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 unicast hash filter(s) on/off for a port::
1097 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
1102 Set the promiscuous mode on for a port or for all ports.
1103 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1105 testpmd> set promisc (port_id|all) (on|off)
1110 Set the allmulti mode for a port or for all ports::
1112 testpmd> set allmulti (port_id|all) (on|off)
1114 Same as the ifconfig (8) option. Controls how multicast packets are handled.
1116 set promisc (for VF)
1117 ~~~~~~~~~~~~~~~~~~~~
1119 Set the unicast promiscuous mode for a VF from PF.
1120 It's supported by Intel i40e NICs now.
1121 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1123 testpmd> set vf promisc (port_id) (vf_id) (on|off)
1125 set allmulticast (for VF)
1126 ~~~~~~~~~~~~~~~~~~~~~~~~~
1128 Set the multicast 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 allmulti (port_id) (vf_id) (on|off)
1134 set tx max bandwidth (for VF)
1135 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1137 Set TX max absolute bandwidth (Mbps) for a VF from PF::
1139 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth)
1141 set tc tx min bandwidth (for VF)
1142 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1144 Set all TCs' TX min relative bandwidth (%) for a VF from PF::
1146 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...)
1148 set tc tx max bandwidth (for VF)
1149 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1151 Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF::
1153 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth)
1155 set tc strict link priority mode
1156 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1158 Set some TCs' strict link priority mode on a physical port::
1160 testpmd> set tx strict-link-priority (port_id) (tc_bitmap)
1162 set tc tx min bandwidth
1163 ~~~~~~~~~~~~~~~~~~~~~~~
1165 Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs::
1167 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...)
1172 Set the link flow control parameter on a port::
1174 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1175 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
1176 autoneg (on|off) (port_id)
1180 * ``high_water`` (integer): High threshold value to trigger XOFF.
1182 * ``low_water`` (integer): Low threshold value to trigger XON.
1184 * ``pause_time`` (integer): Pause quota in the Pause frame.
1186 * ``send_xon`` (0/1): Send XON frame.
1188 * ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
1190 * ``autoneg``: Change the auto-negotiation parameter.
1195 Set the priority flow control parameter on a port::
1197 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1198 (pause_time) (priority) (port_id)
1202 * ``high_water`` (integer): High threshold value.
1204 * ``low_water`` (integer): Low threshold value.
1206 * ``pause_time`` (integer): Pause quota in the Pause frame.
1208 * ``priority`` (0-7): VLAN User Priority.
1213 Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
1215 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
1217 For example, to set rx queue 2 on port 0 to mapping 5::
1219 testpmd>set stat_qmap rx 0 2 5
1221 set xstats-hide-zero
1222 ~~~~~~~~~~~~~~~~~~~~
1224 Set the option to hide zero values for xstats display::
1226 testpmd> set xstats-hide-zero on|off
1230 By default, the zero values are displayed for xstats.
1232 set port - rx/tx (for VF)
1233 ~~~~~~~~~~~~~~~~~~~~~~~~~
1235 Set VF receive/transmit from a port::
1237 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
1239 set port - mac address filter (for VF)
1240 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1242 Add/Remove unicast or multicast MAC addr filter for a VF::
1244 testpmd> set port (port_id) vf (vf_id) (mac_addr) \
1245 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
1247 set port - rx mode(for VF)
1248 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1250 Set the VF receive mode of a port::
1252 testpmd> set port (port_id) vf (vf_id) \
1253 rxmode (AUPE|ROPE|BAM|MPE) (on|off)
1255 The available receive modes are:
1257 * ``AUPE``: Accepts untagged VLAN.
1259 * ``ROPE``: Accepts unicast hash.
1261 * ``BAM``: Accepts broadcast packets.
1263 * ``MPE``: Accepts all multicast packets.
1265 set port - tx_rate (for Queue)
1266 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1268 Set TX rate limitation for a queue on a port::
1270 testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1272 set port - tx_rate (for VF)
1273 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
1275 Set TX rate limitation for queues in VF on a port::
1277 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1279 set port - mirror rule
1280 ~~~~~~~~~~~~~~~~~~~~~~
1282 Set pool or vlan type mirror rule for a port::
1284 testpmd> set port (port_id) mirror-rule (rule_id) \
1285 (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1286 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1288 Set link mirror rule for a port::
1290 testpmd> set port (port_id) mirror-rule (rule_id) \
1291 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1293 For example to enable mirror traffic with vlan 0,1 to pool 0::
1295 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1297 reset port - mirror rule
1298 ~~~~~~~~~~~~~~~~~~~~~~~~
1300 Reset a mirror rule for a port::
1302 testpmd> reset port (port_id) mirror-rule (rule_id)
1307 Set the flush on RX streams before forwarding.
1308 The default is flush ``on``.
1309 Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1311 testpmd> set flush_rx off
1316 Set the bypass mode for the lowest port on bypass enabled NIC::
1318 testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1323 Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1325 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1326 mode (normal|bypass|isolate) (port_id)
1330 * ``timeout``: Enable bypass after watchdog timeout.
1332 * ``os_on``: Enable bypass when OS/board is powered on.
1334 * ``os_off``: Enable bypass when OS/board is powered off.
1336 * ``power_on``: Enable bypass when power supply is turned on.
1338 * ``power_off``: Enable bypass when power supply is turned off.
1344 Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1346 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1351 Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1353 testpmd> show bypass config (port_id)
1358 Set link up for a port::
1360 testpmd> set link-up port (port id)
1365 Set link down for a port::
1367 testpmd> set link-down port (port id)
1372 Enable E-tag insertion for a VF on a port::
1374 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1376 Disable E-tag insertion for a VF on a port::
1378 testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1380 Enable/disable E-tag stripping on a port::
1382 testpmd> E-tag set stripping (on|off) port (port_id)
1384 Enable/disable E-tag based forwarding on a port::
1386 testpmd> E-tag set forwarding (on|off) port (port_id)
1388 Add an E-tag forwarding filter on a port::
1390 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1392 Delete an E-tag forwarding filter on a port::
1393 testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1398 Load a dynamic device personalization (DDP) package::
1400 testpmd> ddp add (port_id) (package_path[,output_path])
1405 Delete a dynamic device personalization package::
1407 testpmd> ddp del (port_id) (package_path)
1412 List all items from the ptype mapping table::
1414 testpmd> ptype mapping get (port_id) (valid_only)
1418 * ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0).
1420 Replace a specific or a group of software defined ptype with a new one::
1422 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type)
1426 * ``target``: A specific software ptype or a mask to represent a group of software ptypes.
1428 * ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1).
1430 * ``pkt_type``: The new software ptype to replace the old ones.
1432 Update hardware defined ptype to software defined packet type mapping table::
1434 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype)
1438 * ``hw_ptype``: hardware ptype as the index of the ptype mapping table.
1440 * ``sw_ptype``: software ptype as the value of the ptype mapping table.
1442 Reset ptype mapping table::
1444 testpmd> ptype mapping reset (port_id)
1449 The following sections show functions for configuring ports.
1453 Port configuration changes only become active when forwarding is started/restarted.
1458 Attach a port specified by pci address or virtual device args::
1460 testpmd> port attach (identifier)
1462 To attach a new pci device, the device should be recognized by kernel first.
1463 Then it should be moved under DPDK management.
1464 Finally the port can be attached to testpmd.
1466 For example, to move a pci device using ixgbe under DPDK management:
1468 .. code-block:: console
1470 # Check the status of the available devices.
1471 ./usertools/dpdk-devbind.py --status
1473 Network devices using DPDK-compatible driver
1474 ============================================
1477 Network devices using kernel driver
1478 ===================================
1479 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1482 # Bind the device to igb_uio.
1483 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1486 # Recheck the status of the devices.
1487 ./usertools/dpdk-devbind.py --status
1488 Network devices using DPDK-compatible driver
1489 ============================================
1490 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1492 To attach a port created by virtual device, above steps are not needed.
1494 For example, to attach a port whose pci address is 0000:0a:00.0.
1496 .. code-block:: console
1498 testpmd> port attach 0000:0a:00.0
1499 Attaching a new port...
1500 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1501 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1502 EAL: PCI memory mapped at 0x7f83bfa00000
1503 EAL: PCI memory mapped at 0x7f83bfa80000
1504 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1505 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1506 Port 0 is attached. Now total ports is 1
1509 For example, to attach a port created by pcap PMD.
1511 .. code-block:: console
1513 testpmd> port attach net_pcap0
1514 Attaching a new port...
1515 PMD: Initializing pmd_pcap for net_pcap0
1516 PMD: Creating pcap-backed ethdev on numa socket 0
1517 Port 0 is attached. Now total ports is 1
1520 In this case, identifier is ``net_pcap0``.
1521 This identifier format is the same as ``--vdev`` format of DPDK applications.
1523 For example, to re-attach a bonded port which has been previously detached,
1524 the mode and slave parameters must be given.
1526 .. code-block:: console
1528 testpmd> port attach net_bond_0,mode=0,slave=1
1529 Attaching a new port...
1530 EAL: Initializing pmd_bond for net_bond_0
1531 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1532 Port 0 is attached. Now total ports is 1
1539 Detach a specific port::
1541 testpmd> port detach (port_id)
1543 Before detaching a port, the port should be stopped and closed.
1545 For example, to detach a pci device port 0.
1547 .. code-block:: console
1549 testpmd> port stop 0
1552 testpmd> port close 0
1556 testpmd> port detach 0
1558 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1559 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1560 EAL: PCI memory unmapped at 0x7f83bfa00000
1561 EAL: PCI memory unmapped at 0x7f83bfa80000
1565 For example, to detach a virtual device port 0.
1567 .. code-block:: console
1569 testpmd> port stop 0
1572 testpmd> port close 0
1576 testpmd> port detach 0
1578 PMD: Closing pcap ethdev on numa socket 0
1579 Port 'net_pcap0' is detached. Now total ports is 0
1582 To remove a pci device completely from the system, first detach the port from testpmd.
1583 Then the device should be moved under kernel management.
1584 Finally the device can be removed using kernel pci hotplug functionality.
1586 For example, to move a pci device under kernel management:
1588 .. code-block:: console
1590 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1592 ./usertools/dpdk-devbind.py --status
1594 Network devices using DPDK-compatible driver
1595 ============================================
1598 Network devices using kernel driver
1599 ===================================
1600 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1602 To remove a port created by a virtual device, above steps are not needed.
1607 Start all ports or a specific port::
1609 testpmd> port start (port_id|all)
1614 Stop all ports or a specific port::
1616 testpmd> port stop (port_id|all)
1621 Close all ports or a specific port::
1623 testpmd> port close (port_id|all)
1625 port start/stop queue
1626 ~~~~~~~~~~~~~~~~~~~~~
1628 Start/stop a rx/tx queue on a specific port::
1630 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1632 Only take effect when port is started.
1637 Set the speed and duplex mode for all ports or a specific port::
1639 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1640 duplex (half|full|auto)
1642 port config - queues/descriptors
1643 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1645 Set number of queues/descriptors for rxq, txq, rxd and txd::
1647 testpmd> port config all (rxq|txq|rxd|txd) (value)
1649 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1651 port config - max-pkt-len
1652 ~~~~~~~~~~~~~~~~~~~~~~~~~
1654 Set the maximum packet length::
1656 testpmd> port config all max-pkt-len (value)
1658 This is equivalent to the ``--max-pkt-len`` command-line option.
1660 port config - CRC Strip
1661 ~~~~~~~~~~~~~~~~~~~~~~~
1663 Set hardware CRC stripping on or off for all ports::
1665 testpmd> port config all crc-strip (on|off)
1667 CRC stripping is on by default.
1669 The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option.
1671 port config - scatter
1672 ~~~~~~~~~~~~~~~~~~~~~~~
1674 Set RX scatter mode on or off for all ports::
1676 testpmd> port config all scatter (on|off)
1678 RX scatter mode is off by default.
1680 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1682 port config - RX Checksum
1683 ~~~~~~~~~~~~~~~~~~~~~~~~~
1685 Set hardware RX checksum offload to on or off for all ports::
1687 testpmd> port config all rx-cksum (on|off)
1689 Checksum offload is off by default.
1691 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1696 Set hardware VLAN on or off for all ports::
1698 testpmd> port config all hw-vlan (on|off)
1700 Hardware VLAN is on by default.
1702 The ``off`` option is equivalent to the ``--disable-hw-vlan`` command-line option.
1704 port config - VLAN filter
1705 ~~~~~~~~~~~~~~~~~~~~~~~~~
1707 Set hardware VLAN filter on or off for all ports::
1709 testpmd> port config all hw-vlan-filter (on|off)
1711 Hardware VLAN filter is on by default.
1713 The ``off`` option is equivalent to the ``--disable-hw-vlan-filter`` command-line option.
1715 port config - VLAN strip
1716 ~~~~~~~~~~~~~~~~~~~~~~~~
1718 Set hardware VLAN strip on or off for all ports::
1720 testpmd> port config all hw-vlan-strip (on|off)
1722 Hardware VLAN strip is on by default.
1724 The ``off`` option is equivalent to the ``--disable-hw-vlan-strip`` command-line option.
1726 port config - VLAN extend
1727 ~~~~~~~~~~~~~~~~~~~~~~~~~
1729 Set hardware VLAN extend on or off for all ports::
1731 testpmd> port config all hw-vlan-extend (on|off)
1733 Hardware VLAN extend is off by default.
1735 The ``off`` option is equivalent to the ``--disable-hw-vlan-extend`` command-line option.
1737 port config - Drop Packets
1738 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1740 Set packet drop for packets with no descriptors on or off for all ports::
1742 testpmd> port config all drop-en (on|off)
1744 Packet dropping for packets with no descriptors is off by default.
1746 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1751 Set the RSS (Receive Side Scaling) mode on or off::
1753 testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1755 RSS is on by default.
1757 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1759 port config - RSS Reta
1760 ~~~~~~~~~~~~~~~~~~~~~~
1762 Set the RSS (Receive Side Scaling) redirection table::
1764 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1769 Set the DCB mode for an individual port::
1771 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1773 The traffic class should be 4 or 8.
1778 Set the number of packets per burst::
1780 testpmd> port config all burst (value)
1782 This is equivalent to the ``--burst`` command-line option.
1784 port config - Threshold
1785 ~~~~~~~~~~~~~~~~~~~~~~~
1787 Set thresholds for TX/RX queues::
1789 testpmd> port config all (threshold) (value)
1791 Where the threshold type can be:
1793 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1795 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1797 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1799 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1801 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1803 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1805 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1807 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1809 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1811 These threshold options are also available from the command-line.
1816 Set the value of ether-type for E-tag::
1818 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1820 Enable/disable the E-tag support::
1822 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1824 port config pctype mapping
1825 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1827 Reset pctype mapping table::
1829 testpmd> port config (port_id) pctype mapping reset
1831 Update hardware defined pctype to software defined flow type mapping table::
1833 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id)
1837 * ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table.
1839 * ``flow_type_id``: software flow type id as the index of the pctype mapping table.
1842 Link Bonding Functions
1843 ----------------------
1845 The Link Bonding functions make it possible to dynamically create and
1846 manage link bonding devices from within testpmd interactive prompt.
1848 create bonded device
1849 ~~~~~~~~~~~~~~~~~~~~
1851 Create a new bonding device::
1853 testpmd> create bonded device (mode) (socket)
1855 For example, to create a bonded device in mode 1 on socket 0::
1857 testpmd> create bonded 1 0
1858 created new bonded device (port X)
1863 Adds Ethernet device to a Link Bonding device::
1865 testpmd> add bonding slave (slave id) (port id)
1867 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
1869 testpmd> add bonding slave 6 10
1872 remove bonding slave
1873 ~~~~~~~~~~~~~~~~~~~~
1875 Removes an Ethernet slave device from a Link Bonding device::
1877 testpmd> remove bonding slave (slave id) (port id)
1879 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
1881 testpmd> remove bonding slave 6 10
1886 Set the Link Bonding mode of a Link Bonding device::
1888 testpmd> set bonding mode (value) (port id)
1890 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
1892 testpmd> set bonding mode 3 10
1897 Set an Ethernet slave device as the primary device on a Link Bonding device::
1899 testpmd> set bonding primary (slave id) (port id)
1901 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
1903 testpmd> set bonding primary 6 10
1908 Set the MAC address of a Link Bonding device::
1910 testpmd> set bonding mac (port id) (mac)
1912 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
1914 testpmd> set bonding mac 10 00:00:00:00:00:01
1916 set bonding xmit_balance_policy
1917 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1919 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
1921 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
1923 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
1925 testpmd> set bonding xmit_balance_policy 10 l34
1928 set bonding mon_period
1929 ~~~~~~~~~~~~~~~~~~~~~~
1931 Set the link status monitoring polling period in milliseconds for a bonding device.
1933 This adds support for PMD slave devices which do not support link status interrupts.
1934 When the mon_period is set to a value greater than 0 then all PMD's which do not support
1935 link status ISR will be queried every polling interval to check if their link status has changed::
1937 testpmd> set bonding mon_period (port_id) (value)
1939 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
1941 testpmd> set bonding mon_period 5 150
1944 set bonding lacp dedicated_queue
1945 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1947 Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic
1948 when in mode 4 (link-aggregration-802.3ad)::
1950 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable)
1953 set bonding agg_mode
1954 ~~~~~~~~~~~~~~~~~~~~
1956 Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad)::
1958 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable)
1964 Show the current configuration of a Link Bonding device::
1966 testpmd> show bonding config (port id)
1969 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
1970 in balance mode with a transmission policy of layer 2+3::
1972 testpmd> show bonding config 9
1974 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
1976 Active Slaves (3): [1 3 4]
1983 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
1984 This is mainly useful for debugging purposes.
1985 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
1986 and fields that can be accessed.
1991 Display the value of a port register::
1993 testpmd> read reg (port_id) (address)
1995 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
1997 testpmd> read reg 0 0xEE00
1998 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
2003 Display a port register bit field::
2005 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
2007 For example, reading the lowest two bits from the register in the example above::
2009 testpmd> read regfield 0 0xEE00 0 1
2010 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
2015 Display a single port register bit::
2017 testpmd> read regbit (port_id) (address) (bit_x)
2019 For example, reading the lowest bit from the register in the example above::
2021 testpmd> read regbit 0 0xEE00 0
2022 port 0 PCI register at offset 0xEE00: bit 0=1
2027 Set the value of a port register::
2029 testpmd> write reg (port_id) (address) (value)
2031 For example, to clear a register::
2033 testpmd> write reg 0 0xEE00 0x0
2034 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
2039 Set bit field of a port register::
2041 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
2043 For example, writing to the register cleared in the example above::
2045 testpmd> write regfield 0 0xEE00 0 1 2
2046 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
2051 Set single bit value of a port register::
2053 testpmd> write regbit (port_id) (address) (bit_x) (value)
2055 For example, to set the high bit in the register from the example above::
2057 testpmd> write regbit 0 0xEE00 31 1
2058 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
2060 Traffic Metering and Policing
2061 -----------------------------
2063 The following section shows functions for configuring traffic metering and
2064 policing on the ethernet device through the use of generic ethdev API.
2066 show port traffic management capability
2067 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2069 Show traffic metering and policing capability of the port::
2071 testpmd> show port meter cap (port_id)
2073 add port meter profile (srTCM rfc2967)
2074 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2076 Add meter profile (srTCM rfc2697) to the ethernet device::
2078 testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \
2083 * ``profile_id``: ID for the meter profile.
2084 * ``cir``: Committed Information Rate (CIR) (bytes/second).
2085 * ``cbs``: Committed Burst Size (CBS) (bytes).
2086 * ``ebs``: Excess Burst Size (EBS) (bytes).
2088 add port meter profile (trTCM rfc2968)
2089 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2091 Add meter profile (srTCM rfc2698) to the ethernet device::
2093 testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \
2094 (cir) (pir) (cbs) (pbs)
2098 * ``profile_id``: ID for the meter profile.
2099 * ``cir``: Committed information rate (bytes/second).
2100 * ``pir``: Peak information rate (bytes/second).
2101 * ``cbs``: Committed burst size (bytes).
2102 * ``pbs``: Peak burst size (bytes).
2104 add port meter profile (trTCM rfc4115)
2105 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2107 Add meter profile (trTCM rfc4115) to the ethernet device::
2109 testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \
2110 (cir) (eir) (cbs) (ebs)
2114 * ``profile_id``: ID for the meter profile.
2115 * ``cir``: Committed information rate (bytes/second).
2116 * ``eir``: Excess information rate (bytes/second).
2117 * ``cbs``: Committed burst size (bytes).
2118 * ``ebs``: Excess burst size (bytes).
2120 delete port meter profile
2121 ~~~~~~~~~~~~~~~~~~~~~~~~~
2123 Delete meter profile from the ethernet device::
2125 testpmd> del port meter profile (port_id) (profile_id)
2130 Create new meter object for the ethernet device::
2132 testpmd> create port meter (port_id) (mtr_id) (profile_id) \
2133 (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \
2134 (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\
2139 * ``mtr_id``: meter object ID.
2140 * ``profile_id``: ID for the meter profile.
2141 * ``meter_enable``: When this parameter has a non-zero value, the meter object
2142 gets enabled at the time of creation, otherwise remains disabled.
2143 * ``g_action``: Policer action for the packet with green color.
2144 * ``y_action``: Policer action for the packet with yellow color.
2145 * ``r_action``: Policer action for the packet with red color.
2146 * ``stats_mask``: Mask of statistics counter types to be enabled for the
2148 * ``shared``: When this parameter has a non-zero value, the meter object is
2149 shared by multiple flows. Otherwise, meter object is used by single flow.
2150 * ``use_pre_meter_color``: When this parameter has a non-zero value, the
2151 input color for the current meter object is determined by the latest meter
2152 object in the same flow. Otherwise, the current meter object uses the
2153 *dscp_table* to determine the input color.
2154 * ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input
2155 color, 0 <= x <= 63.
2160 Enable meter for the ethernet device::
2162 testpmd> enable port meter (port_id) (mtr_id)
2167 Disable meter for the ethernet device::
2169 testpmd> disable port meter (port_id) (mtr_id)
2174 Delete meter for the ethernet device::
2176 testpmd> del port meter (port_id) (mtr_id)
2178 Set port meter profile
2179 ~~~~~~~~~~~~~~~~~~~~~~
2181 Set meter profile for the ethernet device::
2183 testpmd> set port meter profile (port_id) (mtr_id) (profile_id)
2185 set port meter dscp table
2186 ~~~~~~~~~~~~~~~~~~~~~~~~~
2188 Set meter dscp table for the ethernet device::
2190 testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \
2191 (dscp_tbl_entry1)...(dscp_tbl_entry63)]
2193 set port meter policer action
2194 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2196 Set meter policer action for the ethernet device::
2198 testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \
2199 (action0) [(action1) (action1)]
2203 * ``action_mask``: Bit mask indicating which policer actions need to be
2204 updated. One or more policer actions can be updated in a single function
2205 invocation. To update the policer action associated with color C, bit
2206 (1 << C) needs to be set in *action_mask* and element at position C
2207 in the *actions* array needs to be valid.
2208 * ``actionx``: Policer action for the color x,
2209 RTE_MTR_GREEN <= x < RTE_MTR_COLORS
2211 set port meter stats mask
2212 ~~~~~~~~~~~~~~~~~~~~~~~~~
2214 Set meter stats mask for the ethernet device::
2216 testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask)
2220 * ``stats_mask``: Bit mask indicating statistics counter types to be enabled.
2222 show port meter stats
2223 ~~~~~~~~~~~~~~~~~~~~~
2225 Show meter stats of the ethernet device::
2227 testpmd> show port meter stats (port_id) (mtr_id) (clear)
2231 * ``clear``: Flag that indicates whether the statistics counters should
2232 be cleared (i.e. set to zero) immediately after they have been read or not.
2237 The following section shows functions for configuring traffic management on
2238 on the ethernet device through the use of generic TM API.
2240 show port traffic management capability
2241 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2243 Show traffic management capability of the port::
2245 testpmd> show port tm cap (port_id)
2247 show port traffic management capability (hierarchy level)
2248 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2250 Show traffic management hierarchy level capability of the port::
2252 testpmd> show port tm cap (port_id) (level_id)
2254 show port traffic management capability (hierarchy node level)
2255 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2257 Show the traffic management hierarchy node capability of the port::
2259 testpmd> show port tm cap (port_id) (node_id)
2261 show port traffic management hierarchy node type
2262 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2264 Show the port traffic management hierarchy node type::
2266 testpmd> show port tm node type (port_id) (node_id)
2268 show port traffic management hierarchy node stats
2269 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2271 Show the port traffic management hierarchy node statistics::
2273 testpmd> show port tm node stats (port_id) (node_id) (clear)
2277 * ``clear``: When this parameter has a non-zero value, the statistics counters
2278 are cleared (i.e. set to zero) immediately after they have been read,
2279 otherwise the statistics counters are left untouched.
2281 Add port traffic management private shaper profile
2282 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2284 Add the port traffic management private shaper profile::
2286 testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \
2287 (tb_rate) (tb_size) (packet_length_adjust)
2291 * ``shaper_profile id``: Shaper profile ID for the new profile.
2292 * ``tb_rate``: Token bucket rate (bytes per second).
2293 * ``tb_size``: Token bucket size (bytes).
2294 * ``packet_length_adjust``: The value (bytes) to be added to the length of
2295 each packet for the purpose of shaping. This parameter value can be used to
2296 correct the packet length with the framing overhead bytes that are consumed
2299 Delete port traffic management private shaper profile
2300 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2302 Delete the port traffic management private shaper::
2304 testpmd> del port tm node shaper profile (port_id) (shaper_profile_id)
2308 * ``shaper_profile id``: Shaper profile ID that needs to be deleted.
2310 Add port traffic management shared shaper
2311 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2313 Create the port traffic management shared shaper::
2315 testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \
2320 * ``shared_shaper_id``: Shared shaper ID to be created.
2321 * ``shaper_profile id``: Shaper profile ID for shared shaper.
2323 Set port traffic management shared shaper
2324 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2326 Update the port traffic management shared shaper::
2328 testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \
2333 * ``shared_shaper_id``: Shared shaper ID to be update.
2334 * ``shaper_profile id``: Shaper profile ID for shared shaper.
2336 Delete port traffic management shared shaper
2337 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2339 Delete the port traffic management shared shaper::
2341 testpmd> del port tm node shared shaper (port_id) (shared_shaper_id)
2345 * ``shared_shaper_id``: Shared shaper ID to be deleted.
2347 Set port traffic management hiearchy node private shaper
2348 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2350 set the port traffic management hierarchy node private shaper::
2352 testpmd> set port tm node shaper profile (port_id) (node_id) \
2357 * ``shaper_profile id``: Private shaper profile ID to be enabled on the
2360 Add port traffic management WRED profile
2361 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2363 Create a new WRED profile::
2365 testpmd> add port tm node wred profile (port_id) (wred_profile_id) \
2366 (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \
2367 (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \
2368 (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r)
2372 * ``wred_profile id``: Identifier for the newly create WRED profile
2373 * ``color_g``: Packet color (green)
2374 * ``min_th_g``: Minimum queue threshold for packet with green color
2375 * ``max_th_g``: Minimum queue threshold for packet with green color
2376 * ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp)
2377 * ``wq_log2_g``: Negated log2 of queue weight (wq)
2378 * ``color_y``: Packet color (yellow)
2379 * ``min_th_y``: Minimum queue threshold for packet with yellow color
2380 * ``max_th_y``: Minimum queue threshold for packet with yellow color
2381 * ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp)
2382 * ``wq_log2_y``: Negated log2 of queue weight (wq)
2383 * ``color_r``: Packet color (red)
2384 * ``min_th_r``: Minimum queue threshold for packet with yellow color
2385 * ``max_th_r``: Minimum queue threshold for packet with yellow color
2386 * ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp)
2387 * ``wq_log2_r``: Negated log2 of queue weight (wq)
2389 Delete port traffic management WRED profile
2390 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2392 Delete the WRED profile::
2394 testpmd> del port tm node wred profile (port_id) (wred_profile_id)
2396 Add port traffic management hierarchy nonleaf node
2397 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2399 Add nonleaf node to port traffic management hiearchy::
2401 testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \
2402 (priority) (weight) (level_id) (shaper_profile_id) \
2403 (n_sp_priorities) (stats_mask) (n_shared_shapers) \
2404 [(shared_shaper_0) (shared_shaper_1) ...] \
2408 * ``parent_node_id``: Node ID of the parent.
2409 * ``priority``: Node priority (highest node priority is zero). This is used by
2410 the SP algorithm running on the parent node for scheduling this node.
2411 * ``weight``: Node weight (lowest weight is one). The node weight is relative
2412 to the weight sum of all siblings that have the same priority. It is used by
2413 the WFQ algorithm running on the parent node for scheduling this node.
2414 * ``level_id``: Hiearchy level of the node.
2415 * ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2417 * ``n_sp_priorities``: Number of strict priorities.
2418 * ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2419 * ``n_shared_shapers``: Number of shared shapers.
2420 * ``shared_shaper_id``: Shared shaper id.
2422 Add port traffic management hierarchy leaf node
2423 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2425 Add leaf node to port traffic management hiearchy::
2427 testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \
2428 (priority) (weight) (level_id) (shaper_profile_id) \
2429 (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \
2430 [(shared_shaper_id) (shared_shaper_id) ...] \
2434 * ``parent_node_id``: Node ID of the parent.
2435 * ``priority``: Node priority (highest node priority is zero). This is used by
2436 the SP algorithm running on the parent node for scheduling this node.
2437 * ``weight``: Node weight (lowest weight is one). The node weight is relative
2438 to the weight sum of all siblings that have the same priority. It is used by
2439 the WFQ algorithm running on the parent node for scheduling this node.
2440 * ``level_id``: Hiearchy level of the node.
2441 * ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2443 * ``cman_mode``: Congestion management mode to be enabled for this node.
2444 * ``wred_profile_id``: WRED profile id to be enabled for this node.
2445 * ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2446 * ``n_shared_shapers``: Number of shared shapers.
2447 * ``shared_shaper_id``: Shared shaper id.
2449 Delete port traffic management hierarchy node
2450 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2452 Delete node from port traffic management hiearchy::
2454 testpmd> del port tm node (port_id) (node_id)
2456 Update port traffic management hierarchy parent node
2457 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2459 Update port traffic management hierarchy parent node::
2461 testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \
2464 This function can only be called after the hierarchy commit invocation. Its
2465 success depends on the port support for this operation, as advertised through
2466 the port capability set. This function is valid for all nodes of the traffic
2467 management hierarchy except root node.
2469 Commit port traffic management hierarchy
2470 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2472 Commit the traffic management hierarchy on the port::
2474 testpmd> port tm hierarchy commit (port_id) (clean_on_fail)
2478 * ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function
2479 call failure. On the other hand, hierarchy is preserved when this parameter
2482 Set port traffic management default hierarchy (tm forwarding mode)
2483 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2485 set the traffic management default hierarchy on the port::
2487 testpmd> set port tm hierarchy default (port_id)
2492 This section details the available filter functions that are available.
2494 Note these functions interface the deprecated legacy filtering framework,
2495 superseded by *rte_flow*. See `Flow rules management`_.
2498 ~~~~~~~~~~~~~~~~~~~~
2500 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
2502 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
2503 ethertype (ether_type) (drop|fwd) queue (queue_id)
2505 The available information parameters are:
2507 * ``port_id``: The port which the Ethertype filter assigned on.
2509 * ``mac_addr``: Compare destination mac address.
2511 * ``mac_ignr``: Ignore destination mac address match.
2513 * ``mac_address``: Destination mac address to match.
2515 * ``ether_type``: The EtherType value want to match,
2516 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
2518 * ``queue_id``: The receive queue associated with this EtherType filter.
2519 It is meaningless when deleting or dropping.
2521 Example, to add/remove an ethertype filter rule::
2523 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
2524 ethertype 0x0806 fwd queue 3
2526 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
2527 ethertype 0x0806 fwd queue 3
2532 Add or delete a 2-tuple filter,
2533 which identifies packets by specific protocol and destination TCP/UDP port
2534 and forwards packets into one of the receive queues::
2536 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
2537 protocol (protocol_value) mask (mask_value) \
2538 tcp_flags (tcp_flags_value) priority (prio_value) \
2541 The available information parameters are:
2543 * ``port_id``: The port which the 2-tuple filter assigned on.
2545 * ``dst_port_value``: Destination port in L4.
2547 * ``protocol_value``: IP L4 protocol.
2549 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
2551 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
2553 * ``prio_value``: Priority of this filter.
2555 * ``queue_id``: The receive queue associated with this 2-tuple filter.
2557 Example, to add/remove an 2tuple filter rule::
2559 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
2560 tcp_flags 0x02 priority 3 queue 3
2562 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
2563 tcp_flags 0x02 priority 3 queue 3
2568 Add or delete a 5-tuple filter,
2569 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
2570 and routes packets into one of the receive queues::
2572 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
2573 (src_address) dst_port (dst_port_value) \
2574 src_port (src_port_value) protocol (protocol_value) \
2575 mask (mask_value) tcp_flags (tcp_flags_value) \
2576 priority (prio_value) queue (queue_id)
2578 The available information parameters are:
2580 * ``port_id``: The port which the 5-tuple filter assigned on.
2582 * ``dst_address``: Destination IP address.
2584 * ``src_address``: Source IP address.
2586 * ``dst_port_value``: TCP/UDP destination port.
2588 * ``src_port_value``: TCP/UDP source port.
2590 * ``protocol_value``: L4 protocol.
2592 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
2594 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
2596 * ``prio_value``: The priority of this filter.
2598 * ``queue_id``: The receive queue associated with this 5-tuple filter.
2600 Example, to add/remove an 5tuple filter rule::
2602 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2603 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2604 flags 0x0 priority 3 queue 3
2606 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2607 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2608 flags 0x0 priority 3 queue 3
2613 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
2615 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
2617 The available information parameters are:
2619 * ``port_id``: The port which the SYN filter assigned on.
2621 * ``high``: This SYN filter has higher priority than other filters.
2623 * ``low``: This SYN filter has lower priority than other filters.
2625 * ``queue_id``: The receive queue associated with this SYN filter
2629 testpmd> syn_filter 0 add priority high queue 3
2634 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
2635 and routed into one of the receive queues::
2637 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
2638 mask (mask_value) priority (prio_value) queue (queue_id)
2640 The available information parameters are:
2642 * ``port_id``: The port which the Flex filter is assigned on.
2644 * ``len_value``: Filter length in bytes, no greater than 128.
2646 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
2648 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
2650 * ``prio_value``: The priority of this filter.
2652 * ``queue_id``: The receive queue associated with this Flex filter.
2656 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
2657 mask 000C priority 3 queue 3
2659 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
2660 mask 000C priority 3 queue 3
2663 .. _testpmd_flow_director:
2665 flow_director_filter
2666 ~~~~~~~~~~~~~~~~~~~~
2668 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
2670 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
2671 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
2673 * Perfect match filters.
2674 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2675 The masked fields are for IP flow.
2677 * Signature filters.
2678 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
2680 * Perfect-mac-vlan match filters.
2681 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2682 The masked fields are for MAC VLAN flow.
2684 * Perfect-tunnel match filters.
2685 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2686 The masked fields are for tunnel flow.
2688 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
2689 per flow type and the flexible payload.
2691 The Flow Director can also mask out parts of all of these fields so that filters
2692 are only applied to certain fields or parts of the fields.
2694 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
2696 # Commands to add flow director filters of different flow types::
2698 flow_director_filter (port_id) mode IP (add|del|update) \
2699 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
2700 src (src_ip_address) dst (dst_ip_address) \
2701 tos (tos_value) proto (proto_value) ttl (ttl_value) \
2702 vlan (vlan_value) flexbytes (flexbytes_value) \
2703 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
2706 flow_director_filter (port_id) mode IP (add|del|update) \
2707 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
2708 src (src_ip_address) (src_port) \
2709 dst (dst_ip_address) (dst_port) \
2710 tos (tos_value) ttl (ttl_value) \
2711 vlan (vlan_value) flexbytes (flexbytes_value) \
2712 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
2715 flow_director_filter (port_id) mode IP (add|del|update) \
2716 flow (ipv4-sctp|ipv6-sctp) \
2717 src (src_ip_address) (src_port) \
2718 dst (dst_ip_address) (dst_port) \
2719 tos (tos_value) ttl (ttl_value) \
2720 tag (verification_tag) vlan (vlan_value) \
2721 flexbytes (flexbytes_value) (drop|fwd) \
2722 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
2724 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
2725 ether (ethertype) flexbytes (flexbytes_value) \
2726 (drop|fwd) pf|vf(vf_id) queue (queue_id)
2729 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
2730 mac (mac_address) vlan (vlan_value) \
2731 flexbytes (flexbytes_value) (drop|fwd) \
2732 queue (queue_id) fd_id (fd_id_value)
2734 flow_director_filter (port_id) mode Tunnel (add|del|update) \
2735 mac (mac_address) vlan (vlan_value) \
2736 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
2737 flexbytes (flexbytes_value) (drop|fwd) \
2738 queue (queue_id) fd_id (fd_id_value)
2740 For example, to add an ipv4-udp flow type filter::
2742 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
2743 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
2744 fwd pf queue 1 fd_id 1
2746 For example, add an ipv4-other flow type filter::
2748 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
2749 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
2750 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
2755 Flush all flow director filters on a device::
2757 testpmd> flush_flow_director (port_id)
2759 Example, to flush all flow director filter on port 0::
2761 testpmd> flush_flow_director 0
2766 Set flow director's input masks::
2768 flow_director_mask (port_id) mode IP vlan (vlan_value) \
2769 src_mask (ipv4_src) (ipv6_src) (src_port) \
2770 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2772 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2774 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2775 mac (mac_value) tunnel-type (tunnel_type_value) \
2776 tunnel-id (tunnel_id_value)
2778 Example, to set flow director mask on port 0::
2780 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2781 src_mask 255.255.255.255 \
2782 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2783 dst_mask 255.255.255.255 \
2784 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2786 flow_director_flex_mask
2787 ~~~~~~~~~~~~~~~~~~~~~~~
2789 set masks of flow director's flexible payload based on certain flow type::
2791 testpmd> flow_director_flex_mask (port_id) \
2792 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2793 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2794 l2_payload|all) (mask)
2796 Example, to set flow director's flex mask for all flow type on port 0::
2798 testpmd> flow_director_flex_mask 0 flow all \
2799 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2802 flow_director_flex_payload
2803 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2805 Configure flexible payload selection::
2807 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2809 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2811 testpmd> flow_director_flex_payload 0 l4 \
2812 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2814 get_sym_hash_ena_per_port
2815 ~~~~~~~~~~~~~~~~~~~~~~~~~
2817 Get symmetric hash enable configuration per port::
2819 get_sym_hash_ena_per_port (port_id)
2821 For example, to get symmetric hash enable configuration of port 1::
2823 testpmd> get_sym_hash_ena_per_port 1
2825 set_sym_hash_ena_per_port
2826 ~~~~~~~~~~~~~~~~~~~~~~~~~
2828 Set symmetric hash enable configuration per port to enable or disable::
2830 set_sym_hash_ena_per_port (port_id) (enable|disable)
2832 For example, to set symmetric hash enable configuration of port 1 to enable::
2834 testpmd> set_sym_hash_ena_per_port 1 enable
2836 get_hash_global_config
2837 ~~~~~~~~~~~~~~~~~~~~~~
2839 Get the global configurations of hash filters::
2841 get_hash_global_config (port_id)
2843 For example, to get the global configurations of hash filters of port 1::
2845 testpmd> get_hash_global_config 1
2847 set_hash_global_config
2848 ~~~~~~~~~~~~~~~~~~~~~~
2850 Set the global configurations of hash filters::
2852 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2853 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2854 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload) \
2857 For example, to enable simple_xor for flow type of ipv6 on port 2::
2859 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2864 Set the input set for hash::
2866 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2867 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2868 l2_payload) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2869 ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2870 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2871 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2872 fld-8th|none) (select|add)
2874 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2876 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2881 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2882 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2884 Set the input set for flow director::
2886 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2887 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2888 l2_payload) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2889 ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2890 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2891 sctp-dst-port|sctp-veri-tag|none) (select|add)
2893 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2895 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2900 Set different GRE key length for input set::
2902 global_config (port_id) gre-key-len (number in bytes)
2904 For example to set GRE key length for input set to 4 bytes on port 0::
2906 testpmd> global_config 0 gre-key-len 4
2909 .. _testpmd_rte_flow:
2911 Flow rules management
2912 ---------------------
2914 Control of the generic flow API (*rte_flow*) is fully exposed through the
2915 ``flow`` command (validation, creation, destruction, queries and operation
2918 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2919 features simultaneously may cause undefined side-effects and is therefore
2925 Because the ``flow`` command uses dynamic tokens to handle the large number
2926 of possible flow rules combinations, its behavior differs slightly from
2927 other commands, in particular:
2929 - Pressing *?* or the *<tab>* key displays contextual help for the current
2930 token, not that of the entire command.
2932 - Optional and repeated parameters are supported (provided they are listed
2933 in the contextual help).
2935 The first parameter stands for the operation mode. Possible operations and
2936 their general syntax are described below. They are covered in detail in the
2939 - Check whether a flow rule can be created::
2941 flow validate {port_id}
2942 [group {group_id}] [priority {level}] [ingress] [egress]
2943 pattern {item} [/ {item} [...]] / end
2944 actions {action} [/ {action} [...]] / end
2946 - Create a flow rule::
2948 flow create {port_id}
2949 [group {group_id}] [priority {level}] [ingress] [egress]
2950 pattern {item} [/ {item} [...]] / end
2951 actions {action} [/ {action} [...]] / end
2953 - Destroy specific flow rules::
2955 flow destroy {port_id} rule {rule_id} [...]
2957 - Destroy all flow rules::
2959 flow flush {port_id}
2961 - Query an existing flow rule::
2963 flow query {port_id} {rule_id} {action}
2965 - List existing flow rules sorted by priority, filtered by group
2968 flow list {port_id} [group {group_id}] [...]
2970 - Restrict ingress traffic to the defined flow rules::
2972 flow isolate {port_id} {boolean}
2974 Validating flow rules
2975 ~~~~~~~~~~~~~~~~~~~~~
2977 ``flow validate`` reports whether a flow rule would be accepted by the
2978 underlying device in its current state but stops short of creating it. It is
2979 bound to ``rte_flow_validate()``::
2981 flow validate {port_id}
2982 [group {group_id}] [priority {level}] [ingress] [egress]
2983 pattern {item} [/ {item} [...]] / end
2984 actions {action} [/ {action} [...]] / end
2986 If successful, it will show::
2990 Otherwise it will show an error message of the form::
2992 Caught error type [...] ([...]): [...]
2994 This command uses the same parameters as ``flow create``, their format is
2995 described in `Creating flow rules`_.
2997 Check whether redirecting any Ethernet packet received on port 0 to RX queue
2998 index 6 is supported::
3000 testpmd> flow validate 0 ingress pattern eth / end
3001 actions queue index 6 / end
3005 Port 0 does not support TCPv6 rules::
3007 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
3009 Caught error type 9 (specific pattern item): Invalid argument
3015 ``flow create`` validates and creates the specified flow rule. It is bound
3016 to ``rte_flow_create()``::
3018 flow create {port_id}
3019 [group {group_id}] [priority {level}] [ingress] [egress]
3020 pattern {item} [/ {item} [...]] / end
3021 actions {action} [/ {action} [...]] / end
3023 If successful, it will return a flow rule ID usable with other commands::
3025 Flow rule #[...] created
3027 Otherwise it will show an error message of the form::
3029 Caught error type [...] ([...]): [...]
3031 Parameters describe in the following order:
3033 - Attributes (*group*, *priority*, *ingress*, *egress* tokens).
3034 - A matching pattern, starting with the *pattern* token and terminated by an
3036 - Actions, starting with the *actions* token and terminated by an *end*
3039 These translate directly to *rte_flow* objects provided as-is to the
3040 underlying functions.
3042 The shortest valid definition only comprises mandatory tokens::
3044 testpmd> flow create 0 pattern end actions end
3046 Note that PMDs may refuse rules that essentially do nothing such as this
3049 **All unspecified object values are automatically initialized to 0.**
3054 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
3055 specified before the ``pattern`` token.
3057 - ``group {group id}``: priority group.
3058 - ``priority {level}``: priority level within group.
3059 - ``ingress``: rule applies to ingress traffic.
3060 - ``egress``: rule applies to egress traffic.
3062 Each instance of an attribute specified several times overrides the previous
3063 value as shown below (group 4 is used)::
3065 testpmd> flow create 0 group 42 group 24 group 4 [...]
3067 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
3069 While not specifying a direction is an error, some rules may allow both
3072 Most rules affect RX therefore contain the ``ingress`` token::
3074 testpmd> flow create 0 ingress pattern [...]
3079 A matching pattern starts after the ``pattern`` token. It is made of pattern
3080 items and is terminated by a mandatory ``end`` item.
3082 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
3083 rte_flow_item_type``).
3085 The ``/`` token is used as a separator between pattern items as shown
3088 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
3090 Note that protocol items like these must be stacked from lowest to highest
3091 layer to make sense. For instance, the following rule is either invalid or
3092 unlikely to match any packet::
3094 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
3096 More information on these restrictions can be found in the *rte_flow*
3099 Several items support additional specification structures, for example
3100 ``ipv4`` allows specifying source and destination addresses as follows::
3102 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
3103 dst is 10.2.0.0 / end [...]
3105 This rule matches all IPv4 traffic with the specified properties.
3107 In this example, ``src`` and ``dst`` are field names of the underlying
3108 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
3109 in a similar fashion.
3111 The ``is`` token means that the subsequent value must be matched exactly,
3112 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
3113 accordingly. Possible assignment tokens are:
3115 - ``is``: match value perfectly (with full bit-mask).
3116 - ``spec``: match value according to configured bit-mask.
3117 - ``last``: specify upper bound to establish a range.
3118 - ``mask``: specify bit-mask with relevant bits set to one.
3119 - ``prefix``: generate bit-mask from a prefix length.
3121 These yield identical results::
3123 ipv4 src is 10.1.1.1
3127 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
3131 ipv4 src spec 10.1.1.1 src prefix 32
3135 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
3139 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
3141 Inclusive ranges can be defined with ``last``::
3143 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
3145 Note that ``mask`` affects both ``spec`` and ``last``::
3147 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
3148 # matches 10.1.0.0 to 10.2.255.255
3150 Properties can be modified multiple times::
3152 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
3156 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
3161 This section lists supported pattern items and their attributes, if any.
3163 - ``end``: end list of pattern items.
3165 - ``void``: no-op pattern item.
3167 - ``invert``: perform actions when pattern does not match.
3169 - ``any``: match any protocol for the current layer.
3171 - ``num {unsigned}``: number of layers covered.
3173 - ``pf``: match packets addressed to the physical function.
3175 - ``vf``: match packets addressed to a virtual function ID.
3177 - ``id {unsigned}``: destination VF ID.
3179 - ``port``: device-specific physical port index to use.
3181 - ``index {unsigned}``: physical port index.
3183 - ``raw``: match an arbitrary byte string.
3185 - ``relative {boolean}``: look for pattern after the previous item.
3186 - ``search {boolean}``: search pattern from offset (see also limit).
3187 - ``offset {integer}``: absolute or relative offset for pattern.
3188 - ``limit {unsigned}``: search area limit for start of pattern.
3189 - ``pattern {string}``: byte string to look for.
3191 - ``eth``: match Ethernet header.
3193 - ``dst {MAC-48}``: destination MAC.
3194 - ``src {MAC-48}``: source MAC.
3195 - ``type {unsigned}``: EtherType.
3197 - ``vlan``: match 802.1Q/ad VLAN tag.
3199 - ``tpid {unsigned}``: tag protocol identifier.
3200 - ``tci {unsigned}``: tag control information.
3201 - ``pcp {unsigned}``: priority code point.
3202 - ``dei {unsigned}``: drop eligible indicator.
3203 - ``vid {unsigned}``: VLAN identifier.
3205 - ``ipv4``: match IPv4 header.
3207 - ``tos {unsigned}``: type of service.
3208 - ``ttl {unsigned}``: time to live.
3209 - ``proto {unsigned}``: next protocol ID.
3210 - ``src {ipv4 address}``: source address.
3211 - ``dst {ipv4 address}``: destination address.
3213 - ``ipv6``: match IPv6 header.
3215 - ``tc {unsigned}``: traffic class.
3216 - ``flow {unsigned}``: flow label.
3217 - ``proto {unsigned}``: protocol (next header).
3218 - ``hop {unsigned}``: hop limit.
3219 - ``src {ipv6 address}``: source address.
3220 - ``dst {ipv6 address}``: destination address.
3222 - ``icmp``: match ICMP header.
3224 - ``type {unsigned}``: ICMP packet type.
3225 - ``code {unsigned}``: ICMP packet code.
3227 - ``udp``: match UDP header.
3229 - ``src {unsigned}``: UDP source port.
3230 - ``dst {unsigned}``: UDP destination port.
3232 - ``tcp``: match TCP header.
3234 - ``src {unsigned}``: TCP source port.
3235 - ``dst {unsigned}``: TCP destination port.
3237 - ``sctp``: match SCTP header.
3239 - ``src {unsigned}``: SCTP source port.
3240 - ``dst {unsigned}``: SCTP destination port.
3241 - ``tag {unsigned}``: validation tag.
3242 - ``cksum {unsigned}``: checksum.
3244 - ``vxlan``: match VXLAN header.
3246 - ``vni {unsigned}``: VXLAN identifier.
3248 - ``e_tag``: match IEEE 802.1BR E-Tag header.
3250 - ``grp_ecid_b {unsigned}``: GRP and E-CID base.
3252 - ``nvgre``: match NVGRE header.
3254 - ``tni {unsigned}``: virtual subnet ID.
3256 - ``mpls``: match MPLS header.
3258 - ``label {unsigned}``: MPLS label.
3260 - ``gre``: match GRE header.
3262 - ``protocol {unsigned}``: protocol type.
3264 - ``fuzzy``: fuzzy pattern match, expect faster than default.
3266 - ``thresh {unsigned}``: accuracy threshold.
3268 - ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header.
3270 - ``teid {unsigned}``: tunnel endpoint identifier.
3275 A list of actions starts after the ``actions`` token in the same fashion as
3276 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
3277 terminated by a mandatory ``end`` action.
3279 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
3280 rte_flow_action_type``).
3282 Dropping all incoming UDPv4 packets can be expressed as follows::
3284 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3287 Several actions have configurable properties which must be specified when
3288 there is no valid default value. For example, ``queue`` requires a target
3291 This rule redirects incoming UDPv4 traffic to queue index 6::
3293 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3294 actions queue index 6 / end
3296 While this one could be rejected by PMDs (unspecified queue index)::
3298 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3301 As defined by *rte_flow*, the list is not ordered, all actions of a given
3302 rule are performed simultaneously. These are equivalent::
3304 queue index 6 / void / mark id 42 / end
3308 void / mark id 42 / queue index 6 / end
3310 All actions in a list should have different types, otherwise only the last
3311 action of a given type is taken into account::
3313 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
3317 drop / drop / drop / end # drop is performed only once
3321 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
3323 Considering they are performed simultaneously, opposite and overlapping
3324 actions can sometimes be combined when the end result is unambiguous::
3326 drop / queue index 6 / end # drop has no effect
3330 drop / dup index 6 / end # same as above
3334 queue index 6 / rss queues 6 7 8 / end # queue has no effect
3338 drop / passthru / end # drop has no effect
3340 Note that PMDs may still refuse such combinations.
3345 This section lists supported actions and their attributes, if any.
3347 - ``end``: end list of actions.
3349 - ``void``: no-op action.
3351 - ``passthru``: let subsequent rule process matched packets.
3353 - ``mark``: attach 32 bit value to packets.
3355 - ``id {unsigned}``: 32 bit value to return with packets.
3357 - ``flag``: flag packets.
3359 - ``queue``: assign packets to a given queue index.
3361 - ``index {unsigned}``: queue index to use.
3363 - ``drop``: drop packets (note: passthru has priority).
3365 - ``count``: enable counters for this rule.
3367 - ``dup``: duplicate packets to a given queue index.
3369 - ``index {unsigned}``: queue index to duplicate packets to.
3371 - ``rss``: spread packets among several queues.
3373 - ``queues [{unsigned} [...]] end``: queue indices to use.
3375 - ``pf``: redirect packets to physical device function.
3377 - ``vf``: redirect packets to virtual device function.
3379 - ``original {boolean}``: use original VF ID if possible.
3380 - ``id {unsigned}``: VF ID to redirect packets to.
3382 Destroying flow rules
3383 ~~~~~~~~~~~~~~~~~~~~~
3385 ``flow destroy`` destroys one or more rules from their rule ID (as returned
3386 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
3387 times as necessary::
3389 flow destroy {port_id} rule {rule_id} [...]
3391 If successful, it will show::
3393 Flow rule #[...] destroyed
3395 It does not report anything for rule IDs that do not exist. The usual error
3396 message is shown when a rule cannot be destroyed::
3398 Caught error type [...] ([...]): [...]
3400 ``flow flush`` destroys all rules on a device and does not take extra
3401 arguments. It is bound to ``rte_flow_flush()``::
3403 flow flush {port_id}
3405 Any errors are reported as above.
3407 Creating several rules and destroying them::
3409 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3410 actions queue index 2 / end
3411 Flow rule #0 created
3412 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3413 actions queue index 3 / end
3414 Flow rule #1 created
3415 testpmd> flow destroy 0 rule 0 rule 1
3416 Flow rule #1 destroyed
3417 Flow rule #0 destroyed
3420 The same result can be achieved using ``flow flush``::
3422 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3423 actions queue index 2 / end
3424 Flow rule #0 created
3425 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3426 actions queue index 3 / end
3427 Flow rule #1 created
3428 testpmd> flow flush 0
3431 Non-existent rule IDs are ignored::
3433 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3434 actions queue index 2 / end
3435 Flow rule #0 created
3436 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3437 actions queue index 3 / end
3438 Flow rule #1 created
3439 testpmd> flow destroy 0 rule 42 rule 10 rule 2
3441 testpmd> flow destroy 0 rule 0
3442 Flow rule #0 destroyed
3448 ``flow query`` queries a specific action of a flow rule having that
3449 ability. Such actions collect information that can be reported using this
3450 command. It is bound to ``rte_flow_query()``::
3452 flow query {port_id} {rule_id} {action}
3454 If successful, it will display either the retrieved data for known actions
3455 or the following message::
3457 Cannot display result for action type [...] ([...])
3459 Otherwise, it will complain either that the rule does not exist or that some
3462 Flow rule #[...] not found
3466 Caught error type [...] ([...]): [...]
3468 Currently only the ``count`` action is supported. This action reports the
3469 number of packets that hit the flow rule and the total number of bytes. Its
3470 output has the following format::
3473 hits_set: [...] # whether "hits" contains a valid value
3474 bytes_set: [...] # whether "bytes" contains a valid value
3475 hits: [...] # number of packets
3476 bytes: [...] # number of bytes
3478 Querying counters for TCPv6 packets redirected to queue 6::
3480 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
3481 actions queue index 6 / count / end
3482 Flow rule #4 created
3483 testpmd> flow query 0 4 count
3494 ``flow list`` lists existing flow rules sorted by priority and optionally
3495 filtered by group identifiers::
3497 flow list {port_id} [group {group_id}] [...]
3499 This command only fails with the following message if the device does not
3504 Output consists of a header line followed by a short description of each
3505 flow rule, one per line. There is no output at all when no flow rules are
3506 configured on the device::
3508 ID Group Prio Attr Rule
3509 [...] [...] [...] [...] [...]
3511 ``Attr`` column flags:
3513 - ``i`` for ``ingress``.
3514 - ``e`` for ``egress``.
3516 Creating several flow rules and listing them::
3518 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3519 actions queue index 6 / end
3520 Flow rule #0 created
3521 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3522 actions queue index 2 / end
3523 Flow rule #1 created
3524 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
3525 actions rss queues 6 7 8 end / end
3526 Flow rule #2 created
3527 testpmd> flow list 0
3528 ID Group Prio Attr Rule
3529 0 0 0 i- ETH IPV4 => QUEUE
3530 1 0 0 i- ETH IPV6 => QUEUE
3531 2 0 5 i- ETH IPV4 UDP => RSS
3534 Rules are sorted by priority (i.e. group ID first, then priority level)::
3536 testpmd> flow list 1
3537 ID Group Prio Attr Rule
3538 0 0 0 i- ETH => COUNT
3539 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3540 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3541 1 24 0 i- ETH IPV4 UDP => QUEUE
3542 4 24 10 i- ETH IPV4 TCP => DROP
3543 3 24 20 i- ETH IPV4 => DROP
3544 2 24 42 i- ETH IPV4 UDP => QUEUE
3545 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3548 Output can be limited to specific groups::
3550 testpmd> flow list 1 group 0 group 63
3551 ID Group Prio Attr Rule
3552 0 0 0 i- ETH => COUNT
3553 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3554 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3555 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3558 Toggling isolated mode
3559 ~~~~~~~~~~~~~~~~~~~~~~
3561 ``flow isolate`` can be used to tell the underlying PMD that ingress traffic
3562 must only be injected from the defined flow rules; that no default traffic
3563 is expected outside those rules and the driver is free to assign more
3564 resources to handle them. It is bound to ``rte_flow_isolate()``::
3566 flow isolate {port_id} {boolean}
3568 If successful, enabling or disabling isolated mode shows either::
3570 Ingress traffic on port [...]
3571 is now restricted to the defined flow rules
3575 Ingress traffic on port [...]
3576 is not restricted anymore to the defined flow rules
3578 Otherwise, in case of error::
3580 Caught error type [...] ([...]): [...]
3582 Mainly due to its side effects, PMDs supporting this mode may not have the
3583 ability to toggle it more than once without reinitializing affected ports
3584 first (e.g. by exiting testpmd).
3586 Enabling isolated mode::
3588 testpmd> flow isolate 0 true
3589 Ingress traffic on port 0 is now restricted to the defined flow rules
3592 Disabling isolated mode::
3594 testpmd> flow isolate 0 false
3595 Ingress traffic on port 0 is not restricted anymore to the defined flow rules
3598 Sample QinQ flow rules
3599 ~~~~~~~~~~~~~~~~~~~~~~
3601 Before creating QinQ rule(s) the following commands should be issued to enable QinQ::
3603 testpmd> port stop 0
3604 testpmd> vlan set qinq on 0
3606 The above command sets the inner and outer TPID's to 0x8100.
3608 To change the TPID's the following commands should be used::
3610 testpmd> vlan set outer tpid 0xa100 0
3611 testpmd> vlan set inner tpid 0x9100 0
3612 testpmd> port start 0
3614 Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM.
3618 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 /
3619 vlan tci is 456 / end actions vf id 1 / queue index 0 / end
3620 Flow rule #0 validated
3622 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 /
3623 vlan tci is 456 / end actions vf id 123 / queue index 0 / end
3624 Flow rule #0 created
3626 testpmd> flow list 0
3627 ID Group Prio Attr Rule
3628 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3630 Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host.
3634 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 /
3635 vlan tci is 654 / end actions pf / queue index 0 / end
3636 Flow rule #1 validated
3638 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 /
3639 vlan tci is 654 / end actions pf / queue index 1 / end
3640 Flow rule #1 created
3642 testpmd> flow list 0
3643 ID Group Prio Attr Rule
3644 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3645 1 0 0 i- ETH VLAN VLAN=>PF QUEUE