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)
867 Set the link flow control parameter on a port::
869 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
870 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
871 autoneg (on|off) (port_id)
875 * ``high_water`` (integer): High threshold value to trigger XOFF.
877 * ``low_water`` (integer): Low threshold value to trigger XON.
879 * ``pause_time`` (integer): Pause quota in the Pause frame.
881 * ``send_xon`` (0/1): Send XON frame.
883 * ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
885 * ``autoneg``: Change the auto-negotiation parameter.
890 Set the priority flow control parameter on a port::
892 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
893 (pause_time) (priority) (port_id)
897 * ``high_water`` (integer): High threshold value.
899 * ``low_water`` (integer): Low threshold value.
901 * ``pause_time`` (integer): Pause quota in the Pause frame.
903 * ``priority`` (0-7): VLAN User Priority.
908 Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
910 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
912 For example, to set rx queue 2 on port 0 to mapping 5::
914 testpmd>set stat_qmap rx 0 2 5
916 set port - rx/tx (for VF)
917 ~~~~~~~~~~~~~~~~~~~~~~~~~
919 Set VF receive/transmit from a port::
921 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
923 set port - mac address filter (for VF)
924 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
926 Add/Remove unicast or multicast MAC addr filter for a VF::
928 testpmd> set port (port_id) vf (vf_id) (mac_addr) \
929 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
931 set port - rx mode(for VF)
932 ~~~~~~~~~~~~~~~~~~~~~~~~~~
934 Set the VF receive mode of a port::
936 testpmd> set port (port_id) vf (vf_id) \
937 rxmode (AUPE|ROPE|BAM|MPE) (on|off)
939 The available receive modes are:
941 * ``AUPE``: Accepts untagged VLAN.
943 * ``ROPE``: Accepts unicast hash.
945 * ``BAM``: Accepts broadcast packets.
947 * ``MPE``: Accepts all multicast packets.
949 set port - tx_rate (for Queue)
950 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
952 Set TX rate limitation for a queue on a port::
954 testpmd> set port (port_id) queue (queue_id) rate (rate_value)
956 set port - tx_rate (for VF)
957 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
959 Set TX rate limitation for queues in VF on a port::
961 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
963 set port - mirror rule
964 ~~~~~~~~~~~~~~~~~~~~~~
966 Set pool or vlan type mirror rule for a port::
968 testpmd> set port (port_id) mirror-rule (rule_id) \
969 (pool-mirror-up|pool-mirror-down|vlan-mirror) \
970 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
972 Set link mirror rule for a port::
974 testpmd> set port (port_id) mirror-rule (rule_id) \
975 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
977 For example to enable mirror traffic with vlan 0,1 to pool 0::
979 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
981 reset port - mirror rule
982 ~~~~~~~~~~~~~~~~~~~~~~~~
984 Reset a mirror rule for a port::
986 testpmd> reset port (port_id) mirror-rule (rule_id)
991 Set the flush on RX streams before forwarding.
992 The default is flush ``on``.
993 Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
995 testpmd> set flush_rx off
1000 Set the bypass mode for the lowest port on bypass enabled NIC::
1002 testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1007 Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1009 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1010 mode (normal|bypass|isolate) (port_id)
1014 * ``timeout``: Enable bypass after watchdog timeout.
1016 * ``os_on``: Enable bypass when OS/board is powered on.
1018 * ``os_off``: Enable bypass when OS/board is powered off.
1020 * ``power_on``: Enable bypass when power supply is turned on.
1022 * ``power_off``: Enable bypass when power supply is turned off.
1028 Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1030 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1035 Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1037 testpmd> show bypass config (port_id)
1042 Set link up for a port::
1044 testpmd> set link-up port (port id)
1049 Set link down for a port::
1051 testpmd> set link-down port (port id)
1056 Enable E-tag insertion for a VF on a port::
1058 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1060 Disable E-tag insertion for a VF on a port::
1062 testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1064 Enable/disable E-tag stripping on a port::
1066 testpmd> E-tag set stripping (on|off) port (port_id)
1068 Enable/disable E-tag based forwarding on a port::
1070 testpmd> E-tag set forwarding (on|off) port (port_id)
1072 Add an E-tag forwarding filter on a port::
1074 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1076 Delete an E-tag forwarding filter on a port::
1077 testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1083 The following sections show functions for configuring ports.
1087 Port configuration changes only become active when forwarding is started/restarted.
1092 Attach a port specified by pci address or virtual device args::
1094 testpmd> port attach (identifier)
1096 To attach a new pci device, the device should be recognized by kernel first.
1097 Then it should be moved under DPDK management.
1098 Finally the port can be attached to testpmd.
1100 For example, to move a pci device using ixgbe under DPDK management:
1102 .. code-block:: console
1104 # Check the status of the available devices.
1105 ./usertools/dpdk-devbind.py --status
1107 Network devices using DPDK-compatible driver
1108 ============================================
1111 Network devices using kernel driver
1112 ===================================
1113 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1116 # Bind the device to igb_uio.
1117 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1120 # Recheck the status of the devices.
1121 ./usertools/dpdk-devbind.py --status
1122 Network devices using DPDK-compatible driver
1123 ============================================
1124 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1126 To attach a port created by virtual device, above steps are not needed.
1128 For example, to attach a port whose pci address is 0000:0a:00.0.
1130 .. code-block:: console
1132 testpmd> port attach 0000:0a:00.0
1133 Attaching a new port...
1134 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1135 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1136 EAL: PCI memory mapped at 0x7f83bfa00000
1137 EAL: PCI memory mapped at 0x7f83bfa80000
1138 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1139 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1140 Port 0 is attached. Now total ports is 1
1143 For example, to attach a port created by pcap PMD.
1145 .. code-block:: console
1147 testpmd> port attach net_pcap0
1148 Attaching a new port...
1149 PMD: Initializing pmd_pcap for net_pcap0
1150 PMD: Creating pcap-backed ethdev on numa socket 0
1151 Port 0 is attached. Now total ports is 1
1154 In this case, identifier is ``net_pcap0``.
1155 This identifier format is the same as ``--vdev`` format of DPDK applications.
1157 For example, to re-attach a bonded port which has been previously detached,
1158 the mode and slave parameters must be given.
1160 .. code-block:: console
1162 testpmd> port attach net_bond_0,mode=0,slave=1
1163 Attaching a new port...
1164 EAL: Initializing pmd_bond for net_bond_0
1165 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1166 Port 0 is attached. Now total ports is 1
1173 Detach a specific port::
1175 testpmd> port detach (port_id)
1177 Before detaching a port, the port should be stopped and closed.
1179 For example, to detach a pci device port 0.
1181 .. code-block:: console
1183 testpmd> port stop 0
1186 testpmd> port close 0
1190 testpmd> port detach 0
1192 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1193 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1194 EAL: PCI memory unmapped at 0x7f83bfa00000
1195 EAL: PCI memory unmapped at 0x7f83bfa80000
1199 For example, to detach a virtual device port 0.
1201 .. code-block:: console
1203 testpmd> port stop 0
1206 testpmd> port close 0
1210 testpmd> port detach 0
1212 PMD: Closing pcap ethdev on numa socket 0
1213 Port 'net_pcap0' is detached. Now total ports is 0
1216 To remove a pci device completely from the system, first detach the port from testpmd.
1217 Then the device should be moved under kernel management.
1218 Finally the device can be removed using kernel pci hotplug functionality.
1220 For example, to move a pci device under kernel management:
1222 .. code-block:: console
1224 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1226 ./usertools/dpdk-devbind.py --status
1228 Network devices using DPDK-compatible driver
1229 ============================================
1232 Network devices using kernel driver
1233 ===================================
1234 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1236 To remove a port created by a virtual device, above steps are not needed.
1241 Start all ports or a specific port::
1243 testpmd> port start (port_id|all)
1248 Stop all ports or a specific port::
1250 testpmd> port stop (port_id|all)
1255 Close all ports or a specific port::
1257 testpmd> port close (port_id|all)
1259 port start/stop queue
1260 ~~~~~~~~~~~~~~~~~~~~~
1262 Start/stop a rx/tx queue on a specific port::
1264 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1266 Only take effect when port is started.
1271 Set the speed and duplex mode for all ports or a specific port::
1273 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1274 duplex (half|full|auto)
1276 port config - queues/descriptors
1277 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1279 Set number of queues/descriptors for rxq, txq, rxd and txd::
1281 testpmd> port config all (rxq|txq|rxd|txd) (value)
1283 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1285 port config - max-pkt-len
1286 ~~~~~~~~~~~~~~~~~~~~~~~~~
1288 Set the maximum packet length::
1290 testpmd> port config all max-pkt-len (value)
1292 This is equivalent to the ``--max-pkt-len`` command-line option.
1294 port config - CRC Strip
1295 ~~~~~~~~~~~~~~~~~~~~~~~
1297 Set hardware CRC stripping on or off for all ports::
1299 testpmd> port config all crc-strip (on|off)
1301 CRC stripping is off by default.
1303 The ``on`` option is equivalent to the ``--crc-strip`` command-line option.
1305 port config - scatter
1306 ~~~~~~~~~~~~~~~~~~~~~~~
1308 Set RX scatter mode on or off for all ports::
1310 testpmd> port config all scatter (on|off)
1312 RX scatter mode is off by default.
1314 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1316 port config - TX queue flags
1317 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1319 Set a hexadecimal bitmap of TX queue flags for all ports::
1321 testpmd> port config all txqflags value
1323 This command is equivalent to the ``--txqflags`` command-line option.
1325 port config - RX Checksum
1326 ~~~~~~~~~~~~~~~~~~~~~~~~~
1328 Set hardware RX checksum offload to on or off for all ports::
1330 testpmd> port config all rx-cksum (on|off)
1332 Checksum offload is off by default.
1334 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1339 Set hardware VLAN on or off for all ports::
1341 testpmd> port config all hw-vlan (on|off)
1343 Hardware VLAN is on by default.
1345 The ``off`` option is equivalent to the ``--disable-hw-vlan`` command-line option.
1347 port config - VLAN filter
1348 ~~~~~~~~~~~~~~~~~~~~~~~~~
1350 Set hardware VLAN filter on or off for all ports::
1352 testpmd> port config all hw-vlan-filter (on|off)
1354 Hardware VLAN filter is on by default.
1356 The ``off`` option is equivalent to the ``--disable-hw-vlan-filter`` command-line option.
1358 port config - VLAN strip
1359 ~~~~~~~~~~~~~~~~~~~~~~~~
1361 Set hardware VLAN strip on or off for all ports::
1363 testpmd> port config all hw-vlan-strip (on|off)
1365 Hardware VLAN strip is on by default.
1367 The ``off`` option is equivalent to the ``--disable-hw-vlan-strip`` command-line option.
1369 port config - VLAN extend
1370 ~~~~~~~~~~~~~~~~~~~~~~~~~
1372 Set hardware VLAN extend on or off for all ports::
1374 testpmd> port config all hw-vlan-extend (on|off)
1376 Hardware VLAN extend is off by default.
1378 The ``off`` option is equivalent to the ``--disable-hw-vlan-extend`` command-line option.
1380 port config - Drop Packets
1381 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1383 Set packet drop for packets with no descriptors on or off for all ports::
1385 testpmd> port config all drop-en (on|off)
1387 Packet dropping for packets with no descriptors is off by default.
1389 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1394 Set the RSS (Receive Side Scaling) mode on or off::
1396 testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1398 RSS is on by default.
1400 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1402 port config - RSS Reta
1403 ~~~~~~~~~~~~~~~~~~~~~~
1405 Set the RSS (Receive Side Scaling) redirection table::
1407 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1412 Set the DCB mode for an individual port::
1414 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1416 The traffic class should be 4 or 8.
1421 Set the number of packets per burst::
1423 testpmd> port config all burst (value)
1425 This is equivalent to the ``--burst`` command-line option.
1427 port config - Threshold
1428 ~~~~~~~~~~~~~~~~~~~~~~~
1430 Set thresholds for TX/RX queues::
1432 testpmd> port config all (threshold) (value)
1434 Where the threshold type can be:
1436 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1438 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1440 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1442 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1444 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1446 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1448 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1450 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1452 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1454 These threshold options are also available from the command-line.
1459 Set the value of ether-type for E-tag::
1461 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1463 Enable/disable the E-tag support::
1465 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1468 Link Bonding Functions
1469 ----------------------
1471 The Link Bonding functions make it possible to dynamically create and
1472 manage link bonding devices from within testpmd interactive prompt.
1474 create bonded device
1475 ~~~~~~~~~~~~~~~~~~~~
1477 Create a new bonding device::
1479 testpmd> create bonded device (mode) (socket)
1481 For example, to create a bonded device in mode 1 on socket 0::
1483 testpmd> create bonded 1 0
1484 created new bonded device (port X)
1489 Adds Ethernet device to a Link Bonding device::
1491 testpmd> add bonding slave (slave id) (port id)
1493 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
1495 testpmd> add bonding slave 6 10
1498 remove bonding slave
1499 ~~~~~~~~~~~~~~~~~~~~
1501 Removes an Ethernet slave device from a Link Bonding device::
1503 testpmd> remove bonding slave (slave id) (port id)
1505 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
1507 testpmd> remove bonding slave 6 10
1512 Set the Link Bonding mode of a Link Bonding device::
1514 testpmd> set bonding mode (value) (port id)
1516 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
1518 testpmd> set bonding mode 3 10
1523 Set an Ethernet slave device as the primary device on a Link Bonding device::
1525 testpmd> set bonding primary (slave id) (port id)
1527 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
1529 testpmd> set bonding primary 6 10
1534 Set the MAC address of a Link Bonding device::
1536 testpmd> set bonding mac (port id) (mac)
1538 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
1540 testpmd> set bonding mac 10 00:00:00:00:00:01
1542 set bonding xmit_balance_policy
1543 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1545 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
1547 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
1549 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
1551 testpmd> set bonding xmit_balance_policy 10 l34
1554 set bonding mon_period
1555 ~~~~~~~~~~~~~~~~~~~~~~
1557 Set the link status monitoring polling period in milliseconds for a bonding device.
1559 This adds support for PMD slave devices which do not support link status interrupts.
1560 When the mon_period is set to a value greater than 0 then all PMD's which do not support
1561 link status ISR will be queried every polling interval to check if their link status has changed::
1563 testpmd> set bonding mon_period (port_id) (value)
1565 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
1567 testpmd> set bonding mon_period 5 150
1573 Show the current configuration of a Link Bonding device::
1575 testpmd> show bonding config (port id)
1578 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
1579 in balance mode with a transmission policy of layer 2+3::
1581 testpmd> show bonding config 9
1583 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
1585 Active Slaves (3): [1 3 4]
1592 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
1593 This is mainly useful for debugging purposes.
1594 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
1595 and fields that can be accessed.
1600 Display the value of a port register::
1602 testpmd> read reg (port_id) (address)
1604 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
1606 testpmd> read reg 0 0xEE00
1607 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
1612 Display a port register bit field::
1614 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
1616 For example, reading the lowest two bits from the register in the example above::
1618 testpmd> read regfield 0 0xEE00 0 1
1619 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
1624 Display a single port register bit::
1626 testpmd> read regbit (port_id) (address) (bit_x)
1628 For example, reading the lowest bit from the register in the example above::
1630 testpmd> read regbit 0 0xEE00 0
1631 port 0 PCI register at offset 0xEE00: bit 0=1
1636 Set the value of a port register::
1638 testpmd> write reg (port_id) (address) (value)
1640 For example, to clear a register::
1642 testpmd> write reg 0 0xEE00 0x0
1643 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
1648 Set bit field of a port register::
1650 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
1652 For example, writing to the register cleared in the example above::
1654 testpmd> write regfield 0 0xEE00 0 1 2
1655 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
1660 Set single bit value of a port register::
1662 testpmd> write regbit (port_id) (address) (bit_x) (value)
1664 For example, to set the high bit in the register from the example above::
1666 testpmd> write regbit 0 0xEE00 31 1
1667 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
1673 This section details the available filter functions that are available.
1675 Note these functions interface the deprecated legacy filtering framework,
1676 superseded by *rte_flow*. See `Flow rules management`_.
1679 ~~~~~~~~~~~~~~~~~~~~
1681 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
1683 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
1684 ethertype (ether_type) (drop|fwd) queue (queue_id)
1686 The available information parameters are:
1688 * ``port_id``: The port which the Ethertype filter assigned on.
1690 * ``mac_addr``: Compare destination mac address.
1692 * ``mac_ignr``: Ignore destination mac address match.
1694 * ``mac_address``: Destination mac address to match.
1696 * ``ether_type``: The EtherType value want to match,
1697 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
1699 * ``queue_id``: The receive queue associated with this EtherType filter.
1700 It is meaningless when deleting or dropping.
1702 Example, to add/remove an ethertype filter rule::
1704 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
1705 ethertype 0x0806 fwd queue 3
1707 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
1708 ethertype 0x0806 fwd queue 3
1713 Add or delete a 2-tuple filter,
1714 which identifies packets by specific protocol and destination TCP/UDP port
1715 and forwards packets into one of the receive queues::
1717 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
1718 protocol (protocol_value) mask (mask_value) \
1719 tcp_flags (tcp_flags_value) priority (prio_value) \
1722 The available information parameters are:
1724 * ``port_id``: The port which the 2-tuple filter assigned on.
1726 * ``dst_port_value``: Destination port in L4.
1728 * ``protocol_value``: IP L4 protocol.
1730 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
1732 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
1734 * ``prio_value``: Priority of this filter.
1736 * ``queue_id``: The receive queue associated with this 2-tuple filter.
1738 Example, to add/remove an 2tuple filter rule::
1740 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
1741 tcp_flags 0x02 priority 3 queue 3
1743 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
1744 tcp_flags 0x02 priority 3 queue 3
1749 Add or delete a 5-tuple filter,
1750 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
1751 and routes packets into one of the receive queues::
1753 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
1754 (src_address) dst_port (dst_port_value) \
1755 src_port (src_port_value) protocol (protocol_value) \
1756 mask (mask_value) tcp_flags (tcp_flags_value) \
1757 priority (prio_value) queue (queue_id)
1759 The available information parameters are:
1761 * ``port_id``: The port which the 5-tuple filter assigned on.
1763 * ``dst_address``: Destination IP address.
1765 * ``src_address``: Source IP address.
1767 * ``dst_port_value``: TCP/UDP destination port.
1769 * ``src_port_value``: TCP/UDP source port.
1771 * ``protocol_value``: L4 protocol.
1773 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
1775 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
1777 * ``prio_value``: The priority of this filter.
1779 * ``queue_id``: The receive queue associated with this 5-tuple filter.
1781 Example, to add/remove an 5tuple filter rule::
1783 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1784 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1785 flags 0x0 priority 3 queue 3
1787 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1788 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1789 flags 0x0 priority 3 queue 3
1794 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
1796 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
1798 The available information parameters are:
1800 * ``port_id``: The port which the SYN filter assigned on.
1802 * ``high``: This SYN filter has higher priority than other filters.
1804 * ``low``: This SYN filter has lower priority than other filters.
1806 * ``queue_id``: The receive queue associated with this SYN filter
1810 testpmd> syn_filter 0 add priority high queue 3
1815 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
1816 and routed into one of the receive queues::
1818 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
1819 mask (mask_value) priority (prio_value) queue (queue_id)
1821 The available information parameters are:
1823 * ``port_id``: The port which the Flex filter is assigned on.
1825 * ``len_value``: Filter length in bytes, no greater than 128.
1827 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
1829 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
1831 * ``prio_value``: The priority of this filter.
1833 * ``queue_id``: The receive queue associated with this Flex filter.
1837 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
1838 mask 000C priority 3 queue 3
1840 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
1841 mask 000C priority 3 queue 3
1844 .. _testpmd_flow_director:
1846 flow_director_filter
1847 ~~~~~~~~~~~~~~~~~~~~
1849 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
1851 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
1852 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
1854 * Perfect match filters.
1855 The hardware checks a match between the masked fields of the received packets and the programmed filters.
1856 The masked fields are for IP flow.
1858 * Signature filters.
1859 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
1861 * Perfect-mac-vlan match filters.
1862 The hardware checks a match between the masked fields of the received packets and the programmed filters.
1863 The masked fields are for MAC VLAN flow.
1865 * Perfect-tunnel match filters.
1866 The hardware checks a match between the masked fields of the received packets and the programmed filters.
1867 The masked fields are for tunnel flow.
1869 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
1870 per flow type and the flexible payload.
1872 The Flow Director can also mask out parts of all of these fields so that filters
1873 are only applied to certain fields or parts of the fields.
1875 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
1877 # Commands to add flow director filters of different flow types::
1879 flow_director_filter (port_id) mode IP (add|del|update) \
1880 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
1881 src (src_ip_address) dst (dst_ip_address) \
1882 tos (tos_value) proto (proto_value) ttl (ttl_value) \
1883 vlan (vlan_value) flexbytes (flexbytes_value) \
1884 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
1887 flow_director_filter (port_id) mode IP (add|del|update) \
1888 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
1889 src (src_ip_address) (src_port) \
1890 dst (dst_ip_address) (dst_port) \
1891 tos (tos_value) ttl (ttl_value) \
1892 vlan (vlan_value) flexbytes (flexbytes_value) \
1893 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
1896 flow_director_filter (port_id) mode IP (add|del|update) \
1897 flow (ipv4-sctp|ipv6-sctp) \
1898 src (src_ip_address) (src_port) \
1899 dst (dst_ip_address) (dst_port) \
1900 tos (tos_value) ttl (ttl_value) \
1901 tag (verification_tag) vlan (vlan_value) \
1902 flexbytes (flexbytes_value) (drop|fwd) \
1903 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
1905 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
1906 ether (ethertype) flexbytes (flexbytes_value) \
1907 (drop|fwd) pf|vf(vf_id) queue (queue_id)
1910 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
1911 mac (mac_address) vlan (vlan_value) \
1912 flexbytes (flexbytes_value) (drop|fwd) \
1913 queue (queue_id) fd_id (fd_id_value)
1915 flow_director_filter (port_id) mode Tunnel (add|del|update) \
1916 mac (mac_address) vlan (vlan_value) \
1917 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
1918 flexbytes (flexbytes_value) (drop|fwd) \
1919 queue (queue_id) fd_id (fd_id_value)
1921 For example, to add an ipv4-udp flow type filter::
1923 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
1924 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
1925 fwd pf queue 1 fd_id 1
1927 For example, add an ipv4-other flow type filter::
1929 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
1930 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
1931 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
1936 Flush all flow director filters on a device::
1938 testpmd> flush_flow_director (port_id)
1940 Example, to flush all flow director filter on port 0::
1942 testpmd> flush_flow_director 0
1947 Set flow director's input masks::
1949 flow_director_mask (port_id) mode IP vlan (vlan_value) \
1950 src_mask (ipv4_src) (ipv6_src) (src_port) \
1951 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
1953 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
1955 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
1956 mac (mac_value) tunnel-type (tunnel_type_value) \
1957 tunnel-id (tunnel_id_value)
1959 Example, to set flow director mask on port 0::
1961 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
1962 src_mask 255.255.255.255 \
1963 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
1964 dst_mask 255.255.255.255 \
1965 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
1967 flow_director_flex_mask
1968 ~~~~~~~~~~~~~~~~~~~~~~~
1970 set masks of flow director's flexible payload based on certain flow type::
1972 testpmd> flow_director_flex_mask (port_id) \
1973 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
1974 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
1975 l2_payload|all) (mask)
1977 Example, to set flow director's flex mask for all flow type on port 0::
1979 testpmd> flow_director_flex_mask 0 flow all \
1980 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
1983 flow_director_flex_payload
1984 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1986 Configure flexible payload selection::
1988 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
1990 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
1992 testpmd> flow_director_flex_payload 0 l4 \
1993 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
1995 get_sym_hash_ena_per_port
1996 ~~~~~~~~~~~~~~~~~~~~~~~~~
1998 Get symmetric hash enable configuration per port::
2000 get_sym_hash_ena_per_port (port_id)
2002 For example, to get symmetric hash enable configuration of port 1::
2004 testpmd> get_sym_hash_ena_per_port 1
2006 set_sym_hash_ena_per_port
2007 ~~~~~~~~~~~~~~~~~~~~~~~~~
2009 Set symmetric hash enable configuration per port to enable or disable::
2011 set_sym_hash_ena_per_port (port_id) (enable|disable)
2013 For example, to set symmetric hash enable configuration of port 1 to enable::
2015 testpmd> set_sym_hash_ena_per_port 1 enable
2017 get_hash_global_config
2018 ~~~~~~~~~~~~~~~~~~~~~~
2020 Get the global configurations of hash filters::
2022 get_hash_global_config (port_id)
2024 For example, to get the global configurations of hash filters of port 1::
2026 testpmd> get_hash_global_config 1
2028 set_hash_global_config
2029 ~~~~~~~~~~~~~~~~~~~~~~
2031 Set the global configurations of hash filters::
2033 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2034 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2035 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload) \
2038 For example, to enable simple_xor for flow type of ipv6 on port 2::
2040 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2045 Set the input set for hash::
2047 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2048 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2049 l2_payload) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2050 ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2051 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2052 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2053 fld-8th|none) (select|add)
2055 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2057 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2062 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2063 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2065 Set the input set for flow director::
2067 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2068 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2069 l2_payload) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2070 ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2071 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2072 sctp-dst-port|sctp-veri-tag|none) (select|add)
2074 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2076 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2081 Set different GRE key length for input set::
2083 global_config (port_id) gre-key-len (number in bytes)
2085 For example to set GRE key length for input set to 4 bytes on port 0::
2087 testpmd> global_config 0 gre-key-len 4
2090 .. _testpmd_rte_flow:
2092 Flow rules management
2093 ---------------------
2095 Control of the generic flow API (*rte_flow*) is fully exposed through the
2096 ``flow`` command (validation, creation, destruction and queries).
2098 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2099 features simultaneously may cause undefined side-effects and is therefore
2105 Because the ``flow`` command uses dynamic tokens to handle the large number
2106 of possible flow rules combinations, its behavior differs slightly from
2107 other commands, in particular:
2109 - Pressing *?* or the *<tab>* key displays contextual help for the current
2110 token, not that of the entire command.
2112 - Optional and repeated parameters are supported (provided they are listed
2113 in the contextual help).
2115 The first parameter stands for the operation mode. Possible operations and
2116 their general syntax are described below. They are covered in detail in the
2119 - Check whether a flow rule can be created::
2121 flow validate {port_id}
2122 [group {group_id}] [priority {level}] [ingress] [egress]
2123 pattern {item} [/ {item} [...]] / end
2124 actions {action} [/ {action} [...]] / end
2126 - Create a flow rule::
2128 flow create {port_id}
2129 [group {group_id}] [priority {level}] [ingress] [egress]
2130 pattern {item} [/ {item} [...]] / end
2131 actions {action} [/ {action} [...]] / end
2133 - Destroy specific flow rules::
2135 flow destroy {port_id} rule {rule_id} [...]
2137 - Destroy all flow rules::
2139 flow flush {port_id}
2141 - Query an existing flow rule::
2143 flow query {port_id} {rule_id} {action}
2145 - List existing flow rules sorted by priority, filtered by group
2148 flow list {port_id} [group {group_id}] [...]
2150 Validating flow rules
2151 ~~~~~~~~~~~~~~~~~~~~~
2153 ``flow validate`` reports whether a flow rule would be accepted by the
2154 underlying device in its current state but stops short of creating it. It is
2155 bound to ``rte_flow_validate()``::
2157 flow validate {port_id}
2158 [group {group_id}] [priority {level}] [ingress] [egress]
2159 pattern {item} [/ {item} [...]] / end
2160 actions {action} [/ {action} [...]] / end
2162 If successful, it will show::
2166 Otherwise it will show an error message of the form::
2168 Caught error type [...] ([...]): [...]
2170 This command uses the same parameters as ``flow create``, their format is
2171 described in `Creating flow rules`_.
2173 Check whether redirecting any Ethernet packet received on port 0 to RX queue
2174 index 6 is supported::
2176 testpmd> flow validate 0 ingress pattern eth / end
2177 actions queue index 6 / end
2181 Port 0 does not support TCPv6 rules::
2183 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
2185 Caught error type 9 (specific pattern item): Invalid argument
2191 ``flow create`` validates and creates the specified flow rule. It is bound
2192 to ``rte_flow_create()``::
2194 flow create {port_id}
2195 [group {group_id}] [priority {level}] [ingress] [egress]
2196 pattern {item} [/ {item} [...]] / end
2197 actions {action} [/ {action} [...]] / end
2199 If successful, it will return a flow rule ID usable with other commands::
2201 Flow rule #[...] created
2203 Otherwise it will show an error message of the form::
2205 Caught error type [...] ([...]): [...]
2207 Parameters describe in the following order:
2209 - Attributes (*group*, *priority*, *ingress*, *egress* tokens).
2210 - A matching pattern, starting with the *pattern* token and terminated by an
2212 - Actions, starting with the *actions* token and terminated by an *end*
2215 These translate directly to *rte_flow* objects provided as-is to the
2216 underlying functions.
2218 The shortest valid definition only comprises mandatory tokens::
2220 testpmd> flow create 0 pattern end actions end
2222 Note that PMDs may refuse rules that essentially do nothing such as this
2225 **All unspecified object values are automatically initialized to 0.**
2230 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
2231 specified before the ``pattern`` token.
2233 - ``group {group id}``: priority group.
2234 - ``priority {level}``: priority level within group.
2235 - ``ingress``: rule applies to ingress traffic.
2236 - ``egress``: rule applies to egress traffic.
2238 Each instance of an attribute specified several times overrides the previous
2239 value as shown below (group 4 is used)::
2241 testpmd> flow create 0 group 42 group 24 group 4 [...]
2243 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
2245 While not specifying a direction is an error, some rules may allow both
2248 Most rules affect RX therefore contain the ``ingress`` token::
2250 testpmd> flow create 0 ingress pattern [...]
2255 A matching pattern starts after the ``pattern`` token. It is made of pattern
2256 items and is terminated by a mandatory ``end`` item.
2258 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
2259 rte_flow_item_type``).
2261 The ``/`` token is used as a separator between pattern items as shown
2264 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
2266 Note that protocol items like these must be stacked from lowest to highest
2267 layer to make sense. For instance, the following rule is either invalid or
2268 unlikely to match any packet::
2270 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
2272 More information on these restrictions can be found in the *rte_flow*
2275 Several items support additional specification structures, for example
2276 ``ipv4`` allows specifying source and destination addresses as follows::
2278 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
2279 dst is 10.2.0.0 / end [...]
2281 This rule matches all IPv4 traffic with the specified properties.
2283 In this example, ``src`` and ``dst`` are field names of the underlying
2284 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
2285 in a similar fashion.
2287 The ``is`` token means that the subsequent value must be matched exactly,
2288 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
2289 accordingly. Possible assignment tokens are:
2291 - ``is``: match value perfectly (with full bit-mask).
2292 - ``spec``: match value according to configured bit-mask.
2293 - ``last``: specify upper bound to establish a range.
2294 - ``mask``: specify bit-mask with relevant bits set to one.
2295 - ``prefix``: generate bit-mask from a prefix length.
2297 These yield identical results::
2299 ipv4 src is 10.1.1.1
2303 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
2307 ipv4 src spec 10.1.1.1 src prefix 32
2311 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
2315 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
2317 Inclusive ranges can be defined with ``last``::
2319 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
2321 Note that ``mask`` affects both ``spec`` and ``last``::
2323 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
2324 # matches 10.1.0.0 to 10.2.255.255
2326 Properties can be modified multiple times::
2328 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
2332 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
2337 This section lists supported pattern items and their attributes, if any.
2339 - ``end``: end list of pattern items.
2341 - ``void``: no-op pattern item.
2343 - ``invert``: perform actions when pattern does not match.
2345 - ``any``: match any protocol for the current layer.
2347 - ``num {unsigned}``: number of layers covered.
2349 - ``pf``: match packets addressed to the physical function.
2351 - ``vf``: match packets addressed to a virtual function ID.
2353 - ``id {unsigned}``: destination VF ID.
2355 - ``port``: device-specific physical port index to use.
2357 - ``index {unsigned}``: physical port index.
2359 - ``raw``: match an arbitrary byte string.
2361 - ``relative {boolean}``: look for pattern after the previous item.
2362 - ``search {boolean}``: search pattern from offset (see also limit).
2363 - ``offset {integer}``: absolute or relative offset for pattern.
2364 - ``limit {unsigned}``: search area limit for start of pattern.
2365 - ``pattern {string}``: byte string to look for.
2367 - ``eth``: match Ethernet header.
2369 - ``dst {MAC-48}``: destination MAC.
2370 - ``src {MAC-48}``: source MAC.
2371 - ``type {unsigned}``: EtherType.
2373 - ``vlan``: match 802.1Q/ad VLAN tag.
2375 - ``tpid {unsigned}``: tag protocol identifier.
2376 - ``tci {unsigned}``: tag control information.
2377 - ``pcp {unsigned}``: priority code point.
2378 - ``dei {unsigned}``: drop eligible indicator.
2379 - ``vid {unsigned}``: VLAN identifier.
2381 - ``ipv4``: match IPv4 header.
2383 - ``tos {unsigned}``: type of service.
2384 - ``ttl {unsigned}``: time to live.
2385 - ``proto {unsigned}``: next protocol ID.
2386 - ``src {ipv4 address}``: source address.
2387 - ``dst {ipv4 address}``: destination address.
2389 - ``ipv6``: match IPv6 header.
2391 - ``tc {unsigned}``: traffic class.
2392 - ``flow {unsigned}``: flow label.
2393 - ``proto {unsigned}``: protocol (next header).
2394 - ``hop {unsigned}``: hop limit.
2395 - ``src {ipv6 address}``: source address.
2396 - ``dst {ipv6 address}``: destination address.
2398 - ``icmp``: match ICMP header.
2400 - ``type {unsigned}``: ICMP packet type.
2401 - ``code {unsigned}``: ICMP packet code.
2403 - ``udp``: match UDP header.
2405 - ``src {unsigned}``: UDP source port.
2406 - ``dst {unsigned}``: UDP destination port.
2408 - ``tcp``: match TCP header.
2410 - ``src {unsigned}``: TCP source port.
2411 - ``dst {unsigned}``: TCP destination port.
2413 - ``sctp``: match SCTP header.
2415 - ``src {unsigned}``: SCTP source port.
2416 - ``dst {unsigned}``: SCTP destination port.
2417 - ``tag {unsigned}``: validation tag.
2418 - ``cksum {unsigned}``: checksum.
2420 - ``vxlan``: match VXLAN header.
2422 - ``vni {unsigned}``: VXLAN identifier.
2427 A list of actions starts after the ``actions`` token in the same fashion as
2428 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
2429 terminated by a mandatory ``end`` action.
2431 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
2432 rte_flow_action_type``).
2434 Dropping all incoming UDPv4 packets can be expressed as follows::
2436 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2439 Several actions have configurable properties which must be specified when
2440 there is no valid default value. For example, ``queue`` requires a target
2443 This rule redirects incoming UDPv4 traffic to queue index 6::
2445 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2446 actions queue index 6 / end
2448 While this one could be rejected by PMDs (unspecified queue index)::
2450 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2453 As defined by *rte_flow*, the list is not ordered, all actions of a given
2454 rule are performed simultaneously. These are equivalent::
2456 queue index 6 / void / mark id 42 / end
2460 void / mark id 42 / queue index 6 / end
2462 All actions in a list should have different types, otherwise only the last
2463 action of a given type is taken into account::
2465 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
2469 drop / drop / drop / end # drop is performed only once
2473 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
2475 Considering they are performed simultaneously, opposite and overlapping
2476 actions can sometimes be combined when the end result is unambiguous::
2478 drop / queue index 6 / end # drop has no effect
2482 drop / dup index 6 / end # same as above
2486 queue index 6 / rss queues 6 7 8 / end # queue has no effect
2490 drop / passthru / end # drop has no effect
2492 Note that PMDs may still refuse such combinations.
2497 This section lists supported actions and their attributes, if any.
2499 - ``end``: end list of actions.
2501 - ``void``: no-op action.
2503 - ``passthru``: let subsequent rule process matched packets.
2505 - ``mark``: attach 32 bit value to packets.
2507 - ``id {unsigned}``: 32 bit value to return with packets.
2509 - ``flag``: flag packets.
2511 - ``queue``: assign packets to a given queue index.
2513 - ``index {unsigned}``: queue index to use.
2515 - ``drop``: drop packets (note: passthru has priority).
2517 - ``count``: enable counters for this rule.
2519 - ``dup``: duplicate packets to a given queue index.
2521 - ``index {unsigned}``: queue index to duplicate packets to.
2523 - ``rss``: spread packets among several queues.
2525 - ``queues [{unsigned} [...]] end``: queue indices to use.
2527 - ``pf``: redirect packets to physical device function.
2529 - ``vf``: redirect packets to virtual device function.
2531 - ``original {boolean}``: use original VF ID if possible.
2532 - ``id {unsigned}``: VF ID to redirect packets to.
2534 Destroying flow rules
2535 ~~~~~~~~~~~~~~~~~~~~~
2537 ``flow destroy`` destroys one or more rules from their rule ID (as returned
2538 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
2539 times as necessary::
2541 flow destroy {port_id} rule {rule_id} [...]
2543 If successful, it will show::
2545 Flow rule #[...] destroyed
2547 It does not report anything for rule IDs that do not exist. The usual error
2548 message is shown when a rule cannot be destroyed::
2550 Caught error type [...] ([...]): [...]
2552 ``flow flush`` destroys all rules on a device and does not take extra
2553 arguments. It is bound to ``rte_flow_flush()``::
2555 flow flush {port_id}
2557 Any errors are reported as above.
2559 Creating several rules and destroying them::
2561 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2562 actions queue index 2 / end
2563 Flow rule #0 created
2564 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2565 actions queue index 3 / end
2566 Flow rule #1 created
2567 testpmd> flow destroy 0 rule 0 rule 1
2568 Flow rule #1 destroyed
2569 Flow rule #0 destroyed
2572 The same result can be achieved using ``flow flush``::
2574 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2575 actions queue index 2 / end
2576 Flow rule #0 created
2577 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2578 actions queue index 3 / end
2579 Flow rule #1 created
2580 testpmd> flow flush 0
2583 Non-existent rule IDs are ignored::
2585 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2586 actions queue index 2 / end
2587 Flow rule #0 created
2588 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2589 actions queue index 3 / end
2590 Flow rule #1 created
2591 testpmd> flow destroy 0 rule 42 rule 10 rule 2
2593 testpmd> flow destroy 0 rule 0
2594 Flow rule #0 destroyed
2600 ``flow query`` queries a specific action of a flow rule having that
2601 ability. Such actions collect information that can be reported using this
2602 command. It is bound to ``rte_flow_query()``::
2604 flow query {port_id} {rule_id} {action}
2606 If successful, it will display either the retrieved data for known actions
2607 or the following message::
2609 Cannot display result for action type [...] ([...])
2611 Otherwise, it will complain either that the rule does not exist or that some
2614 Flow rule #[...] not found
2618 Caught error type [...] ([...]): [...]
2620 Currently only the ``count`` action is supported. This action reports the
2621 number of packets that hit the flow rule and the total number of bytes. Its
2622 output has the following format::
2625 hits_set: [...] # whether "hits" contains a valid value
2626 bytes_set: [...] # whether "bytes" contains a valid value
2627 hits: [...] # number of packets
2628 bytes: [...] # number of bytes
2630 Querying counters for TCPv6 packets redirected to queue 6::
2632 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
2633 actions queue index 6 / count / end
2634 Flow rule #4 created
2635 testpmd> flow query 0 4 count
2646 ``flow list`` lists existing flow rules sorted by priority and optionally
2647 filtered by group identifiers::
2649 flow list {port_id} [group {group_id}] [...]
2651 This command only fails with the following message if the device does not
2656 Output consists of a header line followed by a short description of each
2657 flow rule, one per line. There is no output at all when no flow rules are
2658 configured on the device::
2660 ID Group Prio Attr Rule
2661 [...] [...] [...] [...] [...]
2663 ``Attr`` column flags:
2665 - ``i`` for ``ingress``.
2666 - ``e`` for ``egress``.
2668 Creating several flow rules and listing them::
2670 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2671 actions queue index 6 / end
2672 Flow rule #0 created
2673 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2674 actions queue index 2 / end
2675 Flow rule #1 created
2676 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
2677 actions rss queues 6 7 8 end / end
2678 Flow rule #2 created
2679 testpmd> flow list 0
2680 ID Group Prio Attr Rule
2681 0 0 0 i- ETH IPV4 => QUEUE
2682 1 0 0 i- ETH IPV6 => QUEUE
2683 2 0 5 i- ETH IPV4 UDP => RSS
2686 Rules are sorted by priority (i.e. group ID first, then priority level)::
2688 testpmd> flow list 1
2689 ID Group Prio Attr Rule
2690 0 0 0 i- ETH => COUNT
2691 6 0 500 i- ETH IPV6 TCP => DROP COUNT
2692 5 0 1000 i- ETH IPV6 ICMP => QUEUE
2693 1 24 0 i- ETH IPV4 UDP => QUEUE
2694 4 24 10 i- ETH IPV4 TCP => DROP
2695 3 24 20 i- ETH IPV4 => DROP
2696 2 24 42 i- ETH IPV4 UDP => QUEUE
2697 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
2700 Output can be limited to specific groups::
2702 testpmd> flow list 1 group 0 group 63
2703 ID Group Prio Attr Rule
2704 0 0 0 i- ETH => COUNT
2705 6 0 500 i- ETH IPV6 TCP => DROP COUNT
2706 5 0 1000 i- ETH IPV6 ICMP => QUEUE
2707 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE