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)
924 set tc tx min bandwidth
925 ~~~~~~~~~~~~~~~~~~~~~~~
927 Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs::
929 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...)
934 Set the link flow control parameter on a port::
936 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
937 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
938 autoneg (on|off) (port_id)
942 * ``high_water`` (integer): High threshold value to trigger XOFF.
944 * ``low_water`` (integer): Low threshold value to trigger XON.
946 * ``pause_time`` (integer): Pause quota in the Pause frame.
948 * ``send_xon`` (0/1): Send XON frame.
950 * ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
952 * ``autoneg``: Change the auto-negotiation parameter.
957 Set the priority flow control parameter on a port::
959 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
960 (pause_time) (priority) (port_id)
964 * ``high_water`` (integer): High threshold value.
966 * ``low_water`` (integer): Low threshold value.
968 * ``pause_time`` (integer): Pause quota in the Pause frame.
970 * ``priority`` (0-7): VLAN User Priority.
975 Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
977 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
979 For example, to set rx queue 2 on port 0 to mapping 5::
981 testpmd>set stat_qmap rx 0 2 5
983 set port - rx/tx (for VF)
984 ~~~~~~~~~~~~~~~~~~~~~~~~~
986 Set VF receive/transmit from a port::
988 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
990 set port - mac address filter (for VF)
991 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
993 Add/Remove unicast or multicast MAC addr filter for a VF::
995 testpmd> set port (port_id) vf (vf_id) (mac_addr) \
996 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
998 set port - rx mode(for VF)
999 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1001 Set the VF receive mode of a port::
1003 testpmd> set port (port_id) vf (vf_id) \
1004 rxmode (AUPE|ROPE|BAM|MPE) (on|off)
1006 The available receive modes are:
1008 * ``AUPE``: Accepts untagged VLAN.
1010 * ``ROPE``: Accepts unicast hash.
1012 * ``BAM``: Accepts broadcast packets.
1014 * ``MPE``: Accepts all multicast packets.
1016 set port - tx_rate (for Queue)
1017 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1019 Set TX rate limitation for a queue on a port::
1021 testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1023 set port - tx_rate (for VF)
1024 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
1026 Set TX rate limitation for queues in VF on a port::
1028 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1030 set port - mirror rule
1031 ~~~~~~~~~~~~~~~~~~~~~~
1033 Set pool or vlan type mirror rule for a port::
1035 testpmd> set port (port_id) mirror-rule (rule_id) \
1036 (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1037 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1039 Set link mirror rule for a port::
1041 testpmd> set port (port_id) mirror-rule (rule_id) \
1042 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1044 For example to enable mirror traffic with vlan 0,1 to pool 0::
1046 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1048 reset port - mirror rule
1049 ~~~~~~~~~~~~~~~~~~~~~~~~
1051 Reset a mirror rule for a port::
1053 testpmd> reset port (port_id) mirror-rule (rule_id)
1058 Set the flush on RX streams before forwarding.
1059 The default is flush ``on``.
1060 Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1062 testpmd> set flush_rx off
1067 Set the bypass mode for the lowest port on bypass enabled NIC::
1069 testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1074 Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1076 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1077 mode (normal|bypass|isolate) (port_id)
1081 * ``timeout``: Enable bypass after watchdog timeout.
1083 * ``os_on``: Enable bypass when OS/board is powered on.
1085 * ``os_off``: Enable bypass when OS/board is powered off.
1087 * ``power_on``: Enable bypass when power supply is turned on.
1089 * ``power_off``: Enable bypass when power supply is turned off.
1095 Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1097 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1102 Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1104 testpmd> show bypass config (port_id)
1109 Set link up for a port::
1111 testpmd> set link-up port (port id)
1116 Set link down for a port::
1118 testpmd> set link-down port (port id)
1123 Enable E-tag insertion for a VF on a port::
1125 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1127 Disable E-tag insertion for a VF on a port::
1129 testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1131 Enable/disable E-tag stripping on a port::
1133 testpmd> E-tag set stripping (on|off) port (port_id)
1135 Enable/disable E-tag based forwarding on a port::
1137 testpmd> E-tag set forwarding (on|off) port (port_id)
1139 Add an E-tag forwarding filter on a port::
1141 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1143 Delete an E-tag forwarding filter on a port::
1144 testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1150 The following sections show functions for configuring ports.
1154 Port configuration changes only become active when forwarding is started/restarted.
1159 Attach a port specified by pci address or virtual device args::
1161 testpmd> port attach (identifier)
1163 To attach a new pci device, the device should be recognized by kernel first.
1164 Then it should be moved under DPDK management.
1165 Finally the port can be attached to testpmd.
1167 For example, to move a pci device using ixgbe under DPDK management:
1169 .. code-block:: console
1171 # Check the status of the available devices.
1172 ./usertools/dpdk-devbind.py --status
1174 Network devices using DPDK-compatible driver
1175 ============================================
1178 Network devices using kernel driver
1179 ===================================
1180 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1183 # Bind the device to igb_uio.
1184 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1187 # Recheck the status of the devices.
1188 ./usertools/dpdk-devbind.py --status
1189 Network devices using DPDK-compatible driver
1190 ============================================
1191 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1193 To attach a port created by virtual device, above steps are not needed.
1195 For example, to attach a port whose pci address is 0000:0a:00.0.
1197 .. code-block:: console
1199 testpmd> port attach 0000:0a:00.0
1200 Attaching a new port...
1201 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1202 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1203 EAL: PCI memory mapped at 0x7f83bfa00000
1204 EAL: PCI memory mapped at 0x7f83bfa80000
1205 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1206 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1207 Port 0 is attached. Now total ports is 1
1210 For example, to attach a port created by pcap PMD.
1212 .. code-block:: console
1214 testpmd> port attach net_pcap0
1215 Attaching a new port...
1216 PMD: Initializing pmd_pcap for net_pcap0
1217 PMD: Creating pcap-backed ethdev on numa socket 0
1218 Port 0 is attached. Now total ports is 1
1221 In this case, identifier is ``net_pcap0``.
1222 This identifier format is the same as ``--vdev`` format of DPDK applications.
1224 For example, to re-attach a bonded port which has been previously detached,
1225 the mode and slave parameters must be given.
1227 .. code-block:: console
1229 testpmd> port attach net_bond_0,mode=0,slave=1
1230 Attaching a new port...
1231 EAL: Initializing pmd_bond for net_bond_0
1232 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1233 Port 0 is attached. Now total ports is 1
1240 Detach a specific port::
1242 testpmd> port detach (port_id)
1244 Before detaching a port, the port should be stopped and closed.
1246 For example, to detach a pci device port 0.
1248 .. code-block:: console
1250 testpmd> port stop 0
1253 testpmd> port close 0
1257 testpmd> port detach 0
1259 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1260 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1261 EAL: PCI memory unmapped at 0x7f83bfa00000
1262 EAL: PCI memory unmapped at 0x7f83bfa80000
1266 For example, to detach a virtual device port 0.
1268 .. code-block:: console
1270 testpmd> port stop 0
1273 testpmd> port close 0
1277 testpmd> port detach 0
1279 PMD: Closing pcap ethdev on numa socket 0
1280 Port 'net_pcap0' is detached. Now total ports is 0
1283 To remove a pci device completely from the system, first detach the port from testpmd.
1284 Then the device should be moved under kernel management.
1285 Finally the device can be removed using kernel pci hotplug functionality.
1287 For example, to move a pci device under kernel management:
1289 .. code-block:: console
1291 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1293 ./usertools/dpdk-devbind.py --status
1295 Network devices using DPDK-compatible driver
1296 ============================================
1299 Network devices using kernel driver
1300 ===================================
1301 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1303 To remove a port created by a virtual device, above steps are not needed.
1308 Start all ports or a specific port::
1310 testpmd> port start (port_id|all)
1315 Stop all ports or a specific port::
1317 testpmd> port stop (port_id|all)
1322 Close all ports or a specific port::
1324 testpmd> port close (port_id|all)
1326 port start/stop queue
1327 ~~~~~~~~~~~~~~~~~~~~~
1329 Start/stop a rx/tx queue on a specific port::
1331 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1333 Only take effect when port is started.
1338 Set the speed and duplex mode for all ports or a specific port::
1340 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1341 duplex (half|full|auto)
1343 port config - queues/descriptors
1344 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1346 Set number of queues/descriptors for rxq, txq, rxd and txd::
1348 testpmd> port config all (rxq|txq|rxd|txd) (value)
1350 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1352 port config - max-pkt-len
1353 ~~~~~~~~~~~~~~~~~~~~~~~~~
1355 Set the maximum packet length::
1357 testpmd> port config all max-pkt-len (value)
1359 This is equivalent to the ``--max-pkt-len`` command-line option.
1361 port config - CRC Strip
1362 ~~~~~~~~~~~~~~~~~~~~~~~
1364 Set hardware CRC stripping on or off for all ports::
1366 testpmd> port config all crc-strip (on|off)
1368 CRC stripping is off by default.
1370 The ``on`` option is equivalent to the ``--crc-strip`` command-line option.
1372 port config - scatter
1373 ~~~~~~~~~~~~~~~~~~~~~~~
1375 Set RX scatter mode on or off for all ports::
1377 testpmd> port config all scatter (on|off)
1379 RX scatter mode is off by default.
1381 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1383 port config - TX queue flags
1384 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1386 Set a hexadecimal bitmap of TX queue flags for all ports::
1388 testpmd> port config all txqflags value
1390 This command is equivalent to the ``--txqflags`` command-line option.
1392 port config - RX Checksum
1393 ~~~~~~~~~~~~~~~~~~~~~~~~~
1395 Set hardware RX checksum offload to on or off for all ports::
1397 testpmd> port config all rx-cksum (on|off)
1399 Checksum offload is off by default.
1401 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1406 Set hardware VLAN on or off for all ports::
1408 testpmd> port config all hw-vlan (on|off)
1410 Hardware VLAN is on by default.
1412 The ``off`` option is equivalent to the ``--disable-hw-vlan`` command-line option.
1414 port config - VLAN filter
1415 ~~~~~~~~~~~~~~~~~~~~~~~~~
1417 Set hardware VLAN filter on or off for all ports::
1419 testpmd> port config all hw-vlan-filter (on|off)
1421 Hardware VLAN filter is on by default.
1423 The ``off`` option is equivalent to the ``--disable-hw-vlan-filter`` command-line option.
1425 port config - VLAN strip
1426 ~~~~~~~~~~~~~~~~~~~~~~~~
1428 Set hardware VLAN strip on or off for all ports::
1430 testpmd> port config all hw-vlan-strip (on|off)
1432 Hardware VLAN strip is on by default.
1434 The ``off`` option is equivalent to the ``--disable-hw-vlan-strip`` command-line option.
1436 port config - VLAN extend
1437 ~~~~~~~~~~~~~~~~~~~~~~~~~
1439 Set hardware VLAN extend on or off for all ports::
1441 testpmd> port config all hw-vlan-extend (on|off)
1443 Hardware VLAN extend is off by default.
1445 The ``off`` option is equivalent to the ``--disable-hw-vlan-extend`` command-line option.
1447 port config - Drop Packets
1448 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1450 Set packet drop for packets with no descriptors on or off for all ports::
1452 testpmd> port config all drop-en (on|off)
1454 Packet dropping for packets with no descriptors is off by default.
1456 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1461 Set the RSS (Receive Side Scaling) mode on or off::
1463 testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1465 RSS is on by default.
1467 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1469 port config - RSS Reta
1470 ~~~~~~~~~~~~~~~~~~~~~~
1472 Set the RSS (Receive Side Scaling) redirection table::
1474 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1479 Set the DCB mode for an individual port::
1481 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1483 The traffic class should be 4 or 8.
1488 Set the number of packets per burst::
1490 testpmd> port config all burst (value)
1492 This is equivalent to the ``--burst`` command-line option.
1494 port config - Threshold
1495 ~~~~~~~~~~~~~~~~~~~~~~~
1497 Set thresholds for TX/RX queues::
1499 testpmd> port config all (threshold) (value)
1501 Where the threshold type can be:
1503 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1505 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1507 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1509 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1511 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1513 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1515 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1517 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1519 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1521 These threshold options are also available from the command-line.
1526 Set the value of ether-type for E-tag::
1528 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1530 Enable/disable the E-tag support::
1532 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1535 Link Bonding Functions
1536 ----------------------
1538 The Link Bonding functions make it possible to dynamically create and
1539 manage link bonding devices from within testpmd interactive prompt.
1541 create bonded device
1542 ~~~~~~~~~~~~~~~~~~~~
1544 Create a new bonding device::
1546 testpmd> create bonded device (mode) (socket)
1548 For example, to create a bonded device in mode 1 on socket 0::
1550 testpmd> create bonded 1 0
1551 created new bonded device (port X)
1556 Adds Ethernet device to a Link Bonding device::
1558 testpmd> add bonding slave (slave id) (port id)
1560 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
1562 testpmd> add bonding slave 6 10
1565 remove bonding slave
1566 ~~~~~~~~~~~~~~~~~~~~
1568 Removes an Ethernet slave device from a Link Bonding device::
1570 testpmd> remove bonding slave (slave id) (port id)
1572 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
1574 testpmd> remove bonding slave 6 10
1579 Set the Link Bonding mode of a Link Bonding device::
1581 testpmd> set bonding mode (value) (port id)
1583 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
1585 testpmd> set bonding mode 3 10
1590 Set an Ethernet slave device as the primary device on a Link Bonding device::
1592 testpmd> set bonding primary (slave id) (port id)
1594 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
1596 testpmd> set bonding primary 6 10
1601 Set the MAC address of a Link Bonding device::
1603 testpmd> set bonding mac (port id) (mac)
1605 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
1607 testpmd> set bonding mac 10 00:00:00:00:00:01
1609 set bonding xmit_balance_policy
1610 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1612 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
1614 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
1616 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
1618 testpmd> set bonding xmit_balance_policy 10 l34
1621 set bonding mon_period
1622 ~~~~~~~~~~~~~~~~~~~~~~
1624 Set the link status monitoring polling period in milliseconds for a bonding device.
1626 This adds support for PMD slave devices which do not support link status interrupts.
1627 When the mon_period is set to a value greater than 0 then all PMD's which do not support
1628 link status ISR will be queried every polling interval to check if their link status has changed::
1630 testpmd> set bonding mon_period (port_id) (value)
1632 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
1634 testpmd> set bonding mon_period 5 150
1640 Show the current configuration of a Link Bonding device::
1642 testpmd> show bonding config (port id)
1645 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
1646 in balance mode with a transmission policy of layer 2+3::
1648 testpmd> show bonding config 9
1650 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
1652 Active Slaves (3): [1 3 4]
1659 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
1660 This is mainly useful for debugging purposes.
1661 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
1662 and fields that can be accessed.
1667 Display the value of a port register::
1669 testpmd> read reg (port_id) (address)
1671 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
1673 testpmd> read reg 0 0xEE00
1674 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
1679 Display a port register bit field::
1681 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
1683 For example, reading the lowest two bits from the register in the example above::
1685 testpmd> read regfield 0 0xEE00 0 1
1686 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
1691 Display a single port register bit::
1693 testpmd> read regbit (port_id) (address) (bit_x)
1695 For example, reading the lowest bit from the register in the example above::
1697 testpmd> read regbit 0 0xEE00 0
1698 port 0 PCI register at offset 0xEE00: bit 0=1
1703 Set the value of a port register::
1705 testpmd> write reg (port_id) (address) (value)
1707 For example, to clear a register::
1709 testpmd> write reg 0 0xEE00 0x0
1710 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
1715 Set bit field of a port register::
1717 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
1719 For example, writing to the register cleared in the example above::
1721 testpmd> write regfield 0 0xEE00 0 1 2
1722 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
1727 Set single bit value of a port register::
1729 testpmd> write regbit (port_id) (address) (bit_x) (value)
1731 For example, to set the high bit in the register from the example above::
1733 testpmd> write regbit 0 0xEE00 31 1
1734 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
1740 This section details the available filter functions that are available.
1742 Note these functions interface the deprecated legacy filtering framework,
1743 superseded by *rte_flow*. See `Flow rules management`_.
1746 ~~~~~~~~~~~~~~~~~~~~
1748 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
1750 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
1751 ethertype (ether_type) (drop|fwd) queue (queue_id)
1753 The available information parameters are:
1755 * ``port_id``: The port which the Ethertype filter assigned on.
1757 * ``mac_addr``: Compare destination mac address.
1759 * ``mac_ignr``: Ignore destination mac address match.
1761 * ``mac_address``: Destination mac address to match.
1763 * ``ether_type``: The EtherType value want to match,
1764 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
1766 * ``queue_id``: The receive queue associated with this EtherType filter.
1767 It is meaningless when deleting or dropping.
1769 Example, to add/remove an ethertype filter rule::
1771 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
1772 ethertype 0x0806 fwd queue 3
1774 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
1775 ethertype 0x0806 fwd queue 3
1780 Add or delete a 2-tuple filter,
1781 which identifies packets by specific protocol and destination TCP/UDP port
1782 and forwards packets into one of the receive queues::
1784 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
1785 protocol (protocol_value) mask (mask_value) \
1786 tcp_flags (tcp_flags_value) priority (prio_value) \
1789 The available information parameters are:
1791 * ``port_id``: The port which the 2-tuple filter assigned on.
1793 * ``dst_port_value``: Destination port in L4.
1795 * ``protocol_value``: IP L4 protocol.
1797 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
1799 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
1801 * ``prio_value``: Priority of this filter.
1803 * ``queue_id``: The receive queue associated with this 2-tuple filter.
1805 Example, to add/remove an 2tuple filter rule::
1807 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
1808 tcp_flags 0x02 priority 3 queue 3
1810 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
1811 tcp_flags 0x02 priority 3 queue 3
1816 Add or delete a 5-tuple filter,
1817 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
1818 and routes packets into one of the receive queues::
1820 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
1821 (src_address) dst_port (dst_port_value) \
1822 src_port (src_port_value) protocol (protocol_value) \
1823 mask (mask_value) tcp_flags (tcp_flags_value) \
1824 priority (prio_value) queue (queue_id)
1826 The available information parameters are:
1828 * ``port_id``: The port which the 5-tuple filter assigned on.
1830 * ``dst_address``: Destination IP address.
1832 * ``src_address``: Source IP address.
1834 * ``dst_port_value``: TCP/UDP destination port.
1836 * ``src_port_value``: TCP/UDP source port.
1838 * ``protocol_value``: L4 protocol.
1840 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
1842 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
1844 * ``prio_value``: The priority of this filter.
1846 * ``queue_id``: The receive queue associated with this 5-tuple filter.
1848 Example, to add/remove an 5tuple filter rule::
1850 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1851 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1852 flags 0x0 priority 3 queue 3
1854 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1855 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1856 flags 0x0 priority 3 queue 3
1861 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
1863 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
1865 The available information parameters are:
1867 * ``port_id``: The port which the SYN filter assigned on.
1869 * ``high``: This SYN filter has higher priority than other filters.
1871 * ``low``: This SYN filter has lower priority than other filters.
1873 * ``queue_id``: The receive queue associated with this SYN filter
1877 testpmd> syn_filter 0 add priority high queue 3
1882 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
1883 and routed into one of the receive queues::
1885 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
1886 mask (mask_value) priority (prio_value) queue (queue_id)
1888 The available information parameters are:
1890 * ``port_id``: The port which the Flex filter is assigned on.
1892 * ``len_value``: Filter length in bytes, no greater than 128.
1894 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
1896 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
1898 * ``prio_value``: The priority of this filter.
1900 * ``queue_id``: The receive queue associated with this Flex filter.
1904 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
1905 mask 000C priority 3 queue 3
1907 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
1908 mask 000C priority 3 queue 3
1911 .. _testpmd_flow_director:
1913 flow_director_filter
1914 ~~~~~~~~~~~~~~~~~~~~
1916 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
1918 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
1919 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
1921 * Perfect 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 IP flow.
1925 * Signature filters.
1926 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
1928 * Perfect-mac-vlan match filters.
1929 The hardware checks a match between the masked fields of the received packets and the programmed filters.
1930 The masked fields are for MAC VLAN flow.
1932 * Perfect-tunnel match filters.
1933 The hardware checks a match between the masked fields of the received packets and the programmed filters.
1934 The masked fields are for tunnel flow.
1936 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
1937 per flow type and the flexible payload.
1939 The Flow Director can also mask out parts of all of these fields so that filters
1940 are only applied to certain fields or parts of the fields.
1942 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
1944 # Commands to add flow director filters of different flow types::
1946 flow_director_filter (port_id) mode IP (add|del|update) \
1947 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
1948 src (src_ip_address) dst (dst_ip_address) \
1949 tos (tos_value) proto (proto_value) ttl (ttl_value) \
1950 vlan (vlan_value) flexbytes (flexbytes_value) \
1951 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
1954 flow_director_filter (port_id) mode IP (add|del|update) \
1955 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
1956 src (src_ip_address) (src_port) \
1957 dst (dst_ip_address) (dst_port) \
1958 tos (tos_value) ttl (ttl_value) \
1959 vlan (vlan_value) flexbytes (flexbytes_value) \
1960 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
1963 flow_director_filter (port_id) mode IP (add|del|update) \
1964 flow (ipv4-sctp|ipv6-sctp) \
1965 src (src_ip_address) (src_port) \
1966 dst (dst_ip_address) (dst_port) \
1967 tos (tos_value) ttl (ttl_value) \
1968 tag (verification_tag) vlan (vlan_value) \
1969 flexbytes (flexbytes_value) (drop|fwd) \
1970 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
1972 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
1973 ether (ethertype) flexbytes (flexbytes_value) \
1974 (drop|fwd) pf|vf(vf_id) queue (queue_id)
1977 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
1978 mac (mac_address) vlan (vlan_value) \
1979 flexbytes (flexbytes_value) (drop|fwd) \
1980 queue (queue_id) fd_id (fd_id_value)
1982 flow_director_filter (port_id) mode Tunnel (add|del|update) \
1983 mac (mac_address) vlan (vlan_value) \
1984 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
1985 flexbytes (flexbytes_value) (drop|fwd) \
1986 queue (queue_id) fd_id (fd_id_value)
1988 For example, to add an ipv4-udp flow type filter::
1990 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
1991 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
1992 fwd pf queue 1 fd_id 1
1994 For example, add an ipv4-other flow type filter::
1996 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
1997 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
1998 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
2003 Flush all flow director filters on a device::
2005 testpmd> flush_flow_director (port_id)
2007 Example, to flush all flow director filter on port 0::
2009 testpmd> flush_flow_director 0
2014 Set flow director's input masks::
2016 flow_director_mask (port_id) mode IP vlan (vlan_value) \
2017 src_mask (ipv4_src) (ipv6_src) (src_port) \
2018 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2020 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2022 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2023 mac (mac_value) tunnel-type (tunnel_type_value) \
2024 tunnel-id (tunnel_id_value)
2026 Example, to set flow director mask on port 0::
2028 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2029 src_mask 255.255.255.255 \
2030 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2031 dst_mask 255.255.255.255 \
2032 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2034 flow_director_flex_mask
2035 ~~~~~~~~~~~~~~~~~~~~~~~
2037 set masks of flow director's flexible payload based on certain flow type::
2039 testpmd> flow_director_flex_mask (port_id) \
2040 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2041 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2042 l2_payload|all) (mask)
2044 Example, to set flow director's flex mask for all flow type on port 0::
2046 testpmd> flow_director_flex_mask 0 flow all \
2047 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2050 flow_director_flex_payload
2051 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2053 Configure flexible payload selection::
2055 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2057 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2059 testpmd> flow_director_flex_payload 0 l4 \
2060 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2062 get_sym_hash_ena_per_port
2063 ~~~~~~~~~~~~~~~~~~~~~~~~~
2065 Get symmetric hash enable configuration per port::
2067 get_sym_hash_ena_per_port (port_id)
2069 For example, to get symmetric hash enable configuration of port 1::
2071 testpmd> get_sym_hash_ena_per_port 1
2073 set_sym_hash_ena_per_port
2074 ~~~~~~~~~~~~~~~~~~~~~~~~~
2076 Set symmetric hash enable configuration per port to enable or disable::
2078 set_sym_hash_ena_per_port (port_id) (enable|disable)
2080 For example, to set symmetric hash enable configuration of port 1 to enable::
2082 testpmd> set_sym_hash_ena_per_port 1 enable
2084 get_hash_global_config
2085 ~~~~~~~~~~~~~~~~~~~~~~
2087 Get the global configurations of hash filters::
2089 get_hash_global_config (port_id)
2091 For example, to get the global configurations of hash filters of port 1::
2093 testpmd> get_hash_global_config 1
2095 set_hash_global_config
2096 ~~~~~~~~~~~~~~~~~~~~~~
2098 Set the global configurations of hash filters::
2100 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2101 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2102 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload) \
2105 For example, to enable simple_xor for flow type of ipv6 on port 2::
2107 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2112 Set the input set for hash::
2114 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2115 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2116 l2_payload) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2117 ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2118 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2119 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2120 fld-8th|none) (select|add)
2122 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2124 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2129 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2130 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2132 Set the input set for flow director::
2134 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2135 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2136 l2_payload) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2137 ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2138 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2139 sctp-dst-port|sctp-veri-tag|none) (select|add)
2141 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2143 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2148 Set different GRE key length for input set::
2150 global_config (port_id) gre-key-len (number in bytes)
2152 For example to set GRE key length for input set to 4 bytes on port 0::
2154 testpmd> global_config 0 gre-key-len 4
2157 .. _testpmd_rte_flow:
2159 Flow rules management
2160 ---------------------
2162 Control of the generic flow API (*rte_flow*) is fully exposed through the
2163 ``flow`` command (validation, creation, destruction and queries).
2165 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2166 features simultaneously may cause undefined side-effects and is therefore
2172 Because the ``flow`` command uses dynamic tokens to handle the large number
2173 of possible flow rules combinations, its behavior differs slightly from
2174 other commands, in particular:
2176 - Pressing *?* or the *<tab>* key displays contextual help for the current
2177 token, not that of the entire command.
2179 - Optional and repeated parameters are supported (provided they are listed
2180 in the contextual help).
2182 The first parameter stands for the operation mode. Possible operations and
2183 their general syntax are described below. They are covered in detail in the
2186 - Check whether a flow rule can be created::
2188 flow validate {port_id}
2189 [group {group_id}] [priority {level}] [ingress] [egress]
2190 pattern {item} [/ {item} [...]] / end
2191 actions {action} [/ {action} [...]] / end
2193 - Create a flow rule::
2195 flow create {port_id}
2196 [group {group_id}] [priority {level}] [ingress] [egress]
2197 pattern {item} [/ {item} [...]] / end
2198 actions {action} [/ {action} [...]] / end
2200 - Destroy specific flow rules::
2202 flow destroy {port_id} rule {rule_id} [...]
2204 - Destroy all flow rules::
2206 flow flush {port_id}
2208 - Query an existing flow rule::
2210 flow query {port_id} {rule_id} {action}
2212 - List existing flow rules sorted by priority, filtered by group
2215 flow list {port_id} [group {group_id}] [...]
2217 Validating flow rules
2218 ~~~~~~~~~~~~~~~~~~~~~
2220 ``flow validate`` reports whether a flow rule would be accepted by the
2221 underlying device in its current state but stops short of creating it. It is
2222 bound to ``rte_flow_validate()``::
2224 flow validate {port_id}
2225 [group {group_id}] [priority {level}] [ingress] [egress]
2226 pattern {item} [/ {item} [...]] / end
2227 actions {action} [/ {action} [...]] / end
2229 If successful, it will show::
2233 Otherwise it will show an error message of the form::
2235 Caught error type [...] ([...]): [...]
2237 This command uses the same parameters as ``flow create``, their format is
2238 described in `Creating flow rules`_.
2240 Check whether redirecting any Ethernet packet received on port 0 to RX queue
2241 index 6 is supported::
2243 testpmd> flow validate 0 ingress pattern eth / end
2244 actions queue index 6 / end
2248 Port 0 does not support TCPv6 rules::
2250 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
2252 Caught error type 9 (specific pattern item): Invalid argument
2258 ``flow create`` validates and creates the specified flow rule. It is bound
2259 to ``rte_flow_create()``::
2261 flow create {port_id}
2262 [group {group_id}] [priority {level}] [ingress] [egress]
2263 pattern {item} [/ {item} [...]] / end
2264 actions {action} [/ {action} [...]] / end
2266 If successful, it will return a flow rule ID usable with other commands::
2268 Flow rule #[...] created
2270 Otherwise it will show an error message of the form::
2272 Caught error type [...] ([...]): [...]
2274 Parameters describe in the following order:
2276 - Attributes (*group*, *priority*, *ingress*, *egress* tokens).
2277 - A matching pattern, starting with the *pattern* token and terminated by an
2279 - Actions, starting with the *actions* token and terminated by an *end*
2282 These translate directly to *rte_flow* objects provided as-is to the
2283 underlying functions.
2285 The shortest valid definition only comprises mandatory tokens::
2287 testpmd> flow create 0 pattern end actions end
2289 Note that PMDs may refuse rules that essentially do nothing such as this
2292 **All unspecified object values are automatically initialized to 0.**
2297 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
2298 specified before the ``pattern`` token.
2300 - ``group {group id}``: priority group.
2301 - ``priority {level}``: priority level within group.
2302 - ``ingress``: rule applies to ingress traffic.
2303 - ``egress``: rule applies to egress traffic.
2305 Each instance of an attribute specified several times overrides the previous
2306 value as shown below (group 4 is used)::
2308 testpmd> flow create 0 group 42 group 24 group 4 [...]
2310 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
2312 While not specifying a direction is an error, some rules may allow both
2315 Most rules affect RX therefore contain the ``ingress`` token::
2317 testpmd> flow create 0 ingress pattern [...]
2322 A matching pattern starts after the ``pattern`` token. It is made of pattern
2323 items and is terminated by a mandatory ``end`` item.
2325 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
2326 rte_flow_item_type``).
2328 The ``/`` token is used as a separator between pattern items as shown
2331 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
2333 Note that protocol items like these must be stacked from lowest to highest
2334 layer to make sense. For instance, the following rule is either invalid or
2335 unlikely to match any packet::
2337 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
2339 More information on these restrictions can be found in the *rte_flow*
2342 Several items support additional specification structures, for example
2343 ``ipv4`` allows specifying source and destination addresses as follows::
2345 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
2346 dst is 10.2.0.0 / end [...]
2348 This rule matches all IPv4 traffic with the specified properties.
2350 In this example, ``src`` and ``dst`` are field names of the underlying
2351 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
2352 in a similar fashion.
2354 The ``is`` token means that the subsequent value must be matched exactly,
2355 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
2356 accordingly. Possible assignment tokens are:
2358 - ``is``: match value perfectly (with full bit-mask).
2359 - ``spec``: match value according to configured bit-mask.
2360 - ``last``: specify upper bound to establish a range.
2361 - ``mask``: specify bit-mask with relevant bits set to one.
2362 - ``prefix``: generate bit-mask from a prefix length.
2364 These yield identical results::
2366 ipv4 src is 10.1.1.1
2370 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
2374 ipv4 src spec 10.1.1.1 src prefix 32
2378 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
2382 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
2384 Inclusive ranges can be defined with ``last``::
2386 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
2388 Note that ``mask`` affects both ``spec`` and ``last``::
2390 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
2391 # matches 10.1.0.0 to 10.2.255.255
2393 Properties can be modified multiple times::
2395 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
2399 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
2404 This section lists supported pattern items and their attributes, if any.
2406 - ``end``: end list of pattern items.
2408 - ``void``: no-op pattern item.
2410 - ``invert``: perform actions when pattern does not match.
2412 - ``any``: match any protocol for the current layer.
2414 - ``num {unsigned}``: number of layers covered.
2416 - ``pf``: match packets addressed to the physical function.
2418 - ``vf``: match packets addressed to a virtual function ID.
2420 - ``id {unsigned}``: destination VF ID.
2422 - ``port``: device-specific physical port index to use.
2424 - ``index {unsigned}``: physical port index.
2426 - ``raw``: match an arbitrary byte string.
2428 - ``relative {boolean}``: look for pattern after the previous item.
2429 - ``search {boolean}``: search pattern from offset (see also limit).
2430 - ``offset {integer}``: absolute or relative offset for pattern.
2431 - ``limit {unsigned}``: search area limit for start of pattern.
2432 - ``pattern {string}``: byte string to look for.
2434 - ``eth``: match Ethernet header.
2436 - ``dst {MAC-48}``: destination MAC.
2437 - ``src {MAC-48}``: source MAC.
2438 - ``type {unsigned}``: EtherType.
2440 - ``vlan``: match 802.1Q/ad VLAN tag.
2442 - ``tpid {unsigned}``: tag protocol identifier.
2443 - ``tci {unsigned}``: tag control information.
2444 - ``pcp {unsigned}``: priority code point.
2445 - ``dei {unsigned}``: drop eligible indicator.
2446 - ``vid {unsigned}``: VLAN identifier.
2448 - ``ipv4``: match IPv4 header.
2450 - ``tos {unsigned}``: type of service.
2451 - ``ttl {unsigned}``: time to live.
2452 - ``proto {unsigned}``: next protocol ID.
2453 - ``src {ipv4 address}``: source address.
2454 - ``dst {ipv4 address}``: destination address.
2456 - ``ipv6``: match IPv6 header.
2458 - ``tc {unsigned}``: traffic class.
2459 - ``flow {unsigned}``: flow label.
2460 - ``proto {unsigned}``: protocol (next header).
2461 - ``hop {unsigned}``: hop limit.
2462 - ``src {ipv6 address}``: source address.
2463 - ``dst {ipv6 address}``: destination address.
2465 - ``icmp``: match ICMP header.
2467 - ``type {unsigned}``: ICMP packet type.
2468 - ``code {unsigned}``: ICMP packet code.
2470 - ``udp``: match UDP header.
2472 - ``src {unsigned}``: UDP source port.
2473 - ``dst {unsigned}``: UDP destination port.
2475 - ``tcp``: match TCP header.
2477 - ``src {unsigned}``: TCP source port.
2478 - ``dst {unsigned}``: TCP destination port.
2480 - ``sctp``: match SCTP header.
2482 - ``src {unsigned}``: SCTP source port.
2483 - ``dst {unsigned}``: SCTP destination port.
2484 - ``tag {unsigned}``: validation tag.
2485 - ``cksum {unsigned}``: checksum.
2487 - ``vxlan``: match VXLAN header.
2489 - ``vni {unsigned}``: VXLAN identifier.
2491 - ``mpls``: match MPLS header.
2493 - ``label {unsigned}``: MPLS label.
2495 - ``gre``: match GRE header.
2497 - ``protocol {unsigned}``: protocol type.
2502 A list of actions starts after the ``actions`` token in the same fashion as
2503 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
2504 terminated by a mandatory ``end`` action.
2506 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
2507 rte_flow_action_type``).
2509 Dropping all incoming UDPv4 packets can be expressed as follows::
2511 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2514 Several actions have configurable properties which must be specified when
2515 there is no valid default value. For example, ``queue`` requires a target
2518 This rule redirects incoming UDPv4 traffic to queue index 6::
2520 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2521 actions queue index 6 / end
2523 While this one could be rejected by PMDs (unspecified queue index)::
2525 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2528 As defined by *rte_flow*, the list is not ordered, all actions of a given
2529 rule are performed simultaneously. These are equivalent::
2531 queue index 6 / void / mark id 42 / end
2535 void / mark id 42 / queue index 6 / end
2537 All actions in a list should have different types, otherwise only the last
2538 action of a given type is taken into account::
2540 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
2544 drop / drop / drop / end # drop is performed only once
2548 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
2550 Considering they are performed simultaneously, opposite and overlapping
2551 actions can sometimes be combined when the end result is unambiguous::
2553 drop / queue index 6 / end # drop has no effect
2557 drop / dup index 6 / end # same as above
2561 queue index 6 / rss queues 6 7 8 / end # queue has no effect
2565 drop / passthru / end # drop has no effect
2567 Note that PMDs may still refuse such combinations.
2572 This section lists supported actions and their attributes, if any.
2574 - ``end``: end list of actions.
2576 - ``void``: no-op action.
2578 - ``passthru``: let subsequent rule process matched packets.
2580 - ``mark``: attach 32 bit value to packets.
2582 - ``id {unsigned}``: 32 bit value to return with packets.
2584 - ``flag``: flag packets.
2586 - ``queue``: assign packets to a given queue index.
2588 - ``index {unsigned}``: queue index to use.
2590 - ``drop``: drop packets (note: passthru has priority).
2592 - ``count``: enable counters for this rule.
2594 - ``dup``: duplicate packets to a given queue index.
2596 - ``index {unsigned}``: queue index to duplicate packets to.
2598 - ``rss``: spread packets among several queues.
2600 - ``queues [{unsigned} [...]] end``: queue indices to use.
2602 - ``pf``: redirect packets to physical device function.
2604 - ``vf``: redirect packets to virtual device function.
2606 - ``original {boolean}``: use original VF ID if possible.
2607 - ``id {unsigned}``: VF ID to redirect packets to.
2609 Destroying flow rules
2610 ~~~~~~~~~~~~~~~~~~~~~
2612 ``flow destroy`` destroys one or more rules from their rule ID (as returned
2613 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
2614 times as necessary::
2616 flow destroy {port_id} rule {rule_id} [...]
2618 If successful, it will show::
2620 Flow rule #[...] destroyed
2622 It does not report anything for rule IDs that do not exist. The usual error
2623 message is shown when a rule cannot be destroyed::
2625 Caught error type [...] ([...]): [...]
2627 ``flow flush`` destroys all rules on a device and does not take extra
2628 arguments. It is bound to ``rte_flow_flush()``::
2630 flow flush {port_id}
2632 Any errors are reported as above.
2634 Creating several rules and destroying them::
2636 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2637 actions queue index 2 / end
2638 Flow rule #0 created
2639 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2640 actions queue index 3 / end
2641 Flow rule #1 created
2642 testpmd> flow destroy 0 rule 0 rule 1
2643 Flow rule #1 destroyed
2644 Flow rule #0 destroyed
2647 The same result can be achieved using ``flow flush``::
2649 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2650 actions queue index 2 / end
2651 Flow rule #0 created
2652 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2653 actions queue index 3 / end
2654 Flow rule #1 created
2655 testpmd> flow flush 0
2658 Non-existent rule IDs are ignored::
2660 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2661 actions queue index 2 / end
2662 Flow rule #0 created
2663 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2664 actions queue index 3 / end
2665 Flow rule #1 created
2666 testpmd> flow destroy 0 rule 42 rule 10 rule 2
2668 testpmd> flow destroy 0 rule 0
2669 Flow rule #0 destroyed
2675 ``flow query`` queries a specific action of a flow rule having that
2676 ability. Such actions collect information that can be reported using this
2677 command. It is bound to ``rte_flow_query()``::
2679 flow query {port_id} {rule_id} {action}
2681 If successful, it will display either the retrieved data for known actions
2682 or the following message::
2684 Cannot display result for action type [...] ([...])
2686 Otherwise, it will complain either that the rule does not exist or that some
2689 Flow rule #[...] not found
2693 Caught error type [...] ([...]): [...]
2695 Currently only the ``count`` action is supported. This action reports the
2696 number of packets that hit the flow rule and the total number of bytes. Its
2697 output has the following format::
2700 hits_set: [...] # whether "hits" contains a valid value
2701 bytes_set: [...] # whether "bytes" contains a valid value
2702 hits: [...] # number of packets
2703 bytes: [...] # number of bytes
2705 Querying counters for TCPv6 packets redirected to queue 6::
2707 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
2708 actions queue index 6 / count / end
2709 Flow rule #4 created
2710 testpmd> flow query 0 4 count
2721 ``flow list`` lists existing flow rules sorted by priority and optionally
2722 filtered by group identifiers::
2724 flow list {port_id} [group {group_id}] [...]
2726 This command only fails with the following message if the device does not
2731 Output consists of a header line followed by a short description of each
2732 flow rule, one per line. There is no output at all when no flow rules are
2733 configured on the device::
2735 ID Group Prio Attr Rule
2736 [...] [...] [...] [...] [...]
2738 ``Attr`` column flags:
2740 - ``i`` for ``ingress``.
2741 - ``e`` for ``egress``.
2743 Creating several flow rules and listing them::
2745 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2746 actions queue index 6 / end
2747 Flow rule #0 created
2748 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2749 actions queue index 2 / end
2750 Flow rule #1 created
2751 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
2752 actions rss queues 6 7 8 end / end
2753 Flow rule #2 created
2754 testpmd> flow list 0
2755 ID Group Prio Attr Rule
2756 0 0 0 i- ETH IPV4 => QUEUE
2757 1 0 0 i- ETH IPV6 => QUEUE
2758 2 0 5 i- ETH IPV4 UDP => RSS
2761 Rules are sorted by priority (i.e. group ID first, then priority level)::
2763 testpmd> flow list 1
2764 ID Group Prio Attr Rule
2765 0 0 0 i- ETH => COUNT
2766 6 0 500 i- ETH IPV6 TCP => DROP COUNT
2767 5 0 1000 i- ETH IPV6 ICMP => QUEUE
2768 1 24 0 i- ETH IPV4 UDP => QUEUE
2769 4 24 10 i- ETH IPV4 TCP => DROP
2770 3 24 20 i- ETH IPV4 => DROP
2771 2 24 42 i- ETH IPV4 UDP => QUEUE
2772 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
2775 Output can be limited to specific groups::
2777 testpmd> flow list 1 group 0 group 63
2778 ID Group Prio Attr Rule
2779 0 0 0 i- ETH => COUNT
2780 6 0 500 i- ETH IPV6 TCP => DROP COUNT
2781 5 0 1000 i- ETH IPV6 ICMP => QUEUE
2782 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE