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 X
55 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc all
56 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc X
57 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc 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) (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.
152 .. code-block:: console
154 testpmd> show port info 0
156 ********************* Infos for port 0 *********************
158 MAC address: XX:XX:XX:XX:XX:XX
160 memory allocation on the socket: 0
162 Link speed: 40000 Mbps
163 Link duplex: full-duplex
164 Promiscuous mode: enabled
165 Allmulticast mode: disabled
166 Maximum number of MAC addresses: 64
167 Maximum number of MAC addresses of hash filtering: 0
172 Redirection table size: 512
173 Supported flow types:
193 Display the rss redirection table entry indicated by masks on port X::
195 testpmd> show port (port_id) rss reta (size) (mask0, mask1...)
197 size is used to indicate the hardware supported reta size
202 Display the RSS hash functions and RSS hash key of a port::
204 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]
209 Clear the port statistics for a given port or for all ports::
211 testpmd> clear port (info|stats|xstats|fdir|stat_qmap) (port_id|all)
215 testpmd> clear port stats all
220 Display information for a given port's RX/TX queue::
222 testpmd> show (rxq|txq) info (port_id) (queue_id)
227 Displays the configuration of the application.
228 The configuration comes from the command-line, the runtime or the application defaults::
230 testpmd> show config (rxtx|cores|fwd|txpkts)
232 The available information categories are:
234 * ``rxtx``: RX/TX configuration items.
236 * ``cores``: List of forwarding cores.
238 * ``fwd``: Packet forwarding configuration.
240 * ``txpkts``: Packets to TX configuration.
244 .. code-block:: console
246 testpmd> show config rxtx
248 io packet forwarding - CRC stripping disabled - packets/burst=16
249 nb forwarding cores=2 - nb forwarding ports=1
250 RX queues=1 - RX desc=128 - RX free threshold=0
251 RX threshold registers: pthresh=8 hthresh=8 wthresh=4
252 TX queues=1 - TX desc=512 - TX free threshold=0
253 TX threshold registers: pthresh=36 hthresh=0 wthresh=0
254 TX RS bit threshold=0 - TXQ flags=0x0
259 Set the packet forwarding mode::
261 testpmd> set fwd (io|mac|macswap|flowgen| \
262 rxonly|txonly|csum|icmpecho) (""|retry)
264 ``retry`` can be specified for forwarding engines except ``rx_only``.
266 The available information categories are:
268 * ``io``: Forwards packets "as-is" in I/O mode.
269 This is the fastest possible forwarding operation as it does not access packets data.
270 This is the default mode.
272 * ``mac``: Changes the source and the destination Ethernet addresses of packets before forwarding them.
273 Default application behaviour is to set source Ethernet address to that of the transmitting interface, and destination
274 address to a dummy value (set during init). The user may specify a target destination Ethernet address via the 'eth-peer' or
275 'eth-peer-configfile' command-line options. It is not currently possible to specify a specific source Ethernet address.
277 * ``macswap``: MAC swap forwarding mode.
278 Swaps the source and the destination Ethernet addresses of packets before forwarding them.
280 * ``flowgen``: Multi-flow generation mode.
281 Originates a number of flows (with varying destination IP addresses), and terminate receive traffic.
283 * ``rxonly``: Receives packets but doesn't transmit them.
285 * ``txonly``: Generates and transmits packets without receiving any.
287 * ``csum``: Changes the checksum field with hardware or software methods depending on the offload flags on the packet.
289 * ``icmpecho``: Receives a burst of packets, lookup for IMCP echo requests and, if any, send back ICMP echo replies.
291 * ``ieee1588``: Demonstrate L2 IEEE1588 V2 PTP timestamping for RX and TX. Requires ``CONFIG_RTE_LIBRTE_IEEE1588=y``.
293 Note: TX timestamping is only available in the "Full Featured" TX path. To force ``testpmd`` into this mode set ``--txqflags=0``.
297 testpmd> set fwd rxonly
299 Set rxonly packet forwarding mode
305 Display an RX descriptor for a port RX queue::
307 testpmd> read rxd (port_id) (queue_id) (rxd_id)
311 testpmd> read rxd 0 0 4
312 0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180
317 Display a TX descriptor for a port TX queue::
319 testpmd> read txd (port_id) (queue_id) (txd_id)
323 testpmd> read txd 0 0 4
324 0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C
327 Configuration Functions
328 -----------------------
330 The testpmd application can be configured from the runtime as well as from the command-line.
332 This section details the available configuration functions that are available.
336 Configuration changes only become active when forwarding is started/restarted.
341 Reset forwarding to the default configuration::
348 Set the debug verbosity level::
350 testpmd> set verbose (level)
352 Currently the only available levels are 0 (silent except for error) and 1 (fully verbose).
357 Set the number of ports used by the application:
361 This is equivalent to the ``--nb-ports`` command-line option.
366 Set the number of cores used by the application::
368 testpmd> set nbcore (num)
370 This is equivalent to the ``--nb-cores`` command-line option.
374 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port.
379 Set the forwarding cores hexadecimal mask::
381 testpmd> set coremask (mask)
383 This is equivalent to the ``--coremask`` command-line option.
387 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding.
392 Set the forwarding ports hexadecimal mask::
394 testpmd> set portmask (mask)
396 This is equivalent to the ``--portmask`` command-line option.
401 Set number of packets per burst::
403 testpmd> set burst (num)
405 This is equivalent to the ``--burst command-line`` option.
407 When retry is enabled, the transmit delay time and number of retries can also be set::
409 testpmd> set burst tx delay (microseconds) retry (num)
414 Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode::
416 testpmd> set txpkts (x[,y]*)
418 Where x[,y]* represents a CSV list of values, without white space.
423 Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes::
425 testpmd> set txsplit (off|on|rand)
429 * ``off`` disable packet copy & split for CSUM mode.
431 * ``on`` split outgoing packet into multiple segments. Size of each segment
432 and number of segments per packet is determined by ``set txpkts`` command
435 * ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments.
440 Set the list of forwarding cores::
442 testpmd> set corelist (x[,y]*)
444 For example, to change the forwarding cores:
446 .. code-block:: console
448 testpmd> set corelist 3,1
449 testpmd> show config fwd
451 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled
452 Logical Core 3 (socket 0) forwards packets on 1 streams:
453 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01
454 Logical Core 1 (socket 0) forwards packets on 1 streams:
455 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
459 The cores are used in the same order as specified on the command line.
464 Set the list of forwarding ports::
466 testpmd> set portlist (x[,y]*)
468 For example, to change the port forwarding:
470 .. code-block:: console
472 testpmd> set portlist 0,2,1,3
473 testpmd> show config fwd
475 io packet forwarding - ports=4 - cores=1 - streams=4
476 Logical Core 3 (socket 0) forwards packets on 4 streams:
477 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01
478 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
479 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03
480 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02
485 Enable/disable tx loopback::
487 testpmd> set tx loopback (port_id) (on|off)
492 set drop enable bit for all queues::
494 testpmd> set all queues drop (port_id) (on|off)
496 set split drop enable (for VF)
497 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
499 set split drop enable bit for VF from PF::
501 testpmd> set vf split drop (port_id) (vf_id) (on|off)
503 set mac antispoof (for VF)
504 ~~~~~~~~~~~~~~~~~~~~~~~~~~
506 Set mac antispoof for a VF from the PF::
508 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off)
513 Enable/disable MACsec offload::
515 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off)
516 testpmd> set macsec offload (port_id) off
521 Configure MACsec secure connection (SC)::
523 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi)
527 The pi argument is ignored for tx.
528 Check the NIC Datasheet for hardware limits.
533 Configure MACsec secure association (SA)::
535 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key)
539 The IDX value must be 0 or 1.
540 Check the NIC Datasheet for hardware limits.
545 Set the VLAN strip on a port::
547 testpmd> vlan set strip (on|off) (port_id)
552 Set the VLAN strip for a queue on a port::
554 testpmd> vlan set stripq (on|off) (port_id,queue_id)
556 vlan set stripq (for VF)
557 ~~~~~~~~~~~~~~~~~~~~~~~~
559 Set VLAN strip for all queues in a pool for a VF from the PF::
561 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off)
563 vlan set insert (for VF)
564 ~~~~~~~~~~~~~~~~~~~~~~~~
566 Set VLAN insert for a VF from the PF::
568 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id)
570 vlan set antispoof (for VF)
571 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
573 Set VLAN antispoof for a VF from the PF::
575 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off)
580 Set the VLAN filter on a port::
582 testpmd> vlan set filter (on|off) (port_id)
587 Set the VLAN QinQ (extended queue in queue) on for a port::
589 testpmd> vlan set qinq (on|off) (port_id)
594 Set the inner or outer VLAN TPID for packet filtering on a port::
596 testpmd> vlan set (inner|outer) tpid (value) (port_id)
600 TPID value must be a 16-bit number (value <= 65536).
605 Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID::
607 testpmd> rx_vlan add (vlan_id|all) (port_id)
611 VLAN filter must be set on that port. VLAN ID < 4096.
612 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries
613 in VFTA table. This is important if enabling all vlan_ids.
618 Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID::
620 testpmd> rx_vlan rm (vlan_id|all) (port_id)
625 Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID::
627 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask)
632 Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID::
634 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask)
639 Add a tunnel filter on a port::
641 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \
642 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
643 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
645 The available information categories are:
647 * ``vxlan``: Set tunnel type as VXLAN.
649 * ``nvgre``: Set tunnel type as NVGRE.
651 * ``ipingre``: Set tunnel type as IP-in-GRE.
653 * ``imac-ivlan``: Set filter type as Inner MAC and VLAN.
655 * ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID.
657 * ``imac-tenid``: Set filter type as Inner MAC and tenant ID.
659 * ``imac``: Set filter type as Inner MAC.
661 * ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID.
663 * ``oip``: Set filter type as Outer IP.
665 * ``iip``: Set filter type as Inner IP.
669 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \
670 192.168.2.2 0 ipingre oip 1 1
672 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP.
677 Remove a tunnel filter on a port::
679 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \
680 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
681 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
686 Add an UDP port for VXLAN packet filter on a port::
688 testpmd> rx_vxlan_port add (udp_port) (port_id)
693 Remove an UDP port for VXLAN packet filter on a port::
695 testpmd> rx_vxlan_port rm (udp_port) (port_id)
700 Set hardware insertion of VLAN IDs in packets sent on a port::
702 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer]
704 For example, set a single VLAN ID (5) insertion on port 0::
708 Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1::
716 Set port based hardware insertion of VLAN ID in packets sent on a port::
718 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off)
723 Disable hardware insertion of a VLAN header in packets sent on a port::
725 testpmd> tx_vlan reset (port_id)
730 Select hardware or software calculation of the checksum when
731 transmitting a packet using the ``csum`` forwarding engine::
733 testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id)
737 * ``ip|udp|tcp|sctp`` always relate to the inner layer.
739 * ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized
740 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are
741 supported). See also the ``csum parse-tunnel`` command.
745 Check the NIC Datasheet for hardware limits.
750 Define how tunneled packets should be handled by the csum forward
753 testpmd> csum parse-tunnel (on|off) (tx_port_id)
755 If enabled, the csum forward engine will try to recognize supported
756 tunnel headers (vxlan, gre, ipip).
758 If disabled, treat tunnel packets as non-tunneled packets (a inner
759 header is handled as a packet payload).
763 The port argument is the TX port like in the ``csum set`` command.
767 Consider a packet in packet like the following::
769 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in
771 * If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
772 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the
773 ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``).
775 * If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
776 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``.
781 Display tx checksum offload configuration::
783 testpmd> csum show (port_id)
788 Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine::
790 testpmd> tso set (segsize) (port_id)
794 Check the NIC datasheet for hardware limits.
799 Display the status of TCP Segmentation Offload::
801 testpmd> tso show (port_id)
806 Add an alternative MAC address to a port::
808 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX)
813 Remove a MAC address from a port::
815 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX)
817 mac_addr add (for VF)
818 ~~~~~~~~~~~~~~~~~~~~~
820 Add an alternative MAC address for a VF to a port::
822 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX)
824 mac_addr set (for VF)
825 ~~~~~~~~~~~~~~~~~~~~~
827 Set the MAC address for a VF from the PF::
829 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX)
834 Set the unicast hash filter(s) on/off for a port::
836 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
841 Set the promiscuous mode on for a port or for all ports.
842 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
844 testpmd> set promisc (port_id|all) (on|off)
849 Set the allmulti mode for a port or for all ports::
851 testpmd> set allmulti (port_id|all) (on|off)
853 Same as the ifconfig (8) option. Controls how multicast packets are handled.
858 Set the unicast promiscuous mode for a VF from PF.
859 It's supported by Intel i40e NICs now.
860 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
862 testpmd> set vf promisc (port_id) (vf_id) (on|off)
864 set allmulticast (for VF)
865 ~~~~~~~~~~~~~~~~~~~~~~~~~
867 Set the multicast promiscuous mode for a VF from PF.
868 It's supported by Intel i40e NICs now.
869 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
871 testpmd> set vf allmulti (port_id) (vf_id) (on|off)
876 Set the link flow control parameter on a port::
878 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
879 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
880 autoneg (on|off) (port_id)
884 * ``high_water`` (integer): High threshold value to trigger XOFF.
886 * ``low_water`` (integer): Low threshold value to trigger XON.
888 * ``pause_time`` (integer): Pause quota in the Pause frame.
890 * ``send_xon`` (0/1): Send XON frame.
892 * ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
894 * ``autoneg``: Change the auto-negotiation parameter.
899 Set the priority flow control parameter on a port::
901 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
902 (pause_time) (priority) (port_id)
906 * ``high_water`` (integer): High threshold value.
908 * ``low_water`` (integer): Low threshold value.
910 * ``pause_time`` (integer): Pause quota in the Pause frame.
912 * ``priority`` (0-7): VLAN User Priority.
917 Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
919 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
921 For example, to set rx queue 2 on port 0 to mapping 5::
923 testpmd>set stat_qmap rx 0 2 5
925 set port - rx/tx (for VF)
926 ~~~~~~~~~~~~~~~~~~~~~~~~~
928 Set VF receive/transmit from a port::
930 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
932 set port - mac address filter (for VF)
933 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
935 Add/Remove unicast or multicast MAC addr filter for a VF::
937 testpmd> set port (port_id) vf (vf_id) (mac_addr) \
938 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
940 set port - rx mode(for VF)
941 ~~~~~~~~~~~~~~~~~~~~~~~~~~
943 Set the VF receive mode of a port::
945 testpmd> set port (port_id) vf (vf_id) \
946 rxmode (AUPE|ROPE|BAM|MPE) (on|off)
948 The available receive modes are:
950 * ``AUPE``: Accepts untagged VLAN.
952 * ``ROPE``: Accepts unicast hash.
954 * ``BAM``: Accepts broadcast packets.
956 * ``MPE``: Accepts all multicast packets.
958 set port - tx_rate (for Queue)
959 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
961 Set TX rate limitation for a queue on a port::
963 testpmd> set port (port_id) queue (queue_id) rate (rate_value)
965 set port - tx_rate (for VF)
966 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
968 Set TX rate limitation for queues in VF on a port::
970 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
972 set port - mirror rule
973 ~~~~~~~~~~~~~~~~~~~~~~
975 Set pool or vlan type mirror rule for a port::
977 testpmd> set port (port_id) mirror-rule (rule_id) \
978 (pool-mirror-up|pool-mirror-down|vlan-mirror) \
979 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
981 Set link mirror rule for a port::
983 testpmd> set port (port_id) mirror-rule (rule_id) \
984 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
986 For example to enable mirror traffic with vlan 0,1 to pool 0::
988 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
990 reset port - mirror rule
991 ~~~~~~~~~~~~~~~~~~~~~~~~
993 Reset a mirror rule for a port::
995 testpmd> reset port (port_id) mirror-rule (rule_id)
1000 Set the flush on RX streams before forwarding.
1001 The default is flush ``on``.
1002 Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1004 testpmd> set flush_rx off
1009 Set the bypass mode for the lowest port on bypass enabled NIC::
1011 testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1016 Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1018 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1019 mode (normal|bypass|isolate) (port_id)
1023 * ``timeout``: Enable bypass after watchdog timeout.
1025 * ``os_on``: Enable bypass when OS/board is powered on.
1027 * ``os_off``: Enable bypass when OS/board is powered off.
1029 * ``power_on``: Enable bypass when power supply is turned on.
1031 * ``power_off``: Enable bypass when power supply is turned off.
1037 Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1039 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1044 Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1046 testpmd> show bypass config (port_id)
1051 Set link up for a port::
1053 testpmd> set link-up port (port id)
1058 Set link down for a port::
1060 testpmd> set link-down port (port id)
1065 Enable E-tag insertion for a VF on a port::
1067 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1069 Disable E-tag insertion for a VF on a port::
1071 testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1073 Enable/disable E-tag stripping on a port::
1075 testpmd> E-tag set stripping (on|off) port (port_id)
1077 Enable/disable E-tag based forwarding on a port::
1079 testpmd> E-tag set forwarding (on|off) port (port_id)
1081 Add an E-tag forwarding filter on a port::
1083 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1085 Delete an E-tag forwarding filter on a port::
1086 testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1092 The following sections show functions for configuring ports.
1096 Port configuration changes only become active when forwarding is started/restarted.
1101 Attach a port specified by pci address or virtual device args::
1103 testpmd> port attach (identifier)
1105 To attach a new pci device, the device should be recognized by kernel first.
1106 Then it should be moved under DPDK management.
1107 Finally the port can be attached to testpmd.
1109 For example, to move a pci device using ixgbe under DPDK management:
1111 .. code-block:: console
1113 # Check the status of the available devices.
1114 ./usertools/dpdk-devbind.py --status
1116 Network devices using DPDK-compatible driver
1117 ============================================
1120 Network devices using kernel driver
1121 ===================================
1122 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1125 # Bind the device to igb_uio.
1126 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1129 # Recheck the status of the devices.
1130 ./usertools/dpdk-devbind.py --status
1131 Network devices using DPDK-compatible driver
1132 ============================================
1133 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1135 To attach a port created by virtual device, above steps are not needed.
1137 For example, to attach a port whose pci address is 0000:0a:00.0.
1139 .. code-block:: console
1141 testpmd> port attach 0000:0a:00.0
1142 Attaching a new port...
1143 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1144 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1145 EAL: PCI memory mapped at 0x7f83bfa00000
1146 EAL: PCI memory mapped at 0x7f83bfa80000
1147 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1148 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1149 Port 0 is attached. Now total ports is 1
1152 For example, to attach a port created by pcap PMD.
1154 .. code-block:: console
1156 testpmd> port attach net_pcap0
1157 Attaching a new port...
1158 PMD: Initializing pmd_pcap for net_pcap0
1159 PMD: Creating pcap-backed ethdev on numa socket 0
1160 Port 0 is attached. Now total ports is 1
1163 In this case, identifier is ``net_pcap0``.
1164 This identifier format is the same as ``--vdev`` format of DPDK applications.
1166 For example, to re-attach a bonded port which has been previously detached,
1167 the mode and slave parameters must be given.
1169 .. code-block:: console
1171 testpmd> port attach net_bond_0,mode=0,slave=1
1172 Attaching a new port...
1173 EAL: Initializing pmd_bond for net_bond_0
1174 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1175 Port 0 is attached. Now total ports is 1
1182 Detach a specific port::
1184 testpmd> port detach (port_id)
1186 Before detaching a port, the port should be stopped and closed.
1188 For example, to detach a pci device port 0.
1190 .. code-block:: console
1192 testpmd> port stop 0
1195 testpmd> port close 0
1199 testpmd> port detach 0
1201 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1202 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1203 EAL: PCI memory unmapped at 0x7f83bfa00000
1204 EAL: PCI memory unmapped at 0x7f83bfa80000
1208 For example, to detach a virtual device port 0.
1210 .. code-block:: console
1212 testpmd> port stop 0
1215 testpmd> port close 0
1219 testpmd> port detach 0
1221 PMD: Closing pcap ethdev on numa socket 0
1222 Port 'net_pcap0' is detached. Now total ports is 0
1225 To remove a pci device completely from the system, first detach the port from testpmd.
1226 Then the device should be moved under kernel management.
1227 Finally the device can be removed using kernel pci hotplug functionality.
1229 For example, to move a pci device under kernel management:
1231 .. code-block:: console
1233 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1235 ./usertools/dpdk-devbind.py --status
1237 Network devices using DPDK-compatible driver
1238 ============================================
1241 Network devices using kernel driver
1242 ===================================
1243 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1245 To remove a port created by a virtual device, above steps are not needed.
1250 Start all ports or a specific port::
1252 testpmd> port start (port_id|all)
1257 Stop all ports or a specific port::
1259 testpmd> port stop (port_id|all)
1264 Close all ports or a specific port::
1266 testpmd> port close (port_id|all)
1268 port start/stop queue
1269 ~~~~~~~~~~~~~~~~~~~~~
1271 Start/stop a rx/tx queue on a specific port::
1273 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1275 Only take effect when port is started.
1280 Set the speed and duplex mode for all ports or a specific port::
1282 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1283 duplex (half|full|auto)
1285 port config - queues/descriptors
1286 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1288 Set number of queues/descriptors for rxq, txq, rxd and txd::
1290 testpmd> port config all (rxq|txq|rxd|txd) (value)
1292 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1294 port config - max-pkt-len
1295 ~~~~~~~~~~~~~~~~~~~~~~~~~
1297 Set the maximum packet length::
1299 testpmd> port config all max-pkt-len (value)
1301 This is equivalent to the ``--max-pkt-len`` command-line option.
1303 port config - CRC Strip
1304 ~~~~~~~~~~~~~~~~~~~~~~~
1306 Set hardware CRC stripping on or off for all ports::
1308 testpmd> port config all crc-strip (on|off)
1310 CRC stripping is off by default.
1312 The ``on`` option is equivalent to the ``--crc-strip`` command-line option.
1314 port config - scatter
1315 ~~~~~~~~~~~~~~~~~~~~~~~
1317 Set RX scatter mode on or off for all ports::
1319 testpmd> port config all scatter (on|off)
1321 RX scatter mode is off by default.
1323 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1325 port config - TX queue flags
1326 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1328 Set a hexadecimal bitmap of TX queue flags for all ports::
1330 testpmd> port config all txqflags value
1332 This command is equivalent to the ``--txqflags`` command-line option.
1334 port config - RX Checksum
1335 ~~~~~~~~~~~~~~~~~~~~~~~~~
1337 Set hardware RX checksum offload to on or off for all ports::
1339 testpmd> port config all rx-cksum (on|off)
1341 Checksum offload is off by default.
1343 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1348 Set hardware VLAN on or off for all ports::
1350 testpmd> port config all hw-vlan (on|off)
1352 Hardware VLAN is on by default.
1354 The ``off`` option is equivalent to the ``--disable-hw-vlan`` command-line option.
1356 port config - VLAN filter
1357 ~~~~~~~~~~~~~~~~~~~~~~~~~
1359 Set hardware VLAN filter on or off for all ports::
1361 testpmd> port config all hw-vlan-filter (on|off)
1363 Hardware VLAN filter is on by default.
1365 The ``off`` option is equivalent to the ``--disable-hw-vlan-filter`` command-line option.
1367 port config - VLAN strip
1368 ~~~~~~~~~~~~~~~~~~~~~~~~
1370 Set hardware VLAN strip on or off for all ports::
1372 testpmd> port config all hw-vlan-strip (on|off)
1374 Hardware VLAN strip is on by default.
1376 The ``off`` option is equivalent to the ``--disable-hw-vlan-strip`` command-line option.
1378 port config - VLAN extend
1379 ~~~~~~~~~~~~~~~~~~~~~~~~~
1381 Set hardware VLAN extend on or off for all ports::
1383 testpmd> port config all hw-vlan-extend (on|off)
1385 Hardware VLAN extend is off by default.
1387 The ``off`` option is equivalent to the ``--disable-hw-vlan-extend`` command-line option.
1389 port config - Drop Packets
1390 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1392 Set packet drop for packets with no descriptors on or off for all ports::
1394 testpmd> port config all drop-en (on|off)
1396 Packet dropping for packets with no descriptors is off by default.
1398 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1403 Set the RSS (Receive Side Scaling) mode on or off::
1405 testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1407 RSS is on by default.
1409 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1411 port config - RSS Reta
1412 ~~~~~~~~~~~~~~~~~~~~~~
1414 Set the RSS (Receive Side Scaling) redirection table::
1416 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1421 Set the DCB mode for an individual port::
1423 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1425 The traffic class should be 4 or 8.
1430 Set the number of packets per burst::
1432 testpmd> port config all burst (value)
1434 This is equivalent to the ``--burst`` command-line option.
1436 port config - Threshold
1437 ~~~~~~~~~~~~~~~~~~~~~~~
1439 Set thresholds for TX/RX queues::
1441 testpmd> port config all (threshold) (value)
1443 Where the threshold type can be:
1445 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1447 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1449 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1451 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1453 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1455 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1457 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1459 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1461 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1463 These threshold options are also available from the command-line.
1468 Set the value of ether-type for E-tag::
1470 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1472 Enable/disable the E-tag support::
1474 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1477 Link Bonding Functions
1478 ----------------------
1480 The Link Bonding functions make it possible to dynamically create and
1481 manage link bonding devices from within testpmd interactive prompt.
1483 create bonded device
1484 ~~~~~~~~~~~~~~~~~~~~
1486 Create a new bonding device::
1488 testpmd> create bonded device (mode) (socket)
1490 For example, to create a bonded device in mode 1 on socket 0::
1492 testpmd> create bonded 1 0
1493 created new bonded device (port X)
1498 Adds Ethernet device to a Link Bonding device::
1500 testpmd> add bonding slave (slave id) (port id)
1502 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
1504 testpmd> add bonding slave 6 10
1507 remove bonding slave
1508 ~~~~~~~~~~~~~~~~~~~~
1510 Removes an Ethernet slave device from a Link Bonding device::
1512 testpmd> remove bonding slave (slave id) (port id)
1514 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
1516 testpmd> remove bonding slave 6 10
1521 Set the Link Bonding mode of a Link Bonding device::
1523 testpmd> set bonding mode (value) (port id)
1525 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
1527 testpmd> set bonding mode 3 10
1532 Set an Ethernet slave device as the primary device on a Link Bonding device::
1534 testpmd> set bonding primary (slave id) (port id)
1536 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
1538 testpmd> set bonding primary 6 10
1543 Set the MAC address of a Link Bonding device::
1545 testpmd> set bonding mac (port id) (mac)
1547 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
1549 testpmd> set bonding mac 10 00:00:00:00:00:01
1551 set bonding xmit_balance_policy
1552 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1554 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
1556 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
1558 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
1560 testpmd> set bonding xmit_balance_policy 10 l34
1563 set bonding mon_period
1564 ~~~~~~~~~~~~~~~~~~~~~~
1566 Set the link status monitoring polling period in milliseconds for a bonding device.
1568 This adds support for PMD slave devices which do not support link status interrupts.
1569 When the mon_period is set to a value greater than 0 then all PMD's which do not support
1570 link status ISR will be queried every polling interval to check if their link status has changed::
1572 testpmd> set bonding mon_period (port_id) (value)
1574 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
1576 testpmd> set bonding mon_period 5 150
1582 Show the current configuration of a Link Bonding device::
1584 testpmd> show bonding config (port id)
1587 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
1588 in balance mode with a transmission policy of layer 2+3::
1590 testpmd> show bonding config 9
1592 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
1594 Active Slaves (3): [1 3 4]
1601 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
1602 This is mainly useful for debugging purposes.
1603 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
1604 and fields that can be accessed.
1609 Display the value of a port register::
1611 testpmd> read reg (port_id) (address)
1613 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
1615 testpmd> read reg 0 0xEE00
1616 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
1621 Display a port register bit field::
1623 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
1625 For example, reading the lowest two bits from the register in the example above::
1627 testpmd> read regfield 0 0xEE00 0 1
1628 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
1633 Display a single port register bit::
1635 testpmd> read regbit (port_id) (address) (bit_x)
1637 For example, reading the lowest bit from the register in the example above::
1639 testpmd> read regbit 0 0xEE00 0
1640 port 0 PCI register at offset 0xEE00: bit 0=1
1645 Set the value of a port register::
1647 testpmd> write reg (port_id) (address) (value)
1649 For example, to clear a register::
1651 testpmd> write reg 0 0xEE00 0x0
1652 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
1657 Set bit field of a port register::
1659 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
1661 For example, writing to the register cleared in the example above::
1663 testpmd> write regfield 0 0xEE00 0 1 2
1664 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
1669 Set single bit value of a port register::
1671 testpmd> write regbit (port_id) (address) (bit_x) (value)
1673 For example, to set the high bit in the register from the example above::
1675 testpmd> write regbit 0 0xEE00 31 1
1676 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
1682 This section details the available filter functions that are available.
1684 Note these functions interface the deprecated legacy filtering framework,
1685 superseded by *rte_flow*. See `Flow rules management`_.
1688 ~~~~~~~~~~~~~~~~~~~~
1690 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
1692 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
1693 ethertype (ether_type) (drop|fwd) queue (queue_id)
1695 The available information parameters are:
1697 * ``port_id``: The port which the Ethertype filter assigned on.
1699 * ``mac_addr``: Compare destination mac address.
1701 * ``mac_ignr``: Ignore destination mac address match.
1703 * ``mac_address``: Destination mac address to match.
1705 * ``ether_type``: The EtherType value want to match,
1706 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
1708 * ``queue_id``: The receive queue associated with this EtherType filter.
1709 It is meaningless when deleting or dropping.
1711 Example, to add/remove an ethertype filter rule::
1713 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
1714 ethertype 0x0806 fwd queue 3
1716 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
1717 ethertype 0x0806 fwd queue 3
1722 Add or delete a 2-tuple filter,
1723 which identifies packets by specific protocol and destination TCP/UDP port
1724 and forwards packets into one of the receive queues::
1726 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
1727 protocol (protocol_value) mask (mask_value) \
1728 tcp_flags (tcp_flags_value) priority (prio_value) \
1731 The available information parameters are:
1733 * ``port_id``: The port which the 2-tuple filter assigned on.
1735 * ``dst_port_value``: Destination port in L4.
1737 * ``protocol_value``: IP L4 protocol.
1739 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
1741 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
1743 * ``prio_value``: Priority of this filter.
1745 * ``queue_id``: The receive queue associated with this 2-tuple filter.
1747 Example, to add/remove an 2tuple filter rule::
1749 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
1750 tcp_flags 0x02 priority 3 queue 3
1752 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
1753 tcp_flags 0x02 priority 3 queue 3
1758 Add or delete a 5-tuple filter,
1759 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
1760 and routes packets into one of the receive queues::
1762 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
1763 (src_address) dst_port (dst_port_value) \
1764 src_port (src_port_value) protocol (protocol_value) \
1765 mask (mask_value) tcp_flags (tcp_flags_value) \
1766 priority (prio_value) queue (queue_id)
1768 The available information parameters are:
1770 * ``port_id``: The port which the 5-tuple filter assigned on.
1772 * ``dst_address``: Destination IP address.
1774 * ``src_address``: Source IP address.
1776 * ``dst_port_value``: TCP/UDP destination port.
1778 * ``src_port_value``: TCP/UDP source port.
1780 * ``protocol_value``: L4 protocol.
1782 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
1784 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
1786 * ``prio_value``: The priority of this filter.
1788 * ``queue_id``: The receive queue associated with this 5-tuple filter.
1790 Example, to add/remove an 5tuple filter rule::
1792 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1793 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1794 flags 0x0 priority 3 queue 3
1796 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1797 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1798 flags 0x0 priority 3 queue 3
1803 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
1805 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
1807 The available information parameters are:
1809 * ``port_id``: The port which the SYN filter assigned on.
1811 * ``high``: This SYN filter has higher priority than other filters.
1813 * ``low``: This SYN filter has lower priority than other filters.
1815 * ``queue_id``: The receive queue associated with this SYN filter
1819 testpmd> syn_filter 0 add priority high queue 3
1824 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
1825 and routed into one of the receive queues::
1827 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
1828 mask (mask_value) priority (prio_value) queue (queue_id)
1830 The available information parameters are:
1832 * ``port_id``: The port which the Flex filter is assigned on.
1834 * ``len_value``: Filter length in bytes, no greater than 128.
1836 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
1838 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
1840 * ``prio_value``: The priority of this filter.
1842 * ``queue_id``: The receive queue associated with this Flex filter.
1846 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
1847 mask 000C priority 3 queue 3
1849 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
1850 mask 000C priority 3 queue 3
1853 .. _testpmd_flow_director:
1855 flow_director_filter
1856 ~~~~~~~~~~~~~~~~~~~~
1858 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
1860 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
1861 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
1863 * Perfect match filters.
1864 The hardware checks a match between the masked fields of the received packets and the programmed filters.
1865 The masked fields are for IP flow.
1867 * Signature filters.
1868 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
1870 * Perfect-mac-vlan match filters.
1871 The hardware checks a match between the masked fields of the received packets and the programmed filters.
1872 The masked fields are for MAC VLAN flow.
1874 * Perfect-tunnel match filters.
1875 The hardware checks a match between the masked fields of the received packets and the programmed filters.
1876 The masked fields are for tunnel flow.
1878 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
1879 per flow type and the flexible payload.
1881 The Flow Director can also mask out parts of all of these fields so that filters
1882 are only applied to certain fields or parts of the fields.
1884 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
1886 # Commands to add flow director filters of different flow types::
1888 flow_director_filter (port_id) mode IP (add|del|update) \
1889 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
1890 src (src_ip_address) dst (dst_ip_address) \
1891 tos (tos_value) proto (proto_value) ttl (ttl_value) \
1892 vlan (vlan_value) flexbytes (flexbytes_value) \
1893 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
1896 flow_director_filter (port_id) mode IP (add|del|update) \
1897 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
1898 src (src_ip_address) (src_port) \
1899 dst (dst_ip_address) (dst_port) \
1900 tos (tos_value) ttl (ttl_value) \
1901 vlan (vlan_value) flexbytes (flexbytes_value) \
1902 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
1905 flow_director_filter (port_id) mode IP (add|del|update) \
1906 flow (ipv4-sctp|ipv6-sctp) \
1907 src (src_ip_address) (src_port) \
1908 dst (dst_ip_address) (dst_port) \
1909 tos (tos_value) ttl (ttl_value) \
1910 tag (verification_tag) vlan (vlan_value) \
1911 flexbytes (flexbytes_value) (drop|fwd) \
1912 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
1914 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
1915 ether (ethertype) flexbytes (flexbytes_value) \
1916 (drop|fwd) pf|vf(vf_id) queue (queue_id)
1919 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
1920 mac (mac_address) vlan (vlan_value) \
1921 flexbytes (flexbytes_value) (drop|fwd) \
1922 queue (queue_id) fd_id (fd_id_value)
1924 flow_director_filter (port_id) mode Tunnel (add|del|update) \
1925 mac (mac_address) vlan (vlan_value) \
1926 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
1927 flexbytes (flexbytes_value) (drop|fwd) \
1928 queue (queue_id) fd_id (fd_id_value)
1930 For example, to add an ipv4-udp flow type filter::
1932 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
1933 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
1934 fwd pf queue 1 fd_id 1
1936 For example, add an ipv4-other flow type filter::
1938 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
1939 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
1940 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
1945 Flush all flow director filters on a device::
1947 testpmd> flush_flow_director (port_id)
1949 Example, to flush all flow director filter on port 0::
1951 testpmd> flush_flow_director 0
1956 Set flow director's input masks::
1958 flow_director_mask (port_id) mode IP vlan (vlan_value) \
1959 src_mask (ipv4_src) (ipv6_src) (src_port) \
1960 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
1962 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
1964 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
1965 mac (mac_value) tunnel-type (tunnel_type_value) \
1966 tunnel-id (tunnel_id_value)
1968 Example, to set flow director mask on port 0::
1970 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
1971 src_mask 255.255.255.255 \
1972 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
1973 dst_mask 255.255.255.255 \
1974 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
1976 flow_director_flex_mask
1977 ~~~~~~~~~~~~~~~~~~~~~~~
1979 set masks of flow director's flexible payload based on certain flow type::
1981 testpmd> flow_director_flex_mask (port_id) \
1982 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
1983 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
1984 l2_payload|all) (mask)
1986 Example, to set flow director's flex mask for all flow type on port 0::
1988 testpmd> flow_director_flex_mask 0 flow all \
1989 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
1992 flow_director_flex_payload
1993 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1995 Configure flexible payload selection::
1997 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
1999 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2001 testpmd> flow_director_flex_payload 0 l4 \
2002 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2004 get_sym_hash_ena_per_port
2005 ~~~~~~~~~~~~~~~~~~~~~~~~~
2007 Get symmetric hash enable configuration per port::
2009 get_sym_hash_ena_per_port (port_id)
2011 For example, to get symmetric hash enable configuration of port 1::
2013 testpmd> get_sym_hash_ena_per_port 1
2015 set_sym_hash_ena_per_port
2016 ~~~~~~~~~~~~~~~~~~~~~~~~~
2018 Set symmetric hash enable configuration per port to enable or disable::
2020 set_sym_hash_ena_per_port (port_id) (enable|disable)
2022 For example, to set symmetric hash enable configuration of port 1 to enable::
2024 testpmd> set_sym_hash_ena_per_port 1 enable
2026 get_hash_global_config
2027 ~~~~~~~~~~~~~~~~~~~~~~
2029 Get the global configurations of hash filters::
2031 get_hash_global_config (port_id)
2033 For example, to get the global configurations of hash filters of port 1::
2035 testpmd> get_hash_global_config 1
2037 set_hash_global_config
2038 ~~~~~~~~~~~~~~~~~~~~~~
2040 Set the global configurations of hash filters::
2042 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2043 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2044 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload) \
2047 For example, to enable simple_xor for flow type of ipv6 on port 2::
2049 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2054 Set the input set for hash::
2056 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2057 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2058 l2_payload) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2059 ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2060 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2061 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2062 fld-8th|none) (select|add)
2064 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2066 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2071 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2072 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2074 Set the input set for flow director::
2076 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2077 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2078 l2_payload) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2079 ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2080 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2081 sctp-dst-port|sctp-veri-tag|none) (select|add)
2083 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2085 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2090 Set different GRE key length for input set::
2092 global_config (port_id) gre-key-len (number in bytes)
2094 For example to set GRE key length for input set to 4 bytes on port 0::
2096 testpmd> global_config 0 gre-key-len 4
2099 .. _testpmd_rte_flow:
2101 Flow rules management
2102 ---------------------
2104 Control of the generic flow API (*rte_flow*) is fully exposed through the
2105 ``flow`` command (validation, creation, destruction and queries).
2107 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2108 features simultaneously may cause undefined side-effects and is therefore
2114 Because the ``flow`` command uses dynamic tokens to handle the large number
2115 of possible flow rules combinations, its behavior differs slightly from
2116 other commands, in particular:
2118 - Pressing *?* or the *<tab>* key displays contextual help for the current
2119 token, not that of the entire command.
2121 - Optional and repeated parameters are supported (provided they are listed
2122 in the contextual help).
2124 The first parameter stands for the operation mode. Possible operations and
2125 their general syntax are described below. They are covered in detail in the
2128 - Check whether a flow rule can be created::
2130 flow validate {port_id}
2131 [group {group_id}] [priority {level}] [ingress] [egress]
2132 pattern {item} [/ {item} [...]] / end
2133 actions {action} [/ {action} [...]] / end
2135 - Create a flow rule::
2137 flow create {port_id}
2138 [group {group_id}] [priority {level}] [ingress] [egress]
2139 pattern {item} [/ {item} [...]] / end
2140 actions {action} [/ {action} [...]] / end
2142 - Destroy specific flow rules::
2144 flow destroy {port_id} rule {rule_id} [...]
2146 - Destroy all flow rules::
2148 flow flush {port_id}
2150 - Query an existing flow rule::
2152 flow query {port_id} {rule_id} {action}
2154 - List existing flow rules sorted by priority, filtered by group
2157 flow list {port_id} [group {group_id}] [...]
2159 Validating flow rules
2160 ~~~~~~~~~~~~~~~~~~~~~
2162 ``flow validate`` reports whether a flow rule would be accepted by the
2163 underlying device in its current state but stops short of creating it. It is
2164 bound to ``rte_flow_validate()``::
2166 flow validate {port_id}
2167 [group {group_id}] [priority {level}] [ingress] [egress]
2168 pattern {item} [/ {item} [...]] / end
2169 actions {action} [/ {action} [...]] / end
2171 If successful, it will show::
2175 Otherwise it will show an error message of the form::
2177 Caught error type [...] ([...]): [...]
2179 This command uses the same parameters as ``flow create``, their format is
2180 described in `Creating flow rules`_.
2182 Check whether redirecting any Ethernet packet received on port 0 to RX queue
2183 index 6 is supported::
2185 testpmd> flow validate 0 ingress pattern eth / end
2186 actions queue index 6 / end
2190 Port 0 does not support TCPv6 rules::
2192 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
2194 Caught error type 9 (specific pattern item): Invalid argument
2200 ``flow create`` validates and creates the specified flow rule. It is bound
2201 to ``rte_flow_create()``::
2203 flow create {port_id}
2204 [group {group_id}] [priority {level}] [ingress] [egress]
2205 pattern {item} [/ {item} [...]] / end
2206 actions {action} [/ {action} [...]] / end
2208 If successful, it will return a flow rule ID usable with other commands::
2210 Flow rule #[...] created
2212 Otherwise it will show an error message of the form::
2214 Caught error type [...] ([...]): [...]
2216 Parameters describe in the following order:
2218 - Attributes (*group*, *priority*, *ingress*, *egress* tokens).
2219 - A matching pattern, starting with the *pattern* token and terminated by an
2221 - Actions, starting with the *actions* token and terminated by an *end*
2224 These translate directly to *rte_flow* objects provided as-is to the
2225 underlying functions.
2227 The shortest valid definition only comprises mandatory tokens::
2229 testpmd> flow create 0 pattern end actions end
2231 Note that PMDs may refuse rules that essentially do nothing such as this
2234 **All unspecified object values are automatically initialized to 0.**
2239 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
2240 specified before the ``pattern`` token.
2242 - ``group {group id}``: priority group.
2243 - ``priority {level}``: priority level within group.
2244 - ``ingress``: rule applies to ingress traffic.
2245 - ``egress``: rule applies to egress traffic.
2247 Each instance of an attribute specified several times overrides the previous
2248 value as shown below (group 4 is used)::
2250 testpmd> flow create 0 group 42 group 24 group 4 [...]
2252 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
2254 While not specifying a direction is an error, some rules may allow both
2257 Most rules affect RX therefore contain the ``ingress`` token::
2259 testpmd> flow create 0 ingress pattern [...]
2264 A matching pattern starts after the ``pattern`` token. It is made of pattern
2265 items and is terminated by a mandatory ``end`` item.
2267 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
2268 rte_flow_item_type``).
2270 The ``/`` token is used as a separator between pattern items as shown
2273 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
2275 Note that protocol items like these must be stacked from lowest to highest
2276 layer to make sense. For instance, the following rule is either invalid or
2277 unlikely to match any packet::
2279 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
2281 More information on these restrictions can be found in the *rte_flow*
2284 Several items support additional specification structures, for example
2285 ``ipv4`` allows specifying source and destination addresses as follows::
2287 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
2288 dst is 10.2.0.0 / end [...]
2290 This rule matches all IPv4 traffic with the specified properties.
2292 In this example, ``src`` and ``dst`` are field names of the underlying
2293 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
2294 in a similar fashion.
2296 The ``is`` token means that the subsequent value must be matched exactly,
2297 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
2298 accordingly. Possible assignment tokens are:
2300 - ``is``: match value perfectly (with full bit-mask).
2301 - ``spec``: match value according to configured bit-mask.
2302 - ``last``: specify upper bound to establish a range.
2303 - ``mask``: specify bit-mask with relevant bits set to one.
2304 - ``prefix``: generate bit-mask from a prefix length.
2306 These yield identical results::
2308 ipv4 src is 10.1.1.1
2312 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
2316 ipv4 src spec 10.1.1.1 src prefix 32
2320 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
2324 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
2326 Inclusive ranges can be defined with ``last``::
2328 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
2330 Note that ``mask`` affects both ``spec`` and ``last``::
2332 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
2333 # matches 10.1.0.0 to 10.2.255.255
2335 Properties can be modified multiple times::
2337 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
2341 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
2346 This section lists supported pattern items and their attributes, if any.
2348 - ``end``: end list of pattern items.
2350 - ``void``: no-op pattern item.
2352 - ``invert``: perform actions when pattern does not match.
2354 - ``any``: match any protocol for the current layer.
2356 - ``num {unsigned}``: number of layers covered.
2358 - ``pf``: match packets addressed to the physical function.
2360 - ``vf``: match packets addressed to a virtual function ID.
2362 - ``id {unsigned}``: destination VF ID.
2364 - ``port``: device-specific physical port index to use.
2366 - ``index {unsigned}``: physical port index.
2368 - ``raw``: match an arbitrary byte string.
2370 - ``relative {boolean}``: look for pattern after the previous item.
2371 - ``search {boolean}``: search pattern from offset (see also limit).
2372 - ``offset {integer}``: absolute or relative offset for pattern.
2373 - ``limit {unsigned}``: search area limit for start of pattern.
2374 - ``pattern {string}``: byte string to look for.
2376 - ``eth``: match Ethernet header.
2378 - ``dst {MAC-48}``: destination MAC.
2379 - ``src {MAC-48}``: source MAC.
2380 - ``type {unsigned}``: EtherType.
2382 - ``vlan``: match 802.1Q/ad VLAN tag.
2384 - ``tpid {unsigned}``: tag protocol identifier.
2385 - ``tci {unsigned}``: tag control information.
2386 - ``pcp {unsigned}``: priority code point.
2387 - ``dei {unsigned}``: drop eligible indicator.
2388 - ``vid {unsigned}``: VLAN identifier.
2390 - ``ipv4``: match IPv4 header.
2392 - ``tos {unsigned}``: type of service.
2393 - ``ttl {unsigned}``: time to live.
2394 - ``proto {unsigned}``: next protocol ID.
2395 - ``src {ipv4 address}``: source address.
2396 - ``dst {ipv4 address}``: destination address.
2398 - ``ipv6``: match IPv6 header.
2400 - ``tc {unsigned}``: traffic class.
2401 - ``flow {unsigned}``: flow label.
2402 - ``proto {unsigned}``: protocol (next header).
2403 - ``hop {unsigned}``: hop limit.
2404 - ``src {ipv6 address}``: source address.
2405 - ``dst {ipv6 address}``: destination address.
2407 - ``icmp``: match ICMP header.
2409 - ``type {unsigned}``: ICMP packet type.
2410 - ``code {unsigned}``: ICMP packet code.
2412 - ``udp``: match UDP header.
2414 - ``src {unsigned}``: UDP source port.
2415 - ``dst {unsigned}``: UDP destination port.
2417 - ``tcp``: match TCP header.
2419 - ``src {unsigned}``: TCP source port.
2420 - ``dst {unsigned}``: TCP destination port.
2422 - ``sctp``: match SCTP header.
2424 - ``src {unsigned}``: SCTP source port.
2425 - ``dst {unsigned}``: SCTP destination port.
2426 - ``tag {unsigned}``: validation tag.
2427 - ``cksum {unsigned}``: checksum.
2429 - ``vxlan``: match VXLAN header.
2431 - ``vni {unsigned}``: VXLAN identifier.
2436 A list of actions starts after the ``actions`` token in the same fashion as
2437 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
2438 terminated by a mandatory ``end`` action.
2440 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
2441 rte_flow_action_type``).
2443 Dropping all incoming UDPv4 packets can be expressed as follows::
2445 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2448 Several actions have configurable properties which must be specified when
2449 there is no valid default value. For example, ``queue`` requires a target
2452 This rule redirects incoming UDPv4 traffic to queue index 6::
2454 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2455 actions queue index 6 / end
2457 While this one could be rejected by PMDs (unspecified queue index)::
2459 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2462 As defined by *rte_flow*, the list is not ordered, all actions of a given
2463 rule are performed simultaneously. These are equivalent::
2465 queue index 6 / void / mark id 42 / end
2469 void / mark id 42 / queue index 6 / end
2471 All actions in a list should have different types, otherwise only the last
2472 action of a given type is taken into account::
2474 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
2478 drop / drop / drop / end # drop is performed only once
2482 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
2484 Considering they are performed simultaneously, opposite and overlapping
2485 actions can sometimes be combined when the end result is unambiguous::
2487 drop / queue index 6 / end # drop has no effect
2491 drop / dup index 6 / end # same as above
2495 queue index 6 / rss queues 6 7 8 / end # queue has no effect
2499 drop / passthru / end # drop has no effect
2501 Note that PMDs may still refuse such combinations.
2506 This section lists supported actions and their attributes, if any.
2508 - ``end``: end list of actions.
2510 - ``void``: no-op action.
2512 - ``passthru``: let subsequent rule process matched packets.
2514 - ``mark``: attach 32 bit value to packets.
2516 - ``id {unsigned}``: 32 bit value to return with packets.
2518 - ``flag``: flag packets.
2520 - ``queue``: assign packets to a given queue index.
2522 - ``index {unsigned}``: queue index to use.
2524 - ``drop``: drop packets (note: passthru has priority).
2526 - ``count``: enable counters for this rule.
2528 - ``dup``: duplicate packets to a given queue index.
2530 - ``index {unsigned}``: queue index to duplicate packets to.
2532 - ``rss``: spread packets among several queues.
2534 - ``queues [{unsigned} [...]] end``: queue indices to use.
2536 - ``pf``: redirect packets to physical device function.
2538 - ``vf``: redirect packets to virtual device function.
2540 - ``original {boolean}``: use original VF ID if possible.
2541 - ``id {unsigned}``: VF ID to redirect packets to.
2543 Destroying flow rules
2544 ~~~~~~~~~~~~~~~~~~~~~
2546 ``flow destroy`` destroys one or more rules from their rule ID (as returned
2547 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
2548 times as necessary::
2550 flow destroy {port_id} rule {rule_id} [...]
2552 If successful, it will show::
2554 Flow rule #[...] destroyed
2556 It does not report anything for rule IDs that do not exist. The usual error
2557 message is shown when a rule cannot be destroyed::
2559 Caught error type [...] ([...]): [...]
2561 ``flow flush`` destroys all rules on a device and does not take extra
2562 arguments. It is bound to ``rte_flow_flush()``::
2564 flow flush {port_id}
2566 Any errors are reported as above.
2568 Creating several rules and destroying them::
2570 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2571 actions queue index 2 / end
2572 Flow rule #0 created
2573 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2574 actions queue index 3 / end
2575 Flow rule #1 created
2576 testpmd> flow destroy 0 rule 0 rule 1
2577 Flow rule #1 destroyed
2578 Flow rule #0 destroyed
2581 The same result can be achieved using ``flow flush``::
2583 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2584 actions queue index 2 / end
2585 Flow rule #0 created
2586 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2587 actions queue index 3 / end
2588 Flow rule #1 created
2589 testpmd> flow flush 0
2592 Non-existent rule IDs are ignored::
2594 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2595 actions queue index 2 / end
2596 Flow rule #0 created
2597 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2598 actions queue index 3 / end
2599 Flow rule #1 created
2600 testpmd> flow destroy 0 rule 42 rule 10 rule 2
2602 testpmd> flow destroy 0 rule 0
2603 Flow rule #0 destroyed
2609 ``flow query`` queries a specific action of a flow rule having that
2610 ability. Such actions collect information that can be reported using this
2611 command. It is bound to ``rte_flow_query()``::
2613 flow query {port_id} {rule_id} {action}
2615 If successful, it will display either the retrieved data for known actions
2616 or the following message::
2618 Cannot display result for action type [...] ([...])
2620 Otherwise, it will complain either that the rule does not exist or that some
2623 Flow rule #[...] not found
2627 Caught error type [...] ([...]): [...]
2629 Currently only the ``count`` action is supported. This action reports the
2630 number of packets that hit the flow rule and the total number of bytes. Its
2631 output has the following format::
2634 hits_set: [...] # whether "hits" contains a valid value
2635 bytes_set: [...] # whether "bytes" contains a valid value
2636 hits: [...] # number of packets
2637 bytes: [...] # number of bytes
2639 Querying counters for TCPv6 packets redirected to queue 6::
2641 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
2642 actions queue index 6 / count / end
2643 Flow rule #4 created
2644 testpmd> flow query 0 4 count
2655 ``flow list`` lists existing flow rules sorted by priority and optionally
2656 filtered by group identifiers::
2658 flow list {port_id} [group {group_id}] [...]
2660 This command only fails with the following message if the device does not
2665 Output consists of a header line followed by a short description of each
2666 flow rule, one per line. There is no output at all when no flow rules are
2667 configured on the device::
2669 ID Group Prio Attr Rule
2670 [...] [...] [...] [...] [...]
2672 ``Attr`` column flags:
2674 - ``i`` for ``ingress``.
2675 - ``e`` for ``egress``.
2677 Creating several flow rules and listing them::
2679 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2680 actions queue index 6 / end
2681 Flow rule #0 created
2682 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2683 actions queue index 2 / end
2684 Flow rule #1 created
2685 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
2686 actions rss queues 6 7 8 end / end
2687 Flow rule #2 created
2688 testpmd> flow list 0
2689 ID Group Prio Attr Rule
2690 0 0 0 i- ETH IPV4 => QUEUE
2691 1 0 0 i- ETH IPV6 => QUEUE
2692 2 0 5 i- ETH IPV4 UDP => RSS
2695 Rules are sorted by priority (i.e. group ID first, then priority level)::
2697 testpmd> flow list 1
2698 ID Group Prio Attr Rule
2699 0 0 0 i- ETH => COUNT
2700 6 0 500 i- ETH IPV6 TCP => DROP COUNT
2701 5 0 1000 i- ETH IPV6 ICMP => QUEUE
2702 1 24 0 i- ETH IPV4 UDP => QUEUE
2703 4 24 10 i- ETH IPV4 TCP => DROP
2704 3 24 20 i- ETH IPV4 => DROP
2705 2 24 42 i- ETH IPV4 UDP => QUEUE
2706 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
2709 Output can be limited to specific groups::
2711 testpmd> flow list 1 group 0 group 63
2712 ID Group Prio Attr Rule
2713 0 0 0 i- ETH => COUNT
2714 6 0 500 i- ETH IPV6 TCP => DROP COUNT
2715 5 0 1000 i- ETH IPV6 ICMP => QUEUE
2716 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE