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, ...)
913 Set the link flow control parameter on a port::
915 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
916 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
917 autoneg (on|off) (port_id)
921 * ``high_water`` (integer): High threshold value to trigger XOFF.
923 * ``low_water`` (integer): Low threshold value to trigger XON.
925 * ``pause_time`` (integer): Pause quota in the Pause frame.
927 * ``send_xon`` (0/1): Send XON frame.
929 * ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
931 * ``autoneg``: Change the auto-negotiation parameter.
936 Set the priority flow control parameter on a port::
938 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
939 (pause_time) (priority) (port_id)
943 * ``high_water`` (integer): High threshold value.
945 * ``low_water`` (integer): Low threshold value.
947 * ``pause_time`` (integer): Pause quota in the Pause frame.
949 * ``priority`` (0-7): VLAN User Priority.
954 Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
956 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
958 For example, to set rx queue 2 on port 0 to mapping 5::
960 testpmd>set stat_qmap rx 0 2 5
962 set port - rx/tx (for VF)
963 ~~~~~~~~~~~~~~~~~~~~~~~~~
965 Set VF receive/transmit from a port::
967 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
969 set port - mac address filter (for VF)
970 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
972 Add/Remove unicast or multicast MAC addr filter for a VF::
974 testpmd> set port (port_id) vf (vf_id) (mac_addr) \
975 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
977 set port - rx mode(for VF)
978 ~~~~~~~~~~~~~~~~~~~~~~~~~~
980 Set the VF receive mode of a port::
982 testpmd> set port (port_id) vf (vf_id) \
983 rxmode (AUPE|ROPE|BAM|MPE) (on|off)
985 The available receive modes are:
987 * ``AUPE``: Accepts untagged VLAN.
989 * ``ROPE``: Accepts unicast hash.
991 * ``BAM``: Accepts broadcast packets.
993 * ``MPE``: Accepts all multicast packets.
995 set port - tx_rate (for Queue)
996 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
998 Set TX rate limitation for a queue on a port::
1000 testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1002 set port - tx_rate (for VF)
1003 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
1005 Set TX rate limitation for queues in VF on a port::
1007 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1009 set port - mirror rule
1010 ~~~~~~~~~~~~~~~~~~~~~~
1012 Set pool or vlan type mirror rule for a port::
1014 testpmd> set port (port_id) mirror-rule (rule_id) \
1015 (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1016 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1018 Set link mirror rule for a port::
1020 testpmd> set port (port_id) mirror-rule (rule_id) \
1021 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1023 For example to enable mirror traffic with vlan 0,1 to pool 0::
1025 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1027 reset port - mirror rule
1028 ~~~~~~~~~~~~~~~~~~~~~~~~
1030 Reset a mirror rule for a port::
1032 testpmd> reset port (port_id) mirror-rule (rule_id)
1037 Set the flush on RX streams before forwarding.
1038 The default is flush ``on``.
1039 Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1041 testpmd> set flush_rx off
1046 Set the bypass mode for the lowest port on bypass enabled NIC::
1048 testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1053 Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1055 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1056 mode (normal|bypass|isolate) (port_id)
1060 * ``timeout``: Enable bypass after watchdog timeout.
1062 * ``os_on``: Enable bypass when OS/board is powered on.
1064 * ``os_off``: Enable bypass when OS/board is powered off.
1066 * ``power_on``: Enable bypass when power supply is turned on.
1068 * ``power_off``: Enable bypass when power supply is turned off.
1074 Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1076 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1081 Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1083 testpmd> show bypass config (port_id)
1088 Set link up for a port::
1090 testpmd> set link-up port (port id)
1095 Set link down for a port::
1097 testpmd> set link-down port (port id)
1102 Enable E-tag insertion for a VF on a port::
1104 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1106 Disable E-tag insertion for a VF on a port::
1108 testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1110 Enable/disable E-tag stripping on a port::
1112 testpmd> E-tag set stripping (on|off) port (port_id)
1114 Enable/disable E-tag based forwarding on a port::
1116 testpmd> E-tag set forwarding (on|off) port (port_id)
1118 Add an E-tag forwarding filter on a port::
1120 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1122 Delete an E-tag forwarding filter on a port::
1123 testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1129 The following sections show functions for configuring ports.
1133 Port configuration changes only become active when forwarding is started/restarted.
1138 Attach a port specified by pci address or virtual device args::
1140 testpmd> port attach (identifier)
1142 To attach a new pci device, the device should be recognized by kernel first.
1143 Then it should be moved under DPDK management.
1144 Finally the port can be attached to testpmd.
1146 For example, to move a pci device using ixgbe under DPDK management:
1148 .. code-block:: console
1150 # Check the status of the available devices.
1151 ./usertools/dpdk-devbind.py --status
1153 Network devices using DPDK-compatible driver
1154 ============================================
1157 Network devices using kernel driver
1158 ===================================
1159 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1162 # Bind the device to igb_uio.
1163 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1166 # Recheck the status of the devices.
1167 ./usertools/dpdk-devbind.py --status
1168 Network devices using DPDK-compatible driver
1169 ============================================
1170 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1172 To attach a port created by virtual device, above steps are not needed.
1174 For example, to attach a port whose pci address is 0000:0a:00.0.
1176 .. code-block:: console
1178 testpmd> port attach 0000:0a:00.0
1179 Attaching a new port...
1180 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1181 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1182 EAL: PCI memory mapped at 0x7f83bfa00000
1183 EAL: PCI memory mapped at 0x7f83bfa80000
1184 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1185 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1186 Port 0 is attached. Now total ports is 1
1189 For example, to attach a port created by pcap PMD.
1191 .. code-block:: console
1193 testpmd> port attach net_pcap0
1194 Attaching a new port...
1195 PMD: Initializing pmd_pcap for net_pcap0
1196 PMD: Creating pcap-backed ethdev on numa socket 0
1197 Port 0 is attached. Now total ports is 1
1200 In this case, identifier is ``net_pcap0``.
1201 This identifier format is the same as ``--vdev`` format of DPDK applications.
1203 For example, to re-attach a bonded port which has been previously detached,
1204 the mode and slave parameters must be given.
1206 .. code-block:: console
1208 testpmd> port attach net_bond_0,mode=0,slave=1
1209 Attaching a new port...
1210 EAL: Initializing pmd_bond for net_bond_0
1211 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1212 Port 0 is attached. Now total ports is 1
1219 Detach a specific port::
1221 testpmd> port detach (port_id)
1223 Before detaching a port, the port should be stopped and closed.
1225 For example, to detach a pci device port 0.
1227 .. code-block:: console
1229 testpmd> port stop 0
1232 testpmd> port close 0
1236 testpmd> port detach 0
1238 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1239 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1240 EAL: PCI memory unmapped at 0x7f83bfa00000
1241 EAL: PCI memory unmapped at 0x7f83bfa80000
1245 For example, to detach a virtual device port 0.
1247 .. code-block:: console
1249 testpmd> port stop 0
1252 testpmd> port close 0
1256 testpmd> port detach 0
1258 PMD: Closing pcap ethdev on numa socket 0
1259 Port 'net_pcap0' is detached. Now total ports is 0
1262 To remove a pci device completely from the system, first detach the port from testpmd.
1263 Then the device should be moved under kernel management.
1264 Finally the device can be removed using kernel pci hotplug functionality.
1266 For example, to move a pci device under kernel management:
1268 .. code-block:: console
1270 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1272 ./usertools/dpdk-devbind.py --status
1274 Network devices using DPDK-compatible driver
1275 ============================================
1278 Network devices using kernel driver
1279 ===================================
1280 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1282 To remove a port created by a virtual device, above steps are not needed.
1287 Start all ports or a specific port::
1289 testpmd> port start (port_id|all)
1294 Stop all ports or a specific port::
1296 testpmd> port stop (port_id|all)
1301 Close all ports or a specific port::
1303 testpmd> port close (port_id|all)
1305 port start/stop queue
1306 ~~~~~~~~~~~~~~~~~~~~~
1308 Start/stop a rx/tx queue on a specific port::
1310 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1312 Only take effect when port is started.
1317 Set the speed and duplex mode for all ports or a specific port::
1319 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1320 duplex (half|full|auto)
1322 port config - queues/descriptors
1323 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1325 Set number of queues/descriptors for rxq, txq, rxd and txd::
1327 testpmd> port config all (rxq|txq|rxd|txd) (value)
1329 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1331 port config - max-pkt-len
1332 ~~~~~~~~~~~~~~~~~~~~~~~~~
1334 Set the maximum packet length::
1336 testpmd> port config all max-pkt-len (value)
1338 This is equivalent to the ``--max-pkt-len`` command-line option.
1340 port config - CRC Strip
1341 ~~~~~~~~~~~~~~~~~~~~~~~
1343 Set hardware CRC stripping on or off for all ports::
1345 testpmd> port config all crc-strip (on|off)
1347 CRC stripping is off by default.
1349 The ``on`` option is equivalent to the ``--crc-strip`` command-line option.
1351 port config - scatter
1352 ~~~~~~~~~~~~~~~~~~~~~~~
1354 Set RX scatter mode on or off for all ports::
1356 testpmd> port config all scatter (on|off)
1358 RX scatter mode is off by default.
1360 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1362 port config - TX queue flags
1363 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1365 Set a hexadecimal bitmap of TX queue flags for all ports::
1367 testpmd> port config all txqflags value
1369 This command is equivalent to the ``--txqflags`` command-line option.
1371 port config - RX Checksum
1372 ~~~~~~~~~~~~~~~~~~~~~~~~~
1374 Set hardware RX checksum offload to on or off for all ports::
1376 testpmd> port config all rx-cksum (on|off)
1378 Checksum offload is off by default.
1380 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1385 Set hardware VLAN on or off for all ports::
1387 testpmd> port config all hw-vlan (on|off)
1389 Hardware VLAN is on by default.
1391 The ``off`` option is equivalent to the ``--disable-hw-vlan`` command-line option.
1393 port config - VLAN filter
1394 ~~~~~~~~~~~~~~~~~~~~~~~~~
1396 Set hardware VLAN filter on or off for all ports::
1398 testpmd> port config all hw-vlan-filter (on|off)
1400 Hardware VLAN filter is on by default.
1402 The ``off`` option is equivalent to the ``--disable-hw-vlan-filter`` command-line option.
1404 port config - VLAN strip
1405 ~~~~~~~~~~~~~~~~~~~~~~~~
1407 Set hardware VLAN strip on or off for all ports::
1409 testpmd> port config all hw-vlan-strip (on|off)
1411 Hardware VLAN strip is on by default.
1413 The ``off`` option is equivalent to the ``--disable-hw-vlan-strip`` command-line option.
1415 port config - VLAN extend
1416 ~~~~~~~~~~~~~~~~~~~~~~~~~
1418 Set hardware VLAN extend on or off for all ports::
1420 testpmd> port config all hw-vlan-extend (on|off)
1422 Hardware VLAN extend is off by default.
1424 The ``off`` option is equivalent to the ``--disable-hw-vlan-extend`` command-line option.
1426 port config - Drop Packets
1427 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1429 Set packet drop for packets with no descriptors on or off for all ports::
1431 testpmd> port config all drop-en (on|off)
1433 Packet dropping for packets with no descriptors is off by default.
1435 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1440 Set the RSS (Receive Side Scaling) mode on or off::
1442 testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1444 RSS is on by default.
1446 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1448 port config - RSS Reta
1449 ~~~~~~~~~~~~~~~~~~~~~~
1451 Set the RSS (Receive Side Scaling) redirection table::
1453 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1458 Set the DCB mode for an individual port::
1460 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1462 The traffic class should be 4 or 8.
1467 Set the number of packets per burst::
1469 testpmd> port config all burst (value)
1471 This is equivalent to the ``--burst`` command-line option.
1473 port config - Threshold
1474 ~~~~~~~~~~~~~~~~~~~~~~~
1476 Set thresholds for TX/RX queues::
1478 testpmd> port config all (threshold) (value)
1480 Where the threshold type can be:
1482 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1484 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1486 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1488 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1490 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1492 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1494 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1496 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1498 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1500 These threshold options are also available from the command-line.
1505 Set the value of ether-type for E-tag::
1507 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1509 Enable/disable the E-tag support::
1511 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1514 Link Bonding Functions
1515 ----------------------
1517 The Link Bonding functions make it possible to dynamically create and
1518 manage link bonding devices from within testpmd interactive prompt.
1520 create bonded device
1521 ~~~~~~~~~~~~~~~~~~~~
1523 Create a new bonding device::
1525 testpmd> create bonded device (mode) (socket)
1527 For example, to create a bonded device in mode 1 on socket 0::
1529 testpmd> create bonded 1 0
1530 created new bonded device (port X)
1535 Adds Ethernet device to a Link Bonding device::
1537 testpmd> add bonding slave (slave id) (port id)
1539 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
1541 testpmd> add bonding slave 6 10
1544 remove bonding slave
1545 ~~~~~~~~~~~~~~~~~~~~
1547 Removes an Ethernet slave device from a Link Bonding device::
1549 testpmd> remove bonding slave (slave id) (port id)
1551 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
1553 testpmd> remove bonding slave 6 10
1558 Set the Link Bonding mode of a Link Bonding device::
1560 testpmd> set bonding mode (value) (port id)
1562 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
1564 testpmd> set bonding mode 3 10
1569 Set an Ethernet slave device as the primary device on a Link Bonding device::
1571 testpmd> set bonding primary (slave id) (port id)
1573 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
1575 testpmd> set bonding primary 6 10
1580 Set the MAC address of a Link Bonding device::
1582 testpmd> set bonding mac (port id) (mac)
1584 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
1586 testpmd> set bonding mac 10 00:00:00:00:00:01
1588 set bonding xmit_balance_policy
1589 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1591 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
1593 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
1595 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
1597 testpmd> set bonding xmit_balance_policy 10 l34
1600 set bonding mon_period
1601 ~~~~~~~~~~~~~~~~~~~~~~
1603 Set the link status monitoring polling period in milliseconds for a bonding device.
1605 This adds support for PMD slave devices which do not support link status interrupts.
1606 When the mon_period is set to a value greater than 0 then all PMD's which do not support
1607 link status ISR will be queried every polling interval to check if their link status has changed::
1609 testpmd> set bonding mon_period (port_id) (value)
1611 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
1613 testpmd> set bonding mon_period 5 150
1619 Show the current configuration of a Link Bonding device::
1621 testpmd> show bonding config (port id)
1624 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
1625 in balance mode with a transmission policy of layer 2+3::
1627 testpmd> show bonding config 9
1629 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
1631 Active Slaves (3): [1 3 4]
1638 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
1639 This is mainly useful for debugging purposes.
1640 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
1641 and fields that can be accessed.
1646 Display the value of a port register::
1648 testpmd> read reg (port_id) (address)
1650 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
1652 testpmd> read reg 0 0xEE00
1653 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
1658 Display a port register bit field::
1660 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
1662 For example, reading the lowest two bits from the register in the example above::
1664 testpmd> read regfield 0 0xEE00 0 1
1665 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
1670 Display a single port register bit::
1672 testpmd> read regbit (port_id) (address) (bit_x)
1674 For example, reading the lowest bit from the register in the example above::
1676 testpmd> read regbit 0 0xEE00 0
1677 port 0 PCI register at offset 0xEE00: bit 0=1
1682 Set the value of a port register::
1684 testpmd> write reg (port_id) (address) (value)
1686 For example, to clear a register::
1688 testpmd> write reg 0 0xEE00 0x0
1689 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
1694 Set bit field of a port register::
1696 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
1698 For example, writing to the register cleared in the example above::
1700 testpmd> write regfield 0 0xEE00 0 1 2
1701 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
1706 Set single bit value of a port register::
1708 testpmd> write regbit (port_id) (address) (bit_x) (value)
1710 For example, to set the high bit in the register from the example above::
1712 testpmd> write regbit 0 0xEE00 31 1
1713 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
1719 This section details the available filter functions that are available.
1721 Note these functions interface the deprecated legacy filtering framework,
1722 superseded by *rte_flow*. See `Flow rules management`_.
1725 ~~~~~~~~~~~~~~~~~~~~
1727 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
1729 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
1730 ethertype (ether_type) (drop|fwd) queue (queue_id)
1732 The available information parameters are:
1734 * ``port_id``: The port which the Ethertype filter assigned on.
1736 * ``mac_addr``: Compare destination mac address.
1738 * ``mac_ignr``: Ignore destination mac address match.
1740 * ``mac_address``: Destination mac address to match.
1742 * ``ether_type``: The EtherType value want to match,
1743 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
1745 * ``queue_id``: The receive queue associated with this EtherType filter.
1746 It is meaningless when deleting or dropping.
1748 Example, to add/remove an ethertype filter rule::
1750 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
1751 ethertype 0x0806 fwd queue 3
1753 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
1754 ethertype 0x0806 fwd queue 3
1759 Add or delete a 2-tuple filter,
1760 which identifies packets by specific protocol and destination TCP/UDP port
1761 and forwards packets into one of the receive queues::
1763 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
1764 protocol (protocol_value) mask (mask_value) \
1765 tcp_flags (tcp_flags_value) priority (prio_value) \
1768 The available information parameters are:
1770 * ``port_id``: The port which the 2-tuple filter assigned on.
1772 * ``dst_port_value``: Destination port in L4.
1774 * ``protocol_value``: IP L4 protocol.
1776 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
1778 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
1780 * ``prio_value``: Priority of this filter.
1782 * ``queue_id``: The receive queue associated with this 2-tuple filter.
1784 Example, to add/remove an 2tuple filter rule::
1786 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
1787 tcp_flags 0x02 priority 3 queue 3
1789 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
1790 tcp_flags 0x02 priority 3 queue 3
1795 Add or delete a 5-tuple filter,
1796 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
1797 and routes packets into one of the receive queues::
1799 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
1800 (src_address) dst_port (dst_port_value) \
1801 src_port (src_port_value) protocol (protocol_value) \
1802 mask (mask_value) tcp_flags (tcp_flags_value) \
1803 priority (prio_value) queue (queue_id)
1805 The available information parameters are:
1807 * ``port_id``: The port which the 5-tuple filter assigned on.
1809 * ``dst_address``: Destination IP address.
1811 * ``src_address``: Source IP address.
1813 * ``dst_port_value``: TCP/UDP destination port.
1815 * ``src_port_value``: TCP/UDP source port.
1817 * ``protocol_value``: L4 protocol.
1819 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
1821 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
1823 * ``prio_value``: The priority of this filter.
1825 * ``queue_id``: The receive queue associated with this 5-tuple filter.
1827 Example, to add/remove an 5tuple filter rule::
1829 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1830 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1831 flags 0x0 priority 3 queue 3
1833 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1834 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1835 flags 0x0 priority 3 queue 3
1840 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
1842 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
1844 The available information parameters are:
1846 * ``port_id``: The port which the SYN filter assigned on.
1848 * ``high``: This SYN filter has higher priority than other filters.
1850 * ``low``: This SYN filter has lower priority than other filters.
1852 * ``queue_id``: The receive queue associated with this SYN filter
1856 testpmd> syn_filter 0 add priority high queue 3
1861 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
1862 and routed into one of the receive queues::
1864 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
1865 mask (mask_value) priority (prio_value) queue (queue_id)
1867 The available information parameters are:
1869 * ``port_id``: The port which the Flex filter is assigned on.
1871 * ``len_value``: Filter length in bytes, no greater than 128.
1873 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
1875 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
1877 * ``prio_value``: The priority of this filter.
1879 * ``queue_id``: The receive queue associated with this Flex filter.
1883 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
1884 mask 000C priority 3 queue 3
1886 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
1887 mask 000C priority 3 queue 3
1890 .. _testpmd_flow_director:
1892 flow_director_filter
1893 ~~~~~~~~~~~~~~~~~~~~
1895 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
1897 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
1898 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
1900 * Perfect match filters.
1901 The hardware checks a match between the masked fields of the received packets and the programmed filters.
1902 The masked fields are for IP flow.
1904 * Signature filters.
1905 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
1907 * Perfect-mac-vlan 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 MAC VLAN flow.
1911 * Perfect-tunnel match filters.
1912 The hardware checks a match between the masked fields of the received packets and the programmed filters.
1913 The masked fields are for tunnel flow.
1915 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
1916 per flow type and the flexible payload.
1918 The Flow Director can also mask out parts of all of these fields so that filters
1919 are only applied to certain fields or parts of the fields.
1921 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
1923 # Commands to add flow director filters of different flow types::
1925 flow_director_filter (port_id) mode IP (add|del|update) \
1926 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
1927 src (src_ip_address) dst (dst_ip_address) \
1928 tos (tos_value) proto (proto_value) ttl (ttl_value) \
1929 vlan (vlan_value) flexbytes (flexbytes_value) \
1930 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
1933 flow_director_filter (port_id) mode IP (add|del|update) \
1934 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
1935 src (src_ip_address) (src_port) \
1936 dst (dst_ip_address) (dst_port) \
1937 tos (tos_value) ttl (ttl_value) \
1938 vlan (vlan_value) flexbytes (flexbytes_value) \
1939 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
1942 flow_director_filter (port_id) mode IP (add|del|update) \
1943 flow (ipv4-sctp|ipv6-sctp) \
1944 src (src_ip_address) (src_port) \
1945 dst (dst_ip_address) (dst_port) \
1946 tos (tos_value) ttl (ttl_value) \
1947 tag (verification_tag) vlan (vlan_value) \
1948 flexbytes (flexbytes_value) (drop|fwd) \
1949 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
1951 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
1952 ether (ethertype) flexbytes (flexbytes_value) \
1953 (drop|fwd) pf|vf(vf_id) queue (queue_id)
1956 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
1957 mac (mac_address) vlan (vlan_value) \
1958 flexbytes (flexbytes_value) (drop|fwd) \
1959 queue (queue_id) fd_id (fd_id_value)
1961 flow_director_filter (port_id) mode Tunnel (add|del|update) \
1962 mac (mac_address) vlan (vlan_value) \
1963 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
1964 flexbytes (flexbytes_value) (drop|fwd) \
1965 queue (queue_id) fd_id (fd_id_value)
1967 For example, to add an ipv4-udp flow type filter::
1969 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
1970 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
1971 fwd pf queue 1 fd_id 1
1973 For example, add an ipv4-other flow type filter::
1975 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
1976 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
1977 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
1982 Flush all flow director filters on a device::
1984 testpmd> flush_flow_director (port_id)
1986 Example, to flush all flow director filter on port 0::
1988 testpmd> flush_flow_director 0
1993 Set flow director's input masks::
1995 flow_director_mask (port_id) mode IP vlan (vlan_value) \
1996 src_mask (ipv4_src) (ipv6_src) (src_port) \
1997 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
1999 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2001 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2002 mac (mac_value) tunnel-type (tunnel_type_value) \
2003 tunnel-id (tunnel_id_value)
2005 Example, to set flow director mask on port 0::
2007 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2008 src_mask 255.255.255.255 \
2009 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2010 dst_mask 255.255.255.255 \
2011 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2013 flow_director_flex_mask
2014 ~~~~~~~~~~~~~~~~~~~~~~~
2016 set masks of flow director's flexible payload based on certain flow type::
2018 testpmd> flow_director_flex_mask (port_id) \
2019 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2020 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2021 l2_payload|all) (mask)
2023 Example, to set flow director's flex mask for all flow type on port 0::
2025 testpmd> flow_director_flex_mask 0 flow all \
2026 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2029 flow_director_flex_payload
2030 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2032 Configure flexible payload selection::
2034 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2036 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2038 testpmd> flow_director_flex_payload 0 l4 \
2039 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2041 get_sym_hash_ena_per_port
2042 ~~~~~~~~~~~~~~~~~~~~~~~~~
2044 Get symmetric hash enable configuration per port::
2046 get_sym_hash_ena_per_port (port_id)
2048 For example, to get symmetric hash enable configuration of port 1::
2050 testpmd> get_sym_hash_ena_per_port 1
2052 set_sym_hash_ena_per_port
2053 ~~~~~~~~~~~~~~~~~~~~~~~~~
2055 Set symmetric hash enable configuration per port to enable or disable::
2057 set_sym_hash_ena_per_port (port_id) (enable|disable)
2059 For example, to set symmetric hash enable configuration of port 1 to enable::
2061 testpmd> set_sym_hash_ena_per_port 1 enable
2063 get_hash_global_config
2064 ~~~~~~~~~~~~~~~~~~~~~~
2066 Get the global configurations of hash filters::
2068 get_hash_global_config (port_id)
2070 For example, to get the global configurations of hash filters of port 1::
2072 testpmd> get_hash_global_config 1
2074 set_hash_global_config
2075 ~~~~~~~~~~~~~~~~~~~~~~
2077 Set the global configurations of hash filters::
2079 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2080 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2081 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload) \
2084 For example, to enable simple_xor for flow type of ipv6 on port 2::
2086 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2091 Set the input set for hash::
2093 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2094 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2095 l2_payload) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2096 ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2097 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2098 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2099 fld-8th|none) (select|add)
2101 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2103 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2108 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2109 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2111 Set the input set for flow director::
2113 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2114 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2115 l2_payload) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2116 ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2117 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2118 sctp-dst-port|sctp-veri-tag|none) (select|add)
2120 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2122 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2127 Set different GRE key length for input set::
2129 global_config (port_id) gre-key-len (number in bytes)
2131 For example to set GRE key length for input set to 4 bytes on port 0::
2133 testpmd> global_config 0 gre-key-len 4
2136 .. _testpmd_rte_flow:
2138 Flow rules management
2139 ---------------------
2141 Control of the generic flow API (*rte_flow*) is fully exposed through the
2142 ``flow`` command (validation, creation, destruction and queries).
2144 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2145 features simultaneously may cause undefined side-effects and is therefore
2151 Because the ``flow`` command uses dynamic tokens to handle the large number
2152 of possible flow rules combinations, its behavior differs slightly from
2153 other commands, in particular:
2155 - Pressing *?* or the *<tab>* key displays contextual help for the current
2156 token, not that of the entire command.
2158 - Optional and repeated parameters are supported (provided they are listed
2159 in the contextual help).
2161 The first parameter stands for the operation mode. Possible operations and
2162 their general syntax are described below. They are covered in detail in the
2165 - Check whether a flow rule can be created::
2167 flow validate {port_id}
2168 [group {group_id}] [priority {level}] [ingress] [egress]
2169 pattern {item} [/ {item} [...]] / end
2170 actions {action} [/ {action} [...]] / end
2172 - Create a flow rule::
2174 flow create {port_id}
2175 [group {group_id}] [priority {level}] [ingress] [egress]
2176 pattern {item} [/ {item} [...]] / end
2177 actions {action} [/ {action} [...]] / end
2179 - Destroy specific flow rules::
2181 flow destroy {port_id} rule {rule_id} [...]
2183 - Destroy all flow rules::
2185 flow flush {port_id}
2187 - Query an existing flow rule::
2189 flow query {port_id} {rule_id} {action}
2191 - List existing flow rules sorted by priority, filtered by group
2194 flow list {port_id} [group {group_id}] [...]
2196 Validating flow rules
2197 ~~~~~~~~~~~~~~~~~~~~~
2199 ``flow validate`` reports whether a flow rule would be accepted by the
2200 underlying device in its current state but stops short of creating it. It is
2201 bound to ``rte_flow_validate()``::
2203 flow validate {port_id}
2204 [group {group_id}] [priority {level}] [ingress] [egress]
2205 pattern {item} [/ {item} [...]] / end
2206 actions {action} [/ {action} [...]] / end
2208 If successful, it will show::
2212 Otherwise it will show an error message of the form::
2214 Caught error type [...] ([...]): [...]
2216 This command uses the same parameters as ``flow create``, their format is
2217 described in `Creating flow rules`_.
2219 Check whether redirecting any Ethernet packet received on port 0 to RX queue
2220 index 6 is supported::
2222 testpmd> flow validate 0 ingress pattern eth / end
2223 actions queue index 6 / end
2227 Port 0 does not support TCPv6 rules::
2229 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
2231 Caught error type 9 (specific pattern item): Invalid argument
2237 ``flow create`` validates and creates the specified flow rule. It is bound
2238 to ``rte_flow_create()``::
2240 flow create {port_id}
2241 [group {group_id}] [priority {level}] [ingress] [egress]
2242 pattern {item} [/ {item} [...]] / end
2243 actions {action} [/ {action} [...]] / end
2245 If successful, it will return a flow rule ID usable with other commands::
2247 Flow rule #[...] created
2249 Otherwise it will show an error message of the form::
2251 Caught error type [...] ([...]): [...]
2253 Parameters describe in the following order:
2255 - Attributes (*group*, *priority*, *ingress*, *egress* tokens).
2256 - A matching pattern, starting with the *pattern* token and terminated by an
2258 - Actions, starting with the *actions* token and terminated by an *end*
2261 These translate directly to *rte_flow* objects provided as-is to the
2262 underlying functions.
2264 The shortest valid definition only comprises mandatory tokens::
2266 testpmd> flow create 0 pattern end actions end
2268 Note that PMDs may refuse rules that essentially do nothing such as this
2271 **All unspecified object values are automatically initialized to 0.**
2276 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
2277 specified before the ``pattern`` token.
2279 - ``group {group id}``: priority group.
2280 - ``priority {level}``: priority level within group.
2281 - ``ingress``: rule applies to ingress traffic.
2282 - ``egress``: rule applies to egress traffic.
2284 Each instance of an attribute specified several times overrides the previous
2285 value as shown below (group 4 is used)::
2287 testpmd> flow create 0 group 42 group 24 group 4 [...]
2289 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
2291 While not specifying a direction is an error, some rules may allow both
2294 Most rules affect RX therefore contain the ``ingress`` token::
2296 testpmd> flow create 0 ingress pattern [...]
2301 A matching pattern starts after the ``pattern`` token. It is made of pattern
2302 items and is terminated by a mandatory ``end`` item.
2304 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
2305 rte_flow_item_type``).
2307 The ``/`` token is used as a separator between pattern items as shown
2310 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
2312 Note that protocol items like these must be stacked from lowest to highest
2313 layer to make sense. For instance, the following rule is either invalid or
2314 unlikely to match any packet::
2316 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
2318 More information on these restrictions can be found in the *rte_flow*
2321 Several items support additional specification structures, for example
2322 ``ipv4`` allows specifying source and destination addresses as follows::
2324 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
2325 dst is 10.2.0.0 / end [...]
2327 This rule matches all IPv4 traffic with the specified properties.
2329 In this example, ``src`` and ``dst`` are field names of the underlying
2330 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
2331 in a similar fashion.
2333 The ``is`` token means that the subsequent value must be matched exactly,
2334 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
2335 accordingly. Possible assignment tokens are:
2337 - ``is``: match value perfectly (with full bit-mask).
2338 - ``spec``: match value according to configured bit-mask.
2339 - ``last``: specify upper bound to establish a range.
2340 - ``mask``: specify bit-mask with relevant bits set to one.
2341 - ``prefix``: generate bit-mask from a prefix length.
2343 These yield identical results::
2345 ipv4 src is 10.1.1.1
2349 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
2353 ipv4 src spec 10.1.1.1 src prefix 32
2357 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
2361 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
2363 Inclusive ranges can be defined with ``last``::
2365 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
2367 Note that ``mask`` affects both ``spec`` and ``last``::
2369 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
2370 # matches 10.1.0.0 to 10.2.255.255
2372 Properties can be modified multiple times::
2374 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
2378 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
2383 This section lists supported pattern items and their attributes, if any.
2385 - ``end``: end list of pattern items.
2387 - ``void``: no-op pattern item.
2389 - ``invert``: perform actions when pattern does not match.
2391 - ``any``: match any protocol for the current layer.
2393 - ``num {unsigned}``: number of layers covered.
2395 - ``pf``: match packets addressed to the physical function.
2397 - ``vf``: match packets addressed to a virtual function ID.
2399 - ``id {unsigned}``: destination VF ID.
2401 - ``port``: device-specific physical port index to use.
2403 - ``index {unsigned}``: physical port index.
2405 - ``raw``: match an arbitrary byte string.
2407 - ``relative {boolean}``: look for pattern after the previous item.
2408 - ``search {boolean}``: search pattern from offset (see also limit).
2409 - ``offset {integer}``: absolute or relative offset for pattern.
2410 - ``limit {unsigned}``: search area limit for start of pattern.
2411 - ``pattern {string}``: byte string to look for.
2413 - ``eth``: match Ethernet header.
2415 - ``dst {MAC-48}``: destination MAC.
2416 - ``src {MAC-48}``: source MAC.
2417 - ``type {unsigned}``: EtherType.
2419 - ``vlan``: match 802.1Q/ad VLAN tag.
2421 - ``tpid {unsigned}``: tag protocol identifier.
2422 - ``tci {unsigned}``: tag control information.
2423 - ``pcp {unsigned}``: priority code point.
2424 - ``dei {unsigned}``: drop eligible indicator.
2425 - ``vid {unsigned}``: VLAN identifier.
2427 - ``ipv4``: match IPv4 header.
2429 - ``tos {unsigned}``: type of service.
2430 - ``ttl {unsigned}``: time to live.
2431 - ``proto {unsigned}``: next protocol ID.
2432 - ``src {ipv4 address}``: source address.
2433 - ``dst {ipv4 address}``: destination address.
2435 - ``ipv6``: match IPv6 header.
2437 - ``tc {unsigned}``: traffic class.
2438 - ``flow {unsigned}``: flow label.
2439 - ``proto {unsigned}``: protocol (next header).
2440 - ``hop {unsigned}``: hop limit.
2441 - ``src {ipv6 address}``: source address.
2442 - ``dst {ipv6 address}``: destination address.
2444 - ``icmp``: match ICMP header.
2446 - ``type {unsigned}``: ICMP packet type.
2447 - ``code {unsigned}``: ICMP packet code.
2449 - ``udp``: match UDP header.
2451 - ``src {unsigned}``: UDP source port.
2452 - ``dst {unsigned}``: UDP destination port.
2454 - ``tcp``: match TCP header.
2456 - ``src {unsigned}``: TCP source port.
2457 - ``dst {unsigned}``: TCP destination port.
2459 - ``sctp``: match SCTP header.
2461 - ``src {unsigned}``: SCTP source port.
2462 - ``dst {unsigned}``: SCTP destination port.
2463 - ``tag {unsigned}``: validation tag.
2464 - ``cksum {unsigned}``: checksum.
2466 - ``vxlan``: match VXLAN header.
2468 - ``vni {unsigned}``: VXLAN identifier.
2473 A list of actions starts after the ``actions`` token in the same fashion as
2474 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
2475 terminated by a mandatory ``end`` action.
2477 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
2478 rte_flow_action_type``).
2480 Dropping all incoming UDPv4 packets can be expressed as follows::
2482 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2485 Several actions have configurable properties which must be specified when
2486 there is no valid default value. For example, ``queue`` requires a target
2489 This rule redirects incoming UDPv4 traffic to queue index 6::
2491 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2492 actions queue index 6 / end
2494 While this one could be rejected by PMDs (unspecified queue index)::
2496 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2499 As defined by *rte_flow*, the list is not ordered, all actions of a given
2500 rule are performed simultaneously. These are equivalent::
2502 queue index 6 / void / mark id 42 / end
2506 void / mark id 42 / queue index 6 / end
2508 All actions in a list should have different types, otherwise only the last
2509 action of a given type is taken into account::
2511 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
2515 drop / drop / drop / end # drop is performed only once
2519 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
2521 Considering they are performed simultaneously, opposite and overlapping
2522 actions can sometimes be combined when the end result is unambiguous::
2524 drop / queue index 6 / end # drop has no effect
2528 drop / dup index 6 / end # same as above
2532 queue index 6 / rss queues 6 7 8 / end # queue has no effect
2536 drop / passthru / end # drop has no effect
2538 Note that PMDs may still refuse such combinations.
2543 This section lists supported actions and their attributes, if any.
2545 - ``end``: end list of actions.
2547 - ``void``: no-op action.
2549 - ``passthru``: let subsequent rule process matched packets.
2551 - ``mark``: attach 32 bit value to packets.
2553 - ``id {unsigned}``: 32 bit value to return with packets.
2555 - ``flag``: flag packets.
2557 - ``queue``: assign packets to a given queue index.
2559 - ``index {unsigned}``: queue index to use.
2561 - ``drop``: drop packets (note: passthru has priority).
2563 - ``count``: enable counters for this rule.
2565 - ``dup``: duplicate packets to a given queue index.
2567 - ``index {unsigned}``: queue index to duplicate packets to.
2569 - ``rss``: spread packets among several queues.
2571 - ``queues [{unsigned} [...]] end``: queue indices to use.
2573 - ``pf``: redirect packets to physical device function.
2575 - ``vf``: redirect packets to virtual device function.
2577 - ``original {boolean}``: use original VF ID if possible.
2578 - ``id {unsigned}``: VF ID to redirect packets to.
2580 Destroying flow rules
2581 ~~~~~~~~~~~~~~~~~~~~~
2583 ``flow destroy`` destroys one or more rules from their rule ID (as returned
2584 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
2585 times as necessary::
2587 flow destroy {port_id} rule {rule_id} [...]
2589 If successful, it will show::
2591 Flow rule #[...] destroyed
2593 It does not report anything for rule IDs that do not exist. The usual error
2594 message is shown when a rule cannot be destroyed::
2596 Caught error type [...] ([...]): [...]
2598 ``flow flush`` destroys all rules on a device and does not take extra
2599 arguments. It is bound to ``rte_flow_flush()``::
2601 flow flush {port_id}
2603 Any errors are reported as above.
2605 Creating several rules and destroying them::
2607 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2608 actions queue index 2 / end
2609 Flow rule #0 created
2610 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2611 actions queue index 3 / end
2612 Flow rule #1 created
2613 testpmd> flow destroy 0 rule 0 rule 1
2614 Flow rule #1 destroyed
2615 Flow rule #0 destroyed
2618 The same result can be achieved using ``flow flush``::
2620 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2621 actions queue index 2 / end
2622 Flow rule #0 created
2623 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2624 actions queue index 3 / end
2625 Flow rule #1 created
2626 testpmd> flow flush 0
2629 Non-existent rule IDs are ignored::
2631 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2632 actions queue index 2 / end
2633 Flow rule #0 created
2634 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2635 actions queue index 3 / end
2636 Flow rule #1 created
2637 testpmd> flow destroy 0 rule 42 rule 10 rule 2
2639 testpmd> flow destroy 0 rule 0
2640 Flow rule #0 destroyed
2646 ``flow query`` queries a specific action of a flow rule having that
2647 ability. Such actions collect information that can be reported using this
2648 command. It is bound to ``rte_flow_query()``::
2650 flow query {port_id} {rule_id} {action}
2652 If successful, it will display either the retrieved data for known actions
2653 or the following message::
2655 Cannot display result for action type [...] ([...])
2657 Otherwise, it will complain either that the rule does not exist or that some
2660 Flow rule #[...] not found
2664 Caught error type [...] ([...]): [...]
2666 Currently only the ``count`` action is supported. This action reports the
2667 number of packets that hit the flow rule and the total number of bytes. Its
2668 output has the following format::
2671 hits_set: [...] # whether "hits" contains a valid value
2672 bytes_set: [...] # whether "bytes" contains a valid value
2673 hits: [...] # number of packets
2674 bytes: [...] # number of bytes
2676 Querying counters for TCPv6 packets redirected to queue 6::
2678 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
2679 actions queue index 6 / count / end
2680 Flow rule #4 created
2681 testpmd> flow query 0 4 count
2692 ``flow list`` lists existing flow rules sorted by priority and optionally
2693 filtered by group identifiers::
2695 flow list {port_id} [group {group_id}] [...]
2697 This command only fails with the following message if the device does not
2702 Output consists of a header line followed by a short description of each
2703 flow rule, one per line. There is no output at all when no flow rules are
2704 configured on the device::
2706 ID Group Prio Attr Rule
2707 [...] [...] [...] [...] [...]
2709 ``Attr`` column flags:
2711 - ``i`` for ``ingress``.
2712 - ``e`` for ``egress``.
2714 Creating several flow rules and listing them::
2716 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2717 actions queue index 6 / end
2718 Flow rule #0 created
2719 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2720 actions queue index 2 / end
2721 Flow rule #1 created
2722 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
2723 actions rss queues 6 7 8 end / end
2724 Flow rule #2 created
2725 testpmd> flow list 0
2726 ID Group Prio Attr Rule
2727 0 0 0 i- ETH IPV4 => QUEUE
2728 1 0 0 i- ETH IPV6 => QUEUE
2729 2 0 5 i- ETH IPV4 UDP => RSS
2732 Rules are sorted by priority (i.e. group ID first, then priority level)::
2734 testpmd> flow list 1
2735 ID Group Prio Attr Rule
2736 0 0 0 i- ETH => COUNT
2737 6 0 500 i- ETH IPV6 TCP => DROP COUNT
2738 5 0 1000 i- ETH IPV6 ICMP => QUEUE
2739 1 24 0 i- ETH IPV4 UDP => QUEUE
2740 4 24 10 i- ETH IPV4 TCP => DROP
2741 3 24 20 i- ETH IPV4 => DROP
2742 2 24 42 i- ETH IPV4 UDP => QUEUE
2743 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
2746 Output can be limited to specific groups::
2748 testpmd> flow list 1 group 0 group 63
2749 ID Group Prio Attr Rule
2750 0 0 0 i- ETH => COUNT
2751 6 0 500 i- ETH IPV6 TCP => DROP COUNT
2752 5 0 1000 i- ETH IPV6 ICMP => QUEUE
2753 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE