2 Copyright(c) 2010-2016 Intel Corporation. All rights reserved.
5 Redistribution and use in source and binary forms, with or without
6 modification, are permitted provided that the following conditions
9 * Redistributions of source code must retain the above copyright
10 notice, this list of conditions and the following disclaimer.
11 * Redistributions in binary form must reproduce the above copyright
12 notice, this list of conditions and the following disclaimer in
13 the documentation and/or other materials provided with the
15 * Neither the name of Intel Corporation nor the names of its
16 contributors may be used to endorse or promote products derived
17 from this software without specific prior written permission.
19 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
20 "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
21 LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
22 A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
23 OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
24 SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
25 LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
26 DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
27 THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
28 (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
29 OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
33 Testpmd Runtime Functions
34 =========================
36 Where the testpmd application is started in interactive mode, (``-i|--interactive``),
37 it displays a prompt that can be used to start and stop forwarding,
38 configure the application, display statistics (including the extended NIC
39 statistics aka xstats) , set the Flow Director and other tasks::
43 The testpmd prompt has some, limited, readline support.
44 Common bash command-line functions such as ``Ctrl+a`` and ``Ctrl+e`` to go to the start and end of the prompt line are supported
45 as well as access to the command history via the up-arrow.
47 There is also support for tab completion.
48 If you type a partial command and hit ``<TAB>`` you get a list of the available completions:
50 .. code-block:: console
52 testpmd> show port <TAB>
54 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
55 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
56 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
57 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
63 Some examples in this document are too long to fit on one line are are shown wrapped at `"\\"` for display purposes::
65 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
66 (pause_time) (send_xon) (port_id)
68 In the real ``testpmd>`` prompt these commands should be on a single line.
73 The testpmd has on-line help for the functions that are available at runtime.
74 These are divided into sections and can be accessed using help, help section or help all:
76 .. code-block:: console
80 help control : Start and stop forwarding.
81 help display : Displaying port, stats and config information.
82 help config : Configuration information.
83 help ports : Configuring ports.
84 help registers : Reading and setting port registers.
85 help filters : Filters configuration help.
86 help all : All of the above sections.
95 Start packet forwarding with current configuration::
102 Start packet forwarding with current configuration after sending specified number of bursts of packets::
104 testpmd> start tx_first (""|burst_num)
106 The default burst number is 1 when ``burst_num`` not presented.
111 Stop packet forwarding, and display accumulated statistics::
126 The functions in the following sections are used to display information about the
127 testpmd configuration or the NIC status.
132 Display information for a given port or all ports::
134 testpmd> show port (info|stats|xstats|fdir|stat_qmap|dcb_tc|cap) (port_id|all)
136 The available information categories are:
138 * ``info``: General port information such as MAC address.
140 * ``stats``: RX/TX statistics.
142 * ``xstats``: RX/TX extended NIC statistics.
144 * ``fdir``: Flow Director information and statistics.
146 * ``stat_qmap``: Queue statistics mapping.
148 * ``dcb_tc``: DCB information such as TC mapping.
150 * ``cap``: Supported offload capabilities.
154 .. code-block:: console
156 testpmd> show port info 0
158 ********************* Infos for port 0 *********************
160 MAC address: XX:XX:XX:XX:XX:XX
162 memory allocation on the socket: 0
164 Link speed: 40000 Mbps
165 Link duplex: full-duplex
166 Promiscuous mode: enabled
167 Allmulticast mode: disabled
168 Maximum number of MAC addresses: 64
169 Maximum number of MAC addresses of hash filtering: 0
174 Redirection table size: 512
175 Supported flow types:
195 Display the rss redirection table entry indicated by masks on port X::
197 testpmd> show port (port_id) rss reta (size) (mask0, mask1...)
199 size is used to indicate the hardware supported reta size
204 Display the RSS hash functions and RSS hash key of a port::
206 testpmd> show port (port_id) rss-hash ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2-payload|ipv6-ex|ipv6-tcp-ex|ipv6-udp-ex [key]
211 Clear the port statistics for a given port or for all ports::
213 testpmd> clear port (info|stats|xstats|fdir|stat_qmap) (port_id|all)
217 testpmd> clear port stats all
222 Display information for a given port's RX/TX queue::
224 testpmd> show (rxq|txq) info (port_id) (queue_id)
229 Displays the configuration of the application.
230 The configuration comes from the command-line, the runtime or the application defaults::
232 testpmd> show config (rxtx|cores|fwd|txpkts)
234 The available information categories are:
236 * ``rxtx``: RX/TX configuration items.
238 * ``cores``: List of forwarding cores.
240 * ``fwd``: Packet forwarding configuration.
242 * ``txpkts``: Packets to TX configuration.
246 .. code-block:: console
248 testpmd> show config rxtx
250 io packet forwarding - CRC stripping disabled - packets/burst=16
251 nb forwarding cores=2 - nb forwarding ports=1
252 RX queues=1 - RX desc=128 - RX free threshold=0
253 RX threshold registers: pthresh=8 hthresh=8 wthresh=4
254 TX queues=1 - TX desc=512 - TX free threshold=0
255 TX threshold registers: pthresh=36 hthresh=0 wthresh=0
256 TX RS bit threshold=0 - TXQ flags=0x0
261 Set the packet forwarding mode::
263 testpmd> set fwd (io|mac|macswap|flowgen| \
264 rxonly|txonly|csum|icmpecho) (""|retry)
266 ``retry`` can be specified for forwarding engines except ``rx_only``.
268 The available information categories are:
270 * ``io``: Forwards packets "as-is" in I/O mode.
271 This is the fastest possible forwarding operation as it does not access packets data.
272 This is the default mode.
274 * ``mac``: Changes the source and the destination Ethernet addresses of packets before forwarding them.
275 Default application behaviour is to set source Ethernet address to that of the transmitting interface, and destination
276 address to a dummy value (set during init). The user may specify a target destination Ethernet address via the 'eth-peer' or
277 'eth-peer-configfile' command-line options. It is not currently possible to specify a specific source Ethernet address.
279 * ``macswap``: MAC swap forwarding mode.
280 Swaps the source and the destination Ethernet addresses of packets before forwarding them.
282 * ``flowgen``: Multi-flow generation mode.
283 Originates a number of flows (with varying destination IP addresses), and terminate receive traffic.
285 * ``rxonly``: Receives packets but doesn't transmit them.
287 * ``txonly``: Generates and transmits packets without receiving any.
289 * ``csum``: Changes the checksum field with hardware or software methods depending on the offload flags on the packet.
291 * ``icmpecho``: Receives a burst of packets, lookup for IMCP echo requests and, if any, send back ICMP echo replies.
293 * ``ieee1588``: Demonstrate L2 IEEE1588 V2 PTP timestamping for RX and TX. Requires ``CONFIG_RTE_LIBRTE_IEEE1588=y``.
295 Note: TX timestamping is only available in the "Full Featured" TX path. To force ``testpmd`` into this mode set ``--txqflags=0``.
299 testpmd> set fwd rxonly
301 Set rxonly packet forwarding mode
307 Display an RX descriptor for a port RX queue::
309 testpmd> read rxd (port_id) (queue_id) (rxd_id)
313 testpmd> read rxd 0 0 4
314 0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180
319 Display a TX descriptor for a port TX queue::
321 testpmd> read txd (port_id) (queue_id) (txd_id)
325 testpmd> read txd 0 0 4
326 0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C
329 Configuration Functions
330 -----------------------
332 The testpmd application can be configured from the runtime as well as from the command-line.
334 This section details the available configuration functions that are available.
338 Configuration changes only become active when forwarding is started/restarted.
343 Reset forwarding to the default configuration::
350 Set the debug verbosity level::
352 testpmd> set verbose (level)
354 Currently the only available levels are 0 (silent except for error) and 1 (fully verbose).
359 Set the number of ports used by the application:
363 This is equivalent to the ``--nb-ports`` command-line option.
368 Set the number of cores used by the application::
370 testpmd> set nbcore (num)
372 This is equivalent to the ``--nb-cores`` command-line option.
376 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port.
381 Set the forwarding cores hexadecimal mask::
383 testpmd> set coremask (mask)
385 This is equivalent to the ``--coremask`` command-line option.
389 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding.
394 Set the forwarding ports hexadecimal mask::
396 testpmd> set portmask (mask)
398 This is equivalent to the ``--portmask`` command-line option.
403 Set number of packets per burst::
405 testpmd> set burst (num)
407 This is equivalent to the ``--burst command-line`` option.
409 When retry is enabled, the transmit delay time and number of retries can also be set::
411 testpmd> set burst tx delay (microseconds) retry (num)
416 Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode::
418 testpmd> set txpkts (x[,y]*)
420 Where x[,y]* represents a CSV list of values, without white space.
425 Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes::
427 testpmd> set txsplit (off|on|rand)
431 * ``off`` disable packet copy & split for CSUM mode.
433 * ``on`` split outgoing packet into multiple segments. Size of each segment
434 and number of segments per packet is determined by ``set txpkts`` command
437 * ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments.
442 Set the list of forwarding cores::
444 testpmd> set corelist (x[,y]*)
446 For example, to change the forwarding cores:
448 .. code-block:: console
450 testpmd> set corelist 3,1
451 testpmd> show config fwd
453 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled
454 Logical Core 3 (socket 0) forwards packets on 1 streams:
455 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01
456 Logical Core 1 (socket 0) forwards packets on 1 streams:
457 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
461 The cores are used in the same order as specified on the command line.
466 Set the list of forwarding ports::
468 testpmd> set portlist (x[,y]*)
470 For example, to change the port forwarding:
472 .. code-block:: console
474 testpmd> set portlist 0,2,1,3
475 testpmd> show config fwd
477 io packet forwarding - ports=4 - cores=1 - streams=4
478 Logical Core 3 (socket 0) forwards packets on 4 streams:
479 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01
480 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
481 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03
482 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02
487 Enable/disable tx loopback::
489 testpmd> set tx loopback (port_id) (on|off)
494 set drop enable bit for all queues::
496 testpmd> set all queues drop (port_id) (on|off)
498 set split drop enable (for VF)
499 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
501 set split drop enable bit for VF from PF::
503 testpmd> set vf split drop (port_id) (vf_id) (on|off)
505 set mac antispoof (for VF)
506 ~~~~~~~~~~~~~~~~~~~~~~~~~~
508 Set mac antispoof for a VF from the PF::
510 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off)
515 Enable/disable MACsec offload::
517 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off)
518 testpmd> set macsec offload (port_id) off
523 Configure MACsec secure connection (SC)::
525 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi)
529 The pi argument is ignored for tx.
530 Check the NIC Datasheet for hardware limits.
535 Configure MACsec secure association (SA)::
537 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key)
541 The IDX value must be 0 or 1.
542 Check the NIC Datasheet for hardware limits.
544 set broadcast mode (for VF)
545 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
547 Set broadcast mode for a VF from the PF::
549 testpmd> set vf broadcast (port_id) (vf_id) (on|off)
554 Set the VLAN strip on a port::
556 testpmd> vlan set strip (on|off) (port_id)
561 Set the VLAN strip for a queue on a port::
563 testpmd> vlan set stripq (on|off) (port_id,queue_id)
565 vlan set stripq (for VF)
566 ~~~~~~~~~~~~~~~~~~~~~~~~
568 Set VLAN strip for all queues in a pool for a VF from the PF::
570 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off)
572 vlan set insert (for VF)
573 ~~~~~~~~~~~~~~~~~~~~~~~~
575 Set VLAN insert for a VF from the PF::
577 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id)
579 vlan set tag (for VF)
580 ~~~~~~~~~~~~~~~~~~~~~
582 Set VLAN tag for a VF from the PF::
584 testpmd> set vf vlan tag (port_id) (vf_id) (on|off)
586 vlan set antispoof (for VF)
587 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
589 Set VLAN antispoof for a VF from the PF::
591 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off)
596 Set the VLAN filter on a port::
598 testpmd> vlan set filter (on|off) (port_id)
603 Set the VLAN QinQ (extended queue in queue) on for a port::
605 testpmd> vlan set qinq (on|off) (port_id)
610 Set the inner or outer VLAN TPID for packet filtering on a port::
612 testpmd> vlan set (inner|outer) tpid (value) (port_id)
616 TPID value must be a 16-bit number (value <= 65536).
621 Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID::
623 testpmd> rx_vlan add (vlan_id|all) (port_id)
627 VLAN filter must be set on that port. VLAN ID < 4096.
628 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries
629 in VFTA table. This is important if enabling all vlan_ids.
634 Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID::
636 testpmd> rx_vlan rm (vlan_id|all) (port_id)
641 Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID::
643 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask)
648 Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID::
650 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask)
655 Add a tunnel filter on a port::
657 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \
658 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
659 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
661 The available information categories are:
663 * ``vxlan``: Set tunnel type as VXLAN.
665 * ``nvgre``: Set tunnel type as NVGRE.
667 * ``ipingre``: Set tunnel type as IP-in-GRE.
669 * ``imac-ivlan``: Set filter type as Inner MAC and VLAN.
671 * ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID.
673 * ``imac-tenid``: Set filter type as Inner MAC and tenant ID.
675 * ``imac``: Set filter type as Inner MAC.
677 * ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID.
679 * ``oip``: Set filter type as Outer IP.
681 * ``iip``: Set filter type as Inner IP.
685 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \
686 192.168.2.2 0 ipingre oip 1 1
688 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP.
693 Remove a tunnel filter on a port::
695 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \
696 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
697 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
702 Add an UDP port for VXLAN packet filter on a port::
704 testpmd> rx_vxlan_port add (udp_port) (port_id)
709 Remove an UDP port for VXLAN packet filter on a port::
711 testpmd> rx_vxlan_port rm (udp_port) (port_id)
716 Set hardware insertion of VLAN IDs in packets sent on a port::
718 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer]
720 For example, set a single VLAN ID (5) insertion on port 0::
724 Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1::
732 Set port based hardware insertion of VLAN ID in packets sent on a port::
734 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off)
739 Disable hardware insertion of a VLAN header in packets sent on a port::
741 testpmd> tx_vlan reset (port_id)
746 Select hardware or software calculation of the checksum when
747 transmitting a packet using the ``csum`` forwarding engine::
749 testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id)
753 * ``ip|udp|tcp|sctp`` always relate to the inner layer.
755 * ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized
756 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are
757 supported). See also the ``csum parse-tunnel`` command.
761 Check the NIC Datasheet for hardware limits.
766 Define how tunneled packets should be handled by the csum forward
769 testpmd> csum parse-tunnel (on|off) (tx_port_id)
771 If enabled, the csum forward engine will try to recognize supported
772 tunnel headers (vxlan, gre, ipip).
774 If disabled, treat tunnel packets as non-tunneled packets (a inner
775 header is handled as a packet payload).
779 The port argument is the TX port like in the ``csum set`` command.
783 Consider a packet in packet like the following::
785 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in
787 * If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
788 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the
789 ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``).
791 * If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
792 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``.
797 Display tx checksum offload configuration::
799 testpmd> csum show (port_id)
804 Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine::
806 testpmd> tso set (segsize) (port_id)
810 Check the NIC datasheet for hardware limits.
815 Display the status of TCP Segmentation Offload::
817 testpmd> tso show (port_id)
822 Add an alternative MAC address to a port::
824 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX)
829 Remove a MAC address from a port::
831 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX)
833 mac_addr add (for VF)
834 ~~~~~~~~~~~~~~~~~~~~~
836 Add an alternative MAC address for a VF to a port::
838 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX)
840 mac_addr set (for VF)
841 ~~~~~~~~~~~~~~~~~~~~~
843 Set the MAC address for a VF from the PF::
845 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX)
850 Set the unicast hash filter(s) on/off for a port::
852 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
857 Set the promiscuous mode on for a port or for all ports.
858 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
860 testpmd> set promisc (port_id|all) (on|off)
865 Set the allmulti mode for a port or for all ports::
867 testpmd> set allmulti (port_id|all) (on|off)
869 Same as the ifconfig (8) option. Controls how multicast packets are handled.
874 Set the unicast promiscuous mode for a VF from PF.
875 It's supported by Intel i40e NICs now.
876 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
878 testpmd> set vf promisc (port_id) (vf_id) (on|off)
880 set allmulticast (for VF)
881 ~~~~~~~~~~~~~~~~~~~~~~~~~
883 Set the multicast promiscuous mode for a VF from PF.
884 It's supported by Intel i40e NICs now.
885 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
887 testpmd> set vf allmulti (port_id) (vf_id) (on|off)
892 Set the link flow control parameter on a port::
894 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
895 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
896 autoneg (on|off) (port_id)
900 * ``high_water`` (integer): High threshold value to trigger XOFF.
902 * ``low_water`` (integer): Low threshold value to trigger XON.
904 * ``pause_time`` (integer): Pause quota in the Pause frame.
906 * ``send_xon`` (0/1): Send XON frame.
908 * ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
910 * ``autoneg``: Change the auto-negotiation parameter.
915 Set the priority flow control parameter on a port::
917 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
918 (pause_time) (priority) (port_id)
922 * ``high_water`` (integer): High threshold value.
924 * ``low_water`` (integer): Low threshold value.
926 * ``pause_time`` (integer): Pause quota in the Pause frame.
928 * ``priority`` (0-7): VLAN User Priority.
933 Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
935 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
937 For example, to set rx queue 2 on port 0 to mapping 5::
939 testpmd>set stat_qmap rx 0 2 5
941 set port - rx/tx (for VF)
942 ~~~~~~~~~~~~~~~~~~~~~~~~~
944 Set VF receive/transmit from a port::
946 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
948 set port - mac address filter (for VF)
949 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
951 Add/Remove unicast or multicast MAC addr filter for a VF::
953 testpmd> set port (port_id) vf (vf_id) (mac_addr) \
954 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
956 set port - rx mode(for VF)
957 ~~~~~~~~~~~~~~~~~~~~~~~~~~
959 Set the VF receive mode of a port::
961 testpmd> set port (port_id) vf (vf_id) \
962 rxmode (AUPE|ROPE|BAM|MPE) (on|off)
964 The available receive modes are:
966 * ``AUPE``: Accepts untagged VLAN.
968 * ``ROPE``: Accepts unicast hash.
970 * ``BAM``: Accepts broadcast packets.
972 * ``MPE``: Accepts all multicast packets.
974 set port - tx_rate (for Queue)
975 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
977 Set TX rate limitation for a queue on a port::
979 testpmd> set port (port_id) queue (queue_id) rate (rate_value)
981 set port - tx_rate (for VF)
982 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
984 Set TX rate limitation for queues in VF on a port::
986 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
988 set port - mirror rule
989 ~~~~~~~~~~~~~~~~~~~~~~
991 Set pool or vlan type mirror rule for a port::
993 testpmd> set port (port_id) mirror-rule (rule_id) \
994 (pool-mirror-up|pool-mirror-down|vlan-mirror) \
995 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
997 Set link mirror rule for a port::
999 testpmd> set port (port_id) mirror-rule (rule_id) \
1000 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1002 For example to enable mirror traffic with vlan 0,1 to pool 0::
1004 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1006 reset port - mirror rule
1007 ~~~~~~~~~~~~~~~~~~~~~~~~
1009 Reset a mirror rule for a port::
1011 testpmd> reset port (port_id) mirror-rule (rule_id)
1016 Set the flush on RX streams before forwarding.
1017 The default is flush ``on``.
1018 Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1020 testpmd> set flush_rx off
1025 Set the bypass mode for the lowest port on bypass enabled NIC::
1027 testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1032 Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1034 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1035 mode (normal|bypass|isolate) (port_id)
1039 * ``timeout``: Enable bypass after watchdog timeout.
1041 * ``os_on``: Enable bypass when OS/board is powered on.
1043 * ``os_off``: Enable bypass when OS/board is powered off.
1045 * ``power_on``: Enable bypass when power supply is turned on.
1047 * ``power_off``: Enable bypass when power supply is turned off.
1053 Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1055 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1060 Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1062 testpmd> show bypass config (port_id)
1067 Set link up for a port::
1069 testpmd> set link-up port (port id)
1074 Set link down for a port::
1076 testpmd> set link-down port (port id)
1081 Enable E-tag insertion for a VF on a port::
1083 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1085 Disable E-tag insertion for a VF on a port::
1087 testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1089 Enable/disable E-tag stripping on a port::
1091 testpmd> E-tag set stripping (on|off) port (port_id)
1093 Enable/disable E-tag based forwarding on a port::
1095 testpmd> E-tag set forwarding (on|off) port (port_id)
1097 Add an E-tag forwarding filter on a port::
1099 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1101 Delete an E-tag forwarding filter on a port::
1102 testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1108 The following sections show functions for configuring ports.
1112 Port configuration changes only become active when forwarding is started/restarted.
1117 Attach a port specified by pci address or virtual device args::
1119 testpmd> port attach (identifier)
1121 To attach a new pci device, the device should be recognized by kernel first.
1122 Then it should be moved under DPDK management.
1123 Finally the port can be attached to testpmd.
1125 For example, to move a pci device using ixgbe under DPDK management:
1127 .. code-block:: console
1129 # Check the status of the available devices.
1130 ./usertools/dpdk-devbind.py --status
1132 Network devices using DPDK-compatible driver
1133 ============================================
1136 Network devices using kernel driver
1137 ===================================
1138 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1141 # Bind the device to igb_uio.
1142 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1145 # Recheck the status of the devices.
1146 ./usertools/dpdk-devbind.py --status
1147 Network devices using DPDK-compatible driver
1148 ============================================
1149 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1151 To attach a port created by virtual device, above steps are not needed.
1153 For example, to attach a port whose pci address is 0000:0a:00.0.
1155 .. code-block:: console
1157 testpmd> port attach 0000:0a:00.0
1158 Attaching a new port...
1159 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1160 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1161 EAL: PCI memory mapped at 0x7f83bfa00000
1162 EAL: PCI memory mapped at 0x7f83bfa80000
1163 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1164 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1165 Port 0 is attached. Now total ports is 1
1168 For example, to attach a port created by pcap PMD.
1170 .. code-block:: console
1172 testpmd> port attach net_pcap0
1173 Attaching a new port...
1174 PMD: Initializing pmd_pcap for net_pcap0
1175 PMD: Creating pcap-backed ethdev on numa socket 0
1176 Port 0 is attached. Now total ports is 1
1179 In this case, identifier is ``net_pcap0``.
1180 This identifier format is the same as ``--vdev`` format of DPDK applications.
1182 For example, to re-attach a bonded port which has been previously detached,
1183 the mode and slave parameters must be given.
1185 .. code-block:: console
1187 testpmd> port attach net_bond_0,mode=0,slave=1
1188 Attaching a new port...
1189 EAL: Initializing pmd_bond for net_bond_0
1190 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1191 Port 0 is attached. Now total ports is 1
1198 Detach a specific port::
1200 testpmd> port detach (port_id)
1202 Before detaching a port, the port should be stopped and closed.
1204 For example, to detach a pci device port 0.
1206 .. code-block:: console
1208 testpmd> port stop 0
1211 testpmd> port close 0
1215 testpmd> port detach 0
1217 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1218 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1219 EAL: PCI memory unmapped at 0x7f83bfa00000
1220 EAL: PCI memory unmapped at 0x7f83bfa80000
1224 For example, to detach a virtual device port 0.
1226 .. code-block:: console
1228 testpmd> port stop 0
1231 testpmd> port close 0
1235 testpmd> port detach 0
1237 PMD: Closing pcap ethdev on numa socket 0
1238 Port 'net_pcap0' is detached. Now total ports is 0
1241 To remove a pci device completely from the system, first detach the port from testpmd.
1242 Then the device should be moved under kernel management.
1243 Finally the device can be removed using kernel pci hotplug functionality.
1245 For example, to move a pci device under kernel management:
1247 .. code-block:: console
1249 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1251 ./usertools/dpdk-devbind.py --status
1253 Network devices using DPDK-compatible driver
1254 ============================================
1257 Network devices using kernel driver
1258 ===================================
1259 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1261 To remove a port created by a virtual device, above steps are not needed.
1266 Start all ports or a specific port::
1268 testpmd> port start (port_id|all)
1273 Stop all ports or a specific port::
1275 testpmd> port stop (port_id|all)
1280 Close all ports or a specific port::
1282 testpmd> port close (port_id|all)
1284 port start/stop queue
1285 ~~~~~~~~~~~~~~~~~~~~~
1287 Start/stop a rx/tx queue on a specific port::
1289 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1291 Only take effect when port is started.
1296 Set the speed and duplex mode for all ports or a specific port::
1298 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1299 duplex (half|full|auto)
1301 port config - queues/descriptors
1302 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1304 Set number of queues/descriptors for rxq, txq, rxd and txd::
1306 testpmd> port config all (rxq|txq|rxd|txd) (value)
1308 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1310 port config - max-pkt-len
1311 ~~~~~~~~~~~~~~~~~~~~~~~~~
1313 Set the maximum packet length::
1315 testpmd> port config all max-pkt-len (value)
1317 This is equivalent to the ``--max-pkt-len`` command-line option.
1319 port config - CRC Strip
1320 ~~~~~~~~~~~~~~~~~~~~~~~
1322 Set hardware CRC stripping on or off for all ports::
1324 testpmd> port config all crc-strip (on|off)
1326 CRC stripping is off by default.
1328 The ``on`` option is equivalent to the ``--crc-strip`` command-line option.
1330 port config - scatter
1331 ~~~~~~~~~~~~~~~~~~~~~~~
1333 Set RX scatter mode on or off for all ports::
1335 testpmd> port config all scatter (on|off)
1337 RX scatter mode is off by default.
1339 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1341 port config - TX queue flags
1342 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1344 Set a hexadecimal bitmap of TX queue flags for all ports::
1346 testpmd> port config all txqflags value
1348 This command is equivalent to the ``--txqflags`` command-line option.
1350 port config - RX Checksum
1351 ~~~~~~~~~~~~~~~~~~~~~~~~~
1353 Set hardware RX checksum offload to on or off for all ports::
1355 testpmd> port config all rx-cksum (on|off)
1357 Checksum offload is off by default.
1359 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1364 Set hardware VLAN on or off for all ports::
1366 testpmd> port config all hw-vlan (on|off)
1368 Hardware VLAN is on by default.
1370 The ``off`` option is equivalent to the ``--disable-hw-vlan`` command-line option.
1372 port config - VLAN filter
1373 ~~~~~~~~~~~~~~~~~~~~~~~~~
1375 Set hardware VLAN filter on or off for all ports::
1377 testpmd> port config all hw-vlan-filter (on|off)
1379 Hardware VLAN filter is on by default.
1381 The ``off`` option is equivalent to the ``--disable-hw-vlan-filter`` command-line option.
1383 port config - VLAN strip
1384 ~~~~~~~~~~~~~~~~~~~~~~~~
1386 Set hardware VLAN strip on or off for all ports::
1388 testpmd> port config all hw-vlan-strip (on|off)
1390 Hardware VLAN strip is on by default.
1392 The ``off`` option is equivalent to the ``--disable-hw-vlan-strip`` command-line option.
1394 port config - VLAN extend
1395 ~~~~~~~~~~~~~~~~~~~~~~~~~
1397 Set hardware VLAN extend on or off for all ports::
1399 testpmd> port config all hw-vlan-extend (on|off)
1401 Hardware VLAN extend is off by default.
1403 The ``off`` option is equivalent to the ``--disable-hw-vlan-extend`` command-line option.
1405 port config - Drop Packets
1406 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1408 Set packet drop for packets with no descriptors on or off for all ports::
1410 testpmd> port config all drop-en (on|off)
1412 Packet dropping for packets with no descriptors is off by default.
1414 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1419 Set the RSS (Receive Side Scaling) mode on or off::
1421 testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1423 RSS is on by default.
1425 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1427 port config - RSS Reta
1428 ~~~~~~~~~~~~~~~~~~~~~~
1430 Set the RSS (Receive Side Scaling) redirection table::
1432 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1437 Set the DCB mode for an individual port::
1439 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1441 The traffic class should be 4 or 8.
1446 Set the number of packets per burst::
1448 testpmd> port config all burst (value)
1450 This is equivalent to the ``--burst`` command-line option.
1452 port config - Threshold
1453 ~~~~~~~~~~~~~~~~~~~~~~~
1455 Set thresholds for TX/RX queues::
1457 testpmd> port config all (threshold) (value)
1459 Where the threshold type can be:
1461 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1463 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1465 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1467 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1469 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1471 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1473 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1475 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1477 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1479 These threshold options are also available from the command-line.
1484 Set the value of ether-type for E-tag::
1486 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1488 Enable/disable the E-tag support::
1490 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1493 Link Bonding Functions
1494 ----------------------
1496 The Link Bonding functions make it possible to dynamically create and
1497 manage link bonding devices from within testpmd interactive prompt.
1499 create bonded device
1500 ~~~~~~~~~~~~~~~~~~~~
1502 Create a new bonding device::
1504 testpmd> create bonded device (mode) (socket)
1506 For example, to create a bonded device in mode 1 on socket 0::
1508 testpmd> create bonded 1 0
1509 created new bonded device (port X)
1514 Adds Ethernet device to a Link Bonding device::
1516 testpmd> add bonding slave (slave id) (port id)
1518 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
1520 testpmd> add bonding slave 6 10
1523 remove bonding slave
1524 ~~~~~~~~~~~~~~~~~~~~
1526 Removes an Ethernet slave device from a Link Bonding device::
1528 testpmd> remove bonding slave (slave id) (port id)
1530 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
1532 testpmd> remove bonding slave 6 10
1537 Set the Link Bonding mode of a Link Bonding device::
1539 testpmd> set bonding mode (value) (port id)
1541 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
1543 testpmd> set bonding mode 3 10
1548 Set an Ethernet slave device as the primary device on a Link Bonding device::
1550 testpmd> set bonding primary (slave id) (port id)
1552 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
1554 testpmd> set bonding primary 6 10
1559 Set the MAC address of a Link Bonding device::
1561 testpmd> set bonding mac (port id) (mac)
1563 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
1565 testpmd> set bonding mac 10 00:00:00:00:00:01
1567 set bonding xmit_balance_policy
1568 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1570 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
1572 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
1574 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
1576 testpmd> set bonding xmit_balance_policy 10 l34
1579 set bonding mon_period
1580 ~~~~~~~~~~~~~~~~~~~~~~
1582 Set the link status monitoring polling period in milliseconds for a bonding device.
1584 This adds support for PMD slave devices which do not support link status interrupts.
1585 When the mon_period is set to a value greater than 0 then all PMD's which do not support
1586 link status ISR will be queried every polling interval to check if their link status has changed::
1588 testpmd> set bonding mon_period (port_id) (value)
1590 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
1592 testpmd> set bonding mon_period 5 150
1598 Show the current configuration of a Link Bonding device::
1600 testpmd> show bonding config (port id)
1603 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
1604 in balance mode with a transmission policy of layer 2+3::
1606 testpmd> show bonding config 9
1608 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
1610 Active Slaves (3): [1 3 4]
1617 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
1618 This is mainly useful for debugging purposes.
1619 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
1620 and fields that can be accessed.
1625 Display the value of a port register::
1627 testpmd> read reg (port_id) (address)
1629 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
1631 testpmd> read reg 0 0xEE00
1632 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
1637 Display a port register bit field::
1639 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
1641 For example, reading the lowest two bits from the register in the example above::
1643 testpmd> read regfield 0 0xEE00 0 1
1644 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
1649 Display a single port register bit::
1651 testpmd> read regbit (port_id) (address) (bit_x)
1653 For example, reading the lowest bit from the register in the example above::
1655 testpmd> read regbit 0 0xEE00 0
1656 port 0 PCI register at offset 0xEE00: bit 0=1
1661 Set the value of a port register::
1663 testpmd> write reg (port_id) (address) (value)
1665 For example, to clear a register::
1667 testpmd> write reg 0 0xEE00 0x0
1668 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
1673 Set bit field of a port register::
1675 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
1677 For example, writing to the register cleared in the example above::
1679 testpmd> write regfield 0 0xEE00 0 1 2
1680 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
1685 Set single bit value of a port register::
1687 testpmd> write regbit (port_id) (address) (bit_x) (value)
1689 For example, to set the high bit in the register from the example above::
1691 testpmd> write regbit 0 0xEE00 31 1
1692 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
1698 This section details the available filter functions that are available.
1700 Note these functions interface the deprecated legacy filtering framework,
1701 superseded by *rte_flow*. See `Flow rules management`_.
1704 ~~~~~~~~~~~~~~~~~~~~
1706 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
1708 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
1709 ethertype (ether_type) (drop|fwd) queue (queue_id)
1711 The available information parameters are:
1713 * ``port_id``: The port which the Ethertype filter assigned on.
1715 * ``mac_addr``: Compare destination mac address.
1717 * ``mac_ignr``: Ignore destination mac address match.
1719 * ``mac_address``: Destination mac address to match.
1721 * ``ether_type``: The EtherType value want to match,
1722 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
1724 * ``queue_id``: The receive queue associated with this EtherType filter.
1725 It is meaningless when deleting or dropping.
1727 Example, to add/remove an ethertype filter rule::
1729 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
1730 ethertype 0x0806 fwd queue 3
1732 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
1733 ethertype 0x0806 fwd queue 3
1738 Add or delete a 2-tuple filter,
1739 which identifies packets by specific protocol and destination TCP/UDP port
1740 and forwards packets into one of the receive queues::
1742 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
1743 protocol (protocol_value) mask (mask_value) \
1744 tcp_flags (tcp_flags_value) priority (prio_value) \
1747 The available information parameters are:
1749 * ``port_id``: The port which the 2-tuple filter assigned on.
1751 * ``dst_port_value``: Destination port in L4.
1753 * ``protocol_value``: IP L4 protocol.
1755 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
1757 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
1759 * ``prio_value``: Priority of this filter.
1761 * ``queue_id``: The receive queue associated with this 2-tuple filter.
1763 Example, to add/remove an 2tuple filter rule::
1765 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
1766 tcp_flags 0x02 priority 3 queue 3
1768 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
1769 tcp_flags 0x02 priority 3 queue 3
1774 Add or delete a 5-tuple filter,
1775 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
1776 and routes packets into one of the receive queues::
1778 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
1779 (src_address) dst_port (dst_port_value) \
1780 src_port (src_port_value) protocol (protocol_value) \
1781 mask (mask_value) tcp_flags (tcp_flags_value) \
1782 priority (prio_value) queue (queue_id)
1784 The available information parameters are:
1786 * ``port_id``: The port which the 5-tuple filter assigned on.
1788 * ``dst_address``: Destination IP address.
1790 * ``src_address``: Source IP address.
1792 * ``dst_port_value``: TCP/UDP destination port.
1794 * ``src_port_value``: TCP/UDP source port.
1796 * ``protocol_value``: L4 protocol.
1798 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
1800 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
1802 * ``prio_value``: The priority of this filter.
1804 * ``queue_id``: The receive queue associated with this 5-tuple filter.
1806 Example, to add/remove an 5tuple filter rule::
1808 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1809 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1810 flags 0x0 priority 3 queue 3
1812 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1813 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1814 flags 0x0 priority 3 queue 3
1819 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
1821 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
1823 The available information parameters are:
1825 * ``port_id``: The port which the SYN filter assigned on.
1827 * ``high``: This SYN filter has higher priority than other filters.
1829 * ``low``: This SYN filter has lower priority than other filters.
1831 * ``queue_id``: The receive queue associated with this SYN filter
1835 testpmd> syn_filter 0 add priority high queue 3
1840 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
1841 and routed into one of the receive queues::
1843 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
1844 mask (mask_value) priority (prio_value) queue (queue_id)
1846 The available information parameters are:
1848 * ``port_id``: The port which the Flex filter is assigned on.
1850 * ``len_value``: Filter length in bytes, no greater than 128.
1852 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
1854 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
1856 * ``prio_value``: The priority of this filter.
1858 * ``queue_id``: The receive queue associated with this Flex filter.
1862 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
1863 mask 000C priority 3 queue 3
1865 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
1866 mask 000C priority 3 queue 3
1869 .. _testpmd_flow_director:
1871 flow_director_filter
1872 ~~~~~~~~~~~~~~~~~~~~
1874 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
1876 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
1877 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
1879 * Perfect match filters.
1880 The hardware checks a match between the masked fields of the received packets and the programmed filters.
1881 The masked fields are for IP flow.
1883 * Signature filters.
1884 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
1886 * Perfect-mac-vlan match filters.
1887 The hardware checks a match between the masked fields of the received packets and the programmed filters.
1888 The masked fields are for MAC VLAN flow.
1890 * Perfect-tunnel match filters.
1891 The hardware checks a match between the masked fields of the received packets and the programmed filters.
1892 The masked fields are for tunnel flow.
1894 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
1895 per flow type and the flexible payload.
1897 The Flow Director can also mask out parts of all of these fields so that filters
1898 are only applied to certain fields or parts of the fields.
1900 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
1902 # Commands to add flow director filters of different flow types::
1904 flow_director_filter (port_id) mode IP (add|del|update) \
1905 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
1906 src (src_ip_address) dst (dst_ip_address) \
1907 tos (tos_value) proto (proto_value) ttl (ttl_value) \
1908 vlan (vlan_value) flexbytes (flexbytes_value) \
1909 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
1912 flow_director_filter (port_id) mode IP (add|del|update) \
1913 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
1914 src (src_ip_address) (src_port) \
1915 dst (dst_ip_address) (dst_port) \
1916 tos (tos_value) ttl (ttl_value) \
1917 vlan (vlan_value) flexbytes (flexbytes_value) \
1918 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
1921 flow_director_filter (port_id) mode IP (add|del|update) \
1922 flow (ipv4-sctp|ipv6-sctp) \
1923 src (src_ip_address) (src_port) \
1924 dst (dst_ip_address) (dst_port) \
1925 tos (tos_value) ttl (ttl_value) \
1926 tag (verification_tag) vlan (vlan_value) \
1927 flexbytes (flexbytes_value) (drop|fwd) \
1928 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
1930 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
1931 ether (ethertype) flexbytes (flexbytes_value) \
1932 (drop|fwd) pf|vf(vf_id) queue (queue_id)
1935 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
1936 mac (mac_address) vlan (vlan_value) \
1937 flexbytes (flexbytes_value) (drop|fwd) \
1938 queue (queue_id) fd_id (fd_id_value)
1940 flow_director_filter (port_id) mode Tunnel (add|del|update) \
1941 mac (mac_address) vlan (vlan_value) \
1942 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
1943 flexbytes (flexbytes_value) (drop|fwd) \
1944 queue (queue_id) fd_id (fd_id_value)
1946 For example, to add an ipv4-udp flow type filter::
1948 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
1949 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
1950 fwd pf queue 1 fd_id 1
1952 For example, add an ipv4-other flow type filter::
1954 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
1955 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
1956 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
1961 Flush all flow director filters on a device::
1963 testpmd> flush_flow_director (port_id)
1965 Example, to flush all flow director filter on port 0::
1967 testpmd> flush_flow_director 0
1972 Set flow director's input masks::
1974 flow_director_mask (port_id) mode IP vlan (vlan_value) \
1975 src_mask (ipv4_src) (ipv6_src) (src_port) \
1976 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
1978 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
1980 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
1981 mac (mac_value) tunnel-type (tunnel_type_value) \
1982 tunnel-id (tunnel_id_value)
1984 Example, to set flow director mask on port 0::
1986 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
1987 src_mask 255.255.255.255 \
1988 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
1989 dst_mask 255.255.255.255 \
1990 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
1992 flow_director_flex_mask
1993 ~~~~~~~~~~~~~~~~~~~~~~~
1995 set masks of flow director's flexible payload based on certain flow type::
1997 testpmd> flow_director_flex_mask (port_id) \
1998 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
1999 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2000 l2_payload|all) (mask)
2002 Example, to set flow director's flex mask for all flow type on port 0::
2004 testpmd> flow_director_flex_mask 0 flow all \
2005 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2008 flow_director_flex_payload
2009 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2011 Configure flexible payload selection::
2013 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2015 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2017 testpmd> flow_director_flex_payload 0 l4 \
2018 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2020 get_sym_hash_ena_per_port
2021 ~~~~~~~~~~~~~~~~~~~~~~~~~
2023 Get symmetric hash enable configuration per port::
2025 get_sym_hash_ena_per_port (port_id)
2027 For example, to get symmetric hash enable configuration of port 1::
2029 testpmd> get_sym_hash_ena_per_port 1
2031 set_sym_hash_ena_per_port
2032 ~~~~~~~~~~~~~~~~~~~~~~~~~
2034 Set symmetric hash enable configuration per port to enable or disable::
2036 set_sym_hash_ena_per_port (port_id) (enable|disable)
2038 For example, to set symmetric hash enable configuration of port 1 to enable::
2040 testpmd> set_sym_hash_ena_per_port 1 enable
2042 get_hash_global_config
2043 ~~~~~~~~~~~~~~~~~~~~~~
2045 Get the global configurations of hash filters::
2047 get_hash_global_config (port_id)
2049 For example, to get the global configurations of hash filters of port 1::
2051 testpmd> get_hash_global_config 1
2053 set_hash_global_config
2054 ~~~~~~~~~~~~~~~~~~~~~~
2056 Set the global configurations of hash filters::
2058 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2059 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2060 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload) \
2063 For example, to enable simple_xor for flow type of ipv6 on port 2::
2065 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2070 Set the input set for hash::
2072 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2073 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2074 l2_payload) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2075 ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2076 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2077 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2078 fld-8th|none) (select|add)
2080 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2082 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2087 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2088 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2090 Set the input set for flow director::
2092 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2093 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2094 l2_payload) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2095 ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2096 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2097 sctp-dst-port|sctp-veri-tag|none) (select|add)
2099 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2101 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2106 Set different GRE key length for input set::
2108 global_config (port_id) gre-key-len (number in bytes)
2110 For example to set GRE key length for input set to 4 bytes on port 0::
2112 testpmd> global_config 0 gre-key-len 4
2115 .. _testpmd_rte_flow:
2117 Flow rules management
2118 ---------------------
2120 Control of the generic flow API (*rte_flow*) is fully exposed through the
2121 ``flow`` command (validation, creation, destruction and queries).
2123 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2124 features simultaneously may cause undefined side-effects and is therefore
2130 Because the ``flow`` command uses dynamic tokens to handle the large number
2131 of possible flow rules combinations, its behavior differs slightly from
2132 other commands, in particular:
2134 - Pressing *?* or the *<tab>* key displays contextual help for the current
2135 token, not that of the entire command.
2137 - Optional and repeated parameters are supported (provided they are listed
2138 in the contextual help).
2140 The first parameter stands for the operation mode. Possible operations and
2141 their general syntax are described below. They are covered in detail in the
2144 - Check whether a flow rule can be created::
2146 flow validate {port_id}
2147 [group {group_id}] [priority {level}] [ingress] [egress]
2148 pattern {item} [/ {item} [...]] / end
2149 actions {action} [/ {action} [...]] / end
2151 - Create a flow rule::
2153 flow create {port_id}
2154 [group {group_id}] [priority {level}] [ingress] [egress]
2155 pattern {item} [/ {item} [...]] / end
2156 actions {action} [/ {action} [...]] / end
2158 - Destroy specific flow rules::
2160 flow destroy {port_id} rule {rule_id} [...]
2162 - Destroy all flow rules::
2164 flow flush {port_id}
2166 - Query an existing flow rule::
2168 flow query {port_id} {rule_id} {action}
2170 - List existing flow rules sorted by priority, filtered by group
2173 flow list {port_id} [group {group_id}] [...]
2175 Validating flow rules
2176 ~~~~~~~~~~~~~~~~~~~~~
2178 ``flow validate`` reports whether a flow rule would be accepted by the
2179 underlying device in its current state but stops short of creating it. It is
2180 bound to ``rte_flow_validate()``::
2182 flow validate {port_id}
2183 [group {group_id}] [priority {level}] [ingress] [egress]
2184 pattern {item} [/ {item} [...]] / end
2185 actions {action} [/ {action} [...]] / end
2187 If successful, it will show::
2191 Otherwise it will show an error message of the form::
2193 Caught error type [...] ([...]): [...]
2195 This command uses the same parameters as ``flow create``, their format is
2196 described in `Creating flow rules`_.
2198 Check whether redirecting any Ethernet packet received on port 0 to RX queue
2199 index 6 is supported::
2201 testpmd> flow validate 0 ingress pattern eth / end
2202 actions queue index 6 / end
2206 Port 0 does not support TCPv6 rules::
2208 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
2210 Caught error type 9 (specific pattern item): Invalid argument
2216 ``flow create`` validates and creates the specified flow rule. It is bound
2217 to ``rte_flow_create()``::
2219 flow create {port_id}
2220 [group {group_id}] [priority {level}] [ingress] [egress]
2221 pattern {item} [/ {item} [...]] / end
2222 actions {action} [/ {action} [...]] / end
2224 If successful, it will return a flow rule ID usable with other commands::
2226 Flow rule #[...] created
2228 Otherwise it will show an error message of the form::
2230 Caught error type [...] ([...]): [...]
2232 Parameters describe in the following order:
2234 - Attributes (*group*, *priority*, *ingress*, *egress* tokens).
2235 - A matching pattern, starting with the *pattern* token and terminated by an
2237 - Actions, starting with the *actions* token and terminated by an *end*
2240 These translate directly to *rte_flow* objects provided as-is to the
2241 underlying functions.
2243 The shortest valid definition only comprises mandatory tokens::
2245 testpmd> flow create 0 pattern end actions end
2247 Note that PMDs may refuse rules that essentially do nothing such as this
2250 **All unspecified object values are automatically initialized to 0.**
2255 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
2256 specified before the ``pattern`` token.
2258 - ``group {group id}``: priority group.
2259 - ``priority {level}``: priority level within group.
2260 - ``ingress``: rule applies to ingress traffic.
2261 - ``egress``: rule applies to egress traffic.
2263 Each instance of an attribute specified several times overrides the previous
2264 value as shown below (group 4 is used)::
2266 testpmd> flow create 0 group 42 group 24 group 4 [...]
2268 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
2270 While not specifying a direction is an error, some rules may allow both
2273 Most rules affect RX therefore contain the ``ingress`` token::
2275 testpmd> flow create 0 ingress pattern [...]
2280 A matching pattern starts after the ``pattern`` token. It is made of pattern
2281 items and is terminated by a mandatory ``end`` item.
2283 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
2284 rte_flow_item_type``).
2286 The ``/`` token is used as a separator between pattern items as shown
2289 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
2291 Note that protocol items like these must be stacked from lowest to highest
2292 layer to make sense. For instance, the following rule is either invalid or
2293 unlikely to match any packet::
2295 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
2297 More information on these restrictions can be found in the *rte_flow*
2300 Several items support additional specification structures, for example
2301 ``ipv4`` allows specifying source and destination addresses as follows::
2303 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
2304 dst is 10.2.0.0 / end [...]
2306 This rule matches all IPv4 traffic with the specified properties.
2308 In this example, ``src`` and ``dst`` are field names of the underlying
2309 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
2310 in a similar fashion.
2312 The ``is`` token means that the subsequent value must be matched exactly,
2313 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
2314 accordingly. Possible assignment tokens are:
2316 - ``is``: match value perfectly (with full bit-mask).
2317 - ``spec``: match value according to configured bit-mask.
2318 - ``last``: specify upper bound to establish a range.
2319 - ``mask``: specify bit-mask with relevant bits set to one.
2320 - ``prefix``: generate bit-mask from a prefix length.
2322 These yield identical results::
2324 ipv4 src is 10.1.1.1
2328 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
2332 ipv4 src spec 10.1.1.1 src prefix 32
2336 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
2340 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
2342 Inclusive ranges can be defined with ``last``::
2344 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
2346 Note that ``mask`` affects both ``spec`` and ``last``::
2348 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
2349 # matches 10.1.0.0 to 10.2.255.255
2351 Properties can be modified multiple times::
2353 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
2357 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
2362 This section lists supported pattern items and their attributes, if any.
2364 - ``end``: end list of pattern items.
2366 - ``void``: no-op pattern item.
2368 - ``invert``: perform actions when pattern does not match.
2370 - ``any``: match any protocol for the current layer.
2372 - ``num {unsigned}``: number of layers covered.
2374 - ``pf``: match packets addressed to the physical function.
2376 - ``vf``: match packets addressed to a virtual function ID.
2378 - ``id {unsigned}``: destination VF ID.
2380 - ``port``: device-specific physical port index to use.
2382 - ``index {unsigned}``: physical port index.
2384 - ``raw``: match an arbitrary byte string.
2386 - ``relative {boolean}``: look for pattern after the previous item.
2387 - ``search {boolean}``: search pattern from offset (see also limit).
2388 - ``offset {integer}``: absolute or relative offset for pattern.
2389 - ``limit {unsigned}``: search area limit for start of pattern.
2390 - ``pattern {string}``: byte string to look for.
2392 - ``eth``: match Ethernet header.
2394 - ``dst {MAC-48}``: destination MAC.
2395 - ``src {MAC-48}``: source MAC.
2396 - ``type {unsigned}``: EtherType.
2398 - ``vlan``: match 802.1Q/ad VLAN tag.
2400 - ``tpid {unsigned}``: tag protocol identifier.
2401 - ``tci {unsigned}``: tag control information.
2402 - ``pcp {unsigned}``: priority code point.
2403 - ``dei {unsigned}``: drop eligible indicator.
2404 - ``vid {unsigned}``: VLAN identifier.
2406 - ``ipv4``: match IPv4 header.
2408 - ``tos {unsigned}``: type of service.
2409 - ``ttl {unsigned}``: time to live.
2410 - ``proto {unsigned}``: next protocol ID.
2411 - ``src {ipv4 address}``: source address.
2412 - ``dst {ipv4 address}``: destination address.
2414 - ``ipv6``: match IPv6 header.
2416 - ``tc {unsigned}``: traffic class.
2417 - ``flow {unsigned}``: flow label.
2418 - ``proto {unsigned}``: protocol (next header).
2419 - ``hop {unsigned}``: hop limit.
2420 - ``src {ipv6 address}``: source address.
2421 - ``dst {ipv6 address}``: destination address.
2423 - ``icmp``: match ICMP header.
2425 - ``type {unsigned}``: ICMP packet type.
2426 - ``code {unsigned}``: ICMP packet code.
2428 - ``udp``: match UDP header.
2430 - ``src {unsigned}``: UDP source port.
2431 - ``dst {unsigned}``: UDP destination port.
2433 - ``tcp``: match TCP header.
2435 - ``src {unsigned}``: TCP source port.
2436 - ``dst {unsigned}``: TCP destination port.
2438 - ``sctp``: match SCTP header.
2440 - ``src {unsigned}``: SCTP source port.
2441 - ``dst {unsigned}``: SCTP destination port.
2442 - ``tag {unsigned}``: validation tag.
2443 - ``cksum {unsigned}``: checksum.
2445 - ``vxlan``: match VXLAN header.
2447 - ``vni {unsigned}``: VXLAN identifier.
2452 A list of actions starts after the ``actions`` token in the same fashion as
2453 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
2454 terminated by a mandatory ``end`` action.
2456 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
2457 rte_flow_action_type``).
2459 Dropping all incoming UDPv4 packets can be expressed as follows::
2461 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2464 Several actions have configurable properties which must be specified when
2465 there is no valid default value. For example, ``queue`` requires a target
2468 This rule redirects incoming UDPv4 traffic to queue index 6::
2470 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2471 actions queue index 6 / end
2473 While this one could be rejected by PMDs (unspecified queue index)::
2475 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2478 As defined by *rte_flow*, the list is not ordered, all actions of a given
2479 rule are performed simultaneously. These are equivalent::
2481 queue index 6 / void / mark id 42 / end
2485 void / mark id 42 / queue index 6 / end
2487 All actions in a list should have different types, otherwise only the last
2488 action of a given type is taken into account::
2490 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
2494 drop / drop / drop / end # drop is performed only once
2498 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
2500 Considering they are performed simultaneously, opposite and overlapping
2501 actions can sometimes be combined when the end result is unambiguous::
2503 drop / queue index 6 / end # drop has no effect
2507 drop / dup index 6 / end # same as above
2511 queue index 6 / rss queues 6 7 8 / end # queue has no effect
2515 drop / passthru / end # drop has no effect
2517 Note that PMDs may still refuse such combinations.
2522 This section lists supported actions and their attributes, if any.
2524 - ``end``: end list of actions.
2526 - ``void``: no-op action.
2528 - ``passthru``: let subsequent rule process matched packets.
2530 - ``mark``: attach 32 bit value to packets.
2532 - ``id {unsigned}``: 32 bit value to return with packets.
2534 - ``flag``: flag packets.
2536 - ``queue``: assign packets to a given queue index.
2538 - ``index {unsigned}``: queue index to use.
2540 - ``drop``: drop packets (note: passthru has priority).
2542 - ``count``: enable counters for this rule.
2544 - ``dup``: duplicate packets to a given queue index.
2546 - ``index {unsigned}``: queue index to duplicate packets to.
2548 - ``rss``: spread packets among several queues.
2550 - ``queues [{unsigned} [...]] end``: queue indices to use.
2552 - ``pf``: redirect packets to physical device function.
2554 - ``vf``: redirect packets to virtual device function.
2556 - ``original {boolean}``: use original VF ID if possible.
2557 - ``id {unsigned}``: VF ID to redirect packets to.
2559 Destroying flow rules
2560 ~~~~~~~~~~~~~~~~~~~~~
2562 ``flow destroy`` destroys one or more rules from their rule ID (as returned
2563 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
2564 times as necessary::
2566 flow destroy {port_id} rule {rule_id} [...]
2568 If successful, it will show::
2570 Flow rule #[...] destroyed
2572 It does not report anything for rule IDs that do not exist. The usual error
2573 message is shown when a rule cannot be destroyed::
2575 Caught error type [...] ([...]): [...]
2577 ``flow flush`` destroys all rules on a device and does not take extra
2578 arguments. It is bound to ``rte_flow_flush()``::
2580 flow flush {port_id}
2582 Any errors are reported as above.
2584 Creating several rules and destroying them::
2586 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2587 actions queue index 2 / end
2588 Flow rule #0 created
2589 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2590 actions queue index 3 / end
2591 Flow rule #1 created
2592 testpmd> flow destroy 0 rule 0 rule 1
2593 Flow rule #1 destroyed
2594 Flow rule #0 destroyed
2597 The same result can be achieved using ``flow flush``::
2599 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2600 actions queue index 2 / end
2601 Flow rule #0 created
2602 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2603 actions queue index 3 / end
2604 Flow rule #1 created
2605 testpmd> flow flush 0
2608 Non-existent rule IDs are ignored::
2610 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2611 actions queue index 2 / end
2612 Flow rule #0 created
2613 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2614 actions queue index 3 / end
2615 Flow rule #1 created
2616 testpmd> flow destroy 0 rule 42 rule 10 rule 2
2618 testpmd> flow destroy 0 rule 0
2619 Flow rule #0 destroyed
2625 ``flow query`` queries a specific action of a flow rule having that
2626 ability. Such actions collect information that can be reported using this
2627 command. It is bound to ``rte_flow_query()``::
2629 flow query {port_id} {rule_id} {action}
2631 If successful, it will display either the retrieved data for known actions
2632 or the following message::
2634 Cannot display result for action type [...] ([...])
2636 Otherwise, it will complain either that the rule does not exist or that some
2639 Flow rule #[...] not found
2643 Caught error type [...] ([...]): [...]
2645 Currently only the ``count`` action is supported. This action reports the
2646 number of packets that hit the flow rule and the total number of bytes. Its
2647 output has the following format::
2650 hits_set: [...] # whether "hits" contains a valid value
2651 bytes_set: [...] # whether "bytes" contains a valid value
2652 hits: [...] # number of packets
2653 bytes: [...] # number of bytes
2655 Querying counters for TCPv6 packets redirected to queue 6::
2657 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
2658 actions queue index 6 / count / end
2659 Flow rule #4 created
2660 testpmd> flow query 0 4 count
2671 ``flow list`` lists existing flow rules sorted by priority and optionally
2672 filtered by group identifiers::
2674 flow list {port_id} [group {group_id}] [...]
2676 This command only fails with the following message if the device does not
2681 Output consists of a header line followed by a short description of each
2682 flow rule, one per line. There is no output at all when no flow rules are
2683 configured on the device::
2685 ID Group Prio Attr Rule
2686 [...] [...] [...] [...] [...]
2688 ``Attr`` column flags:
2690 - ``i`` for ``ingress``.
2691 - ``e`` for ``egress``.
2693 Creating several flow rules and listing them::
2695 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2696 actions queue index 6 / end
2697 Flow rule #0 created
2698 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2699 actions queue index 2 / end
2700 Flow rule #1 created
2701 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
2702 actions rss queues 6 7 8 end / end
2703 Flow rule #2 created
2704 testpmd> flow list 0
2705 ID Group Prio Attr Rule
2706 0 0 0 i- ETH IPV4 => QUEUE
2707 1 0 0 i- ETH IPV6 => QUEUE
2708 2 0 5 i- ETH IPV4 UDP => RSS
2711 Rules are sorted by priority (i.e. group ID first, then priority level)::
2713 testpmd> flow list 1
2714 ID Group Prio Attr Rule
2715 0 0 0 i- ETH => COUNT
2716 6 0 500 i- ETH IPV6 TCP => DROP COUNT
2717 5 0 1000 i- ETH IPV6 ICMP => QUEUE
2718 1 24 0 i- ETH IPV4 UDP => QUEUE
2719 4 24 10 i- ETH IPV4 TCP => DROP
2720 3 24 20 i- ETH IPV4 => DROP
2721 2 24 42 i- ETH IPV4 UDP => QUEUE
2722 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
2725 Output can be limited to specific groups::
2727 testpmd> flow list 1 group 0 group 63
2728 ID Group Prio Attr Rule
2729 0 0 0 i- ETH => COUNT
2730 6 0 500 i- ETH IPV6 TCP => DROP COUNT
2731 5 0 1000 i- ETH IPV6 ICMP => QUEUE
2732 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE