2 Copyright(c) 2010-2016 Intel Corporation. All rights reserved.
5 Redistribution and use in source and binary forms, with or without
6 modification, are permitted provided that the following conditions
9 * Redistributions of source code must retain the above copyright
10 notice, this list of conditions and the following disclaimer.
11 * Redistributions in binary form must reproduce the above copyright
12 notice, this list of conditions and the following disclaimer in
13 the documentation and/or other materials provided with the
15 * Neither the name of Intel Corporation nor the names of its
16 contributors may be used to endorse or promote products derived
17 from this software without specific prior written permission.
19 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
20 "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
21 LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
22 A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
23 OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
24 SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
25 LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
26 DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
27 THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
28 (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
29 OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
33 Testpmd Runtime Functions
34 =========================
36 Where the testpmd application is started in interactive mode, (``-i|--interactive``),
37 it displays a prompt that can be used to start and stop forwarding,
38 configure the application, display statistics (including the extended NIC
39 statistics aka xstats) , set the Flow Director and other tasks::
43 The testpmd prompt has some, limited, readline support.
44 Common bash command-line functions such as ``Ctrl+a`` and ``Ctrl+e`` to go to the start and end of the prompt line are supported
45 as well as access to the command history via the up-arrow.
47 There is also support for tab completion.
48 If you type a partial command and hit ``<TAB>`` you get a list of the available completions:
50 .. code-block:: console
52 testpmd> show port <TAB>
54 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
55 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
56 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
57 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
63 Some examples in this document are too long to fit on one line are are shown wrapped at `"\\"` for display purposes::
65 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
66 (pause_time) (send_xon) (port_id)
68 In the real ``testpmd>`` prompt these commands should be on a single line.
73 The testpmd has on-line help for the functions that are available at runtime.
74 These are divided into sections and can be accessed using help, help section or help all:
76 .. code-block:: console
80 help control : Start and stop forwarding.
81 help display : Displaying port, stats and config information.
82 help config : Configuration information.
83 help ports : Configuring ports.
84 help registers : Reading and setting port registers.
85 help filters : Filters configuration help.
86 help all : All of the above sections.
89 Command File Functions
90 ----------------------
92 To facilitate loading large number of commands or to avoid cutting and pasting where not
93 practical or possible testpmd supports alternative methods for executing commands.
95 * If started with the ``--cmdline-file=FILENAME`` command line argument testpmd
96 will execute all CLI commands contained within the file immediately before
97 starting packet forwarding or entering interactive mode.
99 .. code-block:: console
101 ./testpmd -n4 -r2 ... -- -i --cmdline-file=/home/ubuntu/flow-create-commands.txt
102 Interactive-mode selected
103 CLI commands to be read from /home/ubuntu/flow-create-commands.txt
104 Configuring Port 0 (socket 0)
105 Port 0: 7C:FE:90:CB:74:CE
106 Configuring Port 1 (socket 0)
107 Port 1: 7C:FE:90:CB:74:CA
108 Checking link statuses...
109 Port 0 Link Up - speed 10000 Mbps - full-duplex
110 Port 1 Link Up - speed 10000 Mbps - full-duplex
116 Flow rule #498 created
117 Flow rule #499 created
118 Read all CLI commands from /home/ubuntu/flow-create-commands.txt
122 * At run-time additional commands can be loaded in bulk by invoking the ``load FILENAME``
125 .. code-block:: console
127 testpmd> load /home/ubuntu/flow-create-commands.txt
132 Flow rule #498 created
133 Flow rule #499 created
134 Read all CLI commands from /home/ubuntu/flow-create-commands.txt
138 In all cases output from any included command will be displayed as standard output.
139 Execution will continue until the end of the file is reached regardless of
140 whether any errors occur. The end user must examine the output to determine if
141 any failures occurred.
150 Start packet forwarding with current configuration::
157 Start packet forwarding with current configuration after sending specified number of bursts of packets::
159 testpmd> start tx_first (""|burst_num)
161 The default burst number is 1 when ``burst_num`` not presented.
166 Stop packet forwarding, and display accumulated statistics::
181 The functions in the following sections are used to display information about the
182 testpmd configuration or the NIC status.
187 Display information for a given port or all ports::
189 testpmd> show port (info|stats|xstats|fdir|stat_qmap|dcb_tc|cap) (port_id|all)
191 The available information categories are:
193 * ``info``: General port information such as MAC address.
195 * ``stats``: RX/TX statistics.
197 * ``xstats``: RX/TX extended NIC statistics.
199 * ``fdir``: Flow Director information and statistics.
201 * ``stat_qmap``: Queue statistics mapping.
203 * ``dcb_tc``: DCB information such as TC mapping.
205 * ``cap``: Supported offload capabilities.
209 .. code-block:: console
211 testpmd> show port info 0
213 ********************* Infos for port 0 *********************
215 MAC address: XX:XX:XX:XX:XX:XX
217 memory allocation on the socket: 0
219 Link speed: 40000 Mbps
220 Link duplex: full-duplex
221 Promiscuous mode: enabled
222 Allmulticast mode: disabled
223 Maximum number of MAC addresses: 64
224 Maximum number of MAC addresses of hash filtering: 0
229 Redirection table size: 512
230 Supported flow types:
250 Display the rss redirection table entry indicated by masks on port X::
252 testpmd> show port (port_id) rss reta (size) (mask0, mask1...)
254 size is used to indicate the hardware supported reta size
259 Display the RSS hash functions and RSS hash key of a port::
261 testpmd> show port (port_id) rss-hash ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2-payload|ipv6-ex|ipv6-tcp-ex|ipv6-udp-ex [key]
266 Clear the port statistics for a given port or for all ports::
268 testpmd> clear port (info|stats|xstats|fdir|stat_qmap) (port_id|all)
272 testpmd> clear port stats all
277 Display information for a given port's RX/TX queue::
279 testpmd> show (rxq|txq) info (port_id) (queue_id)
284 Displays the configuration of the application.
285 The configuration comes from the command-line, the runtime or the application defaults::
287 testpmd> show config (rxtx|cores|fwd|txpkts)
289 The available information categories are:
291 * ``rxtx``: RX/TX configuration items.
293 * ``cores``: List of forwarding cores.
295 * ``fwd``: Packet forwarding configuration.
297 * ``txpkts``: Packets to TX configuration.
301 .. code-block:: console
303 testpmd> show config rxtx
305 io packet forwarding - CRC stripping disabled - packets/burst=16
306 nb forwarding cores=2 - nb forwarding ports=1
307 RX queues=1 - RX desc=128 - RX free threshold=0
308 RX threshold registers: pthresh=8 hthresh=8 wthresh=4
309 TX queues=1 - TX desc=512 - TX free threshold=0
310 TX threshold registers: pthresh=36 hthresh=0 wthresh=0
311 TX RS bit threshold=0 - TXQ flags=0x0
316 Set the packet forwarding mode::
318 testpmd> set fwd (io|mac|macswap|flowgen| \
319 rxonly|txonly|csum|icmpecho) (""|retry)
321 ``retry`` can be specified for forwarding engines except ``rx_only``.
323 The available information categories are:
325 * ``io``: Forwards packets "as-is" in I/O mode.
326 This is the fastest possible forwarding operation as it does not access packets data.
327 This is the default mode.
329 * ``mac``: Changes the source and the destination Ethernet addresses of packets before forwarding them.
330 Default application behaviour is to set source Ethernet address to that of the transmitting interface, and destination
331 address to a dummy value (set during init). The user may specify a target destination Ethernet address via the 'eth-peer' or
332 'eth-peer-configfile' command-line options. It is not currently possible to specify a specific source Ethernet address.
334 * ``macswap``: MAC swap forwarding mode.
335 Swaps the source and the destination Ethernet addresses of packets before forwarding them.
337 * ``flowgen``: Multi-flow generation mode.
338 Originates a number of flows (with varying destination IP addresses), and terminate receive traffic.
340 * ``rxonly``: Receives packets but doesn't transmit them.
342 * ``txonly``: Generates and transmits packets without receiving any.
344 * ``csum``: Changes the checksum field with hardware or software methods depending on the offload flags on the packet.
346 * ``icmpecho``: Receives a burst of packets, lookup for IMCP echo requests and, if any, send back ICMP echo replies.
348 * ``ieee1588``: Demonstrate L2 IEEE1588 V2 PTP timestamping for RX and TX. Requires ``CONFIG_RTE_LIBRTE_IEEE1588=y``.
350 * ``tm``: Traffic Management forwarding mode
351 Demonstrates the use of ethdev traffic management APIs and softnic PMD for
352 QoS traffic management. In this mode, 5-level hierarchical QoS scheduler is
353 available as an default option that can be enabled through CLI. The user can
354 also modify the default hierarchy or specify the new hierarchy through CLI for
355 implementing QoS scheduler. Requires ``CONFIG_RTE_LIBRTE_PMD_SOFTNIC=y`` ``CONFIG_RTE_LIBRTE_SCHED=y``.
357 Note: TX timestamping is only available in the "Full Featured" TX path. To force ``testpmd`` into this mode set ``--txqflags=0``.
361 testpmd> set fwd rxonly
363 Set rxonly packet forwarding mode
369 Display an RX descriptor for a port RX queue::
371 testpmd> read rxd (port_id) (queue_id) (rxd_id)
375 testpmd> read rxd 0 0 4
376 0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180
381 Display a TX descriptor for a port TX queue::
383 testpmd> read txd (port_id) (queue_id) (txd_id)
387 testpmd> read txd 0 0 4
388 0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C
393 Get loaded dynamic device personalization (DDP) package info list::
395 testpmd> ddp get list (port_id)
400 Display information about dynamic device personalization (DDP) profile::
402 testpmd> ddp get info (profile_path)
407 Display VF statistics::
409 testpmd> show vf stats (port_id) (vf_id)
414 Reset VF statistics::
416 testpmd> clear vf stats (port_id) (vf_id)
418 show port pctype mapping
419 ~~~~~~~~~~~~~~~~~~~~~~~~
421 List all items from the pctype mapping table::
423 testpmd> show port (port_id) pctype mapping
426 Configuration Functions
427 -----------------------
429 The testpmd application can be configured from the runtime as well as from the command-line.
431 This section details the available configuration functions that are available.
435 Configuration changes only become active when forwarding is started/restarted.
440 Reset forwarding to the default configuration::
447 Set the debug verbosity level::
449 testpmd> set verbose (level)
451 Currently the only available levels are 0 (silent except for error) and 1 (fully verbose).
456 Set the number of ports used by the application:
460 This is equivalent to the ``--nb-ports`` command-line option.
465 Set the number of cores used by the application::
467 testpmd> set nbcore (num)
469 This is equivalent to the ``--nb-cores`` command-line option.
473 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port.
478 Set the forwarding cores hexadecimal mask::
480 testpmd> set coremask (mask)
482 This is equivalent to the ``--coremask`` command-line option.
486 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding.
491 Set the forwarding ports hexadecimal mask::
493 testpmd> set portmask (mask)
495 This is equivalent to the ``--portmask`` command-line option.
500 Set number of packets per burst::
502 testpmd> set burst (num)
504 This is equivalent to the ``--burst command-line`` option.
506 When retry is enabled, the transmit delay time and number of retries can also be set::
508 testpmd> set burst tx delay (microseconds) retry (num)
513 Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode::
515 testpmd> set txpkts (x[,y]*)
517 Where x[,y]* represents a CSV list of values, without white space.
522 Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes::
524 testpmd> set txsplit (off|on|rand)
528 * ``off`` disable packet copy & split for CSUM mode.
530 * ``on`` split outgoing packet into multiple segments. Size of each segment
531 and number of segments per packet is determined by ``set txpkts`` command
534 * ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments.
539 Set the list of forwarding cores::
541 testpmd> set corelist (x[,y]*)
543 For example, to change the forwarding cores:
545 .. code-block:: console
547 testpmd> set corelist 3,1
548 testpmd> show config fwd
550 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled
551 Logical Core 3 (socket 0) forwards packets on 1 streams:
552 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01
553 Logical Core 1 (socket 0) forwards packets on 1 streams:
554 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
558 The cores are used in the same order as specified on the command line.
563 Set the list of forwarding ports::
565 testpmd> set portlist (x[,y]*)
567 For example, to change the port forwarding:
569 .. code-block:: console
571 testpmd> set portlist 0,2,1,3
572 testpmd> show config fwd
574 io packet forwarding - ports=4 - cores=1 - streams=4
575 Logical Core 3 (socket 0) forwards packets on 4 streams:
576 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01
577 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
578 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03
579 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02
584 Enable/disable tx loopback::
586 testpmd> set tx loopback (port_id) (on|off)
591 set drop enable bit for all queues::
593 testpmd> set all queues drop (port_id) (on|off)
595 set split drop enable (for VF)
596 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
598 set split drop enable bit for VF from PF::
600 testpmd> set vf split drop (port_id) (vf_id) (on|off)
602 set mac antispoof (for VF)
603 ~~~~~~~~~~~~~~~~~~~~~~~~~~
605 Set mac antispoof for a VF from the PF::
607 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off)
612 Enable/disable MACsec offload::
614 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off)
615 testpmd> set macsec offload (port_id) off
620 Configure MACsec secure connection (SC)::
622 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi)
626 The pi argument is ignored for tx.
627 Check the NIC Datasheet for hardware limits.
632 Configure MACsec secure association (SA)::
634 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key)
638 The IDX value must be 0 or 1.
639 Check the NIC Datasheet for hardware limits.
641 set broadcast mode (for VF)
642 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
644 Set broadcast mode for a VF from the PF::
646 testpmd> set vf broadcast (port_id) (vf_id) (on|off)
651 Set the VLAN strip on a port::
653 testpmd> vlan set strip (on|off) (port_id)
658 Set the VLAN strip for a queue on a port::
660 testpmd> vlan set stripq (on|off) (port_id,queue_id)
662 vlan set stripq (for VF)
663 ~~~~~~~~~~~~~~~~~~~~~~~~
665 Set VLAN strip for all queues in a pool for a VF from the PF::
667 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off)
669 vlan set insert (for VF)
670 ~~~~~~~~~~~~~~~~~~~~~~~~
672 Set VLAN insert for a VF from the PF::
674 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id)
676 vlan set tag (for VF)
677 ~~~~~~~~~~~~~~~~~~~~~
679 Set VLAN tag for a VF from the PF::
681 testpmd> set vf vlan tag (port_id) (vf_id) (on|off)
683 vlan set antispoof (for VF)
684 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
686 Set VLAN antispoof for a VF from the PF::
688 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off)
693 Set the VLAN filter on a port::
695 testpmd> vlan set filter (on|off) (port_id)
700 Set the VLAN QinQ (extended queue in queue) on for a port::
702 testpmd> vlan set qinq (on|off) (port_id)
707 Set the inner or outer VLAN TPID for packet filtering on a port::
709 testpmd> vlan set (inner|outer) tpid (value) (port_id)
713 TPID value must be a 16-bit number (value <= 65536).
718 Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID::
720 testpmd> rx_vlan add (vlan_id|all) (port_id)
724 VLAN filter must be set on that port. VLAN ID < 4096.
725 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries
726 in VFTA table. This is important if enabling all vlan_ids.
731 Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID::
733 testpmd> rx_vlan rm (vlan_id|all) (port_id)
738 Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID::
740 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask)
745 Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID::
747 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask)
752 Add a tunnel filter on a port::
754 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \
755 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
756 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
758 The available information categories are:
760 * ``vxlan``: Set tunnel type as VXLAN.
762 * ``nvgre``: Set tunnel type as NVGRE.
764 * ``ipingre``: Set tunnel type as IP-in-GRE.
766 * ``imac-ivlan``: Set filter type as Inner MAC and VLAN.
768 * ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID.
770 * ``imac-tenid``: Set filter type as Inner MAC and tenant ID.
772 * ``imac``: Set filter type as Inner MAC.
774 * ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID.
776 * ``oip``: Set filter type as Outer IP.
778 * ``iip``: Set filter type as Inner IP.
782 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \
783 192.168.2.2 0 ipingre oip 1 1
785 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP.
790 Remove a tunnel filter on a port::
792 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \
793 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
794 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
799 Add an UDP port for VXLAN packet filter on a port::
801 testpmd> rx_vxlan_port add (udp_port) (port_id)
806 Remove an UDP port for VXLAN packet filter on a port::
808 testpmd> rx_vxlan_port rm (udp_port) (port_id)
813 Set hardware insertion of VLAN IDs in packets sent on a port::
815 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer]
817 For example, set a single VLAN ID (5) insertion on port 0::
821 Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1::
829 Set port based hardware insertion of VLAN ID in packets sent on a port::
831 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off)
836 Disable hardware insertion of a VLAN header in packets sent on a port::
838 testpmd> tx_vlan reset (port_id)
843 Select hardware or software calculation of the checksum when
844 transmitting a packet using the ``csum`` forwarding engine::
846 testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id)
850 * ``ip|udp|tcp|sctp`` always relate to the inner layer.
852 * ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized
853 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are
854 supported). See also the ``csum parse-tunnel`` command.
858 Check the NIC Datasheet for hardware limits.
863 Set RSS queue region span on a port::
865 testpmd> set port (port_id) queue-region region_id (value) \
866 queue_start_index (value) queue_num (value)
868 Set flowtype mapping on a RSS queue region on a port::
870 testpmd> set port (port_id) queue-region region_id (value) flowtype (value)
874 * For the flowtype(pctype) of packet,the specific index for each type has
875 been defined in file i40e_type.h as enum i40e_filter_pctype.
877 Set user priority mapping on a RSS queue region on a port::
879 testpmd> set port (port_id) queue-region UP (value) region_id (value)
881 Flush all queue region related configuration on a port::
883 testpmd> set port (port_id) queue-region flush (on|off)
887 * "on"is just an enable function which server for other configuration,
888 it is for all configuration about queue region from up layer,
889 at first will only keep in DPDK softwarestored in driver,
890 only after "flush on", it commit all configuration to HW.
891 "off" is just clean all configuration about queue region just now,
892 and restore all to DPDK i40e driver default config when start up.
894 Show all queue region related configuration info on a port::
896 testpmd> show port (port_id) queue-region
900 Queue region only support on PF by now, so these command is
901 only for configuration of queue region on PF port.
906 Define how tunneled packets should be handled by the csum forward
909 testpmd> csum parse-tunnel (on|off) (tx_port_id)
911 If enabled, the csum forward engine will try to recognize supported
912 tunnel headers (vxlan, gre, ipip).
914 If disabled, treat tunnel packets as non-tunneled packets (a inner
915 header is handled as a packet payload).
919 The port argument is the TX port like in the ``csum set`` command.
923 Consider a packet in packet like the following::
925 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in
927 * If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
928 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the
929 ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``).
931 * If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
932 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``.
937 Display tx checksum offload configuration::
939 testpmd> csum show (port_id)
944 Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine::
946 testpmd> tso set (segsize) (port_id)
950 Check the NIC datasheet for hardware limits.
955 Display the status of TCP Segmentation Offload::
957 testpmd> tso show (port_id)
962 Enable or disable GRO in ``csum`` forwarding engine::
964 testpmd> set port <port_id> gro on|off
966 If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4
967 packets received from the given port.
969 If disabled, packets received from the given port won't be performed
970 GRO. By default, GRO is disabled for all ports.
974 When enable GRO for a port, TCP/IPv4 packets received from the port
975 will be performed GRO. After GRO, all merged packets have bad
976 checksums, since the GRO library doesn't re-calculate checksums for
977 the merged packets. Therefore, if users want the merged packets to
978 have correct checksums, please select HW IP checksum calculation and
979 HW TCP checksum calculation for the port which the merged packets are
985 Display GRO configuration for a given port::
987 testpmd> show port <port_id> gro
992 Set the cycle to flush the GROed packets from reassembly tables::
994 testpmd> set gro flush <cycles>
996 When enable GRO, the csum forwarding engine performs GRO on received
997 packets, and the GROed packets are stored in reassembly tables. Users
998 can use this command to determine when the GROed packets are flushed
999 from the reassembly tables.
1001 The ``cycles`` is measured in GRO operation times. The csum forwarding
1002 engine flushes the GROed packets from the tables every ``cycles`` GRO
1005 By default, the value of ``cycles`` is 1, which means flush GROed packets
1006 from the reassembly tables as soon as one GRO operation finishes. The value
1007 of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``.
1009 Please note that the large value of ``cycles`` may cause the poor TCP/IP
1010 stack performance. Because the GROed packets are delayed to arrive the
1011 stack, thus causing more duplicated ACKs and TCP retransmissions.
1016 Toggle per-port GSO support in ``csum`` forwarding engine::
1018 testpmd> set port <port_id> gso on|off
1020 If enabled, the csum forwarding engine will perform GSO on supported IPv4
1021 packets, transmitted on the given port.
1023 If disabled, packets transmitted on the given port will not undergo GSO.
1024 By default, GSO is disabled for all ports.
1028 When GSO is enabled on a port, supported IPv4 packets transmitted on that
1029 port undergo GSO. Afterwards, the segmented packets are represented by
1030 multi-segment mbufs; however, the csum forwarding engine doesn't calculation
1031 of checksums for GSO'd segments in SW. As a result, if users want correct
1032 checksums in GSO segments, they should enable HW checksum calculation for
1035 For example, HW checksum calculation for VxLAN GSO'd packets may be enabled
1036 by setting the following options in the csum forwarding engine:
1038 testpmd> csum set outer_ip hw <port_id>
1040 testpmd> csum set ip hw <port_id>
1042 testpmd> csum set tcp hw <port_id>
1047 Set the maximum GSO segment size (measured in bytes), which includes the
1048 packet header and the packet payload for GSO-enabled ports (global)::
1050 testpmd> set gso segsz <length>
1055 Display the status of Generic Segmentation Offload for a given port::
1057 testpmd> show port <port_id> gso
1062 Add an alternative MAC address to a port::
1064 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX)
1069 Remove a MAC address from a port::
1071 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX)
1073 mac_addr add (for VF)
1074 ~~~~~~~~~~~~~~~~~~~~~
1076 Add an alternative MAC address for a VF to a port::
1078 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX)
1083 Set the default MAC address for a port::
1085 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX)
1087 mac_addr set (for VF)
1088 ~~~~~~~~~~~~~~~~~~~~~
1090 Set the MAC address for a VF from the PF::
1092 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX)
1097 Set the unicast hash filter(s) on/off for a port::
1099 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
1104 Set the promiscuous mode on for a port or for all ports.
1105 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1107 testpmd> set promisc (port_id|all) (on|off)
1112 Set the allmulti mode for a port or for all ports::
1114 testpmd> set allmulti (port_id|all) (on|off)
1116 Same as the ifconfig (8) option. Controls how multicast packets are handled.
1118 set promisc (for VF)
1119 ~~~~~~~~~~~~~~~~~~~~
1121 Set the unicast promiscuous mode for a VF from PF.
1122 It's supported by Intel i40e NICs now.
1123 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1125 testpmd> set vf promisc (port_id) (vf_id) (on|off)
1127 set allmulticast (for VF)
1128 ~~~~~~~~~~~~~~~~~~~~~~~~~
1130 Set the multicast promiscuous mode for a VF from PF.
1131 It's supported by Intel i40e NICs now.
1132 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1134 testpmd> set vf allmulti (port_id) (vf_id) (on|off)
1136 set tx max bandwidth (for VF)
1137 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1139 Set TX max absolute bandwidth (Mbps) for a VF from PF::
1141 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth)
1143 set tc tx min bandwidth (for VF)
1144 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1146 Set all TCs' TX min relative bandwidth (%) for a VF from PF::
1148 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...)
1150 set tc tx max bandwidth (for VF)
1151 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1153 Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF::
1155 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth)
1157 set tc strict link priority mode
1158 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1160 Set some TCs' strict link priority mode on a physical port::
1162 testpmd> set tx strict-link-priority (port_id) (tc_bitmap)
1164 set tc tx min bandwidth
1165 ~~~~~~~~~~~~~~~~~~~~~~~
1167 Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs::
1169 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...)
1174 Set the link flow control parameter on a port::
1176 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1177 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
1178 autoneg (on|off) (port_id)
1182 * ``high_water`` (integer): High threshold value to trigger XOFF.
1184 * ``low_water`` (integer): Low threshold value to trigger XON.
1186 * ``pause_time`` (integer): Pause quota in the Pause frame.
1188 * ``send_xon`` (0/1): Send XON frame.
1190 * ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
1192 * ``autoneg``: Change the auto-negotiation parameter.
1197 Set the priority flow control parameter on a port::
1199 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1200 (pause_time) (priority) (port_id)
1204 * ``high_water`` (integer): High threshold value.
1206 * ``low_water`` (integer): Low threshold value.
1208 * ``pause_time`` (integer): Pause quota in the Pause frame.
1210 * ``priority`` (0-7): VLAN User Priority.
1215 Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
1217 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
1219 For example, to set rx queue 2 on port 0 to mapping 5::
1221 testpmd>set stat_qmap rx 0 2 5
1223 set xstats-hide-zero
1224 ~~~~~~~~~~~~~~~~~~~~
1226 Set the option to hide zero values for xstats display::
1228 testpmd> set xstats-hide-zero on|off
1232 By default, the zero values are displayed for xstats.
1234 set port - rx/tx (for VF)
1235 ~~~~~~~~~~~~~~~~~~~~~~~~~
1237 Set VF receive/transmit from a port::
1239 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
1241 set port - mac address filter (for VF)
1242 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1244 Add/Remove unicast or multicast MAC addr filter for a VF::
1246 testpmd> set port (port_id) vf (vf_id) (mac_addr) \
1247 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
1249 set port - rx mode(for VF)
1250 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1252 Set the VF receive mode of a port::
1254 testpmd> set port (port_id) vf (vf_id) \
1255 rxmode (AUPE|ROPE|BAM|MPE) (on|off)
1257 The available receive modes are:
1259 * ``AUPE``: Accepts untagged VLAN.
1261 * ``ROPE``: Accepts unicast hash.
1263 * ``BAM``: Accepts broadcast packets.
1265 * ``MPE``: Accepts all multicast packets.
1267 set port - tx_rate (for Queue)
1268 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1270 Set TX rate limitation for a queue on a port::
1272 testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1274 set port - tx_rate (for VF)
1275 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
1277 Set TX rate limitation for queues in VF on a port::
1279 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1281 set port - mirror rule
1282 ~~~~~~~~~~~~~~~~~~~~~~
1284 Set pool or vlan type mirror rule for a port::
1286 testpmd> set port (port_id) mirror-rule (rule_id) \
1287 (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1288 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1290 Set link mirror rule for a port::
1292 testpmd> set port (port_id) mirror-rule (rule_id) \
1293 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1295 For example to enable mirror traffic with vlan 0,1 to pool 0::
1297 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1299 reset port - mirror rule
1300 ~~~~~~~~~~~~~~~~~~~~~~~~
1302 Reset a mirror rule for a port::
1304 testpmd> reset port (port_id) mirror-rule (rule_id)
1309 Set the flush on RX streams before forwarding.
1310 The default is flush ``on``.
1311 Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1313 testpmd> set flush_rx off
1318 Set the bypass mode for the lowest port on bypass enabled NIC::
1320 testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1325 Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1327 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1328 mode (normal|bypass|isolate) (port_id)
1332 * ``timeout``: Enable bypass after watchdog timeout.
1334 * ``os_on``: Enable bypass when OS/board is powered on.
1336 * ``os_off``: Enable bypass when OS/board is powered off.
1338 * ``power_on``: Enable bypass when power supply is turned on.
1340 * ``power_off``: Enable bypass when power supply is turned off.
1346 Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1348 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1353 Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1355 testpmd> show bypass config (port_id)
1360 Set link up for a port::
1362 testpmd> set link-up port (port id)
1367 Set link down for a port::
1369 testpmd> set link-down port (port id)
1374 Enable E-tag insertion for a VF on a port::
1376 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1378 Disable E-tag insertion for a VF on a port::
1380 testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1382 Enable/disable E-tag stripping on a port::
1384 testpmd> E-tag set stripping (on|off) port (port_id)
1386 Enable/disable E-tag based forwarding on a port::
1388 testpmd> E-tag set forwarding (on|off) port (port_id)
1390 Add an E-tag forwarding filter on a port::
1392 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1394 Delete an E-tag forwarding filter on a port::
1395 testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1400 Load a dynamic device personalization (DDP) package::
1402 testpmd> ddp add (port_id) (package_path[,output_path])
1407 Delete a dynamic device personalization package::
1409 testpmd> ddp del (port_id) (package_path)
1414 List all items from the ptype mapping table::
1416 testpmd> ptype mapping get (port_id) (valid_only)
1420 * ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0).
1422 Replace a specific or a group of software defined ptype with a new one::
1424 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type)
1428 * ``target``: A specific software ptype or a mask to represent a group of software ptypes.
1430 * ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1).
1432 * ``pkt_type``: The new software ptype to replace the old ones.
1434 Update hardware defined ptype to software defined packet type mapping table::
1436 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype)
1440 * ``hw_ptype``: hardware ptype as the index of the ptype mapping table.
1442 * ``sw_ptype``: software ptype as the value of the ptype mapping table.
1444 Reset ptype mapping table::
1446 testpmd> ptype mapping reset (port_id)
1451 The following sections show functions for configuring ports.
1455 Port configuration changes only become active when forwarding is started/restarted.
1460 Attach a port specified by pci address or virtual device args::
1462 testpmd> port attach (identifier)
1464 To attach a new pci device, the device should be recognized by kernel first.
1465 Then it should be moved under DPDK management.
1466 Finally the port can be attached to testpmd.
1468 For example, to move a pci device using ixgbe under DPDK management:
1470 .. code-block:: console
1472 # Check the status of the available devices.
1473 ./usertools/dpdk-devbind.py --status
1475 Network devices using DPDK-compatible driver
1476 ============================================
1479 Network devices using kernel driver
1480 ===================================
1481 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1484 # Bind the device to igb_uio.
1485 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1488 # Recheck the status of the devices.
1489 ./usertools/dpdk-devbind.py --status
1490 Network devices using DPDK-compatible driver
1491 ============================================
1492 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1494 To attach a port created by virtual device, above steps are not needed.
1496 For example, to attach a port whose pci address is 0000:0a:00.0.
1498 .. code-block:: console
1500 testpmd> port attach 0000:0a:00.0
1501 Attaching a new port...
1502 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1503 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1504 EAL: PCI memory mapped at 0x7f83bfa00000
1505 EAL: PCI memory mapped at 0x7f83bfa80000
1506 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1507 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1508 Port 0 is attached. Now total ports is 1
1511 For example, to attach a port created by pcap PMD.
1513 .. code-block:: console
1515 testpmd> port attach net_pcap0
1516 Attaching a new port...
1517 PMD: Initializing pmd_pcap for net_pcap0
1518 PMD: Creating pcap-backed ethdev on numa socket 0
1519 Port 0 is attached. Now total ports is 1
1522 In this case, identifier is ``net_pcap0``.
1523 This identifier format is the same as ``--vdev`` format of DPDK applications.
1525 For example, to re-attach a bonded port which has been previously detached,
1526 the mode and slave parameters must be given.
1528 .. code-block:: console
1530 testpmd> port attach net_bond_0,mode=0,slave=1
1531 Attaching a new port...
1532 EAL: Initializing pmd_bond for net_bond_0
1533 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1534 Port 0 is attached. Now total ports is 1
1541 Detach a specific port::
1543 testpmd> port detach (port_id)
1545 Before detaching a port, the port should be stopped and closed.
1547 For example, to detach a pci device port 0.
1549 .. code-block:: console
1551 testpmd> port stop 0
1554 testpmd> port close 0
1558 testpmd> port detach 0
1560 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1561 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1562 EAL: PCI memory unmapped at 0x7f83bfa00000
1563 EAL: PCI memory unmapped at 0x7f83bfa80000
1567 For example, to detach a virtual device port 0.
1569 .. code-block:: console
1571 testpmd> port stop 0
1574 testpmd> port close 0
1578 testpmd> port detach 0
1580 PMD: Closing pcap ethdev on numa socket 0
1581 Port 'net_pcap0' is detached. Now total ports is 0
1584 To remove a pci device completely from the system, first detach the port from testpmd.
1585 Then the device should be moved under kernel management.
1586 Finally the device can be removed using kernel pci hotplug functionality.
1588 For example, to move a pci device under kernel management:
1590 .. code-block:: console
1592 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1594 ./usertools/dpdk-devbind.py --status
1596 Network devices using DPDK-compatible driver
1597 ============================================
1600 Network devices using kernel driver
1601 ===================================
1602 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1604 To remove a port created by a virtual device, above steps are not needed.
1609 Start all ports or a specific port::
1611 testpmd> port start (port_id|all)
1616 Stop all ports or a specific port::
1618 testpmd> port stop (port_id|all)
1623 Close all ports or a specific port::
1625 testpmd> port close (port_id|all)
1627 port start/stop queue
1628 ~~~~~~~~~~~~~~~~~~~~~
1630 Start/stop a rx/tx queue on a specific port::
1632 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1634 Only take effect when port is started.
1639 Set the speed and duplex mode for all ports or a specific port::
1641 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1642 duplex (half|full|auto)
1644 port config - queues/descriptors
1645 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1647 Set number of queues/descriptors for rxq, txq, rxd and txd::
1649 testpmd> port config all (rxq|txq|rxd|txd) (value)
1651 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1653 port config - max-pkt-len
1654 ~~~~~~~~~~~~~~~~~~~~~~~~~
1656 Set the maximum packet length::
1658 testpmd> port config all max-pkt-len (value)
1660 This is equivalent to the ``--max-pkt-len`` command-line option.
1662 port config - CRC Strip
1663 ~~~~~~~~~~~~~~~~~~~~~~~
1665 Set hardware CRC stripping on or off for all ports::
1667 testpmd> port config all crc-strip (on|off)
1669 CRC stripping is on by default.
1671 The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option.
1673 port config - scatter
1674 ~~~~~~~~~~~~~~~~~~~~~~~
1676 Set RX scatter mode on or off for all ports::
1678 testpmd> port config all scatter (on|off)
1680 RX scatter mode is off by default.
1682 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1684 port config - TX queue flags
1685 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1687 Set a hexadecimal bitmap of TX queue flags for all ports::
1689 testpmd> port config all txqflags value
1691 This command is equivalent to the ``--txqflags`` command-line option.
1693 port config - RX Checksum
1694 ~~~~~~~~~~~~~~~~~~~~~~~~~
1696 Set hardware RX checksum offload to on or off for all ports::
1698 testpmd> port config all rx-cksum (on|off)
1700 Checksum offload is off by default.
1702 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1707 Set hardware VLAN on or off for all ports::
1709 testpmd> port config all hw-vlan (on|off)
1711 Hardware VLAN is on by default.
1713 The ``off`` option is equivalent to the ``--disable-hw-vlan`` command-line option.
1715 port config - VLAN filter
1716 ~~~~~~~~~~~~~~~~~~~~~~~~~
1718 Set hardware VLAN filter on or off for all ports::
1720 testpmd> port config all hw-vlan-filter (on|off)
1722 Hardware VLAN filter is on by default.
1724 The ``off`` option is equivalent to the ``--disable-hw-vlan-filter`` command-line option.
1726 port config - VLAN strip
1727 ~~~~~~~~~~~~~~~~~~~~~~~~
1729 Set hardware VLAN strip on or off for all ports::
1731 testpmd> port config all hw-vlan-strip (on|off)
1733 Hardware VLAN strip is on by default.
1735 The ``off`` option is equivalent to the ``--disable-hw-vlan-strip`` command-line option.
1737 port config - VLAN extend
1738 ~~~~~~~~~~~~~~~~~~~~~~~~~
1740 Set hardware VLAN extend on or off for all ports::
1742 testpmd> port config all hw-vlan-extend (on|off)
1744 Hardware VLAN extend is off by default.
1746 The ``off`` option is equivalent to the ``--disable-hw-vlan-extend`` command-line option.
1748 port config - Drop Packets
1749 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1751 Set packet drop for packets with no descriptors on or off for all ports::
1753 testpmd> port config all drop-en (on|off)
1755 Packet dropping for packets with no descriptors is off by default.
1757 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1762 Set the RSS (Receive Side Scaling) mode on or off::
1764 testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1766 RSS is on by default.
1768 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1770 port config - RSS Reta
1771 ~~~~~~~~~~~~~~~~~~~~~~
1773 Set the RSS (Receive Side Scaling) redirection table::
1775 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1780 Set the DCB mode for an individual port::
1782 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1784 The traffic class should be 4 or 8.
1789 Set the number of packets per burst::
1791 testpmd> port config all burst (value)
1793 This is equivalent to the ``--burst`` command-line option.
1795 port config - Threshold
1796 ~~~~~~~~~~~~~~~~~~~~~~~
1798 Set thresholds for TX/RX queues::
1800 testpmd> port config all (threshold) (value)
1802 Where the threshold type can be:
1804 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1806 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1808 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1810 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1812 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1814 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1816 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1818 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1820 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1822 These threshold options are also available from the command-line.
1827 Set the value of ether-type for E-tag::
1829 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1831 Enable/disable the E-tag support::
1833 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1835 port config pctype mapping
1836 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1838 Reset pctype mapping table::
1840 testpmd> port config (port_id) pctype mapping reset
1842 Update hardware defined pctype to software defined flow type mapping table::
1844 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id)
1848 * ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table.
1850 * ``flow_type_id``: software flow type id as the index of the pctype mapping table.
1853 Link Bonding Functions
1854 ----------------------
1856 The Link Bonding functions make it possible to dynamically create and
1857 manage link bonding devices from within testpmd interactive prompt.
1859 create bonded device
1860 ~~~~~~~~~~~~~~~~~~~~
1862 Create a new bonding device::
1864 testpmd> create bonded device (mode) (socket)
1866 For example, to create a bonded device in mode 1 on socket 0::
1868 testpmd> create bonded 1 0
1869 created new bonded device (port X)
1874 Adds Ethernet device to a Link Bonding device::
1876 testpmd> add bonding slave (slave id) (port id)
1878 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
1880 testpmd> add bonding slave 6 10
1883 remove bonding slave
1884 ~~~~~~~~~~~~~~~~~~~~
1886 Removes an Ethernet slave device from a Link Bonding device::
1888 testpmd> remove bonding slave (slave id) (port id)
1890 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
1892 testpmd> remove bonding slave 6 10
1897 Set the Link Bonding mode of a Link Bonding device::
1899 testpmd> set bonding mode (value) (port id)
1901 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
1903 testpmd> set bonding mode 3 10
1908 Set an Ethernet slave device as the primary device on a Link Bonding device::
1910 testpmd> set bonding primary (slave id) (port id)
1912 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
1914 testpmd> set bonding primary 6 10
1919 Set the MAC address of a Link Bonding device::
1921 testpmd> set bonding mac (port id) (mac)
1923 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
1925 testpmd> set bonding mac 10 00:00:00:00:00:01
1927 set bonding xmit_balance_policy
1928 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1930 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
1932 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
1934 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
1936 testpmd> set bonding xmit_balance_policy 10 l34
1939 set bonding mon_period
1940 ~~~~~~~~~~~~~~~~~~~~~~
1942 Set the link status monitoring polling period in milliseconds for a bonding device.
1944 This adds support for PMD slave devices which do not support link status interrupts.
1945 When the mon_period is set to a value greater than 0 then all PMD's which do not support
1946 link status ISR will be queried every polling interval to check if their link status has changed::
1948 testpmd> set bonding mon_period (port_id) (value)
1950 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
1952 testpmd> set bonding mon_period 5 150
1955 set bonding lacp dedicated_queue
1956 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1958 Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic
1959 when in mode 4 (link-aggregration-802.3ad)::
1961 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable)
1964 set bonding agg_mode
1965 ~~~~~~~~~~~~~~~~~~~~
1967 Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad)::
1969 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable)
1975 Show the current configuration of a Link Bonding device::
1977 testpmd> show bonding config (port id)
1980 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
1981 in balance mode with a transmission policy of layer 2+3::
1983 testpmd> show bonding config 9
1985 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
1987 Active Slaves (3): [1 3 4]
1994 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
1995 This is mainly useful for debugging purposes.
1996 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
1997 and fields that can be accessed.
2002 Display the value of a port register::
2004 testpmd> read reg (port_id) (address)
2006 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
2008 testpmd> read reg 0 0xEE00
2009 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
2014 Display a port register bit field::
2016 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
2018 For example, reading the lowest two bits from the register in the example above::
2020 testpmd> read regfield 0 0xEE00 0 1
2021 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
2026 Display a single port register bit::
2028 testpmd> read regbit (port_id) (address) (bit_x)
2030 For example, reading the lowest bit from the register in the example above::
2032 testpmd> read regbit 0 0xEE00 0
2033 port 0 PCI register at offset 0xEE00: bit 0=1
2038 Set the value of a port register::
2040 testpmd> write reg (port_id) (address) (value)
2042 For example, to clear a register::
2044 testpmd> write reg 0 0xEE00 0x0
2045 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
2050 Set bit field of a port register::
2052 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
2054 For example, writing to the register cleared in the example above::
2056 testpmd> write regfield 0 0xEE00 0 1 2
2057 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
2062 Set single bit value of a port register::
2064 testpmd> write regbit (port_id) (address) (bit_x) (value)
2066 For example, to set the high bit in the register from the example above::
2068 testpmd> write regbit 0 0xEE00 31 1
2069 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
2071 Traffic Metering and Policing
2072 -----------------------------
2074 The following section shows functions for configuring traffic metering and
2075 policing on the ethernet device through the use of generic ethdev API.
2077 show port traffic management capability
2078 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2080 Show traffic metering and policing capability of the port::
2082 testpmd> show port meter cap (port_id)
2084 add port meter profile (srTCM rfc2967)
2085 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2087 Add meter profile (srTCM rfc2697) to the ethernet device::
2089 testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \
2094 * ``profile_id``: ID for the meter profile.
2095 * ``cir``: Committed Information Rate (CIR) (bytes/second).
2096 * ``cbs``: Committed Burst Size (CBS) (bytes).
2097 * ``ebs``: Excess Burst Size (EBS) (bytes).
2099 add port meter profile (trTCM rfc2968)
2100 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2102 Add meter profile (srTCM rfc2698) to the ethernet device::
2104 testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \
2105 (cir) (pir) (cbs) (pbs)
2109 * ``profile_id``: ID for the meter profile.
2110 * ``cir``: Committed information rate (bytes/second).
2111 * ``pir``: Peak information rate (bytes/second).
2112 * ``cbs``: Committed burst size (bytes).
2113 * ``pbs``: Peak burst size (bytes).
2115 add port meter profile (trTCM rfc4115)
2116 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2118 Add meter profile (trTCM rfc4115) to the ethernet device::
2120 testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \
2121 (cir) (eir) (cbs) (ebs)
2125 * ``profile_id``: ID for the meter profile.
2126 * ``cir``: Committed information rate (bytes/second).
2127 * ``eir``: Excess information rate (bytes/second).
2128 * ``cbs``: Committed burst size (bytes).
2129 * ``ebs``: Excess burst size (bytes).
2131 delete port meter profile
2132 ~~~~~~~~~~~~~~~~~~~~~~~~~
2134 Delete meter profile from the ethernet device::
2136 testpmd> del port meter profile (port_id) (profile_id)
2141 Create new meter object for the ethernet device::
2143 testpmd> create port meter (port_id) (mtr_id) (profile_id) \
2144 (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \
2145 (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\
2150 * ``mtr_id``: meter object ID.
2151 * ``profile_id``: ID for the meter profile.
2152 * ``meter_enable``: When this parameter has a non-zero value, the meter object
2153 gets enabled at the time of creation, otherwise remains disabled.
2154 * ``g_action``: Policer action for the packet with green color.
2155 * ``y_action``: Policer action for the packet with yellow color.
2156 * ``r_action``: Policer action for the packet with red color.
2157 * ``stats_mask``: Mask of statistics counter types to be enabled for the
2159 * ``shared``: When this parameter has a non-zero value, the meter object is
2160 shared by multiple flows. Otherwise, meter object is used by single flow.
2161 * ``use_pre_meter_color``: When this parameter has a non-zero value, the
2162 input color for the current meter object is determined by the latest meter
2163 object in the same flow. Otherwise, the current meter object uses the
2164 *dscp_table* to determine the input color.
2165 * ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input
2166 color, 0 <= x <= 63.
2171 Enable meter for the ethernet device::
2173 testpmd> enable port meter (port_id) (mtr_id)
2178 Disable meter for the ethernet device::
2180 testpmd> disable port meter (port_id) (mtr_id)
2185 Delete meter for the ethernet device::
2187 testpmd> del port meter (port_id) (mtr_id)
2189 Set port meter profile
2190 ~~~~~~~~~~~~~~~~~~~~~~
2192 Set meter profile for the ethernet device::
2194 testpmd> set port meter profile (port_id) (mtr_id) (profile_id)
2196 set port meter dscp table
2197 ~~~~~~~~~~~~~~~~~~~~~~~~~
2199 Set meter dscp table for the ethernet device::
2201 testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \
2202 (dscp_tbl_entry1)...(dscp_tbl_entry63)]
2204 set port meter policer action
2205 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2207 Set meter policer action for the ethernet device::
2209 testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \
2210 (action0) [(action1) (action1)]
2214 * ``action_mask``: Bit mask indicating which policer actions need to be
2215 updated. One or more policer actions can be updated in a single function
2216 invocation. To update the policer action associated with color C, bit
2217 (1 << C) needs to be set in *action_mask* and element at position C
2218 in the *actions* array needs to be valid.
2219 * ``actionx``: Policer action for the color x,
2220 RTE_MTR_GREEN <= x < RTE_MTR_COLORS
2222 set port meter stats mask
2223 ~~~~~~~~~~~~~~~~~~~~~~~~~
2225 Set meter stats mask for the ethernet device::
2227 testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask)
2231 * ``stats_mask``: Bit mask indicating statistics counter types to be enabled.
2233 show port meter stats
2234 ~~~~~~~~~~~~~~~~~~~~~
2236 Show meter stats of the ethernet device::
2238 testpmd> show port meter stats (port_id) (mtr_id) (clear)
2242 * ``clear``: Flag that indicates whether the statistics counters should
2243 be cleared (i.e. set to zero) immediately after they have been read or not.
2248 The following section shows functions for configuring traffic management on
2249 on the ethernet device through the use of generic TM API.
2251 show port traffic management capability
2252 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2254 Show traffic management capability of the port::
2256 testpmd> show port tm cap (port_id)
2258 show port traffic management capability (hierarchy level)
2259 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2261 Show traffic management hierarchy level capability of the port::
2263 testpmd> show port tm cap (port_id) (level_id)
2265 show port traffic management capability (hierarchy node level)
2266 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2268 Show the traffic management hierarchy node capability of the port::
2270 testpmd> show port tm cap (port_id) (node_id)
2272 show port traffic management hierarchy node type
2273 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2275 Show the port traffic management hierarchy node type::
2277 testpmd> show port tm node type (port_id) (node_id)
2279 show port traffic management hierarchy node stats
2280 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2282 Show the port traffic management hierarchy node statistics::
2284 testpmd> show port tm node stats (port_id) (node_id) (clear)
2288 * ``clear``: When this parameter has a non-zero value, the statistics counters
2289 are cleared (i.e. set to zero) immediately after they have been read,
2290 otherwise the statistics counters are left untouched.
2292 Add port traffic management private shaper profile
2293 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2295 Add the port traffic management private shaper profile::
2297 testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \
2298 (tb_rate) (tb_size) (packet_length_adjust)
2302 * ``shaper_profile id``: Shaper profile ID for the new profile.
2303 * ``tb_rate``: Token bucket rate (bytes per second).
2304 * ``tb_size``: Token bucket size (bytes).
2305 * ``packet_length_adjust``: The value (bytes) to be added to the length of
2306 each packet for the purpose of shaping. This parameter value can be used to
2307 correct the packet length with the framing overhead bytes that are consumed
2310 Delete port traffic management private shaper profile
2311 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2313 Delete the port traffic management private shaper::
2315 testpmd> del port tm node shaper profile (port_id) (shaper_profile_id)
2319 * ``shaper_profile id``: Shaper profile ID that needs to be deleted.
2321 Add port traffic management shared shaper
2322 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2324 Create the port traffic management shared shaper::
2326 testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \
2331 * ``shared_shaper_id``: Shared shaper ID to be created.
2332 * ``shaper_profile id``: Shaper profile ID for shared shaper.
2334 Set port traffic management shared shaper
2335 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2337 Update the port traffic management shared shaper::
2339 testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \
2344 * ``shared_shaper_id``: Shared shaper ID to be update.
2345 * ``shaper_profile id``: Shaper profile ID for shared shaper.
2347 Delete port traffic management shared shaper
2348 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2350 Delete the port traffic management shared shaper::
2352 testpmd> del port tm node shared shaper (port_id) (shared_shaper_id)
2356 * ``shared_shaper_id``: Shared shaper ID to be deleted.
2358 Set port traffic management hiearchy node private shaper
2359 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2361 set the port traffic management hierarchy node private shaper::
2363 testpmd> set port tm node shaper profile (port_id) (node_id) \
2368 * ``shaper_profile id``: Private shaper profile ID to be enabled on the
2371 Add port traffic management WRED profile
2372 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2374 Create a new WRED profile::
2376 testpmd> add port tm node wred profile (port_id) (wred_profile_id) \
2377 (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \
2378 (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \
2379 (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r)
2383 * ``wred_profile id``: Identifier for the newly create WRED profile
2384 * ``color_g``: Packet color (green)
2385 * ``min_th_g``: Minimum queue threshold for packet with green color
2386 * ``max_th_g``: Minimum queue threshold for packet with green color
2387 * ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp)
2388 * ``wq_log2_g``: Negated log2 of queue weight (wq)
2389 * ``color_y``: Packet color (yellow)
2390 * ``min_th_y``: Minimum queue threshold for packet with yellow color
2391 * ``max_th_y``: Minimum queue threshold for packet with yellow color
2392 * ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp)
2393 * ``wq_log2_y``: Negated log2 of queue weight (wq)
2394 * ``color_r``: Packet color (red)
2395 * ``min_th_r``: Minimum queue threshold for packet with yellow color
2396 * ``max_th_r``: Minimum queue threshold for packet with yellow color
2397 * ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp)
2398 * ``wq_log2_r``: Negated log2 of queue weight (wq)
2400 Delete port traffic management WRED profile
2401 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2403 Delete the WRED profile::
2405 testpmd> del port tm node wred profile (port_id) (wred_profile_id)
2407 Add port traffic management hierarchy nonleaf node
2408 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2410 Add nonleaf node to port traffic management hiearchy::
2412 testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \
2413 (priority) (weight) (level_id) (shaper_profile_id) \
2414 (n_sp_priorities) (stats_mask) (n_shared_shapers) \
2415 [(shared_shaper_0) (shared_shaper_1) ...] \
2419 * ``parent_node_id``: Node ID of the parent.
2420 * ``priority``: Node priority (highest node priority is zero). This is used by
2421 the SP algorithm running on the parent node for scheduling this node.
2422 * ``weight``: Node weight (lowest weight is one). The node weight is relative
2423 to the weight sum of all siblings that have the same priority. It is used by
2424 the WFQ algorithm running on the parent node for scheduling this node.
2425 * ``level_id``: Hiearchy level of the node.
2426 * ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2428 * ``n_sp_priorities``: Number of strict priorities.
2429 * ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2430 * ``n_shared_shapers``: Number of shared shapers.
2431 * ``shared_shaper_id``: Shared shaper id.
2433 Add port traffic management hierarchy leaf node
2434 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2436 Add leaf node to port traffic management hiearchy::
2438 testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \
2439 (priority) (weight) (level_id) (shaper_profile_id) \
2440 (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \
2441 [(shared_shaper_id) (shared_shaper_id) ...] \
2445 * ``parent_node_id``: Node ID of the parent.
2446 * ``priority``: Node priority (highest node priority is zero). This is used by
2447 the SP algorithm running on the parent node for scheduling this node.
2448 * ``weight``: Node weight (lowest weight is one). The node weight is relative
2449 to the weight sum of all siblings that have the same priority. It is used by
2450 the WFQ algorithm running on the parent node for scheduling this node.
2451 * ``level_id``: Hiearchy level of the node.
2452 * ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2454 * ``cman_mode``: Congestion management mode to be enabled for this node.
2455 * ``wred_profile_id``: WRED profile id to be enabled for this node.
2456 * ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2457 * ``n_shared_shapers``: Number of shared shapers.
2458 * ``shared_shaper_id``: Shared shaper id.
2460 Delete port traffic management hierarchy node
2461 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2463 Delete node from port traffic management hiearchy::
2465 testpmd> del port tm node (port_id) (node_id)
2467 Update port traffic management hierarchy parent node
2468 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2470 Update port traffic management hierarchy parent node::
2472 testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \
2475 This function can only be called after the hierarchy commit invocation. Its
2476 success depends on the port support for this operation, as advertised through
2477 the port capability set. This function is valid for all nodes of the traffic
2478 management hierarchy except root node.
2480 Commit port traffic management hierarchy
2481 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2483 Commit the traffic management hierarchy on the port::
2485 testpmd> port tm hierarchy commit (port_id) (clean_on_fail)
2489 * ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function
2490 call failure. On the other hand, hierarchy is preserved when this parameter
2493 Set port traffic management default hierarchy (tm forwarding mode)
2494 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2496 set the traffic management default hierarchy on the port::
2498 testpmd> set port tm hierarchy default (port_id)
2503 This section details the available filter functions that are available.
2505 Note these functions interface the deprecated legacy filtering framework,
2506 superseded by *rte_flow*. See `Flow rules management`_.
2509 ~~~~~~~~~~~~~~~~~~~~
2511 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
2513 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
2514 ethertype (ether_type) (drop|fwd) queue (queue_id)
2516 The available information parameters are:
2518 * ``port_id``: The port which the Ethertype filter assigned on.
2520 * ``mac_addr``: Compare destination mac address.
2522 * ``mac_ignr``: Ignore destination mac address match.
2524 * ``mac_address``: Destination mac address to match.
2526 * ``ether_type``: The EtherType value want to match,
2527 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
2529 * ``queue_id``: The receive queue associated with this EtherType filter.
2530 It is meaningless when deleting or dropping.
2532 Example, to add/remove an ethertype filter rule::
2534 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
2535 ethertype 0x0806 fwd queue 3
2537 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
2538 ethertype 0x0806 fwd queue 3
2543 Add or delete a 2-tuple filter,
2544 which identifies packets by specific protocol and destination TCP/UDP port
2545 and forwards packets into one of the receive queues::
2547 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
2548 protocol (protocol_value) mask (mask_value) \
2549 tcp_flags (tcp_flags_value) priority (prio_value) \
2552 The available information parameters are:
2554 * ``port_id``: The port which the 2-tuple filter assigned on.
2556 * ``dst_port_value``: Destination port in L4.
2558 * ``protocol_value``: IP L4 protocol.
2560 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
2562 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
2564 * ``prio_value``: Priority of this filter.
2566 * ``queue_id``: The receive queue associated with this 2-tuple filter.
2568 Example, to add/remove an 2tuple filter rule::
2570 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
2571 tcp_flags 0x02 priority 3 queue 3
2573 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
2574 tcp_flags 0x02 priority 3 queue 3
2579 Add or delete a 5-tuple filter,
2580 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
2581 and routes packets into one of the receive queues::
2583 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
2584 (src_address) dst_port (dst_port_value) \
2585 src_port (src_port_value) protocol (protocol_value) \
2586 mask (mask_value) tcp_flags (tcp_flags_value) \
2587 priority (prio_value) queue (queue_id)
2589 The available information parameters are:
2591 * ``port_id``: The port which the 5-tuple filter assigned on.
2593 * ``dst_address``: Destination IP address.
2595 * ``src_address``: Source IP address.
2597 * ``dst_port_value``: TCP/UDP destination port.
2599 * ``src_port_value``: TCP/UDP source port.
2601 * ``protocol_value``: L4 protocol.
2603 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
2605 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
2607 * ``prio_value``: The priority of this filter.
2609 * ``queue_id``: The receive queue associated with this 5-tuple filter.
2611 Example, to add/remove an 5tuple filter rule::
2613 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2614 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2615 flags 0x0 priority 3 queue 3
2617 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2618 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2619 flags 0x0 priority 3 queue 3
2624 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
2626 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
2628 The available information parameters are:
2630 * ``port_id``: The port which the SYN filter assigned on.
2632 * ``high``: This SYN filter has higher priority than other filters.
2634 * ``low``: This SYN filter has lower priority than other filters.
2636 * ``queue_id``: The receive queue associated with this SYN filter
2640 testpmd> syn_filter 0 add priority high queue 3
2645 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
2646 and routed into one of the receive queues::
2648 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
2649 mask (mask_value) priority (prio_value) queue (queue_id)
2651 The available information parameters are:
2653 * ``port_id``: The port which the Flex filter is assigned on.
2655 * ``len_value``: Filter length in bytes, no greater than 128.
2657 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
2659 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
2661 * ``prio_value``: The priority of this filter.
2663 * ``queue_id``: The receive queue associated with this Flex filter.
2667 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
2668 mask 000C priority 3 queue 3
2670 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
2671 mask 000C priority 3 queue 3
2674 .. _testpmd_flow_director:
2676 flow_director_filter
2677 ~~~~~~~~~~~~~~~~~~~~
2679 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
2681 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
2682 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
2684 * Perfect 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 IP flow.
2688 * Signature filters.
2689 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
2691 * Perfect-mac-vlan match filters.
2692 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2693 The masked fields are for MAC VLAN flow.
2695 * Perfect-tunnel match filters.
2696 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2697 The masked fields are for tunnel flow.
2699 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
2700 per flow type and the flexible payload.
2702 The Flow Director can also mask out parts of all of these fields so that filters
2703 are only applied to certain fields or parts of the fields.
2705 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
2707 # Commands to add flow director filters of different flow types::
2709 flow_director_filter (port_id) mode IP (add|del|update) \
2710 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
2711 src (src_ip_address) dst (dst_ip_address) \
2712 tos (tos_value) proto (proto_value) ttl (ttl_value) \
2713 vlan (vlan_value) flexbytes (flexbytes_value) \
2714 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
2717 flow_director_filter (port_id) mode IP (add|del|update) \
2718 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
2719 src (src_ip_address) (src_port) \
2720 dst (dst_ip_address) (dst_port) \
2721 tos (tos_value) ttl (ttl_value) \
2722 vlan (vlan_value) flexbytes (flexbytes_value) \
2723 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
2726 flow_director_filter (port_id) mode IP (add|del|update) \
2727 flow (ipv4-sctp|ipv6-sctp) \
2728 src (src_ip_address) (src_port) \
2729 dst (dst_ip_address) (dst_port) \
2730 tos (tos_value) ttl (ttl_value) \
2731 tag (verification_tag) vlan (vlan_value) \
2732 flexbytes (flexbytes_value) (drop|fwd) \
2733 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
2735 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
2736 ether (ethertype) flexbytes (flexbytes_value) \
2737 (drop|fwd) pf|vf(vf_id) queue (queue_id)
2740 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
2741 mac (mac_address) vlan (vlan_value) \
2742 flexbytes (flexbytes_value) (drop|fwd) \
2743 queue (queue_id) fd_id (fd_id_value)
2745 flow_director_filter (port_id) mode Tunnel (add|del|update) \
2746 mac (mac_address) vlan (vlan_value) \
2747 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
2748 flexbytes (flexbytes_value) (drop|fwd) \
2749 queue (queue_id) fd_id (fd_id_value)
2751 For example, to add an ipv4-udp flow type filter::
2753 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
2754 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
2755 fwd pf queue 1 fd_id 1
2757 For example, add an ipv4-other flow type filter::
2759 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
2760 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
2761 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
2766 Flush all flow director filters on a device::
2768 testpmd> flush_flow_director (port_id)
2770 Example, to flush all flow director filter on port 0::
2772 testpmd> flush_flow_director 0
2777 Set flow director's input masks::
2779 flow_director_mask (port_id) mode IP vlan (vlan_value) \
2780 src_mask (ipv4_src) (ipv6_src) (src_port) \
2781 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2783 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2785 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2786 mac (mac_value) tunnel-type (tunnel_type_value) \
2787 tunnel-id (tunnel_id_value)
2789 Example, to set flow director mask on port 0::
2791 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2792 src_mask 255.255.255.255 \
2793 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2794 dst_mask 255.255.255.255 \
2795 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2797 flow_director_flex_mask
2798 ~~~~~~~~~~~~~~~~~~~~~~~
2800 set masks of flow director's flexible payload based on certain flow type::
2802 testpmd> flow_director_flex_mask (port_id) \
2803 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2804 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2805 l2_payload|all) (mask)
2807 Example, to set flow director's flex mask for all flow type on port 0::
2809 testpmd> flow_director_flex_mask 0 flow all \
2810 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2813 flow_director_flex_payload
2814 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2816 Configure flexible payload selection::
2818 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2820 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2822 testpmd> flow_director_flex_payload 0 l4 \
2823 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2825 get_sym_hash_ena_per_port
2826 ~~~~~~~~~~~~~~~~~~~~~~~~~
2828 Get symmetric hash enable configuration per port::
2830 get_sym_hash_ena_per_port (port_id)
2832 For example, to get symmetric hash enable configuration of port 1::
2834 testpmd> get_sym_hash_ena_per_port 1
2836 set_sym_hash_ena_per_port
2837 ~~~~~~~~~~~~~~~~~~~~~~~~~
2839 Set symmetric hash enable configuration per port to enable or disable::
2841 set_sym_hash_ena_per_port (port_id) (enable|disable)
2843 For example, to set symmetric hash enable configuration of port 1 to enable::
2845 testpmd> set_sym_hash_ena_per_port 1 enable
2847 get_hash_global_config
2848 ~~~~~~~~~~~~~~~~~~~~~~
2850 Get the global configurations of hash filters::
2852 get_hash_global_config (port_id)
2854 For example, to get the global configurations of hash filters of port 1::
2856 testpmd> get_hash_global_config 1
2858 set_hash_global_config
2859 ~~~~~~~~~~~~~~~~~~~~~~
2861 Set the global configurations of hash filters::
2863 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2864 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2865 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload) \
2868 For example, to enable simple_xor for flow type of ipv6 on port 2::
2870 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2875 Set the input set for hash::
2877 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2878 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2879 l2_payload) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2880 ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2881 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2882 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2883 fld-8th|none) (select|add)
2885 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2887 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2892 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2893 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2895 Set the input set for flow director::
2897 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2898 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2899 l2_payload) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2900 ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2901 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2902 sctp-dst-port|sctp-veri-tag|none) (select|add)
2904 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2906 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2911 Set different GRE key length for input set::
2913 global_config (port_id) gre-key-len (number in bytes)
2915 For example to set GRE key length for input set to 4 bytes on port 0::
2917 testpmd> global_config 0 gre-key-len 4
2920 .. _testpmd_rte_flow:
2922 Flow rules management
2923 ---------------------
2925 Control of the generic flow API (*rte_flow*) is fully exposed through the
2926 ``flow`` command (validation, creation, destruction, queries and operation
2929 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2930 features simultaneously may cause undefined side-effects and is therefore
2936 Because the ``flow`` command uses dynamic tokens to handle the large number
2937 of possible flow rules combinations, its behavior differs slightly from
2938 other commands, in particular:
2940 - Pressing *?* or the *<tab>* key displays contextual help for the current
2941 token, not that of the entire command.
2943 - Optional and repeated parameters are supported (provided they are listed
2944 in the contextual help).
2946 The first parameter stands for the operation mode. Possible operations and
2947 their general syntax are described below. They are covered in detail in the
2950 - Check whether a flow rule can be created::
2952 flow validate {port_id}
2953 [group {group_id}] [priority {level}] [ingress] [egress]
2954 pattern {item} [/ {item} [...]] / end
2955 actions {action} [/ {action} [...]] / end
2957 - Create a flow rule::
2959 flow create {port_id}
2960 [group {group_id}] [priority {level}] [ingress] [egress]
2961 pattern {item} [/ {item} [...]] / end
2962 actions {action} [/ {action} [...]] / end
2964 - Destroy specific flow rules::
2966 flow destroy {port_id} rule {rule_id} [...]
2968 - Destroy all flow rules::
2970 flow flush {port_id}
2972 - Query an existing flow rule::
2974 flow query {port_id} {rule_id} {action}
2976 - List existing flow rules sorted by priority, filtered by group
2979 flow list {port_id} [group {group_id}] [...]
2981 - Restrict ingress traffic to the defined flow rules::
2983 flow isolate {port_id} {boolean}
2985 Validating flow rules
2986 ~~~~~~~~~~~~~~~~~~~~~
2988 ``flow validate`` reports whether a flow rule would be accepted by the
2989 underlying device in its current state but stops short of creating it. It is
2990 bound to ``rte_flow_validate()``::
2992 flow validate {port_id}
2993 [group {group_id}] [priority {level}] [ingress] [egress]
2994 pattern {item} [/ {item} [...]] / end
2995 actions {action} [/ {action} [...]] / end
2997 If successful, it will show::
3001 Otherwise it will show an error message of the form::
3003 Caught error type [...] ([...]): [...]
3005 This command uses the same parameters as ``flow create``, their format is
3006 described in `Creating flow rules`_.
3008 Check whether redirecting any Ethernet packet received on port 0 to RX queue
3009 index 6 is supported::
3011 testpmd> flow validate 0 ingress pattern eth / end
3012 actions queue index 6 / end
3016 Port 0 does not support TCPv6 rules::
3018 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
3020 Caught error type 9 (specific pattern item): Invalid argument
3026 ``flow create`` validates and creates the specified flow rule. It is bound
3027 to ``rte_flow_create()``::
3029 flow create {port_id}
3030 [group {group_id}] [priority {level}] [ingress] [egress]
3031 pattern {item} [/ {item} [...]] / end
3032 actions {action} [/ {action} [...]] / end
3034 If successful, it will return a flow rule ID usable with other commands::
3036 Flow rule #[...] created
3038 Otherwise it will show an error message of the form::
3040 Caught error type [...] ([...]): [...]
3042 Parameters describe in the following order:
3044 - Attributes (*group*, *priority*, *ingress*, *egress* tokens).
3045 - A matching pattern, starting with the *pattern* token and terminated by an
3047 - Actions, starting with the *actions* token and terminated by an *end*
3050 These translate directly to *rte_flow* objects provided as-is to the
3051 underlying functions.
3053 The shortest valid definition only comprises mandatory tokens::
3055 testpmd> flow create 0 pattern end actions end
3057 Note that PMDs may refuse rules that essentially do nothing such as this
3060 **All unspecified object values are automatically initialized to 0.**
3065 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
3066 specified before the ``pattern`` token.
3068 - ``group {group id}``: priority group.
3069 - ``priority {level}``: priority level within group.
3070 - ``ingress``: rule applies to ingress traffic.
3071 - ``egress``: rule applies to egress traffic.
3073 Each instance of an attribute specified several times overrides the previous
3074 value as shown below (group 4 is used)::
3076 testpmd> flow create 0 group 42 group 24 group 4 [...]
3078 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
3080 While not specifying a direction is an error, some rules may allow both
3083 Most rules affect RX therefore contain the ``ingress`` token::
3085 testpmd> flow create 0 ingress pattern [...]
3090 A matching pattern starts after the ``pattern`` token. It is made of pattern
3091 items and is terminated by a mandatory ``end`` item.
3093 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
3094 rte_flow_item_type``).
3096 The ``/`` token is used as a separator between pattern items as shown
3099 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
3101 Note that protocol items like these must be stacked from lowest to highest
3102 layer to make sense. For instance, the following rule is either invalid or
3103 unlikely to match any packet::
3105 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
3107 More information on these restrictions can be found in the *rte_flow*
3110 Several items support additional specification structures, for example
3111 ``ipv4`` allows specifying source and destination addresses as follows::
3113 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
3114 dst is 10.2.0.0 / end [...]
3116 This rule matches all IPv4 traffic with the specified properties.
3118 In this example, ``src`` and ``dst`` are field names of the underlying
3119 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
3120 in a similar fashion.
3122 The ``is`` token means that the subsequent value must be matched exactly,
3123 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
3124 accordingly. Possible assignment tokens are:
3126 - ``is``: match value perfectly (with full bit-mask).
3127 - ``spec``: match value according to configured bit-mask.
3128 - ``last``: specify upper bound to establish a range.
3129 - ``mask``: specify bit-mask with relevant bits set to one.
3130 - ``prefix``: generate bit-mask from a prefix length.
3132 These yield identical results::
3134 ipv4 src is 10.1.1.1
3138 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
3142 ipv4 src spec 10.1.1.1 src prefix 32
3146 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
3150 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
3152 Inclusive ranges can be defined with ``last``::
3154 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
3156 Note that ``mask`` affects both ``spec`` and ``last``::
3158 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
3159 # matches 10.1.0.0 to 10.2.255.255
3161 Properties can be modified multiple times::
3163 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
3167 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
3172 This section lists supported pattern items and their attributes, if any.
3174 - ``end``: end list of pattern items.
3176 - ``void``: no-op pattern item.
3178 - ``invert``: perform actions when pattern does not match.
3180 - ``any``: match any protocol for the current layer.
3182 - ``num {unsigned}``: number of layers covered.
3184 - ``pf``: match packets addressed to the physical function.
3186 - ``vf``: match packets addressed to a virtual function ID.
3188 - ``id {unsigned}``: destination VF ID.
3190 - ``port``: device-specific physical port index to use.
3192 - ``index {unsigned}``: physical port index.
3194 - ``raw``: match an arbitrary byte string.
3196 - ``relative {boolean}``: look for pattern after the previous item.
3197 - ``search {boolean}``: search pattern from offset (see also limit).
3198 - ``offset {integer}``: absolute or relative offset for pattern.
3199 - ``limit {unsigned}``: search area limit for start of pattern.
3200 - ``pattern {string}``: byte string to look for.
3202 - ``eth``: match Ethernet header.
3204 - ``dst {MAC-48}``: destination MAC.
3205 - ``src {MAC-48}``: source MAC.
3206 - ``type {unsigned}``: EtherType.
3208 - ``vlan``: match 802.1Q/ad VLAN tag.
3210 - ``tpid {unsigned}``: tag protocol identifier.
3211 - ``tci {unsigned}``: tag control information.
3212 - ``pcp {unsigned}``: priority code point.
3213 - ``dei {unsigned}``: drop eligible indicator.
3214 - ``vid {unsigned}``: VLAN identifier.
3216 - ``ipv4``: match IPv4 header.
3218 - ``tos {unsigned}``: type of service.
3219 - ``ttl {unsigned}``: time to live.
3220 - ``proto {unsigned}``: next protocol ID.
3221 - ``src {ipv4 address}``: source address.
3222 - ``dst {ipv4 address}``: destination address.
3224 - ``ipv6``: match IPv6 header.
3226 - ``tc {unsigned}``: traffic class.
3227 - ``flow {unsigned}``: flow label.
3228 - ``proto {unsigned}``: protocol (next header).
3229 - ``hop {unsigned}``: hop limit.
3230 - ``src {ipv6 address}``: source address.
3231 - ``dst {ipv6 address}``: destination address.
3233 - ``icmp``: match ICMP header.
3235 - ``type {unsigned}``: ICMP packet type.
3236 - ``code {unsigned}``: ICMP packet code.
3238 - ``udp``: match UDP header.
3240 - ``src {unsigned}``: UDP source port.
3241 - ``dst {unsigned}``: UDP destination port.
3243 - ``tcp``: match TCP header.
3245 - ``src {unsigned}``: TCP source port.
3246 - ``dst {unsigned}``: TCP destination port.
3248 - ``sctp``: match SCTP header.
3250 - ``src {unsigned}``: SCTP source port.
3251 - ``dst {unsigned}``: SCTP destination port.
3252 - ``tag {unsigned}``: validation tag.
3253 - ``cksum {unsigned}``: checksum.
3255 - ``vxlan``: match VXLAN header.
3257 - ``vni {unsigned}``: VXLAN identifier.
3259 - ``e_tag``: match IEEE 802.1BR E-Tag header.
3261 - ``grp_ecid_b {unsigned}``: GRP and E-CID base.
3263 - ``nvgre``: match NVGRE header.
3265 - ``tni {unsigned}``: virtual subnet ID.
3267 - ``mpls``: match MPLS header.
3269 - ``label {unsigned}``: MPLS label.
3271 - ``gre``: match GRE header.
3273 - ``protocol {unsigned}``: protocol type.
3275 - ``fuzzy``: fuzzy pattern match, expect faster than default.
3277 - ``thresh {unsigned}``: accuracy threshold.
3279 - ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header.
3281 - ``teid {unsigned}``: tunnel endpoint identifier.
3286 A list of actions starts after the ``actions`` token in the same fashion as
3287 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
3288 terminated by a mandatory ``end`` action.
3290 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
3291 rte_flow_action_type``).
3293 Dropping all incoming UDPv4 packets can be expressed as follows::
3295 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3298 Several actions have configurable properties which must be specified when
3299 there is no valid default value. For example, ``queue`` requires a target
3302 This rule redirects incoming UDPv4 traffic to queue index 6::
3304 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3305 actions queue index 6 / end
3307 While this one could be rejected by PMDs (unspecified queue index)::
3309 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3312 As defined by *rte_flow*, the list is not ordered, all actions of a given
3313 rule are performed simultaneously. These are equivalent::
3315 queue index 6 / void / mark id 42 / end
3319 void / mark id 42 / queue index 6 / end
3321 All actions in a list should have different types, otherwise only the last
3322 action of a given type is taken into account::
3324 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
3328 drop / drop / drop / end # drop is performed only once
3332 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
3334 Considering they are performed simultaneously, opposite and overlapping
3335 actions can sometimes be combined when the end result is unambiguous::
3337 drop / queue index 6 / end # drop has no effect
3341 drop / dup index 6 / end # same as above
3345 queue index 6 / rss queues 6 7 8 / end # queue has no effect
3349 drop / passthru / end # drop has no effect
3351 Note that PMDs may still refuse such combinations.
3356 This section lists supported actions and their attributes, if any.
3358 - ``end``: end list of actions.
3360 - ``void``: no-op action.
3362 - ``passthru``: let subsequent rule process matched packets.
3364 - ``mark``: attach 32 bit value to packets.
3366 - ``id {unsigned}``: 32 bit value to return with packets.
3368 - ``flag``: flag packets.
3370 - ``queue``: assign packets to a given queue index.
3372 - ``index {unsigned}``: queue index to use.
3374 - ``drop``: drop packets (note: passthru has priority).
3376 - ``count``: enable counters for this rule.
3378 - ``dup``: duplicate packets to a given queue index.
3380 - ``index {unsigned}``: queue index to duplicate packets to.
3382 - ``rss``: spread packets among several queues.
3384 - ``queues [{unsigned} [...]] end``: queue indices to use.
3386 - ``pf``: redirect packets to physical device function.
3388 - ``vf``: redirect packets to virtual device function.
3390 - ``original {boolean}``: use original VF ID if possible.
3391 - ``id {unsigned}``: VF ID to redirect packets to.
3393 Destroying flow rules
3394 ~~~~~~~~~~~~~~~~~~~~~
3396 ``flow destroy`` destroys one or more rules from their rule ID (as returned
3397 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
3398 times as necessary::
3400 flow destroy {port_id} rule {rule_id} [...]
3402 If successful, it will show::
3404 Flow rule #[...] destroyed
3406 It does not report anything for rule IDs that do not exist. The usual error
3407 message is shown when a rule cannot be destroyed::
3409 Caught error type [...] ([...]): [...]
3411 ``flow flush`` destroys all rules on a device and does not take extra
3412 arguments. It is bound to ``rte_flow_flush()``::
3414 flow flush {port_id}
3416 Any errors are reported as above.
3418 Creating several rules and destroying them::
3420 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3421 actions queue index 2 / end
3422 Flow rule #0 created
3423 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3424 actions queue index 3 / end
3425 Flow rule #1 created
3426 testpmd> flow destroy 0 rule 0 rule 1
3427 Flow rule #1 destroyed
3428 Flow rule #0 destroyed
3431 The same result can be achieved using ``flow flush``::
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 flush 0
3442 Non-existent rule IDs are ignored::
3444 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3445 actions queue index 2 / end
3446 Flow rule #0 created
3447 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3448 actions queue index 3 / end
3449 Flow rule #1 created
3450 testpmd> flow destroy 0 rule 42 rule 10 rule 2
3452 testpmd> flow destroy 0 rule 0
3453 Flow rule #0 destroyed
3459 ``flow query`` queries a specific action of a flow rule having that
3460 ability. Such actions collect information that can be reported using this
3461 command. It is bound to ``rte_flow_query()``::
3463 flow query {port_id} {rule_id} {action}
3465 If successful, it will display either the retrieved data for known actions
3466 or the following message::
3468 Cannot display result for action type [...] ([...])
3470 Otherwise, it will complain either that the rule does not exist or that some
3473 Flow rule #[...] not found
3477 Caught error type [...] ([...]): [...]
3479 Currently only the ``count`` action is supported. This action reports the
3480 number of packets that hit the flow rule and the total number of bytes. Its
3481 output has the following format::
3484 hits_set: [...] # whether "hits" contains a valid value
3485 bytes_set: [...] # whether "bytes" contains a valid value
3486 hits: [...] # number of packets
3487 bytes: [...] # number of bytes
3489 Querying counters for TCPv6 packets redirected to queue 6::
3491 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
3492 actions queue index 6 / count / end
3493 Flow rule #4 created
3494 testpmd> flow query 0 4 count
3505 ``flow list`` lists existing flow rules sorted by priority and optionally
3506 filtered by group identifiers::
3508 flow list {port_id} [group {group_id}] [...]
3510 This command only fails with the following message if the device does not
3515 Output consists of a header line followed by a short description of each
3516 flow rule, one per line. There is no output at all when no flow rules are
3517 configured on the device::
3519 ID Group Prio Attr Rule
3520 [...] [...] [...] [...] [...]
3522 ``Attr`` column flags:
3524 - ``i`` for ``ingress``.
3525 - ``e`` for ``egress``.
3527 Creating several flow rules and listing them::
3529 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3530 actions queue index 6 / end
3531 Flow rule #0 created
3532 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3533 actions queue index 2 / end
3534 Flow rule #1 created
3535 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
3536 actions rss queues 6 7 8 end / end
3537 Flow rule #2 created
3538 testpmd> flow list 0
3539 ID Group Prio Attr Rule
3540 0 0 0 i- ETH IPV4 => QUEUE
3541 1 0 0 i- ETH IPV6 => QUEUE
3542 2 0 5 i- ETH IPV4 UDP => RSS
3545 Rules are sorted by priority (i.e. group ID first, then priority level)::
3547 testpmd> flow list 1
3548 ID Group Prio Attr Rule
3549 0 0 0 i- ETH => COUNT
3550 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3551 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3552 1 24 0 i- ETH IPV4 UDP => QUEUE
3553 4 24 10 i- ETH IPV4 TCP => DROP
3554 3 24 20 i- ETH IPV4 => DROP
3555 2 24 42 i- ETH IPV4 UDP => QUEUE
3556 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3559 Output can be limited to specific groups::
3561 testpmd> flow list 1 group 0 group 63
3562 ID Group Prio Attr Rule
3563 0 0 0 i- ETH => COUNT
3564 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3565 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3566 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3569 Toggling isolated mode
3570 ~~~~~~~~~~~~~~~~~~~~~~
3572 ``flow isolate`` can be used to tell the underlying PMD that ingress traffic
3573 must only be injected from the defined flow rules; that no default traffic
3574 is expected outside those rules and the driver is free to assign more
3575 resources to handle them. It is bound to ``rte_flow_isolate()``::
3577 flow isolate {port_id} {boolean}
3579 If successful, enabling or disabling isolated mode shows either::
3581 Ingress traffic on port [...]
3582 is now restricted to the defined flow rules
3586 Ingress traffic on port [...]
3587 is not restricted anymore to the defined flow rules
3589 Otherwise, in case of error::
3591 Caught error type [...] ([...]): [...]
3593 Mainly due to its side effects, PMDs supporting this mode may not have the
3594 ability to toggle it more than once without reinitializing affected ports
3595 first (e.g. by exiting testpmd).
3597 Enabling isolated mode::
3599 testpmd> flow isolate 0 true
3600 Ingress traffic on port 0 is now restricted to the defined flow rules
3603 Disabling isolated mode::
3605 testpmd> flow isolate 0 false
3606 Ingress traffic on port 0 is not restricted anymore to the defined flow rules
3609 Sample QinQ flow rules
3610 ~~~~~~~~~~~~~~~~~~~~~~
3612 Before creating QinQ rule(s) the following commands should be issued to enable QinQ::
3614 testpmd> port stop 0
3615 testpmd> vlan set qinq on 0
3617 The above command sets the inner and outer TPID's to 0x8100.
3619 To change the TPID's the following commands should be used::
3621 testpmd> vlan set outer tpid 0xa100 0
3622 testpmd> vlan set inner tpid 0x9100 0
3623 testpmd> port start 0
3625 Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM.
3629 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 /
3630 vlan tci is 456 / end actions vf id 1 / queue index 0 / end
3631 Flow rule #0 validated
3633 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 /
3634 vlan tci is 456 / end actions vf id 123 / queue index 0 / end
3635 Flow rule #0 created
3637 testpmd> flow list 0
3638 ID Group Prio Attr Rule
3639 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3641 Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host.
3645 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 /
3646 vlan tci is 654 / end actions pf / queue index 0 / end
3647 Flow rule #1 validated
3649 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 /
3650 vlan tci is 654 / end actions pf / queue index 1 / end
3651 Flow rule #1 created
3653 testpmd> flow list 0
3654 ID Group Prio Attr Rule
3655 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3656 1 0 0 i- ETH VLAN VLAN=>PF QUEUE