2 Copyright(c) 2010-2016 Intel Corporation. All rights reserved.
5 Redistribution and use in source and binary forms, with or without
6 modification, are permitted provided that the following conditions
9 * Redistributions of source code must retain the above copyright
10 notice, this list of conditions and the following disclaimer.
11 * Redistributions in binary form must reproduce the above copyright
12 notice, this list of conditions and the following disclaimer in
13 the documentation and/or other materials provided with the
15 * Neither the name of Intel Corporation nor the names of its
16 contributors may be used to endorse or promote products derived
17 from this software without specific prior written permission.
19 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
20 "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
21 LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
22 A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
23 OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
24 SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
25 LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
26 DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
27 THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
28 (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
29 OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
33 Testpmd Runtime Functions
34 =========================
36 Where the testpmd application is started in interactive mode, (``-i|--interactive``),
37 it displays a prompt that can be used to start and stop forwarding,
38 configure the application, display statistics (including the extended NIC
39 statistics aka xstats) , set the Flow Director and other tasks::
43 The testpmd prompt has some, limited, readline support.
44 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
45 as well as access to the command history via the up-arrow.
47 There is also support for tab completion.
48 If you type a partial command and hit ``<TAB>`` you get a list of the available completions:
50 .. code-block:: console
52 testpmd> show port <TAB>
54 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
55 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
56 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
57 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
63 Some examples in this document are too long to fit on one line are are shown wrapped at `"\\"` for display purposes::
65 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
66 (pause_time) (send_xon) (port_id)
68 In the real ``testpmd>`` prompt these commands should be on a single line.
73 The testpmd has on-line help for the functions that are available at runtime.
74 These are divided into sections and can be accessed using help, help section or help all:
76 .. code-block:: console
80 help control : Start and stop forwarding.
81 help display : Displaying port, stats and config information.
82 help config : Configuration information.
83 help ports : Configuring ports.
84 help registers : Reading and setting port registers.
85 help filters : Filters configuration help.
86 help all : All of the above sections.
95 Start packet forwarding with current configuration::
102 Start packet forwarding with current configuration after sending specified number of bursts of packets::
104 testpmd> start tx_first (""|burst_num)
106 The default burst number is 1 when ``burst_num`` not presented.
111 Stop packet forwarding, and display accumulated statistics::
126 The functions in the following sections are used to display information about the
127 testpmd configuration or the NIC status.
132 Display information for a given port or all ports::
134 testpmd> show port (info|stats|xstats|fdir|stat_qmap|dcb_tc|cap) (port_id|all)
136 The available information categories are:
138 * ``info``: General port information such as MAC address.
140 * ``stats``: RX/TX statistics.
142 * ``xstats``: RX/TX extended NIC statistics.
144 * ``fdir``: Flow Director information and statistics.
146 * ``stat_qmap``: Queue statistics mapping.
148 * ``dcb_tc``: DCB information such as TC mapping.
150 * ``cap``: Supported offload capabilities.
154 .. code-block:: console
156 testpmd> show port info 0
158 ********************* Infos for port 0 *********************
160 MAC address: XX:XX:XX:XX:XX:XX
162 memory allocation on the socket: 0
164 Link speed: 40000 Mbps
165 Link duplex: full-duplex
166 Promiscuous mode: enabled
167 Allmulticast mode: disabled
168 Maximum number of MAC addresses: 64
169 Maximum number of MAC addresses of hash filtering: 0
174 Redirection table size: 512
175 Supported flow types:
195 Display the rss redirection table entry indicated by masks on port X::
197 testpmd> show port (port_id) rss reta (size) (mask0, mask1...)
199 size is used to indicate the hardware supported reta size
204 Display the RSS hash functions and RSS hash key of a port::
206 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]
211 Clear the port statistics for a given port or for all ports::
213 testpmd> clear port (info|stats|xstats|fdir|stat_qmap) (port_id|all)
217 testpmd> clear port stats all
222 Display information for a given port's RX/TX queue::
224 testpmd> show (rxq|txq) info (port_id) (queue_id)
229 Displays the configuration of the application.
230 The configuration comes from the command-line, the runtime or the application defaults::
232 testpmd> show config (rxtx|cores|fwd|txpkts)
234 The available information categories are:
236 * ``rxtx``: RX/TX configuration items.
238 * ``cores``: List of forwarding cores.
240 * ``fwd``: Packet forwarding configuration.
242 * ``txpkts``: Packets to TX configuration.
246 .. code-block:: console
248 testpmd> show config rxtx
250 io packet forwarding - CRC stripping disabled - packets/burst=16
251 nb forwarding cores=2 - nb forwarding ports=1
252 RX queues=1 - RX desc=128 - RX free threshold=0
253 RX threshold registers: pthresh=8 hthresh=8 wthresh=4
254 TX queues=1 - TX desc=512 - TX free threshold=0
255 TX threshold registers: pthresh=36 hthresh=0 wthresh=0
256 TX RS bit threshold=0 - TXQ flags=0x0
261 Set the packet forwarding mode::
263 testpmd> set fwd (io|mac|macswap|flowgen| \
264 rxonly|txonly|csum|icmpecho) (""|retry)
266 ``retry`` can be specified for forwarding engines except ``rx_only``.
268 The available information categories are:
270 * ``io``: Forwards packets "as-is" in I/O mode.
271 This is the fastest possible forwarding operation as it does not access packets data.
272 This is the default mode.
274 * ``mac``: Changes the source and the destination Ethernet addresses of packets before forwarding them.
275 Default application behaviour is to set source Ethernet address to that of the transmitting interface, and destination
276 address to a dummy value (set during init). The user may specify a target destination Ethernet address via the 'eth-peer' or
277 'eth-peer-configfile' command-line options. It is not currently possible to specify a specific source Ethernet address.
279 * ``macswap``: MAC swap forwarding mode.
280 Swaps the source and the destination Ethernet addresses of packets before forwarding them.
282 * ``flowgen``: Multi-flow generation mode.
283 Originates a number of flows (with varying destination IP addresses), and terminate receive traffic.
285 * ``rxonly``: Receives packets but doesn't transmit them.
287 * ``txonly``: Generates and transmits packets without receiving any.
289 * ``csum``: Changes the checksum field with hardware or software methods depending on the offload flags on the packet.
291 * ``icmpecho``: Receives a burst of packets, lookup for IMCP echo requests and, if any, send back ICMP echo replies.
293 * ``ieee1588``: Demonstrate L2 IEEE1588 V2 PTP timestamping for RX and TX. Requires ``CONFIG_RTE_LIBRTE_IEEE1588=y``.
295 Note: TX timestamping is only available in the "Full Featured" TX path. To force ``testpmd`` into this mode set ``--txqflags=0``.
299 testpmd> set fwd rxonly
301 Set rxonly packet forwarding mode
307 Display an RX descriptor for a port RX queue::
309 testpmd> read rxd (port_id) (queue_id) (rxd_id)
313 testpmd> read rxd 0 0 4
314 0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180
319 Display a TX descriptor for a port TX queue::
321 testpmd> read txd (port_id) (queue_id) (txd_id)
325 testpmd> read txd 0 0 4
326 0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C
329 Configuration Functions
330 -----------------------
332 The testpmd application can be configured from the runtime as well as from the command-line.
334 This section details the available configuration functions that are available.
338 Configuration changes only become active when forwarding is started/restarted.
343 Reset forwarding to the default configuration::
350 Set the debug verbosity level::
352 testpmd> set verbose (level)
354 Currently the only available levels are 0 (silent except for error) and 1 (fully verbose).
359 Set the number of ports used by the application:
363 This is equivalent to the ``--nb-ports`` command-line option.
368 Set the number of cores used by the application::
370 testpmd> set nbcore (num)
372 This is equivalent to the ``--nb-cores`` command-line option.
376 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port.
381 Set the forwarding cores hexadecimal mask::
383 testpmd> set coremask (mask)
385 This is equivalent to the ``--coremask`` command-line option.
389 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding.
394 Set the forwarding ports hexadecimal mask::
396 testpmd> set portmask (mask)
398 This is equivalent to the ``--portmask`` command-line option.
403 Set number of packets per burst::
405 testpmd> set burst (num)
407 This is equivalent to the ``--burst command-line`` option.
409 When retry is enabled, the transmit delay time and number of retries can also be set::
411 testpmd> set burst tx delay (microseconds) retry (num)
416 Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode::
418 testpmd> set txpkts (x[,y]*)
420 Where x[,y]* represents a CSV list of values, without white space.
425 Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes::
427 testpmd> set txsplit (off|on|rand)
431 * ``off`` disable packet copy & split for CSUM mode.
433 * ``on`` split outgoing packet into multiple segments. Size of each segment
434 and number of segments per packet is determined by ``set txpkts`` command
437 * ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments.
442 Set the list of forwarding cores::
444 testpmd> set corelist (x[,y]*)
446 For example, to change the forwarding cores:
448 .. code-block:: console
450 testpmd> set corelist 3,1
451 testpmd> show config fwd
453 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled
454 Logical Core 3 (socket 0) forwards packets on 1 streams:
455 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01
456 Logical Core 1 (socket 0) forwards packets on 1 streams:
457 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
461 The cores are used in the same order as specified on the command line.
466 Set the list of forwarding ports::
468 testpmd> set portlist (x[,y]*)
470 For example, to change the port forwarding:
472 .. code-block:: console
474 testpmd> set portlist 0,2,1,3
475 testpmd> show config fwd
477 io packet forwarding - ports=4 - cores=1 - streams=4
478 Logical Core 3 (socket 0) forwards packets on 4 streams:
479 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01
480 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
481 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03
482 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02
487 Enable/disable tx loopback::
489 testpmd> set tx loopback (port_id) (on|off)
494 set drop enable bit for all queues::
496 testpmd> set all queues drop (port_id) (on|off)
498 set split drop enable (for VF)
499 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
501 set split drop enable bit for VF from PF::
503 testpmd> set vf split drop (port_id) (vf_id) (on|off)
505 set mac antispoof (for VF)
506 ~~~~~~~~~~~~~~~~~~~~~~~~~~
508 Set mac antispoof for a VF from the PF::
510 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off)
515 Enable/disable MACsec offload::
517 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off)
518 testpmd> set macsec offload (port_id) off
523 Configure MACsec secure connection (SC)::
525 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi)
529 The pi argument is ignored for tx.
530 Check the NIC Datasheet for hardware limits.
535 Configure MACsec secure association (SA)::
537 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key)
541 The IDX value must be 0 or 1.
542 Check the NIC Datasheet for hardware limits.
544 set broadcast mode (for VF)
545 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
547 Set broadcast mode for a VF from the PF::
549 testpmd> set vf broadcast (port_id) (vf_id) (on|off)
554 Set the VLAN strip on a port::
556 testpmd> vlan set strip (on|off) (port_id)
561 Set the VLAN strip for a queue on a port::
563 testpmd> vlan set stripq (on|off) (port_id,queue_id)
565 vlan set stripq (for VF)
566 ~~~~~~~~~~~~~~~~~~~~~~~~
568 Set VLAN strip for all queues in a pool for a VF from the PF::
570 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off)
572 vlan set insert (for VF)
573 ~~~~~~~~~~~~~~~~~~~~~~~~
575 Set VLAN insert for a VF from the PF::
577 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id)
579 vlan set tag (for VF)
580 ~~~~~~~~~~~~~~~~~~~~~
582 Set VLAN tag for a VF from the PF::
584 testpmd> set vf vlan tag (port_id) (vf_id) (on|off)
586 vlan set antispoof (for VF)
587 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
589 Set VLAN antispoof for a VF from the PF::
591 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off)
596 Set the VLAN filter on a port::
598 testpmd> vlan set filter (on|off) (port_id)
603 Set the VLAN QinQ (extended queue in queue) on for a port::
605 testpmd> vlan set qinq (on|off) (port_id)
610 Set the inner or outer VLAN TPID for packet filtering on a port::
612 testpmd> vlan set (inner|outer) tpid (value) (port_id)
616 TPID value must be a 16-bit number (value <= 65536).
621 Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID::
623 testpmd> rx_vlan add (vlan_id|all) (port_id)
627 VLAN filter must be set on that port. VLAN ID < 4096.
628 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries
629 in VFTA table. This is important if enabling all vlan_ids.
634 Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID::
636 testpmd> rx_vlan rm (vlan_id|all) (port_id)
641 Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID::
643 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask)
648 Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID::
650 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask)
655 Add a tunnel filter on a port::
657 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \
658 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
659 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
661 The available information categories are:
663 * ``vxlan``: Set tunnel type as VXLAN.
665 * ``nvgre``: Set tunnel type as NVGRE.
667 * ``ipingre``: Set tunnel type as IP-in-GRE.
669 * ``imac-ivlan``: Set filter type as Inner MAC and VLAN.
671 * ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID.
673 * ``imac-tenid``: Set filter type as Inner MAC and tenant ID.
675 * ``imac``: Set filter type as Inner MAC.
677 * ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID.
679 * ``oip``: Set filter type as Outer IP.
681 * ``iip``: Set filter type as Inner IP.
685 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \
686 192.168.2.2 0 ipingre oip 1 1
688 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP.
693 Remove a tunnel filter on a port::
695 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \
696 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
697 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
702 Add an UDP port for VXLAN packet filter on a port::
704 testpmd> rx_vxlan_port add (udp_port) (port_id)
709 Remove an UDP port for VXLAN packet filter on a port::
711 testpmd> rx_vxlan_port rm (udp_port) (port_id)
716 Set hardware insertion of VLAN IDs in packets sent on a port::
718 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer]
720 For example, set a single VLAN ID (5) insertion on port 0::
724 Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1::
732 Set port based hardware insertion of VLAN ID in packets sent on a port::
734 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off)
739 Disable hardware insertion of a VLAN header in packets sent on a port::
741 testpmd> tx_vlan reset (port_id)
746 Select hardware or software calculation of the checksum when
747 transmitting a packet using the ``csum`` forwarding engine::
749 testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id)
753 * ``ip|udp|tcp|sctp`` always relate to the inner layer.
755 * ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized
756 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are
757 supported). See also the ``csum parse-tunnel`` command.
761 Check the NIC Datasheet for hardware limits.
766 Define how tunneled packets should be handled by the csum forward
769 testpmd> csum parse-tunnel (on|off) (tx_port_id)
771 If enabled, the csum forward engine will try to recognize supported
772 tunnel headers (vxlan, gre, ipip).
774 If disabled, treat tunnel packets as non-tunneled packets (a inner
775 header is handled as a packet payload).
779 The port argument is the TX port like in the ``csum set`` command.
783 Consider a packet in packet like the following::
785 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in
787 * If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
788 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the
789 ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``).
791 * If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
792 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``.
797 Display tx checksum offload configuration::
799 testpmd> csum show (port_id)
804 Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine::
806 testpmd> tso set (segsize) (port_id)
810 Check the NIC datasheet for hardware limits.
815 Display the status of TCP Segmentation Offload::
817 testpmd> tso show (port_id)
822 Add an alternative MAC address to a port::
824 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX)
829 Remove a MAC address from a port::
831 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX)
833 mac_addr add (for VF)
834 ~~~~~~~~~~~~~~~~~~~~~
836 Add an alternative MAC address for a VF to a port::
838 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX)
843 Set the default MAC address for a port::
845 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX)
847 mac_addr set (for VF)
848 ~~~~~~~~~~~~~~~~~~~~~
850 Set the MAC address for a VF from the PF::
852 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX)
857 Set the unicast hash filter(s) on/off for a port::
859 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
864 Set the promiscuous mode on for a port or for all ports.
865 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
867 testpmd> set promisc (port_id|all) (on|off)
872 Set the allmulti mode for a port or for all ports::
874 testpmd> set allmulti (port_id|all) (on|off)
876 Same as the ifconfig (8) option. Controls how multicast packets are handled.
881 Set the unicast promiscuous mode for a VF from PF.
882 It's supported by Intel i40e NICs now.
883 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
885 testpmd> set vf promisc (port_id) (vf_id) (on|off)
887 set allmulticast (for VF)
888 ~~~~~~~~~~~~~~~~~~~~~~~~~
890 Set the multicast promiscuous mode for a VF from PF.
891 It's supported by Intel i40e NICs now.
892 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
894 testpmd> set vf allmulti (port_id) (vf_id) (on|off)
896 set tx max bandwidth (for VF)
897 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
899 Set TX max absolute bandwidth (Mbps) for a VF from PF::
901 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth)
903 set tc tx min bandwidth (for VF)
904 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
906 Set all TCs' TX min relative bandwidth (%) for a VF from PF::
908 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...)
910 set tc tx max bandwidth (for VF)
911 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
913 Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF::
915 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth)
917 set tc strict link priority mode
918 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
920 Set some TCs' strict link priority mode on a physical port::
922 testpmd> set tx strict-link-priority (port_id) (tc_bitmap)
927 Set the link flow control parameter on a port::
929 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
930 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
931 autoneg (on|off) (port_id)
935 * ``high_water`` (integer): High threshold value to trigger XOFF.
937 * ``low_water`` (integer): Low threshold value to trigger XON.
939 * ``pause_time`` (integer): Pause quota in the Pause frame.
941 * ``send_xon`` (0/1): Send XON frame.
943 * ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
945 * ``autoneg``: Change the auto-negotiation parameter.
950 Set the priority flow control parameter on a port::
952 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
953 (pause_time) (priority) (port_id)
957 * ``high_water`` (integer): High threshold value.
959 * ``low_water`` (integer): Low threshold value.
961 * ``pause_time`` (integer): Pause quota in the Pause frame.
963 * ``priority`` (0-7): VLAN User Priority.
968 Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
970 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
972 For example, to set rx queue 2 on port 0 to mapping 5::
974 testpmd>set stat_qmap rx 0 2 5
976 set port - rx/tx (for VF)
977 ~~~~~~~~~~~~~~~~~~~~~~~~~
979 Set VF receive/transmit from a port::
981 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
983 set port - mac address filter (for VF)
984 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
986 Add/Remove unicast or multicast MAC addr filter for a VF::
988 testpmd> set port (port_id) vf (vf_id) (mac_addr) \
989 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
991 set port - rx mode(for VF)
992 ~~~~~~~~~~~~~~~~~~~~~~~~~~
994 Set the VF receive mode of a port::
996 testpmd> set port (port_id) vf (vf_id) \
997 rxmode (AUPE|ROPE|BAM|MPE) (on|off)
999 The available receive modes are:
1001 * ``AUPE``: Accepts untagged VLAN.
1003 * ``ROPE``: Accepts unicast hash.
1005 * ``BAM``: Accepts broadcast packets.
1007 * ``MPE``: Accepts all multicast packets.
1009 set port - tx_rate (for Queue)
1010 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1012 Set TX rate limitation for a queue on a port::
1014 testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1016 set port - tx_rate (for VF)
1017 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
1019 Set TX rate limitation for queues in VF on a port::
1021 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1023 set port - mirror rule
1024 ~~~~~~~~~~~~~~~~~~~~~~
1026 Set pool or vlan type mirror rule for a port::
1028 testpmd> set port (port_id) mirror-rule (rule_id) \
1029 (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1030 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1032 Set link mirror rule for a port::
1034 testpmd> set port (port_id) mirror-rule (rule_id) \
1035 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1037 For example to enable mirror traffic with vlan 0,1 to pool 0::
1039 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1041 reset port - mirror rule
1042 ~~~~~~~~~~~~~~~~~~~~~~~~
1044 Reset a mirror rule for a port::
1046 testpmd> reset port (port_id) mirror-rule (rule_id)
1051 Set the flush on RX streams before forwarding.
1052 The default is flush ``on``.
1053 Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1055 testpmd> set flush_rx off
1060 Set the bypass mode for the lowest port on bypass enabled NIC::
1062 testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1067 Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1069 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1070 mode (normal|bypass|isolate) (port_id)
1074 * ``timeout``: Enable bypass after watchdog timeout.
1076 * ``os_on``: Enable bypass when OS/board is powered on.
1078 * ``os_off``: Enable bypass when OS/board is powered off.
1080 * ``power_on``: Enable bypass when power supply is turned on.
1082 * ``power_off``: Enable bypass when power supply is turned off.
1088 Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1090 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1095 Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1097 testpmd> show bypass config (port_id)
1102 Set link up for a port::
1104 testpmd> set link-up port (port id)
1109 Set link down for a port::
1111 testpmd> set link-down port (port id)
1116 Enable E-tag insertion for a VF on a port::
1118 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1120 Disable E-tag insertion for a VF on a port::
1122 testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1124 Enable/disable E-tag stripping on a port::
1126 testpmd> E-tag set stripping (on|off) port (port_id)
1128 Enable/disable E-tag based forwarding on a port::
1130 testpmd> E-tag set forwarding (on|off) port (port_id)
1132 Add an E-tag forwarding filter on a port::
1134 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1136 Delete an E-tag forwarding filter on a port::
1137 testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1143 The following sections show functions for configuring ports.
1147 Port configuration changes only become active when forwarding is started/restarted.
1152 Attach a port specified by pci address or virtual device args::
1154 testpmd> port attach (identifier)
1156 To attach a new pci device, the device should be recognized by kernel first.
1157 Then it should be moved under DPDK management.
1158 Finally the port can be attached to testpmd.
1160 For example, to move a pci device using ixgbe under DPDK management:
1162 .. code-block:: console
1164 # Check the status of the available devices.
1165 ./usertools/dpdk-devbind.py --status
1167 Network devices using DPDK-compatible driver
1168 ============================================
1171 Network devices using kernel driver
1172 ===================================
1173 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1176 # Bind the device to igb_uio.
1177 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1180 # Recheck the status of the devices.
1181 ./usertools/dpdk-devbind.py --status
1182 Network devices using DPDK-compatible driver
1183 ============================================
1184 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1186 To attach a port created by virtual device, above steps are not needed.
1188 For example, to attach a port whose pci address is 0000:0a:00.0.
1190 .. code-block:: console
1192 testpmd> port attach 0000:0a:00.0
1193 Attaching a new port...
1194 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1195 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1196 EAL: PCI memory mapped at 0x7f83bfa00000
1197 EAL: PCI memory mapped at 0x7f83bfa80000
1198 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1199 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1200 Port 0 is attached. Now total ports is 1
1203 For example, to attach a port created by pcap PMD.
1205 .. code-block:: console
1207 testpmd> port attach net_pcap0
1208 Attaching a new port...
1209 PMD: Initializing pmd_pcap for net_pcap0
1210 PMD: Creating pcap-backed ethdev on numa socket 0
1211 Port 0 is attached. Now total ports is 1
1214 In this case, identifier is ``net_pcap0``.
1215 This identifier format is the same as ``--vdev`` format of DPDK applications.
1217 For example, to re-attach a bonded port which has been previously detached,
1218 the mode and slave parameters must be given.
1220 .. code-block:: console
1222 testpmd> port attach net_bond_0,mode=0,slave=1
1223 Attaching a new port...
1224 EAL: Initializing pmd_bond for net_bond_0
1225 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1226 Port 0 is attached. Now total ports is 1
1233 Detach a specific port::
1235 testpmd> port detach (port_id)
1237 Before detaching a port, the port should be stopped and closed.
1239 For example, to detach a pci device port 0.
1241 .. code-block:: console
1243 testpmd> port stop 0
1246 testpmd> port close 0
1250 testpmd> port detach 0
1252 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1253 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1254 EAL: PCI memory unmapped at 0x7f83bfa00000
1255 EAL: PCI memory unmapped at 0x7f83bfa80000
1259 For example, to detach a virtual device port 0.
1261 .. code-block:: console
1263 testpmd> port stop 0
1266 testpmd> port close 0
1270 testpmd> port detach 0
1272 PMD: Closing pcap ethdev on numa socket 0
1273 Port 'net_pcap0' is detached. Now total ports is 0
1276 To remove a pci device completely from the system, first detach the port from testpmd.
1277 Then the device should be moved under kernel management.
1278 Finally the device can be removed using kernel pci hotplug functionality.
1280 For example, to move a pci device under kernel management:
1282 .. code-block:: console
1284 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1286 ./usertools/dpdk-devbind.py --status
1288 Network devices using DPDK-compatible driver
1289 ============================================
1292 Network devices using kernel driver
1293 ===================================
1294 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1296 To remove a port created by a virtual device, above steps are not needed.
1301 Start all ports or a specific port::
1303 testpmd> port start (port_id|all)
1308 Stop all ports or a specific port::
1310 testpmd> port stop (port_id|all)
1315 Close all ports or a specific port::
1317 testpmd> port close (port_id|all)
1319 port start/stop queue
1320 ~~~~~~~~~~~~~~~~~~~~~
1322 Start/stop a rx/tx queue on a specific port::
1324 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1326 Only take effect when port is started.
1331 Set the speed and duplex mode for all ports or a specific port::
1333 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1334 duplex (half|full|auto)
1336 port config - queues/descriptors
1337 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1339 Set number of queues/descriptors for rxq, txq, rxd and txd::
1341 testpmd> port config all (rxq|txq|rxd|txd) (value)
1343 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1345 port config - max-pkt-len
1346 ~~~~~~~~~~~~~~~~~~~~~~~~~
1348 Set the maximum packet length::
1350 testpmd> port config all max-pkt-len (value)
1352 This is equivalent to the ``--max-pkt-len`` command-line option.
1354 port config - CRC Strip
1355 ~~~~~~~~~~~~~~~~~~~~~~~
1357 Set hardware CRC stripping on or off for all ports::
1359 testpmd> port config all crc-strip (on|off)
1361 CRC stripping is off by default.
1363 The ``on`` option is equivalent to the ``--crc-strip`` command-line option.
1365 port config - scatter
1366 ~~~~~~~~~~~~~~~~~~~~~~~
1368 Set RX scatter mode on or off for all ports::
1370 testpmd> port config all scatter (on|off)
1372 RX scatter mode is off by default.
1374 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1376 port config - TX queue flags
1377 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1379 Set a hexadecimal bitmap of TX queue flags for all ports::
1381 testpmd> port config all txqflags value
1383 This command is equivalent to the ``--txqflags`` command-line option.
1385 port config - RX Checksum
1386 ~~~~~~~~~~~~~~~~~~~~~~~~~
1388 Set hardware RX checksum offload to on or off for all ports::
1390 testpmd> port config all rx-cksum (on|off)
1392 Checksum offload is off by default.
1394 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1399 Set hardware VLAN on or off for all ports::
1401 testpmd> port config all hw-vlan (on|off)
1403 Hardware VLAN is on by default.
1405 The ``off`` option is equivalent to the ``--disable-hw-vlan`` command-line option.
1407 port config - VLAN filter
1408 ~~~~~~~~~~~~~~~~~~~~~~~~~
1410 Set hardware VLAN filter on or off for all ports::
1412 testpmd> port config all hw-vlan-filter (on|off)
1414 Hardware VLAN filter is on by default.
1416 The ``off`` option is equivalent to the ``--disable-hw-vlan-filter`` command-line option.
1418 port config - VLAN strip
1419 ~~~~~~~~~~~~~~~~~~~~~~~~
1421 Set hardware VLAN strip on or off for all ports::
1423 testpmd> port config all hw-vlan-strip (on|off)
1425 Hardware VLAN strip is on by default.
1427 The ``off`` option is equivalent to the ``--disable-hw-vlan-strip`` command-line option.
1429 port config - VLAN extend
1430 ~~~~~~~~~~~~~~~~~~~~~~~~~
1432 Set hardware VLAN extend on or off for all ports::
1434 testpmd> port config all hw-vlan-extend (on|off)
1436 Hardware VLAN extend is off by default.
1438 The ``off`` option is equivalent to the ``--disable-hw-vlan-extend`` command-line option.
1440 port config - Drop Packets
1441 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1443 Set packet drop for packets with no descriptors on or off for all ports::
1445 testpmd> port config all drop-en (on|off)
1447 Packet dropping for packets with no descriptors is off by default.
1449 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1454 Set the RSS (Receive Side Scaling) mode on or off::
1456 testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1458 RSS is on by default.
1460 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1462 port config - RSS Reta
1463 ~~~~~~~~~~~~~~~~~~~~~~
1465 Set the RSS (Receive Side Scaling) redirection table::
1467 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1472 Set the DCB mode for an individual port::
1474 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1476 The traffic class should be 4 or 8.
1481 Set the number of packets per burst::
1483 testpmd> port config all burst (value)
1485 This is equivalent to the ``--burst`` command-line option.
1487 port config - Threshold
1488 ~~~~~~~~~~~~~~~~~~~~~~~
1490 Set thresholds for TX/RX queues::
1492 testpmd> port config all (threshold) (value)
1494 Where the threshold type can be:
1496 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1498 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1500 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1502 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1504 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1506 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1508 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1510 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1512 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1514 These threshold options are also available from the command-line.
1519 Set the value of ether-type for E-tag::
1521 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1523 Enable/disable the E-tag support::
1525 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1528 Link Bonding Functions
1529 ----------------------
1531 The Link Bonding functions make it possible to dynamically create and
1532 manage link bonding devices from within testpmd interactive prompt.
1534 create bonded device
1535 ~~~~~~~~~~~~~~~~~~~~
1537 Create a new bonding device::
1539 testpmd> create bonded device (mode) (socket)
1541 For example, to create a bonded device in mode 1 on socket 0::
1543 testpmd> create bonded 1 0
1544 created new bonded device (port X)
1549 Adds Ethernet device to a Link Bonding device::
1551 testpmd> add bonding slave (slave id) (port id)
1553 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
1555 testpmd> add bonding slave 6 10
1558 remove bonding slave
1559 ~~~~~~~~~~~~~~~~~~~~
1561 Removes an Ethernet slave device from a Link Bonding device::
1563 testpmd> remove bonding slave (slave id) (port id)
1565 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
1567 testpmd> remove bonding slave 6 10
1572 Set the Link Bonding mode of a Link Bonding device::
1574 testpmd> set bonding mode (value) (port id)
1576 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
1578 testpmd> set bonding mode 3 10
1583 Set an Ethernet slave device as the primary device on a Link Bonding device::
1585 testpmd> set bonding primary (slave id) (port id)
1587 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
1589 testpmd> set bonding primary 6 10
1594 Set the MAC address of a Link Bonding device::
1596 testpmd> set bonding mac (port id) (mac)
1598 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
1600 testpmd> set bonding mac 10 00:00:00:00:00:01
1602 set bonding xmit_balance_policy
1603 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1605 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
1607 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
1609 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
1611 testpmd> set bonding xmit_balance_policy 10 l34
1614 set bonding mon_period
1615 ~~~~~~~~~~~~~~~~~~~~~~
1617 Set the link status monitoring polling period in milliseconds for a bonding device.
1619 This adds support for PMD slave devices which do not support link status interrupts.
1620 When the mon_period is set to a value greater than 0 then all PMD's which do not support
1621 link status ISR will be queried every polling interval to check if their link status has changed::
1623 testpmd> set bonding mon_period (port_id) (value)
1625 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
1627 testpmd> set bonding mon_period 5 150
1633 Show the current configuration of a Link Bonding device::
1635 testpmd> show bonding config (port id)
1638 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
1639 in balance mode with a transmission policy of layer 2+3::
1641 testpmd> show bonding config 9
1643 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
1645 Active Slaves (3): [1 3 4]
1652 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
1653 This is mainly useful for debugging purposes.
1654 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
1655 and fields that can be accessed.
1660 Display the value of a port register::
1662 testpmd> read reg (port_id) (address)
1664 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
1666 testpmd> read reg 0 0xEE00
1667 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
1672 Display a port register bit field::
1674 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
1676 For example, reading the lowest two bits from the register in the example above::
1678 testpmd> read regfield 0 0xEE00 0 1
1679 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
1684 Display a single port register bit::
1686 testpmd> read regbit (port_id) (address) (bit_x)
1688 For example, reading the lowest bit from the register in the example above::
1690 testpmd> read regbit 0 0xEE00 0
1691 port 0 PCI register at offset 0xEE00: bit 0=1
1696 Set the value of a port register::
1698 testpmd> write reg (port_id) (address) (value)
1700 For example, to clear a register::
1702 testpmd> write reg 0 0xEE00 0x0
1703 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
1708 Set bit field of a port register::
1710 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
1712 For example, writing to the register cleared in the example above::
1714 testpmd> write regfield 0 0xEE00 0 1 2
1715 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
1720 Set single bit value of a port register::
1722 testpmd> write regbit (port_id) (address) (bit_x) (value)
1724 For example, to set the high bit in the register from the example above::
1726 testpmd> write regbit 0 0xEE00 31 1
1727 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
1733 This section details the available filter functions that are available.
1735 Note these functions interface the deprecated legacy filtering framework,
1736 superseded by *rte_flow*. See `Flow rules management`_.
1739 ~~~~~~~~~~~~~~~~~~~~
1741 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
1743 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
1744 ethertype (ether_type) (drop|fwd) queue (queue_id)
1746 The available information parameters are:
1748 * ``port_id``: The port which the Ethertype filter assigned on.
1750 * ``mac_addr``: Compare destination mac address.
1752 * ``mac_ignr``: Ignore destination mac address match.
1754 * ``mac_address``: Destination mac address to match.
1756 * ``ether_type``: The EtherType value want to match,
1757 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
1759 * ``queue_id``: The receive queue associated with this EtherType filter.
1760 It is meaningless when deleting or dropping.
1762 Example, to add/remove an ethertype filter rule::
1764 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
1765 ethertype 0x0806 fwd queue 3
1767 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
1768 ethertype 0x0806 fwd queue 3
1773 Add or delete a 2-tuple filter,
1774 which identifies packets by specific protocol and destination TCP/UDP port
1775 and forwards packets into one of the receive queues::
1777 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
1778 protocol (protocol_value) mask (mask_value) \
1779 tcp_flags (tcp_flags_value) priority (prio_value) \
1782 The available information parameters are:
1784 * ``port_id``: The port which the 2-tuple filter assigned on.
1786 * ``dst_port_value``: Destination port in L4.
1788 * ``protocol_value``: IP L4 protocol.
1790 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
1792 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
1794 * ``prio_value``: Priority of this filter.
1796 * ``queue_id``: The receive queue associated with this 2-tuple filter.
1798 Example, to add/remove an 2tuple filter rule::
1800 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
1801 tcp_flags 0x02 priority 3 queue 3
1803 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
1804 tcp_flags 0x02 priority 3 queue 3
1809 Add or delete a 5-tuple filter,
1810 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
1811 and routes packets into one of the receive queues::
1813 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
1814 (src_address) dst_port (dst_port_value) \
1815 src_port (src_port_value) protocol (protocol_value) \
1816 mask (mask_value) tcp_flags (tcp_flags_value) \
1817 priority (prio_value) queue (queue_id)
1819 The available information parameters are:
1821 * ``port_id``: The port which the 5-tuple filter assigned on.
1823 * ``dst_address``: Destination IP address.
1825 * ``src_address``: Source IP address.
1827 * ``dst_port_value``: TCP/UDP destination port.
1829 * ``src_port_value``: TCP/UDP source port.
1831 * ``protocol_value``: L4 protocol.
1833 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
1835 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
1837 * ``prio_value``: The priority of this filter.
1839 * ``queue_id``: The receive queue associated with this 5-tuple filter.
1841 Example, to add/remove an 5tuple filter rule::
1843 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1844 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1845 flags 0x0 priority 3 queue 3
1847 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1848 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1849 flags 0x0 priority 3 queue 3
1854 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
1856 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
1858 The available information parameters are:
1860 * ``port_id``: The port which the SYN filter assigned on.
1862 * ``high``: This SYN filter has higher priority than other filters.
1864 * ``low``: This SYN filter has lower priority than other filters.
1866 * ``queue_id``: The receive queue associated with this SYN filter
1870 testpmd> syn_filter 0 add priority high queue 3
1875 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
1876 and routed into one of the receive queues::
1878 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
1879 mask (mask_value) priority (prio_value) queue (queue_id)
1881 The available information parameters are:
1883 * ``port_id``: The port which the Flex filter is assigned on.
1885 * ``len_value``: Filter length in bytes, no greater than 128.
1887 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
1889 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
1891 * ``prio_value``: The priority of this filter.
1893 * ``queue_id``: The receive queue associated with this Flex filter.
1897 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
1898 mask 000C priority 3 queue 3
1900 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
1901 mask 000C priority 3 queue 3
1904 .. _testpmd_flow_director:
1906 flow_director_filter
1907 ~~~~~~~~~~~~~~~~~~~~
1909 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
1911 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
1912 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
1914 * Perfect match filters.
1915 The hardware checks a match between the masked fields of the received packets and the programmed filters.
1916 The masked fields are for IP flow.
1918 * Signature filters.
1919 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
1921 * Perfect-mac-vlan match filters.
1922 The hardware checks a match between the masked fields of the received packets and the programmed filters.
1923 The masked fields are for MAC VLAN flow.
1925 * Perfect-tunnel match filters.
1926 The hardware checks a match between the masked fields of the received packets and the programmed filters.
1927 The masked fields are for tunnel flow.
1929 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
1930 per flow type and the flexible payload.
1932 The Flow Director can also mask out parts of all of these fields so that filters
1933 are only applied to certain fields or parts of the fields.
1935 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
1937 # Commands to add flow director filters of different flow types::
1939 flow_director_filter (port_id) mode IP (add|del|update) \
1940 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
1941 src (src_ip_address) dst (dst_ip_address) \
1942 tos (tos_value) proto (proto_value) ttl (ttl_value) \
1943 vlan (vlan_value) flexbytes (flexbytes_value) \
1944 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
1947 flow_director_filter (port_id) mode IP (add|del|update) \
1948 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
1949 src (src_ip_address) (src_port) \
1950 dst (dst_ip_address) (dst_port) \
1951 tos (tos_value) ttl (ttl_value) \
1952 vlan (vlan_value) flexbytes (flexbytes_value) \
1953 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
1956 flow_director_filter (port_id) mode IP (add|del|update) \
1957 flow (ipv4-sctp|ipv6-sctp) \
1958 src (src_ip_address) (src_port) \
1959 dst (dst_ip_address) (dst_port) \
1960 tos (tos_value) ttl (ttl_value) \
1961 tag (verification_tag) vlan (vlan_value) \
1962 flexbytes (flexbytes_value) (drop|fwd) \
1963 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
1965 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
1966 ether (ethertype) flexbytes (flexbytes_value) \
1967 (drop|fwd) pf|vf(vf_id) queue (queue_id)
1970 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
1971 mac (mac_address) vlan (vlan_value) \
1972 flexbytes (flexbytes_value) (drop|fwd) \
1973 queue (queue_id) fd_id (fd_id_value)
1975 flow_director_filter (port_id) mode Tunnel (add|del|update) \
1976 mac (mac_address) vlan (vlan_value) \
1977 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
1978 flexbytes (flexbytes_value) (drop|fwd) \
1979 queue (queue_id) fd_id (fd_id_value)
1981 For example, to add an ipv4-udp flow type filter::
1983 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
1984 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
1985 fwd pf queue 1 fd_id 1
1987 For example, add an ipv4-other flow type filter::
1989 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
1990 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
1991 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
1996 Flush all flow director filters on a device::
1998 testpmd> flush_flow_director (port_id)
2000 Example, to flush all flow director filter on port 0::
2002 testpmd> flush_flow_director 0
2007 Set flow director's input masks::
2009 flow_director_mask (port_id) mode IP vlan (vlan_value) \
2010 src_mask (ipv4_src) (ipv6_src) (src_port) \
2011 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2013 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2015 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2016 mac (mac_value) tunnel-type (tunnel_type_value) \
2017 tunnel-id (tunnel_id_value)
2019 Example, to set flow director mask on port 0::
2021 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2022 src_mask 255.255.255.255 \
2023 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2024 dst_mask 255.255.255.255 \
2025 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2027 flow_director_flex_mask
2028 ~~~~~~~~~~~~~~~~~~~~~~~
2030 set masks of flow director's flexible payload based on certain flow type::
2032 testpmd> flow_director_flex_mask (port_id) \
2033 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2034 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2035 l2_payload|all) (mask)
2037 Example, to set flow director's flex mask for all flow type on port 0::
2039 testpmd> flow_director_flex_mask 0 flow all \
2040 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2043 flow_director_flex_payload
2044 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2046 Configure flexible payload selection::
2048 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2050 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2052 testpmd> flow_director_flex_payload 0 l4 \
2053 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2055 get_sym_hash_ena_per_port
2056 ~~~~~~~~~~~~~~~~~~~~~~~~~
2058 Get symmetric hash enable configuration per port::
2060 get_sym_hash_ena_per_port (port_id)
2062 For example, to get symmetric hash enable configuration of port 1::
2064 testpmd> get_sym_hash_ena_per_port 1
2066 set_sym_hash_ena_per_port
2067 ~~~~~~~~~~~~~~~~~~~~~~~~~
2069 Set symmetric hash enable configuration per port to enable or disable::
2071 set_sym_hash_ena_per_port (port_id) (enable|disable)
2073 For example, to set symmetric hash enable configuration of port 1 to enable::
2075 testpmd> set_sym_hash_ena_per_port 1 enable
2077 get_hash_global_config
2078 ~~~~~~~~~~~~~~~~~~~~~~
2080 Get the global configurations of hash filters::
2082 get_hash_global_config (port_id)
2084 For example, to get the global configurations of hash filters of port 1::
2086 testpmd> get_hash_global_config 1
2088 set_hash_global_config
2089 ~~~~~~~~~~~~~~~~~~~~~~
2091 Set the global configurations of hash filters::
2093 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2094 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2095 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload) \
2098 For example, to enable simple_xor for flow type of ipv6 on port 2::
2100 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2105 Set the input set for hash::
2107 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2108 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2109 l2_payload) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2110 ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2111 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2112 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2113 fld-8th|none) (select|add)
2115 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2117 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2122 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2123 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2125 Set the input set for flow director::
2127 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2128 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2129 l2_payload) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2130 ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2131 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2132 sctp-dst-port|sctp-veri-tag|none) (select|add)
2134 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2136 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2141 Set different GRE key length for input set::
2143 global_config (port_id) gre-key-len (number in bytes)
2145 For example to set GRE key length for input set to 4 bytes on port 0::
2147 testpmd> global_config 0 gre-key-len 4
2150 .. _testpmd_rte_flow:
2152 Flow rules management
2153 ---------------------
2155 Control of the generic flow API (*rte_flow*) is fully exposed through the
2156 ``flow`` command (validation, creation, destruction and queries).
2158 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2159 features simultaneously may cause undefined side-effects and is therefore
2165 Because the ``flow`` command uses dynamic tokens to handle the large number
2166 of possible flow rules combinations, its behavior differs slightly from
2167 other commands, in particular:
2169 - Pressing *?* or the *<tab>* key displays contextual help for the current
2170 token, not that of the entire command.
2172 - Optional and repeated parameters are supported (provided they are listed
2173 in the contextual help).
2175 The first parameter stands for the operation mode. Possible operations and
2176 their general syntax are described below. They are covered in detail in the
2179 - Check whether a flow rule can be created::
2181 flow validate {port_id}
2182 [group {group_id}] [priority {level}] [ingress] [egress]
2183 pattern {item} [/ {item} [...]] / end
2184 actions {action} [/ {action} [...]] / end
2186 - Create a flow rule::
2188 flow create {port_id}
2189 [group {group_id}] [priority {level}] [ingress] [egress]
2190 pattern {item} [/ {item} [...]] / end
2191 actions {action} [/ {action} [...]] / end
2193 - Destroy specific flow rules::
2195 flow destroy {port_id} rule {rule_id} [...]
2197 - Destroy all flow rules::
2199 flow flush {port_id}
2201 - Query an existing flow rule::
2203 flow query {port_id} {rule_id} {action}
2205 - List existing flow rules sorted by priority, filtered by group
2208 flow list {port_id} [group {group_id}] [...]
2210 Validating flow rules
2211 ~~~~~~~~~~~~~~~~~~~~~
2213 ``flow validate`` reports whether a flow rule would be accepted by the
2214 underlying device in its current state but stops short of creating it. It is
2215 bound to ``rte_flow_validate()``::
2217 flow validate {port_id}
2218 [group {group_id}] [priority {level}] [ingress] [egress]
2219 pattern {item} [/ {item} [...]] / end
2220 actions {action} [/ {action} [...]] / end
2222 If successful, it will show::
2226 Otherwise it will show an error message of the form::
2228 Caught error type [...] ([...]): [...]
2230 This command uses the same parameters as ``flow create``, their format is
2231 described in `Creating flow rules`_.
2233 Check whether redirecting any Ethernet packet received on port 0 to RX queue
2234 index 6 is supported::
2236 testpmd> flow validate 0 ingress pattern eth / end
2237 actions queue index 6 / end
2241 Port 0 does not support TCPv6 rules::
2243 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
2245 Caught error type 9 (specific pattern item): Invalid argument
2251 ``flow create`` validates and creates the specified flow rule. It is bound
2252 to ``rte_flow_create()``::
2254 flow create {port_id}
2255 [group {group_id}] [priority {level}] [ingress] [egress]
2256 pattern {item} [/ {item} [...]] / end
2257 actions {action} [/ {action} [...]] / end
2259 If successful, it will return a flow rule ID usable with other commands::
2261 Flow rule #[...] created
2263 Otherwise it will show an error message of the form::
2265 Caught error type [...] ([...]): [...]
2267 Parameters describe in the following order:
2269 - Attributes (*group*, *priority*, *ingress*, *egress* tokens).
2270 - A matching pattern, starting with the *pattern* token and terminated by an
2272 - Actions, starting with the *actions* token and terminated by an *end*
2275 These translate directly to *rte_flow* objects provided as-is to the
2276 underlying functions.
2278 The shortest valid definition only comprises mandatory tokens::
2280 testpmd> flow create 0 pattern end actions end
2282 Note that PMDs may refuse rules that essentially do nothing such as this
2285 **All unspecified object values are automatically initialized to 0.**
2290 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
2291 specified before the ``pattern`` token.
2293 - ``group {group id}``: priority group.
2294 - ``priority {level}``: priority level within group.
2295 - ``ingress``: rule applies to ingress traffic.
2296 - ``egress``: rule applies to egress traffic.
2298 Each instance of an attribute specified several times overrides the previous
2299 value as shown below (group 4 is used)::
2301 testpmd> flow create 0 group 42 group 24 group 4 [...]
2303 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
2305 While not specifying a direction is an error, some rules may allow both
2308 Most rules affect RX therefore contain the ``ingress`` token::
2310 testpmd> flow create 0 ingress pattern [...]
2315 A matching pattern starts after the ``pattern`` token. It is made of pattern
2316 items and is terminated by a mandatory ``end`` item.
2318 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
2319 rte_flow_item_type``).
2321 The ``/`` token is used as a separator between pattern items as shown
2324 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
2326 Note that protocol items like these must be stacked from lowest to highest
2327 layer to make sense. For instance, the following rule is either invalid or
2328 unlikely to match any packet::
2330 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
2332 More information on these restrictions can be found in the *rte_flow*
2335 Several items support additional specification structures, for example
2336 ``ipv4`` allows specifying source and destination addresses as follows::
2338 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
2339 dst is 10.2.0.0 / end [...]
2341 This rule matches all IPv4 traffic with the specified properties.
2343 In this example, ``src`` and ``dst`` are field names of the underlying
2344 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
2345 in a similar fashion.
2347 The ``is`` token means that the subsequent value must be matched exactly,
2348 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
2349 accordingly. Possible assignment tokens are:
2351 - ``is``: match value perfectly (with full bit-mask).
2352 - ``spec``: match value according to configured bit-mask.
2353 - ``last``: specify upper bound to establish a range.
2354 - ``mask``: specify bit-mask with relevant bits set to one.
2355 - ``prefix``: generate bit-mask from a prefix length.
2357 These yield identical results::
2359 ipv4 src is 10.1.1.1
2363 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
2367 ipv4 src spec 10.1.1.1 src prefix 32
2371 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
2375 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
2377 Inclusive ranges can be defined with ``last``::
2379 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
2381 Note that ``mask`` affects both ``spec`` and ``last``::
2383 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
2384 # matches 10.1.0.0 to 10.2.255.255
2386 Properties can be modified multiple times::
2388 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
2392 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
2397 This section lists supported pattern items and their attributes, if any.
2399 - ``end``: end list of pattern items.
2401 - ``void``: no-op pattern item.
2403 - ``invert``: perform actions when pattern does not match.
2405 - ``any``: match any protocol for the current layer.
2407 - ``num {unsigned}``: number of layers covered.
2409 - ``pf``: match packets addressed to the physical function.
2411 - ``vf``: match packets addressed to a virtual function ID.
2413 - ``id {unsigned}``: destination VF ID.
2415 - ``port``: device-specific physical port index to use.
2417 - ``index {unsigned}``: physical port index.
2419 - ``raw``: match an arbitrary byte string.
2421 - ``relative {boolean}``: look for pattern after the previous item.
2422 - ``search {boolean}``: search pattern from offset (see also limit).
2423 - ``offset {integer}``: absolute or relative offset for pattern.
2424 - ``limit {unsigned}``: search area limit for start of pattern.
2425 - ``pattern {string}``: byte string to look for.
2427 - ``eth``: match Ethernet header.
2429 - ``dst {MAC-48}``: destination MAC.
2430 - ``src {MAC-48}``: source MAC.
2431 - ``type {unsigned}``: EtherType.
2433 - ``vlan``: match 802.1Q/ad VLAN tag.
2435 - ``tpid {unsigned}``: tag protocol identifier.
2436 - ``tci {unsigned}``: tag control information.
2437 - ``pcp {unsigned}``: priority code point.
2438 - ``dei {unsigned}``: drop eligible indicator.
2439 - ``vid {unsigned}``: VLAN identifier.
2441 - ``ipv4``: match IPv4 header.
2443 - ``tos {unsigned}``: type of service.
2444 - ``ttl {unsigned}``: time to live.
2445 - ``proto {unsigned}``: next protocol ID.
2446 - ``src {ipv4 address}``: source address.
2447 - ``dst {ipv4 address}``: destination address.
2449 - ``ipv6``: match IPv6 header.
2451 - ``tc {unsigned}``: traffic class.
2452 - ``flow {unsigned}``: flow label.
2453 - ``proto {unsigned}``: protocol (next header).
2454 - ``hop {unsigned}``: hop limit.
2455 - ``src {ipv6 address}``: source address.
2456 - ``dst {ipv6 address}``: destination address.
2458 - ``icmp``: match ICMP header.
2460 - ``type {unsigned}``: ICMP packet type.
2461 - ``code {unsigned}``: ICMP packet code.
2463 - ``udp``: match UDP header.
2465 - ``src {unsigned}``: UDP source port.
2466 - ``dst {unsigned}``: UDP destination port.
2468 - ``tcp``: match TCP header.
2470 - ``src {unsigned}``: TCP source port.
2471 - ``dst {unsigned}``: TCP destination port.
2473 - ``sctp``: match SCTP header.
2475 - ``src {unsigned}``: SCTP source port.
2476 - ``dst {unsigned}``: SCTP destination port.
2477 - ``tag {unsigned}``: validation tag.
2478 - ``cksum {unsigned}``: checksum.
2480 - ``vxlan``: match VXLAN header.
2482 - ``vni {unsigned}``: VXLAN identifier.
2487 A list of actions starts after the ``actions`` token in the same fashion as
2488 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
2489 terminated by a mandatory ``end`` action.
2491 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
2492 rte_flow_action_type``).
2494 Dropping all incoming UDPv4 packets can be expressed as follows::
2496 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2499 Several actions have configurable properties which must be specified when
2500 there is no valid default value. For example, ``queue`` requires a target
2503 This rule redirects incoming UDPv4 traffic to queue index 6::
2505 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2506 actions queue index 6 / end
2508 While this one could be rejected by PMDs (unspecified queue index)::
2510 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2513 As defined by *rte_flow*, the list is not ordered, all actions of a given
2514 rule are performed simultaneously. These are equivalent::
2516 queue index 6 / void / mark id 42 / end
2520 void / mark id 42 / queue index 6 / end
2522 All actions in a list should have different types, otherwise only the last
2523 action of a given type is taken into account::
2525 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
2529 drop / drop / drop / end # drop is performed only once
2533 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
2535 Considering they are performed simultaneously, opposite and overlapping
2536 actions can sometimes be combined when the end result is unambiguous::
2538 drop / queue index 6 / end # drop has no effect
2542 drop / dup index 6 / end # same as above
2546 queue index 6 / rss queues 6 7 8 / end # queue has no effect
2550 drop / passthru / end # drop has no effect
2552 Note that PMDs may still refuse such combinations.
2557 This section lists supported actions and their attributes, if any.
2559 - ``end``: end list of actions.
2561 - ``void``: no-op action.
2563 - ``passthru``: let subsequent rule process matched packets.
2565 - ``mark``: attach 32 bit value to packets.
2567 - ``id {unsigned}``: 32 bit value to return with packets.
2569 - ``flag``: flag packets.
2571 - ``queue``: assign packets to a given queue index.
2573 - ``index {unsigned}``: queue index to use.
2575 - ``drop``: drop packets (note: passthru has priority).
2577 - ``count``: enable counters for this rule.
2579 - ``dup``: duplicate packets to a given queue index.
2581 - ``index {unsigned}``: queue index to duplicate packets to.
2583 - ``rss``: spread packets among several queues.
2585 - ``queues [{unsigned} [...]] end``: queue indices to use.
2587 - ``pf``: redirect packets to physical device function.
2589 - ``vf``: redirect packets to virtual device function.
2591 - ``original {boolean}``: use original VF ID if possible.
2592 - ``id {unsigned}``: VF ID to redirect packets to.
2594 Destroying flow rules
2595 ~~~~~~~~~~~~~~~~~~~~~
2597 ``flow destroy`` destroys one or more rules from their rule ID (as returned
2598 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
2599 times as necessary::
2601 flow destroy {port_id} rule {rule_id} [...]
2603 If successful, it will show::
2605 Flow rule #[...] destroyed
2607 It does not report anything for rule IDs that do not exist. The usual error
2608 message is shown when a rule cannot be destroyed::
2610 Caught error type [...] ([...]): [...]
2612 ``flow flush`` destroys all rules on a device and does not take extra
2613 arguments. It is bound to ``rte_flow_flush()``::
2615 flow flush {port_id}
2617 Any errors are reported as above.
2619 Creating several rules and destroying them::
2621 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2622 actions queue index 2 / end
2623 Flow rule #0 created
2624 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2625 actions queue index 3 / end
2626 Flow rule #1 created
2627 testpmd> flow destroy 0 rule 0 rule 1
2628 Flow rule #1 destroyed
2629 Flow rule #0 destroyed
2632 The same result can be achieved using ``flow flush``::
2634 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2635 actions queue index 2 / end
2636 Flow rule #0 created
2637 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2638 actions queue index 3 / end
2639 Flow rule #1 created
2640 testpmd> flow flush 0
2643 Non-existent rule IDs are ignored::
2645 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2646 actions queue index 2 / end
2647 Flow rule #0 created
2648 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2649 actions queue index 3 / end
2650 Flow rule #1 created
2651 testpmd> flow destroy 0 rule 42 rule 10 rule 2
2653 testpmd> flow destroy 0 rule 0
2654 Flow rule #0 destroyed
2660 ``flow query`` queries a specific action of a flow rule having that
2661 ability. Such actions collect information that can be reported using this
2662 command. It is bound to ``rte_flow_query()``::
2664 flow query {port_id} {rule_id} {action}
2666 If successful, it will display either the retrieved data for known actions
2667 or the following message::
2669 Cannot display result for action type [...] ([...])
2671 Otherwise, it will complain either that the rule does not exist or that some
2674 Flow rule #[...] not found
2678 Caught error type [...] ([...]): [...]
2680 Currently only the ``count`` action is supported. This action reports the
2681 number of packets that hit the flow rule and the total number of bytes. Its
2682 output has the following format::
2685 hits_set: [...] # whether "hits" contains a valid value
2686 bytes_set: [...] # whether "bytes" contains a valid value
2687 hits: [...] # number of packets
2688 bytes: [...] # number of bytes
2690 Querying counters for TCPv6 packets redirected to queue 6::
2692 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
2693 actions queue index 6 / count / end
2694 Flow rule #4 created
2695 testpmd> flow query 0 4 count
2706 ``flow list`` lists existing flow rules sorted by priority and optionally
2707 filtered by group identifiers::
2709 flow list {port_id} [group {group_id}] [...]
2711 This command only fails with the following message if the device does not
2716 Output consists of a header line followed by a short description of each
2717 flow rule, one per line. There is no output at all when no flow rules are
2718 configured on the device::
2720 ID Group Prio Attr Rule
2721 [...] [...] [...] [...] [...]
2723 ``Attr`` column flags:
2725 - ``i`` for ``ingress``.
2726 - ``e`` for ``egress``.
2728 Creating several flow rules and listing them::
2730 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2731 actions queue index 6 / end
2732 Flow rule #0 created
2733 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2734 actions queue index 2 / end
2735 Flow rule #1 created
2736 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
2737 actions rss queues 6 7 8 end / end
2738 Flow rule #2 created
2739 testpmd> flow list 0
2740 ID Group Prio Attr Rule
2741 0 0 0 i- ETH IPV4 => QUEUE
2742 1 0 0 i- ETH IPV6 => QUEUE
2743 2 0 5 i- ETH IPV4 UDP => RSS
2746 Rules are sorted by priority (i.e. group ID first, then priority level)::
2748 testpmd> flow list 1
2749 ID Group Prio Attr Rule
2750 0 0 0 i- ETH => COUNT
2751 6 0 500 i- ETH IPV6 TCP => DROP COUNT
2752 5 0 1000 i- ETH IPV6 ICMP => QUEUE
2753 1 24 0 i- ETH IPV4 UDP => QUEUE
2754 4 24 10 i- ETH IPV4 TCP => DROP
2755 3 24 20 i- ETH IPV4 => DROP
2756 2 24 42 i- ETH IPV4 UDP => QUEUE
2757 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
2760 Output can be limited to specific groups::
2762 testpmd> flow list 1 group 0 group 63
2763 ID Group Prio Attr Rule
2764 0 0 0 i- ETH => COUNT
2765 6 0 500 i- ETH IPV6 TCP => DROP COUNT
2766 5 0 1000 i- ETH IPV6 ICMP => QUEUE
2767 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE