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``.
359 testpmd> set fwd rxonly
361 Set rxonly packet forwarding mode
367 Display an RX descriptor for a port RX queue::
369 testpmd> read rxd (port_id) (queue_id) (rxd_id)
373 testpmd> read rxd 0 0 4
374 0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180
379 Display a TX descriptor for a port TX queue::
381 testpmd> read txd (port_id) (queue_id) (txd_id)
385 testpmd> read txd 0 0 4
386 0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C
391 Get loaded dynamic device personalization (DDP) package info list::
393 testpmd> ddp get list (port_id)
398 Display information about dynamic device personalization (DDP) profile::
400 testpmd> ddp get info (profile_path)
405 Display VF statistics::
407 testpmd> show vf stats (port_id) (vf_id)
412 Reset VF statistics::
414 testpmd> clear vf stats (port_id) (vf_id)
416 show port pctype mapping
417 ~~~~~~~~~~~~~~~~~~~~~~~~
419 List all items from the pctype mapping table::
421 testpmd> show port (port_id) pctype mapping
424 Configuration Functions
425 -----------------------
427 The testpmd application can be configured from the runtime as well as from the command-line.
429 This section details the available configuration functions that are available.
433 Configuration changes only become active when forwarding is started/restarted.
438 Reset forwarding to the default configuration::
445 Set the debug verbosity level::
447 testpmd> set verbose (level)
449 Currently the only available levels are 0 (silent except for error) and 1 (fully verbose).
454 Set the number of ports used by the application:
458 This is equivalent to the ``--nb-ports`` command-line option.
463 Set the number of cores used by the application::
465 testpmd> set nbcore (num)
467 This is equivalent to the ``--nb-cores`` command-line option.
471 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port.
476 Set the forwarding cores hexadecimal mask::
478 testpmd> set coremask (mask)
480 This is equivalent to the ``--coremask`` command-line option.
484 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding.
489 Set the forwarding ports hexadecimal mask::
491 testpmd> set portmask (mask)
493 This is equivalent to the ``--portmask`` command-line option.
498 Set number of packets per burst::
500 testpmd> set burst (num)
502 This is equivalent to the ``--burst command-line`` option.
504 When retry is enabled, the transmit delay time and number of retries can also be set::
506 testpmd> set burst tx delay (microseconds) retry (num)
511 Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode::
513 testpmd> set txpkts (x[,y]*)
515 Where x[,y]* represents a CSV list of values, without white space.
520 Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes::
522 testpmd> set txsplit (off|on|rand)
526 * ``off`` disable packet copy & split for CSUM mode.
528 * ``on`` split outgoing packet into multiple segments. Size of each segment
529 and number of segments per packet is determined by ``set txpkts`` command
532 * ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments.
537 Set the list of forwarding cores::
539 testpmd> set corelist (x[,y]*)
541 For example, to change the forwarding cores:
543 .. code-block:: console
545 testpmd> set corelist 3,1
546 testpmd> show config fwd
548 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled
549 Logical Core 3 (socket 0) forwards packets on 1 streams:
550 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01
551 Logical Core 1 (socket 0) forwards packets on 1 streams:
552 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
556 The cores are used in the same order as specified on the command line.
561 Set the list of forwarding ports::
563 testpmd> set portlist (x[,y]*)
565 For example, to change the port forwarding:
567 .. code-block:: console
569 testpmd> set portlist 0,2,1,3
570 testpmd> show config fwd
572 io packet forwarding - ports=4 - cores=1 - streams=4
573 Logical Core 3 (socket 0) forwards packets on 4 streams:
574 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01
575 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
576 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03
577 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02
582 Enable/disable tx loopback::
584 testpmd> set tx loopback (port_id) (on|off)
589 set drop enable bit for all queues::
591 testpmd> set all queues drop (port_id) (on|off)
593 set split drop enable (for VF)
594 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
596 set split drop enable bit for VF from PF::
598 testpmd> set vf split drop (port_id) (vf_id) (on|off)
600 set mac antispoof (for VF)
601 ~~~~~~~~~~~~~~~~~~~~~~~~~~
603 Set mac antispoof for a VF from the PF::
605 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off)
610 Enable/disable MACsec offload::
612 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off)
613 testpmd> set macsec offload (port_id) off
618 Configure MACsec secure connection (SC)::
620 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi)
624 The pi argument is ignored for tx.
625 Check the NIC Datasheet for hardware limits.
630 Configure MACsec secure association (SA)::
632 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key)
636 The IDX value must be 0 or 1.
637 Check the NIC Datasheet for hardware limits.
639 set broadcast mode (for VF)
640 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
642 Set broadcast mode for a VF from the PF::
644 testpmd> set vf broadcast (port_id) (vf_id) (on|off)
649 Set the VLAN strip on a port::
651 testpmd> vlan set strip (on|off) (port_id)
656 Set the VLAN strip for a queue on a port::
658 testpmd> vlan set stripq (on|off) (port_id,queue_id)
660 vlan set stripq (for VF)
661 ~~~~~~~~~~~~~~~~~~~~~~~~
663 Set VLAN strip for all queues in a pool for a VF from the PF::
665 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off)
667 vlan set insert (for VF)
668 ~~~~~~~~~~~~~~~~~~~~~~~~
670 Set VLAN insert for a VF from the PF::
672 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id)
674 vlan set tag (for VF)
675 ~~~~~~~~~~~~~~~~~~~~~
677 Set VLAN tag for a VF from the PF::
679 testpmd> set vf vlan tag (port_id) (vf_id) (on|off)
681 vlan set antispoof (for VF)
682 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
684 Set VLAN antispoof for a VF from the PF::
686 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off)
691 Set the VLAN filter on a port::
693 testpmd> vlan set filter (on|off) (port_id)
698 Set the VLAN QinQ (extended queue in queue) on for a port::
700 testpmd> vlan set qinq (on|off) (port_id)
705 Set the inner or outer VLAN TPID for packet filtering on a port::
707 testpmd> vlan set (inner|outer) tpid (value) (port_id)
711 TPID value must be a 16-bit number (value <= 65536).
716 Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID::
718 testpmd> rx_vlan add (vlan_id|all) (port_id)
722 VLAN filter must be set on that port. VLAN ID < 4096.
723 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries
724 in VFTA table. This is important if enabling all vlan_ids.
729 Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID::
731 testpmd> rx_vlan rm (vlan_id|all) (port_id)
736 Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID::
738 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask)
743 Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID::
745 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask)
750 Add a tunnel filter on a port::
752 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \
753 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
754 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
756 The available information categories are:
758 * ``vxlan``: Set tunnel type as VXLAN.
760 * ``nvgre``: Set tunnel type as NVGRE.
762 * ``ipingre``: Set tunnel type as IP-in-GRE.
764 * ``imac-ivlan``: Set filter type as Inner MAC and VLAN.
766 * ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID.
768 * ``imac-tenid``: Set filter type as Inner MAC and tenant ID.
770 * ``imac``: Set filter type as Inner MAC.
772 * ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID.
774 * ``oip``: Set filter type as Outer IP.
776 * ``iip``: Set filter type as Inner IP.
780 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \
781 192.168.2.2 0 ipingre oip 1 1
783 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP.
788 Remove a tunnel filter on a port::
790 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \
791 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
792 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
797 Add an UDP port for VXLAN packet filter on a port::
799 testpmd> rx_vxlan_port add (udp_port) (port_id)
804 Remove an UDP port for VXLAN packet filter on a port::
806 testpmd> rx_vxlan_port rm (udp_port) (port_id)
811 Set hardware insertion of VLAN IDs in packets sent on a port::
813 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer]
815 For example, set a single VLAN ID (5) insertion on port 0::
819 Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1::
827 Set port based hardware insertion of VLAN ID in packets sent on a port::
829 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off)
834 Disable hardware insertion of a VLAN header in packets sent on a port::
836 testpmd> tx_vlan reset (port_id)
841 Select hardware or software calculation of the checksum when
842 transmitting a packet using the ``csum`` forwarding engine::
844 testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id)
848 * ``ip|udp|tcp|sctp`` always relate to the inner layer.
850 * ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized
851 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are
852 supported). See also the ``csum parse-tunnel`` command.
856 Check the NIC Datasheet for hardware limits.
861 Set RSS queue region span on a port::
863 testpmd> set port (port_id) queue-region region_id (value) \
864 queue_start_index (value) queue_num (value)
866 Set flowtype mapping on a RSS queue region on a port::
868 testpmd> set port (port_id) queue-region region_id (value) flowtype (value)
872 * For the flowtype(pctype) of packet,the specific index for each type has
873 been defined in file i40e_type.h as enum i40e_filter_pctype.
875 Set user priority mapping on a RSS queue region on a port::
877 testpmd> set port (port_id) queue-region UP (value) region_id (value)
879 Flush all queue region related configuration on a port::
881 testpmd> set port (port_id) queue-region flush (on|off)
885 * "on"is just an enable function which server for other configuration,
886 it is for all configuration about queue region from up layer,
887 at first will only keep in DPDK softwarestored in driver,
888 only after "flush on", it commit all configuration to HW.
889 "off" is just clean all configuration about queue region just now,
890 and restore all to DPDK i40e driver default config when start up.
892 Show all queue region related configuration info on a port::
894 testpmd> show port (port_id) queue-region
898 Queue region only support on PF by now, so these command is
899 only for configuration of queue region on PF port.
904 Define how tunneled packets should be handled by the csum forward
907 testpmd> csum parse-tunnel (on|off) (tx_port_id)
909 If enabled, the csum forward engine will try to recognize supported
910 tunnel headers (vxlan, gre, ipip).
912 If disabled, treat tunnel packets as non-tunneled packets (a inner
913 header is handled as a packet payload).
917 The port argument is the TX port like in the ``csum set`` command.
921 Consider a packet in packet like the following::
923 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in
925 * If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
926 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the
927 ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``).
929 * If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
930 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``.
935 Display tx checksum offload configuration::
937 testpmd> csum show (port_id)
942 Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine::
944 testpmd> tso set (segsize) (port_id)
948 Check the NIC datasheet for hardware limits.
953 Display the status of TCP Segmentation Offload::
955 testpmd> tso show (port_id)
960 Enable or disable GRO in ``csum`` forwarding engine::
962 testpmd> set port <port_id> gro on|off
964 If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4
965 packets received from the given port.
967 If disabled, packets received from the given port won't be performed
968 GRO. By default, GRO is disabled for all ports.
972 When enable GRO for a port, TCP/IPv4 packets received from the port
973 will be performed GRO. After GRO, all merged packets have bad
974 checksums, since the GRO library doesn't re-calculate checksums for
975 the merged packets. Therefore, if users want the merged packets to
976 have correct checksums, please select HW IP checksum calculation and
977 HW TCP checksum calculation for the port which the merged packets are
983 Display GRO configuration for a given port::
985 testpmd> show port <port_id> gro
990 Set the cycle to flush the GROed packets from reassembly tables::
992 testpmd> set gro flush <cycles>
994 When enable GRO, the csum forwarding engine performs GRO on received
995 packets, and the GROed packets are stored in reassembly tables. Users
996 can use this command to determine when the GROed packets are flushed
997 from the reassembly tables.
999 The ``cycles`` is measured in GRO operation times. The csum forwarding
1000 engine flushes the GROed packets from the tables every ``cycles`` GRO
1003 By default, the value of ``cycles`` is 1, which means flush GROed packets
1004 from the reassembly tables as soon as one GRO operation finishes. The value
1005 of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``.
1007 Please note that the large value of ``cycles`` may cause the poor TCP/IP
1008 stack performance. Because the GROed packets are delayed to arrive the
1009 stack, thus causing more duplicated ACKs and TCP retransmissions.
1014 Toggle per-port GSO support in ``csum`` forwarding engine::
1016 testpmd> set port <port_id> gso on|off
1018 If enabled, the csum forwarding engine will perform GSO on supported IPv4
1019 packets, transmitted on the given port.
1021 If disabled, packets transmitted on the given port will not undergo GSO.
1022 By default, GSO is disabled for all ports.
1026 When GSO is enabled on a port, supported IPv4 packets transmitted on that
1027 port undergo GSO. Afterwards, the segmented packets are represented by
1028 multi-segment mbufs; however, the csum forwarding engine doesn't calculation
1029 of checksums for GSO'd segments in SW. As a result, if users want correct
1030 checksums in GSO segments, they should enable HW checksum calculation for
1033 For example, HW checksum calculation for VxLAN GSO'd packets may be enabled
1034 by setting the following options in the csum forwarding engine:
1036 testpmd> csum set outer_ip hw <port_id>
1038 testpmd> csum set ip hw <port_id>
1040 testpmd> csum set tcp hw <port_id>
1045 Set the maximum GSO segment size (measured in bytes), which includes the
1046 packet header and the packet payload for GSO-enabled ports (global)::
1048 testpmd> set gso segsz <length>
1053 Display the status of Generic Segmentation Offload for a given port::
1055 testpmd> show port <port_id> gso
1060 Add an alternative MAC address to a port::
1062 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX)
1067 Remove a MAC address from a port::
1069 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX)
1071 mac_addr add (for VF)
1072 ~~~~~~~~~~~~~~~~~~~~~
1074 Add an alternative MAC address for a VF to a port::
1076 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX)
1081 Set the default MAC address for a port::
1083 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX)
1085 mac_addr set (for VF)
1086 ~~~~~~~~~~~~~~~~~~~~~
1088 Set the MAC address for a VF from the PF::
1090 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX)
1095 Set the forwarding peer address for certain port::
1097 testpmd> set eth-peer (port_id) (perr_addr)
1099 This is equivalent to the ``--eth-peer`` command-line option.
1104 Set the unicast hash filter(s) on/off for a port::
1106 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
1111 Set the promiscuous mode on for a port or for all ports.
1112 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1114 testpmd> set promisc (port_id|all) (on|off)
1119 Set the allmulti mode for a port or for all ports::
1121 testpmd> set allmulti (port_id|all) (on|off)
1123 Same as the ifconfig (8) option. Controls how multicast packets are handled.
1125 set promisc (for VF)
1126 ~~~~~~~~~~~~~~~~~~~~
1128 Set the unicast promiscuous mode for a VF from PF.
1129 It's supported by Intel i40e NICs now.
1130 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1132 testpmd> set vf promisc (port_id) (vf_id) (on|off)
1134 set allmulticast (for VF)
1135 ~~~~~~~~~~~~~~~~~~~~~~~~~
1137 Set the multicast promiscuous mode for a VF from PF.
1138 It's supported by Intel i40e NICs now.
1139 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1141 testpmd> set vf allmulti (port_id) (vf_id) (on|off)
1143 set tx max bandwidth (for VF)
1144 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1146 Set TX max absolute bandwidth (Mbps) for a VF from PF::
1148 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth)
1150 set tc tx min bandwidth (for VF)
1151 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1153 Set all TCs' TX min relative bandwidth (%) for a VF from PF::
1155 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...)
1157 set tc tx max bandwidth (for VF)
1158 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1160 Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF::
1162 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth)
1164 set tc strict link priority mode
1165 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1167 Set some TCs' strict link priority mode on a physical port::
1169 testpmd> set tx strict-link-priority (port_id) (tc_bitmap)
1171 set tc tx min bandwidth
1172 ~~~~~~~~~~~~~~~~~~~~~~~
1174 Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs::
1176 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...)
1181 Set the link flow control parameter on a port::
1183 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1184 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
1185 autoneg (on|off) (port_id)
1189 * ``high_water`` (integer): High threshold value to trigger XOFF.
1191 * ``low_water`` (integer): Low threshold value to trigger XON.
1193 * ``pause_time`` (integer): Pause quota in the Pause frame.
1195 * ``send_xon`` (0/1): Send XON frame.
1197 * ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
1199 * ``autoneg``: Change the auto-negotiation parameter.
1204 Set the priority flow control parameter on a port::
1206 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1207 (pause_time) (priority) (port_id)
1211 * ``high_water`` (integer): High threshold value.
1213 * ``low_water`` (integer): Low threshold value.
1215 * ``pause_time`` (integer): Pause quota in the Pause frame.
1217 * ``priority`` (0-7): VLAN User Priority.
1222 Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
1224 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
1226 For example, to set rx queue 2 on port 0 to mapping 5::
1228 testpmd>set stat_qmap rx 0 2 5
1230 set xstats-hide-zero
1231 ~~~~~~~~~~~~~~~~~~~~
1233 Set the option to hide zero values for xstats display::
1235 testpmd> set xstats-hide-zero on|off
1239 By default, the zero values are displayed for xstats.
1241 set port - rx/tx (for VF)
1242 ~~~~~~~~~~~~~~~~~~~~~~~~~
1244 Set VF receive/transmit from a port::
1246 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
1248 set port - mac address filter (for VF)
1249 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1251 Add/Remove unicast or multicast MAC addr filter for a VF::
1253 testpmd> set port (port_id) vf (vf_id) (mac_addr) \
1254 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
1256 set port - rx mode(for VF)
1257 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1259 Set the VF receive mode of a port::
1261 testpmd> set port (port_id) vf (vf_id) \
1262 rxmode (AUPE|ROPE|BAM|MPE) (on|off)
1264 The available receive modes are:
1266 * ``AUPE``: Accepts untagged VLAN.
1268 * ``ROPE``: Accepts unicast hash.
1270 * ``BAM``: Accepts broadcast packets.
1272 * ``MPE``: Accepts all multicast packets.
1274 set port - tx_rate (for Queue)
1275 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1277 Set TX rate limitation for a queue on a port::
1279 testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1281 set port - tx_rate (for VF)
1282 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
1284 Set TX rate limitation for queues in VF on a port::
1286 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1288 set port - mirror rule
1289 ~~~~~~~~~~~~~~~~~~~~~~
1291 Set pool or vlan type mirror rule for a port::
1293 testpmd> set port (port_id) mirror-rule (rule_id) \
1294 (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1295 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1297 Set link mirror rule for a port::
1299 testpmd> set port (port_id) mirror-rule (rule_id) \
1300 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1302 For example to enable mirror traffic with vlan 0,1 to pool 0::
1304 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1306 reset port - mirror rule
1307 ~~~~~~~~~~~~~~~~~~~~~~~~
1309 Reset a mirror rule for a port::
1311 testpmd> reset port (port_id) mirror-rule (rule_id)
1316 Set the flush on RX streams before forwarding.
1317 The default is flush ``on``.
1318 Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1320 testpmd> set flush_rx off
1325 Set the bypass mode for the lowest port on bypass enabled NIC::
1327 testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1332 Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1334 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1335 mode (normal|bypass|isolate) (port_id)
1339 * ``timeout``: Enable bypass after watchdog timeout.
1341 * ``os_on``: Enable bypass when OS/board is powered on.
1343 * ``os_off``: Enable bypass when OS/board is powered off.
1345 * ``power_on``: Enable bypass when power supply is turned on.
1347 * ``power_off``: Enable bypass when power supply is turned off.
1353 Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1355 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1360 Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1362 testpmd> show bypass config (port_id)
1367 Set link up for a port::
1369 testpmd> set link-up port (port id)
1374 Set link down for a port::
1376 testpmd> set link-down port (port id)
1381 Enable E-tag insertion for a VF on a port::
1383 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1385 Disable E-tag insertion for a VF on a port::
1387 testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1389 Enable/disable E-tag stripping on a port::
1391 testpmd> E-tag set stripping (on|off) port (port_id)
1393 Enable/disable E-tag based forwarding on a port::
1395 testpmd> E-tag set forwarding (on|off) port (port_id)
1397 Add an E-tag forwarding filter on a port::
1399 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1401 Delete an E-tag forwarding filter on a port::
1402 testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1407 Load a dynamic device personalization (DDP) package::
1409 testpmd> ddp add (port_id) (package_path[,output_path])
1414 Delete a dynamic device personalization package::
1416 testpmd> ddp del (port_id) (package_path)
1421 List all items from the ptype mapping table::
1423 testpmd> ptype mapping get (port_id) (valid_only)
1427 * ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0).
1429 Replace a specific or a group of software defined ptype with a new one::
1431 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type)
1435 * ``target``: A specific software ptype or a mask to represent a group of software ptypes.
1437 * ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1).
1439 * ``pkt_type``: The new software ptype to replace the old ones.
1441 Update hardware defined ptype to software defined packet type mapping table::
1443 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype)
1447 * ``hw_ptype``: hardware ptype as the index of the ptype mapping table.
1449 * ``sw_ptype``: software ptype as the value of the ptype mapping table.
1451 Reset ptype mapping table::
1453 testpmd> ptype mapping reset (port_id)
1458 The following sections show functions for configuring ports.
1462 Port configuration changes only become active when forwarding is started/restarted.
1467 Attach a port specified by pci address or virtual device args::
1469 testpmd> port attach (identifier)
1471 To attach a new pci device, the device should be recognized by kernel first.
1472 Then it should be moved under DPDK management.
1473 Finally the port can be attached to testpmd.
1475 For example, to move a pci device using ixgbe under DPDK management:
1477 .. code-block:: console
1479 # Check the status of the available devices.
1480 ./usertools/dpdk-devbind.py --status
1482 Network devices using DPDK-compatible driver
1483 ============================================
1486 Network devices using kernel driver
1487 ===================================
1488 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1491 # Bind the device to igb_uio.
1492 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1495 # Recheck the status of the devices.
1496 ./usertools/dpdk-devbind.py --status
1497 Network devices using DPDK-compatible driver
1498 ============================================
1499 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1501 To attach a port created by virtual device, above steps are not needed.
1503 For example, to attach a port whose pci address is 0000:0a:00.0.
1505 .. code-block:: console
1507 testpmd> port attach 0000:0a:00.0
1508 Attaching a new port...
1509 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1510 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1511 EAL: PCI memory mapped at 0x7f83bfa00000
1512 EAL: PCI memory mapped at 0x7f83bfa80000
1513 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1514 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1515 Port 0 is attached. Now total ports is 1
1518 For example, to attach a port created by pcap PMD.
1520 .. code-block:: console
1522 testpmd> port attach net_pcap0
1523 Attaching a new port...
1524 PMD: Initializing pmd_pcap for net_pcap0
1525 PMD: Creating pcap-backed ethdev on numa socket 0
1526 Port 0 is attached. Now total ports is 1
1529 In this case, identifier is ``net_pcap0``.
1530 This identifier format is the same as ``--vdev`` format of DPDK applications.
1532 For example, to re-attach a bonded port which has been previously detached,
1533 the mode and slave parameters must be given.
1535 .. code-block:: console
1537 testpmd> port attach net_bond_0,mode=0,slave=1
1538 Attaching a new port...
1539 EAL: Initializing pmd_bond for net_bond_0
1540 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1541 Port 0 is attached. Now total ports is 1
1548 Detach a specific port::
1550 testpmd> port detach (port_id)
1552 Before detaching a port, the port should be stopped and closed.
1554 For example, to detach a pci device port 0.
1556 .. code-block:: console
1558 testpmd> port stop 0
1561 testpmd> port close 0
1565 testpmd> port detach 0
1567 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1568 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1569 EAL: PCI memory unmapped at 0x7f83bfa00000
1570 EAL: PCI memory unmapped at 0x7f83bfa80000
1574 For example, to detach a virtual device port 0.
1576 .. code-block:: console
1578 testpmd> port stop 0
1581 testpmd> port close 0
1585 testpmd> port detach 0
1587 PMD: Closing pcap ethdev on numa socket 0
1588 Port 'net_pcap0' is detached. Now total ports is 0
1591 To remove a pci device completely from the system, first detach the port from testpmd.
1592 Then the device should be moved under kernel management.
1593 Finally the device can be removed using kernel pci hotplug functionality.
1595 For example, to move a pci device under kernel management:
1597 .. code-block:: console
1599 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1601 ./usertools/dpdk-devbind.py --status
1603 Network devices using DPDK-compatible driver
1604 ============================================
1607 Network devices using kernel driver
1608 ===================================
1609 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1611 To remove a port created by a virtual device, above steps are not needed.
1616 Start all ports or a specific port::
1618 testpmd> port start (port_id|all)
1623 Stop all ports or a specific port::
1625 testpmd> port stop (port_id|all)
1630 Close all ports or a specific port::
1632 testpmd> port close (port_id|all)
1634 port start/stop queue
1635 ~~~~~~~~~~~~~~~~~~~~~
1637 Start/stop a rx/tx queue on a specific port::
1639 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1641 Only take effect when port is started.
1646 Set the speed and duplex mode for all ports or a specific port::
1648 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1649 duplex (half|full|auto)
1651 port config - queues/descriptors
1652 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1654 Set number of queues/descriptors for rxq, txq, rxd and txd::
1656 testpmd> port config all (rxq|txq|rxd|txd) (value)
1658 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1660 port config - max-pkt-len
1661 ~~~~~~~~~~~~~~~~~~~~~~~~~
1663 Set the maximum packet length::
1665 testpmd> port config all max-pkt-len (value)
1667 This is equivalent to the ``--max-pkt-len`` command-line option.
1669 port config - CRC Strip
1670 ~~~~~~~~~~~~~~~~~~~~~~~
1672 Set hardware CRC stripping on or off for all ports::
1674 testpmd> port config all crc-strip (on|off)
1676 CRC stripping is on by default.
1678 The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option.
1680 port config - scatter
1681 ~~~~~~~~~~~~~~~~~~~~~~~
1683 Set RX scatter mode on or off for all ports::
1685 testpmd> port config all scatter (on|off)
1687 RX scatter mode is off by default.
1689 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1691 port config - RX Checksum
1692 ~~~~~~~~~~~~~~~~~~~~~~~~~
1694 Set hardware RX checksum offload to on or off for all ports::
1696 testpmd> port config all rx-cksum (on|off)
1698 Checksum offload is off by default.
1700 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1705 Set hardware VLAN on or off for all ports::
1707 testpmd> port config all hw-vlan (on|off)
1709 Hardware VLAN is on by default.
1711 The ``off`` option is equivalent to the ``--disable-hw-vlan`` command-line option.
1713 port config - VLAN filter
1714 ~~~~~~~~~~~~~~~~~~~~~~~~~
1716 Set hardware VLAN filter on or off for all ports::
1718 testpmd> port config all hw-vlan-filter (on|off)
1720 Hardware VLAN filter is on by default.
1722 The ``off`` option is equivalent to the ``--disable-hw-vlan-filter`` command-line option.
1724 port config - VLAN strip
1725 ~~~~~~~~~~~~~~~~~~~~~~~~
1727 Set hardware VLAN strip on or off for all ports::
1729 testpmd> port config all hw-vlan-strip (on|off)
1731 Hardware VLAN strip is on by default.
1733 The ``off`` option is equivalent to the ``--disable-hw-vlan-strip`` command-line option.
1735 port config - VLAN extend
1736 ~~~~~~~~~~~~~~~~~~~~~~~~~
1738 Set hardware VLAN extend on or off for all ports::
1740 testpmd> port config all hw-vlan-extend (on|off)
1742 Hardware VLAN extend is off by default.
1744 The ``off`` option is equivalent to the ``--disable-hw-vlan-extend`` command-line option.
1746 port config - Drop Packets
1747 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1749 Set packet drop for packets with no descriptors on or off for all ports::
1751 testpmd> port config all drop-en (on|off)
1753 Packet dropping for packets with no descriptors is off by default.
1755 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1760 Set the RSS (Receive Side Scaling) mode on or off::
1762 testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1764 RSS is on by default.
1766 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1768 port config - RSS Reta
1769 ~~~~~~~~~~~~~~~~~~~~~~
1771 Set the RSS (Receive Side Scaling) redirection table::
1773 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1778 Set the DCB mode for an individual port::
1780 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1782 The traffic class should be 4 or 8.
1787 Set the number of packets per burst::
1789 testpmd> port config all burst (value)
1791 This is equivalent to the ``--burst`` command-line option.
1793 port config - Threshold
1794 ~~~~~~~~~~~~~~~~~~~~~~~
1796 Set thresholds for TX/RX queues::
1798 testpmd> port config all (threshold) (value)
1800 Where the threshold type can be:
1802 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1804 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1806 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1808 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1810 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1812 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1814 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1816 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1818 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1820 These threshold options are also available from the command-line.
1825 Set the value of ether-type for E-tag::
1827 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1829 Enable/disable the E-tag support::
1831 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1833 port config pctype mapping
1834 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1836 Reset pctype mapping table::
1838 testpmd> port config (port_id) pctype mapping reset
1840 Update hardware defined pctype to software defined flow type mapping table::
1842 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id)
1846 * ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table.
1848 * ``flow_type_id``: software flow type id as the index of the pctype mapping table.
1851 Link Bonding Functions
1852 ----------------------
1854 The Link Bonding functions make it possible to dynamically create and
1855 manage link bonding devices from within testpmd interactive prompt.
1857 create bonded device
1858 ~~~~~~~~~~~~~~~~~~~~
1860 Create a new bonding device::
1862 testpmd> create bonded device (mode) (socket)
1864 For example, to create a bonded device in mode 1 on socket 0::
1866 testpmd> create bonded 1 0
1867 created new bonded device (port X)
1872 Adds Ethernet device to a Link Bonding device::
1874 testpmd> add bonding slave (slave id) (port id)
1876 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
1878 testpmd> add bonding slave 6 10
1881 remove bonding slave
1882 ~~~~~~~~~~~~~~~~~~~~
1884 Removes an Ethernet slave device from a Link Bonding device::
1886 testpmd> remove bonding slave (slave id) (port id)
1888 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
1890 testpmd> remove bonding slave 6 10
1895 Set the Link Bonding mode of a Link Bonding device::
1897 testpmd> set bonding mode (value) (port id)
1899 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
1901 testpmd> set bonding mode 3 10
1906 Set an Ethernet slave device as the primary device on a Link Bonding device::
1908 testpmd> set bonding primary (slave id) (port id)
1910 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
1912 testpmd> set bonding primary 6 10
1917 Set the MAC address of a Link Bonding device::
1919 testpmd> set bonding mac (port id) (mac)
1921 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
1923 testpmd> set bonding mac 10 00:00:00:00:00:01
1925 set bonding xmit_balance_policy
1926 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1928 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
1930 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
1932 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
1934 testpmd> set bonding xmit_balance_policy 10 l34
1937 set bonding mon_period
1938 ~~~~~~~~~~~~~~~~~~~~~~
1940 Set the link status monitoring polling period in milliseconds for a bonding device.
1942 This adds support for PMD slave devices which do not support link status interrupts.
1943 When the mon_period is set to a value greater than 0 then all PMD's which do not support
1944 link status ISR will be queried every polling interval to check if their link status has changed::
1946 testpmd> set bonding mon_period (port_id) (value)
1948 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
1950 testpmd> set bonding mon_period 5 150
1953 set bonding lacp dedicated_queue
1954 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1956 Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic
1957 when in mode 4 (link-aggregration-802.3ad)::
1959 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable)
1962 set bonding agg_mode
1963 ~~~~~~~~~~~~~~~~~~~~
1965 Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad)::
1967 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable)
1973 Show the current configuration of a Link Bonding device::
1975 testpmd> show bonding config (port id)
1978 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
1979 in balance mode with a transmission policy of layer 2+3::
1981 testpmd> show bonding config 9
1983 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
1985 Active Slaves (3): [1 3 4]
1992 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
1993 This is mainly useful for debugging purposes.
1994 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
1995 and fields that can be accessed.
2000 Display the value of a port register::
2002 testpmd> read reg (port_id) (address)
2004 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
2006 testpmd> read reg 0 0xEE00
2007 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
2012 Display a port register bit field::
2014 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
2016 For example, reading the lowest two bits from the register in the example above::
2018 testpmd> read regfield 0 0xEE00 0 1
2019 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
2024 Display a single port register bit::
2026 testpmd> read regbit (port_id) (address) (bit_x)
2028 For example, reading the lowest bit from the register in the example above::
2030 testpmd> read regbit 0 0xEE00 0
2031 port 0 PCI register at offset 0xEE00: bit 0=1
2036 Set the value of a port register::
2038 testpmd> write reg (port_id) (address) (value)
2040 For example, to clear a register::
2042 testpmd> write reg 0 0xEE00 0x0
2043 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
2048 Set bit field of a port register::
2050 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
2052 For example, writing to the register cleared in the example above::
2054 testpmd> write regfield 0 0xEE00 0 1 2
2055 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
2060 Set single bit value of a port register::
2062 testpmd> write regbit (port_id) (address) (bit_x) (value)
2064 For example, to set the high bit in the register from the example above::
2066 testpmd> write regbit 0 0xEE00 31 1
2067 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
2069 Traffic Metering and Policing
2070 -----------------------------
2072 The following section shows functions for configuring traffic metering and
2073 policing on the ethernet device through the use of generic ethdev API.
2075 show port traffic management capability
2076 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2078 Show traffic metering and policing capability of the port::
2080 testpmd> show port meter cap (port_id)
2082 add port meter profile (srTCM rfc2967)
2083 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2085 Add meter profile (srTCM rfc2697) to the ethernet device::
2087 testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \
2092 * ``profile_id``: ID for the meter profile.
2093 * ``cir``: Committed Information Rate (CIR) (bytes/second).
2094 * ``cbs``: Committed Burst Size (CBS) (bytes).
2095 * ``ebs``: Excess Burst Size (EBS) (bytes).
2097 add port meter profile (trTCM rfc2968)
2098 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2100 Add meter profile (srTCM rfc2698) to the ethernet device::
2102 testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \
2103 (cir) (pir) (cbs) (pbs)
2107 * ``profile_id``: ID for the meter profile.
2108 * ``cir``: Committed information rate (bytes/second).
2109 * ``pir``: Peak information rate (bytes/second).
2110 * ``cbs``: Committed burst size (bytes).
2111 * ``pbs``: Peak burst size (bytes).
2113 add port meter profile (trTCM rfc4115)
2114 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2116 Add meter profile (trTCM rfc4115) to the ethernet device::
2118 testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \
2119 (cir) (eir) (cbs) (ebs)
2123 * ``profile_id``: ID for the meter profile.
2124 * ``cir``: Committed information rate (bytes/second).
2125 * ``eir``: Excess information rate (bytes/second).
2126 * ``cbs``: Committed burst size (bytes).
2127 * ``ebs``: Excess burst size (bytes).
2129 delete port meter profile
2130 ~~~~~~~~~~~~~~~~~~~~~~~~~
2132 Delete meter profile from the ethernet device::
2134 testpmd> del port meter profile (port_id) (profile_id)
2139 Create new meter object for the ethernet device::
2141 testpmd> create port meter (port_id) (mtr_id) (profile_id) \
2142 (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \
2143 (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\
2148 * ``mtr_id``: meter object ID.
2149 * ``profile_id``: ID for the meter profile.
2150 * ``meter_enable``: When this parameter has a non-zero value, the meter object
2151 gets enabled at the time of creation, otherwise remains disabled.
2152 * ``g_action``: Policer action for the packet with green color.
2153 * ``y_action``: Policer action for the packet with yellow color.
2154 * ``r_action``: Policer action for the packet with red color.
2155 * ``stats_mask``: Mask of statistics counter types to be enabled for the
2157 * ``shared``: When this parameter has a non-zero value, the meter object is
2158 shared by multiple flows. Otherwise, meter object is used by single flow.
2159 * ``use_pre_meter_color``: When this parameter has a non-zero value, the
2160 input color for the current meter object is determined by the latest meter
2161 object in the same flow. Otherwise, the current meter object uses the
2162 *dscp_table* to determine the input color.
2163 * ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input
2164 color, 0 <= x <= 63.
2169 Enable meter for the ethernet device::
2171 testpmd> enable port meter (port_id) (mtr_id)
2176 Disable meter for the ethernet device::
2178 testpmd> disable port meter (port_id) (mtr_id)
2183 Delete meter for the ethernet device::
2185 testpmd> del port meter (port_id) (mtr_id)
2187 Set port meter profile
2188 ~~~~~~~~~~~~~~~~~~~~~~
2190 Set meter profile for the ethernet device::
2192 testpmd> set port meter profile (port_id) (mtr_id) (profile_id)
2194 set port meter dscp table
2195 ~~~~~~~~~~~~~~~~~~~~~~~~~
2197 Set meter dscp table for the ethernet device::
2199 testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \
2200 (dscp_tbl_entry1)...(dscp_tbl_entry63)]
2202 set port meter policer action
2203 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2205 Set meter policer action for the ethernet device::
2207 testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \
2208 (action0) [(action1) (action1)]
2212 * ``action_mask``: Bit mask indicating which policer actions need to be
2213 updated. One or more policer actions can be updated in a single function
2214 invocation. To update the policer action associated with color C, bit
2215 (1 << C) needs to be set in *action_mask* and element at position C
2216 in the *actions* array needs to be valid.
2217 * ``actionx``: Policer action for the color x,
2218 RTE_MTR_GREEN <= x < RTE_MTR_COLORS
2220 set port meter stats mask
2221 ~~~~~~~~~~~~~~~~~~~~~~~~~
2223 Set meter stats mask for the ethernet device::
2225 testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask)
2229 * ``stats_mask``: Bit mask indicating statistics counter types to be enabled.
2231 show port meter stats
2232 ~~~~~~~~~~~~~~~~~~~~~
2234 Show meter stats of the ethernet device::
2236 testpmd> show port meter stats (port_id) (mtr_id) (clear)
2240 * ``clear``: Flag that indicates whether the statistics counters should
2241 be cleared (i.e. set to zero) immediately after they have been read or not.
2246 The following section shows functions for configuring traffic management on
2247 on the ethernet device through the use of generic TM API.
2249 show port traffic management capability
2250 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2252 Show traffic management capability of the port::
2254 testpmd> show port tm cap (port_id)
2256 show port traffic management capability (hierarchy level)
2257 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2259 Show traffic management hierarchy level capability of the port::
2261 testpmd> show port tm cap (port_id) (level_id)
2263 show port traffic management capability (hierarchy node level)
2264 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2266 Show the traffic management hierarchy node capability of the port::
2268 testpmd> show port tm cap (port_id) (node_id)
2270 show port traffic management hierarchy node type
2271 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2273 Show the port traffic management hierarchy node type::
2275 testpmd> show port tm node type (port_id) (node_id)
2277 show port traffic management hierarchy node stats
2278 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2280 Show the port traffic management hierarchy node statistics::
2282 testpmd> show port tm node stats (port_id) (node_id) (clear)
2286 * ``clear``: When this parameter has a non-zero value, the statistics counters
2287 are cleared (i.e. set to zero) immediately after they have been read,
2288 otherwise the statistics counters are left untouched.
2290 Add port traffic management private shaper profile
2291 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2293 Add the port traffic management private shaper profile::
2295 testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \
2296 (tb_rate) (tb_size) (packet_length_adjust)
2300 * ``shaper_profile id``: Shaper profile ID for the new profile.
2301 * ``tb_rate``: Token bucket rate (bytes per second).
2302 * ``tb_size``: Token bucket size (bytes).
2303 * ``packet_length_adjust``: The value (bytes) to be added to the length of
2304 each packet for the purpose of shaping. This parameter value can be used to
2305 correct the packet length with the framing overhead bytes that are consumed
2308 Delete port traffic management private shaper profile
2309 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2311 Delete the port traffic management private shaper::
2313 testpmd> del port tm node shaper profile (port_id) (shaper_profile_id)
2317 * ``shaper_profile id``: Shaper profile ID that needs to be deleted.
2319 Add port traffic management shared shaper
2320 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2322 Create the port traffic management shared shaper::
2324 testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \
2329 * ``shared_shaper_id``: Shared shaper ID to be created.
2330 * ``shaper_profile id``: Shaper profile ID for shared shaper.
2332 Set port traffic management shared shaper
2333 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2335 Update the port traffic management shared shaper::
2337 testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \
2342 * ``shared_shaper_id``: Shared shaper ID to be update.
2343 * ``shaper_profile id``: Shaper profile ID for shared shaper.
2345 Delete port traffic management shared shaper
2346 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2348 Delete the port traffic management shared shaper::
2350 testpmd> del port tm node shared shaper (port_id) (shared_shaper_id)
2354 * ``shared_shaper_id``: Shared shaper ID to be deleted.
2356 Set port traffic management hiearchy node private shaper
2357 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2359 set the port traffic management hierarchy node private shaper::
2361 testpmd> set port tm node shaper profile (port_id) (node_id) \
2366 * ``shaper_profile id``: Private shaper profile ID to be enabled on the
2369 Add port traffic management WRED profile
2370 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2372 Create a new WRED profile::
2374 testpmd> add port tm node wred profile (port_id) (wred_profile_id) \
2375 (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \
2376 (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \
2377 (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r)
2381 * ``wred_profile id``: Identifier for the newly create WRED profile
2382 * ``color_g``: Packet color (green)
2383 * ``min_th_g``: Minimum queue threshold for packet with green color
2384 * ``max_th_g``: Minimum queue threshold for packet with green color
2385 * ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp)
2386 * ``wq_log2_g``: Negated log2 of queue weight (wq)
2387 * ``color_y``: Packet color (yellow)
2388 * ``min_th_y``: Minimum queue threshold for packet with yellow color
2389 * ``max_th_y``: Minimum queue threshold for packet with yellow color
2390 * ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp)
2391 * ``wq_log2_y``: Negated log2 of queue weight (wq)
2392 * ``color_r``: Packet color (red)
2393 * ``min_th_r``: Minimum queue threshold for packet with yellow color
2394 * ``max_th_r``: Minimum queue threshold for packet with yellow color
2395 * ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp)
2396 * ``wq_log2_r``: Negated log2 of queue weight (wq)
2398 Delete port traffic management WRED profile
2399 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2401 Delete the WRED profile::
2403 testpmd> del port tm node wred profile (port_id) (wred_profile_id)
2405 Add port traffic management hierarchy nonleaf node
2406 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2408 Add nonleaf node to port traffic management hiearchy::
2410 testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \
2411 (priority) (weight) (level_id) (shaper_profile_id) \
2412 (n_sp_priorities) (stats_mask) (n_shared_shapers) \
2413 [(shared_shaper_0) (shared_shaper_1) ...] \
2417 * ``parent_node_id``: Node ID of the parent.
2418 * ``priority``: Node priority (highest node priority is zero). This is used by
2419 the SP algorithm running on the parent node for scheduling this node.
2420 * ``weight``: Node weight (lowest weight is one). The node weight is relative
2421 to the weight sum of all siblings that have the same priority. It is used by
2422 the WFQ algorithm running on the parent node for scheduling this node.
2423 * ``level_id``: Hiearchy level of the node.
2424 * ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2426 * ``n_sp_priorities``: Number of strict priorities.
2427 * ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2428 * ``n_shared_shapers``: Number of shared shapers.
2429 * ``shared_shaper_id``: Shared shaper id.
2431 Add port traffic management hierarchy leaf node
2432 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2434 Add leaf node to port traffic management hiearchy::
2436 testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \
2437 (priority) (weight) (level_id) (shaper_profile_id) \
2438 (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \
2439 [(shared_shaper_id) (shared_shaper_id) ...] \
2443 * ``parent_node_id``: Node ID of the parent.
2444 * ``priority``: Node priority (highest node priority is zero). This is used by
2445 the SP algorithm running on the parent node for scheduling this node.
2446 * ``weight``: Node weight (lowest weight is one). The node weight is relative
2447 to the weight sum of all siblings that have the same priority. It is used by
2448 the WFQ algorithm running on the parent node for scheduling this node.
2449 * ``level_id``: Hiearchy level of the node.
2450 * ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2452 * ``cman_mode``: Congestion management mode to be enabled for this node.
2453 * ``wred_profile_id``: WRED profile id to be enabled for this node.
2454 * ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2455 * ``n_shared_shapers``: Number of shared shapers.
2456 * ``shared_shaper_id``: Shared shaper id.
2458 Delete port traffic management hierarchy node
2459 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2461 Delete node from port traffic management hiearchy::
2463 testpmd> del port tm node (port_id) (node_id)
2465 Update port traffic management hierarchy parent node
2466 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2468 Update port traffic management hierarchy parent node::
2470 testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \
2473 This function can only be called after the hierarchy commit invocation. Its
2474 success depends on the port support for this operation, as advertised through
2475 the port capability set. This function is valid for all nodes of the traffic
2476 management hierarchy except root node.
2478 Commit port traffic management hierarchy
2479 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2481 Commit the traffic management hierarchy on the port::
2483 testpmd> port tm hierarchy commit (port_id) (clean_on_fail)
2487 * ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function
2488 call failure. On the other hand, hierarchy is preserved when this parameter
2491 Set port traffic management default hierarchy (tm forwarding mode)
2492 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2494 set the traffic management default hierarchy on the port::
2496 testpmd> set port tm hierarchy default (port_id)
2501 This section details the available filter functions that are available.
2503 Note these functions interface the deprecated legacy filtering framework,
2504 superseded by *rte_flow*. See `Flow rules management`_.
2507 ~~~~~~~~~~~~~~~~~~~~
2509 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
2511 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
2512 ethertype (ether_type) (drop|fwd) queue (queue_id)
2514 The available information parameters are:
2516 * ``port_id``: The port which the Ethertype filter assigned on.
2518 * ``mac_addr``: Compare destination mac address.
2520 * ``mac_ignr``: Ignore destination mac address match.
2522 * ``mac_address``: Destination mac address to match.
2524 * ``ether_type``: The EtherType value want to match,
2525 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
2527 * ``queue_id``: The receive queue associated with this EtherType filter.
2528 It is meaningless when deleting or dropping.
2530 Example, to add/remove an ethertype filter rule::
2532 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
2533 ethertype 0x0806 fwd queue 3
2535 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
2536 ethertype 0x0806 fwd queue 3
2541 Add or delete a 2-tuple filter,
2542 which identifies packets by specific protocol and destination TCP/UDP port
2543 and forwards packets into one of the receive queues::
2545 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
2546 protocol (protocol_value) mask (mask_value) \
2547 tcp_flags (tcp_flags_value) priority (prio_value) \
2550 The available information parameters are:
2552 * ``port_id``: The port which the 2-tuple filter assigned on.
2554 * ``dst_port_value``: Destination port in L4.
2556 * ``protocol_value``: IP L4 protocol.
2558 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
2560 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
2562 * ``prio_value``: Priority of this filter.
2564 * ``queue_id``: The receive queue associated with this 2-tuple filter.
2566 Example, to add/remove an 2tuple filter rule::
2568 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
2569 tcp_flags 0x02 priority 3 queue 3
2571 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
2572 tcp_flags 0x02 priority 3 queue 3
2577 Add or delete a 5-tuple filter,
2578 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
2579 and routes packets into one of the receive queues::
2581 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
2582 (src_address) dst_port (dst_port_value) \
2583 src_port (src_port_value) protocol (protocol_value) \
2584 mask (mask_value) tcp_flags (tcp_flags_value) \
2585 priority (prio_value) queue (queue_id)
2587 The available information parameters are:
2589 * ``port_id``: The port which the 5-tuple filter assigned on.
2591 * ``dst_address``: Destination IP address.
2593 * ``src_address``: Source IP address.
2595 * ``dst_port_value``: TCP/UDP destination port.
2597 * ``src_port_value``: TCP/UDP source port.
2599 * ``protocol_value``: L4 protocol.
2601 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
2603 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
2605 * ``prio_value``: The priority of this filter.
2607 * ``queue_id``: The receive queue associated with this 5-tuple filter.
2609 Example, to add/remove an 5tuple filter rule::
2611 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2612 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2613 flags 0x0 priority 3 queue 3
2615 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2616 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2617 flags 0x0 priority 3 queue 3
2622 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
2624 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
2626 The available information parameters are:
2628 * ``port_id``: The port which the SYN filter assigned on.
2630 * ``high``: This SYN filter has higher priority than other filters.
2632 * ``low``: This SYN filter has lower priority than other filters.
2634 * ``queue_id``: The receive queue associated with this SYN filter
2638 testpmd> syn_filter 0 add priority high queue 3
2643 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
2644 and routed into one of the receive queues::
2646 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
2647 mask (mask_value) priority (prio_value) queue (queue_id)
2649 The available information parameters are:
2651 * ``port_id``: The port which the Flex filter is assigned on.
2653 * ``len_value``: Filter length in bytes, no greater than 128.
2655 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
2657 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
2659 * ``prio_value``: The priority of this filter.
2661 * ``queue_id``: The receive queue associated with this Flex filter.
2665 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
2666 mask 000C priority 3 queue 3
2668 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
2669 mask 000C priority 3 queue 3
2672 .. _testpmd_flow_director:
2674 flow_director_filter
2675 ~~~~~~~~~~~~~~~~~~~~
2677 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
2679 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
2680 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
2682 * Perfect match filters.
2683 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2684 The masked fields are for IP flow.
2686 * Signature filters.
2687 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
2689 * Perfect-mac-vlan match filters.
2690 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2691 The masked fields are for MAC VLAN flow.
2693 * Perfect-tunnel match filters.
2694 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2695 The masked fields are for tunnel flow.
2697 * Perfect-raw-flow-type match filters.
2698 The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet.
2699 The masked fields are specified by input sets.
2701 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
2702 per flow type and the flexible payload.
2704 The Flow Director can also mask out parts of all of these fields so that filters
2705 are only applied to certain fields or parts of the fields.
2707 Note that for raw flow type mode the source and destination fields in the
2708 raw packet buffer need to be presented in a reversed order with respect
2709 to the expected received packets.
2710 For example: IP source and destination addresses or TCP/UDP/SCTP
2711 source and destination ports
2713 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
2715 # Commands to add flow director filters of different flow types::
2717 flow_director_filter (port_id) mode IP (add|del|update) \
2718 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
2719 src (src_ip_address) dst (dst_ip_address) \
2720 tos (tos_value) proto (proto_value) ttl (ttl_value) \
2721 vlan (vlan_value) flexbytes (flexbytes_value) \
2722 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
2725 flow_director_filter (port_id) mode IP (add|del|update) \
2726 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
2727 src (src_ip_address) (src_port) \
2728 dst (dst_ip_address) (dst_port) \
2729 tos (tos_value) ttl (ttl_value) \
2730 vlan (vlan_value) flexbytes (flexbytes_value) \
2731 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
2734 flow_director_filter (port_id) mode IP (add|del|update) \
2735 flow (ipv4-sctp|ipv6-sctp) \
2736 src (src_ip_address) (src_port) \
2737 dst (dst_ip_address) (dst_port) \
2738 tos (tos_value) ttl (ttl_value) \
2739 tag (verification_tag) vlan (vlan_value) \
2740 flexbytes (flexbytes_value) (drop|fwd) \
2741 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
2743 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
2744 ether (ethertype) flexbytes (flexbytes_value) \
2745 (drop|fwd) pf|vf(vf_id) queue (queue_id)
2748 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
2749 mac (mac_address) vlan (vlan_value) \
2750 flexbytes (flexbytes_value) (drop|fwd) \
2751 queue (queue_id) fd_id (fd_id_value)
2753 flow_director_filter (port_id) mode Tunnel (add|del|update) \
2754 mac (mac_address) vlan (vlan_value) \
2755 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
2756 flexbytes (flexbytes_value) (drop|fwd) \
2757 queue (queue_id) fd_id (fd_id_value)
2759 flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \
2760 (drop|fwd) queue (queue_id) fd_id (fd_id_value) \
2761 packet (packet file name)
2763 For example, to add an ipv4-udp flow type filter::
2765 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
2766 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
2767 fwd pf queue 1 fd_id 1
2769 For example, add an ipv4-other flow type filter::
2771 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
2772 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
2773 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
2778 Flush all flow director filters on a device::
2780 testpmd> flush_flow_director (port_id)
2782 Example, to flush all flow director filter on port 0::
2784 testpmd> flush_flow_director 0
2789 Set flow director's input masks::
2791 flow_director_mask (port_id) mode IP vlan (vlan_value) \
2792 src_mask (ipv4_src) (ipv6_src) (src_port) \
2793 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2795 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2797 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2798 mac (mac_value) tunnel-type (tunnel_type_value) \
2799 tunnel-id (tunnel_id_value)
2801 Example, to set flow director mask on port 0::
2803 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2804 src_mask 255.255.255.255 \
2805 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2806 dst_mask 255.255.255.255 \
2807 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2809 flow_director_flex_mask
2810 ~~~~~~~~~~~~~~~~~~~~~~~
2812 set masks of flow director's flexible payload based on certain flow type::
2814 testpmd> flow_director_flex_mask (port_id) \
2815 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2816 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2817 l2_payload|all) (mask)
2819 Example, to set flow director's flex mask for all flow type on port 0::
2821 testpmd> flow_director_flex_mask 0 flow all \
2822 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2825 flow_director_flex_payload
2826 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2828 Configure flexible payload selection::
2830 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2832 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2834 testpmd> flow_director_flex_payload 0 l4 \
2835 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2837 get_sym_hash_ena_per_port
2838 ~~~~~~~~~~~~~~~~~~~~~~~~~
2840 Get symmetric hash enable configuration per port::
2842 get_sym_hash_ena_per_port (port_id)
2844 For example, to get symmetric hash enable configuration of port 1::
2846 testpmd> get_sym_hash_ena_per_port 1
2848 set_sym_hash_ena_per_port
2849 ~~~~~~~~~~~~~~~~~~~~~~~~~
2851 Set symmetric hash enable configuration per port to enable or disable::
2853 set_sym_hash_ena_per_port (port_id) (enable|disable)
2855 For example, to set symmetric hash enable configuration of port 1 to enable::
2857 testpmd> set_sym_hash_ena_per_port 1 enable
2859 get_hash_global_config
2860 ~~~~~~~~~~~~~~~~~~~~~~
2862 Get the global configurations of hash filters::
2864 get_hash_global_config (port_id)
2866 For example, to get the global configurations of hash filters of port 1::
2868 testpmd> get_hash_global_config 1
2870 set_hash_global_config
2871 ~~~~~~~~~~~~~~~~~~~~~~
2873 Set the global configurations of hash filters::
2875 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2876 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2877 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \
2880 For example, to enable simple_xor for flow type of ipv6 on port 2::
2882 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2887 Set the input set for hash::
2889 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2890 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2891 l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
2892 ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2893 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2894 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2895 fld-8th|none) (select|add)
2897 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2899 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2904 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2905 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2907 Set the input set for flow director::
2909 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2910 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2911 l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
2912 ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2913 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2914 sctp-dst-port|sctp-veri-tag|none) (select|add)
2916 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2918 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2923 Set different GRE key length for input set::
2925 global_config (port_id) gre-key-len (number in bytes)
2927 For example to set GRE key length for input set to 4 bytes on port 0::
2929 testpmd> global_config 0 gre-key-len 4
2932 .. _testpmd_rte_flow:
2934 Flow rules management
2935 ---------------------
2937 Control of the generic flow API (*rte_flow*) is fully exposed through the
2938 ``flow`` command (validation, creation, destruction, queries and operation
2941 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2942 features simultaneously may cause undefined side-effects and is therefore
2948 Because the ``flow`` command uses dynamic tokens to handle the large number
2949 of possible flow rules combinations, its behavior differs slightly from
2950 other commands, in particular:
2952 - Pressing *?* or the *<tab>* key displays contextual help for the current
2953 token, not that of the entire command.
2955 - Optional and repeated parameters are supported (provided they are listed
2956 in the contextual help).
2958 The first parameter stands for the operation mode. Possible operations and
2959 their general syntax are described below. They are covered in detail in the
2962 - Check whether a flow rule can be created::
2964 flow validate {port_id}
2965 [group {group_id}] [priority {level}] [ingress] [egress]
2966 pattern {item} [/ {item} [...]] / end
2967 actions {action} [/ {action} [...]] / end
2969 - Create a flow rule::
2971 flow create {port_id}
2972 [group {group_id}] [priority {level}] [ingress] [egress]
2973 pattern {item} [/ {item} [...]] / end
2974 actions {action} [/ {action} [...]] / end
2976 - Destroy specific flow rules::
2978 flow destroy {port_id} rule {rule_id} [...]
2980 - Destroy all flow rules::
2982 flow flush {port_id}
2984 - Query an existing flow rule::
2986 flow query {port_id} {rule_id} {action}
2988 - List existing flow rules sorted by priority, filtered by group
2991 flow list {port_id} [group {group_id}] [...]
2993 - Restrict ingress traffic to the defined flow rules::
2995 flow isolate {port_id} {boolean}
2997 Validating flow rules
2998 ~~~~~~~~~~~~~~~~~~~~~
3000 ``flow validate`` reports whether a flow rule would be accepted by the
3001 underlying device in its current state but stops short of creating it. It is
3002 bound to ``rte_flow_validate()``::
3004 flow validate {port_id}
3005 [group {group_id}] [priority {level}] [ingress] [egress]
3006 pattern {item} [/ {item} [...]] / end
3007 actions {action} [/ {action} [...]] / end
3009 If successful, it will show::
3013 Otherwise it will show an error message of the form::
3015 Caught error type [...] ([...]): [...]
3017 This command uses the same parameters as ``flow create``, their format is
3018 described in `Creating flow rules`_.
3020 Check whether redirecting any Ethernet packet received on port 0 to RX queue
3021 index 6 is supported::
3023 testpmd> flow validate 0 ingress pattern eth / end
3024 actions queue index 6 / end
3028 Port 0 does not support TCPv6 rules::
3030 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
3032 Caught error type 9 (specific pattern item): Invalid argument
3038 ``flow create`` validates and creates the specified flow rule. It is bound
3039 to ``rte_flow_create()``::
3041 flow create {port_id}
3042 [group {group_id}] [priority {level}] [ingress] [egress]
3043 pattern {item} [/ {item} [...]] / end
3044 actions {action} [/ {action} [...]] / end
3046 If successful, it will return a flow rule ID usable with other commands::
3048 Flow rule #[...] created
3050 Otherwise it will show an error message of the form::
3052 Caught error type [...] ([...]): [...]
3054 Parameters describe in the following order:
3056 - Attributes (*group*, *priority*, *ingress*, *egress* tokens).
3057 - A matching pattern, starting with the *pattern* token and terminated by an
3059 - Actions, starting with the *actions* token and terminated by an *end*
3062 These translate directly to *rte_flow* objects provided as-is to the
3063 underlying functions.
3065 The shortest valid definition only comprises mandatory tokens::
3067 testpmd> flow create 0 pattern end actions end
3069 Note that PMDs may refuse rules that essentially do nothing such as this
3072 **All unspecified object values are automatically initialized to 0.**
3077 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
3078 specified before the ``pattern`` token.
3080 - ``group {group id}``: priority group.
3081 - ``priority {level}``: priority level within group.
3082 - ``ingress``: rule applies to ingress traffic.
3083 - ``egress``: rule applies to egress traffic.
3085 Each instance of an attribute specified several times overrides the previous
3086 value as shown below (group 4 is used)::
3088 testpmd> flow create 0 group 42 group 24 group 4 [...]
3090 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
3092 While not specifying a direction is an error, some rules may allow both
3095 Most rules affect RX therefore contain the ``ingress`` token::
3097 testpmd> flow create 0 ingress pattern [...]
3102 A matching pattern starts after the ``pattern`` token. It is made of pattern
3103 items and is terminated by a mandatory ``end`` item.
3105 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
3106 rte_flow_item_type``).
3108 The ``/`` token is used as a separator between pattern items as shown
3111 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
3113 Note that protocol items like these must be stacked from lowest to highest
3114 layer to make sense. For instance, the following rule is either invalid or
3115 unlikely to match any packet::
3117 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
3119 More information on these restrictions can be found in the *rte_flow*
3122 Several items support additional specification structures, for example
3123 ``ipv4`` allows specifying source and destination addresses as follows::
3125 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
3126 dst is 10.2.0.0 / end [...]
3128 This rule matches all IPv4 traffic with the specified properties.
3130 In this example, ``src`` and ``dst`` are field names of the underlying
3131 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
3132 in a similar fashion.
3134 The ``is`` token means that the subsequent value must be matched exactly,
3135 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
3136 accordingly. Possible assignment tokens are:
3138 - ``is``: match value perfectly (with full bit-mask).
3139 - ``spec``: match value according to configured bit-mask.
3140 - ``last``: specify upper bound to establish a range.
3141 - ``mask``: specify bit-mask with relevant bits set to one.
3142 - ``prefix``: generate bit-mask from a prefix length.
3144 These yield identical results::
3146 ipv4 src is 10.1.1.1
3150 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
3154 ipv4 src spec 10.1.1.1 src prefix 32
3158 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
3162 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
3164 Inclusive ranges can be defined with ``last``::
3166 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
3168 Note that ``mask`` affects both ``spec`` and ``last``::
3170 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
3171 # matches 10.1.0.0 to 10.2.255.255
3173 Properties can be modified multiple times::
3175 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
3179 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
3184 This section lists supported pattern items and their attributes, if any.
3186 - ``end``: end list of pattern items.
3188 - ``void``: no-op pattern item.
3190 - ``invert``: perform actions when pattern does not match.
3192 - ``any``: match any protocol for the current layer.
3194 - ``num {unsigned}``: number of layers covered.
3196 - ``pf``: match packets addressed to the physical function.
3198 - ``vf``: match packets addressed to a virtual function ID.
3200 - ``id {unsigned}``: destination VF ID.
3202 - ``port``: device-specific physical port index to use.
3204 - ``index {unsigned}``: physical port index.
3206 - ``raw``: match an arbitrary byte string.
3208 - ``relative {boolean}``: look for pattern after the previous item.
3209 - ``search {boolean}``: search pattern from offset (see also limit).
3210 - ``offset {integer}``: absolute or relative offset for pattern.
3211 - ``limit {unsigned}``: search area limit for start of pattern.
3212 - ``pattern {string}``: byte string to look for.
3214 - ``eth``: match Ethernet header.
3216 - ``dst {MAC-48}``: destination MAC.
3217 - ``src {MAC-48}``: source MAC.
3218 - ``type {unsigned}``: EtherType.
3220 - ``vlan``: match 802.1Q/ad VLAN tag.
3222 - ``tpid {unsigned}``: tag protocol identifier.
3223 - ``tci {unsigned}``: tag control information.
3224 - ``pcp {unsigned}``: priority code point.
3225 - ``dei {unsigned}``: drop eligible indicator.
3226 - ``vid {unsigned}``: VLAN identifier.
3228 - ``ipv4``: match IPv4 header.
3230 - ``tos {unsigned}``: type of service.
3231 - ``ttl {unsigned}``: time to live.
3232 - ``proto {unsigned}``: next protocol ID.
3233 - ``src {ipv4 address}``: source address.
3234 - ``dst {ipv4 address}``: destination address.
3236 - ``ipv6``: match IPv6 header.
3238 - ``tc {unsigned}``: traffic class.
3239 - ``flow {unsigned}``: flow label.
3240 - ``proto {unsigned}``: protocol (next header).
3241 - ``hop {unsigned}``: hop limit.
3242 - ``src {ipv6 address}``: source address.
3243 - ``dst {ipv6 address}``: destination address.
3245 - ``icmp``: match ICMP header.
3247 - ``type {unsigned}``: ICMP packet type.
3248 - ``code {unsigned}``: ICMP packet code.
3250 - ``udp``: match UDP header.
3252 - ``src {unsigned}``: UDP source port.
3253 - ``dst {unsigned}``: UDP destination port.
3255 - ``tcp``: match TCP header.
3257 - ``src {unsigned}``: TCP source port.
3258 - ``dst {unsigned}``: TCP destination port.
3260 - ``sctp``: match SCTP header.
3262 - ``src {unsigned}``: SCTP source port.
3263 - ``dst {unsigned}``: SCTP destination port.
3264 - ``tag {unsigned}``: validation tag.
3265 - ``cksum {unsigned}``: checksum.
3267 - ``vxlan``: match VXLAN header.
3269 - ``vni {unsigned}``: VXLAN identifier.
3271 - ``e_tag``: match IEEE 802.1BR E-Tag header.
3273 - ``grp_ecid_b {unsigned}``: GRP and E-CID base.
3275 - ``nvgre``: match NVGRE header.
3277 - ``tni {unsigned}``: virtual subnet ID.
3279 - ``mpls``: match MPLS header.
3281 - ``label {unsigned}``: MPLS label.
3283 - ``gre``: match GRE header.
3285 - ``protocol {unsigned}``: protocol type.
3287 - ``fuzzy``: fuzzy pattern match, expect faster than default.
3289 - ``thresh {unsigned}``: accuracy threshold.
3291 - ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header.
3293 - ``teid {unsigned}``: tunnel endpoint identifier.
3295 - ``geneve``: match GENEVE header.
3297 - ``vni {unsigned}``: virtual network identifier.
3298 - ``protocol {unsigned}``: protocol type.
3303 A list of actions starts after the ``actions`` token in the same fashion as
3304 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
3305 terminated by a mandatory ``end`` action.
3307 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
3308 rte_flow_action_type``).
3310 Dropping all incoming UDPv4 packets can be expressed as follows::
3312 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3315 Several actions have configurable properties which must be specified when
3316 there is no valid default value. For example, ``queue`` requires a target
3319 This rule redirects incoming UDPv4 traffic to queue index 6::
3321 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3322 actions queue index 6 / end
3324 While this one could be rejected by PMDs (unspecified queue index)::
3326 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3329 As defined by *rte_flow*, the list is not ordered, all actions of a given
3330 rule are performed simultaneously. These are equivalent::
3332 queue index 6 / void / mark id 42 / end
3336 void / mark id 42 / queue index 6 / end
3338 All actions in a list should have different types, otherwise only the last
3339 action of a given type is taken into account::
3341 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
3345 drop / drop / drop / end # drop is performed only once
3349 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
3351 Considering they are performed simultaneously, opposite and overlapping
3352 actions can sometimes be combined when the end result is unambiguous::
3354 drop / queue index 6 / end # drop has no effect
3358 drop / dup index 6 / end # same as above
3362 queue index 6 / rss queues 6 7 8 / end # queue has no effect
3366 drop / passthru / end # drop has no effect
3368 Note that PMDs may still refuse such combinations.
3373 This section lists supported actions and their attributes, if any.
3375 - ``end``: end list of actions.
3377 - ``void``: no-op action.
3379 - ``passthru``: let subsequent rule process matched packets.
3381 - ``mark``: attach 32 bit value to packets.
3383 - ``id {unsigned}``: 32 bit value to return with packets.
3385 - ``flag``: flag packets.
3387 - ``queue``: assign packets to a given queue index.
3389 - ``index {unsigned}``: queue index to use.
3391 - ``drop``: drop packets (note: passthru has priority).
3393 - ``count``: enable counters for this rule.
3395 - ``dup``: duplicate packets to a given queue index.
3397 - ``index {unsigned}``: queue index to duplicate packets to.
3399 - ``rss``: spread packets among several queues.
3401 - ``queues [{unsigned} [...]] end``: queue indices to use.
3403 - ``pf``: redirect packets to physical device function.
3405 - ``vf``: redirect packets to virtual device function.
3407 - ``original {boolean}``: use original VF ID if possible.
3408 - ``id {unsigned}``: VF ID to redirect packets to.
3410 Destroying flow rules
3411 ~~~~~~~~~~~~~~~~~~~~~
3413 ``flow destroy`` destroys one or more rules from their rule ID (as returned
3414 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
3415 times as necessary::
3417 flow destroy {port_id} rule {rule_id} [...]
3419 If successful, it will show::
3421 Flow rule #[...] destroyed
3423 It does not report anything for rule IDs that do not exist. The usual error
3424 message is shown when a rule cannot be destroyed::
3426 Caught error type [...] ([...]): [...]
3428 ``flow flush`` destroys all rules on a device and does not take extra
3429 arguments. It is bound to ``rte_flow_flush()``::
3431 flow flush {port_id}
3433 Any errors are reported as above.
3435 Creating several rules and destroying them::
3437 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3438 actions queue index 2 / end
3439 Flow rule #0 created
3440 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3441 actions queue index 3 / end
3442 Flow rule #1 created
3443 testpmd> flow destroy 0 rule 0 rule 1
3444 Flow rule #1 destroyed
3445 Flow rule #0 destroyed
3448 The same result can be achieved using ``flow flush``::
3450 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3451 actions queue index 2 / end
3452 Flow rule #0 created
3453 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3454 actions queue index 3 / end
3455 Flow rule #1 created
3456 testpmd> flow flush 0
3459 Non-existent rule IDs are ignored::
3461 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3462 actions queue index 2 / end
3463 Flow rule #0 created
3464 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3465 actions queue index 3 / end
3466 Flow rule #1 created
3467 testpmd> flow destroy 0 rule 42 rule 10 rule 2
3469 testpmd> flow destroy 0 rule 0
3470 Flow rule #0 destroyed
3476 ``flow query`` queries a specific action of a flow rule having that
3477 ability. Such actions collect information that can be reported using this
3478 command. It is bound to ``rte_flow_query()``::
3480 flow query {port_id} {rule_id} {action}
3482 If successful, it will display either the retrieved data for known actions
3483 or the following message::
3485 Cannot display result for action type [...] ([...])
3487 Otherwise, it will complain either that the rule does not exist or that some
3490 Flow rule #[...] not found
3494 Caught error type [...] ([...]): [...]
3496 Currently only the ``count`` action is supported. This action reports the
3497 number of packets that hit the flow rule and the total number of bytes. Its
3498 output has the following format::
3501 hits_set: [...] # whether "hits" contains a valid value
3502 bytes_set: [...] # whether "bytes" contains a valid value
3503 hits: [...] # number of packets
3504 bytes: [...] # number of bytes
3506 Querying counters for TCPv6 packets redirected to queue 6::
3508 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
3509 actions queue index 6 / count / end
3510 Flow rule #4 created
3511 testpmd> flow query 0 4 count
3522 ``flow list`` lists existing flow rules sorted by priority and optionally
3523 filtered by group identifiers::
3525 flow list {port_id} [group {group_id}] [...]
3527 This command only fails with the following message if the device does not
3532 Output consists of a header line followed by a short description of each
3533 flow rule, one per line. There is no output at all when no flow rules are
3534 configured on the device::
3536 ID Group Prio Attr Rule
3537 [...] [...] [...] [...] [...]
3539 ``Attr`` column flags:
3541 - ``i`` for ``ingress``.
3542 - ``e`` for ``egress``.
3544 Creating several flow rules and listing them::
3546 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3547 actions queue index 6 / end
3548 Flow rule #0 created
3549 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3550 actions queue index 2 / end
3551 Flow rule #1 created
3552 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
3553 actions rss queues 6 7 8 end / end
3554 Flow rule #2 created
3555 testpmd> flow list 0
3556 ID Group Prio Attr Rule
3557 0 0 0 i- ETH IPV4 => QUEUE
3558 1 0 0 i- ETH IPV6 => QUEUE
3559 2 0 5 i- ETH IPV4 UDP => RSS
3562 Rules are sorted by priority (i.e. group ID first, then priority level)::
3564 testpmd> flow list 1
3565 ID Group Prio Attr Rule
3566 0 0 0 i- ETH => COUNT
3567 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3568 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3569 1 24 0 i- ETH IPV4 UDP => QUEUE
3570 4 24 10 i- ETH IPV4 TCP => DROP
3571 3 24 20 i- ETH IPV4 => DROP
3572 2 24 42 i- ETH IPV4 UDP => QUEUE
3573 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3576 Output can be limited to specific groups::
3578 testpmd> flow list 1 group 0 group 63
3579 ID Group Prio Attr Rule
3580 0 0 0 i- ETH => COUNT
3581 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3582 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3583 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3586 Toggling isolated mode
3587 ~~~~~~~~~~~~~~~~~~~~~~
3589 ``flow isolate`` can be used to tell the underlying PMD that ingress traffic
3590 must only be injected from the defined flow rules; that no default traffic
3591 is expected outside those rules and the driver is free to assign more
3592 resources to handle them. It is bound to ``rte_flow_isolate()``::
3594 flow isolate {port_id} {boolean}
3596 If successful, enabling or disabling isolated mode shows either::
3598 Ingress traffic on port [...]
3599 is now restricted to the defined flow rules
3603 Ingress traffic on port [...]
3604 is not restricted anymore to the defined flow rules
3606 Otherwise, in case of error::
3608 Caught error type [...] ([...]): [...]
3610 Mainly due to its side effects, PMDs supporting this mode may not have the
3611 ability to toggle it more than once without reinitializing affected ports
3612 first (e.g. by exiting testpmd).
3614 Enabling isolated mode::
3616 testpmd> flow isolate 0 true
3617 Ingress traffic on port 0 is now restricted to the defined flow rules
3620 Disabling isolated mode::
3622 testpmd> flow isolate 0 false
3623 Ingress traffic on port 0 is not restricted anymore to the defined flow rules
3626 Sample QinQ flow rules
3627 ~~~~~~~~~~~~~~~~~~~~~~
3629 Before creating QinQ rule(s) the following commands should be issued to enable QinQ::
3631 testpmd> port stop 0
3632 testpmd> vlan set qinq on 0
3634 The above command sets the inner and outer TPID's to 0x8100.
3636 To change the TPID's the following commands should be used::
3638 testpmd> vlan set outer tpid 0xa100 0
3639 testpmd> vlan set inner tpid 0x9100 0
3640 testpmd> port start 0
3642 Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM.
3646 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 /
3647 vlan tci is 456 / end actions vf id 1 / queue index 0 / end
3648 Flow rule #0 validated
3650 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 /
3651 vlan tci is 456 / end actions vf id 123 / queue index 0 / end
3652 Flow rule #0 created
3654 testpmd> flow list 0
3655 ID Group Prio Attr Rule
3656 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3658 Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host.
3662 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 /
3663 vlan tci is 654 / end actions pf / queue index 0 / end
3664 Flow rule #1 validated
3666 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 /
3667 vlan tci is 654 / end actions pf / queue index 1 / end
3668 Flow rule #1 created
3670 testpmd> flow list 0
3671 ID Group Prio Attr Rule
3672 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3673 1 0 0 i- ETH VLAN VLAN=>PF QUEUE