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)
899 Set the link flow control parameter on a port::
901 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
902 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
903 autoneg (on|off) (port_id)
907 * ``high_water`` (integer): High threshold value to trigger XOFF.
909 * ``low_water`` (integer): Low threshold value to trigger XON.
911 * ``pause_time`` (integer): Pause quota in the Pause frame.
913 * ``send_xon`` (0/1): Send XON frame.
915 * ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
917 * ``autoneg``: Change the auto-negotiation parameter.
922 Set the priority flow control parameter on a port::
924 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
925 (pause_time) (priority) (port_id)
929 * ``high_water`` (integer): High threshold value.
931 * ``low_water`` (integer): Low threshold value.
933 * ``pause_time`` (integer): Pause quota in the Pause frame.
935 * ``priority`` (0-7): VLAN User Priority.
940 Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
942 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
944 For example, to set rx queue 2 on port 0 to mapping 5::
946 testpmd>set stat_qmap rx 0 2 5
948 set port - rx/tx (for VF)
949 ~~~~~~~~~~~~~~~~~~~~~~~~~
951 Set VF receive/transmit from a port::
953 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
955 set port - mac address filter (for VF)
956 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
958 Add/Remove unicast or multicast MAC addr filter for a VF::
960 testpmd> set port (port_id) vf (vf_id) (mac_addr) \
961 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
963 set port - rx mode(for VF)
964 ~~~~~~~~~~~~~~~~~~~~~~~~~~
966 Set the VF receive mode of a port::
968 testpmd> set port (port_id) vf (vf_id) \
969 rxmode (AUPE|ROPE|BAM|MPE) (on|off)
971 The available receive modes are:
973 * ``AUPE``: Accepts untagged VLAN.
975 * ``ROPE``: Accepts unicast hash.
977 * ``BAM``: Accepts broadcast packets.
979 * ``MPE``: Accepts all multicast packets.
981 set port - tx_rate (for Queue)
982 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
984 Set TX rate limitation for a queue on a port::
986 testpmd> set port (port_id) queue (queue_id) rate (rate_value)
988 set port - tx_rate (for VF)
989 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
991 Set TX rate limitation for queues in VF on a port::
993 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
995 set port - mirror rule
996 ~~~~~~~~~~~~~~~~~~~~~~
998 Set pool or vlan type mirror rule for a port::
1000 testpmd> set port (port_id) mirror-rule (rule_id) \
1001 (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1002 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1004 Set link mirror rule for a port::
1006 testpmd> set port (port_id) mirror-rule (rule_id) \
1007 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1009 For example to enable mirror traffic with vlan 0,1 to pool 0::
1011 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1013 reset port - mirror rule
1014 ~~~~~~~~~~~~~~~~~~~~~~~~
1016 Reset a mirror rule for a port::
1018 testpmd> reset port (port_id) mirror-rule (rule_id)
1023 Set the flush on RX streams before forwarding.
1024 The default is flush ``on``.
1025 Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1027 testpmd> set flush_rx off
1032 Set the bypass mode for the lowest port on bypass enabled NIC::
1034 testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1039 Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1041 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1042 mode (normal|bypass|isolate) (port_id)
1046 * ``timeout``: Enable bypass after watchdog timeout.
1048 * ``os_on``: Enable bypass when OS/board is powered on.
1050 * ``os_off``: Enable bypass when OS/board is powered off.
1052 * ``power_on``: Enable bypass when power supply is turned on.
1054 * ``power_off``: Enable bypass when power supply is turned off.
1060 Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1062 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1067 Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1069 testpmd> show bypass config (port_id)
1074 Set link up for a port::
1076 testpmd> set link-up port (port id)
1081 Set link down for a port::
1083 testpmd> set link-down port (port id)
1088 Enable E-tag insertion for a VF on a port::
1090 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1092 Disable E-tag insertion for a VF on a port::
1094 testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1096 Enable/disable E-tag stripping on a port::
1098 testpmd> E-tag set stripping (on|off) port (port_id)
1100 Enable/disable E-tag based forwarding on a port::
1102 testpmd> E-tag set forwarding (on|off) port (port_id)
1104 Add an E-tag forwarding filter on a port::
1106 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1108 Delete an E-tag forwarding filter on a port::
1109 testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1115 The following sections show functions for configuring ports.
1119 Port configuration changes only become active when forwarding is started/restarted.
1124 Attach a port specified by pci address or virtual device args::
1126 testpmd> port attach (identifier)
1128 To attach a new pci device, the device should be recognized by kernel first.
1129 Then it should be moved under DPDK management.
1130 Finally the port can be attached to testpmd.
1132 For example, to move a pci device using ixgbe under DPDK management:
1134 .. code-block:: console
1136 # Check the status of the available devices.
1137 ./usertools/dpdk-devbind.py --status
1139 Network devices using DPDK-compatible driver
1140 ============================================
1143 Network devices using kernel driver
1144 ===================================
1145 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1148 # Bind the device to igb_uio.
1149 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1152 # Recheck the status of the devices.
1153 ./usertools/dpdk-devbind.py --status
1154 Network devices using DPDK-compatible driver
1155 ============================================
1156 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1158 To attach a port created by virtual device, above steps are not needed.
1160 For example, to attach a port whose pci address is 0000:0a:00.0.
1162 .. code-block:: console
1164 testpmd> port attach 0000:0a:00.0
1165 Attaching a new port...
1166 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1167 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1168 EAL: PCI memory mapped at 0x7f83bfa00000
1169 EAL: PCI memory mapped at 0x7f83bfa80000
1170 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1171 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1172 Port 0 is attached. Now total ports is 1
1175 For example, to attach a port created by pcap PMD.
1177 .. code-block:: console
1179 testpmd> port attach net_pcap0
1180 Attaching a new port...
1181 PMD: Initializing pmd_pcap for net_pcap0
1182 PMD: Creating pcap-backed ethdev on numa socket 0
1183 Port 0 is attached. Now total ports is 1
1186 In this case, identifier is ``net_pcap0``.
1187 This identifier format is the same as ``--vdev`` format of DPDK applications.
1189 For example, to re-attach a bonded port which has been previously detached,
1190 the mode and slave parameters must be given.
1192 .. code-block:: console
1194 testpmd> port attach net_bond_0,mode=0,slave=1
1195 Attaching a new port...
1196 EAL: Initializing pmd_bond for net_bond_0
1197 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1198 Port 0 is attached. Now total ports is 1
1205 Detach a specific port::
1207 testpmd> port detach (port_id)
1209 Before detaching a port, the port should be stopped and closed.
1211 For example, to detach a pci device port 0.
1213 .. code-block:: console
1215 testpmd> port stop 0
1218 testpmd> port close 0
1222 testpmd> port detach 0
1224 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1225 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1226 EAL: PCI memory unmapped at 0x7f83bfa00000
1227 EAL: PCI memory unmapped at 0x7f83bfa80000
1231 For example, to detach a virtual device port 0.
1233 .. code-block:: console
1235 testpmd> port stop 0
1238 testpmd> port close 0
1242 testpmd> port detach 0
1244 PMD: Closing pcap ethdev on numa socket 0
1245 Port 'net_pcap0' is detached. Now total ports is 0
1248 To remove a pci device completely from the system, first detach the port from testpmd.
1249 Then the device should be moved under kernel management.
1250 Finally the device can be removed using kernel pci hotplug functionality.
1252 For example, to move a pci device under kernel management:
1254 .. code-block:: console
1256 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1258 ./usertools/dpdk-devbind.py --status
1260 Network devices using DPDK-compatible driver
1261 ============================================
1264 Network devices using kernel driver
1265 ===================================
1266 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1268 To remove a port created by a virtual device, above steps are not needed.
1273 Start all ports or a specific port::
1275 testpmd> port start (port_id|all)
1280 Stop all ports or a specific port::
1282 testpmd> port stop (port_id|all)
1287 Close all ports or a specific port::
1289 testpmd> port close (port_id|all)
1291 port start/stop queue
1292 ~~~~~~~~~~~~~~~~~~~~~
1294 Start/stop a rx/tx queue on a specific port::
1296 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1298 Only take effect when port is started.
1303 Set the speed and duplex mode for all ports or a specific port::
1305 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1306 duplex (half|full|auto)
1308 port config - queues/descriptors
1309 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1311 Set number of queues/descriptors for rxq, txq, rxd and txd::
1313 testpmd> port config all (rxq|txq|rxd|txd) (value)
1315 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1317 port config - max-pkt-len
1318 ~~~~~~~~~~~~~~~~~~~~~~~~~
1320 Set the maximum packet length::
1322 testpmd> port config all max-pkt-len (value)
1324 This is equivalent to the ``--max-pkt-len`` command-line option.
1326 port config - CRC Strip
1327 ~~~~~~~~~~~~~~~~~~~~~~~
1329 Set hardware CRC stripping on or off for all ports::
1331 testpmd> port config all crc-strip (on|off)
1333 CRC stripping is off by default.
1335 The ``on`` option is equivalent to the ``--crc-strip`` command-line option.
1337 port config - scatter
1338 ~~~~~~~~~~~~~~~~~~~~~~~
1340 Set RX scatter mode on or off for all ports::
1342 testpmd> port config all scatter (on|off)
1344 RX scatter mode is off by default.
1346 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1348 port config - TX queue flags
1349 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1351 Set a hexadecimal bitmap of TX queue flags for all ports::
1353 testpmd> port config all txqflags value
1355 This command is equivalent to the ``--txqflags`` command-line option.
1357 port config - RX Checksum
1358 ~~~~~~~~~~~~~~~~~~~~~~~~~
1360 Set hardware RX checksum offload to on or off for all ports::
1362 testpmd> port config all rx-cksum (on|off)
1364 Checksum offload is off by default.
1366 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1371 Set hardware VLAN on or off for all ports::
1373 testpmd> port config all hw-vlan (on|off)
1375 Hardware VLAN is on by default.
1377 The ``off`` option is equivalent to the ``--disable-hw-vlan`` command-line option.
1379 port config - VLAN filter
1380 ~~~~~~~~~~~~~~~~~~~~~~~~~
1382 Set hardware VLAN filter on or off for all ports::
1384 testpmd> port config all hw-vlan-filter (on|off)
1386 Hardware VLAN filter is on by default.
1388 The ``off`` option is equivalent to the ``--disable-hw-vlan-filter`` command-line option.
1390 port config - VLAN strip
1391 ~~~~~~~~~~~~~~~~~~~~~~~~
1393 Set hardware VLAN strip on or off for all ports::
1395 testpmd> port config all hw-vlan-strip (on|off)
1397 Hardware VLAN strip is on by default.
1399 The ``off`` option is equivalent to the ``--disable-hw-vlan-strip`` command-line option.
1401 port config - VLAN extend
1402 ~~~~~~~~~~~~~~~~~~~~~~~~~
1404 Set hardware VLAN extend on or off for all ports::
1406 testpmd> port config all hw-vlan-extend (on|off)
1408 Hardware VLAN extend is off by default.
1410 The ``off`` option is equivalent to the ``--disable-hw-vlan-extend`` command-line option.
1412 port config - Drop Packets
1413 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1415 Set packet drop for packets with no descriptors on or off for all ports::
1417 testpmd> port config all drop-en (on|off)
1419 Packet dropping for packets with no descriptors is off by default.
1421 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1426 Set the RSS (Receive Side Scaling) mode on or off::
1428 testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1430 RSS is on by default.
1432 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1434 port config - RSS Reta
1435 ~~~~~~~~~~~~~~~~~~~~~~
1437 Set the RSS (Receive Side Scaling) redirection table::
1439 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1444 Set the DCB mode for an individual port::
1446 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1448 The traffic class should be 4 or 8.
1453 Set the number of packets per burst::
1455 testpmd> port config all burst (value)
1457 This is equivalent to the ``--burst`` command-line option.
1459 port config - Threshold
1460 ~~~~~~~~~~~~~~~~~~~~~~~
1462 Set thresholds for TX/RX queues::
1464 testpmd> port config all (threshold) (value)
1466 Where the threshold type can be:
1468 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1470 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1472 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1474 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1476 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1478 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1480 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1482 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1484 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1486 These threshold options are also available from the command-line.
1491 Set the value of ether-type for E-tag::
1493 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1495 Enable/disable the E-tag support::
1497 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1500 Link Bonding Functions
1501 ----------------------
1503 The Link Bonding functions make it possible to dynamically create and
1504 manage link bonding devices from within testpmd interactive prompt.
1506 create bonded device
1507 ~~~~~~~~~~~~~~~~~~~~
1509 Create a new bonding device::
1511 testpmd> create bonded device (mode) (socket)
1513 For example, to create a bonded device in mode 1 on socket 0::
1515 testpmd> create bonded 1 0
1516 created new bonded device (port X)
1521 Adds Ethernet device to a Link Bonding device::
1523 testpmd> add bonding slave (slave id) (port id)
1525 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
1527 testpmd> add bonding slave 6 10
1530 remove bonding slave
1531 ~~~~~~~~~~~~~~~~~~~~
1533 Removes an Ethernet slave device from a Link Bonding device::
1535 testpmd> remove bonding slave (slave id) (port id)
1537 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
1539 testpmd> remove bonding slave 6 10
1544 Set the Link Bonding mode of a Link Bonding device::
1546 testpmd> set bonding mode (value) (port id)
1548 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
1550 testpmd> set bonding mode 3 10
1555 Set an Ethernet slave device as the primary device on a Link Bonding device::
1557 testpmd> set bonding primary (slave id) (port id)
1559 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
1561 testpmd> set bonding primary 6 10
1566 Set the MAC address of a Link Bonding device::
1568 testpmd> set bonding mac (port id) (mac)
1570 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
1572 testpmd> set bonding mac 10 00:00:00:00:00:01
1574 set bonding xmit_balance_policy
1575 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1577 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
1579 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
1581 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
1583 testpmd> set bonding xmit_balance_policy 10 l34
1586 set bonding mon_period
1587 ~~~~~~~~~~~~~~~~~~~~~~
1589 Set the link status monitoring polling period in milliseconds for a bonding device.
1591 This adds support for PMD slave devices which do not support link status interrupts.
1592 When the mon_period is set to a value greater than 0 then all PMD's which do not support
1593 link status ISR will be queried every polling interval to check if their link status has changed::
1595 testpmd> set bonding mon_period (port_id) (value)
1597 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
1599 testpmd> set bonding mon_period 5 150
1605 Show the current configuration of a Link Bonding device::
1607 testpmd> show bonding config (port id)
1610 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
1611 in balance mode with a transmission policy of layer 2+3::
1613 testpmd> show bonding config 9
1615 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
1617 Active Slaves (3): [1 3 4]
1624 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
1625 This is mainly useful for debugging purposes.
1626 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
1627 and fields that can be accessed.
1632 Display the value of a port register::
1634 testpmd> read reg (port_id) (address)
1636 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
1638 testpmd> read reg 0 0xEE00
1639 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
1644 Display a port register bit field::
1646 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
1648 For example, reading the lowest two bits from the register in the example above::
1650 testpmd> read regfield 0 0xEE00 0 1
1651 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
1656 Display a single port register bit::
1658 testpmd> read regbit (port_id) (address) (bit_x)
1660 For example, reading the lowest bit from the register in the example above::
1662 testpmd> read regbit 0 0xEE00 0
1663 port 0 PCI register at offset 0xEE00: bit 0=1
1668 Set the value of a port register::
1670 testpmd> write reg (port_id) (address) (value)
1672 For example, to clear a register::
1674 testpmd> write reg 0 0xEE00 0x0
1675 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
1680 Set bit field of a port register::
1682 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
1684 For example, writing to the register cleared in the example above::
1686 testpmd> write regfield 0 0xEE00 0 1 2
1687 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
1692 Set single bit value of a port register::
1694 testpmd> write regbit (port_id) (address) (bit_x) (value)
1696 For example, to set the high bit in the register from the example above::
1698 testpmd> write regbit 0 0xEE00 31 1
1699 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
1705 This section details the available filter functions that are available.
1707 Note these functions interface the deprecated legacy filtering framework,
1708 superseded by *rte_flow*. See `Flow rules management`_.
1711 ~~~~~~~~~~~~~~~~~~~~
1713 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
1715 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
1716 ethertype (ether_type) (drop|fwd) queue (queue_id)
1718 The available information parameters are:
1720 * ``port_id``: The port which the Ethertype filter assigned on.
1722 * ``mac_addr``: Compare destination mac address.
1724 * ``mac_ignr``: Ignore destination mac address match.
1726 * ``mac_address``: Destination mac address to match.
1728 * ``ether_type``: The EtherType value want to match,
1729 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
1731 * ``queue_id``: The receive queue associated with this EtherType filter.
1732 It is meaningless when deleting or dropping.
1734 Example, to add/remove an ethertype filter rule::
1736 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
1737 ethertype 0x0806 fwd queue 3
1739 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
1740 ethertype 0x0806 fwd queue 3
1745 Add or delete a 2-tuple filter,
1746 which identifies packets by specific protocol and destination TCP/UDP port
1747 and forwards packets into one of the receive queues::
1749 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
1750 protocol (protocol_value) mask (mask_value) \
1751 tcp_flags (tcp_flags_value) priority (prio_value) \
1754 The available information parameters are:
1756 * ``port_id``: The port which the 2-tuple filter assigned on.
1758 * ``dst_port_value``: Destination port in L4.
1760 * ``protocol_value``: IP L4 protocol.
1762 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
1764 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
1766 * ``prio_value``: Priority of this filter.
1768 * ``queue_id``: The receive queue associated with this 2-tuple filter.
1770 Example, to add/remove an 2tuple filter rule::
1772 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
1773 tcp_flags 0x02 priority 3 queue 3
1775 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
1776 tcp_flags 0x02 priority 3 queue 3
1781 Add or delete a 5-tuple filter,
1782 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
1783 and routes packets into one of the receive queues::
1785 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
1786 (src_address) dst_port (dst_port_value) \
1787 src_port (src_port_value) protocol (protocol_value) \
1788 mask (mask_value) tcp_flags (tcp_flags_value) \
1789 priority (prio_value) queue (queue_id)
1791 The available information parameters are:
1793 * ``port_id``: The port which the 5-tuple filter assigned on.
1795 * ``dst_address``: Destination IP address.
1797 * ``src_address``: Source IP address.
1799 * ``dst_port_value``: TCP/UDP destination port.
1801 * ``src_port_value``: TCP/UDP source port.
1803 * ``protocol_value``: L4 protocol.
1805 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
1807 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
1809 * ``prio_value``: The priority of this filter.
1811 * ``queue_id``: The receive queue associated with this 5-tuple filter.
1813 Example, to add/remove an 5tuple filter rule::
1815 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1816 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1817 flags 0x0 priority 3 queue 3
1819 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1820 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1821 flags 0x0 priority 3 queue 3
1826 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
1828 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
1830 The available information parameters are:
1832 * ``port_id``: The port which the SYN filter assigned on.
1834 * ``high``: This SYN filter has higher priority than other filters.
1836 * ``low``: This SYN filter has lower priority than other filters.
1838 * ``queue_id``: The receive queue associated with this SYN filter
1842 testpmd> syn_filter 0 add priority high queue 3
1847 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
1848 and routed into one of the receive queues::
1850 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
1851 mask (mask_value) priority (prio_value) queue (queue_id)
1853 The available information parameters are:
1855 * ``port_id``: The port which the Flex filter is assigned on.
1857 * ``len_value``: Filter length in bytes, no greater than 128.
1859 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
1861 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
1863 * ``prio_value``: The priority of this filter.
1865 * ``queue_id``: The receive queue associated with this Flex filter.
1869 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
1870 mask 000C priority 3 queue 3
1872 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
1873 mask 000C priority 3 queue 3
1876 .. _testpmd_flow_director:
1878 flow_director_filter
1879 ~~~~~~~~~~~~~~~~~~~~
1881 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
1883 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
1884 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
1886 * Perfect match filters.
1887 The hardware checks a match between the masked fields of the received packets and the programmed filters.
1888 The masked fields are for IP flow.
1890 * Signature filters.
1891 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
1893 * Perfect-mac-vlan 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 MAC VLAN flow.
1897 * Perfect-tunnel match filters.
1898 The hardware checks a match between the masked fields of the received packets and the programmed filters.
1899 The masked fields are for tunnel flow.
1901 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
1902 per flow type and the flexible payload.
1904 The Flow Director can also mask out parts of all of these fields so that filters
1905 are only applied to certain fields or parts of the fields.
1907 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
1909 # Commands to add flow director filters of different flow types::
1911 flow_director_filter (port_id) mode IP (add|del|update) \
1912 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
1913 src (src_ip_address) dst (dst_ip_address) \
1914 tos (tos_value) proto (proto_value) ttl (ttl_value) \
1915 vlan (vlan_value) flexbytes (flexbytes_value) \
1916 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
1919 flow_director_filter (port_id) mode IP (add|del|update) \
1920 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
1921 src (src_ip_address) (src_port) \
1922 dst (dst_ip_address) (dst_port) \
1923 tos (tos_value) ttl (ttl_value) \
1924 vlan (vlan_value) flexbytes (flexbytes_value) \
1925 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
1928 flow_director_filter (port_id) mode IP (add|del|update) \
1929 flow (ipv4-sctp|ipv6-sctp) \
1930 src (src_ip_address) (src_port) \
1931 dst (dst_ip_address) (dst_port) \
1932 tos (tos_value) ttl (ttl_value) \
1933 tag (verification_tag) vlan (vlan_value) \
1934 flexbytes (flexbytes_value) (drop|fwd) \
1935 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
1937 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
1938 ether (ethertype) flexbytes (flexbytes_value) \
1939 (drop|fwd) pf|vf(vf_id) queue (queue_id)
1942 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
1943 mac (mac_address) vlan (vlan_value) \
1944 flexbytes (flexbytes_value) (drop|fwd) \
1945 queue (queue_id) fd_id (fd_id_value)
1947 flow_director_filter (port_id) mode Tunnel (add|del|update) \
1948 mac (mac_address) vlan (vlan_value) \
1949 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
1950 flexbytes (flexbytes_value) (drop|fwd) \
1951 queue (queue_id) fd_id (fd_id_value)
1953 For example, to add an ipv4-udp flow type filter::
1955 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
1956 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
1957 fwd pf queue 1 fd_id 1
1959 For example, add an ipv4-other flow type filter::
1961 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
1962 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
1963 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
1968 Flush all flow director filters on a device::
1970 testpmd> flush_flow_director (port_id)
1972 Example, to flush all flow director filter on port 0::
1974 testpmd> flush_flow_director 0
1979 Set flow director's input masks::
1981 flow_director_mask (port_id) mode IP vlan (vlan_value) \
1982 src_mask (ipv4_src) (ipv6_src) (src_port) \
1983 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
1985 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
1987 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
1988 mac (mac_value) tunnel-type (tunnel_type_value) \
1989 tunnel-id (tunnel_id_value)
1991 Example, to set flow director mask on port 0::
1993 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
1994 src_mask 255.255.255.255 \
1995 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
1996 dst_mask 255.255.255.255 \
1997 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
1999 flow_director_flex_mask
2000 ~~~~~~~~~~~~~~~~~~~~~~~
2002 set masks of flow director's flexible payload based on certain flow type::
2004 testpmd> flow_director_flex_mask (port_id) \
2005 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2006 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2007 l2_payload|all) (mask)
2009 Example, to set flow director's flex mask for all flow type on port 0::
2011 testpmd> flow_director_flex_mask 0 flow all \
2012 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2015 flow_director_flex_payload
2016 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2018 Configure flexible payload selection::
2020 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2022 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2024 testpmd> flow_director_flex_payload 0 l4 \
2025 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2027 get_sym_hash_ena_per_port
2028 ~~~~~~~~~~~~~~~~~~~~~~~~~
2030 Get symmetric hash enable configuration per port::
2032 get_sym_hash_ena_per_port (port_id)
2034 For example, to get symmetric hash enable configuration of port 1::
2036 testpmd> get_sym_hash_ena_per_port 1
2038 set_sym_hash_ena_per_port
2039 ~~~~~~~~~~~~~~~~~~~~~~~~~
2041 Set symmetric hash enable configuration per port to enable or disable::
2043 set_sym_hash_ena_per_port (port_id) (enable|disable)
2045 For example, to set symmetric hash enable configuration of port 1 to enable::
2047 testpmd> set_sym_hash_ena_per_port 1 enable
2049 get_hash_global_config
2050 ~~~~~~~~~~~~~~~~~~~~~~
2052 Get the global configurations of hash filters::
2054 get_hash_global_config (port_id)
2056 For example, to get the global configurations of hash filters of port 1::
2058 testpmd> get_hash_global_config 1
2060 set_hash_global_config
2061 ~~~~~~~~~~~~~~~~~~~~~~
2063 Set the global configurations of hash filters::
2065 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2066 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2067 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload) \
2070 For example, to enable simple_xor for flow type of ipv6 on port 2::
2072 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2077 Set the input set for hash::
2079 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2080 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2081 l2_payload) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2082 ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2083 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2084 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2085 fld-8th|none) (select|add)
2087 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2089 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2094 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2095 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2097 Set the input set for flow director::
2099 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2100 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2101 l2_payload) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2102 ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2103 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2104 sctp-dst-port|sctp-veri-tag|none) (select|add)
2106 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2108 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2113 Set different GRE key length for input set::
2115 global_config (port_id) gre-key-len (number in bytes)
2117 For example to set GRE key length for input set to 4 bytes on port 0::
2119 testpmd> global_config 0 gre-key-len 4
2122 .. _testpmd_rte_flow:
2124 Flow rules management
2125 ---------------------
2127 Control of the generic flow API (*rte_flow*) is fully exposed through the
2128 ``flow`` command (validation, creation, destruction and queries).
2130 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2131 features simultaneously may cause undefined side-effects and is therefore
2137 Because the ``flow`` command uses dynamic tokens to handle the large number
2138 of possible flow rules combinations, its behavior differs slightly from
2139 other commands, in particular:
2141 - Pressing *?* or the *<tab>* key displays contextual help for the current
2142 token, not that of the entire command.
2144 - Optional and repeated parameters are supported (provided they are listed
2145 in the contextual help).
2147 The first parameter stands for the operation mode. Possible operations and
2148 their general syntax are described below. They are covered in detail in the
2151 - Check whether a flow rule can be created::
2153 flow validate {port_id}
2154 [group {group_id}] [priority {level}] [ingress] [egress]
2155 pattern {item} [/ {item} [...]] / end
2156 actions {action} [/ {action} [...]] / end
2158 - Create a flow rule::
2160 flow create {port_id}
2161 [group {group_id}] [priority {level}] [ingress] [egress]
2162 pattern {item} [/ {item} [...]] / end
2163 actions {action} [/ {action} [...]] / end
2165 - Destroy specific flow rules::
2167 flow destroy {port_id} rule {rule_id} [...]
2169 - Destroy all flow rules::
2171 flow flush {port_id}
2173 - Query an existing flow rule::
2175 flow query {port_id} {rule_id} {action}
2177 - List existing flow rules sorted by priority, filtered by group
2180 flow list {port_id} [group {group_id}] [...]
2182 Validating flow rules
2183 ~~~~~~~~~~~~~~~~~~~~~
2185 ``flow validate`` reports whether a flow rule would be accepted by the
2186 underlying device in its current state but stops short of creating it. It is
2187 bound to ``rte_flow_validate()``::
2189 flow validate {port_id}
2190 [group {group_id}] [priority {level}] [ingress] [egress]
2191 pattern {item} [/ {item} [...]] / end
2192 actions {action} [/ {action} [...]] / end
2194 If successful, it will show::
2198 Otherwise it will show an error message of the form::
2200 Caught error type [...] ([...]): [...]
2202 This command uses the same parameters as ``flow create``, their format is
2203 described in `Creating flow rules`_.
2205 Check whether redirecting any Ethernet packet received on port 0 to RX queue
2206 index 6 is supported::
2208 testpmd> flow validate 0 ingress pattern eth / end
2209 actions queue index 6 / end
2213 Port 0 does not support TCPv6 rules::
2215 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
2217 Caught error type 9 (specific pattern item): Invalid argument
2223 ``flow create`` validates and creates the specified flow rule. It is bound
2224 to ``rte_flow_create()``::
2226 flow create {port_id}
2227 [group {group_id}] [priority {level}] [ingress] [egress]
2228 pattern {item} [/ {item} [...]] / end
2229 actions {action} [/ {action} [...]] / end
2231 If successful, it will return a flow rule ID usable with other commands::
2233 Flow rule #[...] created
2235 Otherwise it will show an error message of the form::
2237 Caught error type [...] ([...]): [...]
2239 Parameters describe in the following order:
2241 - Attributes (*group*, *priority*, *ingress*, *egress* tokens).
2242 - A matching pattern, starting with the *pattern* token and terminated by an
2244 - Actions, starting with the *actions* token and terminated by an *end*
2247 These translate directly to *rte_flow* objects provided as-is to the
2248 underlying functions.
2250 The shortest valid definition only comprises mandatory tokens::
2252 testpmd> flow create 0 pattern end actions end
2254 Note that PMDs may refuse rules that essentially do nothing such as this
2257 **All unspecified object values are automatically initialized to 0.**
2262 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
2263 specified before the ``pattern`` token.
2265 - ``group {group id}``: priority group.
2266 - ``priority {level}``: priority level within group.
2267 - ``ingress``: rule applies to ingress traffic.
2268 - ``egress``: rule applies to egress traffic.
2270 Each instance of an attribute specified several times overrides the previous
2271 value as shown below (group 4 is used)::
2273 testpmd> flow create 0 group 42 group 24 group 4 [...]
2275 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
2277 While not specifying a direction is an error, some rules may allow both
2280 Most rules affect RX therefore contain the ``ingress`` token::
2282 testpmd> flow create 0 ingress pattern [...]
2287 A matching pattern starts after the ``pattern`` token. It is made of pattern
2288 items and is terminated by a mandatory ``end`` item.
2290 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
2291 rte_flow_item_type``).
2293 The ``/`` token is used as a separator between pattern items as shown
2296 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
2298 Note that protocol items like these must be stacked from lowest to highest
2299 layer to make sense. For instance, the following rule is either invalid or
2300 unlikely to match any packet::
2302 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
2304 More information on these restrictions can be found in the *rte_flow*
2307 Several items support additional specification structures, for example
2308 ``ipv4`` allows specifying source and destination addresses as follows::
2310 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
2311 dst is 10.2.0.0 / end [...]
2313 This rule matches all IPv4 traffic with the specified properties.
2315 In this example, ``src`` and ``dst`` are field names of the underlying
2316 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
2317 in a similar fashion.
2319 The ``is`` token means that the subsequent value must be matched exactly,
2320 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
2321 accordingly. Possible assignment tokens are:
2323 - ``is``: match value perfectly (with full bit-mask).
2324 - ``spec``: match value according to configured bit-mask.
2325 - ``last``: specify upper bound to establish a range.
2326 - ``mask``: specify bit-mask with relevant bits set to one.
2327 - ``prefix``: generate bit-mask from a prefix length.
2329 These yield identical results::
2331 ipv4 src is 10.1.1.1
2335 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
2339 ipv4 src spec 10.1.1.1 src prefix 32
2343 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
2347 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
2349 Inclusive ranges can be defined with ``last``::
2351 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
2353 Note that ``mask`` affects both ``spec`` and ``last``::
2355 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
2356 # matches 10.1.0.0 to 10.2.255.255
2358 Properties can be modified multiple times::
2360 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
2364 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
2369 This section lists supported pattern items and their attributes, if any.
2371 - ``end``: end list of pattern items.
2373 - ``void``: no-op pattern item.
2375 - ``invert``: perform actions when pattern does not match.
2377 - ``any``: match any protocol for the current layer.
2379 - ``num {unsigned}``: number of layers covered.
2381 - ``pf``: match packets addressed to the physical function.
2383 - ``vf``: match packets addressed to a virtual function ID.
2385 - ``id {unsigned}``: destination VF ID.
2387 - ``port``: device-specific physical port index to use.
2389 - ``index {unsigned}``: physical port index.
2391 - ``raw``: match an arbitrary byte string.
2393 - ``relative {boolean}``: look for pattern after the previous item.
2394 - ``search {boolean}``: search pattern from offset (see also limit).
2395 - ``offset {integer}``: absolute or relative offset for pattern.
2396 - ``limit {unsigned}``: search area limit for start of pattern.
2397 - ``pattern {string}``: byte string to look for.
2399 - ``eth``: match Ethernet header.
2401 - ``dst {MAC-48}``: destination MAC.
2402 - ``src {MAC-48}``: source MAC.
2403 - ``type {unsigned}``: EtherType.
2405 - ``vlan``: match 802.1Q/ad VLAN tag.
2407 - ``tpid {unsigned}``: tag protocol identifier.
2408 - ``tci {unsigned}``: tag control information.
2409 - ``pcp {unsigned}``: priority code point.
2410 - ``dei {unsigned}``: drop eligible indicator.
2411 - ``vid {unsigned}``: VLAN identifier.
2413 - ``ipv4``: match IPv4 header.
2415 - ``tos {unsigned}``: type of service.
2416 - ``ttl {unsigned}``: time to live.
2417 - ``proto {unsigned}``: next protocol ID.
2418 - ``src {ipv4 address}``: source address.
2419 - ``dst {ipv4 address}``: destination address.
2421 - ``ipv6``: match IPv6 header.
2423 - ``tc {unsigned}``: traffic class.
2424 - ``flow {unsigned}``: flow label.
2425 - ``proto {unsigned}``: protocol (next header).
2426 - ``hop {unsigned}``: hop limit.
2427 - ``src {ipv6 address}``: source address.
2428 - ``dst {ipv6 address}``: destination address.
2430 - ``icmp``: match ICMP header.
2432 - ``type {unsigned}``: ICMP packet type.
2433 - ``code {unsigned}``: ICMP packet code.
2435 - ``udp``: match UDP header.
2437 - ``src {unsigned}``: UDP source port.
2438 - ``dst {unsigned}``: UDP destination port.
2440 - ``tcp``: match TCP header.
2442 - ``src {unsigned}``: TCP source port.
2443 - ``dst {unsigned}``: TCP destination port.
2445 - ``sctp``: match SCTP header.
2447 - ``src {unsigned}``: SCTP source port.
2448 - ``dst {unsigned}``: SCTP destination port.
2449 - ``tag {unsigned}``: validation tag.
2450 - ``cksum {unsigned}``: checksum.
2452 - ``vxlan``: match VXLAN header.
2454 - ``vni {unsigned}``: VXLAN identifier.
2459 A list of actions starts after the ``actions`` token in the same fashion as
2460 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
2461 terminated by a mandatory ``end`` action.
2463 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
2464 rte_flow_action_type``).
2466 Dropping all incoming UDPv4 packets can be expressed as follows::
2468 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2471 Several actions have configurable properties which must be specified when
2472 there is no valid default value. For example, ``queue`` requires a target
2475 This rule redirects incoming UDPv4 traffic to queue index 6::
2477 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2478 actions queue index 6 / end
2480 While this one could be rejected by PMDs (unspecified queue index)::
2482 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2485 As defined by *rte_flow*, the list is not ordered, all actions of a given
2486 rule are performed simultaneously. These are equivalent::
2488 queue index 6 / void / mark id 42 / end
2492 void / mark id 42 / queue index 6 / end
2494 All actions in a list should have different types, otherwise only the last
2495 action of a given type is taken into account::
2497 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
2501 drop / drop / drop / end # drop is performed only once
2505 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
2507 Considering they are performed simultaneously, opposite and overlapping
2508 actions can sometimes be combined when the end result is unambiguous::
2510 drop / queue index 6 / end # drop has no effect
2514 drop / dup index 6 / end # same as above
2518 queue index 6 / rss queues 6 7 8 / end # queue has no effect
2522 drop / passthru / end # drop has no effect
2524 Note that PMDs may still refuse such combinations.
2529 This section lists supported actions and their attributes, if any.
2531 - ``end``: end list of actions.
2533 - ``void``: no-op action.
2535 - ``passthru``: let subsequent rule process matched packets.
2537 - ``mark``: attach 32 bit value to packets.
2539 - ``id {unsigned}``: 32 bit value to return with packets.
2541 - ``flag``: flag packets.
2543 - ``queue``: assign packets to a given queue index.
2545 - ``index {unsigned}``: queue index to use.
2547 - ``drop``: drop packets (note: passthru has priority).
2549 - ``count``: enable counters for this rule.
2551 - ``dup``: duplicate packets to a given queue index.
2553 - ``index {unsigned}``: queue index to duplicate packets to.
2555 - ``rss``: spread packets among several queues.
2557 - ``queues [{unsigned} [...]] end``: queue indices to use.
2559 - ``pf``: redirect packets to physical device function.
2561 - ``vf``: redirect packets to virtual device function.
2563 - ``original {boolean}``: use original VF ID if possible.
2564 - ``id {unsigned}``: VF ID to redirect packets to.
2566 Destroying flow rules
2567 ~~~~~~~~~~~~~~~~~~~~~
2569 ``flow destroy`` destroys one or more rules from their rule ID (as returned
2570 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
2571 times as necessary::
2573 flow destroy {port_id} rule {rule_id} [...]
2575 If successful, it will show::
2577 Flow rule #[...] destroyed
2579 It does not report anything for rule IDs that do not exist. The usual error
2580 message is shown when a rule cannot be destroyed::
2582 Caught error type [...] ([...]): [...]
2584 ``flow flush`` destroys all rules on a device and does not take extra
2585 arguments. It is bound to ``rte_flow_flush()``::
2587 flow flush {port_id}
2589 Any errors are reported as above.
2591 Creating several rules and destroying them::
2593 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2594 actions queue index 2 / end
2595 Flow rule #0 created
2596 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2597 actions queue index 3 / end
2598 Flow rule #1 created
2599 testpmd> flow destroy 0 rule 0 rule 1
2600 Flow rule #1 destroyed
2601 Flow rule #0 destroyed
2604 The same result can be achieved using ``flow flush``::
2606 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2607 actions queue index 2 / end
2608 Flow rule #0 created
2609 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2610 actions queue index 3 / end
2611 Flow rule #1 created
2612 testpmd> flow flush 0
2615 Non-existent rule IDs are ignored::
2617 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2618 actions queue index 2 / end
2619 Flow rule #0 created
2620 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2621 actions queue index 3 / end
2622 Flow rule #1 created
2623 testpmd> flow destroy 0 rule 42 rule 10 rule 2
2625 testpmd> flow destroy 0 rule 0
2626 Flow rule #0 destroyed
2632 ``flow query`` queries a specific action of a flow rule having that
2633 ability. Such actions collect information that can be reported using this
2634 command. It is bound to ``rte_flow_query()``::
2636 flow query {port_id} {rule_id} {action}
2638 If successful, it will display either the retrieved data for known actions
2639 or the following message::
2641 Cannot display result for action type [...] ([...])
2643 Otherwise, it will complain either that the rule does not exist or that some
2646 Flow rule #[...] not found
2650 Caught error type [...] ([...]): [...]
2652 Currently only the ``count`` action is supported. This action reports the
2653 number of packets that hit the flow rule and the total number of bytes. Its
2654 output has the following format::
2657 hits_set: [...] # whether "hits" contains a valid value
2658 bytes_set: [...] # whether "bytes" contains a valid value
2659 hits: [...] # number of packets
2660 bytes: [...] # number of bytes
2662 Querying counters for TCPv6 packets redirected to queue 6::
2664 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
2665 actions queue index 6 / count / end
2666 Flow rule #4 created
2667 testpmd> flow query 0 4 count
2678 ``flow list`` lists existing flow rules sorted by priority and optionally
2679 filtered by group identifiers::
2681 flow list {port_id} [group {group_id}] [...]
2683 This command only fails with the following message if the device does not
2688 Output consists of a header line followed by a short description of each
2689 flow rule, one per line. There is no output at all when no flow rules are
2690 configured on the device::
2692 ID Group Prio Attr Rule
2693 [...] [...] [...] [...] [...]
2695 ``Attr`` column flags:
2697 - ``i`` for ``ingress``.
2698 - ``e`` for ``egress``.
2700 Creating several flow rules and listing them::
2702 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2703 actions queue index 6 / end
2704 Flow rule #0 created
2705 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2706 actions queue index 2 / end
2707 Flow rule #1 created
2708 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
2709 actions rss queues 6 7 8 end / end
2710 Flow rule #2 created
2711 testpmd> flow list 0
2712 ID Group Prio Attr Rule
2713 0 0 0 i- ETH IPV4 => QUEUE
2714 1 0 0 i- ETH IPV6 => QUEUE
2715 2 0 5 i- ETH IPV4 UDP => RSS
2718 Rules are sorted by priority (i.e. group ID first, then priority level)::
2720 testpmd> flow list 1
2721 ID Group Prio Attr Rule
2722 0 0 0 i- ETH => COUNT
2723 6 0 500 i- ETH IPV6 TCP => DROP COUNT
2724 5 0 1000 i- ETH IPV6 ICMP => QUEUE
2725 1 24 0 i- ETH IPV4 UDP => QUEUE
2726 4 24 10 i- ETH IPV4 TCP => DROP
2727 3 24 20 i- ETH IPV4 => DROP
2728 2 24 42 i- ETH IPV4 UDP => QUEUE
2729 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
2732 Output can be limited to specific groups::
2734 testpmd> flow list 1 group 0 group 63
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 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE