_flow*
3280 Several items support additional specification structures, for example
3281 ``ipv4`` allows specifying source and destination addresses as follows::
3283 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
3284 dst is 10.2.0.0 / end [...]
3286 This rule matches all IPv4 traffic with the specified properties.
3288 In this example, ``src`` and ``dst`` are field names of the underlying
3289 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
3290 in a similar fashion.
3292 The ``is`` token means that the subsequent value must be matched exactly,
3293 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
3294 accordingly. Possible assignment tokens are:
3296 - ``is``: match value perfectly (with full bit-mask).
3297 - ``spec``: match value according to configured bit-mask.
3298 - ``last``: specify upper bound to establish a range.
3299 - ``mask``: specify bit-mask with relevant bits set to one.
3300 - ``prefix``: generate bit-mask from a prefix length.
3302 These yield identical results::
3304 ipv4 src is 10.1.1.1
3308 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
3312 ipv4 src spec 10.1.1.1 src prefix 32
3316 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
3320 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
3322 Inclusive ranges can be defined with ``last``::
3324 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
3326 Note that ``mask`` affects both ``spec`` and ``last``::
3328 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
3329 # matches 10.1.0.0 to 10.2.255.255
3331 Properties can be modified multiple times::
3333 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
3337 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
3342 This section lists supported pattern items and their attributes, if any.
3344 - ``end``: end list of pattern items.
3346 - ``void``: no-op pattern item.
3348 - ``invert``: perform actions when pattern does not match.
3350 - ``any``: match any protocol for the current layer.
3352 - ``num {unsigned}``: number of layers covered.
3354 - ``pf``: match traffic from/to the physical function.
3356 - ``vf``: match traffic from/to a virtual function ID.
3358 - ``id {unsigned}``: VF ID.
3360 - ``phy_port``: match traffic from/to a specific physical port.
3362 - ``index {unsigned}``: physical port index.
3364 - ``port_id``: match traffic from/to a given DPDK port ID.
3366 - ``id {unsigned}``: DPDK port ID.
3368 - ``mark``: match value set in previously matched flow rule using the mark action.
3370 - ``id {unsigned}``: arbitrary integer value.
3372 - ``raw``: match an arbitrary byte string.
3374 - ``relative {boolean}``: look for pattern after the previous item.
3375 - ``search {boolean}``: search pattern from offset (see also limit).
3376 - ``offset {integer}``: absolute or relative offset for pattern.
3377 - ``limit {unsigned}``: search area limit for start of pattern.
3378 - ``pattern {string}``: byte string to look for.
3380 - ``eth``: match Ethernet header.
3382 - ``dst {MAC-48}``: destination MAC.
3383 - ``src {MAC-48}``: source MAC.
3384 - ``type {unsigned}``: EtherType or TPID.
3386 - ``vlan``: match 802.1Q/ad VLAN tag.
3388 - ``tci {unsigned}``: tag control information.
3389 - ``pcp {unsigned}``: priority code point.
3390 - ``dei {unsigned}``: drop eligible indicator.
3391 - ``vid {unsigned}``: VLAN identifier.
3392 - ``inner_type {unsigned}``: inner EtherType or TPID.
3394 - ``ipv4``: match IPv4 header.
3396 - ``tos {unsigned}``: type of service.
3397 - ``ttl {unsigned}``: time to live.
3398 - ``proto {unsigned}``: next protocol ID.
3399 - ``src {ipv4 address}``: source address.
3400 - ``dst {ipv4 address}``: destination address.
3402 - ``ipv6``: match IPv6 header.
3404 - ``tc {unsigned}``: traffic class.
3405 - ``flow {unsigned}``: flow label.
3406 - ``proto {unsigned}``: protocol (next header).
3407 - ``hop {unsigned}``: hop limit.
3408 - ``src {ipv6 address}``: source address.
3409 - ``dst {ipv6 address}``: destination address.
3411 - ``icmp``: match ICMP header.
3413 - ``type {unsigned}``: ICMP packet type.
3414 - ``code {unsigned}``: ICMP packet code.
3416 - ``udp``: match UDP header.
3418 - ``src {unsigned}``: UDP source port.
3419 - ``dst {unsigned}``: UDP destination port.
3421 - ``tcp``: match TCP header.
3423 - ``src {unsigned}``: TCP source port.
3424 - ``dst {unsigned}``: TCP destination port.
3426 - ``sctp``: match SCTP header.
3428 - ``src {unsigned}``: SCTP source port.
3429 - ``dst {unsigned}``: SCTP destination port.
3430 - ``tag {unsigned}``: validation tag.
3431 - ``cksum {unsigned}``: checksum.
3433 - ``vxlan``: match VXLAN header.
3435 - ``vni {unsigned}``: VXLAN identifier.
3437 - ``e_tag``: match IEEE 802.1BR E-Tag header.
3439 - ``grp_ecid_b {unsigned}``: GRP and E-CID base.
3441 - ``nvgre``: match NVGRE header.
3443 - ``tni {unsigned}``: virtual subnet ID.
3445 - ``mpls``: match MPLS header.
3447 - ``label {unsigned}``: MPLS label.
3449 - ``gre``: match GRE header.
3451 - ``protocol {unsigned}``: protocol type.
3453 - ``fuzzy``: fuzzy pattern match, expect faster than default.
3455 - ``thresh {unsigned}``: accuracy threshold.
3457 - ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header.
3459 - ``teid {unsigned}``: tunnel endpoint identifier.
3461 - ``geneve``: match GENEVE header.
3463 - ``vni {unsigned}``: virtual network identifier.
3464 - ``protocol {unsigned}``: protocol type.
3466 - ``vxlan-gpe``: match VXLAN-GPE header.
3468 - ``vni {unsigned}``: VXLAN-GPE identifier.
3470 - ``arp_eth_ipv4``: match ARP header for Ethernet/IPv4.
3472 - ``sha {MAC-48}``: sender hardware address.
3473 - ``spa {ipv4 address}``: sender IPv4 address.
3474 - ``tha {MAC-48}``: target hardware address.
3475 - ``tpa {ipv4 address}``: target IPv4 address.
3477 - ``ipv6_ext``: match presence of any IPv6 extension header.
3479 - ``next_hdr {unsigned}``: next header.
3481 - ``icmp6``: match any ICMPv6 header.
3483 - ``type {unsigned}``: ICMPv6 type.
3484 - ``code {unsigned}``: ICMPv6 code.
3486 - ``icmp6_nd_ns``: match ICMPv6 neighbor discovery solicitation.
3488 - ``target_addr {ipv6 address}``: target address.
3490 - ``icmp6_nd_na``: match ICMPv6 neighbor discovery advertisement.
3492 - ``target_addr {ipv6 address}``: target address.
3494 - ``icmp6_nd_opt``: match presence of any ICMPv6 neighbor discovery option.
3496 - ``type {unsigned}``: ND option type.
3498 - ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery source Ethernet
3499 link-layer address option.
3501 - ``sla {MAC-48}``: source Ethernet LLA.
3503 - ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery target Ethernet
3504 link-layer address option.
3506 - ``tla {MAC-48}``: target Ethernet LLA.
3511 A list of actions starts after the ``actions`` token in the same fashion as
3512 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
3513 terminated by a mandatory ``end`` action.
3515 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
3516 rte_flow_action_type``).
3518 Dropping all incoming UDPv4 packets can be expressed as follows::
3520 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3523 Several actions have configurable properties which must be specified when
3524 there is no valid default value. For example, ``queue`` requires a target
3527 This rule redirects incoming UDPv4 traffic to queue index 6::
3529 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3530 actions queue index 6 / end
3532 While this one could be rejected by PMDs (unspecified queue index)::
3534 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3537 As defined by *rte_flow*, the list is not ordered, all actions of a given
3538 rule are performed simultaneously. These are equivalent::
3540 queue index 6 / void / mark id 42 / end
3544 void / mark id 42 / queue index 6 / end
3546 All actions in a list should have different types, otherwise only the last
3547 action of a given type is taken into account::
3549 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
3553 drop / drop / drop / end # drop is performed only once
3557 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
3559 Considering they are performed simultaneously, opposite and overlapping
3560 actions can sometimes be combined when the end result is unambiguous::
3562 drop / queue index 6 / end # drop has no effect
3566 queue index 6 / rss queues 6 7 8 / end # queue has no effect
3570 drop / passthru / end # drop has no effect
3572 Note that PMDs may still refuse such combinations.
3577 This section lists supported actions and their attributes, if any.
3579 - ``end``: end list of actions.
3581 - ``void``: no-op action.
3583 - ``passthru``: let subsequent rule process matched packets.
3585 - ``jump``: redirect traffic to group on device.
3587 - ``group {unsigned}``: group to redirect to.
3589 - ``mark``: attach 32 bit value to packets.
3591 - ``id {unsigned}``: 32 bit value to return with packets.
3593 - ``flag``: flag packets.
3595 - ``queue``: assign packets to a given queue index.
3597 - ``index {unsigned}``: queue index to use.
3599 - ``drop``: drop packets (note: passthru has priority).
3601 - ``count``: enable counters for this rule.
3603 - ``rss``: spread packets among several queues.
3605 - ``func {hash function}``: RSS hash function to apply, allowed tokens are
3606 the same as `set_hash_global_config`_.
3608 - ``level {unsigned}``: encapsulation level for ``types``.
3610 - ``types [{RSS hash type} [...]] end``: specific RSS hash types, allowed
3611 tokens are the same as `set_hash_input_set`_, except that an empty list
3612 does not disable RSS but instead requests unspecified "best-effort"
3615 - ``key {string}``: RSS hash key, overrides ``key_len``.
3617 - ``key_len {unsigned}``: RSS hash key length in bytes, can be used in
3618 conjunction with ``key`` to pad or truncate it.
3620 - ``queues [{unsigned} [...]] end``: queue indices to use.
3622 - ``pf``: direct traffic to physical function.
3624 - ``vf``: direct traffic to a virtual function ID.
3626 - ``original {boolean}``: use original VF ID if possible.
3627 - ``id {unsigned}``: VF ID.
3629 - ``phy_port``: direct packets to physical port index.
3631 - ``original {boolean}``: use original port index if possible.
3632 - ``index {unsigned}``: physical port index.
3634 - ``port_id``: direct matching traffic to a given DPDK port ID.
3636 - ``original {boolean}``: use original DPDK port ID if possible.
3637 - ``id {unsigned}``: DPDK port ID.
3639 - ``of_set_mpls_ttl``: OpenFlow's ``OFPAT_SET_MPLS_TTL``.
3641 - ``mpls_ttl``: MPLS TTL.
3643 - ``of_dec_mpls_ttl``: OpenFlow's ``OFPAT_DEC_MPLS_TTL``.
3645 - ``of_set_nw_ttl``: OpenFlow's ``OFPAT_SET_NW_TTL``.
3647 - ``nw_ttl``: IP TTL.
3649 - ``of_dec_nw_ttl``: OpenFlow's ``OFPAT_DEC_NW_TTL``.
3651 - ``of_copy_ttl_out``: OpenFlow's ``OFPAT_COPY_TTL_OUT``.
3653 - ``of_copy_ttl_in``: OpenFlow's ``OFPAT_COPY_TTL_IN``.
3655 - ``of_pop_vlan``: OpenFlow's ``OFPAT_POP_VLAN``.
3657 - ``of_push_vlan``: OpenFlow's ``OFPAT_PUSH_VLAN``.
3659 - ``ethertype``: Ethertype.
3661 - ``of_set_vlan_vid``: OpenFlow's ``OFPAT_SET_VLAN_VID``.
3663 - ``vlan_vid``: VLAN id.
3665 - ``of_set_vlan_pcp``: OpenFlow's ``OFPAT_SET_VLAN_PCP``.
3667 - ``vlan_pcp``: VLAN priority.
3669 - ``of_pop_mpls``: OpenFlow's ``OFPAT_POP_MPLS``.
3671 - ``ethertype``: Ethertype.
3673 - ``of_push_mpls``: OpenFlow's ``OFPAT_PUSH_MPLS``.
3675 - ``ethertype``: Ethertype.
3677 - ``vxlan_encap``: Performs a VXLAN encapsulation, outer layer configuration
3678 is done through `Config VXLAN Encap outer layers`_.
3680 - ``vxlan_decap``: Performs a decapsulation action by stripping all headers of
3681 the VXLAN tunnel network overlay from the matched flow.
3683 Destroying flow rules
3684 ~~~~~~~~~~~~~~~~~~~~~
3686 ``flow destroy`` destroys one or more rules from their rule ID (as returned
3687 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
3688 times as necessary::
3690 flow destroy {port_id} rule {rule_id} [...]
3692 If successful, it will show::
3694 Flow rule #[...] destroyed
3696 It does not report anything for rule IDs that do not exist. The usual error
3697 message is shown when a rule cannot be destroyed::
3699 Caught error type [...] ([...]): [...]
3701 ``flow flush`` destroys all rules on a device and does not take extra
3702 arguments. It is bound to ``rte_flow_flush()``::
3704 flow flush {port_id}
3706 Any errors are reported as above.
3708 Creating several rules and destroying them::
3710 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3711 actions queue index 2 / end
3712 Flow rule #0 created
3713 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3714 actions queue index 3 / end
3715 Flow rule #1 created
3716 testpmd> flow destroy 0 rule 0 rule 1
3717 Flow rule #1 destroyed
3718 Flow rule #0 destroyed
3721 The same result can be achieved using ``flow flush``::
3723 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3724 actions queue index 2 / end
3725 Flow rule #0 created
3726 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3727 actions queue index 3 / end
3728 Flow rule #1 created
3729 testpmd> flow flush 0
3732 Non-existent rule IDs are ignored::
3734 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3735 actions queue index 2 / end
3736 Flow rule #0 created
3737 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3738 actions queue index 3 / end
3739 Flow rule #1 created
3740 testpmd> flow destroy 0 rule 42 rule 10 rule 2
3742 testpmd> flow destroy 0 rule 0
3743 Flow rule #0 destroyed
3749 ``flow query`` queries a specific action of a flow rule having that
3750 ability. Such actions collect information that can be reported using this
3751 command. It is bound to ``rte_flow_query()``::
3753 flow query {port_id} {rule_id} {action}
3755 If successful, it will display either the retrieved data for known actions
3756 or the following message::
3758 Cannot display result for action type [...] ([...])
3760 Otherwise, it will complain either that the rule does not exist or that some
3763 Flow rule #[...] not found
3767 Caught error type [...] ([...]): [...]
3769 Currently only the ``count`` action is supported. This action reports the
3770 number of packets that hit the flow rule and the total number of bytes. Its
3771 output has the following format::
3774 hits_set: [...] # whether "hits" contains a valid value
3775 bytes_set: [...] # whether "bytes" contains a valid value
3776 hits: [...] # number of packets
3777 bytes: [...] # number of bytes
3779 Querying counters for TCPv6 packets redirected to queue 6::
3781 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
3782 actions queue index 6 / count / end
3783 Flow rule #4 created
3784 testpmd> flow query 0 4 count
3795 ``flow list`` lists existing flow rules sorted by priority and optionally
3796 filtered by group identifiers::
3798 flow list {port_id} [group {group_id}] [...]
3800 This command only fails with the following message if the device does not
3805 Output consists of a header line followed by a short description of each
3806 flow rule, one per line. There is no output at all when no flow rules are
3807 configured on the device::
3809 ID Group Prio Attr Rule
3810 [...] [...] [...] [...] [...]
3812 ``Attr`` column flags:
3814 - ``i`` for ``ingress``.
3815 - ``e`` for ``egress``.
3817 Creating several flow rules and listing them::
3819 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3820 actions queue index 6 / end
3821 Flow rule #0 created
3822 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3823 actions queue index 2 / end
3824 Flow rule #1 created
3825 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
3826 actions rss queues 6 7 8 end / end
3827 Flow rule #2 created
3828 testpmd> flow list 0
3829 ID Group Prio Attr Rule
3830 0 0 0 i- ETH IPV4 => QUEUE
3831 1 0 0 i- ETH IPV6 => QUEUE
3832 2 0 5 i- ETH IPV4 UDP => RSS
3835 Rules are sorted by priority (i.e. group ID first, then priority level)::
3837 testpmd> flow list 1
3838 ID Group Prio Attr Rule
3839 0 0 0 i- ETH => COUNT
3840 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3841 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3842 1 24 0 i- ETH IPV4 UDP => QUEUE
3843 4 24 10 i- ETH IPV4 TCP => DROP
3844 3 24 20 i- ETH IPV4 => DROP
3845 2 24 42 i- ETH IPV4 UDP => QUEUE
3846 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3849 Output can be limited to specific groups::
3851 testpmd> flow list 1 group 0 group 63
3852 ID Group Prio Attr Rule
3853 0 0 0 i- ETH => COUNT
3854 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3855 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3856 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3859 Toggling isolated mode
3860 ~~~~~~~~~~~~~~~~~~~~~~
3862 ``flow isolate`` can be used to tell the underlying PMD that ingress traffic
3863 must only be injected from the defined flow rules; that no default traffic
3864 is expected outside those rules and the driver is free to assign more
3865 resources to handle them. It is bound to ``rte_flow_isolate()``::
3867 flow isolate {port_id} {boolean}
3869 If successful, enabling or disabling isolated mode shows either::
3871 Ingress traffic on port [...]
3872 is now restricted to the defined flow rules
3876 Ingress traffic on port [...]
3877 is not restricted anymore to the defined flow rules
3879 Otherwise, in case of error::
3881 Caught error type [...] ([...]): [...]
3883 Mainly due to its side effects, PMDs supporting this mode may not have the
3884 ability to toggle it more than once without reinitializing affected ports
3885 first (e.g. by exiting testpmd).
3887 Enabling isolated mode::
3889 testpmd> flow isolate 0 true
3890 Ingress traffic on port 0 is now restricted to the defined flow rules
3893 Disabling isolated mode::
3895 testpmd> flow isolate 0 false
3896 Ingress traffic on port 0 is not restricted anymore to the defined flow rules
3899 Sample QinQ flow rules
3900 ~~~~~~~~~~~~~~~~~~~~~~
3902 Before creating QinQ rule(s) the following commands should be issued to enable QinQ::
3904 testpmd> port stop 0
3905 testpmd> vlan set qinq on 0
3907 The above command sets the inner and outer TPID's to 0x8100.
3909 To change the TPID's the following commands should be used::
3911 testpmd> vlan set outer tpid 0xa100 0
3912 testpmd> vlan set inner tpid 0x9100 0
3913 testpmd> port start 0
3915 Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM.
3919 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 /
3920 vlan tci is 456 / end actions vf id 1 / queue index 0 / end
3921 Flow rule #0 validated
3923 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 /
3924 vlan tci is 456 / end actions vf id 123 / queue index 0 / end
3925 Flow rule #0 created
3927 testpmd> flow list 0
3928 ID Group Prio Attr Rule
3929 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3931 Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host.
3935 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 /
3936 vlan tci is 654 / end actions pf / queue index 0 / end
3937 Flow rule #1 validated
3939 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 /
3940 vlan tci is 654 / end actions pf / queue index 1 / end
3941 Flow rule #1 created
3943 testpmd> flow list 0
3944 ID Group Prio Attr Rule
3945 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3946 1 0 0 i- ETH VLAN VLAN=>PF QUEUE
3948 Sample VXLAN encapsulation rule
3949 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
3951 VXLAN encapsulation outer layer has default value pre-configured in testpmd
3952 source code, those can be changed by using the following commands
3954 IPv4 VXLAN outer header::
3956 testpmd> set vxlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src 127.0.0.1
3957 ip-dst 128.0.0.1 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
3958 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
3961 testpmd> set vxlan-with-vlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src
3962 127.0.0.1 ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11
3963 eth-dst 22:22:22:22:22:22
3964 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
3967 IPv6 VXLAN outer header::
3969 testpmd> set vxlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4 ip-src ::1
3970 ip-dst ::2222 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
3971 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
3974 testpmd> set vxlan-with-vlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4
3975 ip-src ::1 ip-dst ::2222 vlan-tci 34 eth-src 11:11:11:11:11:11
3976 eth-dst 22:22:22:22:22:22
3977 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
3983 The following sections show functions to load/unload eBPF based filters.
3988 Load an eBPF program as a callback for partciular RX/TX queue::
3990 testpmd> bpf-load rx|tx (portid) (queueid) (load-flags) (bpf-prog-filename)
3992 The available load-flags are:
3994 * ``J``: use JIT generated native code, otherwise BPF interpreter will be used.
3996 * ``M``: assume input parameter is a pointer to rte_mbuf, otherwise assume it is a pointer to first segment's data.
4002 You'll need clang v3.7 or above to build bpf program you'd like to load
4006 .. code-block:: console
4009 clang -O2 -target bpf -c t1.c
4011 Then to load (and JIT compile) t1.o at RX queue 0, port 1::
4013 .. code-block:: console
4015 testpmd> bpf-load rx 1 0 J ./dpdk.org/test/bpf/t1.o
4017 To load (not JITed) t1.o at TX queue 0, port 0::
4019 .. code-block:: console
4021 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o
4026 Unload previously loaded eBPF program for partciular RX/TX queue::
4028 testpmd> bpf-unload rx|tx (portid) (queueid)
4030 For example to unload BPF filter from TX queue 0, port 0:
4032 .. code-block:: console
4034 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o