1 .. SPDX-License-Identifier: BSD-3-Clause
2 Copyright(c) 2010-2016 Intel Corporation.
6 Testpmd Runtime Functions
7 =========================
9 Where the testpmd application is started in interactive mode, (``-i|--interactive``),
10 it displays a prompt that can be used to start and stop forwarding,
11 configure the application, display statistics (including the extended NIC
12 statistics aka xstats) , set the Flow Director and other tasks::
16 The testpmd prompt has some, limited, readline support.
17 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
18 as well as access to the command history via the up-arrow.
20 There is also support for tab completion.
21 If you type a partial command and hit ``<TAB>`` you get a list of the available completions:
23 .. code-block:: console
25 testpmd> show port <TAB>
27 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
28 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
29 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
30 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
36 Some examples in this document are too long to fit on one line are are shown wrapped at `"\\"` for display purposes::
38 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
39 (pause_time) (send_xon) (port_id)
41 In the real ``testpmd>`` prompt these commands should be on a single line.
46 The testpmd has on-line help for the functions that are available at runtime.
47 These are divided into sections and can be accessed using help, help section or help all:
49 .. code-block:: console
53 help control : Start and stop forwarding.
54 help display : Displaying port, stats and config information.
55 help config : Configuration information.
56 help ports : Configuring ports.
57 help registers : Reading and setting port registers.
58 help filters : Filters configuration help.
59 help all : All of the above sections.
62 Command File Functions
63 ----------------------
65 To facilitate loading large number of commands or to avoid cutting and pasting where not
66 practical or possible testpmd supports alternative methods for executing commands.
68 * If started with the ``--cmdline-file=FILENAME`` command line argument testpmd
69 will execute all CLI commands contained within the file immediately before
70 starting packet forwarding or entering interactive mode.
72 .. code-block:: console
74 ./testpmd -n4 -r2 ... -- -i --cmdline-file=/home/ubuntu/flow-create-commands.txt
75 Interactive-mode selected
76 CLI commands to be read from /home/ubuntu/flow-create-commands.txt
77 Configuring Port 0 (socket 0)
78 Port 0: 7C:FE:90:CB:74:CE
79 Configuring Port 1 (socket 0)
80 Port 1: 7C:FE:90:CB:74:CA
81 Checking link statuses...
82 Port 0 Link Up - speed 10000 Mbps - full-duplex
83 Port 1 Link Up - speed 10000 Mbps - full-duplex
89 Flow rule #498 created
90 Flow rule #499 created
91 Read all CLI commands from /home/ubuntu/flow-create-commands.txt
95 * At run-time additional commands can be loaded in bulk by invoking the ``load FILENAME``
98 .. code-block:: console
100 testpmd> load /home/ubuntu/flow-create-commands.txt
105 Flow rule #498 created
106 Flow rule #499 created
107 Read all CLI commands from /home/ubuntu/flow-create-commands.txt
111 In all cases output from any included command will be displayed as standard output.
112 Execution will continue until the end of the file is reached regardless of
113 whether any errors occur. The end user must examine the output to determine if
114 any failures occurred.
123 Start packet forwarding with current configuration::
130 Start packet forwarding with current configuration after sending specified number of bursts of packets::
132 testpmd> start tx_first (""|burst_num)
134 The default burst number is 1 when ``burst_num`` not presented.
139 Stop packet forwarding, and display accumulated statistics::
154 The functions in the following sections are used to display information about the
155 testpmd configuration or the NIC status.
160 Display information for a given port or all ports::
162 testpmd> show port (info|summary|stats|xstats|fdir|stat_qmap|dcb_tc|cap) (port_id|all)
164 The available information categories are:
166 * ``info``: General port information such as MAC address.
168 * ``summary``: Brief port summary such as Device Name, Driver Name etc.
170 * ``stats``: RX/TX statistics.
172 * ``xstats``: RX/TX extended NIC statistics.
174 * ``fdir``: Flow Director information and statistics.
176 * ``stat_qmap``: Queue statistics mapping.
178 * ``dcb_tc``: DCB information such as TC mapping.
180 * ``cap``: Supported offload capabilities.
184 .. code-block:: console
186 testpmd> show port info 0
188 ********************* Infos for port 0 *********************
190 MAC address: XX:XX:XX:XX:XX:XX
192 memory allocation on the socket: 0
194 Link speed: 40000 Mbps
195 Link duplex: full-duplex
196 Promiscuous mode: enabled
197 Allmulticast mode: disabled
198 Maximum number of MAC addresses: 64
199 Maximum number of MAC addresses of hash filtering: 0
204 Redirection table size: 512
205 Supported flow types:
225 Display the rss redirection table entry indicated by masks on port X::
227 testpmd> show port (port_id) rss reta (size) (mask0, mask1...)
229 size is used to indicate the hardware supported reta size
234 Display the RSS hash functions and RSS hash key of a port::
236 testpmd> show port (port_id) rss-hash [key]
241 Clear the port statistics for a given port or for all ports::
243 testpmd> clear port (info|stats|xstats|fdir|stat_qmap) (port_id|all)
247 testpmd> clear port stats all
252 Display information for a given port's RX/TX queue::
254 testpmd> show (rxq|txq) info (port_id) (queue_id)
259 Displays the configuration of the application.
260 The configuration comes from the command-line, the runtime or the application defaults::
262 testpmd> show config (rxtx|cores|fwd|txpkts)
264 The available information categories are:
266 * ``rxtx``: RX/TX configuration items.
268 * ``cores``: List of forwarding cores.
270 * ``fwd``: Packet forwarding configuration.
272 * ``txpkts``: Packets to TX configuration.
276 .. code-block:: console
278 testpmd> show config rxtx
280 io packet forwarding - CRC stripping disabled - packets/burst=16
281 nb forwarding cores=2 - nb forwarding ports=1
282 RX queues=1 - RX desc=128 - RX free threshold=0
283 RX threshold registers: pthresh=8 hthresh=8 wthresh=4
284 TX queues=1 - TX desc=512 - TX free threshold=0
285 TX threshold registers: pthresh=36 hthresh=0 wthresh=0
286 TX RS bit threshold=0 - TXQ flags=0x0
291 Set the packet forwarding mode::
293 testpmd> set fwd (io|mac|macswap|flowgen| \
294 rxonly|txonly|csum|icmpecho|noisy) (""|retry)
296 ``retry`` can be specified for forwarding engines except ``rx_only``.
298 The available information categories are:
300 * ``io``: Forwards packets "as-is" in I/O mode.
301 This is the fastest possible forwarding operation as it does not access packets data.
302 This is the default mode.
304 * ``mac``: Changes the source and the destination Ethernet addresses of packets before forwarding them.
305 Default application behaviour is to set source Ethernet address to that of the transmitting interface, and destination
306 address to a dummy value (set during init). The user may specify a target destination Ethernet address via the 'eth-peer' or
307 'eth-peer-configfile' command-line options. It is not currently possible to specify a specific source Ethernet address.
309 * ``macswap``: MAC swap forwarding mode.
310 Swaps the source and the destination Ethernet addresses of packets before forwarding them.
312 * ``flowgen``: Multi-flow generation mode.
313 Originates a number of flows (with varying destination IP addresses), and terminate receive traffic.
315 * ``rxonly``: Receives packets but doesn't transmit them.
317 * ``txonly``: Generates and transmits packets without receiving any.
319 * ``csum``: Changes the checksum field with hardware or software methods depending on the offload flags on the packet.
321 * ``icmpecho``: Receives a burst of packets, lookup for IMCP echo requests and, if any, send back ICMP echo replies.
323 * ``ieee1588``: Demonstrate L2 IEEE1588 V2 PTP timestamping for RX and TX. Requires ``CONFIG_RTE_LIBRTE_IEEE1588=y``.
325 * ``softnic``: Demonstrates the softnic forwarding operation. In this mode, packet forwarding is
326 similar to I/O mode except for the fact that packets are loopback to the softnic ports only. Therefore, portmask parameter should be set to softnic port only. The various software based custom NIC pipelines specified through the softnic firmware (DPDK packet framework script) can be tested in this mode. Furthermore, it allows to build 5-level hierarchical QoS scheduler as a default option that can be enabled through CLI once testpmd application is initialised. The user can modify the default scheduler hierarchy or can specify the new QoS Scheduler hierarchy through CLI. Requires ``CONFIG_RTE_LIBRTE_PMD_SOFTNIC=y``.
328 * ``noisy``: Noisy neighbour simulation.
329 Simulate more realistic behavior of a guest machine engaged in receiving
330 and sending packets performing Virtual Network Function (VNF).
334 testpmd> set fwd rxonly
336 Set rxonly packet forwarding mode
342 Display an RX descriptor for a port RX queue::
344 testpmd> read rxd (port_id) (queue_id) (rxd_id)
348 testpmd> read rxd 0 0 4
349 0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180
354 Display a TX descriptor for a port TX queue::
356 testpmd> read txd (port_id) (queue_id) (txd_id)
360 testpmd> read txd 0 0 4
361 0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C
366 Get loaded dynamic device personalization (DDP) package info list::
368 testpmd> ddp get list (port_id)
373 Display information about dynamic device personalization (DDP) profile::
375 testpmd> ddp get info (profile_path)
380 Display VF statistics::
382 testpmd> show vf stats (port_id) (vf_id)
387 Reset VF statistics::
389 testpmd> clear vf stats (port_id) (vf_id)
391 show port pctype mapping
392 ~~~~~~~~~~~~~~~~~~~~~~~~
394 List all items from the pctype mapping table::
396 testpmd> show port (port_id) pctype mapping
398 show rx offloading capabilities
399 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
401 List all per queue and per port Rx offloading capabilities of a port::
403 testpmd> show port (port_id) rx_offload capabilities
405 show rx offloading configuration
406 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
408 List port level and all queue level Rx offloading configuration::
410 testpmd> show port (port_id) rx_offload configuration
412 show tx offloading capabilities
413 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
415 List all per queue and per port Tx offloading capabilities of a port::
417 testpmd> show port (port_id) tx_offload capabilities
419 show tx offloading configuration
420 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
422 List port level and all queue level Tx offloading configuration::
424 testpmd> show port (port_id) tx_offload configuration
426 show tx metadata setting
427 ~~~~~~~~~~~~~~~~~~~~~~~~
429 Show Tx metadata value set for a specific port::
431 testpmd> show port (port_id) tx_metadata
433 Configuration Functions
434 -----------------------
436 The testpmd application can be configured from the runtime as well as from the command-line.
438 This section details the available configuration functions that are available.
442 Configuration changes only become active when forwarding is started/restarted.
447 Reset forwarding to the default configuration::
454 Set the debug verbosity level::
456 testpmd> set verbose (level)
458 Available levels are as following:
460 * ``0`` silent except for error.
461 * ``1`` fully verbose except for Tx packets.
462 * ``2`` fully verbose except for Rx packets.
463 * ``> 2`` fully verbose.
468 Set the log level for a log type::
470 testpmd> set log global|(type) (level)
474 * ``type`` is the log name.
476 * ``level`` is the log level.
478 For example, to change the global log level::
479 testpmd> set log global (level)
481 Regexes can also be used for type. To change log level of user1, user2 and user3::
482 testpmd> set log user[1-3] (level)
487 Set the number of ports used by the application:
491 This is equivalent to the ``--nb-ports`` command-line option.
496 Set the number of cores used by the application::
498 testpmd> set nbcore (num)
500 This is equivalent to the ``--nb-cores`` command-line option.
504 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port.
509 Set the forwarding cores hexadecimal mask::
511 testpmd> set coremask (mask)
513 This is equivalent to the ``--coremask`` command-line option.
517 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding.
522 Set the forwarding ports hexadecimal mask::
524 testpmd> set portmask (mask)
526 This is equivalent to the ``--portmask`` command-line option.
531 Set number of packets per burst::
533 testpmd> set burst (num)
535 This is equivalent to the ``--burst command-line`` option.
537 When retry is enabled, the transmit delay time and number of retries can also be set::
539 testpmd> set burst tx delay (microseconds) retry (num)
544 Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode::
546 testpmd> set txpkts (x[,y]*)
548 Where x[,y]* represents a CSV list of values, without white space.
553 Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes::
555 testpmd> set txsplit (off|on|rand)
559 * ``off`` disable packet copy & split for CSUM mode.
561 * ``on`` split outgoing packet into multiple segments. Size of each segment
562 and number of segments per packet is determined by ``set txpkts`` command
565 * ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments.
570 Set the list of forwarding cores::
572 testpmd> set corelist (x[,y]*)
574 For example, to change the forwarding cores:
576 .. code-block:: console
578 testpmd> set corelist 3,1
579 testpmd> show config fwd
581 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled
582 Logical Core 3 (socket 0) forwards packets on 1 streams:
583 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01
584 Logical Core 1 (socket 0) forwards packets on 1 streams:
585 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
589 The cores are used in the same order as specified on the command line.
594 Set the list of forwarding ports::
596 testpmd> set portlist (x[,y]*)
598 For example, to change the port forwarding:
600 .. code-block:: console
602 testpmd> set portlist 0,2,1,3
603 testpmd> show config fwd
605 io packet forwarding - ports=4 - cores=1 - streams=4
606 Logical Core 3 (socket 0) forwards packets on 4 streams:
607 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01
608 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
609 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03
610 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02
615 Enable/disable tx loopback::
617 testpmd> set tx loopback (port_id) (on|off)
622 set drop enable bit for all queues::
624 testpmd> set all queues drop (port_id) (on|off)
626 set split drop enable (for VF)
627 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
629 set split drop enable bit for VF from PF::
631 testpmd> set vf split drop (port_id) (vf_id) (on|off)
633 set mac antispoof (for VF)
634 ~~~~~~~~~~~~~~~~~~~~~~~~~~
636 Set mac antispoof for a VF from the PF::
638 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off)
643 Enable/disable MACsec offload::
645 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off)
646 testpmd> set macsec offload (port_id) off
651 Configure MACsec secure connection (SC)::
653 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi)
657 The pi argument is ignored for tx.
658 Check the NIC Datasheet for hardware limits.
663 Configure MACsec secure association (SA)::
665 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key)
669 The IDX value must be 0 or 1.
670 Check the NIC Datasheet for hardware limits.
672 set broadcast mode (for VF)
673 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
675 Set broadcast mode for a VF from the PF::
677 testpmd> set vf broadcast (port_id) (vf_id) (on|off)
682 Set the VLAN strip on a port::
684 testpmd> vlan set strip (on|off) (port_id)
689 Set the VLAN strip for a queue on a port::
691 testpmd> vlan set stripq (on|off) (port_id,queue_id)
693 vlan set stripq (for VF)
694 ~~~~~~~~~~~~~~~~~~~~~~~~
696 Set VLAN strip for all queues in a pool for a VF from the PF::
698 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off)
700 vlan set insert (for VF)
701 ~~~~~~~~~~~~~~~~~~~~~~~~
703 Set VLAN insert for a VF from the PF::
705 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id)
707 vlan set tag (for VF)
708 ~~~~~~~~~~~~~~~~~~~~~
710 Set VLAN tag for a VF from the PF::
712 testpmd> set vf vlan tag (port_id) (vf_id) (on|off)
714 vlan set antispoof (for VF)
715 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
717 Set VLAN antispoof for a VF from the PF::
719 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off)
724 Set the VLAN filter on a port::
726 testpmd> vlan set filter (on|off) (port_id)
731 Set the VLAN QinQ (extended queue in queue) on for a port::
733 testpmd> vlan set qinq (on|off) (port_id)
738 Set the inner or outer VLAN TPID for packet filtering on a port::
740 testpmd> vlan set (inner|outer) tpid (value) (port_id)
744 TPID value must be a 16-bit number (value <= 65536).
749 Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID::
751 testpmd> rx_vlan add (vlan_id|all) (port_id)
755 VLAN filter must be set on that port. VLAN ID < 4096.
756 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries
757 in VFTA table. This is important if enabling all vlan_ids.
762 Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID::
764 testpmd> rx_vlan rm (vlan_id|all) (port_id)
769 Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID::
771 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask)
776 Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID::
778 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask)
783 Add a tunnel filter on a port::
785 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \
786 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
787 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
789 The available information categories are:
791 * ``vxlan``: Set tunnel type as VXLAN.
793 * ``nvgre``: Set tunnel type as NVGRE.
795 * ``ipingre``: Set tunnel type as IP-in-GRE.
797 * ``imac-ivlan``: Set filter type as Inner MAC and VLAN.
799 * ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID.
801 * ``imac-tenid``: Set filter type as Inner MAC and tenant ID.
803 * ``imac``: Set filter type as Inner MAC.
805 * ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID.
807 * ``oip``: Set filter type as Outer IP.
809 * ``iip``: Set filter type as Inner IP.
813 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \
814 192.168.2.2 0 ipingre oip 1 1
816 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP.
821 Remove a tunnel filter on a port::
823 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \
824 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
825 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
830 Add an UDP port for VXLAN packet filter on a port::
832 testpmd> rx_vxlan_port add (udp_port) (port_id)
837 Remove an UDP port for VXLAN packet filter on a port::
839 testpmd> rx_vxlan_port rm (udp_port) (port_id)
844 Set hardware insertion of VLAN IDs in packets sent on a port::
846 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer]
848 For example, set a single VLAN ID (5) insertion on port 0::
852 Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1::
860 Set port based hardware insertion of VLAN ID in packets sent on a port::
862 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off)
867 Disable hardware insertion of a VLAN header in packets sent on a port::
869 testpmd> tx_vlan reset (port_id)
874 Select hardware or software calculation of the checksum when
875 transmitting a packet using the ``csum`` forwarding engine::
877 testpmd> csum set (ip|udp|tcp|sctp|outer-ip|outer-udp) (hw|sw) (port_id)
881 * ``ip|udp|tcp|sctp`` always relate to the inner layer.
883 * ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized
884 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are
885 supported). See also the ``csum parse-tunnel`` command.
887 * ``outer-udp`` relates to the outer UDP layer in the case where the packet is recognized
888 as a tunnel packet by the forwarding engine (vxlan, vxlan-gpe are
889 supported). See also the ``csum parse-tunnel`` command.
893 Check the NIC Datasheet for hardware limits.
898 Set RSS queue region span on a port::
900 testpmd> set port (port_id) queue-region region_id (value) \
901 queue_start_index (value) queue_num (value)
903 Set flowtype mapping on a RSS queue region on a port::
905 testpmd> set port (port_id) queue-region region_id (value) flowtype (value)
909 * For the flowtype(pctype) of packet,the specific index for each type has
910 been defined in file i40e_type.h as enum i40e_filter_pctype.
912 Set user priority mapping on a RSS queue region on a port::
914 testpmd> set port (port_id) queue-region UP (value) region_id (value)
916 Flush all queue region related configuration on a port::
918 testpmd> set port (port_id) queue-region flush (on|off)
922 * "on"is just an enable function which server for other configuration,
923 it is for all configuration about queue region from up layer,
924 at first will only keep in DPDK softwarestored in driver,
925 only after "flush on", it commit all configuration to HW.
926 "off" is just clean all configuration about queue region just now,
927 and restore all to DPDK i40e driver default config when start up.
929 Show all queue region related configuration info on a port::
931 testpmd> show port (port_id) queue-region
935 Queue region only support on PF by now, so these command is
936 only for configuration of queue region on PF port.
941 Define how tunneled packets should be handled by the csum forward
944 testpmd> csum parse-tunnel (on|off) (tx_port_id)
946 If enabled, the csum forward engine will try to recognize supported
947 tunnel headers (vxlan, gre, ipip).
949 If disabled, treat tunnel packets as non-tunneled packets (a inner
950 header is handled as a packet payload).
954 The port argument is the TX port like in the ``csum set`` command.
958 Consider a packet in packet like the following::
960 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in
962 * If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
963 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the
964 ``outer-ip|outer-udp`` parameter relates to the outer headers (here ``ipv4_out`` and ``udp_out``).
966 * If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
967 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``.
972 Display tx checksum offload configuration::
974 testpmd> csum show (port_id)
979 Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine::
981 testpmd> tso set (segsize) (port_id)
985 Check the NIC datasheet for hardware limits.
990 Display the status of TCP Segmentation Offload::
992 testpmd> tso show (port_id)
997 Enable or disable GRO in ``csum`` forwarding engine::
999 testpmd> set port <port_id> gro on|off
1001 If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4
1002 packets received from the given port.
1004 If disabled, packets received from the given port won't be performed
1005 GRO. By default, GRO is disabled for all ports.
1009 When enable GRO for a port, TCP/IPv4 packets received from the port
1010 will be performed GRO. After GRO, all merged packets have bad
1011 checksums, since the GRO library doesn't re-calculate checksums for
1012 the merged packets. Therefore, if users want the merged packets to
1013 have correct checksums, please select HW IP checksum calculation and
1014 HW TCP checksum calculation for the port which the merged packets are
1020 Display GRO configuration for a given port::
1022 testpmd> show port <port_id> gro
1027 Set the cycle to flush the GROed packets from reassembly tables::
1029 testpmd> set gro flush <cycles>
1031 When enable GRO, the csum forwarding engine performs GRO on received
1032 packets, and the GROed packets are stored in reassembly tables. Users
1033 can use this command to determine when the GROed packets are flushed
1034 from the reassembly tables.
1036 The ``cycles`` is measured in GRO operation times. The csum forwarding
1037 engine flushes the GROed packets from the tables every ``cycles`` GRO
1040 By default, the value of ``cycles`` is 1, which means flush GROed packets
1041 from the reassembly tables as soon as one GRO operation finishes. The value
1042 of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``.
1044 Please note that the large value of ``cycles`` may cause the poor TCP/IP
1045 stack performance. Because the GROed packets are delayed to arrive the
1046 stack, thus causing more duplicated ACKs and TCP retransmissions.
1051 Toggle per-port GSO support in ``csum`` forwarding engine::
1053 testpmd> set port <port_id> gso on|off
1055 If enabled, the csum forwarding engine will perform GSO on supported IPv4
1056 packets, transmitted on the given port.
1058 If disabled, packets transmitted on the given port will not undergo GSO.
1059 By default, GSO is disabled for all ports.
1063 When GSO is enabled on a port, supported IPv4 packets transmitted on that
1064 port undergo GSO. Afterwards, the segmented packets are represented by
1065 multi-segment mbufs; however, the csum forwarding engine doesn't calculation
1066 of checksums for GSO'd segments in SW. As a result, if users want correct
1067 checksums in GSO segments, they should enable HW checksum calculation for
1070 For example, HW checksum calculation for VxLAN GSO'd packets may be enabled
1071 by setting the following options in the csum forwarding engine:
1073 testpmd> csum set outer_ip hw <port_id>
1075 testpmd> csum set ip hw <port_id>
1077 testpmd> csum set tcp hw <port_id>
1079 UDP GSO is the same as IP fragmentation, which treats the UDP header
1080 as the payload and does not modify it during segmentation. That is,
1081 after UDP GSO, only the first output fragment has the original UDP
1082 header. Therefore, users need to enable HW IP checksum calculation
1083 and SW UDP checksum calculation for GSO-enabled ports, if they want
1084 correct checksums for UDP/IPv4 packets.
1089 Set the maximum GSO segment size (measured in bytes), which includes the
1090 packet header and the packet payload for GSO-enabled ports (global)::
1092 testpmd> set gso segsz <length>
1097 Display the status of Generic Segmentation Offload for a given port::
1099 testpmd> show port <port_id> gso
1104 Add an alternative MAC address to a port::
1106 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX)
1111 Remove a MAC address from a port::
1113 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX)
1115 mac_addr add (for VF)
1116 ~~~~~~~~~~~~~~~~~~~~~
1118 Add an alternative MAC address for a VF to a port::
1120 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX)
1125 Set the default MAC address for a port::
1127 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX)
1129 mac_addr set (for VF)
1130 ~~~~~~~~~~~~~~~~~~~~~
1132 Set the MAC address for a VF from the PF::
1134 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX)
1139 Set the forwarding peer address for certain port::
1141 testpmd> set eth-peer (port_id) (perr_addr)
1143 This is equivalent to the ``--eth-peer`` command-line option.
1148 Set the unicast hash filter(s) on/off for a port::
1150 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
1155 Set the promiscuous mode on for a port or for all ports.
1156 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1158 testpmd> set promisc (port_id|all) (on|off)
1163 Set the allmulti mode for a port or for all ports::
1165 testpmd> set allmulti (port_id|all) (on|off)
1167 Same as the ifconfig (8) option. Controls how multicast packets are handled.
1169 set promisc (for VF)
1170 ~~~~~~~~~~~~~~~~~~~~
1172 Set the unicast promiscuous mode for a VF from PF.
1173 It's supported by Intel i40e NICs now.
1174 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1176 testpmd> set vf promisc (port_id) (vf_id) (on|off)
1178 set allmulticast (for VF)
1179 ~~~~~~~~~~~~~~~~~~~~~~~~~
1181 Set the multicast promiscuous mode for a VF from PF.
1182 It's supported by Intel i40e NICs now.
1183 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1185 testpmd> set vf allmulti (port_id) (vf_id) (on|off)
1187 set tx max bandwidth (for VF)
1188 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1190 Set TX max absolute bandwidth (Mbps) for a VF from PF::
1192 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth)
1194 set tc tx min bandwidth (for VF)
1195 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1197 Set all TCs' TX min relative bandwidth (%) for a VF from PF::
1199 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...)
1201 set tc tx max bandwidth (for VF)
1202 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1204 Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF::
1206 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth)
1208 set tc strict link priority mode
1209 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1211 Set some TCs' strict link priority mode on a physical port::
1213 testpmd> set tx strict-link-priority (port_id) (tc_bitmap)
1215 set tc tx min bandwidth
1216 ~~~~~~~~~~~~~~~~~~~~~~~
1218 Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs::
1220 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...)
1225 Set the link flow control parameter on a port::
1227 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1228 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
1229 autoneg (on|off) (port_id)
1233 * ``high_water`` (integer): High threshold value to trigger XOFF.
1235 * ``low_water`` (integer): Low threshold value to trigger XON.
1237 * ``pause_time`` (integer): Pause quota in the Pause frame.
1239 * ``send_xon`` (0/1): Send XON frame.
1241 * ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
1243 * ``autoneg``: Change the auto-negotiation parameter.
1248 Set the priority flow control parameter on a port::
1250 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1251 (pause_time) (priority) (port_id)
1255 * ``high_water`` (integer): High threshold value.
1257 * ``low_water`` (integer): Low threshold value.
1259 * ``pause_time`` (integer): Pause quota in the Pause frame.
1261 * ``priority`` (0-7): VLAN User Priority.
1266 Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
1268 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
1270 For example, to set rx queue 2 on port 0 to mapping 5::
1272 testpmd>set stat_qmap rx 0 2 5
1274 set xstats-hide-zero
1275 ~~~~~~~~~~~~~~~~~~~~
1277 Set the option to hide zero values for xstats display::
1279 testpmd> set xstats-hide-zero on|off
1283 By default, the zero values are displayed for xstats.
1285 set port - rx/tx (for VF)
1286 ~~~~~~~~~~~~~~~~~~~~~~~~~
1288 Set VF receive/transmit from a port::
1290 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
1292 set port - mac address filter (for VF)
1293 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1295 Add/Remove unicast or multicast MAC addr filter for a VF::
1297 testpmd> set port (port_id) vf (vf_id) (mac_addr) \
1298 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
1300 set port - rx mode(for VF)
1301 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1303 Set the VF receive mode of a port::
1305 testpmd> set port (port_id) vf (vf_id) \
1306 rxmode (AUPE|ROPE|BAM|MPE) (on|off)
1308 The available receive modes are:
1310 * ``AUPE``: Accepts untagged VLAN.
1312 * ``ROPE``: Accepts unicast hash.
1314 * ``BAM``: Accepts broadcast packets.
1316 * ``MPE``: Accepts all multicast packets.
1318 set port - tx_rate (for Queue)
1319 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1321 Set TX rate limitation for a queue on a port::
1323 testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1325 set port - tx_rate (for VF)
1326 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
1328 Set TX rate limitation for queues in VF on a port::
1330 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1332 set port - mirror rule
1333 ~~~~~~~~~~~~~~~~~~~~~~
1335 Set pool or vlan type mirror rule for a port::
1337 testpmd> set port (port_id) mirror-rule (rule_id) \
1338 (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1339 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1341 Set link mirror rule for a port::
1343 testpmd> set port (port_id) mirror-rule (rule_id) \
1344 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1346 For example to enable mirror traffic with vlan 0,1 to pool 0::
1348 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1350 reset port - mirror rule
1351 ~~~~~~~~~~~~~~~~~~~~~~~~
1353 Reset a mirror rule for a port::
1355 testpmd> reset port (port_id) mirror-rule (rule_id)
1360 Set the flush on RX streams before forwarding.
1361 The default is flush ``on``.
1362 Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1364 testpmd> set flush_rx off
1369 Set the bypass mode for the lowest port on bypass enabled NIC::
1371 testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1376 Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1378 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1379 mode (normal|bypass|isolate) (port_id)
1383 * ``timeout``: Enable bypass after watchdog timeout.
1385 * ``os_on``: Enable bypass when OS/board is powered on.
1387 * ``os_off``: Enable bypass when OS/board is powered off.
1389 * ``power_on``: Enable bypass when power supply is turned on.
1391 * ``power_off``: Enable bypass when power supply is turned off.
1397 Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1399 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1404 Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1406 testpmd> show bypass config (port_id)
1411 Set link up for a port::
1413 testpmd> set link-up port (port id)
1418 Set link down for a port::
1420 testpmd> set link-down port (port id)
1425 Enable E-tag insertion for a VF on a port::
1427 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1429 Disable E-tag insertion for a VF on a port::
1431 testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1433 Enable/disable E-tag stripping on a port::
1435 testpmd> E-tag set stripping (on|off) port (port_id)
1437 Enable/disable E-tag based forwarding on a port::
1439 testpmd> E-tag set forwarding (on|off) port (port_id)
1441 Add an E-tag forwarding filter on a port::
1443 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1445 Delete an E-tag forwarding filter on a port::
1446 testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1451 Load a dynamic device personalization (DDP) profile and store backup profile::
1453 testpmd> ddp add (port_id) (profile_path[,backup_profile_path])
1458 Delete a dynamic device personalization profile and restore backup profile::
1460 testpmd> ddp del (port_id) (backup_profile_path)
1465 List all items from the ptype mapping table::
1467 testpmd> ptype mapping get (port_id) (valid_only)
1471 * ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0).
1473 Replace a specific or a group of software defined ptype with a new one::
1475 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type)
1479 * ``target``: A specific software ptype or a mask to represent a group of software ptypes.
1481 * ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1).
1483 * ``pkt_type``: The new software ptype to replace the old ones.
1485 Update hardware defined ptype to software defined packet type mapping table::
1487 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype)
1491 * ``hw_ptype``: hardware ptype as the index of the ptype mapping table.
1493 * ``sw_ptype``: software ptype as the value of the ptype mapping table.
1495 Reset ptype mapping table::
1497 testpmd> ptype mapping reset (port_id)
1499 config per port Rx offloading
1500 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1502 Enable or disable a per port Rx offloading on all Rx queues of a port::
1504 testpmd> port config (port_id) rx_offload (offloading) on|off
1506 * ``offloading``: can be any of these offloading capability:
1507 vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro,
1508 qinq_strip, outer_ipv4_cksum, macsec_strip,
1509 header_split, vlan_filter, vlan_extend, jumbo_frame,
1510 crc_strip, scatter, timestamp, security, keep_crc
1512 This command should be run when the port is stopped, or else it will fail.
1514 config per queue Rx offloading
1515 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1517 Enable or disable a per queue Rx offloading only on a specific Rx queue::
1519 testpmd> port (port_id) rxq (queue_id) rx_offload (offloading) on|off
1521 * ``offloading``: can be any of these offloading capability:
1522 vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro,
1523 qinq_strip, outer_ipv4_cksum, macsec_strip,
1524 header_split, vlan_filter, vlan_extend, jumbo_frame,
1525 crc_strip, scatter, timestamp, security, keep_crc
1527 This command should be run when the port is stopped, or else it will fail.
1529 config per port Tx offloading
1530 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1532 Enable or disable a per port Tx offloading on all Tx queues of a port::
1534 testpmd> port config (port_id) tx_offload (offloading) on|off
1536 * ``offloading``: can be any of these offloading capability:
1537 vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum,
1538 sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum,
1539 qinq_insert, vxlan_tnl_tso, gre_tnl_tso,
1540 ipip_tnl_tso, geneve_tnl_tso, macsec_insert,
1541 mt_lockfree, multi_segs, mbuf_fast_free, security,
1544 This command should be run when the port is stopped, or else it will fail.
1546 config per queue Tx offloading
1547 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1549 Enable or disable a per queue Tx offloading only on a specific Tx queue::
1551 testpmd> port (port_id) txq (queue_id) tx_offload (offloading) on|off
1553 * ``offloading``: can be any of these offloading capability:
1554 vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum,
1555 sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum,
1556 qinq_insert, vxlan_tnl_tso, gre_tnl_tso,
1557 ipip_tnl_tso, geneve_tnl_tso, macsec_insert,
1558 mt_lockfree, multi_segs, mbuf_fast_free, security
1560 This command should be run when the port is stopped, or else it will fail.
1562 Config VXLAN Encap outer layers
1563 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1565 Configure the outer layer to encapsulate a packet inside a VXLAN tunnel::
1567 set vxlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \
1568 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) eth-src (eth-src) \
1571 set vxlan-with-vlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \
1572 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) vlan-tci (vlan-tci) \
1573 eth-src (eth-src) eth-dst (eth-dst)
1575 Those command will set an internal configuration inside testpmd, any following
1576 flow rule using the action vxlan_encap will use the last configuration set.
1577 To have a different encapsulation header, one of those commands must be called
1578 before the flow rule creation.
1580 Config NVGRE Encap outer layers
1581 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1583 Configure the outer layer to encapsulate a packet inside a NVGRE tunnel::
1585 set nvgre ip-version (ipv4|ipv6) tni (tni) ip-src (ip-src) ip-dst (ip-dst) \
1586 eth-src (eth-src) eth-dst (eth-dst)
1587 set nvgre-with-vlan ip-version (ipv4|ipv6) tni (tni) ip-src (ip-src) \
1588 ip-dst (ip-dst) vlan-tci (vlan-tci) eth-src (eth-src) eth-dst (eth-dst)
1590 Those command will set an internal configuration inside testpmd, any following
1591 flow rule using the action nvgre_encap will use the last configuration set.
1592 To have a different encapsulation header, one of those commands must be called
1593 before the flow rule creation.
1598 The following sections show functions for configuring ports.
1602 Port configuration changes only become active when forwarding is started/restarted.
1607 Attach a port specified by pci address or virtual device args::
1609 testpmd> port attach (identifier)
1611 To attach a new pci device, the device should be recognized by kernel first.
1612 Then it should be moved under DPDK management.
1613 Finally the port can be attached to testpmd.
1615 For example, to move a pci device using ixgbe under DPDK management:
1617 .. code-block:: console
1619 # Check the status of the available devices.
1620 ./usertools/dpdk-devbind.py --status
1622 Network devices using DPDK-compatible driver
1623 ============================================
1626 Network devices using kernel driver
1627 ===================================
1628 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1631 # Bind the device to igb_uio.
1632 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1635 # Recheck the status of the devices.
1636 ./usertools/dpdk-devbind.py --status
1637 Network devices using DPDK-compatible driver
1638 ============================================
1639 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1641 To attach a port created by virtual device, above steps are not needed.
1643 For example, to attach a port whose pci address is 0000:0a:00.0.
1645 .. code-block:: console
1647 testpmd> port attach 0000:0a:00.0
1648 Attaching a new port...
1649 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1650 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1651 EAL: PCI memory mapped at 0x7f83bfa00000
1652 EAL: PCI memory mapped at 0x7f83bfa80000
1653 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1654 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1655 Port 0 is attached. Now total ports is 1
1658 For example, to attach a port created by pcap PMD.
1660 .. code-block:: console
1662 testpmd> port attach net_pcap0
1663 Attaching a new port...
1664 PMD: Initializing pmd_pcap for net_pcap0
1665 PMD: Creating pcap-backed ethdev on numa socket 0
1666 Port 0 is attached. Now total ports is 1
1669 In this case, identifier is ``net_pcap0``.
1670 This identifier format is the same as ``--vdev`` format of DPDK applications.
1672 For example, to re-attach a bonded port which has been previously detached,
1673 the mode and slave parameters must be given.
1675 .. code-block:: console
1677 testpmd> port attach net_bond_0,mode=0,slave=1
1678 Attaching a new port...
1679 EAL: Initializing pmd_bond for net_bond_0
1680 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1681 Port 0 is attached. Now total ports is 1
1688 Detach a specific port::
1690 testpmd> port detach (port_id)
1692 Before detaching a port, the port should be stopped and closed.
1694 For example, to detach a pci device port 0.
1696 .. code-block:: console
1698 testpmd> port stop 0
1701 testpmd> port close 0
1705 testpmd> port detach 0
1707 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1708 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1709 EAL: PCI memory unmapped at 0x7f83bfa00000
1710 EAL: PCI memory unmapped at 0x7f83bfa80000
1714 For example, to detach a virtual device port 0.
1716 .. code-block:: console
1718 testpmd> port stop 0
1721 testpmd> port close 0
1725 testpmd> port detach 0
1727 PMD: Closing pcap ethdev on numa socket 0
1728 Port 'net_pcap0' is detached. Now total ports is 0
1731 To remove a pci device completely from the system, first detach the port from testpmd.
1732 Then the device should be moved under kernel management.
1733 Finally the device can be removed using kernel pci hotplug functionality.
1735 For example, to move a pci device under kernel management:
1737 .. code-block:: console
1739 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1741 ./usertools/dpdk-devbind.py --status
1743 Network devices using DPDK-compatible driver
1744 ============================================
1747 Network devices using kernel driver
1748 ===================================
1749 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1751 To remove a port created by a virtual device, above steps are not needed.
1756 Start all ports or a specific port::
1758 testpmd> port start (port_id|all)
1763 Stop all ports or a specific port::
1765 testpmd> port stop (port_id|all)
1770 Close all ports or a specific port::
1772 testpmd> port close (port_id|all)
1774 port config - queue ring size
1775 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1777 Configure a rx/tx queue ring size::
1779 testpmd> port (port_id) (rxq|txq) (queue_id) ring_size (value)
1781 Only take effect after command that (re-)start the port or command that setup specific queue.
1783 port start/stop queue
1784 ~~~~~~~~~~~~~~~~~~~~~
1786 Start/stop a rx/tx queue on a specific port::
1788 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1790 port config - queue deferred start
1791 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1793 Switch on/off deferred start of a specific port queue::
1795 testpmd> port (port_id) (rxq|txq) (queue_id) deferred_start (on|off)
1798 ~~~~~~~~~~~~~~~~~~~~~
1800 Setup a rx/tx queue on a specific port::
1802 testpmd> port (port_id) (rxq|txq) (queue_id) setup
1804 Only take effect when port is started.
1809 Set the speed and duplex mode for all ports or a specific port::
1811 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1812 duplex (half|full|auto)
1814 port config - queues/descriptors
1815 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1817 Set number of queues/descriptors for rxq, txq, rxd and txd::
1819 testpmd> port config all (rxq|txq|rxd|txd) (value)
1821 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1823 port config - max-pkt-len
1824 ~~~~~~~~~~~~~~~~~~~~~~~~~
1826 Set the maximum packet length::
1828 testpmd> port config all max-pkt-len (value)
1830 This is equivalent to the ``--max-pkt-len`` command-line option.
1832 port config - CRC Strip
1833 ~~~~~~~~~~~~~~~~~~~~~~~
1835 Set hardware CRC stripping on or off for all ports::
1837 testpmd> port config all crc-strip (on|off)
1839 CRC stripping is on by default.
1841 The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option.
1843 port config - scatter
1844 ~~~~~~~~~~~~~~~~~~~~~~~
1846 Set RX scatter mode on or off for all ports::
1848 testpmd> port config all scatter (on|off)
1850 RX scatter mode is off by default.
1852 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1854 port config - RX Checksum
1855 ~~~~~~~~~~~~~~~~~~~~~~~~~
1857 Set hardware RX checksum offload to on or off for all ports::
1859 testpmd> port config all rx-cksum (on|off)
1861 Checksum offload is off by default.
1863 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1868 Set hardware VLAN on or off for all ports::
1870 testpmd> port config all hw-vlan (on|off)
1872 Hardware VLAN is off by default.
1874 The ``on`` option is equivalent to the ``--enable-hw-vlan`` command-line option.
1876 port config - VLAN filter
1877 ~~~~~~~~~~~~~~~~~~~~~~~~~
1879 Set hardware VLAN filter on or off for all ports::
1881 testpmd> port config all hw-vlan-filter (on|off)
1883 Hardware VLAN filter is off by default.
1885 The ``on`` option is equivalent to the ``--enable-hw-vlan-filter`` command-line option.
1887 port config - VLAN strip
1888 ~~~~~~~~~~~~~~~~~~~~~~~~
1890 Set hardware VLAN strip on or off for all ports::
1892 testpmd> port config all hw-vlan-strip (on|off)
1894 Hardware VLAN strip is off by default.
1896 The ``on`` option is equivalent to the ``--enable-hw-vlan-strip`` command-line option.
1898 port config - VLAN extend
1899 ~~~~~~~~~~~~~~~~~~~~~~~~~
1901 Set hardware VLAN extend on or off for all ports::
1903 testpmd> port config all hw-vlan-extend (on|off)
1905 Hardware VLAN extend is off by default.
1907 The ``on`` option is equivalent to the ``--enable-hw-vlan-extend`` command-line option.
1909 port config - Drop Packets
1910 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1912 Set packet drop for packets with no descriptors on or off for all ports::
1914 testpmd> port config all drop-en (on|off)
1916 Packet dropping for packets with no descriptors is off by default.
1918 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1923 Set the RSS (Receive Side Scaling) mode on or off::
1925 testpmd> port config all rss (all|default|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1927 RSS is on by default.
1929 The ``all`` option is equivalent to ip|tcp|udp|sctp|ether.
1930 The ``default`` option enables all supported RSS types reported by device info.
1931 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1933 port config - RSS Reta
1934 ~~~~~~~~~~~~~~~~~~~~~~
1936 Set the RSS (Receive Side Scaling) redirection table::
1938 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1943 Set the DCB mode for an individual port::
1945 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1947 The traffic class should be 4 or 8.
1952 Set the number of packets per burst::
1954 testpmd> port config all burst (value)
1956 This is equivalent to the ``--burst`` command-line option.
1958 port config - Threshold
1959 ~~~~~~~~~~~~~~~~~~~~~~~
1961 Set thresholds for TX/RX queues::
1963 testpmd> port config all (threshold) (value)
1965 Where the threshold type can be:
1967 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1969 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1971 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1973 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1975 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1977 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1979 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1981 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1983 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1985 These threshold options are also available from the command-line.
1990 Set the value of ether-type for E-tag::
1992 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1994 Enable/disable the E-tag support::
1996 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1998 port config pctype mapping
1999 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2001 Reset pctype mapping table::
2003 testpmd> port config (port_id) pctype mapping reset
2005 Update hardware defined pctype to software defined flow type mapping table::
2007 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id)
2011 * ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table.
2013 * ``flow_type_id``: software flow type id as the index of the pctype mapping table.
2015 port config input set
2016 ~~~~~~~~~~~~~~~~~~~~~
2018 Config RSS/FDIR/FDIR flexible payload input set for some pctype::
2019 testpmd> port config (port_id) pctype (pctype_id) \
2020 (hash_inset|fdir_inset|fdir_flx_inset) \
2021 (get|set|clear) field (field_idx)
2023 Clear RSS/FDIR/FDIR flexible payload input set for some pctype::
2024 testpmd> port config (port_id) pctype (pctype_id) \
2025 (hash_inset|fdir_inset|fdir_flx_inset) clear all
2029 * ``pctype_id``: hardware packet classification types.
2030 * ``field_idx``: hardware field index.
2032 port config udp_tunnel_port
2033 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
2035 Add/remove UDP tunnel port for VXLAN/GENEVE tunneling protocols::
2036 testpmd> port config (port_id) udp_tunnel_port add|rm vxlan|geneve (udp_port)
2038 port config tx_metadata
2039 ~~~~~~~~~~~~~~~~~~~~~~~
2041 Set Tx metadata value per port.
2042 testpmd will add this value to any Tx packet sent from this port::
2044 testpmd> port config (port_id) tx_metadata (value)
2046 Link Bonding Functions
2047 ----------------------
2049 The Link Bonding functions make it possible to dynamically create and
2050 manage link bonding devices from within testpmd interactive prompt.
2052 create bonded device
2053 ~~~~~~~~~~~~~~~~~~~~
2055 Create a new bonding device::
2057 testpmd> create bonded device (mode) (socket)
2059 For example, to create a bonded device in mode 1 on socket 0::
2061 testpmd> create bonded device 1 0
2062 created new bonded device (port X)
2067 Adds Ethernet device to a Link Bonding device::
2069 testpmd> add bonding slave (slave id) (port id)
2071 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
2073 testpmd> add bonding slave 6 10
2076 remove bonding slave
2077 ~~~~~~~~~~~~~~~~~~~~
2079 Removes an Ethernet slave device from a Link Bonding device::
2081 testpmd> remove bonding slave (slave id) (port id)
2083 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
2085 testpmd> remove bonding slave 6 10
2090 Set the Link Bonding mode of a Link Bonding device::
2092 testpmd> set bonding mode (value) (port id)
2094 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
2096 testpmd> set bonding mode 3 10
2101 Set an Ethernet slave device as the primary device on a Link Bonding device::
2103 testpmd> set bonding primary (slave id) (port id)
2105 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
2107 testpmd> set bonding primary 6 10
2112 Set the MAC address of a Link Bonding device::
2114 testpmd> set bonding mac (port id) (mac)
2116 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
2118 testpmd> set bonding mac 10 00:00:00:00:00:01
2120 set bonding xmit_balance_policy
2121 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2123 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
2125 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
2127 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
2129 testpmd> set bonding xmit_balance_policy 10 l34
2132 set bonding mon_period
2133 ~~~~~~~~~~~~~~~~~~~~~~
2135 Set the link status monitoring polling period in milliseconds for a bonding device.
2137 This adds support for PMD slave devices which do not support link status interrupts.
2138 When the mon_period is set to a value greater than 0 then all PMD's which do not support
2139 link status ISR will be queried every polling interval to check if their link status has changed::
2141 testpmd> set bonding mon_period (port_id) (value)
2143 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
2145 testpmd> set bonding mon_period 5 150
2148 set bonding lacp dedicated_queue
2149 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2151 Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic
2152 when in mode 4 (link-aggregration-802.3ad)::
2154 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable)
2157 set bonding agg_mode
2158 ~~~~~~~~~~~~~~~~~~~~
2160 Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad)::
2162 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable)
2168 Show the current configuration of a Link Bonding device::
2170 testpmd> show bonding config (port id)
2173 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
2174 in balance mode with a transmission policy of layer 2+3::
2176 testpmd> show bonding config 9
2178 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
2180 Active Slaves (3): [1 3 4]
2187 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
2188 This is mainly useful for debugging purposes.
2189 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
2190 and fields that can be accessed.
2195 Display the value of a port register::
2197 testpmd> read reg (port_id) (address)
2199 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
2201 testpmd> read reg 0 0xEE00
2202 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
2207 Display a port register bit field::
2209 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
2211 For example, reading the lowest two bits from the register in the example above::
2213 testpmd> read regfield 0 0xEE00 0 1
2214 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
2219 Display a single port register bit::
2221 testpmd> read regbit (port_id) (address) (bit_x)
2223 For example, reading the lowest bit from the register in the example above::
2225 testpmd> read regbit 0 0xEE00 0
2226 port 0 PCI register at offset 0xEE00: bit 0=1
2231 Set the value of a port register::
2233 testpmd> write reg (port_id) (address) (value)
2235 For example, to clear a register::
2237 testpmd> write reg 0 0xEE00 0x0
2238 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
2243 Set bit field of a port register::
2245 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
2247 For example, writing to the register cleared in the example above::
2249 testpmd> write regfield 0 0xEE00 0 1 2
2250 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
2255 Set single bit value of a port register::
2257 testpmd> write regbit (port_id) (address) (bit_x) (value)
2259 For example, to set the high bit in the register from the example above::
2261 testpmd> write regbit 0 0xEE00 31 1
2262 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
2264 Traffic Metering and Policing
2265 -----------------------------
2267 The following section shows functions for configuring traffic metering and
2268 policing on the ethernet device through the use of generic ethdev API.
2270 show port traffic management capability
2271 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2273 Show traffic metering and policing capability of the port::
2275 testpmd> show port meter cap (port_id)
2277 add port meter profile (srTCM rfc2967)
2278 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2280 Add meter profile (srTCM rfc2697) to the ethernet device::
2282 testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \
2287 * ``profile_id``: ID for the meter profile.
2288 * ``cir``: Committed Information Rate (CIR) (bytes/second).
2289 * ``cbs``: Committed Burst Size (CBS) (bytes).
2290 * ``ebs``: Excess Burst Size (EBS) (bytes).
2292 add port meter profile (trTCM rfc2968)
2293 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2295 Add meter profile (srTCM rfc2698) to the ethernet device::
2297 testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \
2298 (cir) (pir) (cbs) (pbs)
2302 * ``profile_id``: ID for the meter profile.
2303 * ``cir``: Committed information rate (bytes/second).
2304 * ``pir``: Peak information rate (bytes/second).
2305 * ``cbs``: Committed burst size (bytes).
2306 * ``pbs``: Peak burst size (bytes).
2308 add port meter profile (trTCM rfc4115)
2309 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2311 Add meter profile (trTCM rfc4115) to the ethernet device::
2313 testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \
2314 (cir) (eir) (cbs) (ebs)
2318 * ``profile_id``: ID for the meter profile.
2319 * ``cir``: Committed information rate (bytes/second).
2320 * ``eir``: Excess information rate (bytes/second).
2321 * ``cbs``: Committed burst size (bytes).
2322 * ``ebs``: Excess burst size (bytes).
2324 delete port meter profile
2325 ~~~~~~~~~~~~~~~~~~~~~~~~~
2327 Delete meter profile from the ethernet device::
2329 testpmd> del port meter profile (port_id) (profile_id)
2334 Create new meter object for the ethernet device::
2336 testpmd> create port meter (port_id) (mtr_id) (profile_id) \
2337 (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \
2338 (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\
2343 * ``mtr_id``: meter object ID.
2344 * ``profile_id``: ID for the meter profile.
2345 * ``meter_enable``: When this parameter has a non-zero value, the meter object
2346 gets enabled at the time of creation, otherwise remains disabled.
2347 * ``g_action``: Policer action for the packet with green color.
2348 * ``y_action``: Policer action for the packet with yellow color.
2349 * ``r_action``: Policer action for the packet with red color.
2350 * ``stats_mask``: Mask of statistics counter types to be enabled for the
2352 * ``shared``: When this parameter has a non-zero value, the meter object is
2353 shared by multiple flows. Otherwise, meter object is used by single flow.
2354 * ``use_pre_meter_color``: When this parameter has a non-zero value, the
2355 input color for the current meter object is determined by the latest meter
2356 object in the same flow. Otherwise, the current meter object uses the
2357 *dscp_table* to determine the input color.
2358 * ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input
2359 color, 0 <= x <= 63.
2364 Enable meter for the ethernet device::
2366 testpmd> enable port meter (port_id) (mtr_id)
2371 Disable meter for the ethernet device::
2373 testpmd> disable port meter (port_id) (mtr_id)
2378 Delete meter for the ethernet device::
2380 testpmd> del port meter (port_id) (mtr_id)
2382 Set port meter profile
2383 ~~~~~~~~~~~~~~~~~~~~~~
2385 Set meter profile for the ethernet device::
2387 testpmd> set port meter profile (port_id) (mtr_id) (profile_id)
2389 set port meter dscp table
2390 ~~~~~~~~~~~~~~~~~~~~~~~~~
2392 Set meter dscp table for the ethernet device::
2394 testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \
2395 (dscp_tbl_entry1)...(dscp_tbl_entry63)]
2397 set port meter policer action
2398 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2400 Set meter policer action for the ethernet device::
2402 testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \
2403 (action0) [(action1) (action1)]
2407 * ``action_mask``: Bit mask indicating which policer actions need to be
2408 updated. One or more policer actions can be updated in a single function
2409 invocation. To update the policer action associated with color C, bit
2410 (1 << C) needs to be set in *action_mask* and element at position C
2411 in the *actions* array needs to be valid.
2412 * ``actionx``: Policer action for the color x,
2413 RTE_MTR_GREEN <= x < RTE_MTR_COLORS
2415 set port meter stats mask
2416 ~~~~~~~~~~~~~~~~~~~~~~~~~
2418 Set meter stats mask for the ethernet device::
2420 testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask)
2424 * ``stats_mask``: Bit mask indicating statistics counter types to be enabled.
2426 show port meter stats
2427 ~~~~~~~~~~~~~~~~~~~~~
2429 Show meter stats of the ethernet device::
2431 testpmd> show port meter stats (port_id) (mtr_id) (clear)
2435 * ``clear``: Flag that indicates whether the statistics counters should
2436 be cleared (i.e. set to zero) immediately after they have been read or not.
2441 The following section shows functions for configuring traffic management on
2442 on the ethernet device through the use of generic TM API.
2444 show port traffic management capability
2445 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2447 Show traffic management capability of the port::
2449 testpmd> show port tm cap (port_id)
2451 show port traffic management capability (hierarchy level)
2452 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2454 Show traffic management hierarchy level capability of the port::
2456 testpmd> show port tm level cap (port_id) (level_id)
2458 show port traffic management capability (hierarchy node level)
2459 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2461 Show the traffic management hierarchy node capability of the port::
2463 testpmd> show port tm node cap (port_id) (node_id)
2465 show port traffic management hierarchy node type
2466 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2468 Show the port traffic management hierarchy node type::
2470 testpmd> show port tm node type (port_id) (node_id)
2472 show port traffic management hierarchy node stats
2473 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2475 Show the port traffic management hierarchy node statistics::
2477 testpmd> show port tm node stats (port_id) (node_id) (clear)
2481 * ``clear``: When this parameter has a non-zero value, the statistics counters
2482 are cleared (i.e. set to zero) immediately after they have been read,
2483 otherwise the statistics counters are left untouched.
2485 Add port traffic management private shaper profile
2486 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2488 Add the port traffic management private shaper profile::
2490 testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \
2491 (tb_rate) (tb_size) (packet_length_adjust)
2495 * ``shaper_profile id``: Shaper profile ID for the new profile.
2496 * ``tb_rate``: Token bucket rate (bytes per second).
2497 * ``tb_size``: Token bucket size (bytes).
2498 * ``packet_length_adjust``: The value (bytes) to be added to the length of
2499 each packet for the purpose of shaping. This parameter value can be used to
2500 correct the packet length with the framing overhead bytes that are consumed
2503 Delete port traffic management private shaper profile
2504 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2506 Delete the port traffic management private shaper::
2508 testpmd> del port tm node shaper profile (port_id) (shaper_profile_id)
2512 * ``shaper_profile id``: Shaper profile ID that needs to be deleted.
2514 Add port traffic management shared shaper
2515 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2517 Create the port traffic management shared shaper::
2519 testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \
2524 * ``shared_shaper_id``: Shared shaper ID to be created.
2525 * ``shaper_profile id``: Shaper profile ID for shared shaper.
2527 Set port traffic management shared shaper
2528 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2530 Update the port traffic management shared shaper::
2532 testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \
2537 * ``shared_shaper_id``: Shared shaper ID to be update.
2538 * ``shaper_profile id``: Shaper profile ID for shared shaper.
2540 Delete port traffic management shared shaper
2541 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2543 Delete the port traffic management shared shaper::
2545 testpmd> del port tm node shared shaper (port_id) (shared_shaper_id)
2549 * ``shared_shaper_id``: Shared shaper ID to be deleted.
2551 Set port traffic management hiearchy node private shaper
2552 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2554 set the port traffic management hierarchy node private shaper::
2556 testpmd> set port tm node shaper profile (port_id) (node_id) \
2561 * ``shaper_profile id``: Private shaper profile ID to be enabled on the
2564 Add port traffic management WRED profile
2565 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2567 Create a new WRED profile::
2569 testpmd> add port tm node wred profile (port_id) (wred_profile_id) \
2570 (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \
2571 (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \
2572 (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r)
2576 * ``wred_profile id``: Identifier for the newly create WRED profile
2577 * ``color_g``: Packet color (green)
2578 * ``min_th_g``: Minimum queue threshold for packet with green color
2579 * ``max_th_g``: Minimum queue threshold for packet with green color
2580 * ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp)
2581 * ``wq_log2_g``: Negated log2 of queue weight (wq)
2582 * ``color_y``: Packet color (yellow)
2583 * ``min_th_y``: Minimum queue threshold for packet with yellow color
2584 * ``max_th_y``: Minimum queue threshold for packet with yellow color
2585 * ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp)
2586 * ``wq_log2_y``: Negated log2 of queue weight (wq)
2587 * ``color_r``: Packet color (red)
2588 * ``min_th_r``: Minimum queue threshold for packet with yellow color
2589 * ``max_th_r``: Minimum queue threshold for packet with yellow color
2590 * ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp)
2591 * ``wq_log2_r``: Negated log2 of queue weight (wq)
2593 Delete port traffic management WRED profile
2594 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2596 Delete the WRED profile::
2598 testpmd> del port tm node wred profile (port_id) (wred_profile_id)
2600 Add port traffic management hierarchy nonleaf node
2601 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2603 Add nonleaf node to port traffic management hiearchy::
2605 testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \
2606 (priority) (weight) (level_id) (shaper_profile_id) \
2607 (n_sp_priorities) (stats_mask) (n_shared_shapers) \
2608 [(shared_shaper_0) (shared_shaper_1) ...] \
2612 * ``parent_node_id``: Node ID of the parent.
2613 * ``priority``: Node priority (highest node priority is zero). This is used by
2614 the SP algorithm running on the parent node for scheduling this node.
2615 * ``weight``: Node weight (lowest weight is one). The node weight is relative
2616 to the weight sum of all siblings that have the same priority. It is used by
2617 the WFQ algorithm running on the parent node for scheduling this node.
2618 * ``level_id``: Hiearchy level of the node.
2619 * ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2621 * ``n_sp_priorities``: Number of strict priorities.
2622 * ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2623 * ``n_shared_shapers``: Number of shared shapers.
2624 * ``shared_shaper_id``: Shared shaper id.
2626 Add port traffic management hierarchy leaf node
2627 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2629 Add leaf node to port traffic management hiearchy::
2631 testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \
2632 (priority) (weight) (level_id) (shaper_profile_id) \
2633 (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \
2634 [(shared_shaper_id) (shared_shaper_id) ...] \
2638 * ``parent_node_id``: Node ID of the parent.
2639 * ``priority``: Node priority (highest node priority is zero). This is used by
2640 the SP algorithm running on the parent node for scheduling this node.
2641 * ``weight``: Node weight (lowest weight is one). The node weight is relative
2642 to the weight sum of all siblings that have the same priority. It is used by
2643 the WFQ algorithm running on the parent node for scheduling this node.
2644 * ``level_id``: Hiearchy level of the node.
2645 * ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2647 * ``cman_mode``: Congestion management mode to be enabled for this node.
2648 * ``wred_profile_id``: WRED profile id to be enabled for this node.
2649 * ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2650 * ``n_shared_shapers``: Number of shared shapers.
2651 * ``shared_shaper_id``: Shared shaper id.
2653 Delete port traffic management hierarchy node
2654 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2656 Delete node from port traffic management hiearchy::
2658 testpmd> del port tm node (port_id) (node_id)
2660 Update port traffic management hierarchy parent node
2661 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2663 Update port traffic management hierarchy parent node::
2665 testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \
2668 This function can only be called after the hierarchy commit invocation. Its
2669 success depends on the port support for this operation, as advertised through
2670 the port capability set. This function is valid for all nodes of the traffic
2671 management hierarchy except root node.
2673 Suspend port traffic management hierarchy node
2674 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2676 testpmd> suspend port tm node (port_id) (node_id)
2678 Resume port traffic management hierarchy node
2679 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2681 testpmd> resume port tm node (port_id) (node_id)
2683 Commit port traffic management hierarchy
2684 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2686 Commit the traffic management hierarchy on the port::
2688 testpmd> port tm hierarchy commit (port_id) (clean_on_fail)
2692 * ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function
2693 call failure. On the other hand, hierarchy is preserved when this parameter
2696 Set port traffic management mark VLAN dei
2697 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2699 Enables/Disables the traffic management marking on the port for VLAN packets::
2701 testpmd> set port tm mark vlan_dei <port_id> <green> <yellow> <red>
2705 * ``port_id``: The port which on which VLAN packets marked as ``green`` or
2706 ``yellow`` or ``red`` will have dei bit enabled
2708 * ``green`` enable 1, disable 0 marking for dei bit of VLAN packets marked as green
2710 * ``yellow`` enable 1, disable 0 marking for dei bit of VLAN packets marked as yellow
2712 * ``red`` enable 1, disable 0 marking for dei bit of VLAN packets marked as red
2714 Set port traffic management mark IP dscp
2715 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2717 Enables/Disables the traffic management marking on the port for IP dscp packets::
2719 testpmd> set port tm mark ip_dscp <port_id> <green> <yellow> <red>
2723 * ``port_id``: The port which on which IP packets marked as ``green`` or
2724 ``yellow`` or ``red`` will have IP dscp bits updated
2726 * ``green`` enable 1, disable 0 marking IP dscp to low drop precedence for green packets
2728 * ``yellow`` enable 1, disable 0 marking IP dscp to medium drop precedence for yellow packets
2730 * ``red`` enable 1, disable 0 marking IP dscp to high drop precedence for red packets
2732 Set port traffic management mark IP ecn
2733 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2735 Enables/Disables the traffic management marking on the port for IP ecn packets::
2737 testpmd> set port tm mark ip_ecn <port_id> <green> <yellow> <red>
2741 * ``port_id``: The port which on which IP packets marked as ``green`` or
2742 ``yellow`` or ``red`` will have IP ecn bits updated
2744 * ``green`` enable 1, disable 0 marking IP ecn for green marked packets with ecn of 2'b01 or 2'b10
2745 to ecn of 2'b11 when IP is caring TCP or SCTP
2747 * ``yellow`` enable 1, disable 0 marking IP ecn for yellow marked packets with ecn of 2'b01 or 2'b10
2748 to ecn of 2'b11 when IP is caring TCP or SCTP
2750 * ``red`` enable 1, disable 0 marking IP ecn for yellow marked packets with ecn of 2'b01 or 2'b10
2751 to ecn of 2'b11 when IP is caring TCP or SCTP
2753 Set port traffic management default hierarchy (softnic forwarding mode)
2754 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2756 set the traffic management default hierarchy on the port::
2758 testpmd> set port tm hierarchy default (port_id)
2763 This section details the available filter functions that are available.
2765 Note these functions interface the deprecated legacy filtering framework,
2766 superseded by *rte_flow*. See `Flow rules management`_.
2769 ~~~~~~~~~~~~~~~~~~~~
2771 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
2773 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
2774 ethertype (ether_type) (drop|fwd) queue (queue_id)
2776 The available information parameters are:
2778 * ``port_id``: The port which the Ethertype filter assigned on.
2780 * ``mac_addr``: Compare destination mac address.
2782 * ``mac_ignr``: Ignore destination mac address match.
2784 * ``mac_address``: Destination mac address to match.
2786 * ``ether_type``: The EtherType value want to match,
2787 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
2789 * ``queue_id``: The receive queue associated with this EtherType filter.
2790 It is meaningless when deleting or dropping.
2792 Example, to add/remove an ethertype filter rule::
2794 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
2795 ethertype 0x0806 fwd queue 3
2797 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
2798 ethertype 0x0806 fwd queue 3
2803 Add or delete a 2-tuple filter,
2804 which identifies packets by specific protocol and destination TCP/UDP port
2805 and forwards packets into one of the receive queues::
2807 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
2808 protocol (protocol_value) mask (mask_value) \
2809 tcp_flags (tcp_flags_value) priority (prio_value) \
2812 The available information parameters are:
2814 * ``port_id``: The port which the 2-tuple filter assigned on.
2816 * ``dst_port_value``: Destination port in L4.
2818 * ``protocol_value``: IP L4 protocol.
2820 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
2822 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
2824 * ``prio_value``: Priority of this filter.
2826 * ``queue_id``: The receive queue associated with this 2-tuple filter.
2828 Example, to add/remove an 2tuple filter rule::
2830 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
2831 tcp_flags 0x02 priority 3 queue 3
2833 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
2834 tcp_flags 0x02 priority 3 queue 3
2839 Add or delete a 5-tuple filter,
2840 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
2841 and routes packets into one of the receive queues::
2843 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
2844 (src_address) dst_port (dst_port_value) \
2845 src_port (src_port_value) protocol (protocol_value) \
2846 mask (mask_value) tcp_flags (tcp_flags_value) \
2847 priority (prio_value) queue (queue_id)
2849 The available information parameters are:
2851 * ``port_id``: The port which the 5-tuple filter assigned on.
2853 * ``dst_address``: Destination IP address.
2855 * ``src_address``: Source IP address.
2857 * ``dst_port_value``: TCP/UDP destination port.
2859 * ``src_port_value``: TCP/UDP source port.
2861 * ``protocol_value``: L4 protocol.
2863 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
2865 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
2867 * ``prio_value``: The priority of this filter.
2869 * ``queue_id``: The receive queue associated with this 5-tuple filter.
2871 Example, to add/remove an 5tuple filter rule::
2873 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2874 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2875 flags 0x0 priority 3 queue 3
2877 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2878 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2879 flags 0x0 priority 3 queue 3
2884 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
2886 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
2888 The available information parameters are:
2890 * ``port_id``: The port which the SYN filter assigned on.
2892 * ``high``: This SYN filter has higher priority than other filters.
2894 * ``low``: This SYN filter has lower priority than other filters.
2896 * ``queue_id``: The receive queue associated with this SYN filter
2900 testpmd> syn_filter 0 add priority high queue 3
2905 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
2906 and routed into one of the receive queues::
2908 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
2909 mask (mask_value) priority (prio_value) queue (queue_id)
2911 The available information parameters are:
2913 * ``port_id``: The port which the Flex filter is assigned on.
2915 * ``len_value``: Filter length in bytes, no greater than 128.
2917 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
2919 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
2921 * ``prio_value``: The priority of this filter.
2923 * ``queue_id``: The receive queue associated with this Flex filter.
2927 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
2928 mask 000C priority 3 queue 3
2930 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
2931 mask 000C priority 3 queue 3
2934 .. _testpmd_flow_director:
2936 flow_director_filter
2937 ~~~~~~~~~~~~~~~~~~~~
2939 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
2941 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
2942 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
2944 * Perfect match filters.
2945 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2946 The masked fields are for IP flow.
2948 * Signature filters.
2949 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
2951 * Perfect-mac-vlan match filters.
2952 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2953 The masked fields are for MAC VLAN flow.
2955 * Perfect-tunnel match filters.
2956 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2957 The masked fields are for tunnel flow.
2959 * Perfect-raw-flow-type match filters.
2960 The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet.
2961 The masked fields are specified by input sets.
2963 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
2964 per flow type and the flexible payload.
2966 The Flow Director can also mask out parts of all of these fields so that filters
2967 are only applied to certain fields or parts of the fields.
2969 Note that for raw flow type mode the source and destination fields in the
2970 raw packet buffer need to be presented in a reversed order with respect
2971 to the expected received packets.
2972 For example: IP source and destination addresses or TCP/UDP/SCTP
2973 source and destination ports
2975 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
2977 # Commands to add flow director filters of different flow types::
2979 flow_director_filter (port_id) mode IP (add|del|update) \
2980 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
2981 src (src_ip_address) dst (dst_ip_address) \
2982 tos (tos_value) proto (proto_value) ttl (ttl_value) \
2983 vlan (vlan_value) flexbytes (flexbytes_value) \
2984 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
2987 flow_director_filter (port_id) mode IP (add|del|update) \
2988 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
2989 src (src_ip_address) (src_port) \
2990 dst (dst_ip_address) (dst_port) \
2991 tos (tos_value) ttl (ttl_value) \
2992 vlan (vlan_value) flexbytes (flexbytes_value) \
2993 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
2996 flow_director_filter (port_id) mode IP (add|del|update) \
2997 flow (ipv4-sctp|ipv6-sctp) \
2998 src (src_ip_address) (src_port) \
2999 dst (dst_ip_address) (dst_port) \
3000 tos (tos_value) ttl (ttl_value) \
3001 tag (verification_tag) vlan (vlan_value) \
3002 flexbytes (flexbytes_value) (drop|fwd) \
3003 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
3005 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
3006 ether (ethertype) flexbytes (flexbytes_value) \
3007 (drop|fwd) pf|vf(vf_id) queue (queue_id)
3010 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
3011 mac (mac_address) vlan (vlan_value) \
3012 flexbytes (flexbytes_value) (drop|fwd) \
3013 queue (queue_id) fd_id (fd_id_value)
3015 flow_director_filter (port_id) mode Tunnel (add|del|update) \
3016 mac (mac_address) vlan (vlan_value) \
3017 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
3018 flexbytes (flexbytes_value) (drop|fwd) \
3019 queue (queue_id) fd_id (fd_id_value)
3021 flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \
3022 (drop|fwd) queue (queue_id) fd_id (fd_id_value) \
3023 packet (packet file name)
3025 For example, to add an ipv4-udp flow type filter::
3027 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
3028 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
3029 fwd pf queue 1 fd_id 1
3031 For example, add an ipv4-other flow type filter::
3033 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
3034 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
3035 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
3040 Flush all flow director filters on a device::
3042 testpmd> flush_flow_director (port_id)
3044 Example, to flush all flow director filter on port 0::
3046 testpmd> flush_flow_director 0
3051 Set flow director's input masks::
3053 flow_director_mask (port_id) mode IP vlan (vlan_value) \
3054 src_mask (ipv4_src) (ipv6_src) (src_port) \
3055 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
3057 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
3059 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
3060 mac (mac_value) tunnel-type (tunnel_type_value) \
3061 tunnel-id (tunnel_id_value)
3063 Example, to set flow director mask on port 0::
3065 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
3066 src_mask 255.255.255.255 \
3067 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
3068 dst_mask 255.255.255.255 \
3069 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
3071 flow_director_flex_mask
3072 ~~~~~~~~~~~~~~~~~~~~~~~
3074 set masks of flow director's flexible payload based on certain flow type::
3076 testpmd> flow_director_flex_mask (port_id) \
3077 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
3078 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
3079 l2_payload|all) (mask)
3081 Example, to set flow director's flex mask for all flow type on port 0::
3083 testpmd> flow_director_flex_mask 0 flow all \
3084 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
3087 flow_director_flex_payload
3088 ~~~~~~~~~~~~~~~~~~~~~~~~~~
3090 Configure flexible payload selection::
3092 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
3094 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
3096 testpmd> flow_director_flex_payload 0 l4 \
3097 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
3099 get_sym_hash_ena_per_port
3100 ~~~~~~~~~~~~~~~~~~~~~~~~~
3102 Get symmetric hash enable configuration per port::
3104 get_sym_hash_ena_per_port (port_id)
3106 For example, to get symmetric hash enable configuration of port 1::
3108 testpmd> get_sym_hash_ena_per_port 1
3110 set_sym_hash_ena_per_port
3111 ~~~~~~~~~~~~~~~~~~~~~~~~~
3113 Set symmetric hash enable configuration per port to enable or disable::
3115 set_sym_hash_ena_per_port (port_id) (enable|disable)
3117 For example, to set symmetric hash enable configuration of port 1 to enable::
3119 testpmd> set_sym_hash_ena_per_port 1 enable
3121 get_hash_global_config
3122 ~~~~~~~~~~~~~~~~~~~~~~
3124 Get the global configurations of hash filters::
3126 get_hash_global_config (port_id)
3128 For example, to get the global configurations of hash filters of port 1::
3130 testpmd> get_hash_global_config 1
3132 set_hash_global_config
3133 ~~~~~~~~~~~~~~~~~~~~~~
3135 Set the global configurations of hash filters::
3137 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
3138 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
3139 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \
3142 For example, to enable simple_xor for flow type of ipv6 on port 2::
3144 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
3149 Set the input set for hash::
3151 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
3152 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
3153 l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
3154 ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
3155 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
3156 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
3157 fld-8th|none) (select|add)
3159 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
3161 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
3166 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
3167 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
3169 Set the input set for flow director::
3171 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
3172 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
3173 l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
3174 ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
3175 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
3176 sctp-dst-port|sctp-veri-tag|none) (select|add)
3178 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
3180 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
3185 Set different GRE key length for input set::
3187 global_config (port_id) gre-key-len (number in bytes)
3189 For example to set GRE key length for input set to 4 bytes on port 0::
3191 testpmd> global_config 0 gre-key-len 4
3194 .. _testpmd_rte_flow:
3196 Flow rules management
3197 ---------------------
3199 Control of the generic flow API (*rte_flow*) is fully exposed through the
3200 ``flow`` command (validation, creation, destruction, queries and operation
3203 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
3204 features simultaneously may cause undefined side-effects and is therefore
3210 Because the ``flow`` command uses dynamic tokens to handle the large number
3211 of possible flow rules combinations, its behavior differs slightly from
3212 other commands, in particular:
3214 - Pressing *?* or the *<tab>* key displays contextual help for the current
3215 token, not that of the entire command.
3217 - Optional and repeated parameters are supported (provided they are listed
3218 in the contextual help).
3220 The first parameter stands for the operation mode. Possible operations and
3221 their general syntax are described below. They are covered in detail in the
3224 - Check whether a flow rule can be created::
3226 flow validate {port_id}
3227 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3228 pattern {item} [/ {item} [...]] / end
3229 actions {action} [/ {action} [...]] / end
3231 - Create a flow rule::
3233 flow create {port_id}
3234 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3235 pattern {item} [/ {item} [...]] / end
3236 actions {action} [/ {action} [...]] / end
3238 - Destroy specific flow rules::
3240 flow destroy {port_id} rule {rule_id} [...]
3242 - Destroy all flow rules::
3244 flow flush {port_id}
3246 - Query an existing flow rule::
3248 flow query {port_id} {rule_id} {action}
3250 - List existing flow rules sorted by priority, filtered by group
3253 flow list {port_id} [group {group_id}] [...]
3255 - Restrict ingress traffic to the defined flow rules::
3257 flow isolate {port_id} {boolean}
3259 Validating flow rules
3260 ~~~~~~~~~~~~~~~~~~~~~
3262 ``flow validate`` reports whether a flow rule would be accepted by the
3263 underlying device in its current state but stops short of creating it. It is
3264 bound to ``rte_flow_validate()``::
3266 flow validate {port_id}
3267 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3268 pattern {item} [/ {item} [...]] / end
3269 actions {action} [/ {action} [...]] / end
3271 If successful, it will show::
3275 Otherwise it will show an error message of the form::
3277 Caught error type [...] ([...]): [...]
3279 This command uses the same parameters as ``flow create``, their format is
3280 described in `Creating flow rules`_.
3282 Check whether redirecting any Ethernet packet received on port 0 to RX queue
3283 index 6 is supported::
3285 testpmd> flow validate 0 ingress pattern eth / end
3286 actions queue index 6 / end
3290 Port 0 does not support TCPv6 rules::
3292 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
3294 Caught error type 9 (specific pattern item): Invalid argument
3300 ``flow create`` validates and creates the specified flow rule. It is bound
3301 to ``rte_flow_create()``::
3303 flow create {port_id}
3304 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3305 pattern {item} [/ {item} [...]] / end
3306 actions {action} [/ {action} [...]] / end
3308 If successful, it will return a flow rule ID usable with other commands::
3310 Flow rule #[...] created
3312 Otherwise it will show an error message of the form::
3314 Caught error type [...] ([...]): [...]
3316 Parameters describe in the following order:
3318 - Attributes (*group*, *priority*, *ingress*, *egress*, *transfer* tokens).
3319 - A matching pattern, starting with the *pattern* token and terminated by an
3321 - Actions, starting with the *actions* token and terminated by an *end*
3324 These translate directly to *rte_flow* objects provided as-is to the
3325 underlying functions.
3327 The shortest valid definition only comprises mandatory tokens::
3329 testpmd> flow create 0 pattern end actions end
3331 Note that PMDs may refuse rules that essentially do nothing such as this
3334 **All unspecified object values are automatically initialized to 0.**
3339 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
3340 specified before the ``pattern`` token.
3342 - ``group {group id}``: priority group.
3343 - ``priority {level}``: priority level within group.
3344 - ``ingress``: rule applies to ingress traffic.
3345 - ``egress``: rule applies to egress traffic.
3346 - ``transfer``: apply rule directly to endpoints found in pattern.
3348 Each instance of an attribute specified several times overrides the previous
3349 value as shown below (group 4 is used)::
3351 testpmd> flow create 0 group 42 group 24 group 4 [...]
3353 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
3355 While not specifying a direction is an error, some rules may allow both
3358 Most rules affect RX therefore contain the ``ingress`` token::
3360 testpmd> flow create 0 ingress pattern [...]
3365 A matching pattern starts after the ``pattern`` token. It is made of pattern
3366 items and is terminated by a mandatory ``end`` item.
3368 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
3369 rte_flow_item_type``).
3371 The ``/`` token is used as a separator between pattern items as shown
3374 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
3376 Note that protocol items like these must be stacked from lowest to highest
3377 layer to make sense. For instance, the following rule is either invalid or
3378 unlikely to match any packet::
3380 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
3382 More information on these restrictions can be found in the *rte_flow*
3385 Several items support additional specification structures, for example
3386 ``ipv4`` allows specifying source and destination addresses as follows::
3388 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
3389 dst is 10.2.0.0 / end [...]
3391 This rule matches all IPv4 traffic with the specified properties.
3393 In this example, ``src`` and ``dst`` are field names of the underlying
3394 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
3395 in a similar fashion.
3397 The ``is`` token means that the subsequent value must be matched exactly,
3398 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
3399 accordingly. Possible assignment tokens are:
3401 - ``is``: match value perfectly (with full bit-mask).
3402 - ``spec``: match value according to configured bit-mask.
3403 - ``last``: specify upper bound to establish a range.
3404 - ``mask``: specify bit-mask with relevant bits set to one.
3405 - ``prefix``: generate bit-mask from a prefix length.
3407 These yield identical results::
3409 ipv4 src is 10.1.1.1
3413 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
3417 ipv4 src spec 10.1.1.1 src prefix 32
3421 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
3425 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
3427 Inclusive ranges can be defined with ``last``::
3429 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
3431 Note that ``mask`` affects both ``spec`` and ``last``::
3433 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
3434 # matches 10.1.0.0 to 10.2.255.255
3436 Properties can be modified multiple times::
3438 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
3442 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
3447 This section lists supported pattern items and their attributes, if any.
3449 - ``end``: end list of pattern items.
3451 - ``void``: no-op pattern item.
3453 - ``invert``: perform actions when pattern does not match.
3455 - ``any``: match any protocol for the current layer.
3457 - ``num {unsigned}``: number of layers covered.
3459 - ``pf``: match traffic from/to the physical function.
3461 - ``vf``: match traffic from/to a virtual function ID.
3463 - ``id {unsigned}``: VF ID.
3465 - ``phy_port``: match traffic from/to a specific physical port.
3467 - ``index {unsigned}``: physical port index.
3469 - ``port_id``: match traffic from/to a given DPDK port ID.
3471 - ``id {unsigned}``: DPDK port ID.
3473 - ``mark``: match value set in previously matched flow rule using the mark action.
3475 - ``id {unsigned}``: arbitrary integer value.
3477 - ``raw``: match an arbitrary byte string.
3479 - ``relative {boolean}``: look for pattern after the previous item.
3480 - ``search {boolean}``: search pattern from offset (see also limit).
3481 - ``offset {integer}``: absolute or relative offset for pattern.
3482 - ``limit {unsigned}``: search area limit for start of pattern.
3483 - ``pattern {string}``: byte string to look for.
3485 - ``eth``: match Ethernet header.
3487 - ``dst {MAC-48}``: destination MAC.
3488 - ``src {MAC-48}``: source MAC.
3489 - ``type {unsigned}``: EtherType or TPID.
3491 - ``vlan``: match 802.1Q/ad VLAN tag.
3493 - ``tci {unsigned}``: tag control information.
3494 - ``pcp {unsigned}``: priority code point.
3495 - ``dei {unsigned}``: drop eligible indicator.
3496 - ``vid {unsigned}``: VLAN identifier.
3497 - ``inner_type {unsigned}``: inner EtherType or TPID.
3499 - ``ipv4``: match IPv4 header.
3501 - ``tos {unsigned}``: type of service.
3502 - ``ttl {unsigned}``: time to live.
3503 - ``proto {unsigned}``: next protocol ID.
3504 - ``src {ipv4 address}``: source address.
3505 - ``dst {ipv4 address}``: destination address.
3507 - ``ipv6``: match IPv6 header.
3509 - ``tc {unsigned}``: traffic class.
3510 - ``flow {unsigned}``: flow label.
3511 - ``proto {unsigned}``: protocol (next header).
3512 - ``hop {unsigned}``: hop limit.
3513 - ``src {ipv6 address}``: source address.
3514 - ``dst {ipv6 address}``: destination address.
3516 - ``icmp``: match ICMP header.
3518 - ``type {unsigned}``: ICMP packet type.
3519 - ``code {unsigned}``: ICMP packet code.
3521 - ``udp``: match UDP header.
3523 - ``src {unsigned}``: UDP source port.
3524 - ``dst {unsigned}``: UDP destination port.
3526 - ``tcp``: match TCP header.
3528 - ``src {unsigned}``: TCP source port.
3529 - ``dst {unsigned}``: TCP destination port.
3531 - ``sctp``: match SCTP header.
3533 - ``src {unsigned}``: SCTP source port.
3534 - ``dst {unsigned}``: SCTP destination port.
3535 - ``tag {unsigned}``: validation tag.
3536 - ``cksum {unsigned}``: checksum.
3538 - ``vxlan``: match VXLAN header.
3540 - ``vni {unsigned}``: VXLAN identifier.
3542 - ``e_tag``: match IEEE 802.1BR E-Tag header.
3544 - ``grp_ecid_b {unsigned}``: GRP and E-CID base.
3546 - ``nvgre``: match NVGRE header.
3548 - ``tni {unsigned}``: virtual subnet ID.
3550 - ``mpls``: match MPLS header.
3552 - ``label {unsigned}``: MPLS label.
3554 - ``gre``: match GRE header.
3556 - ``protocol {unsigned}``: protocol type.
3558 - ``fuzzy``: fuzzy pattern match, expect faster than default.
3560 - ``thresh {unsigned}``: accuracy threshold.
3562 - ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header.
3564 - ``teid {unsigned}``: tunnel endpoint identifier.
3566 - ``geneve``: match GENEVE header.
3568 - ``vni {unsigned}``: virtual network identifier.
3569 - ``protocol {unsigned}``: protocol type.
3571 - ``vxlan-gpe``: match VXLAN-GPE header.
3573 - ``vni {unsigned}``: VXLAN-GPE identifier.
3575 - ``arp_eth_ipv4``: match ARP header for Ethernet/IPv4.
3577 - ``sha {MAC-48}``: sender hardware address.
3578 - ``spa {ipv4 address}``: sender IPv4 address.
3579 - ``tha {MAC-48}``: target hardware address.
3580 - ``tpa {ipv4 address}``: target IPv4 address.
3582 - ``ipv6_ext``: match presence of any IPv6 extension header.
3584 - ``next_hdr {unsigned}``: next header.
3586 - ``icmp6``: match any ICMPv6 header.
3588 - ``type {unsigned}``: ICMPv6 type.
3589 - ``code {unsigned}``: ICMPv6 code.
3591 - ``icmp6_nd_ns``: match ICMPv6 neighbor discovery solicitation.
3593 - ``target_addr {ipv6 address}``: target address.
3595 - ``icmp6_nd_na``: match ICMPv6 neighbor discovery advertisement.
3597 - ``target_addr {ipv6 address}``: target address.
3599 - ``icmp6_nd_opt``: match presence of any ICMPv6 neighbor discovery option.
3601 - ``type {unsigned}``: ND option type.
3603 - ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery source Ethernet
3604 link-layer address option.
3606 - ``sla {MAC-48}``: source Ethernet LLA.
3608 - ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery target Ethernet
3609 link-layer address option.
3611 - ``tla {MAC-48}``: target Ethernet LLA.
3613 - ``meta``: match application specific metadata.
3615 - ``data {unsigned}``: metadata value.
3620 A list of actions starts after the ``actions`` token in the same fashion as
3621 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
3622 terminated by a mandatory ``end`` action.
3624 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
3625 rte_flow_action_type``).
3627 Dropping all incoming UDPv4 packets can be expressed as follows::
3629 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3632 Several actions have configurable properties which must be specified when
3633 there is no valid default value. For example, ``queue`` requires a target
3636 This rule redirects incoming UDPv4 traffic to queue index 6::
3638 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3639 actions queue index 6 / end
3641 While this one could be rejected by PMDs (unspecified queue index)::
3643 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3646 As defined by *rte_flow*, the list is not ordered, all actions of a given
3647 rule are performed simultaneously. These are equivalent::
3649 queue index 6 / void / mark id 42 / end
3653 void / mark id 42 / queue index 6 / end
3655 All actions in a list should have different types, otherwise only the last
3656 action of a given type is taken into account::
3658 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
3662 drop / drop / drop / end # drop is performed only once
3666 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
3668 Considering they are performed simultaneously, opposite and overlapping
3669 actions can sometimes be combined when the end result is unambiguous::
3671 drop / queue index 6 / end # drop has no effect
3675 queue index 6 / rss queues 6 7 8 / end # queue has no effect
3679 drop / passthru / end # drop has no effect
3681 Note that PMDs may still refuse such combinations.
3686 This section lists supported actions and their attributes, if any.
3688 - ``end``: end list of actions.
3690 - ``void``: no-op action.
3692 - ``passthru``: let subsequent rule process matched packets.
3694 - ``jump``: redirect traffic to group on device.
3696 - ``group {unsigned}``: group to redirect to.
3698 - ``mark``: attach 32 bit value to packets.
3700 - ``id {unsigned}``: 32 bit value to return with packets.
3702 - ``flag``: flag packets.
3704 - ``queue``: assign packets to a given queue index.
3706 - ``index {unsigned}``: queue index to use.
3708 - ``drop``: drop packets (note: passthru has priority).
3710 - ``count``: enable counters for this rule.
3712 - ``rss``: spread packets among several queues.
3714 - ``func {hash function}``: RSS hash function to apply, allowed tokens are
3715 the same as `set_hash_global_config`_.
3717 - ``level {unsigned}``: encapsulation level for ``types``.
3719 - ``types [{RSS hash type} [...]] end``: specific RSS hash types, allowed
3720 tokens are the same as `set_hash_input_set`_, except that an empty list
3721 does not disable RSS but instead requests unspecified "best-effort"
3724 - ``key {string}``: RSS hash key, overrides ``key_len``.
3726 - ``key_len {unsigned}``: RSS hash key length in bytes, can be used in
3727 conjunction with ``key`` to pad or truncate it.
3729 - ``queues [{unsigned} [...]] end``: queue indices to use.
3731 - ``pf``: direct traffic to physical function.
3733 - ``vf``: direct traffic to a virtual function ID.
3735 - ``original {boolean}``: use original VF ID if possible.
3736 - ``id {unsigned}``: VF ID.
3738 - ``phy_port``: direct packets to physical port index.
3740 - ``original {boolean}``: use original port index if possible.
3741 - ``index {unsigned}``: physical port index.
3743 - ``port_id``: direct matching traffic to a given DPDK port ID.
3745 - ``original {boolean}``: use original DPDK port ID if possible.
3746 - ``id {unsigned}``: DPDK port ID.
3748 - ``of_set_mpls_ttl``: OpenFlow's ``OFPAT_SET_MPLS_TTL``.
3750 - ``mpls_ttl``: MPLS TTL.
3752 - ``of_dec_mpls_ttl``: OpenFlow's ``OFPAT_DEC_MPLS_TTL``.
3754 - ``of_set_nw_ttl``: OpenFlow's ``OFPAT_SET_NW_TTL``.
3756 - ``nw_ttl``: IP TTL.
3758 - ``of_dec_nw_ttl``: OpenFlow's ``OFPAT_DEC_NW_TTL``.
3760 - ``of_copy_ttl_out``: OpenFlow's ``OFPAT_COPY_TTL_OUT``.
3762 - ``of_copy_ttl_in``: OpenFlow's ``OFPAT_COPY_TTL_IN``.
3764 - ``of_pop_vlan``: OpenFlow's ``OFPAT_POP_VLAN``.
3766 - ``of_push_vlan``: OpenFlow's ``OFPAT_PUSH_VLAN``.
3768 - ``ethertype``: Ethertype.
3770 - ``of_set_vlan_vid``: OpenFlow's ``OFPAT_SET_VLAN_VID``.
3772 - ``vlan_vid``: VLAN id.
3774 - ``of_set_vlan_pcp``: OpenFlow's ``OFPAT_SET_VLAN_PCP``.
3776 - ``vlan_pcp``: VLAN priority.
3778 - ``of_pop_mpls``: OpenFlow's ``OFPAT_POP_MPLS``.
3780 - ``ethertype``: Ethertype.
3782 - ``of_push_mpls``: OpenFlow's ``OFPAT_PUSH_MPLS``.
3784 - ``ethertype``: Ethertype.
3786 - ``vxlan_encap``: Performs a VXLAN encapsulation, outer layer configuration
3787 is done through `Config VXLAN Encap outer layers`_.
3789 - ``vxlan_decap``: Performs a decapsulation action by stripping all headers of
3790 the VXLAN tunnel network overlay from the matched flow.
3792 - ``nvgre_encap``: Performs a NVGRE encapsulation, outer layer configuration
3793 is done through `Config NVGRE Encap outer layers`_.
3795 - ``nvgre_decap``: Performs a decapsulation action by stripping all headers of
3796 the NVGRE tunnel network overlay from the matched flow.
3798 - ``set_ipv4_src``: Set a new IPv4 source address in the outermost IPv4 header.
3800 - ``ipv4_addr``: New IPv4 source address.
3802 - ``set_ipv4_dst``: Set a new IPv4 destination address in the outermost IPv4
3805 - ``ipv4_addr``: New IPv4 destination address.
3807 - ``set_ipv6_src``: Set a new IPv6 source address in the outermost IPv6 header.
3809 - ``ipv6_addr``: New IPv6 source address.
3811 - ``set_ipv6_dst``: Set a new IPv6 destination address in the outermost IPv6
3814 - ``ipv6_addr``: New IPv6 destination address.
3816 - ``of_set_tp_src``: Set a new source port number in the outermost TCP/UDP
3819 - ``port``: New TCP/UDP source port number.
3821 - ``of_set_tp_dst``: Set a new destination port number in the outermost TCP/UDP
3824 - ``port``: New TCP/UDP destination port number.
3826 - ``mac_swap``: Swap the source and destination MAC addresses in the outermost
3829 - ``dec_ttl``: Performs a decrease TTL value action
3831 - ``set_ttl``: Set TTL value with specificed value
3832 - ``ttl_value {unsigned}``: The new TTL value to be set
3834 - ``set_mac_src``: set source MAC address
3836 - ``mac_addr {MAC-48}``: new source MAC address
3838 - ``set_mac_dst``: set destination MAC address
3840 - ``mac_addr {MAC-48}``: new destination MAC address
3842 Destroying flow rules
3843 ~~~~~~~~~~~~~~~~~~~~~
3845 ``flow destroy`` destroys one or more rules from their rule ID (as returned
3846 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
3847 times as necessary::
3849 flow destroy {port_id} rule {rule_id} [...]
3851 If successful, it will show::
3853 Flow rule #[...] destroyed
3855 It does not report anything for rule IDs that do not exist. The usual error
3856 message is shown when a rule cannot be destroyed::
3858 Caught error type [...] ([...]): [...]
3860 ``flow flush`` destroys all rules on a device and does not take extra
3861 arguments. It is bound to ``rte_flow_flush()``::
3863 flow flush {port_id}
3865 Any errors are reported as above.
3867 Creating several rules and destroying them::
3869 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3870 actions queue index 2 / end
3871 Flow rule #0 created
3872 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3873 actions queue index 3 / end
3874 Flow rule #1 created
3875 testpmd> flow destroy 0 rule 0 rule 1
3876 Flow rule #1 destroyed
3877 Flow rule #0 destroyed
3880 The same result can be achieved using ``flow flush``::
3882 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3883 actions queue index 2 / end
3884 Flow rule #0 created
3885 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3886 actions queue index 3 / end
3887 Flow rule #1 created
3888 testpmd> flow flush 0
3891 Non-existent rule IDs are ignored::
3893 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3894 actions queue index 2 / end
3895 Flow rule #0 created
3896 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3897 actions queue index 3 / end
3898 Flow rule #1 created
3899 testpmd> flow destroy 0 rule 42 rule 10 rule 2
3901 testpmd> flow destroy 0 rule 0
3902 Flow rule #0 destroyed
3908 ``flow query`` queries a specific action of a flow rule having that
3909 ability. Such actions collect information that can be reported using this
3910 command. It is bound to ``rte_flow_query()``::
3912 flow query {port_id} {rule_id} {action}
3914 If successful, it will display either the retrieved data for known actions
3915 or the following message::
3917 Cannot display result for action type [...] ([...])
3919 Otherwise, it will complain either that the rule does not exist or that some
3922 Flow rule #[...] not found
3926 Caught error type [...] ([...]): [...]
3928 Currently only the ``count`` action is supported. This action reports the
3929 number of packets that hit the flow rule and the total number of bytes. Its
3930 output has the following format::
3933 hits_set: [...] # whether "hits" contains a valid value
3934 bytes_set: [...] # whether "bytes" contains a valid value
3935 hits: [...] # number of packets
3936 bytes: [...] # number of bytes
3938 Querying counters for TCPv6 packets redirected to queue 6::
3940 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
3941 actions queue index 6 / count / end
3942 Flow rule #4 created
3943 testpmd> flow query 0 4 count
3954 ``flow list`` lists existing flow rules sorted by priority and optionally
3955 filtered by group identifiers::
3957 flow list {port_id} [group {group_id}] [...]
3959 This command only fails with the following message if the device does not
3964 Output consists of a header line followed by a short description of each
3965 flow rule, one per line. There is no output at all when no flow rules are
3966 configured on the device::
3968 ID Group Prio Attr Rule
3969 [...] [...] [...] [...] [...]
3971 ``Attr`` column flags:
3973 - ``i`` for ``ingress``.
3974 - ``e`` for ``egress``.
3976 Creating several flow rules and listing them::
3978 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3979 actions queue index 6 / end
3980 Flow rule #0 created
3981 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3982 actions queue index 2 / end
3983 Flow rule #1 created
3984 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
3985 actions rss queues 6 7 8 end / end
3986 Flow rule #2 created
3987 testpmd> flow list 0
3988 ID Group Prio Attr Rule
3989 0 0 0 i- ETH IPV4 => QUEUE
3990 1 0 0 i- ETH IPV6 => QUEUE
3991 2 0 5 i- ETH IPV4 UDP => RSS
3994 Rules are sorted by priority (i.e. group ID first, then priority level)::
3996 testpmd> flow list 1
3997 ID Group Prio Attr Rule
3998 0 0 0 i- ETH => COUNT
3999 6 0 500 i- ETH IPV6 TCP => DROP COUNT
4000 5 0 1000 i- ETH IPV6 ICMP => QUEUE
4001 1 24 0 i- ETH IPV4 UDP => QUEUE
4002 4 24 10 i- ETH IPV4 TCP => DROP
4003 3 24 20 i- ETH IPV4 => DROP
4004 2 24 42 i- ETH IPV4 UDP => QUEUE
4005 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
4008 Output can be limited to specific groups::
4010 testpmd> flow list 1 group 0 group 63
4011 ID Group Prio Attr Rule
4012 0 0 0 i- ETH => COUNT
4013 6 0 500 i- ETH IPV6 TCP => DROP COUNT
4014 5 0 1000 i- ETH IPV6 ICMP => QUEUE
4015 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
4018 Toggling isolated mode
4019 ~~~~~~~~~~~~~~~~~~~~~~
4021 ``flow isolate`` can be used to tell the underlying PMD that ingress traffic
4022 must only be injected from the defined flow rules; that no default traffic
4023 is expected outside those rules and the driver is free to assign more
4024 resources to handle them. It is bound to ``rte_flow_isolate()``::
4026 flow isolate {port_id} {boolean}
4028 If successful, enabling or disabling isolated mode shows either::
4030 Ingress traffic on port [...]
4031 is now restricted to the defined flow rules
4035 Ingress traffic on port [...]
4036 is not restricted anymore to the defined flow rules
4038 Otherwise, in case of error::
4040 Caught error type [...] ([...]): [...]
4042 Mainly due to its side effects, PMDs supporting this mode may not have the
4043 ability to toggle it more than once without reinitializing affected ports
4044 first (e.g. by exiting testpmd).
4046 Enabling isolated mode::
4048 testpmd> flow isolate 0 true
4049 Ingress traffic on port 0 is now restricted to the defined flow rules
4052 Disabling isolated mode::
4054 testpmd> flow isolate 0 false
4055 Ingress traffic on port 0 is not restricted anymore to the defined flow rules
4058 Sample QinQ flow rules
4059 ~~~~~~~~~~~~~~~~~~~~~~
4061 Before creating QinQ rule(s) the following commands should be issued to enable QinQ::
4063 testpmd> port stop 0
4064 testpmd> vlan set qinq on 0
4066 The above command sets the inner and outer TPID's to 0x8100.
4068 To change the TPID's the following commands should be used::
4070 testpmd> vlan set outer tpid 0xa100 0
4071 testpmd> vlan set inner tpid 0x9100 0
4072 testpmd> port start 0
4074 Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM.
4078 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 /
4079 vlan tci is 456 / end actions vf id 1 / queue index 0 / end
4080 Flow rule #0 validated
4082 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 /
4083 vlan tci is 456 / end actions vf id 123 / queue index 0 / end
4084 Flow rule #0 created
4086 testpmd> flow list 0
4087 ID Group Prio Attr Rule
4088 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
4090 Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host.
4094 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 /
4095 vlan tci is 654 / end actions pf / queue index 0 / end
4096 Flow rule #1 validated
4098 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 /
4099 vlan tci is 654 / end actions pf / queue index 1 / end
4100 Flow rule #1 created
4102 testpmd> flow list 0
4103 ID Group Prio Attr Rule
4104 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
4105 1 0 0 i- ETH VLAN VLAN=>PF QUEUE
4107 Sample VXLAN encapsulation rule
4108 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
4110 VXLAN encapsulation outer layer has default value pre-configured in testpmd
4111 source code, those can be changed by using the following commands
4113 IPv4 VXLAN outer header::
4115 testpmd> set vxlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src 127.0.0.1
4116 ip-dst 128.0.0.1 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
4117 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
4120 testpmd> set vxlan-with-vlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src
4121 127.0.0.1 ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11
4122 eth-dst 22:22:22:22:22:22
4123 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
4126 IPv6 VXLAN outer header::
4128 testpmd> set vxlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4 ip-src ::1
4129 ip-dst ::2222 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
4130 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
4133 testpmd> set vxlan-with-vlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4
4134 ip-src ::1 ip-dst ::2222 vlan-tci 34 eth-src 11:11:11:11:11:11
4135 eth-dst 22:22:22:22:22:22
4136 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
4139 Sample NVGRE encapsulation rule
4140 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
4142 NVGRE encapsulation outer layer has default value pre-configured in testpmd
4143 source code, those can be changed by using the following commands
4145 IPv4 NVGRE outer header::
4147 testpmd> set nvgre ip-version ipv4 tni 4 ip-src 127.0.0.1 ip-dst 128.0.0.1
4148 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
4149 testpmd> flow create 0 ingress pattern end actions nvgre_encap /
4152 testpmd> set nvgre-with-vlan ip-version ipv4 tni 4 ip-src 127.0.0.1
4153 ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11
4154 eth-dst 22:22:22:22:22:22
4155 testpmd> flow create 0 ingress pattern end actions nvgre_encap /
4158 IPv6 NVGRE outer header::
4160 testpmd> set nvgre ip-version ipv6 tni 4 ip-src ::1 ip-dst ::2222
4161 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
4162 testpmd> flow create 0 ingress pattern end actions nvgre_encap /
4165 testpmd> set nvgre-with-vlan ip-version ipv6 tni 4 ip-src ::1 ip-dst ::2222
4166 vlan-tci 34 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
4167 testpmd> flow create 0 ingress pattern end actions nvgre_encap /
4173 The following sections show functions to load/unload eBPF based filters.
4178 Load an eBPF program as a callback for partciular RX/TX queue::
4180 testpmd> bpf-load rx|tx (portid) (queueid) (load-flags) (bpf-prog-filename)
4182 The available load-flags are:
4184 * ``J``: use JIT generated native code, otherwise BPF interpreter will be used.
4186 * ``M``: assume input parameter is a pointer to rte_mbuf, otherwise assume it is a pointer to first segment's data.
4192 You'll need clang v3.7 or above to build bpf program you'd like to load
4196 .. code-block:: console
4199 clang -O2 -target bpf -c t1.c
4201 Then to load (and JIT compile) t1.o at RX queue 0, port 1::
4203 .. code-block:: console
4205 testpmd> bpf-load rx 1 0 J ./dpdk.org/test/bpf/t1.o
4207 To load (not JITed) t1.o at TX queue 0, port 0::
4209 .. code-block:: console
4211 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o
4216 Unload previously loaded eBPF program for partciular RX/TX queue::
4218 testpmd> bpf-unload rx|tx (portid) (queueid)
4220 For example to unload BPF filter from TX queue 0, port 0:
4222 .. code-block:: console
4224 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o