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)
1162 List all items from the ptype mapping table::
1164 testpmd> ptype mapping get (port_id) (valid_only)
1168 * ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0).
1170 Replace a specific or a group of software defined ptype with a new one::
1172 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type)
1176 * ``target``: A specific software ptype or a mask to represent a group of software ptypes.
1178 * ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1).
1180 * ``pkt_type``: The new software ptype to replace the old ones.
1182 Update hardware defined ptype to software defined packet type mapping table::
1184 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype)
1188 * ``hw_ptype``: hardware ptype as the index of the ptype mapping table.
1190 * ``sw_ptype``: software ptype as the value of the ptype mapping table.
1192 Reset ptype mapping table::
1194 testpmd> ptype mapping reset (port_id)
1199 The following sections show functions for configuring ports.
1203 Port configuration changes only become active when forwarding is started/restarted.
1208 Attach a port specified by pci address or virtual device args::
1210 testpmd> port attach (identifier)
1212 To attach a new pci device, the device should be recognized by kernel first.
1213 Then it should be moved under DPDK management.
1214 Finally the port can be attached to testpmd.
1216 For example, to move a pci device using ixgbe under DPDK management:
1218 .. code-block:: console
1220 # Check the status of the available devices.
1221 ./usertools/dpdk-devbind.py --status
1223 Network devices using DPDK-compatible driver
1224 ============================================
1227 Network devices using kernel driver
1228 ===================================
1229 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1232 # Bind the device to igb_uio.
1233 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1236 # Recheck the status of the devices.
1237 ./usertools/dpdk-devbind.py --status
1238 Network devices using DPDK-compatible driver
1239 ============================================
1240 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1242 To attach a port created by virtual device, above steps are not needed.
1244 For example, to attach a port whose pci address is 0000:0a:00.0.
1246 .. code-block:: console
1248 testpmd> port attach 0000:0a:00.0
1249 Attaching a new port...
1250 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1251 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1252 EAL: PCI memory mapped at 0x7f83bfa00000
1253 EAL: PCI memory mapped at 0x7f83bfa80000
1254 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1255 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1256 Port 0 is attached. Now total ports is 1
1259 For example, to attach a port created by pcap PMD.
1261 .. code-block:: console
1263 testpmd> port attach net_pcap0
1264 Attaching a new port...
1265 PMD: Initializing pmd_pcap for net_pcap0
1266 PMD: Creating pcap-backed ethdev on numa socket 0
1267 Port 0 is attached. Now total ports is 1
1270 In this case, identifier is ``net_pcap0``.
1271 This identifier format is the same as ``--vdev`` format of DPDK applications.
1273 For example, to re-attach a bonded port which has been previously detached,
1274 the mode and slave parameters must be given.
1276 .. code-block:: console
1278 testpmd> port attach net_bond_0,mode=0,slave=1
1279 Attaching a new port...
1280 EAL: Initializing pmd_bond for net_bond_0
1281 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1282 Port 0 is attached. Now total ports is 1
1289 Detach a specific port::
1291 testpmd> port detach (port_id)
1293 Before detaching a port, the port should be stopped and closed.
1295 For example, to detach a pci device port 0.
1297 .. code-block:: console
1299 testpmd> port stop 0
1302 testpmd> port close 0
1306 testpmd> port detach 0
1308 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1309 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1310 EAL: PCI memory unmapped at 0x7f83bfa00000
1311 EAL: PCI memory unmapped at 0x7f83bfa80000
1315 For example, to detach a virtual device port 0.
1317 .. code-block:: console
1319 testpmd> port stop 0
1322 testpmd> port close 0
1326 testpmd> port detach 0
1328 PMD: Closing pcap ethdev on numa socket 0
1329 Port 'net_pcap0' is detached. Now total ports is 0
1332 To remove a pci device completely from the system, first detach the port from testpmd.
1333 Then the device should be moved under kernel management.
1334 Finally the device can be removed using kernel pci hotplug functionality.
1336 For example, to move a pci device under kernel management:
1338 .. code-block:: console
1340 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1342 ./usertools/dpdk-devbind.py --status
1344 Network devices using DPDK-compatible driver
1345 ============================================
1348 Network devices using kernel driver
1349 ===================================
1350 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1352 To remove a port created by a virtual device, above steps are not needed.
1357 Start all ports or a specific port::
1359 testpmd> port start (port_id|all)
1364 Stop all ports or a specific port::
1366 testpmd> port stop (port_id|all)
1371 Close all ports or a specific port::
1373 testpmd> port close (port_id|all)
1375 port start/stop queue
1376 ~~~~~~~~~~~~~~~~~~~~~
1378 Start/stop a rx/tx queue on a specific port::
1380 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1382 Only take effect when port is started.
1387 Set the speed and duplex mode for all ports or a specific port::
1389 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1390 duplex (half|full|auto)
1392 port config - queues/descriptors
1393 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1395 Set number of queues/descriptors for rxq, txq, rxd and txd::
1397 testpmd> port config all (rxq|txq|rxd|txd) (value)
1399 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1401 port config - max-pkt-len
1402 ~~~~~~~~~~~~~~~~~~~~~~~~~
1404 Set the maximum packet length::
1406 testpmd> port config all max-pkt-len (value)
1408 This is equivalent to the ``--max-pkt-len`` command-line option.
1410 port config - CRC Strip
1411 ~~~~~~~~~~~~~~~~~~~~~~~
1413 Set hardware CRC stripping on or off for all ports::
1415 testpmd> port config all crc-strip (on|off)
1417 CRC stripping is on by default.
1419 The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option.
1421 port config - scatter
1422 ~~~~~~~~~~~~~~~~~~~~~~~
1424 Set RX scatter mode on or off for all ports::
1426 testpmd> port config all scatter (on|off)
1428 RX scatter mode is off by default.
1430 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1432 port config - TX queue flags
1433 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1435 Set a hexadecimal bitmap of TX queue flags for all ports::
1437 testpmd> port config all txqflags value
1439 This command is equivalent to the ``--txqflags`` command-line option.
1441 port config - RX Checksum
1442 ~~~~~~~~~~~~~~~~~~~~~~~~~
1444 Set hardware RX checksum offload to on or off for all ports::
1446 testpmd> port config all rx-cksum (on|off)
1448 Checksum offload is off by default.
1450 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1455 Set hardware VLAN on or off for all ports::
1457 testpmd> port config all hw-vlan (on|off)
1459 Hardware VLAN is on by default.
1461 The ``off`` option is equivalent to the ``--disable-hw-vlan`` command-line option.
1463 port config - VLAN filter
1464 ~~~~~~~~~~~~~~~~~~~~~~~~~
1466 Set hardware VLAN filter on or off for all ports::
1468 testpmd> port config all hw-vlan-filter (on|off)
1470 Hardware VLAN filter is on by default.
1472 The ``off`` option is equivalent to the ``--disable-hw-vlan-filter`` command-line option.
1474 port config - VLAN strip
1475 ~~~~~~~~~~~~~~~~~~~~~~~~
1477 Set hardware VLAN strip on or off for all ports::
1479 testpmd> port config all hw-vlan-strip (on|off)
1481 Hardware VLAN strip is on by default.
1483 The ``off`` option is equivalent to the ``--disable-hw-vlan-strip`` command-line option.
1485 port config - VLAN extend
1486 ~~~~~~~~~~~~~~~~~~~~~~~~~
1488 Set hardware VLAN extend on or off for all ports::
1490 testpmd> port config all hw-vlan-extend (on|off)
1492 Hardware VLAN extend is off by default.
1494 The ``off`` option is equivalent to the ``--disable-hw-vlan-extend`` command-line option.
1496 port config - Drop Packets
1497 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1499 Set packet drop for packets with no descriptors on or off for all ports::
1501 testpmd> port config all drop-en (on|off)
1503 Packet dropping for packets with no descriptors is off by default.
1505 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1510 Set the RSS (Receive Side Scaling) mode on or off::
1512 testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1514 RSS is on by default.
1516 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1518 port config - RSS Reta
1519 ~~~~~~~~~~~~~~~~~~~~~~
1521 Set the RSS (Receive Side Scaling) redirection table::
1523 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1528 Set the DCB mode for an individual port::
1530 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1532 The traffic class should be 4 or 8.
1537 Set the number of packets per burst::
1539 testpmd> port config all burst (value)
1541 This is equivalent to the ``--burst`` command-line option.
1543 port config - Threshold
1544 ~~~~~~~~~~~~~~~~~~~~~~~
1546 Set thresholds for TX/RX queues::
1548 testpmd> port config all (threshold) (value)
1550 Where the threshold type can be:
1552 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1554 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1556 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1558 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1560 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1562 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1564 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1566 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1568 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1570 These threshold options are also available from the command-line.
1575 Set the value of ether-type for E-tag::
1577 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1579 Enable/disable the E-tag support::
1581 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1584 Link Bonding Functions
1585 ----------------------
1587 The Link Bonding functions make it possible to dynamically create and
1588 manage link bonding devices from within testpmd interactive prompt.
1590 create bonded device
1591 ~~~~~~~~~~~~~~~~~~~~
1593 Create a new bonding device::
1595 testpmd> create bonded device (mode) (socket)
1597 For example, to create a bonded device in mode 1 on socket 0::
1599 testpmd> create bonded 1 0
1600 created new bonded device (port X)
1605 Adds Ethernet device to a Link Bonding device::
1607 testpmd> add bonding slave (slave id) (port id)
1609 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
1611 testpmd> add bonding slave 6 10
1614 remove bonding slave
1615 ~~~~~~~~~~~~~~~~~~~~
1617 Removes an Ethernet slave device from a Link Bonding device::
1619 testpmd> remove bonding slave (slave id) (port id)
1621 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
1623 testpmd> remove bonding slave 6 10
1628 Set the Link Bonding mode of a Link Bonding device::
1630 testpmd> set bonding mode (value) (port id)
1632 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
1634 testpmd> set bonding mode 3 10
1639 Set an Ethernet slave device as the primary device on a Link Bonding device::
1641 testpmd> set bonding primary (slave id) (port id)
1643 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
1645 testpmd> set bonding primary 6 10
1650 Set the MAC address of a Link Bonding device::
1652 testpmd> set bonding mac (port id) (mac)
1654 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
1656 testpmd> set bonding mac 10 00:00:00:00:00:01
1658 set bonding xmit_balance_policy
1659 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1661 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
1663 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
1665 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
1667 testpmd> set bonding xmit_balance_policy 10 l34
1670 set bonding mon_period
1671 ~~~~~~~~~~~~~~~~~~~~~~
1673 Set the link status monitoring polling period in milliseconds for a bonding device.
1675 This adds support for PMD slave devices which do not support link status interrupts.
1676 When the mon_period is set to a value greater than 0 then all PMD's which do not support
1677 link status ISR will be queried every polling interval to check if their link status has changed::
1679 testpmd> set bonding mon_period (port_id) (value)
1681 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
1683 testpmd> set bonding mon_period 5 150
1689 Show the current configuration of a Link Bonding device::
1691 testpmd> show bonding config (port id)
1694 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
1695 in balance mode with a transmission policy of layer 2+3::
1697 testpmd> show bonding config 9
1699 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
1701 Active Slaves (3): [1 3 4]
1708 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
1709 This is mainly useful for debugging purposes.
1710 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
1711 and fields that can be accessed.
1716 Display the value of a port register::
1718 testpmd> read reg (port_id) (address)
1720 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
1722 testpmd> read reg 0 0xEE00
1723 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
1728 Display a port register bit field::
1730 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
1732 For example, reading the lowest two bits from the register in the example above::
1734 testpmd> read regfield 0 0xEE00 0 1
1735 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
1740 Display a single port register bit::
1742 testpmd> read regbit (port_id) (address) (bit_x)
1744 For example, reading the lowest bit from the register in the example above::
1746 testpmd> read regbit 0 0xEE00 0
1747 port 0 PCI register at offset 0xEE00: bit 0=1
1752 Set the value of a port register::
1754 testpmd> write reg (port_id) (address) (value)
1756 For example, to clear a register::
1758 testpmd> write reg 0 0xEE00 0x0
1759 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
1764 Set bit field of a port register::
1766 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
1768 For example, writing to the register cleared in the example above::
1770 testpmd> write regfield 0 0xEE00 0 1 2
1771 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
1776 Set single bit value of a port register::
1778 testpmd> write regbit (port_id) (address) (bit_x) (value)
1780 For example, to set the high bit in the register from the example above::
1782 testpmd> write regbit 0 0xEE00 31 1
1783 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
1789 This section details the available filter functions that are available.
1791 Note these functions interface the deprecated legacy filtering framework,
1792 superseded by *rte_flow*. See `Flow rules management`_.
1795 ~~~~~~~~~~~~~~~~~~~~
1797 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
1799 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
1800 ethertype (ether_type) (drop|fwd) queue (queue_id)
1802 The available information parameters are:
1804 * ``port_id``: The port which the Ethertype filter assigned on.
1806 * ``mac_addr``: Compare destination mac address.
1808 * ``mac_ignr``: Ignore destination mac address match.
1810 * ``mac_address``: Destination mac address to match.
1812 * ``ether_type``: The EtherType value want to match,
1813 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
1815 * ``queue_id``: The receive queue associated with this EtherType filter.
1816 It is meaningless when deleting or dropping.
1818 Example, to add/remove an ethertype filter rule::
1820 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
1821 ethertype 0x0806 fwd queue 3
1823 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
1824 ethertype 0x0806 fwd queue 3
1829 Add or delete a 2-tuple filter,
1830 which identifies packets by specific protocol and destination TCP/UDP port
1831 and forwards packets into one of the receive queues::
1833 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
1834 protocol (protocol_value) mask (mask_value) \
1835 tcp_flags (tcp_flags_value) priority (prio_value) \
1838 The available information parameters are:
1840 * ``port_id``: The port which the 2-tuple filter assigned on.
1842 * ``dst_port_value``: Destination port in L4.
1844 * ``protocol_value``: IP L4 protocol.
1846 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
1848 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
1850 * ``prio_value``: Priority of this filter.
1852 * ``queue_id``: The receive queue associated with this 2-tuple filter.
1854 Example, to add/remove an 2tuple filter rule::
1856 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
1857 tcp_flags 0x02 priority 3 queue 3
1859 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
1860 tcp_flags 0x02 priority 3 queue 3
1865 Add or delete a 5-tuple filter,
1866 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
1867 and routes packets into one of the receive queues::
1869 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
1870 (src_address) dst_port (dst_port_value) \
1871 src_port (src_port_value) protocol (protocol_value) \
1872 mask (mask_value) tcp_flags (tcp_flags_value) \
1873 priority (prio_value) queue (queue_id)
1875 The available information parameters are:
1877 * ``port_id``: The port which the 5-tuple filter assigned on.
1879 * ``dst_address``: Destination IP address.
1881 * ``src_address``: Source IP address.
1883 * ``dst_port_value``: TCP/UDP destination port.
1885 * ``src_port_value``: TCP/UDP source port.
1887 * ``protocol_value``: L4 protocol.
1889 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
1891 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
1893 * ``prio_value``: The priority of this filter.
1895 * ``queue_id``: The receive queue associated with this 5-tuple filter.
1897 Example, to add/remove an 5tuple filter rule::
1899 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1900 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1901 flags 0x0 priority 3 queue 3
1903 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1904 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1905 flags 0x0 priority 3 queue 3
1910 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
1912 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
1914 The available information parameters are:
1916 * ``port_id``: The port which the SYN filter assigned on.
1918 * ``high``: This SYN filter has higher priority than other filters.
1920 * ``low``: This SYN filter has lower priority than other filters.
1922 * ``queue_id``: The receive queue associated with this SYN filter
1926 testpmd> syn_filter 0 add priority high queue 3
1931 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
1932 and routed into one of the receive queues::
1934 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
1935 mask (mask_value) priority (prio_value) queue (queue_id)
1937 The available information parameters are:
1939 * ``port_id``: The port which the Flex filter is assigned on.
1941 * ``len_value``: Filter length in bytes, no greater than 128.
1943 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
1945 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
1947 * ``prio_value``: The priority of this filter.
1949 * ``queue_id``: The receive queue associated with this Flex filter.
1953 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
1954 mask 000C priority 3 queue 3
1956 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
1957 mask 000C priority 3 queue 3
1960 .. _testpmd_flow_director:
1962 flow_director_filter
1963 ~~~~~~~~~~~~~~~~~~~~
1965 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
1967 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
1968 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
1970 * Perfect match filters.
1971 The hardware checks a match between the masked fields of the received packets and the programmed filters.
1972 The masked fields are for IP flow.
1974 * Signature filters.
1975 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
1977 * Perfect-mac-vlan match filters.
1978 The hardware checks a match between the masked fields of the received packets and the programmed filters.
1979 The masked fields are for MAC VLAN flow.
1981 * Perfect-tunnel match filters.
1982 The hardware checks a match between the masked fields of the received packets and the programmed filters.
1983 The masked fields are for tunnel flow.
1985 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
1986 per flow type and the flexible payload.
1988 The Flow Director can also mask out parts of all of these fields so that filters
1989 are only applied to certain fields or parts of the fields.
1991 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
1993 # Commands to add flow director filters of different flow types::
1995 flow_director_filter (port_id) mode IP (add|del|update) \
1996 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
1997 src (src_ip_address) dst (dst_ip_address) \
1998 tos (tos_value) proto (proto_value) ttl (ttl_value) \
1999 vlan (vlan_value) flexbytes (flexbytes_value) \
2000 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
2003 flow_director_filter (port_id) mode IP (add|del|update) \
2004 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
2005 src (src_ip_address) (src_port) \
2006 dst (dst_ip_address) (dst_port) \
2007 tos (tos_value) ttl (ttl_value) \
2008 vlan (vlan_value) flexbytes (flexbytes_value) \
2009 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
2012 flow_director_filter (port_id) mode IP (add|del|update) \
2013 flow (ipv4-sctp|ipv6-sctp) \
2014 src (src_ip_address) (src_port) \
2015 dst (dst_ip_address) (dst_port) \
2016 tos (tos_value) ttl (ttl_value) \
2017 tag (verification_tag) vlan (vlan_value) \
2018 flexbytes (flexbytes_value) (drop|fwd) \
2019 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
2021 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
2022 ether (ethertype) flexbytes (flexbytes_value) \
2023 (drop|fwd) pf|vf(vf_id) queue (queue_id)
2026 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
2027 mac (mac_address) vlan (vlan_value) \
2028 flexbytes (flexbytes_value) (drop|fwd) \
2029 queue (queue_id) fd_id (fd_id_value)
2031 flow_director_filter (port_id) mode Tunnel (add|del|update) \
2032 mac (mac_address) vlan (vlan_value) \
2033 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
2034 flexbytes (flexbytes_value) (drop|fwd) \
2035 queue (queue_id) fd_id (fd_id_value)
2037 For example, to add an ipv4-udp flow type filter::
2039 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
2040 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
2041 fwd pf queue 1 fd_id 1
2043 For example, add an ipv4-other flow type filter::
2045 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
2046 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
2047 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
2052 Flush all flow director filters on a device::
2054 testpmd> flush_flow_director (port_id)
2056 Example, to flush all flow director filter on port 0::
2058 testpmd> flush_flow_director 0
2063 Set flow director's input masks::
2065 flow_director_mask (port_id) mode IP vlan (vlan_value) \
2066 src_mask (ipv4_src) (ipv6_src) (src_port) \
2067 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2069 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2071 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2072 mac (mac_value) tunnel-type (tunnel_type_value) \
2073 tunnel-id (tunnel_id_value)
2075 Example, to set flow director mask on port 0::
2077 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2078 src_mask 255.255.255.255 \
2079 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2080 dst_mask 255.255.255.255 \
2081 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2083 flow_director_flex_mask
2084 ~~~~~~~~~~~~~~~~~~~~~~~
2086 set masks of flow director's flexible payload based on certain flow type::
2088 testpmd> flow_director_flex_mask (port_id) \
2089 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2090 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2091 l2_payload|all) (mask)
2093 Example, to set flow director's flex mask for all flow type on port 0::
2095 testpmd> flow_director_flex_mask 0 flow all \
2096 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2099 flow_director_flex_payload
2100 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2102 Configure flexible payload selection::
2104 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2106 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2108 testpmd> flow_director_flex_payload 0 l4 \
2109 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2111 get_sym_hash_ena_per_port
2112 ~~~~~~~~~~~~~~~~~~~~~~~~~
2114 Get symmetric hash enable configuration per port::
2116 get_sym_hash_ena_per_port (port_id)
2118 For example, to get symmetric hash enable configuration of port 1::
2120 testpmd> get_sym_hash_ena_per_port 1
2122 set_sym_hash_ena_per_port
2123 ~~~~~~~~~~~~~~~~~~~~~~~~~
2125 Set symmetric hash enable configuration per port to enable or disable::
2127 set_sym_hash_ena_per_port (port_id) (enable|disable)
2129 For example, to set symmetric hash enable configuration of port 1 to enable::
2131 testpmd> set_sym_hash_ena_per_port 1 enable
2133 get_hash_global_config
2134 ~~~~~~~~~~~~~~~~~~~~~~
2136 Get the global configurations of hash filters::
2138 get_hash_global_config (port_id)
2140 For example, to get the global configurations of hash filters of port 1::
2142 testpmd> get_hash_global_config 1
2144 set_hash_global_config
2145 ~~~~~~~~~~~~~~~~~~~~~~
2147 Set the global configurations of hash filters::
2149 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2150 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2151 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload) \
2154 For example, to enable simple_xor for flow type of ipv6 on port 2::
2156 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2161 Set the input set for hash::
2163 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2164 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2165 l2_payload) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2166 ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2167 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2168 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2169 fld-8th|none) (select|add)
2171 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2173 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2178 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2179 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2181 Set the input set for flow director::
2183 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2184 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2185 l2_payload) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2186 ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2187 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2188 sctp-dst-port|sctp-veri-tag|none) (select|add)
2190 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2192 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2197 Set different GRE key length for input set::
2199 global_config (port_id) gre-key-len (number in bytes)
2201 For example to set GRE key length for input set to 4 bytes on port 0::
2203 testpmd> global_config 0 gre-key-len 4
2206 .. _testpmd_rte_flow:
2208 Flow rules management
2209 ---------------------
2211 Control of the generic flow API (*rte_flow*) is fully exposed through the
2212 ``flow`` command (validation, creation, destruction and queries).
2214 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2215 features simultaneously may cause undefined side-effects and is therefore
2221 Because the ``flow`` command uses dynamic tokens to handle the large number
2222 of possible flow rules combinations, its behavior differs slightly from
2223 other commands, in particular:
2225 - Pressing *?* or the *<tab>* key displays contextual help for the current
2226 token, not that of the entire command.
2228 - Optional and repeated parameters are supported (provided they are listed
2229 in the contextual help).
2231 The first parameter stands for the operation mode. Possible operations and
2232 their general syntax are described below. They are covered in detail in the
2235 - Check whether a flow rule can be created::
2237 flow validate {port_id}
2238 [group {group_id}] [priority {level}] [ingress] [egress]
2239 pattern {item} [/ {item} [...]] / end
2240 actions {action} [/ {action} [...]] / end
2242 - Create a flow rule::
2244 flow create {port_id}
2245 [group {group_id}] [priority {level}] [ingress] [egress]
2246 pattern {item} [/ {item} [...]] / end
2247 actions {action} [/ {action} [...]] / end
2249 - Destroy specific flow rules::
2251 flow destroy {port_id} rule {rule_id} [...]
2253 - Destroy all flow rules::
2255 flow flush {port_id}
2257 - Query an existing flow rule::
2259 flow query {port_id} {rule_id} {action}
2261 - List existing flow rules sorted by priority, filtered by group
2264 flow list {port_id} [group {group_id}] [...]
2266 Validating flow rules
2267 ~~~~~~~~~~~~~~~~~~~~~
2269 ``flow validate`` reports whether a flow rule would be accepted by the
2270 underlying device in its current state but stops short of creating it. It is
2271 bound to ``rte_flow_validate()``::
2273 flow validate {port_id}
2274 [group {group_id}] [priority {level}] [ingress] [egress]
2275 pattern {item} [/ {item} [...]] / end
2276 actions {action} [/ {action} [...]] / end
2278 If successful, it will show::
2282 Otherwise it will show an error message of the form::
2284 Caught error type [...] ([...]): [...]
2286 This command uses the same parameters as ``flow create``, their format is
2287 described in `Creating flow rules`_.
2289 Check whether redirecting any Ethernet packet received on port 0 to RX queue
2290 index 6 is supported::
2292 testpmd> flow validate 0 ingress pattern eth / end
2293 actions queue index 6 / end
2297 Port 0 does not support TCPv6 rules::
2299 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
2301 Caught error type 9 (specific pattern item): Invalid argument
2307 ``flow create`` validates and creates the specified flow rule. It is bound
2308 to ``rte_flow_create()``::
2310 flow create {port_id}
2311 [group {group_id}] [priority {level}] [ingress] [egress]
2312 pattern {item} [/ {item} [...]] / end
2313 actions {action} [/ {action} [...]] / end
2315 If successful, it will return a flow rule ID usable with other commands::
2317 Flow rule #[...] created
2319 Otherwise it will show an error message of the form::
2321 Caught error type [...] ([...]): [...]
2323 Parameters describe in the following order:
2325 - Attributes (*group*, *priority*, *ingress*, *egress* tokens).
2326 - A matching pattern, starting with the *pattern* token and terminated by an
2328 - Actions, starting with the *actions* token and terminated by an *end*
2331 These translate directly to *rte_flow* objects provided as-is to the
2332 underlying functions.
2334 The shortest valid definition only comprises mandatory tokens::
2336 testpmd> flow create 0 pattern end actions end
2338 Note that PMDs may refuse rules that essentially do nothing such as this
2341 **All unspecified object values are automatically initialized to 0.**
2346 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
2347 specified before the ``pattern`` token.
2349 - ``group {group id}``: priority group.
2350 - ``priority {level}``: priority level within group.
2351 - ``ingress``: rule applies to ingress traffic.
2352 - ``egress``: rule applies to egress traffic.
2354 Each instance of an attribute specified several times overrides the previous
2355 value as shown below (group 4 is used)::
2357 testpmd> flow create 0 group 42 group 24 group 4 [...]
2359 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
2361 While not specifying a direction is an error, some rules may allow both
2364 Most rules affect RX therefore contain the ``ingress`` token::
2366 testpmd> flow create 0 ingress pattern [...]
2371 A matching pattern starts after the ``pattern`` token. It is made of pattern
2372 items and is terminated by a mandatory ``end`` item.
2374 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
2375 rte_flow_item_type``).
2377 The ``/`` token is used as a separator between pattern items as shown
2380 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
2382 Note that protocol items like these must be stacked from lowest to highest
2383 layer to make sense. For instance, the following rule is either invalid or
2384 unlikely to match any packet::
2386 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
2388 More information on these restrictions can be found in the *rte_flow*
2391 Several items support additional specification structures, for example
2392 ``ipv4`` allows specifying source and destination addresses as follows::
2394 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
2395 dst is 10.2.0.0 / end [...]
2397 This rule matches all IPv4 traffic with the specified properties.
2399 In this example, ``src`` and ``dst`` are field names of the underlying
2400 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
2401 in a similar fashion.
2403 The ``is`` token means that the subsequent value must be matched exactly,
2404 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
2405 accordingly. Possible assignment tokens are:
2407 - ``is``: match value perfectly (with full bit-mask).
2408 - ``spec``: match value according to configured bit-mask.
2409 - ``last``: specify upper bound to establish a range.
2410 - ``mask``: specify bit-mask with relevant bits set to one.
2411 - ``prefix``: generate bit-mask from a prefix length.
2413 These yield identical results::
2415 ipv4 src is 10.1.1.1
2419 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
2423 ipv4 src spec 10.1.1.1 src prefix 32
2427 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
2431 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
2433 Inclusive ranges can be defined with ``last``::
2435 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
2437 Note that ``mask`` affects both ``spec`` and ``last``::
2439 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
2440 # matches 10.1.0.0 to 10.2.255.255
2442 Properties can be modified multiple times::
2444 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
2448 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
2453 This section lists supported pattern items and their attributes, if any.
2455 - ``end``: end list of pattern items.
2457 - ``void``: no-op pattern item.
2459 - ``invert``: perform actions when pattern does not match.
2461 - ``any``: match any protocol for the current layer.
2463 - ``num {unsigned}``: number of layers covered.
2465 - ``pf``: match packets addressed to the physical function.
2467 - ``vf``: match packets addressed to a virtual function ID.
2469 - ``id {unsigned}``: destination VF ID.
2471 - ``port``: device-specific physical port index to use.
2473 - ``index {unsigned}``: physical port index.
2475 - ``raw``: match an arbitrary byte string.
2477 - ``relative {boolean}``: look for pattern after the previous item.
2478 - ``search {boolean}``: search pattern from offset (see also limit).
2479 - ``offset {integer}``: absolute or relative offset for pattern.
2480 - ``limit {unsigned}``: search area limit for start of pattern.
2481 - ``pattern {string}``: byte string to look for.
2483 - ``eth``: match Ethernet header.
2485 - ``dst {MAC-48}``: destination MAC.
2486 - ``src {MAC-48}``: source MAC.
2487 - ``type {unsigned}``: EtherType.
2489 - ``vlan``: match 802.1Q/ad VLAN tag.
2491 - ``tpid {unsigned}``: tag protocol identifier.
2492 - ``tci {unsigned}``: tag control information.
2493 - ``pcp {unsigned}``: priority code point.
2494 - ``dei {unsigned}``: drop eligible indicator.
2495 - ``vid {unsigned}``: VLAN identifier.
2497 - ``ipv4``: match IPv4 header.
2499 - ``tos {unsigned}``: type of service.
2500 - ``ttl {unsigned}``: time to live.
2501 - ``proto {unsigned}``: next protocol ID.
2502 - ``src {ipv4 address}``: source address.
2503 - ``dst {ipv4 address}``: destination address.
2505 - ``ipv6``: match IPv6 header.
2507 - ``tc {unsigned}``: traffic class.
2508 - ``flow {unsigned}``: flow label.
2509 - ``proto {unsigned}``: protocol (next header).
2510 - ``hop {unsigned}``: hop limit.
2511 - ``src {ipv6 address}``: source address.
2512 - ``dst {ipv6 address}``: destination address.
2514 - ``icmp``: match ICMP header.
2516 - ``type {unsigned}``: ICMP packet type.
2517 - ``code {unsigned}``: ICMP packet code.
2519 - ``udp``: match UDP header.
2521 - ``src {unsigned}``: UDP source port.
2522 - ``dst {unsigned}``: UDP destination port.
2524 - ``tcp``: match TCP header.
2526 - ``src {unsigned}``: TCP source port.
2527 - ``dst {unsigned}``: TCP destination port.
2529 - ``sctp``: match SCTP header.
2531 - ``src {unsigned}``: SCTP source port.
2532 - ``dst {unsigned}``: SCTP destination port.
2533 - ``tag {unsigned}``: validation tag.
2534 - ``cksum {unsigned}``: checksum.
2536 - ``vxlan``: match VXLAN header.
2538 - ``vni {unsigned}``: VXLAN identifier.
2540 - ``mpls``: match MPLS header.
2542 - ``label {unsigned}``: MPLS label.
2544 - ``gre``: match GRE header.
2546 - ``protocol {unsigned}``: protocol type.
2551 A list of actions starts after the ``actions`` token in the same fashion as
2552 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
2553 terminated by a mandatory ``end`` action.
2555 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
2556 rte_flow_action_type``).
2558 Dropping all incoming UDPv4 packets can be expressed as follows::
2560 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2563 Several actions have configurable properties which must be specified when
2564 there is no valid default value. For example, ``queue`` requires a target
2567 This rule redirects incoming UDPv4 traffic to queue index 6::
2569 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2570 actions queue index 6 / end
2572 While this one could be rejected by PMDs (unspecified queue index)::
2574 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2577 As defined by *rte_flow*, the list is not ordered, all actions of a given
2578 rule are performed simultaneously. These are equivalent::
2580 queue index 6 / void / mark id 42 / end
2584 void / mark id 42 / queue index 6 / end
2586 All actions in a list should have different types, otherwise only the last
2587 action of a given type is taken into account::
2589 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
2593 drop / drop / drop / end # drop is performed only once
2597 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
2599 Considering they are performed simultaneously, opposite and overlapping
2600 actions can sometimes be combined when the end result is unambiguous::
2602 drop / queue index 6 / end # drop has no effect
2606 drop / dup index 6 / end # same as above
2610 queue index 6 / rss queues 6 7 8 / end # queue has no effect
2614 drop / passthru / end # drop has no effect
2616 Note that PMDs may still refuse such combinations.
2621 This section lists supported actions and their attributes, if any.
2623 - ``end``: end list of actions.
2625 - ``void``: no-op action.
2627 - ``passthru``: let subsequent rule process matched packets.
2629 - ``mark``: attach 32 bit value to packets.
2631 - ``id {unsigned}``: 32 bit value to return with packets.
2633 - ``flag``: flag packets.
2635 - ``queue``: assign packets to a given queue index.
2637 - ``index {unsigned}``: queue index to use.
2639 - ``drop``: drop packets (note: passthru has priority).
2641 - ``count``: enable counters for this rule.
2643 - ``dup``: duplicate packets to a given queue index.
2645 - ``index {unsigned}``: queue index to duplicate packets to.
2647 - ``rss``: spread packets among several queues.
2649 - ``queues [{unsigned} [...]] end``: queue indices to use.
2651 - ``pf``: redirect packets to physical device function.
2653 - ``vf``: redirect packets to virtual device function.
2655 - ``original {boolean}``: use original VF ID if possible.
2656 - ``id {unsigned}``: VF ID to redirect packets to.
2658 Destroying flow rules
2659 ~~~~~~~~~~~~~~~~~~~~~
2661 ``flow destroy`` destroys one or more rules from their rule ID (as returned
2662 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
2663 times as necessary::
2665 flow destroy {port_id} rule {rule_id} [...]
2667 If successful, it will show::
2669 Flow rule #[...] destroyed
2671 It does not report anything for rule IDs that do not exist. The usual error
2672 message is shown when a rule cannot be destroyed::
2674 Caught error type [...] ([...]): [...]
2676 ``flow flush`` destroys all rules on a device and does not take extra
2677 arguments. It is bound to ``rte_flow_flush()``::
2679 flow flush {port_id}
2681 Any errors are reported as above.
2683 Creating several rules and destroying them::
2685 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2686 actions queue index 2 / end
2687 Flow rule #0 created
2688 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2689 actions queue index 3 / end
2690 Flow rule #1 created
2691 testpmd> flow destroy 0 rule 0 rule 1
2692 Flow rule #1 destroyed
2693 Flow rule #0 destroyed
2696 The same result can be achieved using ``flow flush``::
2698 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2699 actions queue index 2 / end
2700 Flow rule #0 created
2701 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2702 actions queue index 3 / end
2703 Flow rule #1 created
2704 testpmd> flow flush 0
2707 Non-existent rule IDs are ignored::
2709 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2710 actions queue index 2 / end
2711 Flow rule #0 created
2712 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2713 actions queue index 3 / end
2714 Flow rule #1 created
2715 testpmd> flow destroy 0 rule 42 rule 10 rule 2
2717 testpmd> flow destroy 0 rule 0
2718 Flow rule #0 destroyed
2724 ``flow query`` queries a specific action of a flow rule having that
2725 ability. Such actions collect information that can be reported using this
2726 command. It is bound to ``rte_flow_query()``::
2728 flow query {port_id} {rule_id} {action}
2730 If successful, it will display either the retrieved data for known actions
2731 or the following message::
2733 Cannot display result for action type [...] ([...])
2735 Otherwise, it will complain either that the rule does not exist or that some
2738 Flow rule #[...] not found
2742 Caught error type [...] ([...]): [...]
2744 Currently only the ``count`` action is supported. This action reports the
2745 number of packets that hit the flow rule and the total number of bytes. Its
2746 output has the following format::
2749 hits_set: [...] # whether "hits" contains a valid value
2750 bytes_set: [...] # whether "bytes" contains a valid value
2751 hits: [...] # number of packets
2752 bytes: [...] # number of bytes
2754 Querying counters for TCPv6 packets redirected to queue 6::
2756 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
2757 actions queue index 6 / count / end
2758 Flow rule #4 created
2759 testpmd> flow query 0 4 count
2770 ``flow list`` lists existing flow rules sorted by priority and optionally
2771 filtered by group identifiers::
2773 flow list {port_id} [group {group_id}] [...]
2775 This command only fails with the following message if the device does not
2780 Output consists of a header line followed by a short description of each
2781 flow rule, one per line. There is no output at all when no flow rules are
2782 configured on the device::
2784 ID Group Prio Attr Rule
2785 [...] [...] [...] [...] [...]
2787 ``Attr`` column flags:
2789 - ``i`` for ``ingress``.
2790 - ``e`` for ``egress``.
2792 Creating several flow rules and listing them::
2794 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2795 actions queue index 6 / end
2796 Flow rule #0 created
2797 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2798 actions queue index 2 / end
2799 Flow rule #1 created
2800 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
2801 actions rss queues 6 7 8 end / end
2802 Flow rule #2 created
2803 testpmd> flow list 0
2804 ID Group Prio Attr Rule
2805 0 0 0 i- ETH IPV4 => QUEUE
2806 1 0 0 i- ETH IPV6 => QUEUE
2807 2 0 5 i- ETH IPV4 UDP => RSS
2810 Rules are sorted by priority (i.e. group ID first, then priority level)::
2812 testpmd> flow list 1
2813 ID Group Prio Attr Rule
2814 0 0 0 i- ETH => COUNT
2815 6 0 500 i- ETH IPV6 TCP => DROP COUNT
2816 5 0 1000 i- ETH IPV6 ICMP => QUEUE
2817 1 24 0 i- ETH IPV4 UDP => QUEUE
2818 4 24 10 i- ETH IPV4 TCP => DROP
2819 3 24 20 i- ETH IPV4 => DROP
2820 2 24 42 i- ETH IPV4 UDP => QUEUE
2821 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
2824 Output can be limited to specific groups::
2826 testpmd> flow list 1 group 0 group 63
2827 ID Group Prio Attr Rule
2828 0 0 0 i- ETH => COUNT
2829 6 0 500 i- ETH IPV6 TCP => DROP COUNT
2830 5 0 1000 i- ETH IPV6 ICMP => QUEUE
2831 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
2834 Sample QinQ flow rules
2835 ~~~~~~~~~~~~~~~~~~~~~~
2837 Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM.
2841 testpmd> flow validate 0 ingress pattern eth / vlan tpid is 0x8100 tci is 4 /
2842 vlan tpid is 0x8100 tci is 5 / end actions vf id 1 / queue index 0 / end
2843 Flow rule #0 validated
2845 testpmd> flow create 0 ingress pattern eth / vlan tpid is 0x8100 tci is 4 /
2846 vlan tpid is 0x8100 tci is 5 / end actions vf id 1 / queue index 0 / end
2847 Flow rule #0 created
2849 testpmd> flow list 0
2850 ID Group Prio Attr Rule
2851 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
2853 Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host.
2857 testpmd> flow validate 0 ingress pattern eth / vlan tpid is 0x8100 tci is 6 /
2858 vlan tpid is 0x8100 tci is 7 / end actions pf / queue index 0 / end
2859 Flow rule #1 validated
2861 testpmd> flow create 0 ingress pattern eth / vlan tpid is 0x8100 tci is 6 /
2862 vlan tpid is 0x8100 tci is 7 / end actions pf / queue index 1 / end
2863 Flow rule #1 created
2865 testpmd> flow list 0
2866 ID Group Prio Attr Rule
2867 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
2868 1 0 0 i- ETH VLAN VLAN=>PF QUEUE
2870 After creating QinQ rule(s) the following command should be issued to enable QinQ::
2872 testpmd> vlan set qinq on 0