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
331 Display VF statistics::
333 testpmd> show vf stats (port_id) (vf_id)
338 Reset VF statistics::
340 testpmd> clear vf stats (port_id) (vf_id)
342 Configuration Functions
343 -----------------------
345 The testpmd application can be configured from the runtime as well as from the command-line.
347 This section details the available configuration functions that are available.
351 Configuration changes only become active when forwarding is started/restarted.
356 Reset forwarding to the default configuration::
363 Set the debug verbosity level::
365 testpmd> set verbose (level)
367 Currently the only available levels are 0 (silent except for error) and 1 (fully verbose).
372 Set the number of ports used by the application:
376 This is equivalent to the ``--nb-ports`` command-line option.
381 Set the number of cores used by the application::
383 testpmd> set nbcore (num)
385 This is equivalent to the ``--nb-cores`` command-line option.
389 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port.
394 Set the forwarding cores hexadecimal mask::
396 testpmd> set coremask (mask)
398 This is equivalent to the ``--coremask`` command-line option.
402 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding.
407 Set the forwarding ports hexadecimal mask::
409 testpmd> set portmask (mask)
411 This is equivalent to the ``--portmask`` command-line option.
416 Set number of packets per burst::
418 testpmd> set burst (num)
420 This is equivalent to the ``--burst command-line`` option.
422 When retry is enabled, the transmit delay time and number of retries can also be set::
424 testpmd> set burst tx delay (microseconds) retry (num)
429 Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode::
431 testpmd> set txpkts (x[,y]*)
433 Where x[,y]* represents a CSV list of values, without white space.
438 Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes::
440 testpmd> set txsplit (off|on|rand)
444 * ``off`` disable packet copy & split for CSUM mode.
446 * ``on`` split outgoing packet into multiple segments. Size of each segment
447 and number of segments per packet is determined by ``set txpkts`` command
450 * ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments.
455 Set the list of forwarding cores::
457 testpmd> set corelist (x[,y]*)
459 For example, to change the forwarding cores:
461 .. code-block:: console
463 testpmd> set corelist 3,1
464 testpmd> show config fwd
466 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled
467 Logical Core 3 (socket 0) forwards packets on 1 streams:
468 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01
469 Logical Core 1 (socket 0) forwards packets on 1 streams:
470 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
474 The cores are used in the same order as specified on the command line.
479 Set the list of forwarding ports::
481 testpmd> set portlist (x[,y]*)
483 For example, to change the port forwarding:
485 .. code-block:: console
487 testpmd> set portlist 0,2,1,3
488 testpmd> show config fwd
490 io packet forwarding - ports=4 - cores=1 - streams=4
491 Logical Core 3 (socket 0) forwards packets on 4 streams:
492 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01
493 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
494 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03
495 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02
500 Enable/disable tx loopback::
502 testpmd> set tx loopback (port_id) (on|off)
507 set drop enable bit for all queues::
509 testpmd> set all queues drop (port_id) (on|off)
511 set split drop enable (for VF)
512 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
514 set split drop enable bit for VF from PF::
516 testpmd> set vf split drop (port_id) (vf_id) (on|off)
518 set mac antispoof (for VF)
519 ~~~~~~~~~~~~~~~~~~~~~~~~~~
521 Set mac antispoof for a VF from the PF::
523 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off)
528 Enable/disable MACsec offload::
530 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off)
531 testpmd> set macsec offload (port_id) off
536 Configure MACsec secure connection (SC)::
538 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi)
542 The pi argument is ignored for tx.
543 Check the NIC Datasheet for hardware limits.
548 Configure MACsec secure association (SA)::
550 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key)
554 The IDX value must be 0 or 1.
555 Check the NIC Datasheet for hardware limits.
557 set broadcast mode (for VF)
558 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
560 Set broadcast mode for a VF from the PF::
562 testpmd> set vf broadcast (port_id) (vf_id) (on|off)
567 Set the VLAN strip on a port::
569 testpmd> vlan set strip (on|off) (port_id)
574 Set the VLAN strip for a queue on a port::
576 testpmd> vlan set stripq (on|off) (port_id,queue_id)
578 vlan set stripq (for VF)
579 ~~~~~~~~~~~~~~~~~~~~~~~~
581 Set VLAN strip for all queues in a pool for a VF from the PF::
583 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off)
585 vlan set insert (for VF)
586 ~~~~~~~~~~~~~~~~~~~~~~~~
588 Set VLAN insert for a VF from the PF::
590 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id)
592 vlan set tag (for VF)
593 ~~~~~~~~~~~~~~~~~~~~~
595 Set VLAN tag for a VF from the PF::
597 testpmd> set vf vlan tag (port_id) (vf_id) (on|off)
599 vlan set antispoof (for VF)
600 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
602 Set VLAN antispoof for a VF from the PF::
604 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off)
609 Set the VLAN filter on a port::
611 testpmd> vlan set filter (on|off) (port_id)
616 Set the VLAN QinQ (extended queue in queue) on for a port::
618 testpmd> vlan set qinq (on|off) (port_id)
623 Set the inner or outer VLAN TPID for packet filtering on a port::
625 testpmd> vlan set (inner|outer) tpid (value) (port_id)
629 TPID value must be a 16-bit number (value <= 65536).
634 Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID::
636 testpmd> rx_vlan add (vlan_id|all) (port_id)
640 VLAN filter must be set on that port. VLAN ID < 4096.
641 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries
642 in VFTA table. This is important if enabling all vlan_ids.
647 Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID::
649 testpmd> rx_vlan rm (vlan_id|all) (port_id)
654 Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID::
656 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask)
661 Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID::
663 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask)
668 Add a tunnel filter on a port::
670 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \
671 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
672 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
674 The available information categories are:
676 * ``vxlan``: Set tunnel type as VXLAN.
678 * ``nvgre``: Set tunnel type as NVGRE.
680 * ``ipingre``: Set tunnel type as IP-in-GRE.
682 * ``imac-ivlan``: Set filter type as Inner MAC and VLAN.
684 * ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID.
686 * ``imac-tenid``: Set filter type as Inner MAC and tenant ID.
688 * ``imac``: Set filter type as Inner MAC.
690 * ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID.
692 * ``oip``: Set filter type as Outer IP.
694 * ``iip``: Set filter type as Inner IP.
698 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \
699 192.168.2.2 0 ipingre oip 1 1
701 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP.
706 Remove a tunnel filter on a port::
708 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \
709 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
710 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
715 Add an UDP port for VXLAN packet filter on a port::
717 testpmd> rx_vxlan_port add (udp_port) (port_id)
722 Remove an UDP port for VXLAN packet filter on a port::
724 testpmd> rx_vxlan_port rm (udp_port) (port_id)
729 Set hardware insertion of VLAN IDs in packets sent on a port::
731 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer]
733 For example, set a single VLAN ID (5) insertion on port 0::
737 Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1::
745 Set port based hardware insertion of VLAN ID in packets sent on a port::
747 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off)
752 Disable hardware insertion of a VLAN header in packets sent on a port::
754 testpmd> tx_vlan reset (port_id)
759 Select hardware or software calculation of the checksum when
760 transmitting a packet using the ``csum`` forwarding engine::
762 testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id)
766 * ``ip|udp|tcp|sctp`` always relate to the inner layer.
768 * ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized
769 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are
770 supported). See also the ``csum parse-tunnel`` command.
774 Check the NIC Datasheet for hardware limits.
779 Define how tunneled packets should be handled by the csum forward
782 testpmd> csum parse-tunnel (on|off) (tx_port_id)
784 If enabled, the csum forward engine will try to recognize supported
785 tunnel headers (vxlan, gre, ipip).
787 If disabled, treat tunnel packets as non-tunneled packets (a inner
788 header is handled as a packet payload).
792 The port argument is the TX port like in the ``csum set`` command.
796 Consider a packet in packet like the following::
798 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in
800 * If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
801 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the
802 ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``).
804 * If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
805 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``.
810 Display tx checksum offload configuration::
812 testpmd> csum show (port_id)
817 Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine::
819 testpmd> tso set (segsize) (port_id)
823 Check the NIC datasheet for hardware limits.
828 Display the status of TCP Segmentation Offload::
830 testpmd> tso show (port_id)
835 Add an alternative MAC address to a port::
837 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX)
842 Remove a MAC address from a port::
844 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX)
846 mac_addr add (for VF)
847 ~~~~~~~~~~~~~~~~~~~~~
849 Add an alternative MAC address for a VF to a port::
851 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX)
856 Set the default MAC address for a port::
858 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX)
860 mac_addr set (for VF)
861 ~~~~~~~~~~~~~~~~~~~~~
863 Set the MAC address for a VF from the PF::
865 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX)
870 Set the unicast hash filter(s) on/off for a port::
872 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
877 Set the promiscuous mode on for a port or for all ports.
878 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
880 testpmd> set promisc (port_id|all) (on|off)
885 Set the allmulti mode for a port or for all ports::
887 testpmd> set allmulti (port_id|all) (on|off)
889 Same as the ifconfig (8) option. Controls how multicast packets are handled.
894 Set the unicast promiscuous mode for a VF from PF.
895 It's supported by Intel i40e NICs now.
896 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
898 testpmd> set vf promisc (port_id) (vf_id) (on|off)
900 set allmulticast (for VF)
901 ~~~~~~~~~~~~~~~~~~~~~~~~~
903 Set the multicast promiscuous mode for a VF from PF.
904 It's supported by Intel i40e NICs now.
905 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
907 testpmd> set vf allmulti (port_id) (vf_id) (on|off)
909 set tx max bandwidth (for VF)
910 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
912 Set TX max absolute bandwidth (Mbps) for a VF from PF::
914 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth)
916 set tc tx min bandwidth (for VF)
917 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
919 Set all TCs' TX min relative bandwidth (%) for a VF from PF::
921 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...)
923 set tc tx max bandwidth (for VF)
924 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
926 Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF::
928 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth)
930 set tc strict link priority mode
931 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
933 Set some TCs' strict link priority mode on a physical port::
935 testpmd> set tx strict-link-priority (port_id) (tc_bitmap)
937 set tc tx min bandwidth
938 ~~~~~~~~~~~~~~~~~~~~~~~
940 Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs::
942 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...)
947 Set the link flow control parameter on a port::
949 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
950 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
951 autoneg (on|off) (port_id)
955 * ``high_water`` (integer): High threshold value to trigger XOFF.
957 * ``low_water`` (integer): Low threshold value to trigger XON.
959 * ``pause_time`` (integer): Pause quota in the Pause frame.
961 * ``send_xon`` (0/1): Send XON frame.
963 * ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
965 * ``autoneg``: Change the auto-negotiation parameter.
970 Set the priority flow control parameter on a port::
972 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
973 (pause_time) (priority) (port_id)
977 * ``high_water`` (integer): High threshold value.
979 * ``low_water`` (integer): Low threshold value.
981 * ``pause_time`` (integer): Pause quota in the Pause frame.
983 * ``priority`` (0-7): VLAN User Priority.
988 Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
990 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
992 For example, to set rx queue 2 on port 0 to mapping 5::
994 testpmd>set stat_qmap rx 0 2 5
996 set port - rx/tx (for VF)
997 ~~~~~~~~~~~~~~~~~~~~~~~~~
999 Set VF receive/transmit from a port::
1001 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
1003 set port - mac address filter (for VF)
1004 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1006 Add/Remove unicast or multicast MAC addr filter for a VF::
1008 testpmd> set port (port_id) vf (vf_id) (mac_addr) \
1009 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
1011 set port - rx mode(for VF)
1012 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1014 Set the VF receive mode of a port::
1016 testpmd> set port (port_id) vf (vf_id) \
1017 rxmode (AUPE|ROPE|BAM|MPE) (on|off)
1019 The available receive modes are:
1021 * ``AUPE``: Accepts untagged VLAN.
1023 * ``ROPE``: Accepts unicast hash.
1025 * ``BAM``: Accepts broadcast packets.
1027 * ``MPE``: Accepts all multicast packets.
1029 set port - tx_rate (for Queue)
1030 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1032 Set TX rate limitation for a queue on a port::
1034 testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1036 set port - tx_rate (for VF)
1037 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
1039 Set TX rate limitation for queues in VF on a port::
1041 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1043 set port - mirror rule
1044 ~~~~~~~~~~~~~~~~~~~~~~
1046 Set pool or vlan type mirror rule for a port::
1048 testpmd> set port (port_id) mirror-rule (rule_id) \
1049 (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1050 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1052 Set link mirror rule for a port::
1054 testpmd> set port (port_id) mirror-rule (rule_id) \
1055 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1057 For example to enable mirror traffic with vlan 0,1 to pool 0::
1059 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1061 reset port - mirror rule
1062 ~~~~~~~~~~~~~~~~~~~~~~~~
1064 Reset a mirror rule for a port::
1066 testpmd> reset port (port_id) mirror-rule (rule_id)
1071 Set the flush on RX streams before forwarding.
1072 The default is flush ``on``.
1073 Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1075 testpmd> set flush_rx off
1080 Set the bypass mode for the lowest port on bypass enabled NIC::
1082 testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1087 Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1089 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1090 mode (normal|bypass|isolate) (port_id)
1094 * ``timeout``: Enable bypass after watchdog timeout.
1096 * ``os_on``: Enable bypass when OS/board is powered on.
1098 * ``os_off``: Enable bypass when OS/board is powered off.
1100 * ``power_on``: Enable bypass when power supply is turned on.
1102 * ``power_off``: Enable bypass when power supply is turned off.
1108 Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1110 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1115 Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1117 testpmd> show bypass config (port_id)
1122 Set link up for a port::
1124 testpmd> set link-up port (port id)
1129 Set link down for a port::
1131 testpmd> set link-down port (port id)
1136 Enable E-tag insertion for a VF on a port::
1138 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1140 Disable E-tag insertion for a VF on a port::
1142 testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1144 Enable/disable E-tag stripping on a port::
1146 testpmd> E-tag set stripping (on|off) port (port_id)
1148 Enable/disable E-tag based forwarding on a port::
1150 testpmd> E-tag set forwarding (on|off) port (port_id)
1152 Add an E-tag forwarding filter on a port::
1154 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1156 Delete an E-tag forwarding filter on a port::
1157 testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1163 The following sections show functions for configuring ports.
1167 Port configuration changes only become active when forwarding is started/restarted.
1172 Attach a port specified by pci address or virtual device args::
1174 testpmd> port attach (identifier)
1176 To attach a new pci device, the device should be recognized by kernel first.
1177 Then it should be moved under DPDK management.
1178 Finally the port can be attached to testpmd.
1180 For example, to move a pci device using ixgbe under DPDK management:
1182 .. code-block:: console
1184 # Check the status of the available devices.
1185 ./usertools/dpdk-devbind.py --status
1187 Network devices using DPDK-compatible driver
1188 ============================================
1191 Network devices using kernel driver
1192 ===================================
1193 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1196 # Bind the device to igb_uio.
1197 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1200 # Recheck the status of the devices.
1201 ./usertools/dpdk-devbind.py --status
1202 Network devices using DPDK-compatible driver
1203 ============================================
1204 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1206 To attach a port created by virtual device, above steps are not needed.
1208 For example, to attach a port whose pci address is 0000:0a:00.0.
1210 .. code-block:: console
1212 testpmd> port attach 0000:0a:00.0
1213 Attaching a new port...
1214 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1215 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1216 EAL: PCI memory mapped at 0x7f83bfa00000
1217 EAL: PCI memory mapped at 0x7f83bfa80000
1218 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1219 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1220 Port 0 is attached. Now total ports is 1
1223 For example, to attach a port created by pcap PMD.
1225 .. code-block:: console
1227 testpmd> port attach net_pcap0
1228 Attaching a new port...
1229 PMD: Initializing pmd_pcap for net_pcap0
1230 PMD: Creating pcap-backed ethdev on numa socket 0
1231 Port 0 is attached. Now total ports is 1
1234 In this case, identifier is ``net_pcap0``.
1235 This identifier format is the same as ``--vdev`` format of DPDK applications.
1237 For example, to re-attach a bonded port which has been previously detached,
1238 the mode and slave parameters must be given.
1240 .. code-block:: console
1242 testpmd> port attach net_bond_0,mode=0,slave=1
1243 Attaching a new port...
1244 EAL: Initializing pmd_bond for net_bond_0
1245 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1246 Port 0 is attached. Now total ports is 1
1253 Detach a specific port::
1255 testpmd> port detach (port_id)
1257 Before detaching a port, the port should be stopped and closed.
1259 For example, to detach a pci device port 0.
1261 .. code-block:: console
1263 testpmd> port stop 0
1266 testpmd> port close 0
1270 testpmd> port detach 0
1272 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1273 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1274 EAL: PCI memory unmapped at 0x7f83bfa00000
1275 EAL: PCI memory unmapped at 0x7f83bfa80000
1279 For example, to detach a virtual device port 0.
1281 .. code-block:: console
1283 testpmd> port stop 0
1286 testpmd> port close 0
1290 testpmd> port detach 0
1292 PMD: Closing pcap ethdev on numa socket 0
1293 Port 'net_pcap0' is detached. Now total ports is 0
1296 To remove a pci device completely from the system, first detach the port from testpmd.
1297 Then the device should be moved under kernel management.
1298 Finally the device can be removed using kernel pci hotplug functionality.
1300 For example, to move a pci device under kernel management:
1302 .. code-block:: console
1304 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1306 ./usertools/dpdk-devbind.py --status
1308 Network devices using DPDK-compatible driver
1309 ============================================
1312 Network devices using kernel driver
1313 ===================================
1314 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1316 To remove a port created by a virtual device, above steps are not needed.
1321 Start all ports or a specific port::
1323 testpmd> port start (port_id|all)
1328 Stop all ports or a specific port::
1330 testpmd> port stop (port_id|all)
1335 Close all ports or a specific port::
1337 testpmd> port close (port_id|all)
1339 port start/stop queue
1340 ~~~~~~~~~~~~~~~~~~~~~
1342 Start/stop a rx/tx queue on a specific port::
1344 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1346 Only take effect when port is started.
1351 Set the speed and duplex mode for all ports or a specific port::
1353 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1354 duplex (half|full|auto)
1356 port config - queues/descriptors
1357 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1359 Set number of queues/descriptors for rxq, txq, rxd and txd::
1361 testpmd> port config all (rxq|txq|rxd|txd) (value)
1363 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1365 port config - max-pkt-len
1366 ~~~~~~~~~~~~~~~~~~~~~~~~~
1368 Set the maximum packet length::
1370 testpmd> port config all max-pkt-len (value)
1372 This is equivalent to the ``--max-pkt-len`` command-line option.
1374 port config - CRC Strip
1375 ~~~~~~~~~~~~~~~~~~~~~~~
1377 Set hardware CRC stripping on or off for all ports::
1379 testpmd> port config all crc-strip (on|off)
1381 CRC stripping is on by default.
1383 The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option.
1385 port config - scatter
1386 ~~~~~~~~~~~~~~~~~~~~~~~
1388 Set RX scatter mode on or off for all ports::
1390 testpmd> port config all scatter (on|off)
1392 RX scatter mode is off by default.
1394 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1396 port config - TX queue flags
1397 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1399 Set a hexadecimal bitmap of TX queue flags for all ports::
1401 testpmd> port config all txqflags value
1403 This command is equivalent to the ``--txqflags`` command-line option.
1405 port config - RX Checksum
1406 ~~~~~~~~~~~~~~~~~~~~~~~~~
1408 Set hardware RX checksum offload to on or off for all ports::
1410 testpmd> port config all rx-cksum (on|off)
1412 Checksum offload is off by default.
1414 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1419 Set hardware VLAN on or off for all ports::
1421 testpmd> port config all hw-vlan (on|off)
1423 Hardware VLAN is on by default.
1425 The ``off`` option is equivalent to the ``--disable-hw-vlan`` command-line option.
1427 port config - VLAN filter
1428 ~~~~~~~~~~~~~~~~~~~~~~~~~
1430 Set hardware VLAN filter on or off for all ports::
1432 testpmd> port config all hw-vlan-filter (on|off)
1434 Hardware VLAN filter is on by default.
1436 The ``off`` option is equivalent to the ``--disable-hw-vlan-filter`` command-line option.
1438 port config - VLAN strip
1439 ~~~~~~~~~~~~~~~~~~~~~~~~
1441 Set hardware VLAN strip on or off for all ports::
1443 testpmd> port config all hw-vlan-strip (on|off)
1445 Hardware VLAN strip is on by default.
1447 The ``off`` option is equivalent to the ``--disable-hw-vlan-strip`` command-line option.
1449 port config - VLAN extend
1450 ~~~~~~~~~~~~~~~~~~~~~~~~~
1452 Set hardware VLAN extend on or off for all ports::
1454 testpmd> port config all hw-vlan-extend (on|off)
1456 Hardware VLAN extend is off by default.
1458 The ``off`` option is equivalent to the ``--disable-hw-vlan-extend`` command-line option.
1460 port config - Drop Packets
1461 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1463 Set packet drop for packets with no descriptors on or off for all ports::
1465 testpmd> port config all drop-en (on|off)
1467 Packet dropping for packets with no descriptors is off by default.
1469 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1474 Set the RSS (Receive Side Scaling) mode on or off::
1476 testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1478 RSS is on by default.
1480 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1482 port config - RSS Reta
1483 ~~~~~~~~~~~~~~~~~~~~~~
1485 Set the RSS (Receive Side Scaling) redirection table::
1487 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1492 Set the DCB mode for an individual port::
1494 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1496 The traffic class should be 4 or 8.
1501 Set the number of packets per burst::
1503 testpmd> port config all burst (value)
1505 This is equivalent to the ``--burst`` command-line option.
1507 port config - Threshold
1508 ~~~~~~~~~~~~~~~~~~~~~~~
1510 Set thresholds for TX/RX queues::
1512 testpmd> port config all (threshold) (value)
1514 Where the threshold type can be:
1516 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1518 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1520 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1522 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1524 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1526 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1528 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1530 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1532 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1534 These threshold options are also available from the command-line.
1539 Set the value of ether-type for E-tag::
1541 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1543 Enable/disable the E-tag support::
1545 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1548 Link Bonding Functions
1549 ----------------------
1551 The Link Bonding functions make it possible to dynamically create and
1552 manage link bonding devices from within testpmd interactive prompt.
1554 create bonded device
1555 ~~~~~~~~~~~~~~~~~~~~
1557 Create a new bonding device::
1559 testpmd> create bonded device (mode) (socket)
1561 For example, to create a bonded device in mode 1 on socket 0::
1563 testpmd> create bonded 1 0
1564 created new bonded device (port X)
1569 Adds Ethernet device to a Link Bonding device::
1571 testpmd> add bonding slave (slave id) (port id)
1573 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
1575 testpmd> add bonding slave 6 10
1578 remove bonding slave
1579 ~~~~~~~~~~~~~~~~~~~~
1581 Removes an Ethernet slave device from a Link Bonding device::
1583 testpmd> remove bonding slave (slave id) (port id)
1585 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
1587 testpmd> remove bonding slave 6 10
1592 Set the Link Bonding mode of a Link Bonding device::
1594 testpmd> set bonding mode (value) (port id)
1596 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
1598 testpmd> set bonding mode 3 10
1603 Set an Ethernet slave device as the primary device on a Link Bonding device::
1605 testpmd> set bonding primary (slave id) (port id)
1607 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
1609 testpmd> set bonding primary 6 10
1614 Set the MAC address of a Link Bonding device::
1616 testpmd> set bonding mac (port id) (mac)
1618 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
1620 testpmd> set bonding mac 10 00:00:00:00:00:01
1622 set bonding xmit_balance_policy
1623 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1625 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
1627 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
1629 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
1631 testpmd> set bonding xmit_balance_policy 10 l34
1634 set bonding mon_period
1635 ~~~~~~~~~~~~~~~~~~~~~~
1637 Set the link status monitoring polling period in milliseconds for a bonding device.
1639 This adds support for PMD slave devices which do not support link status interrupts.
1640 When the mon_period is set to a value greater than 0 then all PMD's which do not support
1641 link status ISR will be queried every polling interval to check if their link status has changed::
1643 testpmd> set bonding mon_period (port_id) (value)
1645 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
1647 testpmd> set bonding mon_period 5 150
1653 Show the current configuration of a Link Bonding device::
1655 testpmd> show bonding config (port id)
1658 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
1659 in balance mode with a transmission policy of layer 2+3::
1661 testpmd> show bonding config 9
1663 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
1665 Active Slaves (3): [1 3 4]
1672 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
1673 This is mainly useful for debugging purposes.
1674 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
1675 and fields that can be accessed.
1680 Display the value of a port register::
1682 testpmd> read reg (port_id) (address)
1684 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
1686 testpmd> read reg 0 0xEE00
1687 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
1692 Display a port register bit field::
1694 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
1696 For example, reading the lowest two bits from the register in the example above::
1698 testpmd> read regfield 0 0xEE00 0 1
1699 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
1704 Display a single port register bit::
1706 testpmd> read regbit (port_id) (address) (bit_x)
1708 For example, reading the lowest bit from the register in the example above::
1710 testpmd> read regbit 0 0xEE00 0
1711 port 0 PCI register at offset 0xEE00: bit 0=1
1716 Set the value of a port register::
1718 testpmd> write reg (port_id) (address) (value)
1720 For example, to clear a register::
1722 testpmd> write reg 0 0xEE00 0x0
1723 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
1728 Set bit field of a port register::
1730 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
1732 For example, writing to the register cleared in the example above::
1734 testpmd> write regfield 0 0xEE00 0 1 2
1735 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
1740 Set single bit value of a port register::
1742 testpmd> write regbit (port_id) (address) (bit_x) (value)
1744 For example, to set the high bit in the register from the example above::
1746 testpmd> write regbit 0 0xEE00 31 1
1747 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
1753 This section details the available filter functions that are available.
1755 Note these functions interface the deprecated legacy filtering framework,
1756 superseded by *rte_flow*. See `Flow rules management`_.
1759 ~~~~~~~~~~~~~~~~~~~~
1761 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
1763 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
1764 ethertype (ether_type) (drop|fwd) queue (queue_id)
1766 The available information parameters are:
1768 * ``port_id``: The port which the Ethertype filter assigned on.
1770 * ``mac_addr``: Compare destination mac address.
1772 * ``mac_ignr``: Ignore destination mac address match.
1774 * ``mac_address``: Destination mac address to match.
1776 * ``ether_type``: The EtherType value want to match,
1777 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
1779 * ``queue_id``: The receive queue associated with this EtherType filter.
1780 It is meaningless when deleting or dropping.
1782 Example, to add/remove an ethertype filter rule::
1784 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
1785 ethertype 0x0806 fwd queue 3
1787 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
1788 ethertype 0x0806 fwd queue 3
1793 Add or delete a 2-tuple filter,
1794 which identifies packets by specific protocol and destination TCP/UDP port
1795 and forwards packets into one of the receive queues::
1797 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
1798 protocol (protocol_value) mask (mask_value) \
1799 tcp_flags (tcp_flags_value) priority (prio_value) \
1802 The available information parameters are:
1804 * ``port_id``: The port which the 2-tuple filter assigned on.
1806 * ``dst_port_value``: Destination port in L4.
1808 * ``protocol_value``: IP L4 protocol.
1810 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
1812 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
1814 * ``prio_value``: Priority of this filter.
1816 * ``queue_id``: The receive queue associated with this 2-tuple filter.
1818 Example, to add/remove an 2tuple filter rule::
1820 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
1821 tcp_flags 0x02 priority 3 queue 3
1823 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
1824 tcp_flags 0x02 priority 3 queue 3
1829 Add or delete a 5-tuple filter,
1830 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
1831 and routes packets into one of the receive queues::
1833 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
1834 (src_address) dst_port (dst_port_value) \
1835 src_port (src_port_value) protocol (protocol_value) \
1836 mask (mask_value) tcp_flags (tcp_flags_value) \
1837 priority (prio_value) queue (queue_id)
1839 The available information parameters are:
1841 * ``port_id``: The port which the 5-tuple filter assigned on.
1843 * ``dst_address``: Destination IP address.
1845 * ``src_address``: Source IP address.
1847 * ``dst_port_value``: TCP/UDP destination port.
1849 * ``src_port_value``: TCP/UDP source port.
1851 * ``protocol_value``: L4 protocol.
1853 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
1855 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
1857 * ``prio_value``: The priority of this filter.
1859 * ``queue_id``: The receive queue associated with this 5-tuple filter.
1861 Example, to add/remove an 5tuple filter rule::
1863 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1864 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1865 flags 0x0 priority 3 queue 3
1867 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1868 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1869 flags 0x0 priority 3 queue 3
1874 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
1876 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
1878 The available information parameters are:
1880 * ``port_id``: The port which the SYN filter assigned on.
1882 * ``high``: This SYN filter has higher priority than other filters.
1884 * ``low``: This SYN filter has lower priority than other filters.
1886 * ``queue_id``: The receive queue associated with this SYN filter
1890 testpmd> syn_filter 0 add priority high queue 3
1895 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
1896 and routed into one of the receive queues::
1898 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
1899 mask (mask_value) priority (prio_value) queue (queue_id)
1901 The available information parameters are:
1903 * ``port_id``: The port which the Flex filter is assigned on.
1905 * ``len_value``: Filter length in bytes, no greater than 128.
1907 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
1909 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
1911 * ``prio_value``: The priority of this filter.
1913 * ``queue_id``: The receive queue associated with this Flex filter.
1917 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
1918 mask 000C priority 3 queue 3
1920 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
1921 mask 000C priority 3 queue 3
1924 .. _testpmd_flow_director:
1926 flow_director_filter
1927 ~~~~~~~~~~~~~~~~~~~~
1929 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
1931 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
1932 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
1934 * Perfect match filters.
1935 The hardware checks a match between the masked fields of the received packets and the programmed filters.
1936 The masked fields are for IP flow.
1938 * Signature filters.
1939 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
1941 * Perfect-mac-vlan match filters.
1942 The hardware checks a match between the masked fields of the received packets and the programmed filters.
1943 The masked fields are for MAC VLAN flow.
1945 * Perfect-tunnel match filters.
1946 The hardware checks a match between the masked fields of the received packets and the programmed filters.
1947 The masked fields are for tunnel flow.
1949 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
1950 per flow type and the flexible payload.
1952 The Flow Director can also mask out parts of all of these fields so that filters
1953 are only applied to certain fields or parts of the fields.
1955 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
1957 # Commands to add flow director filters of different flow types::
1959 flow_director_filter (port_id) mode IP (add|del|update) \
1960 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
1961 src (src_ip_address) dst (dst_ip_address) \
1962 tos (tos_value) proto (proto_value) ttl (ttl_value) \
1963 vlan (vlan_value) flexbytes (flexbytes_value) \
1964 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
1967 flow_director_filter (port_id) mode IP (add|del|update) \
1968 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
1969 src (src_ip_address) (src_port) \
1970 dst (dst_ip_address) (dst_port) \
1971 tos (tos_value) ttl (ttl_value) \
1972 vlan (vlan_value) flexbytes (flexbytes_value) \
1973 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
1976 flow_director_filter (port_id) mode IP (add|del|update) \
1977 flow (ipv4-sctp|ipv6-sctp) \
1978 src (src_ip_address) (src_port) \
1979 dst (dst_ip_address) (dst_port) \
1980 tos (tos_value) ttl (ttl_value) \
1981 tag (verification_tag) vlan (vlan_value) \
1982 flexbytes (flexbytes_value) (drop|fwd) \
1983 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
1985 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
1986 ether (ethertype) flexbytes (flexbytes_value) \
1987 (drop|fwd) pf|vf(vf_id) queue (queue_id)
1990 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
1991 mac (mac_address) vlan (vlan_value) \
1992 flexbytes (flexbytes_value) (drop|fwd) \
1993 queue (queue_id) fd_id (fd_id_value)
1995 flow_director_filter (port_id) mode Tunnel (add|del|update) \
1996 mac (mac_address) vlan (vlan_value) \
1997 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
1998 flexbytes (flexbytes_value) (drop|fwd) \
1999 queue (queue_id) fd_id (fd_id_value)
2001 For example, to add an ipv4-udp flow type filter::
2003 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
2004 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
2005 fwd pf queue 1 fd_id 1
2007 For example, add an ipv4-other flow type filter::
2009 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
2010 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
2011 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
2016 Flush all flow director filters on a device::
2018 testpmd> flush_flow_director (port_id)
2020 Example, to flush all flow director filter on port 0::
2022 testpmd> flush_flow_director 0
2027 Set flow director's input masks::
2029 flow_director_mask (port_id) mode IP vlan (vlan_value) \
2030 src_mask (ipv4_src) (ipv6_src) (src_port) \
2031 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2033 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2035 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2036 mac (mac_value) tunnel-type (tunnel_type_value) \
2037 tunnel-id (tunnel_id_value)
2039 Example, to set flow director mask on port 0::
2041 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2042 src_mask 255.255.255.255 \
2043 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2044 dst_mask 255.255.255.255 \
2045 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2047 flow_director_flex_mask
2048 ~~~~~~~~~~~~~~~~~~~~~~~
2050 set masks of flow director's flexible payload based on certain flow type::
2052 testpmd> flow_director_flex_mask (port_id) \
2053 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2054 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2055 l2_payload|all) (mask)
2057 Example, to set flow director's flex mask for all flow type on port 0::
2059 testpmd> flow_director_flex_mask 0 flow all \
2060 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2063 flow_director_flex_payload
2064 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2066 Configure flexible payload selection::
2068 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2070 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2072 testpmd> flow_director_flex_payload 0 l4 \
2073 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2075 get_sym_hash_ena_per_port
2076 ~~~~~~~~~~~~~~~~~~~~~~~~~
2078 Get symmetric hash enable configuration per port::
2080 get_sym_hash_ena_per_port (port_id)
2082 For example, to get symmetric hash enable configuration of port 1::
2084 testpmd> get_sym_hash_ena_per_port 1
2086 set_sym_hash_ena_per_port
2087 ~~~~~~~~~~~~~~~~~~~~~~~~~
2089 Set symmetric hash enable configuration per port to enable or disable::
2091 set_sym_hash_ena_per_port (port_id) (enable|disable)
2093 For example, to set symmetric hash enable configuration of port 1 to enable::
2095 testpmd> set_sym_hash_ena_per_port 1 enable
2097 get_hash_global_config
2098 ~~~~~~~~~~~~~~~~~~~~~~
2100 Get the global configurations of hash filters::
2102 get_hash_global_config (port_id)
2104 For example, to get the global configurations of hash filters of port 1::
2106 testpmd> get_hash_global_config 1
2108 set_hash_global_config
2109 ~~~~~~~~~~~~~~~~~~~~~~
2111 Set the global configurations of hash filters::
2113 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2114 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2115 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload) \
2118 For example, to enable simple_xor for flow type of ipv6 on port 2::
2120 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2125 Set the input set for hash::
2127 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2128 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2129 l2_payload) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2130 ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2131 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2132 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2133 fld-8th|none) (select|add)
2135 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2137 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2142 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2143 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2145 Set the input set for flow director::
2147 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2148 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2149 l2_payload) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2150 ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2151 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2152 sctp-dst-port|sctp-veri-tag|none) (select|add)
2154 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2156 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2161 Set different GRE key length for input set::
2163 global_config (port_id) gre-key-len (number in bytes)
2165 For example to set GRE key length for input set to 4 bytes on port 0::
2167 testpmd> global_config 0 gre-key-len 4
2170 .. _testpmd_rte_flow:
2172 Flow rules management
2173 ---------------------
2175 Control of the generic flow API (*rte_flow*) is fully exposed through the
2176 ``flow`` command (validation, creation, destruction and queries).
2178 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2179 features simultaneously may cause undefined side-effects and is therefore
2185 Because the ``flow`` command uses dynamic tokens to handle the large number
2186 of possible flow rules combinations, its behavior differs slightly from
2187 other commands, in particular:
2189 - Pressing *?* or the *<tab>* key displays contextual help for the current
2190 token, not that of the entire command.
2192 - Optional and repeated parameters are supported (provided they are listed
2193 in the contextual help).
2195 The first parameter stands for the operation mode. Possible operations and
2196 their general syntax are described below. They are covered in detail in the
2199 - Check whether a flow rule can be created::
2201 flow validate {port_id}
2202 [group {group_id}] [priority {level}] [ingress] [egress]
2203 pattern {item} [/ {item} [...]] / end
2204 actions {action} [/ {action} [...]] / end
2206 - Create a flow rule::
2208 flow create {port_id}
2209 [group {group_id}] [priority {level}] [ingress] [egress]
2210 pattern {item} [/ {item} [...]] / end
2211 actions {action} [/ {action} [...]] / end
2213 - Destroy specific flow rules::
2215 flow destroy {port_id} rule {rule_id} [...]
2217 - Destroy all flow rules::
2219 flow flush {port_id}
2221 - Query an existing flow rule::
2223 flow query {port_id} {rule_id} {action}
2225 - List existing flow rules sorted by priority, filtered by group
2228 flow list {port_id} [group {group_id}] [...]
2230 Validating flow rules
2231 ~~~~~~~~~~~~~~~~~~~~~
2233 ``flow validate`` reports whether a flow rule would be accepted by the
2234 underlying device in its current state but stops short of creating it. It is
2235 bound to ``rte_flow_validate()``::
2237 flow validate {port_id}
2238 [group {group_id}] [priority {level}] [ingress] [egress]
2239 pattern {item} [/ {item} [...]] / end
2240 actions {action} [/ {action} [...]] / end
2242 If successful, it will show::
2246 Otherwise it will show an error message of the form::
2248 Caught error type [...] ([...]): [...]
2250 This command uses the same parameters as ``flow create``, their format is
2251 described in `Creating flow rules`_.
2253 Check whether redirecting any Ethernet packet received on port 0 to RX queue
2254 index 6 is supported::
2256 testpmd> flow validate 0 ingress pattern eth / end
2257 actions queue index 6 / end
2261 Port 0 does not support TCPv6 rules::
2263 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
2265 Caught error type 9 (specific pattern item): Invalid argument
2271 ``flow create`` validates and creates the specified flow rule. It is bound
2272 to ``rte_flow_create()``::
2274 flow create {port_id}
2275 [group {group_id}] [priority {level}] [ingress] [egress]
2276 pattern {item} [/ {item} [...]] / end
2277 actions {action} [/ {action} [...]] / end
2279 If successful, it will return a flow rule ID usable with other commands::
2281 Flow rule #[...] created
2283 Otherwise it will show an error message of the form::
2285 Caught error type [...] ([...]): [...]
2287 Parameters describe in the following order:
2289 - Attributes (*group*, *priority*, *ingress*, *egress* tokens).
2290 - A matching pattern, starting with the *pattern* token and terminated by an
2292 - Actions, starting with the *actions* token and terminated by an *end*
2295 These translate directly to *rte_flow* objects provided as-is to the
2296 underlying functions.
2298 The shortest valid definition only comprises mandatory tokens::
2300 testpmd> flow create 0 pattern end actions end
2302 Note that PMDs may refuse rules that essentially do nothing such as this
2305 **All unspecified object values are automatically initialized to 0.**
2310 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
2311 specified before the ``pattern`` token.
2313 - ``group {group id}``: priority group.
2314 - ``priority {level}``: priority level within group.
2315 - ``ingress``: rule applies to ingress traffic.
2316 - ``egress``: rule applies to egress traffic.
2318 Each instance of an attribute specified several times overrides the previous
2319 value as shown below (group 4 is used)::
2321 testpmd> flow create 0 group 42 group 24 group 4 [...]
2323 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
2325 While not specifying a direction is an error, some rules may allow both
2328 Most rules affect RX therefore contain the ``ingress`` token::
2330 testpmd> flow create 0 ingress pattern [...]
2335 A matching pattern starts after the ``pattern`` token. It is made of pattern
2336 items and is terminated by a mandatory ``end`` item.
2338 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
2339 rte_flow_item_type``).
2341 The ``/`` token is used as a separator between pattern items as shown
2344 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
2346 Note that protocol items like these must be stacked from lowest to highest
2347 layer to make sense. For instance, the following rule is either invalid or
2348 unlikely to match any packet::
2350 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
2352 More information on these restrictions can be found in the *rte_flow*
2355 Several items support additional specification structures, for example
2356 ``ipv4`` allows specifying source and destination addresses as follows::
2358 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
2359 dst is 10.2.0.0 / end [...]
2361 This rule matches all IPv4 traffic with the specified properties.
2363 In this example, ``src`` and ``dst`` are field names of the underlying
2364 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
2365 in a similar fashion.
2367 The ``is`` token means that the subsequent value must be matched exactly,
2368 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
2369 accordingly. Possible assignment tokens are:
2371 - ``is``: match value perfectly (with full bit-mask).
2372 - ``spec``: match value according to configured bit-mask.
2373 - ``last``: specify upper bound to establish a range.
2374 - ``mask``: specify bit-mask with relevant bits set to one.
2375 - ``prefix``: generate bit-mask from a prefix length.
2377 These yield identical results::
2379 ipv4 src is 10.1.1.1
2383 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
2387 ipv4 src spec 10.1.1.1 src prefix 32
2391 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
2395 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
2397 Inclusive ranges can be defined with ``last``::
2399 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
2401 Note that ``mask`` affects both ``spec`` and ``last``::
2403 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
2404 # matches 10.1.0.0 to 10.2.255.255
2406 Properties can be modified multiple times::
2408 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
2412 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
2417 This section lists supported pattern items and their attributes, if any.
2419 - ``end``: end list of pattern items.
2421 - ``void``: no-op pattern item.
2423 - ``invert``: perform actions when pattern does not match.
2425 - ``any``: match any protocol for the current layer.
2427 - ``num {unsigned}``: number of layers covered.
2429 - ``pf``: match packets addressed to the physical function.
2431 - ``vf``: match packets addressed to a virtual function ID.
2433 - ``id {unsigned}``: destination VF ID.
2435 - ``port``: device-specific physical port index to use.
2437 - ``index {unsigned}``: physical port index.
2439 - ``raw``: match an arbitrary byte string.
2441 - ``relative {boolean}``: look for pattern after the previous item.
2442 - ``search {boolean}``: search pattern from offset (see also limit).
2443 - ``offset {integer}``: absolute or relative offset for pattern.
2444 - ``limit {unsigned}``: search area limit for start of pattern.
2445 - ``pattern {string}``: byte string to look for.
2447 - ``eth``: match Ethernet header.
2449 - ``dst {MAC-48}``: destination MAC.
2450 - ``src {MAC-48}``: source MAC.
2451 - ``type {unsigned}``: EtherType.
2453 - ``vlan``: match 802.1Q/ad VLAN tag.
2455 - ``tpid {unsigned}``: tag protocol identifier.
2456 - ``tci {unsigned}``: tag control information.
2457 - ``pcp {unsigned}``: priority code point.
2458 - ``dei {unsigned}``: drop eligible indicator.
2459 - ``vid {unsigned}``: VLAN identifier.
2461 - ``ipv4``: match IPv4 header.
2463 - ``tos {unsigned}``: type of service.
2464 - ``ttl {unsigned}``: time to live.
2465 - ``proto {unsigned}``: next protocol ID.
2466 - ``src {ipv4 address}``: source address.
2467 - ``dst {ipv4 address}``: destination address.
2469 - ``ipv6``: match IPv6 header.
2471 - ``tc {unsigned}``: traffic class.
2472 - ``flow {unsigned}``: flow label.
2473 - ``proto {unsigned}``: protocol (next header).
2474 - ``hop {unsigned}``: hop limit.
2475 - ``src {ipv6 address}``: source address.
2476 - ``dst {ipv6 address}``: destination address.
2478 - ``icmp``: match ICMP header.
2480 - ``type {unsigned}``: ICMP packet type.
2481 - ``code {unsigned}``: ICMP packet code.
2483 - ``udp``: match UDP header.
2485 - ``src {unsigned}``: UDP source port.
2486 - ``dst {unsigned}``: UDP destination port.
2488 - ``tcp``: match TCP header.
2490 - ``src {unsigned}``: TCP source port.
2491 - ``dst {unsigned}``: TCP destination port.
2493 - ``sctp``: match SCTP header.
2495 - ``src {unsigned}``: SCTP source port.
2496 - ``dst {unsigned}``: SCTP destination port.
2497 - ``tag {unsigned}``: validation tag.
2498 - ``cksum {unsigned}``: checksum.
2500 - ``vxlan``: match VXLAN header.
2502 - ``vni {unsigned}``: VXLAN identifier.
2504 - ``mpls``: match MPLS header.
2506 - ``label {unsigned}``: MPLS label.
2508 - ``gre``: match GRE header.
2510 - ``protocol {unsigned}``: protocol type.
2515 A list of actions starts after the ``actions`` token in the same fashion as
2516 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
2517 terminated by a mandatory ``end`` action.
2519 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
2520 rte_flow_action_type``).
2522 Dropping all incoming UDPv4 packets can be expressed as follows::
2524 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2527 Several actions have configurable properties which must be specified when
2528 there is no valid default value. For example, ``queue`` requires a target
2531 This rule redirects incoming UDPv4 traffic to queue index 6::
2533 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2534 actions queue index 6 / end
2536 While this one could be rejected by PMDs (unspecified queue index)::
2538 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2541 As defined by *rte_flow*, the list is not ordered, all actions of a given
2542 rule are performed simultaneously. These are equivalent::
2544 queue index 6 / void / mark id 42 / end
2548 void / mark id 42 / queue index 6 / end
2550 All actions in a list should have different types, otherwise only the last
2551 action of a given type is taken into account::
2553 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
2557 drop / drop / drop / end # drop is performed only once
2561 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
2563 Considering they are performed simultaneously, opposite and overlapping
2564 actions can sometimes be combined when the end result is unambiguous::
2566 drop / queue index 6 / end # drop has no effect
2570 drop / dup index 6 / end # same as above
2574 queue index 6 / rss queues 6 7 8 / end # queue has no effect
2578 drop / passthru / end # drop has no effect
2580 Note that PMDs may still refuse such combinations.
2585 This section lists supported actions and their attributes, if any.
2587 - ``end``: end list of actions.
2589 - ``void``: no-op action.
2591 - ``passthru``: let subsequent rule process matched packets.
2593 - ``mark``: attach 32 bit value to packets.
2595 - ``id {unsigned}``: 32 bit value to return with packets.
2597 - ``flag``: flag packets.
2599 - ``queue``: assign packets to a given queue index.
2601 - ``index {unsigned}``: queue index to use.
2603 - ``drop``: drop packets (note: passthru has priority).
2605 - ``count``: enable counters for this rule.
2607 - ``dup``: duplicate packets to a given queue index.
2609 - ``index {unsigned}``: queue index to duplicate packets to.
2611 - ``rss``: spread packets among several queues.
2613 - ``queues [{unsigned} [...]] end``: queue indices to use.
2615 - ``pf``: redirect packets to physical device function.
2617 - ``vf``: redirect packets to virtual device function.
2619 - ``original {boolean}``: use original VF ID if possible.
2620 - ``id {unsigned}``: VF ID to redirect packets to.
2622 Destroying flow rules
2623 ~~~~~~~~~~~~~~~~~~~~~
2625 ``flow destroy`` destroys one or more rules from their rule ID (as returned
2626 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
2627 times as necessary::
2629 flow destroy {port_id} rule {rule_id} [...]
2631 If successful, it will show::
2633 Flow rule #[...] destroyed
2635 It does not report anything for rule IDs that do not exist. The usual error
2636 message is shown when a rule cannot be destroyed::
2638 Caught error type [...] ([...]): [...]
2640 ``flow flush`` destroys all rules on a device and does not take extra
2641 arguments. It is bound to ``rte_flow_flush()``::
2643 flow flush {port_id}
2645 Any errors are reported as above.
2647 Creating several rules and destroying them::
2649 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2650 actions queue index 2 / end
2651 Flow rule #0 created
2652 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2653 actions queue index 3 / end
2654 Flow rule #1 created
2655 testpmd> flow destroy 0 rule 0 rule 1
2656 Flow rule #1 destroyed
2657 Flow rule #0 destroyed
2660 The same result can be achieved using ``flow flush``::
2662 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2663 actions queue index 2 / end
2664 Flow rule #0 created
2665 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2666 actions queue index 3 / end
2667 Flow rule #1 created
2668 testpmd> flow flush 0
2671 Non-existent rule IDs are ignored::
2673 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2674 actions queue index 2 / end
2675 Flow rule #0 created
2676 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2677 actions queue index 3 / end
2678 Flow rule #1 created
2679 testpmd> flow destroy 0 rule 42 rule 10 rule 2
2681 testpmd> flow destroy 0 rule 0
2682 Flow rule #0 destroyed
2688 ``flow query`` queries a specific action of a flow rule having that
2689 ability. Such actions collect information that can be reported using this
2690 command. It is bound to ``rte_flow_query()``::
2692 flow query {port_id} {rule_id} {action}
2694 If successful, it will display either the retrieved data for known actions
2695 or the following message::
2697 Cannot display result for action type [...] ([...])
2699 Otherwise, it will complain either that the rule does not exist or that some
2702 Flow rule #[...] not found
2706 Caught error type [...] ([...]): [...]
2708 Currently only the ``count`` action is supported. This action reports the
2709 number of packets that hit the flow rule and the total number of bytes. Its
2710 output has the following format::
2713 hits_set: [...] # whether "hits" contains a valid value
2714 bytes_set: [...] # whether "bytes" contains a valid value
2715 hits: [...] # number of packets
2716 bytes: [...] # number of bytes
2718 Querying counters for TCPv6 packets redirected to queue 6::
2720 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
2721 actions queue index 6 / count / end
2722 Flow rule #4 created
2723 testpmd> flow query 0 4 count
2734 ``flow list`` lists existing flow rules sorted by priority and optionally
2735 filtered by group identifiers::
2737 flow list {port_id} [group {group_id}] [...]
2739 This command only fails with the following message if the device does not
2744 Output consists of a header line followed by a short description of each
2745 flow rule, one per line. There is no output at all when no flow rules are
2746 configured on the device::
2748 ID Group Prio Attr Rule
2749 [...] [...] [...] [...] [...]
2751 ``Attr`` column flags:
2753 - ``i`` for ``ingress``.
2754 - ``e`` for ``egress``.
2756 Creating several flow rules and listing them::
2758 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2759 actions queue index 6 / end
2760 Flow rule #0 created
2761 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2762 actions queue index 2 / end
2763 Flow rule #1 created
2764 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
2765 actions rss queues 6 7 8 end / end
2766 Flow rule #2 created
2767 testpmd> flow list 0
2768 ID Group Prio Attr Rule
2769 0 0 0 i- ETH IPV4 => QUEUE
2770 1 0 0 i- ETH IPV6 => QUEUE
2771 2 0 5 i- ETH IPV4 UDP => RSS
2774 Rules are sorted by priority (i.e. group ID first, then priority level)::
2776 testpmd> flow list 1
2777 ID Group Prio Attr Rule
2778 0 0 0 i- ETH => COUNT
2779 6 0 500 i- ETH IPV6 TCP => DROP COUNT
2780 5 0 1000 i- ETH IPV6 ICMP => QUEUE
2781 1 24 0 i- ETH IPV4 UDP => QUEUE
2782 4 24 10 i- ETH IPV4 TCP => DROP
2783 3 24 20 i- ETH IPV4 => DROP
2784 2 24 42 i- ETH IPV4 UDP => QUEUE
2785 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
2788 Output can be limited to specific groups::
2790 testpmd> flow list 1 group 0 group 63
2791 ID Group Prio Attr Rule
2792 0 0 0 i- ETH => COUNT
2793 6 0 500 i- ETH IPV6 TCP => DROP COUNT
2794 5 0 1000 i- ETH IPV6 ICMP => QUEUE
2795 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE