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)
906 Set the link flow control parameter on a port::
908 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
909 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
910 autoneg (on|off) (port_id)
914 * ``high_water`` (integer): High threshold value to trigger XOFF.
916 * ``low_water`` (integer): Low threshold value to trigger XON.
918 * ``pause_time`` (integer): Pause quota in the Pause frame.
920 * ``send_xon`` (0/1): Send XON frame.
922 * ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
924 * ``autoneg``: Change the auto-negotiation parameter.
929 Set the priority flow control parameter on a port::
931 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
932 (pause_time) (priority) (port_id)
936 * ``high_water`` (integer): High threshold value.
938 * ``low_water`` (integer): Low threshold value.
940 * ``pause_time`` (integer): Pause quota in the Pause frame.
942 * ``priority`` (0-7): VLAN User Priority.
947 Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
949 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
951 For example, to set rx queue 2 on port 0 to mapping 5::
953 testpmd>set stat_qmap rx 0 2 5
955 set port - rx/tx (for VF)
956 ~~~~~~~~~~~~~~~~~~~~~~~~~
958 Set VF receive/transmit from a port::
960 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
962 set port - mac address filter (for VF)
963 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
965 Add/Remove unicast or multicast MAC addr filter for a VF::
967 testpmd> set port (port_id) vf (vf_id) (mac_addr) \
968 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
970 set port - rx mode(for VF)
971 ~~~~~~~~~~~~~~~~~~~~~~~~~~
973 Set the VF receive mode of a port::
975 testpmd> set port (port_id) vf (vf_id) \
976 rxmode (AUPE|ROPE|BAM|MPE) (on|off)
978 The available receive modes are:
980 * ``AUPE``: Accepts untagged VLAN.
982 * ``ROPE``: Accepts unicast hash.
984 * ``BAM``: Accepts broadcast packets.
986 * ``MPE``: Accepts all multicast packets.
988 set port - tx_rate (for Queue)
989 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
991 Set TX rate limitation for a queue on a port::
993 testpmd> set port (port_id) queue (queue_id) rate (rate_value)
995 set port - tx_rate (for VF)
996 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
998 Set TX rate limitation for queues in VF on a port::
1000 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1002 set port - mirror rule
1003 ~~~~~~~~~~~~~~~~~~~~~~
1005 Set pool or vlan type mirror rule for a port::
1007 testpmd> set port (port_id) mirror-rule (rule_id) \
1008 (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1009 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1011 Set link mirror rule for a port::
1013 testpmd> set port (port_id) mirror-rule (rule_id) \
1014 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1016 For example to enable mirror traffic with vlan 0,1 to pool 0::
1018 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1020 reset port - mirror rule
1021 ~~~~~~~~~~~~~~~~~~~~~~~~
1023 Reset a mirror rule for a port::
1025 testpmd> reset port (port_id) mirror-rule (rule_id)
1030 Set the flush on RX streams before forwarding.
1031 The default is flush ``on``.
1032 Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1034 testpmd> set flush_rx off
1039 Set the bypass mode for the lowest port on bypass enabled NIC::
1041 testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1046 Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1048 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1049 mode (normal|bypass|isolate) (port_id)
1053 * ``timeout``: Enable bypass after watchdog timeout.
1055 * ``os_on``: Enable bypass when OS/board is powered on.
1057 * ``os_off``: Enable bypass when OS/board is powered off.
1059 * ``power_on``: Enable bypass when power supply is turned on.
1061 * ``power_off``: Enable bypass when power supply is turned off.
1067 Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1069 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1074 Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1076 testpmd> show bypass config (port_id)
1081 Set link up for a port::
1083 testpmd> set link-up port (port id)
1088 Set link down for a port::
1090 testpmd> set link-down port (port id)
1095 Enable E-tag insertion for a VF on a port::
1097 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1099 Disable E-tag insertion for a VF on a port::
1101 testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1103 Enable/disable E-tag stripping on a port::
1105 testpmd> E-tag set stripping (on|off) port (port_id)
1107 Enable/disable E-tag based forwarding on a port::
1109 testpmd> E-tag set forwarding (on|off) port (port_id)
1111 Add an E-tag forwarding filter on a port::
1113 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1115 Delete an E-tag forwarding filter on a port::
1116 testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1122 The following sections show functions for configuring ports.
1126 Port configuration changes only become active when forwarding is started/restarted.
1131 Attach a port specified by pci address or virtual device args::
1133 testpmd> port attach (identifier)
1135 To attach a new pci device, the device should be recognized by kernel first.
1136 Then it should be moved under DPDK management.
1137 Finally the port can be attached to testpmd.
1139 For example, to move a pci device using ixgbe under DPDK management:
1141 .. code-block:: console
1143 # Check the status of the available devices.
1144 ./usertools/dpdk-devbind.py --status
1146 Network devices using DPDK-compatible driver
1147 ============================================
1150 Network devices using kernel driver
1151 ===================================
1152 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1155 # Bind the device to igb_uio.
1156 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1159 # Recheck the status of the devices.
1160 ./usertools/dpdk-devbind.py --status
1161 Network devices using DPDK-compatible driver
1162 ============================================
1163 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1165 To attach a port created by virtual device, above steps are not needed.
1167 For example, to attach a port whose pci address is 0000:0a:00.0.
1169 .. code-block:: console
1171 testpmd> port attach 0000:0a:00.0
1172 Attaching a new port...
1173 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1174 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1175 EAL: PCI memory mapped at 0x7f83bfa00000
1176 EAL: PCI memory mapped at 0x7f83bfa80000
1177 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1178 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1179 Port 0 is attached. Now total ports is 1
1182 For example, to attach a port created by pcap PMD.
1184 .. code-block:: console
1186 testpmd> port attach net_pcap0
1187 Attaching a new port...
1188 PMD: Initializing pmd_pcap for net_pcap0
1189 PMD: Creating pcap-backed ethdev on numa socket 0
1190 Port 0 is attached. Now total ports is 1
1193 In this case, identifier is ``net_pcap0``.
1194 This identifier format is the same as ``--vdev`` format of DPDK applications.
1196 For example, to re-attach a bonded port which has been previously detached,
1197 the mode and slave parameters must be given.
1199 .. code-block:: console
1201 testpmd> port attach net_bond_0,mode=0,slave=1
1202 Attaching a new port...
1203 EAL: Initializing pmd_bond for net_bond_0
1204 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1205 Port 0 is attached. Now total ports is 1
1212 Detach a specific port::
1214 testpmd> port detach (port_id)
1216 Before detaching a port, the port should be stopped and closed.
1218 For example, to detach a pci device port 0.
1220 .. code-block:: console
1222 testpmd> port stop 0
1225 testpmd> port close 0
1229 testpmd> port detach 0
1231 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1232 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1233 EAL: PCI memory unmapped at 0x7f83bfa00000
1234 EAL: PCI memory unmapped at 0x7f83bfa80000
1238 For example, to detach a virtual device port 0.
1240 .. code-block:: console
1242 testpmd> port stop 0
1245 testpmd> port close 0
1249 testpmd> port detach 0
1251 PMD: Closing pcap ethdev on numa socket 0
1252 Port 'net_pcap0' is detached. Now total ports is 0
1255 To remove a pci device completely from the system, first detach the port from testpmd.
1256 Then the device should be moved under kernel management.
1257 Finally the device can be removed using kernel pci hotplug functionality.
1259 For example, to move a pci device under kernel management:
1261 .. code-block:: console
1263 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1265 ./usertools/dpdk-devbind.py --status
1267 Network devices using DPDK-compatible driver
1268 ============================================
1271 Network devices using kernel driver
1272 ===================================
1273 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1275 To remove a port created by a virtual device, above steps are not needed.
1280 Start all ports or a specific port::
1282 testpmd> port start (port_id|all)
1287 Stop all ports or a specific port::
1289 testpmd> port stop (port_id|all)
1294 Close all ports or a specific port::
1296 testpmd> port close (port_id|all)
1298 port start/stop queue
1299 ~~~~~~~~~~~~~~~~~~~~~
1301 Start/stop a rx/tx queue on a specific port::
1303 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1305 Only take effect when port is started.
1310 Set the speed and duplex mode for all ports or a specific port::
1312 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1313 duplex (half|full|auto)
1315 port config - queues/descriptors
1316 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1318 Set number of queues/descriptors for rxq, txq, rxd and txd::
1320 testpmd> port config all (rxq|txq|rxd|txd) (value)
1322 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1324 port config - max-pkt-len
1325 ~~~~~~~~~~~~~~~~~~~~~~~~~
1327 Set the maximum packet length::
1329 testpmd> port config all max-pkt-len (value)
1331 This is equivalent to the ``--max-pkt-len`` command-line option.
1333 port config - CRC Strip
1334 ~~~~~~~~~~~~~~~~~~~~~~~
1336 Set hardware CRC stripping on or off for all ports::
1338 testpmd> port config all crc-strip (on|off)
1340 CRC stripping is off by default.
1342 The ``on`` option is equivalent to the ``--crc-strip`` command-line option.
1344 port config - scatter
1345 ~~~~~~~~~~~~~~~~~~~~~~~
1347 Set RX scatter mode on or off for all ports::
1349 testpmd> port config all scatter (on|off)
1351 RX scatter mode is off by default.
1353 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1355 port config - TX queue flags
1356 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1358 Set a hexadecimal bitmap of TX queue flags for all ports::
1360 testpmd> port config all txqflags value
1362 This command is equivalent to the ``--txqflags`` command-line option.
1364 port config - RX Checksum
1365 ~~~~~~~~~~~~~~~~~~~~~~~~~
1367 Set hardware RX checksum offload to on or off for all ports::
1369 testpmd> port config all rx-cksum (on|off)
1371 Checksum offload is off by default.
1373 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1378 Set hardware VLAN on or off for all ports::
1380 testpmd> port config all hw-vlan (on|off)
1382 Hardware VLAN is on by default.
1384 The ``off`` option is equivalent to the ``--disable-hw-vlan`` command-line option.
1386 port config - VLAN filter
1387 ~~~~~~~~~~~~~~~~~~~~~~~~~
1389 Set hardware VLAN filter on or off for all ports::
1391 testpmd> port config all hw-vlan-filter (on|off)
1393 Hardware VLAN filter is on by default.
1395 The ``off`` option is equivalent to the ``--disable-hw-vlan-filter`` command-line option.
1397 port config - VLAN strip
1398 ~~~~~~~~~~~~~~~~~~~~~~~~
1400 Set hardware VLAN strip on or off for all ports::
1402 testpmd> port config all hw-vlan-strip (on|off)
1404 Hardware VLAN strip is on by default.
1406 The ``off`` option is equivalent to the ``--disable-hw-vlan-strip`` command-line option.
1408 port config - VLAN extend
1409 ~~~~~~~~~~~~~~~~~~~~~~~~~
1411 Set hardware VLAN extend on or off for all ports::
1413 testpmd> port config all hw-vlan-extend (on|off)
1415 Hardware VLAN extend is off by default.
1417 The ``off`` option is equivalent to the ``--disable-hw-vlan-extend`` command-line option.
1419 port config - Drop Packets
1420 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1422 Set packet drop for packets with no descriptors on or off for all ports::
1424 testpmd> port config all drop-en (on|off)
1426 Packet dropping for packets with no descriptors is off by default.
1428 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1433 Set the RSS (Receive Side Scaling) mode on or off::
1435 testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1437 RSS is on by default.
1439 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1441 port config - RSS Reta
1442 ~~~~~~~~~~~~~~~~~~~~~~
1444 Set the RSS (Receive Side Scaling) redirection table::
1446 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1451 Set the DCB mode for an individual port::
1453 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1455 The traffic class should be 4 or 8.
1460 Set the number of packets per burst::
1462 testpmd> port config all burst (value)
1464 This is equivalent to the ``--burst`` command-line option.
1466 port config - Threshold
1467 ~~~~~~~~~~~~~~~~~~~~~~~
1469 Set thresholds for TX/RX queues::
1471 testpmd> port config all (threshold) (value)
1473 Where the threshold type can be:
1475 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1477 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1479 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1481 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1483 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1485 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1487 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1489 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1491 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1493 These threshold options are also available from the command-line.
1498 Set the value of ether-type for E-tag::
1500 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1502 Enable/disable the E-tag support::
1504 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1507 Link Bonding Functions
1508 ----------------------
1510 The Link Bonding functions make it possible to dynamically create and
1511 manage link bonding devices from within testpmd interactive prompt.
1513 create bonded device
1514 ~~~~~~~~~~~~~~~~~~~~
1516 Create a new bonding device::
1518 testpmd> create bonded device (mode) (socket)
1520 For example, to create a bonded device in mode 1 on socket 0::
1522 testpmd> create bonded 1 0
1523 created new bonded device (port X)
1528 Adds Ethernet device to a Link Bonding device::
1530 testpmd> add bonding slave (slave id) (port id)
1532 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
1534 testpmd> add bonding slave 6 10
1537 remove bonding slave
1538 ~~~~~~~~~~~~~~~~~~~~
1540 Removes an Ethernet slave device from a Link Bonding device::
1542 testpmd> remove bonding slave (slave id) (port id)
1544 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
1546 testpmd> remove bonding slave 6 10
1551 Set the Link Bonding mode of a Link Bonding device::
1553 testpmd> set bonding mode (value) (port id)
1555 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
1557 testpmd> set bonding mode 3 10
1562 Set an Ethernet slave device as the primary device on a Link Bonding device::
1564 testpmd> set bonding primary (slave id) (port id)
1566 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
1568 testpmd> set bonding primary 6 10
1573 Set the MAC address of a Link Bonding device::
1575 testpmd> set bonding mac (port id) (mac)
1577 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
1579 testpmd> set bonding mac 10 00:00:00:00:00:01
1581 set bonding xmit_balance_policy
1582 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1584 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
1586 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
1588 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
1590 testpmd> set bonding xmit_balance_policy 10 l34
1593 set bonding mon_period
1594 ~~~~~~~~~~~~~~~~~~~~~~
1596 Set the link status monitoring polling period in milliseconds for a bonding device.
1598 This adds support for PMD slave devices which do not support link status interrupts.
1599 When the mon_period is set to a value greater than 0 then all PMD's which do not support
1600 link status ISR will be queried every polling interval to check if their link status has changed::
1602 testpmd> set bonding mon_period (port_id) (value)
1604 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
1606 testpmd> set bonding mon_period 5 150
1612 Show the current configuration of a Link Bonding device::
1614 testpmd> show bonding config (port id)
1617 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
1618 in balance mode with a transmission policy of layer 2+3::
1620 testpmd> show bonding config 9
1622 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
1624 Active Slaves (3): [1 3 4]
1631 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
1632 This is mainly useful for debugging purposes.
1633 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
1634 and fields that can be accessed.
1639 Display the value of a port register::
1641 testpmd> read reg (port_id) (address)
1643 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
1645 testpmd> read reg 0 0xEE00
1646 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
1651 Display a port register bit field::
1653 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
1655 For example, reading the lowest two bits from the register in the example above::
1657 testpmd> read regfield 0 0xEE00 0 1
1658 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
1663 Display a single port register bit::
1665 testpmd> read regbit (port_id) (address) (bit_x)
1667 For example, reading the lowest bit from the register in the example above::
1669 testpmd> read regbit 0 0xEE00 0
1670 port 0 PCI register at offset 0xEE00: bit 0=1
1675 Set the value of a port register::
1677 testpmd> write reg (port_id) (address) (value)
1679 For example, to clear a register::
1681 testpmd> write reg 0 0xEE00 0x0
1682 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
1687 Set bit field of a port register::
1689 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
1691 For example, writing to the register cleared in the example above::
1693 testpmd> write regfield 0 0xEE00 0 1 2
1694 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
1699 Set single bit value of a port register::
1701 testpmd> write regbit (port_id) (address) (bit_x) (value)
1703 For example, to set the high bit in the register from the example above::
1705 testpmd> write regbit 0 0xEE00 31 1
1706 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
1712 This section details the available filter functions that are available.
1714 Note these functions interface the deprecated legacy filtering framework,
1715 superseded by *rte_flow*. See `Flow rules management`_.
1718 ~~~~~~~~~~~~~~~~~~~~
1720 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
1722 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
1723 ethertype (ether_type) (drop|fwd) queue (queue_id)
1725 The available information parameters are:
1727 * ``port_id``: The port which the Ethertype filter assigned on.
1729 * ``mac_addr``: Compare destination mac address.
1731 * ``mac_ignr``: Ignore destination mac address match.
1733 * ``mac_address``: Destination mac address to match.
1735 * ``ether_type``: The EtherType value want to match,
1736 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
1738 * ``queue_id``: The receive queue associated with this EtherType filter.
1739 It is meaningless when deleting or dropping.
1741 Example, to add/remove an ethertype filter rule::
1743 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
1744 ethertype 0x0806 fwd queue 3
1746 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
1747 ethertype 0x0806 fwd queue 3
1752 Add or delete a 2-tuple filter,
1753 which identifies packets by specific protocol and destination TCP/UDP port
1754 and forwards packets into one of the receive queues::
1756 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
1757 protocol (protocol_value) mask (mask_value) \
1758 tcp_flags (tcp_flags_value) priority (prio_value) \
1761 The available information parameters are:
1763 * ``port_id``: The port which the 2-tuple filter assigned on.
1765 * ``dst_port_value``: Destination port in L4.
1767 * ``protocol_value``: IP L4 protocol.
1769 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
1771 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
1773 * ``prio_value``: Priority of this filter.
1775 * ``queue_id``: The receive queue associated with this 2-tuple filter.
1777 Example, to add/remove an 2tuple filter rule::
1779 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
1780 tcp_flags 0x02 priority 3 queue 3
1782 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
1783 tcp_flags 0x02 priority 3 queue 3
1788 Add or delete a 5-tuple filter,
1789 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
1790 and routes packets into one of the receive queues::
1792 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
1793 (src_address) dst_port (dst_port_value) \
1794 src_port (src_port_value) protocol (protocol_value) \
1795 mask (mask_value) tcp_flags (tcp_flags_value) \
1796 priority (prio_value) queue (queue_id)
1798 The available information parameters are:
1800 * ``port_id``: The port which the 5-tuple filter assigned on.
1802 * ``dst_address``: Destination IP address.
1804 * ``src_address``: Source IP address.
1806 * ``dst_port_value``: TCP/UDP destination port.
1808 * ``src_port_value``: TCP/UDP source port.
1810 * ``protocol_value``: L4 protocol.
1812 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
1814 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
1816 * ``prio_value``: The priority of this filter.
1818 * ``queue_id``: The receive queue associated with this 5-tuple filter.
1820 Example, to add/remove an 5tuple filter rule::
1822 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1823 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1824 flags 0x0 priority 3 queue 3
1826 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1827 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1828 flags 0x0 priority 3 queue 3
1833 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
1835 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
1837 The available information parameters are:
1839 * ``port_id``: The port which the SYN filter assigned on.
1841 * ``high``: This SYN filter has higher priority than other filters.
1843 * ``low``: This SYN filter has lower priority than other filters.
1845 * ``queue_id``: The receive queue associated with this SYN filter
1849 testpmd> syn_filter 0 add priority high queue 3
1854 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
1855 and routed into one of the receive queues::
1857 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
1858 mask (mask_value) priority (prio_value) queue (queue_id)
1860 The available information parameters are:
1862 * ``port_id``: The port which the Flex filter is assigned on.
1864 * ``len_value``: Filter length in bytes, no greater than 128.
1866 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
1868 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
1870 * ``prio_value``: The priority of this filter.
1872 * ``queue_id``: The receive queue associated with this Flex filter.
1876 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
1877 mask 000C priority 3 queue 3
1879 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
1880 mask 000C priority 3 queue 3
1883 .. _testpmd_flow_director:
1885 flow_director_filter
1886 ~~~~~~~~~~~~~~~~~~~~
1888 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
1890 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
1891 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
1893 * Perfect match filters.
1894 The hardware checks a match between the masked fields of the received packets and the programmed filters.
1895 The masked fields are for IP flow.
1897 * Signature filters.
1898 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
1900 * Perfect-mac-vlan 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 MAC VLAN flow.
1904 * Perfect-tunnel match filters.
1905 The hardware checks a match between the masked fields of the received packets and the programmed filters.
1906 The masked fields are for tunnel flow.
1908 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
1909 per flow type and the flexible payload.
1911 The Flow Director can also mask out parts of all of these fields so that filters
1912 are only applied to certain fields or parts of the fields.
1914 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
1916 # Commands to add flow director filters of different flow types::
1918 flow_director_filter (port_id) mode IP (add|del|update) \
1919 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
1920 src (src_ip_address) dst (dst_ip_address) \
1921 tos (tos_value) proto (proto_value) ttl (ttl_value) \
1922 vlan (vlan_value) flexbytes (flexbytes_value) \
1923 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
1926 flow_director_filter (port_id) mode IP (add|del|update) \
1927 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
1928 src (src_ip_address) (src_port) \
1929 dst (dst_ip_address) (dst_port) \
1930 tos (tos_value) ttl (ttl_value) \
1931 vlan (vlan_value) flexbytes (flexbytes_value) \
1932 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
1935 flow_director_filter (port_id) mode IP (add|del|update) \
1936 flow (ipv4-sctp|ipv6-sctp) \
1937 src (src_ip_address) (src_port) \
1938 dst (dst_ip_address) (dst_port) \
1939 tos (tos_value) ttl (ttl_value) \
1940 tag (verification_tag) vlan (vlan_value) \
1941 flexbytes (flexbytes_value) (drop|fwd) \
1942 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
1944 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
1945 ether (ethertype) flexbytes (flexbytes_value) \
1946 (drop|fwd) pf|vf(vf_id) queue (queue_id)
1949 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
1950 mac (mac_address) vlan (vlan_value) \
1951 flexbytes (flexbytes_value) (drop|fwd) \
1952 queue (queue_id) fd_id (fd_id_value)
1954 flow_director_filter (port_id) mode Tunnel (add|del|update) \
1955 mac (mac_address) vlan (vlan_value) \
1956 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
1957 flexbytes (flexbytes_value) (drop|fwd) \
1958 queue (queue_id) fd_id (fd_id_value)
1960 For example, to add an ipv4-udp flow type filter::
1962 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
1963 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
1964 fwd pf queue 1 fd_id 1
1966 For example, add an ipv4-other flow type filter::
1968 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
1969 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
1970 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
1975 Flush all flow director filters on a device::
1977 testpmd> flush_flow_director (port_id)
1979 Example, to flush all flow director filter on port 0::
1981 testpmd> flush_flow_director 0
1986 Set flow director's input masks::
1988 flow_director_mask (port_id) mode IP vlan (vlan_value) \
1989 src_mask (ipv4_src) (ipv6_src) (src_port) \
1990 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
1992 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
1994 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
1995 mac (mac_value) tunnel-type (tunnel_type_value) \
1996 tunnel-id (tunnel_id_value)
1998 Example, to set flow director mask on port 0::
2000 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2001 src_mask 255.255.255.255 \
2002 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2003 dst_mask 255.255.255.255 \
2004 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2006 flow_director_flex_mask
2007 ~~~~~~~~~~~~~~~~~~~~~~~
2009 set masks of flow director's flexible payload based on certain flow type::
2011 testpmd> flow_director_flex_mask (port_id) \
2012 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2013 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2014 l2_payload|all) (mask)
2016 Example, to set flow director's flex mask for all flow type on port 0::
2018 testpmd> flow_director_flex_mask 0 flow all \
2019 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2022 flow_director_flex_payload
2023 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2025 Configure flexible payload selection::
2027 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2029 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2031 testpmd> flow_director_flex_payload 0 l4 \
2032 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2034 get_sym_hash_ena_per_port
2035 ~~~~~~~~~~~~~~~~~~~~~~~~~
2037 Get symmetric hash enable configuration per port::
2039 get_sym_hash_ena_per_port (port_id)
2041 For example, to get symmetric hash enable configuration of port 1::
2043 testpmd> get_sym_hash_ena_per_port 1
2045 set_sym_hash_ena_per_port
2046 ~~~~~~~~~~~~~~~~~~~~~~~~~
2048 Set symmetric hash enable configuration per port to enable or disable::
2050 set_sym_hash_ena_per_port (port_id) (enable|disable)
2052 For example, to set symmetric hash enable configuration of port 1 to enable::
2054 testpmd> set_sym_hash_ena_per_port 1 enable
2056 get_hash_global_config
2057 ~~~~~~~~~~~~~~~~~~~~~~
2059 Get the global configurations of hash filters::
2061 get_hash_global_config (port_id)
2063 For example, to get the global configurations of hash filters of port 1::
2065 testpmd> get_hash_global_config 1
2067 set_hash_global_config
2068 ~~~~~~~~~~~~~~~~~~~~~~
2070 Set the global configurations of hash filters::
2072 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2073 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2074 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload) \
2077 For example, to enable simple_xor for flow type of ipv6 on port 2::
2079 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2084 Set the input set for hash::
2086 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2087 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2088 l2_payload) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2089 ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2090 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2091 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2092 fld-8th|none) (select|add)
2094 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2096 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2101 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2102 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2104 Set the input set for flow director::
2106 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2107 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2108 l2_payload) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2109 ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2110 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2111 sctp-dst-port|sctp-veri-tag|none) (select|add)
2113 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2115 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2120 Set different GRE key length for input set::
2122 global_config (port_id) gre-key-len (number in bytes)
2124 For example to set GRE key length for input set to 4 bytes on port 0::
2126 testpmd> global_config 0 gre-key-len 4
2129 .. _testpmd_rte_flow:
2131 Flow rules management
2132 ---------------------
2134 Control of the generic flow API (*rte_flow*) is fully exposed through the
2135 ``flow`` command (validation, creation, destruction and queries).
2137 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2138 features simultaneously may cause undefined side-effects and is therefore
2144 Because the ``flow`` command uses dynamic tokens to handle the large number
2145 of possible flow rules combinations, its behavior differs slightly from
2146 other commands, in particular:
2148 - Pressing *?* or the *<tab>* key displays contextual help for the current
2149 token, not that of the entire command.
2151 - Optional and repeated parameters are supported (provided they are listed
2152 in the contextual help).
2154 The first parameter stands for the operation mode. Possible operations and
2155 their general syntax are described below. They are covered in detail in the
2158 - Check whether a flow rule can be created::
2160 flow validate {port_id}
2161 [group {group_id}] [priority {level}] [ingress] [egress]
2162 pattern {item} [/ {item} [...]] / end
2163 actions {action} [/ {action} [...]] / end
2165 - Create a flow rule::
2167 flow create {port_id}
2168 [group {group_id}] [priority {level}] [ingress] [egress]
2169 pattern {item} [/ {item} [...]] / end
2170 actions {action} [/ {action} [...]] / end
2172 - Destroy specific flow rules::
2174 flow destroy {port_id} rule {rule_id} [...]
2176 - Destroy all flow rules::
2178 flow flush {port_id}
2180 - Query an existing flow rule::
2182 flow query {port_id} {rule_id} {action}
2184 - List existing flow rules sorted by priority, filtered by group
2187 flow list {port_id} [group {group_id}] [...]
2189 Validating flow rules
2190 ~~~~~~~~~~~~~~~~~~~~~
2192 ``flow validate`` reports whether a flow rule would be accepted by the
2193 underlying device in its current state but stops short of creating it. It is
2194 bound to ``rte_flow_validate()``::
2196 flow validate {port_id}
2197 [group {group_id}] [priority {level}] [ingress] [egress]
2198 pattern {item} [/ {item} [...]] / end
2199 actions {action} [/ {action} [...]] / end
2201 If successful, it will show::
2205 Otherwise it will show an error message of the form::
2207 Caught error type [...] ([...]): [...]
2209 This command uses the same parameters as ``flow create``, their format is
2210 described in `Creating flow rules`_.
2212 Check whether redirecting any Ethernet packet received on port 0 to RX queue
2213 index 6 is supported::
2215 testpmd> flow validate 0 ingress pattern eth / end
2216 actions queue index 6 / end
2220 Port 0 does not support TCPv6 rules::
2222 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
2224 Caught error type 9 (specific pattern item): Invalid argument
2230 ``flow create`` validates and creates the specified flow rule. It is bound
2231 to ``rte_flow_create()``::
2233 flow create {port_id}
2234 [group {group_id}] [priority {level}] [ingress] [egress]
2235 pattern {item} [/ {item} [...]] / end
2236 actions {action} [/ {action} [...]] / end
2238 If successful, it will return a flow rule ID usable with other commands::
2240 Flow rule #[...] created
2242 Otherwise it will show an error message of the form::
2244 Caught error type [...] ([...]): [...]
2246 Parameters describe in the following order:
2248 - Attributes (*group*, *priority*, *ingress*, *egress* tokens).
2249 - A matching pattern, starting with the *pattern* token and terminated by an
2251 - Actions, starting with the *actions* token and terminated by an *end*
2254 These translate directly to *rte_flow* objects provided as-is to the
2255 underlying functions.
2257 The shortest valid definition only comprises mandatory tokens::
2259 testpmd> flow create 0 pattern end actions end
2261 Note that PMDs may refuse rules that essentially do nothing such as this
2264 **All unspecified object values are automatically initialized to 0.**
2269 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
2270 specified before the ``pattern`` token.
2272 - ``group {group id}``: priority group.
2273 - ``priority {level}``: priority level within group.
2274 - ``ingress``: rule applies to ingress traffic.
2275 - ``egress``: rule applies to egress traffic.
2277 Each instance of an attribute specified several times overrides the previous
2278 value as shown below (group 4 is used)::
2280 testpmd> flow create 0 group 42 group 24 group 4 [...]
2282 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
2284 While not specifying a direction is an error, some rules may allow both
2287 Most rules affect RX therefore contain the ``ingress`` token::
2289 testpmd> flow create 0 ingress pattern [...]
2294 A matching pattern starts after the ``pattern`` token. It is made of pattern
2295 items and is terminated by a mandatory ``end`` item.
2297 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
2298 rte_flow_item_type``).
2300 The ``/`` token is used as a separator between pattern items as shown
2303 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
2305 Note that protocol items like these must be stacked from lowest to highest
2306 layer to make sense. For instance, the following rule is either invalid or
2307 unlikely to match any packet::
2309 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
2311 More information on these restrictions can be found in the *rte_flow*
2314 Several items support additional specification structures, for example
2315 ``ipv4`` allows specifying source and destination addresses as follows::
2317 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
2318 dst is 10.2.0.0 / end [...]
2320 This rule matches all IPv4 traffic with the specified properties.
2322 In this example, ``src`` and ``dst`` are field names of the underlying
2323 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
2324 in a similar fashion.
2326 The ``is`` token means that the subsequent value must be matched exactly,
2327 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
2328 accordingly. Possible assignment tokens are:
2330 - ``is``: match value perfectly (with full bit-mask).
2331 - ``spec``: match value according to configured bit-mask.
2332 - ``last``: specify upper bound to establish a range.
2333 - ``mask``: specify bit-mask with relevant bits set to one.
2334 - ``prefix``: generate bit-mask from a prefix length.
2336 These yield identical results::
2338 ipv4 src is 10.1.1.1
2342 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
2346 ipv4 src spec 10.1.1.1 src prefix 32
2350 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
2354 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
2356 Inclusive ranges can be defined with ``last``::
2358 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
2360 Note that ``mask`` affects both ``spec`` and ``last``::
2362 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
2363 # matches 10.1.0.0 to 10.2.255.255
2365 Properties can be modified multiple times::
2367 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
2371 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
2376 This section lists supported pattern items and their attributes, if any.
2378 - ``end``: end list of pattern items.
2380 - ``void``: no-op pattern item.
2382 - ``invert``: perform actions when pattern does not match.
2384 - ``any``: match any protocol for the current layer.
2386 - ``num {unsigned}``: number of layers covered.
2388 - ``pf``: match packets addressed to the physical function.
2390 - ``vf``: match packets addressed to a virtual function ID.
2392 - ``id {unsigned}``: destination VF ID.
2394 - ``port``: device-specific physical port index to use.
2396 - ``index {unsigned}``: physical port index.
2398 - ``raw``: match an arbitrary byte string.
2400 - ``relative {boolean}``: look for pattern after the previous item.
2401 - ``search {boolean}``: search pattern from offset (see also limit).
2402 - ``offset {integer}``: absolute or relative offset for pattern.
2403 - ``limit {unsigned}``: search area limit for start of pattern.
2404 - ``pattern {string}``: byte string to look for.
2406 - ``eth``: match Ethernet header.
2408 - ``dst {MAC-48}``: destination MAC.
2409 - ``src {MAC-48}``: source MAC.
2410 - ``type {unsigned}``: EtherType.
2412 - ``vlan``: match 802.1Q/ad VLAN tag.
2414 - ``tpid {unsigned}``: tag protocol identifier.
2415 - ``tci {unsigned}``: tag control information.
2416 - ``pcp {unsigned}``: priority code point.
2417 - ``dei {unsigned}``: drop eligible indicator.
2418 - ``vid {unsigned}``: VLAN identifier.
2420 - ``ipv4``: match IPv4 header.
2422 - ``tos {unsigned}``: type of service.
2423 - ``ttl {unsigned}``: time to live.
2424 - ``proto {unsigned}``: next protocol ID.
2425 - ``src {ipv4 address}``: source address.
2426 - ``dst {ipv4 address}``: destination address.
2428 - ``ipv6``: match IPv6 header.
2430 - ``tc {unsigned}``: traffic class.
2431 - ``flow {unsigned}``: flow label.
2432 - ``proto {unsigned}``: protocol (next header).
2433 - ``hop {unsigned}``: hop limit.
2434 - ``src {ipv6 address}``: source address.
2435 - ``dst {ipv6 address}``: destination address.
2437 - ``icmp``: match ICMP header.
2439 - ``type {unsigned}``: ICMP packet type.
2440 - ``code {unsigned}``: ICMP packet code.
2442 - ``udp``: match UDP header.
2444 - ``src {unsigned}``: UDP source port.
2445 - ``dst {unsigned}``: UDP destination port.
2447 - ``tcp``: match TCP header.
2449 - ``src {unsigned}``: TCP source port.
2450 - ``dst {unsigned}``: TCP destination port.
2452 - ``sctp``: match SCTP header.
2454 - ``src {unsigned}``: SCTP source port.
2455 - ``dst {unsigned}``: SCTP destination port.
2456 - ``tag {unsigned}``: validation tag.
2457 - ``cksum {unsigned}``: checksum.
2459 - ``vxlan``: match VXLAN header.
2461 - ``vni {unsigned}``: VXLAN identifier.
2466 A list of actions starts after the ``actions`` token in the same fashion as
2467 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
2468 terminated by a mandatory ``end`` action.
2470 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
2471 rte_flow_action_type``).
2473 Dropping all incoming UDPv4 packets can be expressed as follows::
2475 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2478 Several actions have configurable properties which must be specified when
2479 there is no valid default value. For example, ``queue`` requires a target
2482 This rule redirects incoming UDPv4 traffic to queue index 6::
2484 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2485 actions queue index 6 / end
2487 While this one could be rejected by PMDs (unspecified queue index)::
2489 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2492 As defined by *rte_flow*, the list is not ordered, all actions of a given
2493 rule are performed simultaneously. These are equivalent::
2495 queue index 6 / void / mark id 42 / end
2499 void / mark id 42 / queue index 6 / end
2501 All actions in a list should have different types, otherwise only the last
2502 action of a given type is taken into account::
2504 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
2508 drop / drop / drop / end # drop is performed only once
2512 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
2514 Considering they are performed simultaneously, opposite and overlapping
2515 actions can sometimes be combined when the end result is unambiguous::
2517 drop / queue index 6 / end # drop has no effect
2521 drop / dup index 6 / end # same as above
2525 queue index 6 / rss queues 6 7 8 / end # queue has no effect
2529 drop / passthru / end # drop has no effect
2531 Note that PMDs may still refuse such combinations.
2536 This section lists supported actions and their attributes, if any.
2538 - ``end``: end list of actions.
2540 - ``void``: no-op action.
2542 - ``passthru``: let subsequent rule process matched packets.
2544 - ``mark``: attach 32 bit value to packets.
2546 - ``id {unsigned}``: 32 bit value to return with packets.
2548 - ``flag``: flag packets.
2550 - ``queue``: assign packets to a given queue index.
2552 - ``index {unsigned}``: queue index to use.
2554 - ``drop``: drop packets (note: passthru has priority).
2556 - ``count``: enable counters for this rule.
2558 - ``dup``: duplicate packets to a given queue index.
2560 - ``index {unsigned}``: queue index to duplicate packets to.
2562 - ``rss``: spread packets among several queues.
2564 - ``queues [{unsigned} [...]] end``: queue indices to use.
2566 - ``pf``: redirect packets to physical device function.
2568 - ``vf``: redirect packets to virtual device function.
2570 - ``original {boolean}``: use original VF ID if possible.
2571 - ``id {unsigned}``: VF ID to redirect packets to.
2573 Destroying flow rules
2574 ~~~~~~~~~~~~~~~~~~~~~
2576 ``flow destroy`` destroys one or more rules from their rule ID (as returned
2577 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
2578 times as necessary::
2580 flow destroy {port_id} rule {rule_id} [...]
2582 If successful, it will show::
2584 Flow rule #[...] destroyed
2586 It does not report anything for rule IDs that do not exist. The usual error
2587 message is shown when a rule cannot be destroyed::
2589 Caught error type [...] ([...]): [...]
2591 ``flow flush`` destroys all rules on a device and does not take extra
2592 arguments. It is bound to ``rte_flow_flush()``::
2594 flow flush {port_id}
2596 Any errors are reported as above.
2598 Creating several rules and destroying them::
2600 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2601 actions queue index 2 / end
2602 Flow rule #0 created
2603 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2604 actions queue index 3 / end
2605 Flow rule #1 created
2606 testpmd> flow destroy 0 rule 0 rule 1
2607 Flow rule #1 destroyed
2608 Flow rule #0 destroyed
2611 The same result can be achieved using ``flow flush``::
2613 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2614 actions queue index 2 / end
2615 Flow rule #0 created
2616 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2617 actions queue index 3 / end
2618 Flow rule #1 created
2619 testpmd> flow flush 0
2622 Non-existent rule IDs are ignored::
2624 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2625 actions queue index 2 / end
2626 Flow rule #0 created
2627 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2628 actions queue index 3 / end
2629 Flow rule #1 created
2630 testpmd> flow destroy 0 rule 42 rule 10 rule 2
2632 testpmd> flow destroy 0 rule 0
2633 Flow rule #0 destroyed
2639 ``flow query`` queries a specific action of a flow rule having that
2640 ability. Such actions collect information that can be reported using this
2641 command. It is bound to ``rte_flow_query()``::
2643 flow query {port_id} {rule_id} {action}
2645 If successful, it will display either the retrieved data for known actions
2646 or the following message::
2648 Cannot display result for action type [...] ([...])
2650 Otherwise, it will complain either that the rule does not exist or that some
2653 Flow rule #[...] not found
2657 Caught error type [...] ([...]): [...]
2659 Currently only the ``count`` action is supported. This action reports the
2660 number of packets that hit the flow rule and the total number of bytes. Its
2661 output has the following format::
2664 hits_set: [...] # whether "hits" contains a valid value
2665 bytes_set: [...] # whether "bytes" contains a valid value
2666 hits: [...] # number of packets
2667 bytes: [...] # number of bytes
2669 Querying counters for TCPv6 packets redirected to queue 6::
2671 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
2672 actions queue index 6 / count / end
2673 Flow rule #4 created
2674 testpmd> flow query 0 4 count
2685 ``flow list`` lists existing flow rules sorted by priority and optionally
2686 filtered by group identifiers::
2688 flow list {port_id} [group {group_id}] [...]
2690 This command only fails with the following message if the device does not
2695 Output consists of a header line followed by a short description of each
2696 flow rule, one per line. There is no output at all when no flow rules are
2697 configured on the device::
2699 ID Group Prio Attr Rule
2700 [...] [...] [...] [...] [...]
2702 ``Attr`` column flags:
2704 - ``i`` for ``ingress``.
2705 - ``e`` for ``egress``.
2707 Creating several flow rules and listing them::
2709 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2710 actions queue index 6 / end
2711 Flow rule #0 created
2712 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2713 actions queue index 2 / end
2714 Flow rule #1 created
2715 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
2716 actions rss queues 6 7 8 end / end
2717 Flow rule #2 created
2718 testpmd> flow list 0
2719 ID Group Prio Attr Rule
2720 0 0 0 i- ETH IPV4 => QUEUE
2721 1 0 0 i- ETH IPV6 => QUEUE
2722 2 0 5 i- ETH IPV4 UDP => RSS
2725 Rules are sorted by priority (i.e. group ID first, then priority level)::
2727 testpmd> flow list 1
2728 ID Group Prio Attr Rule
2729 0 0 0 i- ETH => COUNT
2730 6 0 500 i- ETH IPV6 TCP => DROP COUNT
2731 5 0 1000 i- ETH IPV6 ICMP => QUEUE
2732 1 24 0 i- ETH IPV4 UDP => QUEUE
2733 4 24 10 i- ETH IPV4 TCP => DROP
2734 3 24 20 i- ETH IPV4 => DROP
2735 2 24 42 i- ETH IPV4 UDP => QUEUE
2736 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
2739 Output can be limited to specific groups::
2741 testpmd> flow list 1 group 0 group 63
2742 ID Group Prio Attr Rule
2743 0 0 0 i- ETH => COUNT
2744 6 0 500 i- ETH IPV6 TCP => DROP COUNT
2745 5 0 1000 i- ETH IPV6 ICMP => QUEUE
2746 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE