2 Copyright(c) 2010-2016 Intel Corporation. All rights reserved.
5 Redistribution and use in source and binary forms, with or without
6 modification, are permitted provided that the following conditions
9 * Redistributions of source code must retain the above copyright
10 notice, this list of conditions and the following disclaimer.
11 * Redistributions in binary form must reproduce the above copyright
12 notice, this list of conditions and the following disclaimer in
13 the documentation and/or other materials provided with the
15 * Neither the name of Intel Corporation nor the names of its
16 contributors may be used to endorse or promote products derived
17 from this software without specific prior written permission.
19 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
20 "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
21 LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
22 A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
23 OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
24 SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
25 LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
26 DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
27 THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
28 (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
29 OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
33 Testpmd Runtime Functions
34 =========================
36 Where the testpmd application is started in interactive mode, (``-i|--interactive``),
37 it displays a prompt that can be used to start and stop forwarding,
38 configure the application, display statistics (including the extended NIC
39 statistics aka xstats) , set the Flow Director and other tasks::
43 The testpmd prompt has some, limited, readline support.
44 Common bash command-line functions such as ``Ctrl+a`` and ``Ctrl+e`` to go to the start and end of the prompt line are supported
45 as well as access to the command history via the up-arrow.
47 There is also support for tab completion.
48 If you type a partial command and hit ``<TAB>`` you get a list of the available completions:
50 .. code-block:: console
52 testpmd> show port <TAB>
54 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
55 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
56 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
57 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
63 Some examples in this document are too long to fit on one line are are shown wrapped at `"\\"` for display purposes::
65 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
66 (pause_time) (send_xon) (port_id)
68 In the real ``testpmd>`` prompt these commands should be on a single line.
73 The testpmd has on-line help for the functions that are available at runtime.
74 These are divided into sections and can be accessed using help, help section or help all:
76 .. code-block:: console
80 help control : Start and stop forwarding.
81 help display : Displaying port, stats and config information.
82 help config : Configuration information.
83 help ports : Configuring ports.
84 help registers : Reading and setting port registers.
85 help filters : Filters configuration help.
86 help all : All of the above sections.
89 Command File Functions
90 ----------------------
92 To facilitate loading large number of commands or to avoid cutting and pasting where not
93 practical or possible testpmd supports alternative methods for executing commands.
95 * If started with the ``--cmdline-file=FILENAME`` command line argument testpmd
96 will execute all CLI commands contained within the file immediately before
97 starting packet forwarding or entering interactive mode.
99 .. code-block:: console
101 ./testpmd -n4 -r2 ... -- -i --cmdline-file=/home/ubuntu/flow-create-commands.txt
102 Interactive-mode selected
103 CLI commands to be read from /home/ubuntu/flow-create-commands.txt
104 Configuring Port 0 (socket 0)
105 Port 0: 7C:FE:90:CB:74:CE
106 Configuring Port 1 (socket 0)
107 Port 1: 7C:FE:90:CB:74:CA
108 Checking link statuses...
109 Port 0 Link Up - speed 10000 Mbps - full-duplex
110 Port 1 Link Up - speed 10000 Mbps - full-duplex
116 Flow rule #498 created
117 Flow rule #499 created
118 Read all CLI commands from /home/ubuntu/flow-create-commands.txt
122 * At run-time additional commands can be loaded in bulk by invoking the ``load FILENAME``
125 .. code-block:: console
127 testpmd> load /home/ubuntu/flow-create-commands.txt
132 Flow rule #498 created
133 Flow rule #499 created
134 Read all CLI commands from /home/ubuntu/flow-create-commands.txt
138 In all cases output from any included command will be displayed as standard output.
139 Execution will continue until the end of the file is reached regardless of
140 whether any errors occur. The end user must examine the output to determine if
141 any failures occurred.
150 Start packet forwarding with current configuration::
157 Start packet forwarding with current configuration after sending specified number of bursts of packets::
159 testpmd> start tx_first (""|burst_num)
161 The default burst number is 1 when ``burst_num`` not presented.
166 Stop packet forwarding, and display accumulated statistics::
181 The functions in the following sections are used to display information about the
182 testpmd configuration or the NIC status.
187 Display information for a given port or all ports::
189 testpmd> show port (info|stats|xstats|fdir|stat_qmap|dcb_tc|cap) (port_id|all)
191 The available information categories are:
193 * ``info``: General port information such as MAC address.
195 * ``stats``: RX/TX statistics.
197 * ``xstats``: RX/TX extended NIC statistics.
199 * ``fdir``: Flow Director information and statistics.
201 * ``stat_qmap``: Queue statistics mapping.
203 * ``dcb_tc``: DCB information such as TC mapping.
205 * ``cap``: Supported offload capabilities.
209 .. code-block:: console
211 testpmd> show port info 0
213 ********************* Infos for port 0 *********************
215 MAC address: XX:XX:XX:XX:XX:XX
217 memory allocation on the socket: 0
219 Link speed: 40000 Mbps
220 Link duplex: full-duplex
221 Promiscuous mode: enabled
222 Allmulticast mode: disabled
223 Maximum number of MAC addresses: 64
224 Maximum number of MAC addresses of hash filtering: 0
229 Redirection table size: 512
230 Supported flow types:
250 Display the rss redirection table entry indicated by masks on port X::
252 testpmd> show port (port_id) rss reta (size) (mask0, mask1...)
254 size is used to indicate the hardware supported reta size
259 Display the RSS hash functions and RSS hash key of a port::
261 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]
266 Clear the port statistics for a given port or for all ports::
268 testpmd> clear port (info|stats|xstats|fdir|stat_qmap) (port_id|all)
272 testpmd> clear port stats all
277 Display information for a given port's RX/TX queue::
279 testpmd> show (rxq|txq) info (port_id) (queue_id)
284 Displays the configuration of the application.
285 The configuration comes from the command-line, the runtime or the application defaults::
287 testpmd> show config (rxtx|cores|fwd|txpkts)
289 The available information categories are:
291 * ``rxtx``: RX/TX configuration items.
293 * ``cores``: List of forwarding cores.
295 * ``fwd``: Packet forwarding configuration.
297 * ``txpkts``: Packets to TX configuration.
301 .. code-block:: console
303 testpmd> show config rxtx
305 io packet forwarding - CRC stripping disabled - packets/burst=16
306 nb forwarding cores=2 - nb forwarding ports=1
307 RX queues=1 - RX desc=128 - RX free threshold=0
308 RX threshold registers: pthresh=8 hthresh=8 wthresh=4
309 TX queues=1 - TX desc=512 - TX free threshold=0
310 TX threshold registers: pthresh=36 hthresh=0 wthresh=0
311 TX RS bit threshold=0 - TXQ flags=0x0
316 Set the packet forwarding mode::
318 testpmd> set fwd (io|mac|macswap|flowgen| \
319 rxonly|txonly|csum|icmpecho) (""|retry)
321 ``retry`` can be specified for forwarding engines except ``rx_only``.
323 The available information categories are:
325 * ``io``: Forwards packets "as-is" in I/O mode.
326 This is the fastest possible forwarding operation as it does not access packets data.
327 This is the default mode.
329 * ``mac``: Changes the source and the destination Ethernet addresses of packets before forwarding them.
330 Default application behaviour is to set source Ethernet address to that of the transmitting interface, and destination
331 address to a dummy value (set during init). The user may specify a target destination Ethernet address via the 'eth-peer' or
332 'eth-peer-configfile' command-line options. It is not currently possible to specify a specific source Ethernet address.
334 * ``macswap``: MAC swap forwarding mode.
335 Swaps the source and the destination Ethernet addresses of packets before forwarding them.
337 * ``flowgen``: Multi-flow generation mode.
338 Originates a number of flows (with varying destination IP addresses), and terminate receive traffic.
340 * ``rxonly``: Receives packets but doesn't transmit them.
342 * ``txonly``: Generates and transmits packets without receiving any.
344 * ``csum``: Changes the checksum field with hardware or software methods depending on the offload flags on the packet.
346 * ``icmpecho``: Receives a burst of packets, lookup for IMCP echo requests and, if any, send back ICMP echo replies.
348 * ``ieee1588``: Demonstrate L2 IEEE1588 V2 PTP timestamping for RX and TX. Requires ``CONFIG_RTE_LIBRTE_IEEE1588=y``.
350 * ``tm``: Traffic Management forwarding mode
351 Demonstrates the use of ethdev traffic management APIs and softnic PMD for
352 QoS traffic management. In this mode, 5-level hierarchical QoS scheduler is
353 available as an default option that can be enabled through CLI. The user can
354 also modify the default hierarchy or specify the new hierarchy through CLI for
355 implementing QoS scheduler. Requires ``CONFIG_RTE_LIBRTE_PMD_SOFTNIC=y`` ``CONFIG_RTE_LIBRTE_SCHED=y``.
357 Note: TX timestamping is only available in the "Full Featured" TX path. To force ``testpmd`` into this mode set ``--txqflags=0``.
361 testpmd> set fwd rxonly
363 Set rxonly packet forwarding mode
369 Display an RX descriptor for a port RX queue::
371 testpmd> read rxd (port_id) (queue_id) (rxd_id)
375 testpmd> read rxd 0 0 4
376 0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180
381 Display a TX descriptor for a port TX queue::
383 testpmd> read txd (port_id) (queue_id) (txd_id)
387 testpmd> read txd 0 0 4
388 0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C
393 Get loaded dynamic device personalization (DDP) package info list::
395 testpmd> ddp get list (port_id)
400 Display information about dynamic device personalization (DDP) profile::
402 testpmd> ddp get info (profile_path)
407 Display VF statistics::
409 testpmd> show vf stats (port_id) (vf_id)
414 Reset VF statistics::
416 testpmd> clear vf stats (port_id) (vf_id)
418 show port pctype mapping
419 ~~~~~~~~~~~~~~~~~~~~~~~~
421 List all items from the pctype mapping table::
423 testpmd> show port (port_id) pctype mapping
426 Configuration Functions
427 -----------------------
429 The testpmd application can be configured from the runtime as well as from the command-line.
431 This section details the available configuration functions that are available.
435 Configuration changes only become active when forwarding is started/restarted.
440 Reset forwarding to the default configuration::
447 Set the debug verbosity level::
449 testpmd> set verbose (level)
451 Currently the only available levels are 0 (silent except for error) and 1 (fully verbose).
456 Set the number of ports used by the application:
460 This is equivalent to the ``--nb-ports`` command-line option.
465 Set the number of cores used by the application::
467 testpmd> set nbcore (num)
469 This is equivalent to the ``--nb-cores`` command-line option.
473 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port.
478 Set the forwarding cores hexadecimal mask::
480 testpmd> set coremask (mask)
482 This is equivalent to the ``--coremask`` command-line option.
486 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding.
491 Set the forwarding ports hexadecimal mask::
493 testpmd> set portmask (mask)
495 This is equivalent to the ``--portmask`` command-line option.
500 Set number of packets per burst::
502 testpmd> set burst (num)
504 This is equivalent to the ``--burst command-line`` option.
506 When retry is enabled, the transmit delay time and number of retries can also be set::
508 testpmd> set burst tx delay (microseconds) retry (num)
513 Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode::
515 testpmd> set txpkts (x[,y]*)
517 Where x[,y]* represents a CSV list of values, without white space.
522 Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes::
524 testpmd> set txsplit (off|on|rand)
528 * ``off`` disable packet copy & split for CSUM mode.
530 * ``on`` split outgoing packet into multiple segments. Size of each segment
531 and number of segments per packet is determined by ``set txpkts`` command
534 * ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments.
539 Set the list of forwarding cores::
541 testpmd> set corelist (x[,y]*)
543 For example, to change the forwarding cores:
545 .. code-block:: console
547 testpmd> set corelist 3,1
548 testpmd> show config fwd
550 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled
551 Logical Core 3 (socket 0) forwards packets on 1 streams:
552 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01
553 Logical Core 1 (socket 0) forwards packets on 1 streams:
554 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
558 The cores are used in the same order as specified on the command line.
563 Set the list of forwarding ports::
565 testpmd> set portlist (x[,y]*)
567 For example, to change the port forwarding:
569 .. code-block:: console
571 testpmd> set portlist 0,2,1,3
572 testpmd> show config fwd
574 io packet forwarding - ports=4 - cores=1 - streams=4
575 Logical Core 3 (socket 0) forwards packets on 4 streams:
576 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01
577 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
578 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03
579 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02
584 Enable/disable tx loopback::
586 testpmd> set tx loopback (port_id) (on|off)
591 set drop enable bit for all queues::
593 testpmd> set all queues drop (port_id) (on|off)
595 set split drop enable (for VF)
596 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
598 set split drop enable bit for VF from PF::
600 testpmd> set vf split drop (port_id) (vf_id) (on|off)
602 set mac antispoof (for VF)
603 ~~~~~~~~~~~~~~~~~~~~~~~~~~
605 Set mac antispoof for a VF from the PF::
607 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off)
612 Enable/disable MACsec offload::
614 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off)
615 testpmd> set macsec offload (port_id) off
620 Configure MACsec secure connection (SC)::
622 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi)
626 The pi argument is ignored for tx.
627 Check the NIC Datasheet for hardware limits.
632 Configure MACsec secure association (SA)::
634 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key)
638 The IDX value must be 0 or 1.
639 Check the NIC Datasheet for hardware limits.
641 set broadcast mode (for VF)
642 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
644 Set broadcast mode for a VF from the PF::
646 testpmd> set vf broadcast (port_id) (vf_id) (on|off)
651 Set the VLAN strip on a port::
653 testpmd> vlan set strip (on|off) (port_id)
658 Set the VLAN strip for a queue on a port::
660 testpmd> vlan set stripq (on|off) (port_id,queue_id)
662 vlan set stripq (for VF)
663 ~~~~~~~~~~~~~~~~~~~~~~~~
665 Set VLAN strip for all queues in a pool for a VF from the PF::
667 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off)
669 vlan set insert (for VF)
670 ~~~~~~~~~~~~~~~~~~~~~~~~
672 Set VLAN insert for a VF from the PF::
674 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id)
676 vlan set tag (for VF)
677 ~~~~~~~~~~~~~~~~~~~~~
679 Set VLAN tag for a VF from the PF::
681 testpmd> set vf vlan tag (port_id) (vf_id) (on|off)
683 vlan set antispoof (for VF)
684 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
686 Set VLAN antispoof for a VF from the PF::
688 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off)
693 Set the VLAN filter on a port::
695 testpmd> vlan set filter (on|off) (port_id)
700 Set the VLAN QinQ (extended queue in queue) on for a port::
702 testpmd> vlan set qinq (on|off) (port_id)
707 Set the inner or outer VLAN TPID for packet filtering on a port::
709 testpmd> vlan set (inner|outer) tpid (value) (port_id)
713 TPID value must be a 16-bit number (value <= 65536).
718 Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID::
720 testpmd> rx_vlan add (vlan_id|all) (port_id)
724 VLAN filter must be set on that port. VLAN ID < 4096.
725 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries
726 in VFTA table. This is important if enabling all vlan_ids.
731 Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID::
733 testpmd> rx_vlan rm (vlan_id|all) (port_id)
738 Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID::
740 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask)
745 Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID::
747 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask)
752 Add a tunnel filter on a port::
754 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \
755 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
756 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
758 The available information categories are:
760 * ``vxlan``: Set tunnel type as VXLAN.
762 * ``nvgre``: Set tunnel type as NVGRE.
764 * ``ipingre``: Set tunnel type as IP-in-GRE.
766 * ``imac-ivlan``: Set filter type as Inner MAC and VLAN.
768 * ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID.
770 * ``imac-tenid``: Set filter type as Inner MAC and tenant ID.
772 * ``imac``: Set filter type as Inner MAC.
774 * ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID.
776 * ``oip``: Set filter type as Outer IP.
778 * ``iip``: Set filter type as Inner IP.
782 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \
783 192.168.2.2 0 ipingre oip 1 1
785 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP.
790 Remove a tunnel filter on a port::
792 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \
793 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
794 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
799 Add an UDP port for VXLAN packet filter on a port::
801 testpmd> rx_vxlan_port add (udp_port) (port_id)
806 Remove an UDP port for VXLAN packet filter on a port::
808 testpmd> rx_vxlan_port rm (udp_port) (port_id)
813 Set hardware insertion of VLAN IDs in packets sent on a port::
815 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer]
817 For example, set a single VLAN ID (5) insertion on port 0::
821 Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1::
829 Set port based hardware insertion of VLAN ID in packets sent on a port::
831 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off)
836 Disable hardware insertion of a VLAN header in packets sent on a port::
838 testpmd> tx_vlan reset (port_id)
843 Select hardware or software calculation of the checksum when
844 transmitting a packet using the ``csum`` forwarding engine::
846 testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id)
850 * ``ip|udp|tcp|sctp`` always relate to the inner layer.
852 * ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized
853 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are
854 supported). See also the ``csum parse-tunnel`` command.
858 Check the NIC Datasheet for hardware limits.
863 Define how tunneled packets should be handled by the csum forward
866 testpmd> csum parse-tunnel (on|off) (tx_port_id)
868 If enabled, the csum forward engine will try to recognize supported
869 tunnel headers (vxlan, gre, ipip).
871 If disabled, treat tunnel packets as non-tunneled packets (a inner
872 header is handled as a packet payload).
876 The port argument is the TX port like in the ``csum set`` command.
880 Consider a packet in packet like the following::
882 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in
884 * If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
885 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the
886 ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``).
888 * If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
889 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``.
894 Display tx checksum offload configuration::
896 testpmd> csum show (port_id)
901 Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine::
903 testpmd> tso set (segsize) (port_id)
907 Check the NIC datasheet for hardware limits.
912 Display the status of TCP Segmentation Offload::
914 testpmd> tso show (port_id)
919 Enable or disable GRO in ``csum`` forwarding engine::
921 testpmd> set port <port_id> gro on|off
923 If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4
924 packets received from the given port.
926 If disabled, packets received from the given port won't be performed
927 GRO. By default, GRO is disabled for all ports.
931 When enable GRO for a port, TCP/IPv4 packets received from the port
932 will be performed GRO. After GRO, all merged packets have bad
933 checksums, since the GRO library doesn't re-calculate checksums for
934 the merged packets. Therefore, if users want the merged packets to
935 have correct checksums, please select HW IP checksum calculation and
936 HW TCP checksum calculation for the port which the merged packets are
942 Display GRO configuration for a given port::
944 testpmd> show port <port_id> gro
949 Set the cycle to flush the GROed packets from reassembly tables::
951 testpmd> set gro flush <cycles>
953 When enable GRO, the csum forwarding engine performs GRO on received
954 packets, and the GROed packets are stored in reassembly tables. Users
955 can use this command to determine when the GROed packets are flushed
956 from the reassembly tables.
958 The ``cycles`` is measured in GRO operation times. The csum forwarding
959 engine flushes the GROed packets from the tables every ``cycles`` GRO
962 By default, the value of ``cycles`` is 1, which means flush GROed packets
963 from the reassembly tables as soon as one GRO operation finishes. The value
964 of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``.
966 Please note that the large value of ``cycles`` may cause the poor TCP/IP
967 stack performance. Because the GROed packets are delayed to arrive the
968 stack, thus causing more duplicated ACKs and TCP retransmissions.
973 Toggle per-port GSO support in ``csum`` forwarding engine::
975 testpmd> set port <port_id> gso on|off
977 If enabled, the csum forwarding engine will perform GSO on supported IPv4
978 packets, transmitted on the given port.
980 If disabled, packets transmitted on the given port will not undergo GSO.
981 By default, GSO is disabled for all ports.
985 When GSO is enabled on a port, supported IPv4 packets transmitted on that
986 port undergo GSO. Afterwards, the segmented packets are represented by
987 multi-segment mbufs; however, the csum forwarding engine doesn't calculation
988 of checksums for GSO'd segments in SW. As a result, if users want correct
989 checksums in GSO segments, they should enable HW checksum calculation for
992 For example, HW checksum calculation for VxLAN GSO'd packets may be enabled
993 by setting the following options in the csum forwarding engine:
995 testpmd> csum set outer_ip hw <port_id>
997 testpmd> csum set ip hw <port_id>
999 testpmd> csum set tcp hw <port_id>
1004 Set the maximum GSO segment size (measured in bytes), which includes the
1005 packet header and the packet payload for GSO-enabled ports (global)::
1007 testpmd> set gso segsz <length>
1012 Display the status of Generic Segmentation Offload for a given port::
1014 testpmd> show port <port_id> gso
1019 Add an alternative MAC address to a port::
1021 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX)
1026 Remove a MAC address from a port::
1028 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX)
1030 mac_addr add (for VF)
1031 ~~~~~~~~~~~~~~~~~~~~~
1033 Add an alternative MAC address for a VF to a port::
1035 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX)
1040 Set the default MAC address for a port::
1042 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX)
1044 mac_addr set (for VF)
1045 ~~~~~~~~~~~~~~~~~~~~~
1047 Set the MAC address for a VF from the PF::
1049 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX)
1054 Set the unicast hash filter(s) on/off for a port::
1056 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
1061 Set the promiscuous mode on for a port or for all ports.
1062 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1064 testpmd> set promisc (port_id|all) (on|off)
1069 Set the allmulti mode for a port or for all ports::
1071 testpmd> set allmulti (port_id|all) (on|off)
1073 Same as the ifconfig (8) option. Controls how multicast packets are handled.
1075 set promisc (for VF)
1076 ~~~~~~~~~~~~~~~~~~~~
1078 Set the unicast promiscuous mode for a VF from PF.
1079 It's supported by Intel i40e NICs now.
1080 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1082 testpmd> set vf promisc (port_id) (vf_id) (on|off)
1084 set allmulticast (for VF)
1085 ~~~~~~~~~~~~~~~~~~~~~~~~~
1087 Set the multicast promiscuous mode for a VF from PF.
1088 It's supported by Intel i40e NICs now.
1089 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1091 testpmd> set vf allmulti (port_id) (vf_id) (on|off)
1093 set tx max bandwidth (for VF)
1094 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1096 Set TX max absolute bandwidth (Mbps) for a VF from PF::
1098 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth)
1100 set tc tx min bandwidth (for VF)
1101 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1103 Set all TCs' TX min relative bandwidth (%) for a VF from PF::
1105 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...)
1107 set tc tx max bandwidth (for VF)
1108 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1110 Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF::
1112 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth)
1114 set tc strict link priority mode
1115 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1117 Set some TCs' strict link priority mode on a physical port::
1119 testpmd> set tx strict-link-priority (port_id) (tc_bitmap)
1121 set tc tx min bandwidth
1122 ~~~~~~~~~~~~~~~~~~~~~~~
1124 Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs::
1126 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...)
1131 Set the link flow control parameter on a port::
1133 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1134 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
1135 autoneg (on|off) (port_id)
1139 * ``high_water`` (integer): High threshold value to trigger XOFF.
1141 * ``low_water`` (integer): Low threshold value to trigger XON.
1143 * ``pause_time`` (integer): Pause quota in the Pause frame.
1145 * ``send_xon`` (0/1): Send XON frame.
1147 * ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
1149 * ``autoneg``: Change the auto-negotiation parameter.
1154 Set the priority flow control parameter on a port::
1156 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1157 (pause_time) (priority) (port_id)
1161 * ``high_water`` (integer): High threshold value.
1163 * ``low_water`` (integer): Low threshold value.
1165 * ``pause_time`` (integer): Pause quota in the Pause frame.
1167 * ``priority`` (0-7): VLAN User Priority.
1172 Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
1174 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
1176 For example, to set rx queue 2 on port 0 to mapping 5::
1178 testpmd>set stat_qmap rx 0 2 5
1180 set port - rx/tx (for VF)
1181 ~~~~~~~~~~~~~~~~~~~~~~~~~
1183 Set VF receive/transmit from a port::
1185 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
1187 set port - mac address filter (for VF)
1188 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1190 Add/Remove unicast or multicast MAC addr filter for a VF::
1192 testpmd> set port (port_id) vf (vf_id) (mac_addr) \
1193 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
1195 set port - rx mode(for VF)
1196 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1198 Set the VF receive mode of a port::
1200 testpmd> set port (port_id) vf (vf_id) \
1201 rxmode (AUPE|ROPE|BAM|MPE) (on|off)
1203 The available receive modes are:
1205 * ``AUPE``: Accepts untagged VLAN.
1207 * ``ROPE``: Accepts unicast hash.
1209 * ``BAM``: Accepts broadcast packets.
1211 * ``MPE``: Accepts all multicast packets.
1213 set port - tx_rate (for Queue)
1214 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1216 Set TX rate limitation for a queue on a port::
1218 testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1220 set port - tx_rate (for VF)
1221 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
1223 Set TX rate limitation for queues in VF on a port::
1225 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1227 set port - mirror rule
1228 ~~~~~~~~~~~~~~~~~~~~~~
1230 Set pool or vlan type mirror rule for a port::
1232 testpmd> set port (port_id) mirror-rule (rule_id) \
1233 (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1234 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1236 Set link mirror rule for a port::
1238 testpmd> set port (port_id) mirror-rule (rule_id) \
1239 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1241 For example to enable mirror traffic with vlan 0,1 to pool 0::
1243 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1245 reset port - mirror rule
1246 ~~~~~~~~~~~~~~~~~~~~~~~~
1248 Reset a mirror rule for a port::
1250 testpmd> reset port (port_id) mirror-rule (rule_id)
1255 Set the flush on RX streams before forwarding.
1256 The default is flush ``on``.
1257 Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1259 testpmd> set flush_rx off
1264 Set the bypass mode for the lowest port on bypass enabled NIC::
1266 testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1271 Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1273 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1274 mode (normal|bypass|isolate) (port_id)
1278 * ``timeout``: Enable bypass after watchdog timeout.
1280 * ``os_on``: Enable bypass when OS/board is powered on.
1282 * ``os_off``: Enable bypass when OS/board is powered off.
1284 * ``power_on``: Enable bypass when power supply is turned on.
1286 * ``power_off``: Enable bypass when power supply is turned off.
1292 Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1294 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1299 Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1301 testpmd> show bypass config (port_id)
1306 Set link up for a port::
1308 testpmd> set link-up port (port id)
1313 Set link down for a port::
1315 testpmd> set link-down port (port id)
1320 Enable E-tag insertion for a VF on a port::
1322 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1324 Disable E-tag insertion for a VF on a port::
1326 testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1328 Enable/disable E-tag stripping on a port::
1330 testpmd> E-tag set stripping (on|off) port (port_id)
1332 Enable/disable E-tag based forwarding on a port::
1334 testpmd> E-tag set forwarding (on|off) port (port_id)
1336 Add an E-tag forwarding filter on a port::
1338 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1340 Delete an E-tag forwarding filter on a port::
1341 testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1346 Load a dynamic device personalization (DDP) package::
1348 testpmd> ddp add (port_id) (package_path[,output_path])
1353 Delete a dynamic device personalization package::
1355 testpmd> ddp del (port_id) (package_path)
1360 List all items from the ptype mapping table::
1362 testpmd> ptype mapping get (port_id) (valid_only)
1366 * ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0).
1368 Replace a specific or a group of software defined ptype with a new one::
1370 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type)
1374 * ``target``: A specific software ptype or a mask to represent a group of software ptypes.
1376 * ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1).
1378 * ``pkt_type``: The new software ptype to replace the old ones.
1380 Update hardware defined ptype to software defined packet type mapping table::
1382 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype)
1386 * ``hw_ptype``: hardware ptype as the index of the ptype mapping table.
1388 * ``sw_ptype``: software ptype as the value of the ptype mapping table.
1390 Reset ptype mapping table::
1392 testpmd> ptype mapping reset (port_id)
1397 The following sections show functions for configuring ports.
1401 Port configuration changes only become active when forwarding is started/restarted.
1406 Attach a port specified by pci address or virtual device args::
1408 testpmd> port attach (identifier)
1410 To attach a new pci device, the device should be recognized by kernel first.
1411 Then it should be moved under DPDK management.
1412 Finally the port can be attached to testpmd.
1414 For example, to move a pci device using ixgbe under DPDK management:
1416 .. code-block:: console
1418 # Check the status of the available devices.
1419 ./usertools/dpdk-devbind.py --status
1421 Network devices using DPDK-compatible driver
1422 ============================================
1425 Network devices using kernel driver
1426 ===================================
1427 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1430 # Bind the device to igb_uio.
1431 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1434 # Recheck the status of the devices.
1435 ./usertools/dpdk-devbind.py --status
1436 Network devices using DPDK-compatible driver
1437 ============================================
1438 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1440 To attach a port created by virtual device, above steps are not needed.
1442 For example, to attach a port whose pci address is 0000:0a:00.0.
1444 .. code-block:: console
1446 testpmd> port attach 0000:0a:00.0
1447 Attaching a new port...
1448 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1449 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1450 EAL: PCI memory mapped at 0x7f83bfa00000
1451 EAL: PCI memory mapped at 0x7f83bfa80000
1452 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1453 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1454 Port 0 is attached. Now total ports is 1
1457 For example, to attach a port created by pcap PMD.
1459 .. code-block:: console
1461 testpmd> port attach net_pcap0
1462 Attaching a new port...
1463 PMD: Initializing pmd_pcap for net_pcap0
1464 PMD: Creating pcap-backed ethdev on numa socket 0
1465 Port 0 is attached. Now total ports is 1
1468 In this case, identifier is ``net_pcap0``.
1469 This identifier format is the same as ``--vdev`` format of DPDK applications.
1471 For example, to re-attach a bonded port which has been previously detached,
1472 the mode and slave parameters must be given.
1474 .. code-block:: console
1476 testpmd> port attach net_bond_0,mode=0,slave=1
1477 Attaching a new port...
1478 EAL: Initializing pmd_bond for net_bond_0
1479 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1480 Port 0 is attached. Now total ports is 1
1487 Detach a specific port::
1489 testpmd> port detach (port_id)
1491 Before detaching a port, the port should be stopped and closed.
1493 For example, to detach a pci device port 0.
1495 .. code-block:: console
1497 testpmd> port stop 0
1500 testpmd> port close 0
1504 testpmd> port detach 0
1506 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1507 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1508 EAL: PCI memory unmapped at 0x7f83bfa00000
1509 EAL: PCI memory unmapped at 0x7f83bfa80000
1513 For example, to detach a virtual device port 0.
1515 .. code-block:: console
1517 testpmd> port stop 0
1520 testpmd> port close 0
1524 testpmd> port detach 0
1526 PMD: Closing pcap ethdev on numa socket 0
1527 Port 'net_pcap0' is detached. Now total ports is 0
1530 To remove a pci device completely from the system, first detach the port from testpmd.
1531 Then the device should be moved under kernel management.
1532 Finally the device can be removed using kernel pci hotplug functionality.
1534 For example, to move a pci device under kernel management:
1536 .. code-block:: console
1538 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1540 ./usertools/dpdk-devbind.py --status
1542 Network devices using DPDK-compatible driver
1543 ============================================
1546 Network devices using kernel driver
1547 ===================================
1548 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1550 To remove a port created by a virtual device, above steps are not needed.
1555 Start all ports or a specific port::
1557 testpmd> port start (port_id|all)
1562 Stop all ports or a specific port::
1564 testpmd> port stop (port_id|all)
1569 Close all ports or a specific port::
1571 testpmd> port close (port_id|all)
1573 port start/stop queue
1574 ~~~~~~~~~~~~~~~~~~~~~
1576 Start/stop a rx/tx queue on a specific port::
1578 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1580 Only take effect when port is started.
1585 Set the speed and duplex mode for all ports or a specific port::
1587 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1588 duplex (half|full|auto)
1590 port config - queues/descriptors
1591 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1593 Set number of queues/descriptors for rxq, txq, rxd and txd::
1595 testpmd> port config all (rxq|txq|rxd|txd) (value)
1597 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1599 port config - max-pkt-len
1600 ~~~~~~~~~~~~~~~~~~~~~~~~~
1602 Set the maximum packet length::
1604 testpmd> port config all max-pkt-len (value)
1606 This is equivalent to the ``--max-pkt-len`` command-line option.
1608 port config - CRC Strip
1609 ~~~~~~~~~~~~~~~~~~~~~~~
1611 Set hardware CRC stripping on or off for all ports::
1613 testpmd> port config all crc-strip (on|off)
1615 CRC stripping is on by default.
1617 The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option.
1619 port config - scatter
1620 ~~~~~~~~~~~~~~~~~~~~~~~
1622 Set RX scatter mode on or off for all ports::
1624 testpmd> port config all scatter (on|off)
1626 RX scatter mode is off by default.
1628 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1630 port config - TX queue flags
1631 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1633 Set a hexadecimal bitmap of TX queue flags for all ports::
1635 testpmd> port config all txqflags value
1637 This command is equivalent to the ``--txqflags`` command-line option.
1639 port config - RX Checksum
1640 ~~~~~~~~~~~~~~~~~~~~~~~~~
1642 Set hardware RX checksum offload to on or off for all ports::
1644 testpmd> port config all rx-cksum (on|off)
1646 Checksum offload is off by default.
1648 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1653 Set hardware VLAN on or off for all ports::
1655 testpmd> port config all hw-vlan (on|off)
1657 Hardware VLAN is on by default.
1659 The ``off`` option is equivalent to the ``--disable-hw-vlan`` command-line option.
1661 port config - VLAN filter
1662 ~~~~~~~~~~~~~~~~~~~~~~~~~
1664 Set hardware VLAN filter on or off for all ports::
1666 testpmd> port config all hw-vlan-filter (on|off)
1668 Hardware VLAN filter is on by default.
1670 The ``off`` option is equivalent to the ``--disable-hw-vlan-filter`` command-line option.
1672 port config - VLAN strip
1673 ~~~~~~~~~~~~~~~~~~~~~~~~
1675 Set hardware VLAN strip on or off for all ports::
1677 testpmd> port config all hw-vlan-strip (on|off)
1679 Hardware VLAN strip is on by default.
1681 The ``off`` option is equivalent to the ``--disable-hw-vlan-strip`` command-line option.
1683 port config - VLAN extend
1684 ~~~~~~~~~~~~~~~~~~~~~~~~~
1686 Set hardware VLAN extend on or off for all ports::
1688 testpmd> port config all hw-vlan-extend (on|off)
1690 Hardware VLAN extend is off by default.
1692 The ``off`` option is equivalent to the ``--disable-hw-vlan-extend`` command-line option.
1694 port config - Drop Packets
1695 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1697 Set packet drop for packets with no descriptors on or off for all ports::
1699 testpmd> port config all drop-en (on|off)
1701 Packet dropping for packets with no descriptors is off by default.
1703 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1708 Set the RSS (Receive Side Scaling) mode on or off::
1710 testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1712 RSS is on by default.
1714 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1716 port config - RSS Reta
1717 ~~~~~~~~~~~~~~~~~~~~~~
1719 Set the RSS (Receive Side Scaling) redirection table::
1721 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1726 Set the DCB mode for an individual port::
1728 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1730 The traffic class should be 4 or 8.
1735 Set the number of packets per burst::
1737 testpmd> port config all burst (value)
1739 This is equivalent to the ``--burst`` command-line option.
1741 port config - Threshold
1742 ~~~~~~~~~~~~~~~~~~~~~~~
1744 Set thresholds for TX/RX queues::
1746 testpmd> port config all (threshold) (value)
1748 Where the threshold type can be:
1750 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1752 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1754 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1756 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1758 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1760 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1762 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1764 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1766 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1768 These threshold options are also available from the command-line.
1773 Set the value of ether-type for E-tag::
1775 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1777 Enable/disable the E-tag support::
1779 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1781 port config pctype mapping
1782 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1784 Reset pctype mapping table::
1786 testpmd> port config (port_id) pctype mapping reset
1788 Update hardware defined pctype to software defined flow type mapping table::
1790 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id)
1794 * ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table.
1796 * ``flow_type_id``: software flow type id as the index of the pctype mapping table.
1799 Link Bonding Functions
1800 ----------------------
1802 The Link Bonding functions make it possible to dynamically create and
1803 manage link bonding devices from within testpmd interactive prompt.
1805 create bonded device
1806 ~~~~~~~~~~~~~~~~~~~~
1808 Create a new bonding device::
1810 testpmd> create bonded device (mode) (socket)
1812 For example, to create a bonded device in mode 1 on socket 0::
1814 testpmd> create bonded 1 0
1815 created new bonded device (port X)
1820 Adds Ethernet device to a Link Bonding device::
1822 testpmd> add bonding slave (slave id) (port id)
1824 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
1826 testpmd> add bonding slave 6 10
1829 remove bonding slave
1830 ~~~~~~~~~~~~~~~~~~~~
1832 Removes an Ethernet slave device from a Link Bonding device::
1834 testpmd> remove bonding slave (slave id) (port id)
1836 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
1838 testpmd> remove bonding slave 6 10
1843 Set the Link Bonding mode of a Link Bonding device::
1845 testpmd> set bonding mode (value) (port id)
1847 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
1849 testpmd> set bonding mode 3 10
1854 Set an Ethernet slave device as the primary device on a Link Bonding device::
1856 testpmd> set bonding primary (slave id) (port id)
1858 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
1860 testpmd> set bonding primary 6 10
1865 Set the MAC address of a Link Bonding device::
1867 testpmd> set bonding mac (port id) (mac)
1869 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
1871 testpmd> set bonding mac 10 00:00:00:00:00:01
1873 set bonding xmit_balance_policy
1874 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1876 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
1878 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
1880 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
1882 testpmd> set bonding xmit_balance_policy 10 l34
1885 set bonding mon_period
1886 ~~~~~~~~~~~~~~~~~~~~~~
1888 Set the link status monitoring polling period in milliseconds for a bonding device.
1890 This adds support for PMD slave devices which do not support link status interrupts.
1891 When the mon_period is set to a value greater than 0 then all PMD's which do not support
1892 link status ISR will be queried every polling interval to check if their link status has changed::
1894 testpmd> set bonding mon_period (port_id) (value)
1896 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
1898 testpmd> set bonding mon_period 5 150
1901 set bonding lacp dedicated_queue
1902 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1904 Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic
1905 when in mode 4 (link-aggregration-802.3ad)::
1907 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable)
1910 set bonding agg_mode
1911 ~~~~~~~~~~~~~~~~~~~~
1913 Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad)::
1915 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable)
1921 Show the current configuration of a Link Bonding device::
1923 testpmd> show bonding config (port id)
1926 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
1927 in balance mode with a transmission policy of layer 2+3::
1929 testpmd> show bonding config 9
1931 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
1933 Active Slaves (3): [1 3 4]
1940 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
1941 This is mainly useful for debugging purposes.
1942 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
1943 and fields that can be accessed.
1948 Display the value of a port register::
1950 testpmd> read reg (port_id) (address)
1952 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
1954 testpmd> read reg 0 0xEE00
1955 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
1960 Display a port register bit field::
1962 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
1964 For example, reading the lowest two bits from the register in the example above::
1966 testpmd> read regfield 0 0xEE00 0 1
1967 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
1972 Display a single port register bit::
1974 testpmd> read regbit (port_id) (address) (bit_x)
1976 For example, reading the lowest bit from the register in the example above::
1978 testpmd> read regbit 0 0xEE00 0
1979 port 0 PCI register at offset 0xEE00: bit 0=1
1984 Set the value of a port register::
1986 testpmd> write reg (port_id) (address) (value)
1988 For example, to clear a register::
1990 testpmd> write reg 0 0xEE00 0x0
1991 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
1996 Set bit field of a port register::
1998 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
2000 For example, writing to the register cleared in the example above::
2002 testpmd> write regfield 0 0xEE00 0 1 2
2003 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
2008 Set single bit value of a port register::
2010 testpmd> write regbit (port_id) (address) (bit_x) (value)
2012 For example, to set the high bit in the register from the example above::
2014 testpmd> write regbit 0 0xEE00 31 1
2015 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
2021 The following section shows functions for configuring traffic management on
2022 on the ethernet device through the use of generic TM API.
2024 show port traffic management capability
2025 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2027 Show traffic management capability of the port::
2029 testpmd> show port tm cap (port_id)
2031 show port traffic management capability (hierarchy level)
2032 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2034 Show traffic management hierarchy level capability of the port::
2036 testpmd> show port tm cap (port_id) (level_id)
2038 show port traffic management capability (hierarchy node level)
2039 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2041 Show the traffic management hierarchy node capability of the port::
2043 testpmd> show port tm cap (port_id) (node_id)
2045 show port traffic management hierarchy node type
2046 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2048 Show the port traffic management hierarchy node type::
2050 testpmd> show port tm node type (port_id) (node_id)
2052 show port traffic management hierarchy node stats
2053 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2055 Show the port traffic management hierarchy node statistics::
2057 testpmd> show port tm node stats (port_id) (node_id) (clear)
2061 * ``clear``: When this parameter has a non-zero value, the statistics counters
2062 are cleared (i.e. set to zero) immediately after they have been read,
2063 otherwise the statistics counters are left untouched.
2065 Add port traffic management private shaper profile
2066 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2068 Add the port traffic management private shaper profile::
2070 testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \
2071 (tb_rate) (tb_size) (packet_length_adjust)
2075 * ``shaper_profile id``: Shaper profile ID for the new profile.
2076 * ``tb_rate``: Token bucket rate (bytes per second).
2077 * ``tb_size``: Token bucket size (bytes).
2078 * ``packet_length_adjust``: The value (bytes) to be added to the length of
2079 each packet for the purpose of shaping. This parameter value can be used to
2080 correct the packet length with the framing overhead bytes that are consumed
2083 Delete port traffic management private shaper profile
2084 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2086 Delete the port traffic management private shaper::
2088 testpmd> del port tm node shaper profile (port_id) (shaper_profile_id)
2092 * ``shaper_profile id``: Shaper profile ID that needs to be deleted.
2094 Add port traffic management shared shaper
2095 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2097 Create the port traffic management shared shaper::
2099 testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \
2104 * ``shared_shaper_id``: Shared shaper ID to be created.
2105 * ``shaper_profile id``: Shaper profile ID for shared shaper.
2107 Set port traffic management shared shaper
2108 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2110 Update the port traffic management shared shaper::
2112 testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \
2117 * ``shared_shaper_id``: Shared shaper ID to be update.
2118 * ``shaper_profile id``: Shaper profile ID for shared shaper.
2120 Delete port traffic management shared shaper
2121 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2123 Delete the port traffic management shared shaper::
2125 testpmd> del port tm node shared shaper (port_id) (shared_shaper_id)
2129 * ``shared_shaper_id``: Shared shaper ID to be deleted.
2131 Set port traffic management hiearchy node private shaper
2132 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2134 set the port traffic management hierarchy node private shaper::
2136 testpmd> set port tm node shaper profile (port_id) (node_id) \
2141 * ``shaper_profile id``: Private shaper profile ID to be enabled on the
2144 Add port traffic management WRED profile
2145 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2147 Create a new WRED profile::
2149 testpmd> add port tm node wred profile (port_id) (wred_profile_id) \
2150 (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \
2151 (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \
2152 (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r)
2156 * ``wred_profile id``: Identifier for the newly create WRED profile
2157 * ``color_g``: Packet color (green)
2158 * ``min_th_g``: Minimum queue threshold for packet with green color
2159 * ``max_th_g``: Minimum queue threshold for packet with green color
2160 * ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp)
2161 * ``wq_log2_g``: Negated log2 of queue weight (wq)
2162 * ``color_y``: Packet color (yellow)
2163 * ``min_th_y``: Minimum queue threshold for packet with yellow color
2164 * ``max_th_y``: Minimum queue threshold for packet with yellow color
2165 * ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp)
2166 * ``wq_log2_y``: Negated log2 of queue weight (wq)
2167 * ``color_r``: Packet color (red)
2168 * ``min_th_r``: Minimum queue threshold for packet with yellow color
2169 * ``max_th_r``: Minimum queue threshold for packet with yellow color
2170 * ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp)
2171 * ``wq_log2_r``: Negated log2 of queue weight (wq)
2173 Delete port traffic management WRED profile
2174 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2176 Delete the WRED profile::
2178 testpmd> del port tm node wred profile (port_id) (wred_profile_id)
2180 Add port traffic management hierarchy nonleaf node
2181 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2183 Add nonleaf node to port traffic management hiearchy::
2185 testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \
2186 (priority) (weight) (level_id) (shaper_profile_id) \
2187 (shared_shaper_id) (n_shared_shapers) (n_sp_priorities) (stats_mask) \
2191 * ``parent_node_id``: Node ID of the parent.
2192 * ``priority``: Node priority (highest node priority is zero). This is used by
2193 the SP algorithm running on the parent node for scheduling this node.
2194 * ``weight``: Node weight (lowest weight is one). The node weight is relative
2195 to the weight sum of all siblings that have the same priority. It is used by
2196 the WFQ algorithm running on the parent node for scheduling this node.
2197 * ``level_id``: Hiearchy level of the node.
2198 * ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2200 * ``shared_shaper_id``: Shared shaper id.
2201 * ``n_shared_shapers``: Number of shared shapers.
2202 * ``n_sp_priorities``: Number of strict priorities.
2203 * ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2205 Add port traffic management hierarchy leaf node
2206 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2208 Add leaf node to port traffic management hiearchy::
2210 testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \
2211 (priority) (weight) (level_id) (shaper_profile_id) \
2212 (shared_shaper_id) (n_shared_shapers) (cman_mode) \
2213 (wred_profile_id) (stats_mask) \
2217 * ``parent_node_id``: Node ID of the parent.
2218 * ``priority``: Node priority (highest node priority is zero). This is used by
2219 the SP algorithm running on the parent node for scheduling this node.
2220 * ``weight``: Node weight (lowest weight is one). The node weight is relative
2221 to the weight sum of all siblings that have the same priority. It is used by
2222 the WFQ algorithm running on the parent node for scheduling this node.
2223 * ``level_id``: Hiearchy level of the node.
2224 * ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2226 * ``shared_shaper_id``: Shared shaper id.
2227 * ``n_shared_shapers``: Number of shared shapers.
2228 * ``cman_mode``: Congestion management mode to be enabled for this node.
2229 * ``wred_profile_id``: WRED profile id to be enabled for this node.
2230 * ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2232 Delete port traffic management hierarchy node
2233 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2235 Delete node from port traffic management hiearchy::
2237 testpmd> del port tm node (port_id) (node_id)
2239 Update port traffic management hierarchy parent node
2240 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2242 Update port traffic management hierarchy parent node::
2244 testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \
2247 This function can only be called after the hierarchy commit invocation. Its
2248 success depends on the port support for this operation, as advertised through
2249 the port capability set. This function is valid for all nodes of the traffic
2250 management hierarchy except root node.
2252 Commit port traffic management hierarchy
2253 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2255 Commit the traffic management hierarchy on the port::
2257 testpmd> port tm hierarchy commit (port_id) (clean_on_fail)
2261 * ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function
2262 call failure. On the other hand, hierarchy is preserved when this parameter
2265 Set port traffic management default hierarchy (tm forwarding mode)
2266 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2268 set the traffic management default hierarchy on the port::
2270 testpmd> set port tm hierarchy default (port_id)
2275 This section details the available filter functions that are available.
2277 Note these functions interface the deprecated legacy filtering framework,
2278 superseded by *rte_flow*. See `Flow rules management`_.
2281 ~~~~~~~~~~~~~~~~~~~~
2283 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
2285 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
2286 ethertype (ether_type) (drop|fwd) queue (queue_id)
2288 The available information parameters are:
2290 * ``port_id``: The port which the Ethertype filter assigned on.
2292 * ``mac_addr``: Compare destination mac address.
2294 * ``mac_ignr``: Ignore destination mac address match.
2296 * ``mac_address``: Destination mac address to match.
2298 * ``ether_type``: The EtherType value want to match,
2299 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
2301 * ``queue_id``: The receive queue associated with this EtherType filter.
2302 It is meaningless when deleting or dropping.
2304 Example, to add/remove an ethertype filter rule::
2306 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
2307 ethertype 0x0806 fwd queue 3
2309 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
2310 ethertype 0x0806 fwd queue 3
2315 Add or delete a 2-tuple filter,
2316 which identifies packets by specific protocol and destination TCP/UDP port
2317 and forwards packets into one of the receive queues::
2319 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
2320 protocol (protocol_value) mask (mask_value) \
2321 tcp_flags (tcp_flags_value) priority (prio_value) \
2324 The available information parameters are:
2326 * ``port_id``: The port which the 2-tuple filter assigned on.
2328 * ``dst_port_value``: Destination port in L4.
2330 * ``protocol_value``: IP L4 protocol.
2332 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
2334 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
2336 * ``prio_value``: Priority of this filter.
2338 * ``queue_id``: The receive queue associated with this 2-tuple filter.
2340 Example, to add/remove an 2tuple filter rule::
2342 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
2343 tcp_flags 0x02 priority 3 queue 3
2345 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
2346 tcp_flags 0x02 priority 3 queue 3
2351 Add or delete a 5-tuple filter,
2352 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
2353 and routes packets into one of the receive queues::
2355 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
2356 (src_address) dst_port (dst_port_value) \
2357 src_port (src_port_value) protocol (protocol_value) \
2358 mask (mask_value) tcp_flags (tcp_flags_value) \
2359 priority (prio_value) queue (queue_id)
2361 The available information parameters are:
2363 * ``port_id``: The port which the 5-tuple filter assigned on.
2365 * ``dst_address``: Destination IP address.
2367 * ``src_address``: Source IP address.
2369 * ``dst_port_value``: TCP/UDP destination port.
2371 * ``src_port_value``: TCP/UDP source port.
2373 * ``protocol_value``: L4 protocol.
2375 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
2377 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
2379 * ``prio_value``: The priority of this filter.
2381 * ``queue_id``: The receive queue associated with this 5-tuple filter.
2383 Example, to add/remove an 5tuple filter rule::
2385 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2386 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2387 flags 0x0 priority 3 queue 3
2389 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2390 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2391 flags 0x0 priority 3 queue 3
2396 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
2398 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
2400 The available information parameters are:
2402 * ``port_id``: The port which the SYN filter assigned on.
2404 * ``high``: This SYN filter has higher priority than other filters.
2406 * ``low``: This SYN filter has lower priority than other filters.
2408 * ``queue_id``: The receive queue associated with this SYN filter
2412 testpmd> syn_filter 0 add priority high queue 3
2417 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
2418 and routed into one of the receive queues::
2420 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
2421 mask (mask_value) priority (prio_value) queue (queue_id)
2423 The available information parameters are:
2425 * ``port_id``: The port which the Flex filter is assigned on.
2427 * ``len_value``: Filter length in bytes, no greater than 128.
2429 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
2431 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
2433 * ``prio_value``: The priority of this filter.
2435 * ``queue_id``: The receive queue associated with this Flex filter.
2439 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
2440 mask 000C priority 3 queue 3
2442 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
2443 mask 000C priority 3 queue 3
2446 .. _testpmd_flow_director:
2448 flow_director_filter
2449 ~~~~~~~~~~~~~~~~~~~~
2451 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
2453 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
2454 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
2456 * Perfect match filters.
2457 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2458 The masked fields are for IP flow.
2460 * Signature filters.
2461 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
2463 * Perfect-mac-vlan match filters.
2464 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2465 The masked fields are for MAC VLAN flow.
2467 * Perfect-tunnel match filters.
2468 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2469 The masked fields are for tunnel flow.
2471 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
2472 per flow type and the flexible payload.
2474 The Flow Director can also mask out parts of all of these fields so that filters
2475 are only applied to certain fields or parts of the fields.
2477 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
2479 # Commands to add flow director filters of different flow types::
2481 flow_director_filter (port_id) mode IP (add|del|update) \
2482 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
2483 src (src_ip_address) dst (dst_ip_address) \
2484 tos (tos_value) proto (proto_value) ttl (ttl_value) \
2485 vlan (vlan_value) flexbytes (flexbytes_value) \
2486 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
2489 flow_director_filter (port_id) mode IP (add|del|update) \
2490 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
2491 src (src_ip_address) (src_port) \
2492 dst (dst_ip_address) (dst_port) \
2493 tos (tos_value) ttl (ttl_value) \
2494 vlan (vlan_value) flexbytes (flexbytes_value) \
2495 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
2498 flow_director_filter (port_id) mode IP (add|del|update) \
2499 flow (ipv4-sctp|ipv6-sctp) \
2500 src (src_ip_address) (src_port) \
2501 dst (dst_ip_address) (dst_port) \
2502 tos (tos_value) ttl (ttl_value) \
2503 tag (verification_tag) vlan (vlan_value) \
2504 flexbytes (flexbytes_value) (drop|fwd) \
2505 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
2507 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
2508 ether (ethertype) flexbytes (flexbytes_value) \
2509 (drop|fwd) pf|vf(vf_id) queue (queue_id)
2512 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
2513 mac (mac_address) vlan (vlan_value) \
2514 flexbytes (flexbytes_value) (drop|fwd) \
2515 queue (queue_id) fd_id (fd_id_value)
2517 flow_director_filter (port_id) mode Tunnel (add|del|update) \
2518 mac (mac_address) vlan (vlan_value) \
2519 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
2520 flexbytes (flexbytes_value) (drop|fwd) \
2521 queue (queue_id) fd_id (fd_id_value)
2523 For example, to add an ipv4-udp flow type filter::
2525 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
2526 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
2527 fwd pf queue 1 fd_id 1
2529 For example, add an ipv4-other flow type filter::
2531 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
2532 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
2533 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
2538 Flush all flow director filters on a device::
2540 testpmd> flush_flow_director (port_id)
2542 Example, to flush all flow director filter on port 0::
2544 testpmd> flush_flow_director 0
2549 Set flow director's input masks::
2551 flow_director_mask (port_id) mode IP vlan (vlan_value) \
2552 src_mask (ipv4_src) (ipv6_src) (src_port) \
2553 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2555 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2557 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2558 mac (mac_value) tunnel-type (tunnel_type_value) \
2559 tunnel-id (tunnel_id_value)
2561 Example, to set flow director mask on port 0::
2563 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2564 src_mask 255.255.255.255 \
2565 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2566 dst_mask 255.255.255.255 \
2567 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2569 flow_director_flex_mask
2570 ~~~~~~~~~~~~~~~~~~~~~~~
2572 set masks of flow director's flexible payload based on certain flow type::
2574 testpmd> flow_director_flex_mask (port_id) \
2575 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2576 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2577 l2_payload|all) (mask)
2579 Example, to set flow director's flex mask for all flow type on port 0::
2581 testpmd> flow_director_flex_mask 0 flow all \
2582 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2585 flow_director_flex_payload
2586 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2588 Configure flexible payload selection::
2590 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2592 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2594 testpmd> flow_director_flex_payload 0 l4 \
2595 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2597 get_sym_hash_ena_per_port
2598 ~~~~~~~~~~~~~~~~~~~~~~~~~
2600 Get symmetric hash enable configuration per port::
2602 get_sym_hash_ena_per_port (port_id)
2604 For example, to get symmetric hash enable configuration of port 1::
2606 testpmd> get_sym_hash_ena_per_port 1
2608 set_sym_hash_ena_per_port
2609 ~~~~~~~~~~~~~~~~~~~~~~~~~
2611 Set symmetric hash enable configuration per port to enable or disable::
2613 set_sym_hash_ena_per_port (port_id) (enable|disable)
2615 For example, to set symmetric hash enable configuration of port 1 to enable::
2617 testpmd> set_sym_hash_ena_per_port 1 enable
2619 get_hash_global_config
2620 ~~~~~~~~~~~~~~~~~~~~~~
2622 Get the global configurations of hash filters::
2624 get_hash_global_config (port_id)
2626 For example, to get the global configurations of hash filters of port 1::
2628 testpmd> get_hash_global_config 1
2630 set_hash_global_config
2631 ~~~~~~~~~~~~~~~~~~~~~~
2633 Set the global configurations of hash filters::
2635 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2636 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2637 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload) \
2640 For example, to enable simple_xor for flow type of ipv6 on port 2::
2642 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2647 Set the input set for hash::
2649 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2650 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2651 l2_payload) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2652 ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2653 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2654 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2655 fld-8th|none) (select|add)
2657 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2659 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2664 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2665 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2667 Set the input set for flow director::
2669 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2670 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2671 l2_payload) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2672 ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2673 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2674 sctp-dst-port|sctp-veri-tag|none) (select|add)
2676 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2678 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2683 Set different GRE key length for input set::
2685 global_config (port_id) gre-key-len (number in bytes)
2687 For example to set GRE key length for input set to 4 bytes on port 0::
2689 testpmd> global_config 0 gre-key-len 4
2692 .. _testpmd_rte_flow:
2694 Flow rules management
2695 ---------------------
2697 Control of the generic flow API (*rte_flow*) is fully exposed through the
2698 ``flow`` command (validation, creation, destruction, queries and operation
2701 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2702 features simultaneously may cause undefined side-effects and is therefore
2708 Because the ``flow`` command uses dynamic tokens to handle the large number
2709 of possible flow rules combinations, its behavior differs slightly from
2710 other commands, in particular:
2712 - Pressing *?* or the *<tab>* key displays contextual help for the current
2713 token, not that of the entire command.
2715 - Optional and repeated parameters are supported (provided they are listed
2716 in the contextual help).
2718 The first parameter stands for the operation mode. Possible operations and
2719 their general syntax are described below. They are covered in detail in the
2722 - Check whether a flow rule can be created::
2724 flow validate {port_id}
2725 [group {group_id}] [priority {level}] [ingress] [egress]
2726 pattern {item} [/ {item} [...]] / end
2727 actions {action} [/ {action} [...]] / end
2729 - Create a flow rule::
2731 flow create {port_id}
2732 [group {group_id}] [priority {level}] [ingress] [egress]
2733 pattern {item} [/ {item} [...]] / end
2734 actions {action} [/ {action} [...]] / end
2736 - Destroy specific flow rules::
2738 flow destroy {port_id} rule {rule_id} [...]
2740 - Destroy all flow rules::
2742 flow flush {port_id}
2744 - Query an existing flow rule::
2746 flow query {port_id} {rule_id} {action}
2748 - List existing flow rules sorted by priority, filtered by group
2751 flow list {port_id} [group {group_id}] [...]
2753 - Restrict ingress traffic to the defined flow rules::
2755 flow isolate {port_id} {boolean}
2757 Validating flow rules
2758 ~~~~~~~~~~~~~~~~~~~~~
2760 ``flow validate`` reports whether a flow rule would be accepted by the
2761 underlying device in its current state but stops short of creating it. It is
2762 bound to ``rte_flow_validate()``::
2764 flow validate {port_id}
2765 [group {group_id}] [priority {level}] [ingress] [egress]
2766 pattern {item} [/ {item} [...]] / end
2767 actions {action} [/ {action} [...]] / end
2769 If successful, it will show::
2773 Otherwise it will show an error message of the form::
2775 Caught error type [...] ([...]): [...]
2777 This command uses the same parameters as ``flow create``, their format is
2778 described in `Creating flow rules`_.
2780 Check whether redirecting any Ethernet packet received on port 0 to RX queue
2781 index 6 is supported::
2783 testpmd> flow validate 0 ingress pattern eth / end
2784 actions queue index 6 / end
2788 Port 0 does not support TCPv6 rules::
2790 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
2792 Caught error type 9 (specific pattern item): Invalid argument
2798 ``flow create`` validates and creates the specified flow rule. It is bound
2799 to ``rte_flow_create()``::
2801 flow create {port_id}
2802 [group {group_id}] [priority {level}] [ingress] [egress]
2803 pattern {item} [/ {item} [...]] / end
2804 actions {action} [/ {action} [...]] / end
2806 If successful, it will return a flow rule ID usable with other commands::
2808 Flow rule #[...] created
2810 Otherwise it will show an error message of the form::
2812 Caught error type [...] ([...]): [...]
2814 Parameters describe in the following order:
2816 - Attributes (*group*, *priority*, *ingress*, *egress* tokens).
2817 - A matching pattern, starting with the *pattern* token and terminated by an
2819 - Actions, starting with the *actions* token and terminated by an *end*
2822 These translate directly to *rte_flow* objects provided as-is to the
2823 underlying functions.
2825 The shortest valid definition only comprises mandatory tokens::
2827 testpmd> flow create 0 pattern end actions end
2829 Note that PMDs may refuse rules that essentially do nothing such as this
2832 **All unspecified object values are automatically initialized to 0.**
2837 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
2838 specified before the ``pattern`` token.
2840 - ``group {group id}``: priority group.
2841 - ``priority {level}``: priority level within group.
2842 - ``ingress``: rule applies to ingress traffic.
2843 - ``egress``: rule applies to egress traffic.
2845 Each instance of an attribute specified several times overrides the previous
2846 value as shown below (group 4 is used)::
2848 testpmd> flow create 0 group 42 group 24 group 4 [...]
2850 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
2852 While not specifying a direction is an error, some rules may allow both
2855 Most rules affect RX therefore contain the ``ingress`` token::
2857 testpmd> flow create 0 ingress pattern [...]
2862 A matching pattern starts after the ``pattern`` token. It is made of pattern
2863 items and is terminated by a mandatory ``end`` item.
2865 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
2866 rte_flow_item_type``).
2868 The ``/`` token is used as a separator between pattern items as shown
2871 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
2873 Note that protocol items like these must be stacked from lowest to highest
2874 layer to make sense. For instance, the following rule is either invalid or
2875 unlikely to match any packet::
2877 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
2879 More information on these restrictions can be found in the *rte_flow*
2882 Several items support additional specification structures, for example
2883 ``ipv4`` allows specifying source and destination addresses as follows::
2885 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
2886 dst is 10.2.0.0 / end [...]
2888 This rule matches all IPv4 traffic with the specified properties.
2890 In this example, ``src`` and ``dst`` are field names of the underlying
2891 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
2892 in a similar fashion.
2894 The ``is`` token means that the subsequent value must be matched exactly,
2895 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
2896 accordingly. Possible assignment tokens are:
2898 - ``is``: match value perfectly (with full bit-mask).
2899 - ``spec``: match value according to configured bit-mask.
2900 - ``last``: specify upper bound to establish a range.
2901 - ``mask``: specify bit-mask with relevant bits set to one.
2902 - ``prefix``: generate bit-mask from a prefix length.
2904 These yield identical results::
2906 ipv4 src is 10.1.1.1
2910 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
2914 ipv4 src spec 10.1.1.1 src prefix 32
2918 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
2922 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
2924 Inclusive ranges can be defined with ``last``::
2926 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
2928 Note that ``mask`` affects both ``spec`` and ``last``::
2930 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
2931 # matches 10.1.0.0 to 10.2.255.255
2933 Properties can be modified multiple times::
2935 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
2939 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
2944 This section lists supported pattern items and their attributes, if any.
2946 - ``end``: end list of pattern items.
2948 - ``void``: no-op pattern item.
2950 - ``invert``: perform actions when pattern does not match.
2952 - ``any``: match any protocol for the current layer.
2954 - ``num {unsigned}``: number of layers covered.
2956 - ``pf``: match packets addressed to the physical function.
2958 - ``vf``: match packets addressed to a virtual function ID.
2960 - ``id {unsigned}``: destination VF ID.
2962 - ``port``: device-specific physical port index to use.
2964 - ``index {unsigned}``: physical port index.
2966 - ``raw``: match an arbitrary byte string.
2968 - ``relative {boolean}``: look for pattern after the previous item.
2969 - ``search {boolean}``: search pattern from offset (see also limit).
2970 - ``offset {integer}``: absolute or relative offset for pattern.
2971 - ``limit {unsigned}``: search area limit for start of pattern.
2972 - ``pattern {string}``: byte string to look for.
2974 - ``eth``: match Ethernet header.
2976 - ``dst {MAC-48}``: destination MAC.
2977 - ``src {MAC-48}``: source MAC.
2978 - ``type {unsigned}``: EtherType.
2980 - ``vlan``: match 802.1Q/ad VLAN tag.
2982 - ``tpid {unsigned}``: tag protocol identifier.
2983 - ``tci {unsigned}``: tag control information.
2984 - ``pcp {unsigned}``: priority code point.
2985 - ``dei {unsigned}``: drop eligible indicator.
2986 - ``vid {unsigned}``: VLAN identifier.
2988 - ``ipv4``: match IPv4 header.
2990 - ``tos {unsigned}``: type of service.
2991 - ``ttl {unsigned}``: time to live.
2992 - ``proto {unsigned}``: next protocol ID.
2993 - ``src {ipv4 address}``: source address.
2994 - ``dst {ipv4 address}``: destination address.
2996 - ``ipv6``: match IPv6 header.
2998 - ``tc {unsigned}``: traffic class.
2999 - ``flow {unsigned}``: flow label.
3000 - ``proto {unsigned}``: protocol (next header).
3001 - ``hop {unsigned}``: hop limit.
3002 - ``src {ipv6 address}``: source address.
3003 - ``dst {ipv6 address}``: destination address.
3005 - ``icmp``: match ICMP header.
3007 - ``type {unsigned}``: ICMP packet type.
3008 - ``code {unsigned}``: ICMP packet code.
3010 - ``udp``: match UDP header.
3012 - ``src {unsigned}``: UDP source port.
3013 - ``dst {unsigned}``: UDP destination port.
3015 - ``tcp``: match TCP header.
3017 - ``src {unsigned}``: TCP source port.
3018 - ``dst {unsigned}``: TCP destination port.
3020 - ``sctp``: match SCTP header.
3022 - ``src {unsigned}``: SCTP source port.
3023 - ``dst {unsigned}``: SCTP destination port.
3024 - ``tag {unsigned}``: validation tag.
3025 - ``cksum {unsigned}``: checksum.
3027 - ``vxlan``: match VXLAN header.
3029 - ``vni {unsigned}``: VXLAN identifier.
3031 - ``e_tag``: match IEEE 802.1BR E-Tag header.
3033 - ``grp_ecid_b {unsigned}``: GRP and E-CID base.
3035 - ``nvgre``: match NVGRE header.
3037 - ``tni {unsigned}``: virtual subnet ID.
3039 - ``mpls``: match MPLS header.
3041 - ``label {unsigned}``: MPLS label.
3043 - ``gre``: match GRE header.
3045 - ``protocol {unsigned}``: protocol type.
3047 - ``fuzzy``: fuzzy pattern match, expect faster than default.
3049 - ``thresh {unsigned}``: accuracy threshold.
3051 - ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header.
3053 - ``teid {unsigned}``: tunnel endpoint identifier.
3058 A list of actions starts after the ``actions`` token in the same fashion as
3059 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
3060 terminated by a mandatory ``end`` action.
3062 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
3063 rte_flow_action_type``).
3065 Dropping all incoming UDPv4 packets can be expressed as follows::
3067 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3070 Several actions have configurable properties which must be specified when
3071 there is no valid default value. For example, ``queue`` requires a target
3074 This rule redirects incoming UDPv4 traffic to queue index 6::
3076 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3077 actions queue index 6 / end
3079 While this one could be rejected by PMDs (unspecified queue index)::
3081 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3084 As defined by *rte_flow*, the list is not ordered, all actions of a given
3085 rule are performed simultaneously. These are equivalent::
3087 queue index 6 / void / mark id 42 / end
3091 void / mark id 42 / queue index 6 / end
3093 All actions in a list should have different types, otherwise only the last
3094 action of a given type is taken into account::
3096 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
3100 drop / drop / drop / end # drop is performed only once
3104 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
3106 Considering they are performed simultaneously, opposite and overlapping
3107 actions can sometimes be combined when the end result is unambiguous::
3109 drop / queue index 6 / end # drop has no effect
3113 drop / dup index 6 / end # same as above
3117 queue index 6 / rss queues 6 7 8 / end # queue has no effect
3121 drop / passthru / end # drop has no effect
3123 Note that PMDs may still refuse such combinations.
3128 This section lists supported actions and their attributes, if any.
3130 - ``end``: end list of actions.
3132 - ``void``: no-op action.
3134 - ``passthru``: let subsequent rule process matched packets.
3136 - ``mark``: attach 32 bit value to packets.
3138 - ``id {unsigned}``: 32 bit value to return with packets.
3140 - ``flag``: flag packets.
3142 - ``queue``: assign packets to a given queue index.
3144 - ``index {unsigned}``: queue index to use.
3146 - ``drop``: drop packets (note: passthru has priority).
3148 - ``count``: enable counters for this rule.
3150 - ``dup``: duplicate packets to a given queue index.
3152 - ``index {unsigned}``: queue index to duplicate packets to.
3154 - ``rss``: spread packets among several queues.
3156 - ``queues [{unsigned} [...]] end``: queue indices to use.
3158 - ``pf``: redirect packets to physical device function.
3160 - ``vf``: redirect packets to virtual device function.
3162 - ``original {boolean}``: use original VF ID if possible.
3163 - ``id {unsigned}``: VF ID to redirect packets to.
3165 Destroying flow rules
3166 ~~~~~~~~~~~~~~~~~~~~~
3168 ``flow destroy`` destroys one or more rules from their rule ID (as returned
3169 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
3170 times as necessary::
3172 flow destroy {port_id} rule {rule_id} [...]
3174 If successful, it will show::
3176 Flow rule #[...] destroyed
3178 It does not report anything for rule IDs that do not exist. The usual error
3179 message is shown when a rule cannot be destroyed::
3181 Caught error type [...] ([...]): [...]
3183 ``flow flush`` destroys all rules on a device and does not take extra
3184 arguments. It is bound to ``rte_flow_flush()``::
3186 flow flush {port_id}
3188 Any errors are reported as above.
3190 Creating several rules and destroying them::
3192 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3193 actions queue index 2 / end
3194 Flow rule #0 created
3195 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3196 actions queue index 3 / end
3197 Flow rule #1 created
3198 testpmd> flow destroy 0 rule 0 rule 1
3199 Flow rule #1 destroyed
3200 Flow rule #0 destroyed
3203 The same result can be achieved using ``flow flush``::
3205 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3206 actions queue index 2 / end
3207 Flow rule #0 created
3208 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3209 actions queue index 3 / end
3210 Flow rule #1 created
3211 testpmd> flow flush 0
3214 Non-existent rule IDs are ignored::
3216 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3217 actions queue index 2 / end
3218 Flow rule #0 created
3219 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3220 actions queue index 3 / end
3221 Flow rule #1 created
3222 testpmd> flow destroy 0 rule 42 rule 10 rule 2
3224 testpmd> flow destroy 0 rule 0
3225 Flow rule #0 destroyed
3231 ``flow query`` queries a specific action of a flow rule having that
3232 ability. Such actions collect information that can be reported using this
3233 command. It is bound to ``rte_flow_query()``::
3235 flow query {port_id} {rule_id} {action}
3237 If successful, it will display either the retrieved data for known actions
3238 or the following message::
3240 Cannot display result for action type [...] ([...])
3242 Otherwise, it will complain either that the rule does not exist or that some
3245 Flow rule #[...] not found
3249 Caught error type [...] ([...]): [...]
3251 Currently only the ``count`` action is supported. This action reports the
3252 number of packets that hit the flow rule and the total number of bytes. Its
3253 output has the following format::
3256 hits_set: [...] # whether "hits" contains a valid value
3257 bytes_set: [...] # whether "bytes" contains a valid value
3258 hits: [...] # number of packets
3259 bytes: [...] # number of bytes
3261 Querying counters for TCPv6 packets redirected to queue 6::
3263 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
3264 actions queue index 6 / count / end
3265 Flow rule #4 created
3266 testpmd> flow query 0 4 count
3277 ``flow list`` lists existing flow rules sorted by priority and optionally
3278 filtered by group identifiers::
3280 flow list {port_id} [group {group_id}] [...]
3282 This command only fails with the following message if the device does not
3287 Output consists of a header line followed by a short description of each
3288 flow rule, one per line. There is no output at all when no flow rules are
3289 configured on the device::
3291 ID Group Prio Attr Rule
3292 [...] [...] [...] [...] [...]
3294 ``Attr`` column flags:
3296 - ``i`` for ``ingress``.
3297 - ``e`` for ``egress``.
3299 Creating several flow rules and listing them::
3301 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3302 actions queue index 6 / end
3303 Flow rule #0 created
3304 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3305 actions queue index 2 / end
3306 Flow rule #1 created
3307 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
3308 actions rss queues 6 7 8 end / end
3309 Flow rule #2 created
3310 testpmd> flow list 0
3311 ID Group Prio Attr Rule
3312 0 0 0 i- ETH IPV4 => QUEUE
3313 1 0 0 i- ETH IPV6 => QUEUE
3314 2 0 5 i- ETH IPV4 UDP => RSS
3317 Rules are sorted by priority (i.e. group ID first, then priority level)::
3319 testpmd> flow list 1
3320 ID Group Prio Attr Rule
3321 0 0 0 i- ETH => COUNT
3322 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3323 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3324 1 24 0 i- ETH IPV4 UDP => QUEUE
3325 4 24 10 i- ETH IPV4 TCP => DROP
3326 3 24 20 i- ETH IPV4 => DROP
3327 2 24 42 i- ETH IPV4 UDP => QUEUE
3328 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3331 Output can be limited to specific groups::
3333 testpmd> flow list 1 group 0 group 63
3334 ID Group Prio Attr Rule
3335 0 0 0 i- ETH => COUNT
3336 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3337 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3338 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3341 Toggling isolated mode
3342 ~~~~~~~~~~~~~~~~~~~~~~
3344 ``flow isolate`` can be used to tell the underlying PMD that ingress traffic
3345 must only be injected from the defined flow rules; that no default traffic
3346 is expected outside those rules and the driver is free to assign more
3347 resources to handle them. It is bound to ``rte_flow_isolate()``::
3349 flow isolate {port_id} {boolean}
3351 If successful, enabling or disabling isolated mode shows either::
3353 Ingress traffic on port [...]
3354 is now restricted to the defined flow rules
3358 Ingress traffic on port [...]
3359 is not restricted anymore to the defined flow rules
3361 Otherwise, in case of error::
3363 Caught error type [...] ([...]): [...]
3365 Mainly due to its side effects, PMDs supporting this mode may not have the
3366 ability to toggle it more than once without reinitializing affected ports
3367 first (e.g. by exiting testpmd).
3369 Enabling isolated mode::
3371 testpmd> flow isolate 0 true
3372 Ingress traffic on port 0 is now restricted to the defined flow rules
3375 Disabling isolated mode::
3377 testpmd> flow isolate 0 false
3378 Ingress traffic on port 0 is not restricted anymore to the defined flow rules
3381 Sample QinQ flow rules
3382 ~~~~~~~~~~~~~~~~~~~~~~
3384 Before creating QinQ rule(s) the following commands should be issued to enable QinQ::
3386 testpmd> port stop 0
3387 testpmd> vlan set qinq on 0
3389 The above command sets the inner and outer TPID's to 0x8100.
3391 To change the TPID's the following commands should be used::
3393 testpmd> vlan set outer tpid 0xa100 0
3394 testpmd> vlan set inner tpid 0x9100 0
3395 testpmd> port start 0
3397 Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM.
3401 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 /
3402 vlan tci is 456 / end actions vf id 1 / queue index 0 / end
3403 Flow rule #0 validated
3405 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 /
3406 vlan tci is 456 / end actions vf id 123 / queue index 0 / end
3407 Flow rule #0 created
3409 testpmd> flow list 0
3410 ID Group Prio Attr Rule
3411 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3413 Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host.
3417 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 /
3418 vlan tci is 654 / end actions pf / queue index 0 / end
3419 Flow rule #1 validated
3421 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 /
3422 vlan tci is 654 / end actions pf / queue index 1 / end
3423 Flow rule #1 created
3425 testpmd> flow list 0
3426 ID Group Prio Attr Rule
3427 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3428 1 0 0 i- ETH VLAN VLAN=>PF QUEUE