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)
920 Set the link flow control parameter on a port::
922 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
923 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
924 autoneg (on|off) (port_id)
928 * ``high_water`` (integer): High threshold value to trigger XOFF.
930 * ``low_water`` (integer): Low threshold value to trigger XON.
932 * ``pause_time`` (integer): Pause quota in the Pause frame.
934 * ``send_xon`` (0/1): Send XON frame.
936 * ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
938 * ``autoneg``: Change the auto-negotiation parameter.
943 Set the priority flow control parameter on a port::
945 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
946 (pause_time) (priority) (port_id)
950 * ``high_water`` (integer): High threshold value.
952 * ``low_water`` (integer): Low threshold value.
954 * ``pause_time`` (integer): Pause quota in the Pause frame.
956 * ``priority`` (0-7): VLAN User Priority.
961 Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
963 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
965 For example, to set rx queue 2 on port 0 to mapping 5::
967 testpmd>set stat_qmap rx 0 2 5
969 set port - rx/tx (for VF)
970 ~~~~~~~~~~~~~~~~~~~~~~~~~
972 Set VF receive/transmit from a port::
974 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
976 set port - mac address filter (for VF)
977 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
979 Add/Remove unicast or multicast MAC addr filter for a VF::
981 testpmd> set port (port_id) vf (vf_id) (mac_addr) \
982 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
984 set port - rx mode(for VF)
985 ~~~~~~~~~~~~~~~~~~~~~~~~~~
987 Set the VF receive mode of a port::
989 testpmd> set port (port_id) vf (vf_id) \
990 rxmode (AUPE|ROPE|BAM|MPE) (on|off)
992 The available receive modes are:
994 * ``AUPE``: Accepts untagged VLAN.
996 * ``ROPE``: Accepts unicast hash.
998 * ``BAM``: Accepts broadcast packets.
1000 * ``MPE``: Accepts all multicast packets.
1002 set port - tx_rate (for Queue)
1003 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1005 Set TX rate limitation for a queue on a port::
1007 testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1009 set port - tx_rate (for VF)
1010 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
1012 Set TX rate limitation for queues in VF on a port::
1014 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1016 set port - mirror rule
1017 ~~~~~~~~~~~~~~~~~~~~~~
1019 Set pool or vlan type mirror rule for a port::
1021 testpmd> set port (port_id) mirror-rule (rule_id) \
1022 (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1023 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1025 Set link mirror rule for a port::
1027 testpmd> set port (port_id) mirror-rule (rule_id) \
1028 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1030 For example to enable mirror traffic with vlan 0,1 to pool 0::
1032 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1034 reset port - mirror rule
1035 ~~~~~~~~~~~~~~~~~~~~~~~~
1037 Reset a mirror rule for a port::
1039 testpmd> reset port (port_id) mirror-rule (rule_id)
1044 Set the flush on RX streams before forwarding.
1045 The default is flush ``on``.
1046 Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1048 testpmd> set flush_rx off
1053 Set the bypass mode for the lowest port on bypass enabled NIC::
1055 testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1060 Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1062 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1063 mode (normal|bypass|isolate) (port_id)
1067 * ``timeout``: Enable bypass after watchdog timeout.
1069 * ``os_on``: Enable bypass when OS/board is powered on.
1071 * ``os_off``: Enable bypass when OS/board is powered off.
1073 * ``power_on``: Enable bypass when power supply is turned on.
1075 * ``power_off``: Enable bypass when power supply is turned off.
1081 Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1083 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1088 Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1090 testpmd> show bypass config (port_id)
1095 Set link up for a port::
1097 testpmd> set link-up port (port id)
1102 Set link down for a port::
1104 testpmd> set link-down port (port id)
1109 Enable E-tag insertion for a VF on a port::
1111 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1113 Disable E-tag insertion for a VF on a port::
1115 testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1117 Enable/disable E-tag stripping on a port::
1119 testpmd> E-tag set stripping (on|off) port (port_id)
1121 Enable/disable E-tag based forwarding on a port::
1123 testpmd> E-tag set forwarding (on|off) port (port_id)
1125 Add an E-tag forwarding filter on a port::
1127 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1129 Delete an E-tag forwarding filter on a port::
1130 testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1136 The following sections show functions for configuring ports.
1140 Port configuration changes only become active when forwarding is started/restarted.
1145 Attach a port specified by pci address or virtual device args::
1147 testpmd> port attach (identifier)
1149 To attach a new pci device, the device should be recognized by kernel first.
1150 Then it should be moved under DPDK management.
1151 Finally the port can be attached to testpmd.
1153 For example, to move a pci device using ixgbe under DPDK management:
1155 .. code-block:: console
1157 # Check the status of the available devices.
1158 ./usertools/dpdk-devbind.py --status
1160 Network devices using DPDK-compatible driver
1161 ============================================
1164 Network devices using kernel driver
1165 ===================================
1166 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1169 # Bind the device to igb_uio.
1170 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1173 # Recheck the status of the devices.
1174 ./usertools/dpdk-devbind.py --status
1175 Network devices using DPDK-compatible driver
1176 ============================================
1177 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1179 To attach a port created by virtual device, above steps are not needed.
1181 For example, to attach a port whose pci address is 0000:0a:00.0.
1183 .. code-block:: console
1185 testpmd> port attach 0000:0a:00.0
1186 Attaching a new port...
1187 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1188 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1189 EAL: PCI memory mapped at 0x7f83bfa00000
1190 EAL: PCI memory mapped at 0x7f83bfa80000
1191 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1192 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1193 Port 0 is attached. Now total ports is 1
1196 For example, to attach a port created by pcap PMD.
1198 .. code-block:: console
1200 testpmd> port attach net_pcap0
1201 Attaching a new port...
1202 PMD: Initializing pmd_pcap for net_pcap0
1203 PMD: Creating pcap-backed ethdev on numa socket 0
1204 Port 0 is attached. Now total ports is 1
1207 In this case, identifier is ``net_pcap0``.
1208 This identifier format is the same as ``--vdev`` format of DPDK applications.
1210 For example, to re-attach a bonded port which has been previously detached,
1211 the mode and slave parameters must be given.
1213 .. code-block:: console
1215 testpmd> port attach net_bond_0,mode=0,slave=1
1216 Attaching a new port...
1217 EAL: Initializing pmd_bond for net_bond_0
1218 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1219 Port 0 is attached. Now total ports is 1
1226 Detach a specific port::
1228 testpmd> port detach (port_id)
1230 Before detaching a port, the port should be stopped and closed.
1232 For example, to detach a pci device port 0.
1234 .. code-block:: console
1236 testpmd> port stop 0
1239 testpmd> port close 0
1243 testpmd> port detach 0
1245 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1246 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1247 EAL: PCI memory unmapped at 0x7f83bfa00000
1248 EAL: PCI memory unmapped at 0x7f83bfa80000
1252 For example, to detach a virtual device port 0.
1254 .. code-block:: console
1256 testpmd> port stop 0
1259 testpmd> port close 0
1263 testpmd> port detach 0
1265 PMD: Closing pcap ethdev on numa socket 0
1266 Port 'net_pcap0' is detached. Now total ports is 0
1269 To remove a pci device completely from the system, first detach the port from testpmd.
1270 Then the device should be moved under kernel management.
1271 Finally the device can be removed using kernel pci hotplug functionality.
1273 For example, to move a pci device under kernel management:
1275 .. code-block:: console
1277 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1279 ./usertools/dpdk-devbind.py --status
1281 Network devices using DPDK-compatible driver
1282 ============================================
1285 Network devices using kernel driver
1286 ===================================
1287 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1289 To remove a port created by a virtual device, above steps are not needed.
1294 Start all ports or a specific port::
1296 testpmd> port start (port_id|all)
1301 Stop all ports or a specific port::
1303 testpmd> port stop (port_id|all)
1308 Close all ports or a specific port::
1310 testpmd> port close (port_id|all)
1312 port start/stop queue
1313 ~~~~~~~~~~~~~~~~~~~~~
1315 Start/stop a rx/tx queue on a specific port::
1317 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1319 Only take effect when port is started.
1324 Set the speed and duplex mode for all ports or a specific port::
1326 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1327 duplex (half|full|auto)
1329 port config - queues/descriptors
1330 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1332 Set number of queues/descriptors for rxq, txq, rxd and txd::
1334 testpmd> port config all (rxq|txq|rxd|txd) (value)
1336 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1338 port config - max-pkt-len
1339 ~~~~~~~~~~~~~~~~~~~~~~~~~
1341 Set the maximum packet length::
1343 testpmd> port config all max-pkt-len (value)
1345 This is equivalent to the ``--max-pkt-len`` command-line option.
1347 port config - CRC Strip
1348 ~~~~~~~~~~~~~~~~~~~~~~~
1350 Set hardware CRC stripping on or off for all ports::
1352 testpmd> port config all crc-strip (on|off)
1354 CRC stripping is off by default.
1356 The ``on`` option is equivalent to the ``--crc-strip`` command-line option.
1358 port config - scatter
1359 ~~~~~~~~~~~~~~~~~~~~~~~
1361 Set RX scatter mode on or off for all ports::
1363 testpmd> port config all scatter (on|off)
1365 RX scatter mode is off by default.
1367 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1369 port config - TX queue flags
1370 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1372 Set a hexadecimal bitmap of TX queue flags for all ports::
1374 testpmd> port config all txqflags value
1376 This command is equivalent to the ``--txqflags`` command-line option.
1378 port config - RX Checksum
1379 ~~~~~~~~~~~~~~~~~~~~~~~~~
1381 Set hardware RX checksum offload to on or off for all ports::
1383 testpmd> port config all rx-cksum (on|off)
1385 Checksum offload is off by default.
1387 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1392 Set hardware VLAN on or off for all ports::
1394 testpmd> port config all hw-vlan (on|off)
1396 Hardware VLAN is on by default.
1398 The ``off`` option is equivalent to the ``--disable-hw-vlan`` command-line option.
1400 port config - VLAN filter
1401 ~~~~~~~~~~~~~~~~~~~~~~~~~
1403 Set hardware VLAN filter on or off for all ports::
1405 testpmd> port config all hw-vlan-filter (on|off)
1407 Hardware VLAN filter is on by default.
1409 The ``off`` option is equivalent to the ``--disable-hw-vlan-filter`` command-line option.
1411 port config - VLAN strip
1412 ~~~~~~~~~~~~~~~~~~~~~~~~
1414 Set hardware VLAN strip on or off for all ports::
1416 testpmd> port config all hw-vlan-strip (on|off)
1418 Hardware VLAN strip is on by default.
1420 The ``off`` option is equivalent to the ``--disable-hw-vlan-strip`` command-line option.
1422 port config - VLAN extend
1423 ~~~~~~~~~~~~~~~~~~~~~~~~~
1425 Set hardware VLAN extend on or off for all ports::
1427 testpmd> port config all hw-vlan-extend (on|off)
1429 Hardware VLAN extend is off by default.
1431 The ``off`` option is equivalent to the ``--disable-hw-vlan-extend`` command-line option.
1433 port config - Drop Packets
1434 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1436 Set packet drop for packets with no descriptors on or off for all ports::
1438 testpmd> port config all drop-en (on|off)
1440 Packet dropping for packets with no descriptors is off by default.
1442 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1447 Set the RSS (Receive Side Scaling) mode on or off::
1449 testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1451 RSS is on by default.
1453 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1455 port config - RSS Reta
1456 ~~~~~~~~~~~~~~~~~~~~~~
1458 Set the RSS (Receive Side Scaling) redirection table::
1460 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1465 Set the DCB mode for an individual port::
1467 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1469 The traffic class should be 4 or 8.
1474 Set the number of packets per burst::
1476 testpmd> port config all burst (value)
1478 This is equivalent to the ``--burst`` command-line option.
1480 port config - Threshold
1481 ~~~~~~~~~~~~~~~~~~~~~~~
1483 Set thresholds for TX/RX queues::
1485 testpmd> port config all (threshold) (value)
1487 Where the threshold type can be:
1489 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1491 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1493 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1495 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1497 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1499 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1501 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1503 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1505 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1507 These threshold options are also available from the command-line.
1512 Set the value of ether-type for E-tag::
1514 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1516 Enable/disable the E-tag support::
1518 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1521 Link Bonding Functions
1522 ----------------------
1524 The Link Bonding functions make it possible to dynamically create and
1525 manage link bonding devices from within testpmd interactive prompt.
1527 create bonded device
1528 ~~~~~~~~~~~~~~~~~~~~
1530 Create a new bonding device::
1532 testpmd> create bonded device (mode) (socket)
1534 For example, to create a bonded device in mode 1 on socket 0::
1536 testpmd> create bonded 1 0
1537 created new bonded device (port X)
1542 Adds Ethernet device to a Link Bonding device::
1544 testpmd> add bonding slave (slave id) (port id)
1546 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
1548 testpmd> add bonding slave 6 10
1551 remove bonding slave
1552 ~~~~~~~~~~~~~~~~~~~~
1554 Removes an Ethernet slave device from a Link Bonding device::
1556 testpmd> remove bonding slave (slave id) (port id)
1558 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
1560 testpmd> remove bonding slave 6 10
1565 Set the Link Bonding mode of a Link Bonding device::
1567 testpmd> set bonding mode (value) (port id)
1569 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
1571 testpmd> set bonding mode 3 10
1576 Set an Ethernet slave device as the primary device on a Link Bonding device::
1578 testpmd> set bonding primary (slave id) (port id)
1580 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
1582 testpmd> set bonding primary 6 10
1587 Set the MAC address of a Link Bonding device::
1589 testpmd> set bonding mac (port id) (mac)
1591 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
1593 testpmd> set bonding mac 10 00:00:00:00:00:01
1595 set bonding xmit_balance_policy
1596 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1598 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
1600 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
1602 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
1604 testpmd> set bonding xmit_balance_policy 10 l34
1607 set bonding mon_period
1608 ~~~~~~~~~~~~~~~~~~~~~~
1610 Set the link status monitoring polling period in milliseconds for a bonding device.
1612 This adds support for PMD slave devices which do not support link status interrupts.
1613 When the mon_period is set to a value greater than 0 then all PMD's which do not support
1614 link status ISR will be queried every polling interval to check if their link status has changed::
1616 testpmd> set bonding mon_period (port_id) (value)
1618 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
1620 testpmd> set bonding mon_period 5 150
1626 Show the current configuration of a Link Bonding device::
1628 testpmd> show bonding config (port id)
1631 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
1632 in balance mode with a transmission policy of layer 2+3::
1634 testpmd> show bonding config 9
1636 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
1638 Active Slaves (3): [1 3 4]
1645 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
1646 This is mainly useful for debugging purposes.
1647 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
1648 and fields that can be accessed.
1653 Display the value of a port register::
1655 testpmd> read reg (port_id) (address)
1657 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
1659 testpmd> read reg 0 0xEE00
1660 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
1665 Display a port register bit field::
1667 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
1669 For example, reading the lowest two bits from the register in the example above::
1671 testpmd> read regfield 0 0xEE00 0 1
1672 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
1677 Display a single port register bit::
1679 testpmd> read regbit (port_id) (address) (bit_x)
1681 For example, reading the lowest bit from the register in the example above::
1683 testpmd> read regbit 0 0xEE00 0
1684 port 0 PCI register at offset 0xEE00: bit 0=1
1689 Set the value of a port register::
1691 testpmd> write reg (port_id) (address) (value)
1693 For example, to clear a register::
1695 testpmd> write reg 0 0xEE00 0x0
1696 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
1701 Set bit field of a port register::
1703 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
1705 For example, writing to the register cleared in the example above::
1707 testpmd> write regfield 0 0xEE00 0 1 2
1708 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
1713 Set single bit value of a port register::
1715 testpmd> write regbit (port_id) (address) (bit_x) (value)
1717 For example, to set the high bit in the register from the example above::
1719 testpmd> write regbit 0 0xEE00 31 1
1720 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
1726 This section details the available filter functions that are available.
1728 Note these functions interface the deprecated legacy filtering framework,
1729 superseded by *rte_flow*. See `Flow rules management`_.
1732 ~~~~~~~~~~~~~~~~~~~~
1734 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
1736 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
1737 ethertype (ether_type) (drop|fwd) queue (queue_id)
1739 The available information parameters are:
1741 * ``port_id``: The port which the Ethertype filter assigned on.
1743 * ``mac_addr``: Compare destination mac address.
1745 * ``mac_ignr``: Ignore destination mac address match.
1747 * ``mac_address``: Destination mac address to match.
1749 * ``ether_type``: The EtherType value want to match,
1750 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
1752 * ``queue_id``: The receive queue associated with this EtherType filter.
1753 It is meaningless when deleting or dropping.
1755 Example, to add/remove an ethertype filter rule::
1757 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
1758 ethertype 0x0806 fwd queue 3
1760 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
1761 ethertype 0x0806 fwd queue 3
1766 Add or delete a 2-tuple filter,
1767 which identifies packets by specific protocol and destination TCP/UDP port
1768 and forwards packets into one of the receive queues::
1770 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
1771 protocol (protocol_value) mask (mask_value) \
1772 tcp_flags (tcp_flags_value) priority (prio_value) \
1775 The available information parameters are:
1777 * ``port_id``: The port which the 2-tuple filter assigned on.
1779 * ``dst_port_value``: Destination port in L4.
1781 * ``protocol_value``: IP L4 protocol.
1783 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
1785 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
1787 * ``prio_value``: Priority of this filter.
1789 * ``queue_id``: The receive queue associated with this 2-tuple filter.
1791 Example, to add/remove an 2tuple filter rule::
1793 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
1794 tcp_flags 0x02 priority 3 queue 3
1796 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
1797 tcp_flags 0x02 priority 3 queue 3
1802 Add or delete a 5-tuple filter,
1803 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
1804 and routes packets into one of the receive queues::
1806 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
1807 (src_address) dst_port (dst_port_value) \
1808 src_port (src_port_value) protocol (protocol_value) \
1809 mask (mask_value) tcp_flags (tcp_flags_value) \
1810 priority (prio_value) queue (queue_id)
1812 The available information parameters are:
1814 * ``port_id``: The port which the 5-tuple filter assigned on.
1816 * ``dst_address``: Destination IP address.
1818 * ``src_address``: Source IP address.
1820 * ``dst_port_value``: TCP/UDP destination port.
1822 * ``src_port_value``: TCP/UDP source port.
1824 * ``protocol_value``: L4 protocol.
1826 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
1828 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
1830 * ``prio_value``: The priority of this filter.
1832 * ``queue_id``: The receive queue associated with this 5-tuple filter.
1834 Example, to add/remove an 5tuple filter rule::
1836 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1837 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1838 flags 0x0 priority 3 queue 3
1840 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1841 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1842 flags 0x0 priority 3 queue 3
1847 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
1849 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
1851 The available information parameters are:
1853 * ``port_id``: The port which the SYN filter assigned on.
1855 * ``high``: This SYN filter has higher priority than other filters.
1857 * ``low``: This SYN filter has lower priority than other filters.
1859 * ``queue_id``: The receive queue associated with this SYN filter
1863 testpmd> syn_filter 0 add priority high queue 3
1868 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
1869 and routed into one of the receive queues::
1871 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
1872 mask (mask_value) priority (prio_value) queue (queue_id)
1874 The available information parameters are:
1876 * ``port_id``: The port which the Flex filter is assigned on.
1878 * ``len_value``: Filter length in bytes, no greater than 128.
1880 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
1882 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
1884 * ``prio_value``: The priority of this filter.
1886 * ``queue_id``: The receive queue associated with this Flex filter.
1890 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
1891 mask 000C priority 3 queue 3
1893 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
1894 mask 000C priority 3 queue 3
1897 .. _testpmd_flow_director:
1899 flow_director_filter
1900 ~~~~~~~~~~~~~~~~~~~~
1902 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
1904 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
1905 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
1907 * Perfect match filters.
1908 The hardware checks a match between the masked fields of the received packets and the programmed filters.
1909 The masked fields are for IP flow.
1911 * Signature filters.
1912 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
1914 * Perfect-mac-vlan 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 MAC VLAN flow.
1918 * Perfect-tunnel match filters.
1919 The hardware checks a match between the masked fields of the received packets and the programmed filters.
1920 The masked fields are for tunnel flow.
1922 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
1923 per flow type and the flexible payload.
1925 The Flow Director can also mask out parts of all of these fields so that filters
1926 are only applied to certain fields or parts of the fields.
1928 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
1930 # Commands to add flow director filters of different flow types::
1932 flow_director_filter (port_id) mode IP (add|del|update) \
1933 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
1934 src (src_ip_address) dst (dst_ip_address) \
1935 tos (tos_value) proto (proto_value) ttl (ttl_value) \
1936 vlan (vlan_value) flexbytes (flexbytes_value) \
1937 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
1940 flow_director_filter (port_id) mode IP (add|del|update) \
1941 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
1942 src (src_ip_address) (src_port) \
1943 dst (dst_ip_address) (dst_port) \
1944 tos (tos_value) ttl (ttl_value) \
1945 vlan (vlan_value) flexbytes (flexbytes_value) \
1946 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
1949 flow_director_filter (port_id) mode IP (add|del|update) \
1950 flow (ipv4-sctp|ipv6-sctp) \
1951 src (src_ip_address) (src_port) \
1952 dst (dst_ip_address) (dst_port) \
1953 tos (tos_value) ttl (ttl_value) \
1954 tag (verification_tag) vlan (vlan_value) \
1955 flexbytes (flexbytes_value) (drop|fwd) \
1956 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
1958 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
1959 ether (ethertype) flexbytes (flexbytes_value) \
1960 (drop|fwd) pf|vf(vf_id) queue (queue_id)
1963 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
1964 mac (mac_address) vlan (vlan_value) \
1965 flexbytes (flexbytes_value) (drop|fwd) \
1966 queue (queue_id) fd_id (fd_id_value)
1968 flow_director_filter (port_id) mode Tunnel (add|del|update) \
1969 mac (mac_address) vlan (vlan_value) \
1970 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
1971 flexbytes (flexbytes_value) (drop|fwd) \
1972 queue (queue_id) fd_id (fd_id_value)
1974 For example, to add an ipv4-udp flow type filter::
1976 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
1977 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
1978 fwd pf queue 1 fd_id 1
1980 For example, add an ipv4-other flow type filter::
1982 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
1983 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
1984 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
1989 Flush all flow director filters on a device::
1991 testpmd> flush_flow_director (port_id)
1993 Example, to flush all flow director filter on port 0::
1995 testpmd> flush_flow_director 0
2000 Set flow director's input masks::
2002 flow_director_mask (port_id) mode IP vlan (vlan_value) \
2003 src_mask (ipv4_src) (ipv6_src) (src_port) \
2004 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2006 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2008 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2009 mac (mac_value) tunnel-type (tunnel_type_value) \
2010 tunnel-id (tunnel_id_value)
2012 Example, to set flow director mask on port 0::
2014 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2015 src_mask 255.255.255.255 \
2016 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2017 dst_mask 255.255.255.255 \
2018 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2020 flow_director_flex_mask
2021 ~~~~~~~~~~~~~~~~~~~~~~~
2023 set masks of flow director's flexible payload based on certain flow type::
2025 testpmd> flow_director_flex_mask (port_id) \
2026 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2027 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2028 l2_payload|all) (mask)
2030 Example, to set flow director's flex mask for all flow type on port 0::
2032 testpmd> flow_director_flex_mask 0 flow all \
2033 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2036 flow_director_flex_payload
2037 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2039 Configure flexible payload selection::
2041 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2043 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2045 testpmd> flow_director_flex_payload 0 l4 \
2046 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2048 get_sym_hash_ena_per_port
2049 ~~~~~~~~~~~~~~~~~~~~~~~~~
2051 Get symmetric hash enable configuration per port::
2053 get_sym_hash_ena_per_port (port_id)
2055 For example, to get symmetric hash enable configuration of port 1::
2057 testpmd> get_sym_hash_ena_per_port 1
2059 set_sym_hash_ena_per_port
2060 ~~~~~~~~~~~~~~~~~~~~~~~~~
2062 Set symmetric hash enable configuration per port to enable or disable::
2064 set_sym_hash_ena_per_port (port_id) (enable|disable)
2066 For example, to set symmetric hash enable configuration of port 1 to enable::
2068 testpmd> set_sym_hash_ena_per_port 1 enable
2070 get_hash_global_config
2071 ~~~~~~~~~~~~~~~~~~~~~~
2073 Get the global configurations of hash filters::
2075 get_hash_global_config (port_id)
2077 For example, to get the global configurations of hash filters of port 1::
2079 testpmd> get_hash_global_config 1
2081 set_hash_global_config
2082 ~~~~~~~~~~~~~~~~~~~~~~
2084 Set the global configurations of hash filters::
2086 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2087 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2088 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload) \
2091 For example, to enable simple_xor for flow type of ipv6 on port 2::
2093 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2098 Set the input set for hash::
2100 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2101 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2102 l2_payload) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2103 ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2104 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2105 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2106 fld-8th|none) (select|add)
2108 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2110 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2115 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2116 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2118 Set the input set for flow director::
2120 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2121 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2122 l2_payload) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2123 ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2124 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2125 sctp-dst-port|sctp-veri-tag|none) (select|add)
2127 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2129 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2134 Set different GRE key length for input set::
2136 global_config (port_id) gre-key-len (number in bytes)
2138 For example to set GRE key length for input set to 4 bytes on port 0::
2140 testpmd> global_config 0 gre-key-len 4
2143 .. _testpmd_rte_flow:
2145 Flow rules management
2146 ---------------------
2148 Control of the generic flow API (*rte_flow*) is fully exposed through the
2149 ``flow`` command (validation, creation, destruction and queries).
2151 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2152 features simultaneously may cause undefined side-effects and is therefore
2158 Because the ``flow`` command uses dynamic tokens to handle the large number
2159 of possible flow rules combinations, its behavior differs slightly from
2160 other commands, in particular:
2162 - Pressing *?* or the *<tab>* key displays contextual help for the current
2163 token, not that of the entire command.
2165 - Optional and repeated parameters are supported (provided they are listed
2166 in the contextual help).
2168 The first parameter stands for the operation mode. Possible operations and
2169 their general syntax are described below. They are covered in detail in the
2172 - Check whether a flow rule can be created::
2174 flow validate {port_id}
2175 [group {group_id}] [priority {level}] [ingress] [egress]
2176 pattern {item} [/ {item} [...]] / end
2177 actions {action} [/ {action} [...]] / end
2179 - Create a flow rule::
2181 flow create {port_id}
2182 [group {group_id}] [priority {level}] [ingress] [egress]
2183 pattern {item} [/ {item} [...]] / end
2184 actions {action} [/ {action} [...]] / end
2186 - Destroy specific flow rules::
2188 flow destroy {port_id} rule {rule_id} [...]
2190 - Destroy all flow rules::
2192 flow flush {port_id}
2194 - Query an existing flow rule::
2196 flow query {port_id} {rule_id} {action}
2198 - List existing flow rules sorted by priority, filtered by group
2201 flow list {port_id} [group {group_id}] [...]
2203 Validating flow rules
2204 ~~~~~~~~~~~~~~~~~~~~~
2206 ``flow validate`` reports whether a flow rule would be accepted by the
2207 underlying device in its current state but stops short of creating it. It is
2208 bound to ``rte_flow_validate()``::
2210 flow validate {port_id}
2211 [group {group_id}] [priority {level}] [ingress] [egress]
2212 pattern {item} [/ {item} [...]] / end
2213 actions {action} [/ {action} [...]] / end
2215 If successful, it will show::
2219 Otherwise it will show an error message of the form::
2221 Caught error type [...] ([...]): [...]
2223 This command uses the same parameters as ``flow create``, their format is
2224 described in `Creating flow rules`_.
2226 Check whether redirecting any Ethernet packet received on port 0 to RX queue
2227 index 6 is supported::
2229 testpmd> flow validate 0 ingress pattern eth / end
2230 actions queue index 6 / end
2234 Port 0 does not support TCPv6 rules::
2236 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
2238 Caught error type 9 (specific pattern item): Invalid argument
2244 ``flow create`` validates and creates the specified flow rule. It is bound
2245 to ``rte_flow_create()``::
2247 flow create {port_id}
2248 [group {group_id}] [priority {level}] [ingress] [egress]
2249 pattern {item} [/ {item} [...]] / end
2250 actions {action} [/ {action} [...]] / end
2252 If successful, it will return a flow rule ID usable with other commands::
2254 Flow rule #[...] created
2256 Otherwise it will show an error message of the form::
2258 Caught error type [...] ([...]): [...]
2260 Parameters describe in the following order:
2262 - Attributes (*group*, *priority*, *ingress*, *egress* tokens).
2263 - A matching pattern, starting with the *pattern* token and terminated by an
2265 - Actions, starting with the *actions* token and terminated by an *end*
2268 These translate directly to *rte_flow* objects provided as-is to the
2269 underlying functions.
2271 The shortest valid definition only comprises mandatory tokens::
2273 testpmd> flow create 0 pattern end actions end
2275 Note that PMDs may refuse rules that essentially do nothing such as this
2278 **All unspecified object values are automatically initialized to 0.**
2283 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
2284 specified before the ``pattern`` token.
2286 - ``group {group id}``: priority group.
2287 - ``priority {level}``: priority level within group.
2288 - ``ingress``: rule applies to ingress traffic.
2289 - ``egress``: rule applies to egress traffic.
2291 Each instance of an attribute specified several times overrides the previous
2292 value as shown below (group 4 is used)::
2294 testpmd> flow create 0 group 42 group 24 group 4 [...]
2296 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
2298 While not specifying a direction is an error, some rules may allow both
2301 Most rules affect RX therefore contain the ``ingress`` token::
2303 testpmd> flow create 0 ingress pattern [...]
2308 A matching pattern starts after the ``pattern`` token. It is made of pattern
2309 items and is terminated by a mandatory ``end`` item.
2311 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
2312 rte_flow_item_type``).
2314 The ``/`` token is used as a separator between pattern items as shown
2317 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
2319 Note that protocol items like these must be stacked from lowest to highest
2320 layer to make sense. For instance, the following rule is either invalid or
2321 unlikely to match any packet::
2323 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
2325 More information on these restrictions can be found in the *rte_flow*
2328 Several items support additional specification structures, for example
2329 ``ipv4`` allows specifying source and destination addresses as follows::
2331 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
2332 dst is 10.2.0.0 / end [...]
2334 This rule matches all IPv4 traffic with the specified properties.
2336 In this example, ``src`` and ``dst`` are field names of the underlying
2337 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
2338 in a similar fashion.
2340 The ``is`` token means that the subsequent value must be matched exactly,
2341 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
2342 accordingly. Possible assignment tokens are:
2344 - ``is``: match value perfectly (with full bit-mask).
2345 - ``spec``: match value according to configured bit-mask.
2346 - ``last``: specify upper bound to establish a range.
2347 - ``mask``: specify bit-mask with relevant bits set to one.
2348 - ``prefix``: generate bit-mask from a prefix length.
2350 These yield identical results::
2352 ipv4 src is 10.1.1.1
2356 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
2360 ipv4 src spec 10.1.1.1 src prefix 32
2364 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
2368 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
2370 Inclusive ranges can be defined with ``last``::
2372 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
2374 Note that ``mask`` affects both ``spec`` and ``last``::
2376 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
2377 # matches 10.1.0.0 to 10.2.255.255
2379 Properties can be modified multiple times::
2381 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
2385 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
2390 This section lists supported pattern items and their attributes, if any.
2392 - ``end``: end list of pattern items.
2394 - ``void``: no-op pattern item.
2396 - ``invert``: perform actions when pattern does not match.
2398 - ``any``: match any protocol for the current layer.
2400 - ``num {unsigned}``: number of layers covered.
2402 - ``pf``: match packets addressed to the physical function.
2404 - ``vf``: match packets addressed to a virtual function ID.
2406 - ``id {unsigned}``: destination VF ID.
2408 - ``port``: device-specific physical port index to use.
2410 - ``index {unsigned}``: physical port index.
2412 - ``raw``: match an arbitrary byte string.
2414 - ``relative {boolean}``: look for pattern after the previous item.
2415 - ``search {boolean}``: search pattern from offset (see also limit).
2416 - ``offset {integer}``: absolute or relative offset for pattern.
2417 - ``limit {unsigned}``: search area limit for start of pattern.
2418 - ``pattern {string}``: byte string to look for.
2420 - ``eth``: match Ethernet header.
2422 - ``dst {MAC-48}``: destination MAC.
2423 - ``src {MAC-48}``: source MAC.
2424 - ``type {unsigned}``: EtherType.
2426 - ``vlan``: match 802.1Q/ad VLAN tag.
2428 - ``tpid {unsigned}``: tag protocol identifier.
2429 - ``tci {unsigned}``: tag control information.
2430 - ``pcp {unsigned}``: priority code point.
2431 - ``dei {unsigned}``: drop eligible indicator.
2432 - ``vid {unsigned}``: VLAN identifier.
2434 - ``ipv4``: match IPv4 header.
2436 - ``tos {unsigned}``: type of service.
2437 - ``ttl {unsigned}``: time to live.
2438 - ``proto {unsigned}``: next protocol ID.
2439 - ``src {ipv4 address}``: source address.
2440 - ``dst {ipv4 address}``: destination address.
2442 - ``ipv6``: match IPv6 header.
2444 - ``tc {unsigned}``: traffic class.
2445 - ``flow {unsigned}``: flow label.
2446 - ``proto {unsigned}``: protocol (next header).
2447 - ``hop {unsigned}``: hop limit.
2448 - ``src {ipv6 address}``: source address.
2449 - ``dst {ipv6 address}``: destination address.
2451 - ``icmp``: match ICMP header.
2453 - ``type {unsigned}``: ICMP packet type.
2454 - ``code {unsigned}``: ICMP packet code.
2456 - ``udp``: match UDP header.
2458 - ``src {unsigned}``: UDP source port.
2459 - ``dst {unsigned}``: UDP destination port.
2461 - ``tcp``: match TCP header.
2463 - ``src {unsigned}``: TCP source port.
2464 - ``dst {unsigned}``: TCP destination port.
2466 - ``sctp``: match SCTP header.
2468 - ``src {unsigned}``: SCTP source port.
2469 - ``dst {unsigned}``: SCTP destination port.
2470 - ``tag {unsigned}``: validation tag.
2471 - ``cksum {unsigned}``: checksum.
2473 - ``vxlan``: match VXLAN header.
2475 - ``vni {unsigned}``: VXLAN identifier.
2480 A list of actions starts after the ``actions`` token in the same fashion as
2481 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
2482 terminated by a mandatory ``end`` action.
2484 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
2485 rte_flow_action_type``).
2487 Dropping all incoming UDPv4 packets can be expressed as follows::
2489 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2492 Several actions have configurable properties which must be specified when
2493 there is no valid default value. For example, ``queue`` requires a target
2496 This rule redirects incoming UDPv4 traffic to queue index 6::
2498 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2499 actions queue index 6 / end
2501 While this one could be rejected by PMDs (unspecified queue index)::
2503 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2506 As defined by *rte_flow*, the list is not ordered, all actions of a given
2507 rule are performed simultaneously. These are equivalent::
2509 queue index 6 / void / mark id 42 / end
2513 void / mark id 42 / queue index 6 / end
2515 All actions in a list should have different types, otherwise only the last
2516 action of a given type is taken into account::
2518 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
2522 drop / drop / drop / end # drop is performed only once
2526 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
2528 Considering they are performed simultaneously, opposite and overlapping
2529 actions can sometimes be combined when the end result is unambiguous::
2531 drop / queue index 6 / end # drop has no effect
2535 drop / dup index 6 / end # same as above
2539 queue index 6 / rss queues 6 7 8 / end # queue has no effect
2543 drop / passthru / end # drop has no effect
2545 Note that PMDs may still refuse such combinations.
2550 This section lists supported actions and their attributes, if any.
2552 - ``end``: end list of actions.
2554 - ``void``: no-op action.
2556 - ``passthru``: let subsequent rule process matched packets.
2558 - ``mark``: attach 32 bit value to packets.
2560 - ``id {unsigned}``: 32 bit value to return with packets.
2562 - ``flag``: flag packets.
2564 - ``queue``: assign packets to a given queue index.
2566 - ``index {unsigned}``: queue index to use.
2568 - ``drop``: drop packets (note: passthru has priority).
2570 - ``count``: enable counters for this rule.
2572 - ``dup``: duplicate packets to a given queue index.
2574 - ``index {unsigned}``: queue index to duplicate packets to.
2576 - ``rss``: spread packets among several queues.
2578 - ``queues [{unsigned} [...]] end``: queue indices to use.
2580 - ``pf``: redirect packets to physical device function.
2582 - ``vf``: redirect packets to virtual device function.
2584 - ``original {boolean}``: use original VF ID if possible.
2585 - ``id {unsigned}``: VF ID to redirect packets to.
2587 Destroying flow rules
2588 ~~~~~~~~~~~~~~~~~~~~~
2590 ``flow destroy`` destroys one or more rules from their rule ID (as returned
2591 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
2592 times as necessary::
2594 flow destroy {port_id} rule {rule_id} [...]
2596 If successful, it will show::
2598 Flow rule #[...] destroyed
2600 It does not report anything for rule IDs that do not exist. The usual error
2601 message is shown when a rule cannot be destroyed::
2603 Caught error type [...] ([...]): [...]
2605 ``flow flush`` destroys all rules on a device and does not take extra
2606 arguments. It is bound to ``rte_flow_flush()``::
2608 flow flush {port_id}
2610 Any errors are reported as above.
2612 Creating several rules and destroying them::
2614 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2615 actions queue index 2 / end
2616 Flow rule #0 created
2617 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2618 actions queue index 3 / end
2619 Flow rule #1 created
2620 testpmd> flow destroy 0 rule 0 rule 1
2621 Flow rule #1 destroyed
2622 Flow rule #0 destroyed
2625 The same result can be achieved using ``flow flush``::
2627 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2628 actions queue index 2 / end
2629 Flow rule #0 created
2630 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2631 actions queue index 3 / end
2632 Flow rule #1 created
2633 testpmd> flow flush 0
2636 Non-existent rule IDs are ignored::
2638 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2639 actions queue index 2 / end
2640 Flow rule #0 created
2641 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2642 actions queue index 3 / end
2643 Flow rule #1 created
2644 testpmd> flow destroy 0 rule 42 rule 10 rule 2
2646 testpmd> flow destroy 0 rule 0
2647 Flow rule #0 destroyed
2653 ``flow query`` queries a specific action of a flow rule having that
2654 ability. Such actions collect information that can be reported using this
2655 command. It is bound to ``rte_flow_query()``::
2657 flow query {port_id} {rule_id} {action}
2659 If successful, it will display either the retrieved data for known actions
2660 or the following message::
2662 Cannot display result for action type [...] ([...])
2664 Otherwise, it will complain either that the rule does not exist or that some
2667 Flow rule #[...] not found
2671 Caught error type [...] ([...]): [...]
2673 Currently only the ``count`` action is supported. This action reports the
2674 number of packets that hit the flow rule and the total number of bytes. Its
2675 output has the following format::
2678 hits_set: [...] # whether "hits" contains a valid value
2679 bytes_set: [...] # whether "bytes" contains a valid value
2680 hits: [...] # number of packets
2681 bytes: [...] # number of bytes
2683 Querying counters for TCPv6 packets redirected to queue 6::
2685 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
2686 actions queue index 6 / count / end
2687 Flow rule #4 created
2688 testpmd> flow query 0 4 count
2699 ``flow list`` lists existing flow rules sorted by priority and optionally
2700 filtered by group identifiers::
2702 flow list {port_id} [group {group_id}] [...]
2704 This command only fails with the following message if the device does not
2709 Output consists of a header line followed by a short description of each
2710 flow rule, one per line. There is no output at all when no flow rules are
2711 configured on the device::
2713 ID Group Prio Attr Rule
2714 [...] [...] [...] [...] [...]
2716 ``Attr`` column flags:
2718 - ``i`` for ``ingress``.
2719 - ``e`` for ``egress``.
2721 Creating several flow rules and listing them::
2723 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2724 actions queue index 6 / end
2725 Flow rule #0 created
2726 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2727 actions queue index 2 / end
2728 Flow rule #1 created
2729 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
2730 actions rss queues 6 7 8 end / end
2731 Flow rule #2 created
2732 testpmd> flow list 0
2733 ID Group Prio Attr Rule
2734 0 0 0 i- ETH IPV4 => QUEUE
2735 1 0 0 i- ETH IPV6 => QUEUE
2736 2 0 5 i- ETH IPV4 UDP => RSS
2739 Rules are sorted by priority (i.e. group ID first, then priority level)::
2741 testpmd> flow list 1
2742 ID Group Prio Attr Rule
2743 0 0 0 i- ETH => COUNT
2744 6 0 500 i- ETH IPV6 TCP => DROP COUNT
2745 5 0 1000 i- ETH IPV6 ICMP => QUEUE
2746 1 24 0 i- ETH IPV4 UDP => QUEUE
2747 4 24 10 i- ETH IPV4 TCP => DROP
2748 3 24 20 i- ETH IPV4 => DROP
2749 2 24 42 i- ETH IPV4 UDP => QUEUE
2750 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
2753 Output can be limited to specific groups::
2755 testpmd> flow list 1 group 0 group 63
2756 ID Group Prio Attr Rule
2757 0 0 0 i- ETH => COUNT
2758 6 0 500 i- ETH IPV6 TCP => DROP COUNT
2759 5 0 1000 i- ETH IPV6 ICMP => QUEUE
2760 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE