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
427 Configuration Functions
428 -----------------------
430 The testpmd application can be configured from the runtime as well as from the command-line.
432 This section details the available configuration functions that are available.
436 Configuration changes only become active when forwarding is started/restarted.
441 Reset forwarding to the default configuration::
448 Set the debug verbosity level::
450 testpmd> set verbose (level)
452 Available levels are as following:
454 * ``0`` silent except for error.
455 * ``1`` fully verbose except for Tx packets.
456 * ``2`` fully verbose except for Rx packets.
457 * ``> 2`` fully verbose.
462 Set the log level for a log type::
464 testpmd> set log global|(type) (level)
468 * ``type`` is the log name.
470 * ``level`` is the log level.
472 For example, to change the global log level::
473 testpmd> set log global (level)
475 Regexes can also be used for type. To change log level of user1, user2 and user3::
476 testpmd> set log user[1-3] (level)
481 Set the number of ports used by the application:
485 This is equivalent to the ``--nb-ports`` command-line option.
490 Set the number of cores used by the application::
492 testpmd> set nbcore (num)
494 This is equivalent to the ``--nb-cores`` command-line option.
498 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port.
503 Set the forwarding cores hexadecimal mask::
505 testpmd> set coremask (mask)
507 This is equivalent to the ``--coremask`` command-line option.
511 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding.
516 Set the forwarding ports hexadecimal mask::
518 testpmd> set portmask (mask)
520 This is equivalent to the ``--portmask`` command-line option.
525 Set number of packets per burst::
527 testpmd> set burst (num)
529 This is equivalent to the ``--burst command-line`` option.
531 When retry is enabled, the transmit delay time and number of retries can also be set::
533 testpmd> set burst tx delay (microseconds) retry (num)
538 Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode::
540 testpmd> set txpkts (x[,y]*)
542 Where x[,y]* represents a CSV list of values, without white space.
547 Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes::
549 testpmd> set txsplit (off|on|rand)
553 * ``off`` disable packet copy & split for CSUM mode.
555 * ``on`` split outgoing packet into multiple segments. Size of each segment
556 and number of segments per packet is determined by ``set txpkts`` command
559 * ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments.
564 Set the list of forwarding cores::
566 testpmd> set corelist (x[,y]*)
568 For example, to change the forwarding cores:
570 .. code-block:: console
572 testpmd> set corelist 3,1
573 testpmd> show config fwd
575 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled
576 Logical Core 3 (socket 0) forwards packets on 1 streams:
577 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01
578 Logical Core 1 (socket 0) forwards packets on 1 streams:
579 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
583 The cores are used in the same order as specified on the command line.
588 Set the list of forwarding ports::
590 testpmd> set portlist (x[,y]*)
592 For example, to change the port forwarding:
594 .. code-block:: console
596 testpmd> set portlist 0,2,1,3
597 testpmd> show config fwd
599 io packet forwarding - ports=4 - cores=1 - streams=4
600 Logical Core 3 (socket 0) forwards packets on 4 streams:
601 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01
602 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
603 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03
604 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02
609 Enable/disable tx loopback::
611 testpmd> set tx loopback (port_id) (on|off)
616 set drop enable bit for all queues::
618 testpmd> set all queues drop (port_id) (on|off)
620 set split drop enable (for VF)
621 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
623 set split drop enable bit for VF from PF::
625 testpmd> set vf split drop (port_id) (vf_id) (on|off)
627 set mac antispoof (for VF)
628 ~~~~~~~~~~~~~~~~~~~~~~~~~~
630 Set mac antispoof for a VF from the PF::
632 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off)
637 Enable/disable MACsec offload::
639 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off)
640 testpmd> set macsec offload (port_id) off
645 Configure MACsec secure connection (SC)::
647 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi)
651 The pi argument is ignored for tx.
652 Check the NIC Datasheet for hardware limits.
657 Configure MACsec secure association (SA)::
659 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key)
663 The IDX value must be 0 or 1.
664 Check the NIC Datasheet for hardware limits.
666 set broadcast mode (for VF)
667 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
669 Set broadcast mode for a VF from the PF::
671 testpmd> set vf broadcast (port_id) (vf_id) (on|off)
676 Set the VLAN strip on a port::
678 testpmd> vlan set strip (on|off) (port_id)
683 Set the VLAN strip for a queue on a port::
685 testpmd> vlan set stripq (on|off) (port_id,queue_id)
687 vlan set stripq (for VF)
688 ~~~~~~~~~~~~~~~~~~~~~~~~
690 Set VLAN strip for all queues in a pool for a VF from the PF::
692 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off)
694 vlan set insert (for VF)
695 ~~~~~~~~~~~~~~~~~~~~~~~~
697 Set VLAN insert for a VF from the PF::
699 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id)
701 vlan set tag (for VF)
702 ~~~~~~~~~~~~~~~~~~~~~
704 Set VLAN tag for a VF from the PF::
706 testpmd> set vf vlan tag (port_id) (vf_id) (on|off)
708 vlan set antispoof (for VF)
709 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
711 Set VLAN antispoof for a VF from the PF::
713 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off)
718 Set the VLAN filter on a port::
720 testpmd> vlan set filter (on|off) (port_id)
725 Set the VLAN QinQ (extended queue in queue) on for a port::
727 testpmd> vlan set qinq (on|off) (port_id)
732 Set the inner or outer VLAN TPID for packet filtering on a port::
734 testpmd> vlan set (inner|outer) tpid (value) (port_id)
738 TPID value must be a 16-bit number (value <= 65536).
743 Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID::
745 testpmd> rx_vlan add (vlan_id|all) (port_id)
749 VLAN filter must be set on that port. VLAN ID < 4096.
750 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries
751 in VFTA table. This is important if enabling all vlan_ids.
756 Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID::
758 testpmd> rx_vlan rm (vlan_id|all) (port_id)
763 Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID::
765 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask)
770 Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID::
772 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask)
777 Add a tunnel filter on a port::
779 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \
780 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
781 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
783 The available information categories are:
785 * ``vxlan``: Set tunnel type as VXLAN.
787 * ``nvgre``: Set tunnel type as NVGRE.
789 * ``ipingre``: Set tunnel type as IP-in-GRE.
791 * ``imac-ivlan``: Set filter type as Inner MAC and VLAN.
793 * ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID.
795 * ``imac-tenid``: Set filter type as Inner MAC and tenant ID.
797 * ``imac``: Set filter type as Inner MAC.
799 * ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID.
801 * ``oip``: Set filter type as Outer IP.
803 * ``iip``: Set filter type as Inner IP.
807 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \
808 192.168.2.2 0 ipingre oip 1 1
810 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP.
815 Remove a tunnel filter on a port::
817 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \
818 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
819 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
824 Add an UDP port for VXLAN packet filter on a port::
826 testpmd> rx_vxlan_port add (udp_port) (port_id)
831 Remove an UDP port for VXLAN packet filter on a port::
833 testpmd> rx_vxlan_port rm (udp_port) (port_id)
838 Set hardware insertion of VLAN IDs in packets sent on a port::
840 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer]
842 For example, set a single VLAN ID (5) insertion on port 0::
846 Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1::
854 Set port based hardware insertion of VLAN ID in packets sent on a port::
856 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off)
861 Disable hardware insertion of a VLAN header in packets sent on a port::
863 testpmd> tx_vlan reset (port_id)
868 Select hardware or software calculation of the checksum when
869 transmitting a packet using the ``csum`` forwarding engine::
871 testpmd> csum set (ip|udp|tcp|sctp|outer-ip|outer-udp) (hw|sw) (port_id)
875 * ``ip|udp|tcp|sctp`` always relate to the inner layer.
877 * ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized
878 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are
879 supported). See also the ``csum parse-tunnel`` command.
881 * ``outer-udp`` relates to the outer UDP layer in the case where the packet is recognized
882 as a tunnel packet by the forwarding engine (vxlan, vxlan-gpe are
883 supported). See also the ``csum parse-tunnel`` command.
887 Check the NIC Datasheet for hardware limits.
892 Set RSS queue region span on a port::
894 testpmd> set port (port_id) queue-region region_id (value) \
895 queue_start_index (value) queue_num (value)
897 Set flowtype mapping on a RSS queue region on a port::
899 testpmd> set port (port_id) queue-region region_id (value) flowtype (value)
903 * For the flowtype(pctype) of packet,the specific index for each type has
904 been defined in file i40e_type.h as enum i40e_filter_pctype.
906 Set user priority mapping on a RSS queue region on a port::
908 testpmd> set port (port_id) queue-region UP (value) region_id (value)
910 Flush all queue region related configuration on a port::
912 testpmd> set port (port_id) queue-region flush (on|off)
916 * "on"is just an enable function which server for other configuration,
917 it is for all configuration about queue region from up layer,
918 at first will only keep in DPDK softwarestored in driver,
919 only after "flush on", it commit all configuration to HW.
920 "off" is just clean all configuration about queue region just now,
921 and restore all to DPDK i40e driver default config when start up.
923 Show all queue region related configuration info on a port::
925 testpmd> show port (port_id) queue-region
929 Queue region only support on PF by now, so these command is
930 only for configuration of queue region on PF port.
935 Define how tunneled packets should be handled by the csum forward
938 testpmd> csum parse-tunnel (on|off) (tx_port_id)
940 If enabled, the csum forward engine will try to recognize supported
941 tunnel headers (vxlan, gre, ipip).
943 If disabled, treat tunnel packets as non-tunneled packets (a inner
944 header is handled as a packet payload).
948 The port argument is the TX port like in the ``csum set`` command.
952 Consider a packet in packet like the following::
954 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in
956 * If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
957 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the
958 ``outer-ip|outer-udp`` parameter relates to the outer headers (here ``ipv4_out`` and ``udp_out``).
960 * If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
961 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``.
966 Display tx checksum offload configuration::
968 testpmd> csum show (port_id)
973 Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine::
975 testpmd> tso set (segsize) (port_id)
979 Check the NIC datasheet for hardware limits.
984 Display the status of TCP Segmentation Offload::
986 testpmd> tso show (port_id)
991 Enable or disable GRO in ``csum`` forwarding engine::
993 testpmd> set port <port_id> gro on|off
995 If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4
996 packets received from the given port.
998 If disabled, packets received from the given port won't be performed
999 GRO. By default, GRO is disabled for all ports.
1003 When enable GRO for a port, TCP/IPv4 packets received from the port
1004 will be performed GRO. After GRO, all merged packets have bad
1005 checksums, since the GRO library doesn't re-calculate checksums for
1006 the merged packets. Therefore, if users want the merged packets to
1007 have correct checksums, please select HW IP checksum calculation and
1008 HW TCP checksum calculation for the port which the merged packets are
1014 Display GRO configuration for a given port::
1016 testpmd> show port <port_id> gro
1021 Set the cycle to flush the GROed packets from reassembly tables::
1023 testpmd> set gro flush <cycles>
1025 When enable GRO, the csum forwarding engine performs GRO on received
1026 packets, and the GROed packets are stored in reassembly tables. Users
1027 can use this command to determine when the GROed packets are flushed
1028 from the reassembly tables.
1030 The ``cycles`` is measured in GRO operation times. The csum forwarding
1031 engine flushes the GROed packets from the tables every ``cycles`` GRO
1034 By default, the value of ``cycles`` is 1, which means flush GROed packets
1035 from the reassembly tables as soon as one GRO operation finishes. The value
1036 of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``.
1038 Please note that the large value of ``cycles`` may cause the poor TCP/IP
1039 stack performance. Because the GROed packets are delayed to arrive the
1040 stack, thus causing more duplicated ACKs and TCP retransmissions.
1045 Toggle per-port GSO support in ``csum`` forwarding engine::
1047 testpmd> set port <port_id> gso on|off
1049 If enabled, the csum forwarding engine will perform GSO on supported IPv4
1050 packets, transmitted on the given port.
1052 If disabled, packets transmitted on the given port will not undergo GSO.
1053 By default, GSO is disabled for all ports.
1057 When GSO is enabled on a port, supported IPv4 packets transmitted on that
1058 port undergo GSO. Afterwards, the segmented packets are represented by
1059 multi-segment mbufs; however, the csum forwarding engine doesn't calculation
1060 of checksums for GSO'd segments in SW. As a result, if users want correct
1061 checksums in GSO segments, they should enable HW checksum calculation for
1064 For example, HW checksum calculation for VxLAN GSO'd packets may be enabled
1065 by setting the following options in the csum forwarding engine:
1067 testpmd> csum set outer_ip hw <port_id>
1069 testpmd> csum set ip hw <port_id>
1071 testpmd> csum set tcp hw <port_id>
1073 UDP GSO is the same as IP fragmentation, which treats the UDP header
1074 as the payload and does not modify it during segmentation. That is,
1075 after UDP GSO, only the first output fragment has the original UDP
1076 header. Therefore, users need to enable HW IP checksum calculation
1077 and SW UDP checksum calculation for GSO-enabled ports, if they want
1078 correct checksums for UDP/IPv4 packets.
1083 Set the maximum GSO segment size (measured in bytes), which includes the
1084 packet header and the packet payload for GSO-enabled ports (global)::
1086 testpmd> set gso segsz <length>
1091 Display the status of Generic Segmentation Offload for a given port::
1093 testpmd> show port <port_id> gso
1098 Add an alternative MAC address to a port::
1100 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX)
1105 Remove a MAC address from a port::
1107 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX)
1109 mac_addr add (for VF)
1110 ~~~~~~~~~~~~~~~~~~~~~
1112 Add an alternative MAC address for a VF to a port::
1114 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX)
1119 Set the default MAC address for a port::
1121 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX)
1123 mac_addr set (for VF)
1124 ~~~~~~~~~~~~~~~~~~~~~
1126 Set the MAC address for a VF from the PF::
1128 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX)
1133 Set the forwarding peer address for certain port::
1135 testpmd> set eth-peer (port_id) (perr_addr)
1137 This is equivalent to the ``--eth-peer`` command-line option.
1142 Set the unicast hash filter(s) on/off for a port::
1144 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
1149 Set the promiscuous mode on for a port or for all ports.
1150 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1152 testpmd> set promisc (port_id|all) (on|off)
1157 Set the allmulti mode for a port or for all ports::
1159 testpmd> set allmulti (port_id|all) (on|off)
1161 Same as the ifconfig (8) option. Controls how multicast packets are handled.
1163 set promisc (for VF)
1164 ~~~~~~~~~~~~~~~~~~~~
1166 Set the unicast promiscuous mode for a VF from PF.
1167 It's supported by Intel i40e NICs now.
1168 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1170 testpmd> set vf promisc (port_id) (vf_id) (on|off)
1172 set allmulticast (for VF)
1173 ~~~~~~~~~~~~~~~~~~~~~~~~~
1175 Set the multicast promiscuous mode for a VF from PF.
1176 It's supported by Intel i40e NICs now.
1177 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1179 testpmd> set vf allmulti (port_id) (vf_id) (on|off)
1181 set tx max bandwidth (for VF)
1182 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1184 Set TX max absolute bandwidth (Mbps) for a VF from PF::
1186 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth)
1188 set tc tx min bandwidth (for VF)
1189 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1191 Set all TCs' TX min relative bandwidth (%) for a VF from PF::
1193 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...)
1195 set tc tx max bandwidth (for VF)
1196 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1198 Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF::
1200 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth)
1202 set tc strict link priority mode
1203 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1205 Set some TCs' strict link priority mode on a physical port::
1207 testpmd> set tx strict-link-priority (port_id) (tc_bitmap)
1209 set tc tx min bandwidth
1210 ~~~~~~~~~~~~~~~~~~~~~~~
1212 Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs::
1214 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...)
1219 Set the link flow control parameter on a port::
1221 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1222 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
1223 autoneg (on|off) (port_id)
1227 * ``high_water`` (integer): High threshold value to trigger XOFF.
1229 * ``low_water`` (integer): Low threshold value to trigger XON.
1231 * ``pause_time`` (integer): Pause quota in the Pause frame.
1233 * ``send_xon`` (0/1): Send XON frame.
1235 * ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
1237 * ``autoneg``: Change the auto-negotiation parameter.
1242 Set the priority flow control parameter on a port::
1244 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1245 (pause_time) (priority) (port_id)
1249 * ``high_water`` (integer): High threshold value.
1251 * ``low_water`` (integer): Low threshold value.
1253 * ``pause_time`` (integer): Pause quota in the Pause frame.
1255 * ``priority`` (0-7): VLAN User Priority.
1260 Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
1262 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
1264 For example, to set rx queue 2 on port 0 to mapping 5::
1266 testpmd>set stat_qmap rx 0 2 5
1268 set xstats-hide-zero
1269 ~~~~~~~~~~~~~~~~~~~~
1271 Set the option to hide zero values for xstats display::
1273 testpmd> set xstats-hide-zero on|off
1277 By default, the zero values are displayed for xstats.
1279 set port - rx/tx (for VF)
1280 ~~~~~~~~~~~~~~~~~~~~~~~~~
1282 Set VF receive/transmit from a port::
1284 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
1286 set port - mac address filter (for VF)
1287 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1289 Add/Remove unicast or multicast MAC addr filter for a VF::
1291 testpmd> set port (port_id) vf (vf_id) (mac_addr) \
1292 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
1294 set port - rx mode(for VF)
1295 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1297 Set the VF receive mode of a port::
1299 testpmd> set port (port_id) vf (vf_id) \
1300 rxmode (AUPE|ROPE|BAM|MPE) (on|off)
1302 The available receive modes are:
1304 * ``AUPE``: Accepts untagged VLAN.
1306 * ``ROPE``: Accepts unicast hash.
1308 * ``BAM``: Accepts broadcast packets.
1310 * ``MPE``: Accepts all multicast packets.
1312 set port - tx_rate (for Queue)
1313 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1315 Set TX rate limitation for a queue on a port::
1317 testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1319 set port - tx_rate (for VF)
1320 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
1322 Set TX rate limitation for queues in VF on a port::
1324 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1326 set port - mirror rule
1327 ~~~~~~~~~~~~~~~~~~~~~~
1329 Set pool or vlan type mirror rule for a port::
1331 testpmd> set port (port_id) mirror-rule (rule_id) \
1332 (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1333 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1335 Set link mirror rule for a port::
1337 testpmd> set port (port_id) mirror-rule (rule_id) \
1338 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1340 For example to enable mirror traffic with vlan 0,1 to pool 0::
1342 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1344 reset port - mirror rule
1345 ~~~~~~~~~~~~~~~~~~~~~~~~
1347 Reset a mirror rule for a port::
1349 testpmd> reset port (port_id) mirror-rule (rule_id)
1354 Set the flush on RX streams before forwarding.
1355 The default is flush ``on``.
1356 Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1358 testpmd> set flush_rx off
1363 Set the bypass mode for the lowest port on bypass enabled NIC::
1365 testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1370 Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1372 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1373 mode (normal|bypass|isolate) (port_id)
1377 * ``timeout``: Enable bypass after watchdog timeout.
1379 * ``os_on``: Enable bypass when OS/board is powered on.
1381 * ``os_off``: Enable bypass when OS/board is powered off.
1383 * ``power_on``: Enable bypass when power supply is turned on.
1385 * ``power_off``: Enable bypass when power supply is turned off.
1391 Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1393 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1398 Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1400 testpmd> show bypass config (port_id)
1405 Set link up for a port::
1407 testpmd> set link-up port (port id)
1412 Set link down for a port::
1414 testpmd> set link-down port (port id)
1419 Enable E-tag insertion for a VF on a port::
1421 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1423 Disable E-tag insertion for a VF on a port::
1425 testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1427 Enable/disable E-tag stripping on a port::
1429 testpmd> E-tag set stripping (on|off) port (port_id)
1431 Enable/disable E-tag based forwarding on a port::
1433 testpmd> E-tag set forwarding (on|off) port (port_id)
1435 Add an E-tag forwarding filter on a port::
1437 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1439 Delete an E-tag forwarding filter on a port::
1440 testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1445 Load a dynamic device personalization (DDP) profile and store backup profile::
1447 testpmd> ddp add (port_id) (profile_path[,backup_profile_path])
1452 Delete a dynamic device personalization profile and restore backup profile::
1454 testpmd> ddp del (port_id) (backup_profile_path)
1459 List all items from the ptype mapping table::
1461 testpmd> ptype mapping get (port_id) (valid_only)
1465 * ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0).
1467 Replace a specific or a group of software defined ptype with a new one::
1469 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type)
1473 * ``target``: A specific software ptype or a mask to represent a group of software ptypes.
1475 * ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1).
1477 * ``pkt_type``: The new software ptype to replace the old ones.
1479 Update hardware defined ptype to software defined packet type mapping table::
1481 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype)
1485 * ``hw_ptype``: hardware ptype as the index of the ptype mapping table.
1487 * ``sw_ptype``: software ptype as the value of the ptype mapping table.
1489 Reset ptype mapping table::
1491 testpmd> ptype mapping reset (port_id)
1493 config per port Rx offloading
1494 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1496 Enable or disable a per port Rx offloading on all Rx queues of a port::
1498 testpmd> port config (port_id) rx_offload (offloading) on|off
1500 * ``offloading``: can be any of these offloading capability:
1501 vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro,
1502 qinq_strip, outer_ipv4_cksum, macsec_strip,
1503 header_split, vlan_filter, vlan_extend, jumbo_frame,
1504 crc_strip, scatter, timestamp, security, keep_crc
1506 This command should be run when the port is stopped, or else it will fail.
1508 config per queue Rx offloading
1509 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1511 Enable or disable a per queue Rx offloading only on a specific Rx queue::
1513 testpmd> port (port_id) rxq (queue_id) rx_offload (offloading) on|off
1515 * ``offloading``: can be any of these offloading capability:
1516 vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro,
1517 qinq_strip, outer_ipv4_cksum, macsec_strip,
1518 header_split, vlan_filter, vlan_extend, jumbo_frame,
1519 crc_strip, scatter, timestamp, security, keep_crc
1521 This command should be run when the port is stopped, or else it will fail.
1523 config per port Tx offloading
1524 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1526 Enable or disable a per port Tx offloading on all Tx queues of a port::
1528 testpmd> port config (port_id) tx_offload (offloading) on|off
1530 * ``offloading``: can be any of these offloading capability:
1531 vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum,
1532 sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum,
1533 qinq_insert, vxlan_tnl_tso, gre_tnl_tso,
1534 ipip_tnl_tso, geneve_tnl_tso, macsec_insert,
1535 mt_lockfree, multi_segs, mbuf_fast_free, security
1537 This command should be run when the port is stopped, or else it will fail.
1539 config per queue Tx offloading
1540 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1542 Enable or disable a per queue Tx offloading only on a specific Tx queue::
1544 testpmd> port (port_id) txq (queue_id) tx_offload (offloading) on|off
1546 * ``offloading``: can be any of these offloading capability:
1547 vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum,
1548 sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum,
1549 qinq_insert, vxlan_tnl_tso, gre_tnl_tso,
1550 ipip_tnl_tso, geneve_tnl_tso, macsec_insert,
1551 mt_lockfree, multi_segs, mbuf_fast_free, security
1553 This command should be run when the port is stopped, or else it will fail.
1555 Config VXLAN Encap outer layers
1556 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1558 Configure the outer layer to encapsulate a packet inside a VXLAN tunnel::
1560 set vxlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \
1561 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) eth-src (eth-src) \
1564 set vxlan-with-vlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \
1565 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) vlan-tci (vlan-tci) \
1566 eth-src (eth-src) eth-dst (eth-dst)
1568 Those command will set an internal configuration inside testpmd, any following
1569 flow rule using the action vxlan_encap will use the last configuration set.
1570 To have a different encapsulation header, one of those commands must be called
1571 before the flow rule creation.
1573 Config NVGRE Encap outer layers
1574 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1576 Configure the outer layer to encapsulate a packet inside a NVGRE tunnel::
1578 set nvgre ip-version (ipv4|ipv6) tni (tni) ip-src (ip-src) ip-dst (ip-dst) \
1579 eth-src (eth-src) eth-dst (eth-dst)
1580 set nvgre-with-vlan ip-version (ipv4|ipv6) tni (tni) ip-src (ip-src) \
1581 ip-dst (ip-dst) vlan-tci (vlan-tci) eth-src (eth-src) eth-dst (eth-dst)
1583 Those command will set an internal configuration inside testpmd, any following
1584 flow rule using the action nvgre_encap will use the last configuration set.
1585 To have a different encapsulation header, one of those commands must be called
1586 before the flow rule creation.
1591 The following sections show functions for configuring ports.
1595 Port configuration changes only become active when forwarding is started/restarted.
1600 Attach a port specified by pci address or virtual device args::
1602 testpmd> port attach (identifier)
1604 To attach a new pci device, the device should be recognized by kernel first.
1605 Then it should be moved under DPDK management.
1606 Finally the port can be attached to testpmd.
1608 For example, to move a pci device using ixgbe under DPDK management:
1610 .. code-block:: console
1612 # Check the status of the available devices.
1613 ./usertools/dpdk-devbind.py --status
1615 Network devices using DPDK-compatible driver
1616 ============================================
1619 Network devices using kernel driver
1620 ===================================
1621 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1624 # Bind the device to igb_uio.
1625 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1628 # Recheck the status of the devices.
1629 ./usertools/dpdk-devbind.py --status
1630 Network devices using DPDK-compatible driver
1631 ============================================
1632 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1634 To attach a port created by virtual device, above steps are not needed.
1636 For example, to attach a port whose pci address is 0000:0a:00.0.
1638 .. code-block:: console
1640 testpmd> port attach 0000:0a:00.0
1641 Attaching a new port...
1642 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1643 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1644 EAL: PCI memory mapped at 0x7f83bfa00000
1645 EAL: PCI memory mapped at 0x7f83bfa80000
1646 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1647 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1648 Port 0 is attached. Now total ports is 1
1651 For example, to attach a port created by pcap PMD.
1653 .. code-block:: console
1655 testpmd> port attach net_pcap0
1656 Attaching a new port...
1657 PMD: Initializing pmd_pcap for net_pcap0
1658 PMD: Creating pcap-backed ethdev on numa socket 0
1659 Port 0 is attached. Now total ports is 1
1662 In this case, identifier is ``net_pcap0``.
1663 This identifier format is the same as ``--vdev`` format of DPDK applications.
1665 For example, to re-attach a bonded port which has been previously detached,
1666 the mode and slave parameters must be given.
1668 .. code-block:: console
1670 testpmd> port attach net_bond_0,mode=0,slave=1
1671 Attaching a new port...
1672 EAL: Initializing pmd_bond for net_bond_0
1673 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1674 Port 0 is attached. Now total ports is 1
1681 Detach a specific port::
1683 testpmd> port detach (port_id)
1685 Before detaching a port, the port should be stopped and closed.
1687 For example, to detach a pci device port 0.
1689 .. code-block:: console
1691 testpmd> port stop 0
1694 testpmd> port close 0
1698 testpmd> port detach 0
1700 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1701 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1702 EAL: PCI memory unmapped at 0x7f83bfa00000
1703 EAL: PCI memory unmapped at 0x7f83bfa80000
1707 For example, to detach a virtual device port 0.
1709 .. code-block:: console
1711 testpmd> port stop 0
1714 testpmd> port close 0
1718 testpmd> port detach 0
1720 PMD: Closing pcap ethdev on numa socket 0
1721 Port 'net_pcap0' is detached. Now total ports is 0
1724 To remove a pci device completely from the system, first detach the port from testpmd.
1725 Then the device should be moved under kernel management.
1726 Finally the device can be removed using kernel pci hotplug functionality.
1728 For example, to move a pci device under kernel management:
1730 .. code-block:: console
1732 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1734 ./usertools/dpdk-devbind.py --status
1736 Network devices using DPDK-compatible driver
1737 ============================================
1740 Network devices using kernel driver
1741 ===================================
1742 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1744 To remove a port created by a virtual device, above steps are not needed.
1749 Start all ports or a specific port::
1751 testpmd> port start (port_id|all)
1756 Stop all ports or a specific port::
1758 testpmd> port stop (port_id|all)
1763 Close all ports or a specific port::
1765 testpmd> port close (port_id|all)
1767 port config - queue ring size
1768 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1770 Configure a rx/tx queue ring size::
1772 testpmd> port (port_id) (rxq|txq) (queue_id) ring_size (value)
1774 Only take effect after command that (re-)start the port or command that setup specific queue.
1776 port start/stop queue
1777 ~~~~~~~~~~~~~~~~~~~~~
1779 Start/stop a rx/tx queue on a specific port::
1781 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1783 port config - queue deferred start
1784 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1786 Switch on/off deferred start of a specific port queue::
1788 testpmd> port (port_id) (rxq|txq) (queue_id) deferred_start (on|off)
1791 ~~~~~~~~~~~~~~~~~~~~~
1793 Setup a rx/tx queue on a specific port::
1795 testpmd> port (port_id) (rxq|txq) (queue_id) setup
1797 Only take effect when port is started.
1802 Set the speed and duplex mode for all ports or a specific port::
1804 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1805 duplex (half|full|auto)
1807 port config - queues/descriptors
1808 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1810 Set number of queues/descriptors for rxq, txq, rxd and txd::
1812 testpmd> port config all (rxq|txq|rxd|txd) (value)
1814 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1816 port config - max-pkt-len
1817 ~~~~~~~~~~~~~~~~~~~~~~~~~
1819 Set the maximum packet length::
1821 testpmd> port config all max-pkt-len (value)
1823 This is equivalent to the ``--max-pkt-len`` command-line option.
1825 port config - CRC Strip
1826 ~~~~~~~~~~~~~~~~~~~~~~~
1828 Set hardware CRC stripping on or off for all ports::
1830 testpmd> port config all crc-strip (on|off)
1832 CRC stripping is on by default.
1834 The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option.
1836 port config - scatter
1837 ~~~~~~~~~~~~~~~~~~~~~~~
1839 Set RX scatter mode on or off for all ports::
1841 testpmd> port config all scatter (on|off)
1843 RX scatter mode is off by default.
1845 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1847 port config - RX Checksum
1848 ~~~~~~~~~~~~~~~~~~~~~~~~~
1850 Set hardware RX checksum offload to on or off for all ports::
1852 testpmd> port config all rx-cksum (on|off)
1854 Checksum offload is off by default.
1856 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1861 Set hardware VLAN on or off for all ports::
1863 testpmd> port config all hw-vlan (on|off)
1865 Hardware VLAN is off by default.
1867 The ``on`` option is equivalent to the ``--enable-hw-vlan`` command-line option.
1869 port config - VLAN filter
1870 ~~~~~~~~~~~~~~~~~~~~~~~~~
1872 Set hardware VLAN filter on or off for all ports::
1874 testpmd> port config all hw-vlan-filter (on|off)
1876 Hardware VLAN filter is off by default.
1878 The ``on`` option is equivalent to the ``--enable-hw-vlan-filter`` command-line option.
1880 port config - VLAN strip
1881 ~~~~~~~~~~~~~~~~~~~~~~~~
1883 Set hardware VLAN strip on or off for all ports::
1885 testpmd> port config all hw-vlan-strip (on|off)
1887 Hardware VLAN strip is off by default.
1889 The ``on`` option is equivalent to the ``--enable-hw-vlan-strip`` command-line option.
1891 port config - VLAN extend
1892 ~~~~~~~~~~~~~~~~~~~~~~~~~
1894 Set hardware VLAN extend on or off for all ports::
1896 testpmd> port config all hw-vlan-extend (on|off)
1898 Hardware VLAN extend is off by default.
1900 The ``on`` option is equivalent to the ``--enable-hw-vlan-extend`` command-line option.
1902 port config - Drop Packets
1903 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1905 Set packet drop for packets with no descriptors on or off for all ports::
1907 testpmd> port config all drop-en (on|off)
1909 Packet dropping for packets with no descriptors is off by default.
1911 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1916 Set the RSS (Receive Side Scaling) mode on or off::
1918 testpmd> port config all rss (all|default|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1920 RSS is on by default.
1922 The ``all`` option is equivalent to ip|tcp|udp|sctp|ether.
1923 The ``default`` option enables all supported RSS types reported by device info.
1924 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1926 port config - RSS Reta
1927 ~~~~~~~~~~~~~~~~~~~~~~
1929 Set the RSS (Receive Side Scaling) redirection table::
1931 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1936 Set the DCB mode for an individual port::
1938 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1940 The traffic class should be 4 or 8.
1945 Set the number of packets per burst::
1947 testpmd> port config all burst (value)
1949 This is equivalent to the ``--burst`` command-line option.
1951 port config - Threshold
1952 ~~~~~~~~~~~~~~~~~~~~~~~
1954 Set thresholds for TX/RX queues::
1956 testpmd> port config all (threshold) (value)
1958 Where the threshold type can be:
1960 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1962 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1964 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1966 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1968 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1970 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1972 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1974 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1976 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1978 These threshold options are also available from the command-line.
1983 Set the value of ether-type for E-tag::
1985 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1987 Enable/disable the E-tag support::
1989 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1991 port config pctype mapping
1992 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1994 Reset pctype mapping table::
1996 testpmd> port config (port_id) pctype mapping reset
1998 Update hardware defined pctype to software defined flow type mapping table::
2000 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id)
2004 * ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table.
2006 * ``flow_type_id``: software flow type id as the index of the pctype mapping table.
2008 port config input set
2009 ~~~~~~~~~~~~~~~~~~~~~
2011 Config RSS/FDIR/FDIR flexible payload input set for some pctype::
2012 testpmd> port config (port_id) pctype (pctype_id) \
2013 (hash_inset|fdir_inset|fdir_flx_inset) \
2014 (get|set|clear) field (field_idx)
2016 Clear RSS/FDIR/FDIR flexible payload input set for some pctype::
2017 testpmd> port config (port_id) pctype (pctype_id) \
2018 (hash_inset|fdir_inset|fdir_flx_inset) clear all
2022 * ``pctype_id``: hardware packet classification types.
2023 * ``field_idx``: hardware field index.
2025 port config udp_tunnel_port
2026 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
2028 Add/remove UDP tunnel port for VXLAN/GENEVE tunneling protocols::
2029 testpmd> port config (port_id) udp_tunnel_port add|rm vxlan|geneve (udp_port)
2031 Link Bonding Functions
2032 ----------------------
2034 The Link Bonding functions make it possible to dynamically create and
2035 manage link bonding devices from within testpmd interactive prompt.
2037 create bonded device
2038 ~~~~~~~~~~~~~~~~~~~~
2040 Create a new bonding device::
2042 testpmd> create bonded device (mode) (socket)
2044 For example, to create a bonded device in mode 1 on socket 0::
2046 testpmd> create bonded device 1 0
2047 created new bonded device (port X)
2052 Adds Ethernet device to a Link Bonding device::
2054 testpmd> add bonding slave (slave id) (port id)
2056 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
2058 testpmd> add bonding slave 6 10
2061 remove bonding slave
2062 ~~~~~~~~~~~~~~~~~~~~
2064 Removes an Ethernet slave device from a Link Bonding device::
2066 testpmd> remove bonding slave (slave id) (port id)
2068 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
2070 testpmd> remove bonding slave 6 10
2075 Set the Link Bonding mode of a Link Bonding device::
2077 testpmd> set bonding mode (value) (port id)
2079 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
2081 testpmd> set bonding mode 3 10
2086 Set an Ethernet slave device as the primary device on a Link Bonding device::
2088 testpmd> set bonding primary (slave id) (port id)
2090 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
2092 testpmd> set bonding primary 6 10
2097 Set the MAC address of a Link Bonding device::
2099 testpmd> set bonding mac (port id) (mac)
2101 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
2103 testpmd> set bonding mac 10 00:00:00:00:00:01
2105 set bonding xmit_balance_policy
2106 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2108 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
2110 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
2112 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
2114 testpmd> set bonding xmit_balance_policy 10 l34
2117 set bonding mon_period
2118 ~~~~~~~~~~~~~~~~~~~~~~
2120 Set the link status monitoring polling period in milliseconds for a bonding device.
2122 This adds support for PMD slave devices which do not support link status interrupts.
2123 When the mon_period is set to a value greater than 0 then all PMD's which do not support
2124 link status ISR will be queried every polling interval to check if their link status has changed::
2126 testpmd> set bonding mon_period (port_id) (value)
2128 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
2130 testpmd> set bonding mon_period 5 150
2133 set bonding lacp dedicated_queue
2134 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2136 Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic
2137 when in mode 4 (link-aggregration-802.3ad)::
2139 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable)
2142 set bonding agg_mode
2143 ~~~~~~~~~~~~~~~~~~~~
2145 Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad)::
2147 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable)
2153 Show the current configuration of a Link Bonding device::
2155 testpmd> show bonding config (port id)
2158 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
2159 in balance mode with a transmission policy of layer 2+3::
2161 testpmd> show bonding config 9
2163 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
2165 Active Slaves (3): [1 3 4]
2172 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
2173 This is mainly useful for debugging purposes.
2174 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
2175 and fields that can be accessed.
2180 Display the value of a port register::
2182 testpmd> read reg (port_id) (address)
2184 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
2186 testpmd> read reg 0 0xEE00
2187 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
2192 Display a port register bit field::
2194 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
2196 For example, reading the lowest two bits from the register in the example above::
2198 testpmd> read regfield 0 0xEE00 0 1
2199 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
2204 Display a single port register bit::
2206 testpmd> read regbit (port_id) (address) (bit_x)
2208 For example, reading the lowest bit from the register in the example above::
2210 testpmd> read regbit 0 0xEE00 0
2211 port 0 PCI register at offset 0xEE00: bit 0=1
2216 Set the value of a port register::
2218 testpmd> write reg (port_id) (address) (value)
2220 For example, to clear a register::
2222 testpmd> write reg 0 0xEE00 0x0
2223 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
2228 Set bit field of a port register::
2230 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
2232 For example, writing to the register cleared in the example above::
2234 testpmd> write regfield 0 0xEE00 0 1 2
2235 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
2240 Set single bit value of a port register::
2242 testpmd> write regbit (port_id) (address) (bit_x) (value)
2244 For example, to set the high bit in the register from the example above::
2246 testpmd> write regbit 0 0xEE00 31 1
2247 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
2249 Traffic Metering and Policing
2250 -----------------------------
2252 The following section shows functions for configuring traffic metering and
2253 policing on the ethernet device through the use of generic ethdev API.
2255 show port traffic management capability
2256 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2258 Show traffic metering and policing capability of the port::
2260 testpmd> show port meter cap (port_id)
2262 add port meter profile (srTCM rfc2967)
2263 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2265 Add meter profile (srTCM rfc2697) to the ethernet device::
2267 testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \
2272 * ``profile_id``: ID for the meter profile.
2273 * ``cir``: Committed Information Rate (CIR) (bytes/second).
2274 * ``cbs``: Committed Burst Size (CBS) (bytes).
2275 * ``ebs``: Excess Burst Size (EBS) (bytes).
2277 add port meter profile (trTCM rfc2968)
2278 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2280 Add meter profile (srTCM rfc2698) to the ethernet device::
2282 testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \
2283 (cir) (pir) (cbs) (pbs)
2287 * ``profile_id``: ID for the meter profile.
2288 * ``cir``: Committed information rate (bytes/second).
2289 * ``pir``: Peak information rate (bytes/second).
2290 * ``cbs``: Committed burst size (bytes).
2291 * ``pbs``: Peak burst size (bytes).
2293 add port meter profile (trTCM rfc4115)
2294 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2296 Add meter profile (trTCM rfc4115) to the ethernet device::
2298 testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \
2299 (cir) (eir) (cbs) (ebs)
2303 * ``profile_id``: ID for the meter profile.
2304 * ``cir``: Committed information rate (bytes/second).
2305 * ``eir``: Excess information rate (bytes/second).
2306 * ``cbs``: Committed burst size (bytes).
2307 * ``ebs``: Excess burst size (bytes).
2309 delete port meter profile
2310 ~~~~~~~~~~~~~~~~~~~~~~~~~
2312 Delete meter profile from the ethernet device::
2314 testpmd> del port meter profile (port_id) (profile_id)
2319 Create new meter object for the ethernet device::
2321 testpmd> create port meter (port_id) (mtr_id) (profile_id) \
2322 (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \
2323 (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\
2328 * ``mtr_id``: meter object ID.
2329 * ``profile_id``: ID for the meter profile.
2330 * ``meter_enable``: When this parameter has a non-zero value, the meter object
2331 gets enabled at the time of creation, otherwise remains disabled.
2332 * ``g_action``: Policer action for the packet with green color.
2333 * ``y_action``: Policer action for the packet with yellow color.
2334 * ``r_action``: Policer action for the packet with red color.
2335 * ``stats_mask``: Mask of statistics counter types to be enabled for the
2337 * ``shared``: When this parameter has a non-zero value, the meter object is
2338 shared by multiple flows. Otherwise, meter object is used by single flow.
2339 * ``use_pre_meter_color``: When this parameter has a non-zero value, the
2340 input color for the current meter object is determined by the latest meter
2341 object in the same flow. Otherwise, the current meter object uses the
2342 *dscp_table* to determine the input color.
2343 * ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input
2344 color, 0 <= x <= 63.
2349 Enable meter for the ethernet device::
2351 testpmd> enable port meter (port_id) (mtr_id)
2356 Disable meter for the ethernet device::
2358 testpmd> disable port meter (port_id) (mtr_id)
2363 Delete meter for the ethernet device::
2365 testpmd> del port meter (port_id) (mtr_id)
2367 Set port meter profile
2368 ~~~~~~~~~~~~~~~~~~~~~~
2370 Set meter profile for the ethernet device::
2372 testpmd> set port meter profile (port_id) (mtr_id) (profile_id)
2374 set port meter dscp table
2375 ~~~~~~~~~~~~~~~~~~~~~~~~~
2377 Set meter dscp table for the ethernet device::
2379 testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \
2380 (dscp_tbl_entry1)...(dscp_tbl_entry63)]
2382 set port meter policer action
2383 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2385 Set meter policer action for the ethernet device::
2387 testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \
2388 (action0) [(action1) (action1)]
2392 * ``action_mask``: Bit mask indicating which policer actions need to be
2393 updated. One or more policer actions can be updated in a single function
2394 invocation. To update the policer action associated with color C, bit
2395 (1 << C) needs to be set in *action_mask* and element at position C
2396 in the *actions* array needs to be valid.
2397 * ``actionx``: Policer action for the color x,
2398 RTE_MTR_GREEN <= x < RTE_MTR_COLORS
2400 set port meter stats mask
2401 ~~~~~~~~~~~~~~~~~~~~~~~~~
2403 Set meter stats mask for the ethernet device::
2405 testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask)
2409 * ``stats_mask``: Bit mask indicating statistics counter types to be enabled.
2411 show port meter stats
2412 ~~~~~~~~~~~~~~~~~~~~~
2414 Show meter stats of the ethernet device::
2416 testpmd> show port meter stats (port_id) (mtr_id) (clear)
2420 * ``clear``: Flag that indicates whether the statistics counters should
2421 be cleared (i.e. set to zero) immediately after they have been read or not.
2426 The following section shows functions for configuring traffic management on
2427 on the ethernet device through the use of generic TM API.
2429 show port traffic management capability
2430 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2432 Show traffic management capability of the port::
2434 testpmd> show port tm cap (port_id)
2436 show port traffic management capability (hierarchy level)
2437 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2439 Show traffic management hierarchy level capability of the port::
2441 testpmd> show port tm level cap (port_id) (level_id)
2443 show port traffic management capability (hierarchy node level)
2444 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2446 Show the traffic management hierarchy node capability of the port::
2448 testpmd> show port tm node cap (port_id) (node_id)
2450 show port traffic management hierarchy node type
2451 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2453 Show the port traffic management hierarchy node type::
2455 testpmd> show port tm node type (port_id) (node_id)
2457 show port traffic management hierarchy node stats
2458 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2460 Show the port traffic management hierarchy node statistics::
2462 testpmd> show port tm node stats (port_id) (node_id) (clear)
2466 * ``clear``: When this parameter has a non-zero value, the statistics counters
2467 are cleared (i.e. set to zero) immediately after they have been read,
2468 otherwise the statistics counters are left untouched.
2470 Add port traffic management private shaper profile
2471 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2473 Add the port traffic management private shaper profile::
2475 testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \
2476 (tb_rate) (tb_size) (packet_length_adjust)
2480 * ``shaper_profile id``: Shaper profile ID for the new profile.
2481 * ``tb_rate``: Token bucket rate (bytes per second).
2482 * ``tb_size``: Token bucket size (bytes).
2483 * ``packet_length_adjust``: The value (bytes) to be added to the length of
2484 each packet for the purpose of shaping. This parameter value can be used to
2485 correct the packet length with the framing overhead bytes that are consumed
2488 Delete port traffic management private shaper profile
2489 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2491 Delete the port traffic management private shaper::
2493 testpmd> del port tm node shaper profile (port_id) (shaper_profile_id)
2497 * ``shaper_profile id``: Shaper profile ID that needs to be deleted.
2499 Add port traffic management shared shaper
2500 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2502 Create the port traffic management shared shaper::
2504 testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \
2509 * ``shared_shaper_id``: Shared shaper ID to be created.
2510 * ``shaper_profile id``: Shaper profile ID for shared shaper.
2512 Set port traffic management shared shaper
2513 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2515 Update the port traffic management shared shaper::
2517 testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \
2522 * ``shared_shaper_id``: Shared shaper ID to be update.
2523 * ``shaper_profile id``: Shaper profile ID for shared shaper.
2525 Delete port traffic management shared shaper
2526 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2528 Delete the port traffic management shared shaper::
2530 testpmd> del port tm node shared shaper (port_id) (shared_shaper_id)
2534 * ``shared_shaper_id``: Shared shaper ID to be deleted.
2536 Set port traffic management hiearchy node private shaper
2537 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2539 set the port traffic management hierarchy node private shaper::
2541 testpmd> set port tm node shaper profile (port_id) (node_id) \
2546 * ``shaper_profile id``: Private shaper profile ID to be enabled on the
2549 Add port traffic management WRED profile
2550 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2552 Create a new WRED profile::
2554 testpmd> add port tm node wred profile (port_id) (wred_profile_id) \
2555 (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \
2556 (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \
2557 (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r)
2561 * ``wred_profile id``: Identifier for the newly create WRED profile
2562 * ``color_g``: Packet color (green)
2563 * ``min_th_g``: Minimum queue threshold for packet with green color
2564 * ``max_th_g``: Minimum queue threshold for packet with green color
2565 * ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp)
2566 * ``wq_log2_g``: Negated log2 of queue weight (wq)
2567 * ``color_y``: Packet color (yellow)
2568 * ``min_th_y``: Minimum queue threshold for packet with yellow color
2569 * ``max_th_y``: Minimum queue threshold for packet with yellow color
2570 * ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp)
2571 * ``wq_log2_y``: Negated log2 of queue weight (wq)
2572 * ``color_r``: Packet color (red)
2573 * ``min_th_r``: Minimum queue threshold for packet with yellow color
2574 * ``max_th_r``: Minimum queue threshold for packet with yellow color
2575 * ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp)
2576 * ``wq_log2_r``: Negated log2 of queue weight (wq)
2578 Delete port traffic management WRED profile
2579 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2581 Delete the WRED profile::
2583 testpmd> del port tm node wred profile (port_id) (wred_profile_id)
2585 Add port traffic management hierarchy nonleaf node
2586 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2588 Add nonleaf node to port traffic management hiearchy::
2590 testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \
2591 (priority) (weight) (level_id) (shaper_profile_id) \
2592 (n_sp_priorities) (stats_mask) (n_shared_shapers) \
2593 [(shared_shaper_0) (shared_shaper_1) ...] \
2597 * ``parent_node_id``: Node ID of the parent.
2598 * ``priority``: Node priority (highest node priority is zero). This is used by
2599 the SP algorithm running on the parent node for scheduling this node.
2600 * ``weight``: Node weight (lowest weight is one). The node weight is relative
2601 to the weight sum of all siblings that have the same priority. It is used by
2602 the WFQ algorithm running on the parent node for scheduling this node.
2603 * ``level_id``: Hiearchy level of the node.
2604 * ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2606 * ``n_sp_priorities``: Number of strict priorities.
2607 * ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2608 * ``n_shared_shapers``: Number of shared shapers.
2609 * ``shared_shaper_id``: Shared shaper id.
2611 Add port traffic management hierarchy leaf node
2612 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2614 Add leaf node to port traffic management hiearchy::
2616 testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \
2617 (priority) (weight) (level_id) (shaper_profile_id) \
2618 (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \
2619 [(shared_shaper_id) (shared_shaper_id) ...] \
2623 * ``parent_node_id``: Node ID of the parent.
2624 * ``priority``: Node priority (highest node priority is zero). This is used by
2625 the SP algorithm running on the parent node for scheduling this node.
2626 * ``weight``: Node weight (lowest weight is one). The node weight is relative
2627 to the weight sum of all siblings that have the same priority. It is used by
2628 the WFQ algorithm running on the parent node for scheduling this node.
2629 * ``level_id``: Hiearchy level of the node.
2630 * ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2632 * ``cman_mode``: Congestion management mode to be enabled for this node.
2633 * ``wred_profile_id``: WRED profile id to be enabled for this node.
2634 * ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2635 * ``n_shared_shapers``: Number of shared shapers.
2636 * ``shared_shaper_id``: Shared shaper id.
2638 Delete port traffic management hierarchy node
2639 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2641 Delete node from port traffic management hiearchy::
2643 testpmd> del port tm node (port_id) (node_id)
2645 Update port traffic management hierarchy parent node
2646 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2648 Update port traffic management hierarchy parent node::
2650 testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \
2653 This function can only be called after the hierarchy commit invocation. Its
2654 success depends on the port support for this operation, as advertised through
2655 the port capability set. This function is valid for all nodes of the traffic
2656 management hierarchy except root node.
2658 Suspend port traffic management hierarchy node
2659 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2661 testpmd> suspend port tm node (port_id) (node_id)
2663 Resume port traffic management hierarchy node
2664 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2666 testpmd> resume port tm node (port_id) (node_id)
2668 Commit port traffic management hierarchy
2669 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2671 Commit the traffic management hierarchy on the port::
2673 testpmd> port tm hierarchy commit (port_id) (clean_on_fail)
2677 * ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function
2678 call failure. On the other hand, hierarchy is preserved when this parameter
2681 Set port traffic management mark VLAN dei
2682 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2684 Enables/Disables the traffic management marking on the port for VLAN packets::
2686 testpmd> set port tm mark vlan_dei <port_id> <green> <yellow> <red>
2690 * ``port_id``: The port which on which VLAN packets marked as ``green`` or
2691 ``yellow`` or ``red`` will have dei bit enabled
2693 * ``green`` enable 1, disable 0 marking for dei bit of VLAN packets marked as green
2695 * ``yellow`` enable 1, disable 0 marking for dei bit of VLAN packets marked as yellow
2697 * ``red`` enable 1, disable 0 marking for dei bit of VLAN packets marked as red
2699 Set port traffic management mark IP dscp
2700 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2702 Enables/Disables the traffic management marking on the port for IP dscp packets::
2704 testpmd> set port tm mark ip_dscp <port_id> <green> <yellow> <red>
2708 * ``port_id``: The port which on which IP packets marked as ``green`` or
2709 ``yellow`` or ``red`` will have IP dscp bits updated
2711 * ``green`` enable 1, disable 0 marking IP dscp to low drop precedence for green packets
2713 * ``yellow`` enable 1, disable 0 marking IP dscp to medium drop precedence for yellow packets
2715 * ``red`` enable 1, disable 0 marking IP dscp to high drop precedence for red packets
2717 Set port traffic management mark IP ecn
2718 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2720 Enables/Disables the traffic management marking on the port for IP ecn packets::
2722 testpmd> set port tm mark ip_ecn <port_id> <green> <yellow> <red>
2726 * ``port_id``: The port which on which IP packets marked as ``green`` or
2727 ``yellow`` or ``red`` will have IP ecn bits updated
2729 * ``green`` enable 1, disable 0 marking IP ecn for green marked packets with ecn of 2'b01 or 2'b10
2730 to ecn of 2'b11 when IP is caring TCP or SCTP
2732 * ``yellow`` enable 1, disable 0 marking IP ecn for yellow marked packets with ecn of 2'b01 or 2'b10
2733 to ecn of 2'b11 when IP is caring TCP or SCTP
2735 * ``red`` enable 1, disable 0 marking IP ecn for yellow marked packets with ecn of 2'b01 or 2'b10
2736 to ecn of 2'b11 when IP is caring TCP or SCTP
2738 Set port traffic management default hierarchy (softnic forwarding mode)
2739 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2741 set the traffic management default hierarchy on the port::
2743 testpmd> set port tm hierarchy default (port_id)
2748 This section details the available filter functions that are available.
2750 Note these functions interface the deprecated legacy filtering framework,
2751 superseded by *rte_flow*. See `Flow rules management`_.
2754 ~~~~~~~~~~~~~~~~~~~~
2756 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
2758 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
2759 ethertype (ether_type) (drop|fwd) queue (queue_id)
2761 The available information parameters are:
2763 * ``port_id``: The port which the Ethertype filter assigned on.
2765 * ``mac_addr``: Compare destination mac address.
2767 * ``mac_ignr``: Ignore destination mac address match.
2769 * ``mac_address``: Destination mac address to match.
2771 * ``ether_type``: The EtherType value want to match,
2772 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
2774 * ``queue_id``: The receive queue associated with this EtherType filter.
2775 It is meaningless when deleting or dropping.
2777 Example, to add/remove an ethertype filter rule::
2779 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
2780 ethertype 0x0806 fwd queue 3
2782 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
2783 ethertype 0x0806 fwd queue 3
2788 Add or delete a 2-tuple filter,
2789 which identifies packets by specific protocol and destination TCP/UDP port
2790 and forwards packets into one of the receive queues::
2792 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
2793 protocol (protocol_value) mask (mask_value) \
2794 tcp_flags (tcp_flags_value) priority (prio_value) \
2797 The available information parameters are:
2799 * ``port_id``: The port which the 2-tuple filter assigned on.
2801 * ``dst_port_value``: Destination port in L4.
2803 * ``protocol_value``: IP L4 protocol.
2805 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
2807 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
2809 * ``prio_value``: Priority of this filter.
2811 * ``queue_id``: The receive queue associated with this 2-tuple filter.
2813 Example, to add/remove an 2tuple filter rule::
2815 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
2816 tcp_flags 0x02 priority 3 queue 3
2818 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
2819 tcp_flags 0x02 priority 3 queue 3
2824 Add or delete a 5-tuple filter,
2825 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
2826 and routes packets into one of the receive queues::
2828 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
2829 (src_address) dst_port (dst_port_value) \
2830 src_port (src_port_value) protocol (protocol_value) \
2831 mask (mask_value) tcp_flags (tcp_flags_value) \
2832 priority (prio_value) queue (queue_id)
2834 The available information parameters are:
2836 * ``port_id``: The port which the 5-tuple filter assigned on.
2838 * ``dst_address``: Destination IP address.
2840 * ``src_address``: Source IP address.
2842 * ``dst_port_value``: TCP/UDP destination port.
2844 * ``src_port_value``: TCP/UDP source port.
2846 * ``protocol_value``: L4 protocol.
2848 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
2850 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
2852 * ``prio_value``: The priority of this filter.
2854 * ``queue_id``: The receive queue associated with this 5-tuple filter.
2856 Example, to add/remove an 5tuple filter rule::
2858 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2859 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2860 flags 0x0 priority 3 queue 3
2862 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2863 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2864 flags 0x0 priority 3 queue 3
2869 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
2871 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
2873 The available information parameters are:
2875 * ``port_id``: The port which the SYN filter assigned on.
2877 * ``high``: This SYN filter has higher priority than other filters.
2879 * ``low``: This SYN filter has lower priority than other filters.
2881 * ``queue_id``: The receive queue associated with this SYN filter
2885 testpmd> syn_filter 0 add priority high queue 3
2890 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
2891 and routed into one of the receive queues::
2893 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
2894 mask (mask_value) priority (prio_value) queue (queue_id)
2896 The available information parameters are:
2898 * ``port_id``: The port which the Flex filter is assigned on.
2900 * ``len_value``: Filter length in bytes, no greater than 128.
2902 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
2904 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
2906 * ``prio_value``: The priority of this filter.
2908 * ``queue_id``: The receive queue associated with this Flex filter.
2912 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
2913 mask 000C priority 3 queue 3
2915 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
2916 mask 000C priority 3 queue 3
2919 .. _testpmd_flow_director:
2921 flow_director_filter
2922 ~~~~~~~~~~~~~~~~~~~~
2924 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
2926 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
2927 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
2929 * Perfect match filters.
2930 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2931 The masked fields are for IP flow.
2933 * Signature filters.
2934 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
2936 * Perfect-mac-vlan match filters.
2937 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2938 The masked fields are for MAC VLAN flow.
2940 * Perfect-tunnel match filters.
2941 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2942 The masked fields are for tunnel flow.
2944 * Perfect-raw-flow-type match filters.
2945 The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet.
2946 The masked fields are specified by input sets.
2948 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
2949 per flow type and the flexible payload.
2951 The Flow Director can also mask out parts of all of these fields so that filters
2952 are only applied to certain fields or parts of the fields.
2954 Note that for raw flow type mode the source and destination fields in the
2955 raw packet buffer need to be presented in a reversed order with respect
2956 to the expected received packets.
2957 For example: IP source and destination addresses or TCP/UDP/SCTP
2958 source and destination ports
2960 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
2962 # Commands to add flow director filters of different flow types::
2964 flow_director_filter (port_id) mode IP (add|del|update) \
2965 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
2966 src (src_ip_address) dst (dst_ip_address) \
2967 tos (tos_value) proto (proto_value) ttl (ttl_value) \
2968 vlan (vlan_value) flexbytes (flexbytes_value) \
2969 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
2972 flow_director_filter (port_id) mode IP (add|del|update) \
2973 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
2974 src (src_ip_address) (src_port) \
2975 dst (dst_ip_address) (dst_port) \
2976 tos (tos_value) ttl (ttl_value) \
2977 vlan (vlan_value) flexbytes (flexbytes_value) \
2978 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
2981 flow_director_filter (port_id) mode IP (add|del|update) \
2982 flow (ipv4-sctp|ipv6-sctp) \
2983 src (src_ip_address) (src_port) \
2984 dst (dst_ip_address) (dst_port) \
2985 tos (tos_value) ttl (ttl_value) \
2986 tag (verification_tag) vlan (vlan_value) \
2987 flexbytes (flexbytes_value) (drop|fwd) \
2988 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
2990 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
2991 ether (ethertype) flexbytes (flexbytes_value) \
2992 (drop|fwd) pf|vf(vf_id) queue (queue_id)
2995 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
2996 mac (mac_address) vlan (vlan_value) \
2997 flexbytes (flexbytes_value) (drop|fwd) \
2998 queue (queue_id) fd_id (fd_id_value)
3000 flow_director_filter (port_id) mode Tunnel (add|del|update) \
3001 mac (mac_address) vlan (vlan_value) \
3002 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
3003 flexbytes (flexbytes_value) (drop|fwd) \
3004 queue (queue_id) fd_id (fd_id_value)
3006 flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \
3007 (drop|fwd) queue (queue_id) fd_id (fd_id_value) \
3008 packet (packet file name)
3010 For example, to add an ipv4-udp flow type filter::
3012 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
3013 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
3014 fwd pf queue 1 fd_id 1
3016 For example, add an ipv4-other flow type filter::
3018 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
3019 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
3020 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
3025 Flush all flow director filters on a device::
3027 testpmd> flush_flow_director (port_id)
3029 Example, to flush all flow director filter on port 0::
3031 testpmd> flush_flow_director 0
3036 Set flow director's input masks::
3038 flow_director_mask (port_id) mode IP vlan (vlan_value) \
3039 src_mask (ipv4_src) (ipv6_src) (src_port) \
3040 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
3042 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
3044 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
3045 mac (mac_value) tunnel-type (tunnel_type_value) \
3046 tunnel-id (tunnel_id_value)
3048 Example, to set flow director mask on port 0::
3050 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
3051 src_mask 255.255.255.255 \
3052 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
3053 dst_mask 255.255.255.255 \
3054 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
3056 flow_director_flex_mask
3057 ~~~~~~~~~~~~~~~~~~~~~~~
3059 set masks of flow director's flexible payload based on certain flow type::
3061 testpmd> flow_director_flex_mask (port_id) \
3062 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
3063 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
3064 l2_payload|all) (mask)
3066 Example, to set flow director's flex mask for all flow type on port 0::
3068 testpmd> flow_director_flex_mask 0 flow all \
3069 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
3072 flow_director_flex_payload
3073 ~~~~~~~~~~~~~~~~~~~~~~~~~~
3075 Configure flexible payload selection::
3077 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
3079 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
3081 testpmd> flow_director_flex_payload 0 l4 \
3082 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
3084 get_sym_hash_ena_per_port
3085 ~~~~~~~~~~~~~~~~~~~~~~~~~
3087 Get symmetric hash enable configuration per port::
3089 get_sym_hash_ena_per_port (port_id)
3091 For example, to get symmetric hash enable configuration of port 1::
3093 testpmd> get_sym_hash_ena_per_port 1
3095 set_sym_hash_ena_per_port
3096 ~~~~~~~~~~~~~~~~~~~~~~~~~
3098 Set symmetric hash enable configuration per port to enable or disable::
3100 set_sym_hash_ena_per_port (port_id) (enable|disable)
3102 For example, to set symmetric hash enable configuration of port 1 to enable::
3104 testpmd> set_sym_hash_ena_per_port 1 enable
3106 get_hash_global_config
3107 ~~~~~~~~~~~~~~~~~~~~~~
3109 Get the global configurations of hash filters::
3111 get_hash_global_config (port_id)
3113 For example, to get the global configurations of hash filters of port 1::
3115 testpmd> get_hash_global_config 1
3117 set_hash_global_config
3118 ~~~~~~~~~~~~~~~~~~~~~~
3120 Set the global configurations of hash filters::
3122 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
3123 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
3124 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \
3127 For example, to enable simple_xor for flow type of ipv6 on port 2::
3129 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
3134 Set the input set for hash::
3136 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
3137 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
3138 l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
3139 ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
3140 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
3141 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
3142 fld-8th|none) (select|add)
3144 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
3146 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
3151 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
3152 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
3154 Set the input set for flow director::
3156 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
3157 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
3158 l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
3159 ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
3160 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
3161 sctp-dst-port|sctp-veri-tag|none) (select|add)
3163 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
3165 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
3170 Set different GRE key length for input set::
3172 global_config (port_id) gre-key-len (number in bytes)
3174 For example to set GRE key length for input set to 4 bytes on port 0::
3176 testpmd> global_config 0 gre-key-len 4
3179 .. _testpmd_rte_flow:
3181 Flow rules management
3182 ---------------------
3184 Control of the generic flow API (*rte_flow*) is fully exposed through the
3185 ``flow`` command (validation, creation, destruction, queries and operation
3188 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
3189 features simultaneously may cause undefined side-effects and is therefore
3195 Because the ``flow`` command uses dynamic tokens to handle the large number
3196 of possible flow rules combinations, its behavior differs slightly from
3197 other commands, in particular:
3199 - Pressing *?* or the *<tab>* key displays contextual help for the current
3200 token, not that of the entire command.
3202 - Optional and repeated parameters are supported (provided they are listed
3203 in the contextual help).
3205 The first parameter stands for the operation mode. Possible operations and
3206 their general syntax are described below. They are covered in detail in the
3209 - Check whether a flow rule can be created::
3211 flow validate {port_id}
3212 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3213 pattern {item} [/ {item} [...]] / end
3214 actions {action} [/ {action} [...]] / end
3216 - Create a flow rule::
3218 flow create {port_id}
3219 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3220 pattern {item} [/ {item} [...]] / end
3221 actions {action} [/ {action} [...]] / end
3223 - Destroy specific flow rules::
3225 flow destroy {port_id} rule {rule_id} [...]
3227 - Destroy all flow rules::
3229 flow flush {port_id}
3231 - Query an existing flow rule::
3233 flow query {port_id} {rule_id} {action}
3235 - List existing flow rules sorted by priority, filtered by group
3238 flow list {port_id} [group {group_id}] [...]
3240 - Restrict ingress traffic to the defined flow rules::
3242 flow isolate {port_id} {boolean}
3244 Validating flow rules
3245 ~~~~~~~~~~~~~~~~~~~~~
3247 ``flow validate`` reports whether a flow rule would be accepted by the
3248 underlying device in its current state but stops short of creating it. It is
3249 bound to ``rte_flow_validate()``::
3251 flow validate {port_id}
3252 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3253 pattern {item} [/ {item} [...]] / end
3254 actions {action} [/ {action} [...]] / end
3256 If successful, it will show::
3260 Otherwise it will show an error message of the form::
3262 Caught error type [...] ([...]): [...]
3264 This command uses the same parameters as ``flow create``, their format is
3265 described in `Creating flow rules`_.
3267 Check whether redirecting any Ethernet packet received on port 0 to RX queue
3268 index 6 is supported::
3270 testpmd> flow validate 0 ingress pattern eth / end
3271 actions queue index 6 / end
3275 Port 0 does not support TCPv6 rules::
3277 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
3279 Caught error type 9 (specific pattern item): Invalid argument
3285 ``flow create`` validates and creates the specified flow rule. It is bound
3286 to ``rte_flow_create()``::
3288 flow create {port_id}
3289 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3290 pattern {item} [/ {item} [...]] / end
3291 actions {action} [/ {action} [...]] / end
3293 If successful, it will return a flow rule ID usable with other commands::
3295 Flow rule #[...] created
3297 Otherwise it will show an error message of the form::
3299 Caught error type [...] ([...]): [...]
3301 Parameters describe in the following order:
3303 - Attributes (*group*, *priority*, *ingress*, *egress*, *transfer* tokens).
3304 - A matching pattern, starting with the *pattern* token and terminated by an
3306 - Actions, starting with the *actions* token and terminated by an *end*
3309 These translate directly to *rte_flow* objects provided as-is to the
3310 underlying functions.
3312 The shortest valid definition only comprises mandatory tokens::
3314 testpmd> flow create 0 pattern end actions end
3316 Note that PMDs may refuse rules that essentially do nothing such as this
3319 **All unspecified object values are automatically initialized to 0.**
3324 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
3325 specified before the ``pattern`` token.
3327 - ``group {group id}``: priority group.
3328 - ``priority {level}``: priority level within group.
3329 - ``ingress``: rule applies to ingress traffic.
3330 - ``egress``: rule applies to egress traffic.
3331 - ``transfer``: apply rule directly to endpoints found in pattern.
3333 Each instance of an attribute specified several times overrides the previous
3334 value as shown below (group 4 is used)::
3336 testpmd> flow create 0 group 42 group 24 group 4 [...]
3338 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
3340 While not specifying a direction is an error, some rules may allow both
3343 Most rules affect RX therefore contain the ``ingress`` token::
3345 testpmd> flow create 0 ingress pattern [...]
3350 A matching pattern starts after the ``pattern`` token. It is made of pattern
3351 items and is terminated by a mandatory ``end`` item.
3353 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
3354 rte_flow_item_type``).
3356 The ``/`` token is used as a separator between pattern items as shown
3359 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
3361 Note that protocol items like these must be stacked from lowest to highest
3362 layer to make sense. For instance, the following rule is either invalid or
3363 unlikely to match any packet::
3365 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
3367 More information on these restrictions can be found in the *rte_flow*
3370 Several items support additional specification structures, for example
3371 ``ipv4`` allows specifying source and destination addresses as follows::
3373 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
3374 dst is 10.2.0.0 / end [...]
3376 This rule matches all IPv4 traffic with the specified properties.
3378 In this example, ``src`` and ``dst`` are field names of the underlying
3379 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
3380 in a similar fashion.
3382 The ``is`` token means that the subsequent value must be matched exactly,
3383 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
3384 accordingly. Possible assignment tokens are:
3386 - ``is``: match value perfectly (with full bit-mask).
3387 - ``spec``: match value according to configured bit-mask.
3388 - ``last``: specify upper bound to establish a range.
3389 - ``mask``: specify bit-mask with relevant bits set to one.
3390 - ``prefix``: generate bit-mask from a prefix length.
3392 These yield identical results::
3394 ipv4 src is 10.1.1.1
3398 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
3402 ipv4 src spec 10.1.1.1 src prefix 32
3406 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
3410 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
3412 Inclusive ranges can be defined with ``last``::
3414 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
3416 Note that ``mask`` affects both ``spec`` and ``last``::
3418 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
3419 # matches 10.1.0.0 to 10.2.255.255
3421 Properties can be modified multiple times::
3423 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
3427 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
3432 This section lists supported pattern items and their attributes, if any.
3434 - ``end``: end list of pattern items.
3436 - ``void``: no-op pattern item.
3438 - ``invert``: perform actions when pattern does not match.
3440 - ``any``: match any protocol for the current layer.
3442 - ``num {unsigned}``: number of layers covered.
3444 - ``pf``: match traffic from/to the physical function.
3446 - ``vf``: match traffic from/to a virtual function ID.
3448 - ``id {unsigned}``: VF ID.
3450 - ``phy_port``: match traffic from/to a specific physical port.
3452 - ``index {unsigned}``: physical port index.
3454 - ``port_id``: match traffic from/to a given DPDK port ID.
3456 - ``id {unsigned}``: DPDK port ID.
3458 - ``mark``: match value set in previously matched flow rule using the mark action.
3460 - ``id {unsigned}``: arbitrary integer value.
3462 - ``raw``: match an arbitrary byte string.
3464 - ``relative {boolean}``: look for pattern after the previous item.
3465 - ``search {boolean}``: search pattern from offset (see also limit).
3466 - ``offset {integer}``: absolute or relative offset for pattern.
3467 - ``limit {unsigned}``: search area limit for start of pattern.
3468 - ``pattern {string}``: byte string to look for.
3470 - ``eth``: match Ethernet header.
3472 - ``dst {MAC-48}``: destination MAC.
3473 - ``src {MAC-48}``: source MAC.
3474 - ``type {unsigned}``: EtherType or TPID.
3476 - ``vlan``: match 802.1Q/ad VLAN tag.
3478 - ``tci {unsigned}``: tag control information.
3479 - ``pcp {unsigned}``: priority code point.
3480 - ``dei {unsigned}``: drop eligible indicator.
3481 - ``vid {unsigned}``: VLAN identifier.
3482 - ``inner_type {unsigned}``: inner EtherType or TPID.
3484 - ``ipv4``: match IPv4 header.
3486 - ``tos {unsigned}``: type of service.
3487 - ``ttl {unsigned}``: time to live.
3488 - ``proto {unsigned}``: next protocol ID.
3489 - ``src {ipv4 address}``: source address.
3490 - ``dst {ipv4 address}``: destination address.
3492 - ``ipv6``: match IPv6 header.
3494 - ``tc {unsigned}``: traffic class.
3495 - ``flow {unsigned}``: flow label.
3496 - ``proto {unsigned}``: protocol (next header).
3497 - ``hop {unsigned}``: hop limit.
3498 - ``src {ipv6 address}``: source address.
3499 - ``dst {ipv6 address}``: destination address.
3501 - ``icmp``: match ICMP header.
3503 - ``type {unsigned}``: ICMP packet type.
3504 - ``code {unsigned}``: ICMP packet code.
3506 - ``udp``: match UDP header.
3508 - ``src {unsigned}``: UDP source port.
3509 - ``dst {unsigned}``: UDP destination port.
3511 - ``tcp``: match TCP header.
3513 - ``src {unsigned}``: TCP source port.
3514 - ``dst {unsigned}``: TCP destination port.
3516 - ``sctp``: match SCTP header.
3518 - ``src {unsigned}``: SCTP source port.
3519 - ``dst {unsigned}``: SCTP destination port.
3520 - ``tag {unsigned}``: validation tag.
3521 - ``cksum {unsigned}``: checksum.
3523 - ``vxlan``: match VXLAN header.
3525 - ``vni {unsigned}``: VXLAN identifier.
3527 - ``e_tag``: match IEEE 802.1BR E-Tag header.
3529 - ``grp_ecid_b {unsigned}``: GRP and E-CID base.
3531 - ``nvgre``: match NVGRE header.
3533 - ``tni {unsigned}``: virtual subnet ID.
3535 - ``mpls``: match MPLS header.
3537 - ``label {unsigned}``: MPLS label.
3539 - ``gre``: match GRE header.
3541 - ``protocol {unsigned}``: protocol type.
3543 - ``fuzzy``: fuzzy pattern match, expect faster than default.
3545 - ``thresh {unsigned}``: accuracy threshold.
3547 - ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header.
3549 - ``teid {unsigned}``: tunnel endpoint identifier.
3551 - ``geneve``: match GENEVE header.
3553 - ``vni {unsigned}``: virtual network identifier.
3554 - ``protocol {unsigned}``: protocol type.
3556 - ``vxlan-gpe``: match VXLAN-GPE header.
3558 - ``vni {unsigned}``: VXLAN-GPE identifier.
3560 - ``arp_eth_ipv4``: match ARP header for Ethernet/IPv4.
3562 - ``sha {MAC-48}``: sender hardware address.
3563 - ``spa {ipv4 address}``: sender IPv4 address.
3564 - ``tha {MAC-48}``: target hardware address.
3565 - ``tpa {ipv4 address}``: target IPv4 address.
3567 - ``ipv6_ext``: match presence of any IPv6 extension header.
3569 - ``next_hdr {unsigned}``: next header.
3571 - ``icmp6``: match any ICMPv6 header.
3573 - ``type {unsigned}``: ICMPv6 type.
3574 - ``code {unsigned}``: ICMPv6 code.
3576 - ``icmp6_nd_ns``: match ICMPv6 neighbor discovery solicitation.
3578 - ``target_addr {ipv6 address}``: target address.
3580 - ``icmp6_nd_na``: match ICMPv6 neighbor discovery advertisement.
3582 - ``target_addr {ipv6 address}``: target address.
3584 - ``icmp6_nd_opt``: match presence of any ICMPv6 neighbor discovery option.
3586 - ``type {unsigned}``: ND option type.
3588 - ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery source Ethernet
3589 link-layer address option.
3591 - ``sla {MAC-48}``: source Ethernet LLA.
3593 - ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery target Ethernet
3594 link-layer address option.
3596 - ``tla {MAC-48}``: target Ethernet LLA.
3601 A list of actions starts after the ``actions`` token in the same fashion as
3602 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
3603 terminated by a mandatory ``end`` action.
3605 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
3606 rte_flow_action_type``).
3608 Dropping all incoming UDPv4 packets can be expressed as follows::
3610 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3613 Several actions have configurable properties which must be specified when
3614 there is no valid default value. For example, ``queue`` requires a target
3617 This rule redirects incoming UDPv4 traffic to queue index 6::
3619 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3620 actions queue index 6 / end
3622 While this one could be rejected by PMDs (unspecified queue index)::
3624 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3627 As defined by *rte_flow*, the list is not ordered, all actions of a given
3628 rule are performed simultaneously. These are equivalent::
3630 queue index 6 / void / mark id 42 / end
3634 void / mark id 42 / queue index 6 / end
3636 All actions in a list should have different types, otherwise only the last
3637 action of a given type is taken into account::
3639 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
3643 drop / drop / drop / end # drop is performed only once
3647 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
3649 Considering they are performed simultaneously, opposite and overlapping
3650 actions can sometimes be combined when the end result is unambiguous::
3652 drop / queue index 6 / end # drop has no effect
3656 queue index 6 / rss queues 6 7 8 / end # queue has no effect
3660 drop / passthru / end # drop has no effect
3662 Note that PMDs may still refuse such combinations.
3667 This section lists supported actions and their attributes, if any.
3669 - ``end``: end list of actions.
3671 - ``void``: no-op action.
3673 - ``passthru``: let subsequent rule process matched packets.
3675 - ``jump``: redirect traffic to group on device.
3677 - ``group {unsigned}``: group to redirect to.
3679 - ``mark``: attach 32 bit value to packets.
3681 - ``id {unsigned}``: 32 bit value to return with packets.
3683 - ``flag``: flag packets.
3685 - ``queue``: assign packets to a given queue index.
3687 - ``index {unsigned}``: queue index to use.
3689 - ``drop``: drop packets (note: passthru has priority).
3691 - ``count``: enable counters for this rule.
3693 - ``rss``: spread packets among several queues.
3695 - ``func {hash function}``: RSS hash function to apply, allowed tokens are
3696 the same as `set_hash_global_config`_.
3698 - ``level {unsigned}``: encapsulation level for ``types``.
3700 - ``types [{RSS hash type} [...]] end``: specific RSS hash types, allowed
3701 tokens are the same as `set_hash_input_set`_, except that an empty list
3702 does not disable RSS but instead requests unspecified "best-effort"
3705 - ``key {string}``: RSS hash key, overrides ``key_len``.
3707 - ``key_len {unsigned}``: RSS hash key length in bytes, can be used in
3708 conjunction with ``key`` to pad or truncate it.
3710 - ``queues [{unsigned} [...]] end``: queue indices to use.
3712 - ``pf``: direct traffic to physical function.
3714 - ``vf``: direct traffic to a virtual function ID.
3716 - ``original {boolean}``: use original VF ID if possible.
3717 - ``id {unsigned}``: VF ID.
3719 - ``phy_port``: direct packets to physical port index.
3721 - ``original {boolean}``: use original port index if possible.
3722 - ``index {unsigned}``: physical port index.
3724 - ``port_id``: direct matching traffic to a given DPDK port ID.
3726 - ``original {boolean}``: use original DPDK port ID if possible.
3727 - ``id {unsigned}``: DPDK port ID.
3729 - ``of_set_mpls_ttl``: OpenFlow's ``OFPAT_SET_MPLS_TTL``.
3731 - ``mpls_ttl``: MPLS TTL.
3733 - ``of_dec_mpls_ttl``: OpenFlow's ``OFPAT_DEC_MPLS_TTL``.
3735 - ``of_set_nw_ttl``: OpenFlow's ``OFPAT_SET_NW_TTL``.
3737 - ``nw_ttl``: IP TTL.
3739 - ``of_dec_nw_ttl``: OpenFlow's ``OFPAT_DEC_NW_TTL``.
3741 - ``of_copy_ttl_out``: OpenFlow's ``OFPAT_COPY_TTL_OUT``.
3743 - ``of_copy_ttl_in``: OpenFlow's ``OFPAT_COPY_TTL_IN``.
3745 - ``of_pop_vlan``: OpenFlow's ``OFPAT_POP_VLAN``.
3747 - ``of_push_vlan``: OpenFlow's ``OFPAT_PUSH_VLAN``.
3749 - ``ethertype``: Ethertype.
3751 - ``of_set_vlan_vid``: OpenFlow's ``OFPAT_SET_VLAN_VID``.
3753 - ``vlan_vid``: VLAN id.
3755 - ``of_set_vlan_pcp``: OpenFlow's ``OFPAT_SET_VLAN_PCP``.
3757 - ``vlan_pcp``: VLAN priority.
3759 - ``of_pop_mpls``: OpenFlow's ``OFPAT_POP_MPLS``.
3761 - ``ethertype``: Ethertype.
3763 - ``of_push_mpls``: OpenFlow's ``OFPAT_PUSH_MPLS``.
3765 - ``ethertype``: Ethertype.
3767 - ``vxlan_encap``: Performs a VXLAN encapsulation, outer layer configuration
3768 is done through `Config VXLAN Encap outer layers`_.
3770 - ``vxlan_decap``: Performs a decapsulation action by stripping all headers of
3771 the VXLAN tunnel network overlay from the matched flow.
3773 - ``nvgre_encap``: Performs a NVGRE encapsulation, outer layer configuration
3774 is done through `Config NVGRE Encap outer layers`_.
3776 - ``nvgre_decap``: Performs a decapsulation action by stripping all headers of
3777 the NVGRE tunnel network overlay from the matched flow.
3779 - ``set_ipv4_src``: Set a new IPv4 source address in the outermost IPv4 header.
3781 - ``ipv4_addr``: New IPv4 source address.
3783 - ``set_ipv4_dst``: Set a new IPv4 destination address in the outermost IPv4
3786 - ``ipv4_addr``: New IPv4 destination address.
3788 - ``set_ipv6_src``: Set a new IPv6 source address in the outermost IPv6 header.
3790 - ``ipv6_addr``: New IPv6 source address.
3792 - ``set_ipv6_dst``: Set a new IPv6 destination address in the outermost IPv6
3795 - ``ipv6_addr``: New IPv6 destination address.
3797 - ``of_set_tp_src``: Set a new source port number in the outermost TCP/UDP
3800 - ``port``: New TCP/UDP source port number.
3802 - ``of_set_tp_dst``: Set a new destination port number in the outermost TCP/UDP
3805 - ``port``: New TCP/UDP destination port number.
3807 - ``mac_swap``: Swap the source and destination MAC addresses in the outermost
3810 - ``dec_ttl``: Performs a decrease TTL value action
3812 - ``set_ttl``: Set TTL value with specificed value
3813 - ``ttl_value {unsigned}``: The new TTL value to be set
3815 - ``set_mac_src``: set source MAC address
3817 - ``mac_addr {MAC-48}``: new source MAC address
3819 - ``set_mac_dst``: set destination MAC address
3821 - ``mac_addr {MAC-48}``: new destination MAC address
3823 Destroying flow rules
3824 ~~~~~~~~~~~~~~~~~~~~~
3826 ``flow destroy`` destroys one or more rules from their rule ID (as returned
3827 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
3828 times as necessary::
3830 flow destroy {port_id} rule {rule_id} [...]
3832 If successful, it will show::
3834 Flow rule #[...] destroyed
3836 It does not report anything for rule IDs that do not exist. The usual error
3837 message is shown when a rule cannot be destroyed::
3839 Caught error type [...] ([...]): [...]
3841 ``flow flush`` destroys all rules on a device and does not take extra
3842 arguments. It is bound to ``rte_flow_flush()``::
3844 flow flush {port_id}
3846 Any errors are reported as above.
3848 Creating several rules and destroying them::
3850 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3851 actions queue index 2 / end
3852 Flow rule #0 created
3853 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3854 actions queue index 3 / end
3855 Flow rule #1 created
3856 testpmd> flow destroy 0 rule 0 rule 1
3857 Flow rule #1 destroyed
3858 Flow rule #0 destroyed
3861 The same result can be achieved using ``flow flush``::
3863 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3864 actions queue index 2 / end
3865 Flow rule #0 created
3866 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3867 actions queue index 3 / end
3868 Flow rule #1 created
3869 testpmd> flow flush 0
3872 Non-existent rule IDs are ignored::
3874 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3875 actions queue index 2 / end
3876 Flow rule #0 created
3877 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3878 actions queue index 3 / end
3879 Flow rule #1 created
3880 testpmd> flow destroy 0 rule 42 rule 10 rule 2
3882 testpmd> flow destroy 0 rule 0
3883 Flow rule #0 destroyed
3889 ``flow query`` queries a specific action of a flow rule having that
3890 ability. Such actions collect information that can be reported using this
3891 command. It is bound to ``rte_flow_query()``::
3893 flow query {port_id} {rule_id} {action}
3895 If successful, it will display either the retrieved data for known actions
3896 or the following message::
3898 Cannot display result for action type [...] ([...])
3900 Otherwise, it will complain either that the rule does not exist or that some
3903 Flow rule #[...] not found
3907 Caught error type [...] ([...]): [...]
3909 Currently only the ``count`` action is supported. This action reports the
3910 number of packets that hit the flow rule and the total number of bytes. Its
3911 output has the following format::
3914 hits_set: [...] # whether "hits" contains a valid value
3915 bytes_set: [...] # whether "bytes" contains a valid value
3916 hits: [...] # number of packets
3917 bytes: [...] # number of bytes
3919 Querying counters for TCPv6 packets redirected to queue 6::
3921 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
3922 actions queue index 6 / count / end
3923 Flow rule #4 created
3924 testpmd> flow query 0 4 count
3935 ``flow list`` lists existing flow rules sorted by priority and optionally
3936 filtered by group identifiers::
3938 flow list {port_id} [group {group_id}] [...]
3940 This command only fails with the following message if the device does not
3945 Output consists of a header line followed by a short description of each
3946 flow rule, one per line. There is no output at all when no flow rules are
3947 configured on the device::
3949 ID Group Prio Attr Rule
3950 [...] [...] [...] [...] [...]
3952 ``Attr`` column flags:
3954 - ``i`` for ``ingress``.
3955 - ``e`` for ``egress``.
3957 Creating several flow rules and listing them::
3959 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3960 actions queue index 6 / end
3961 Flow rule #0 created
3962 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3963 actions queue index 2 / end
3964 Flow rule #1 created
3965 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
3966 actions rss queues 6 7 8 end / end
3967 Flow rule #2 created
3968 testpmd> flow list 0
3969 ID Group Prio Attr Rule
3970 0 0 0 i- ETH IPV4 => QUEUE
3971 1 0 0 i- ETH IPV6 => QUEUE
3972 2 0 5 i- ETH IPV4 UDP => RSS
3975 Rules are sorted by priority (i.e. group ID first, then priority level)::
3977 testpmd> flow list 1
3978 ID Group Prio Attr Rule
3979 0 0 0 i- ETH => COUNT
3980 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3981 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3982 1 24 0 i- ETH IPV4 UDP => QUEUE
3983 4 24 10 i- ETH IPV4 TCP => DROP
3984 3 24 20 i- ETH IPV4 => DROP
3985 2 24 42 i- ETH IPV4 UDP => QUEUE
3986 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3989 Output can be limited to specific groups::
3991 testpmd> flow list 1 group 0 group 63
3992 ID Group Prio Attr Rule
3993 0 0 0 i- ETH => COUNT
3994 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3995 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3996 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3999 Toggling isolated mode
4000 ~~~~~~~~~~~~~~~~~~~~~~
4002 ``flow isolate`` can be used to tell the underlying PMD that ingress traffic
4003 must only be injected from the defined flow rules; that no default traffic
4004 is expected outside those rules and the driver is free to assign more
4005 resources to handle them. It is bound to ``rte_flow_isolate()``::
4007 flow isolate {port_id} {boolean}
4009 If successful, enabling or disabling isolated mode shows either::
4011 Ingress traffic on port [...]
4012 is now restricted to the defined flow rules
4016 Ingress traffic on port [...]
4017 is not restricted anymore to the defined flow rules
4019 Otherwise, in case of error::
4021 Caught error type [...] ([...]): [...]
4023 Mainly due to its side effects, PMDs supporting this mode may not have the
4024 ability to toggle it more than once without reinitializing affected ports
4025 first (e.g. by exiting testpmd).
4027 Enabling isolated mode::
4029 testpmd> flow isolate 0 true
4030 Ingress traffic on port 0 is now restricted to the defined flow rules
4033 Disabling isolated mode::
4035 testpmd> flow isolate 0 false
4036 Ingress traffic on port 0 is not restricted anymore to the defined flow rules
4039 Sample QinQ flow rules
4040 ~~~~~~~~~~~~~~~~~~~~~~
4042 Before creating QinQ rule(s) the following commands should be issued to enable QinQ::
4044 testpmd> port stop 0
4045 testpmd> vlan set qinq on 0
4047 The above command sets the inner and outer TPID's to 0x8100.
4049 To change the TPID's the following commands should be used::
4051 testpmd> vlan set outer tpid 0xa100 0
4052 testpmd> vlan set inner tpid 0x9100 0
4053 testpmd> port start 0
4055 Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM.
4059 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 /
4060 vlan tci is 456 / end actions vf id 1 / queue index 0 / end
4061 Flow rule #0 validated
4063 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 /
4064 vlan tci is 456 / end actions vf id 123 / queue index 0 / end
4065 Flow rule #0 created
4067 testpmd> flow list 0
4068 ID Group Prio Attr Rule
4069 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
4071 Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host.
4075 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 /
4076 vlan tci is 654 / end actions pf / queue index 0 / end
4077 Flow rule #1 validated
4079 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 /
4080 vlan tci is 654 / end actions pf / queue index 1 / end
4081 Flow rule #1 created
4083 testpmd> flow list 0
4084 ID Group Prio Attr Rule
4085 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
4086 1 0 0 i- ETH VLAN VLAN=>PF QUEUE
4088 Sample VXLAN encapsulation rule
4089 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
4091 VXLAN encapsulation outer layer has default value pre-configured in testpmd
4092 source code, those can be changed by using the following commands
4094 IPv4 VXLAN outer header::
4096 testpmd> set vxlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src 127.0.0.1
4097 ip-dst 128.0.0.1 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
4098 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
4101 testpmd> set vxlan-with-vlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src
4102 127.0.0.1 ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11
4103 eth-dst 22:22:22:22:22:22
4104 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
4107 IPv6 VXLAN outer header::
4109 testpmd> set vxlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4 ip-src ::1
4110 ip-dst ::2222 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
4111 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
4114 testpmd> set vxlan-with-vlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4
4115 ip-src ::1 ip-dst ::2222 vlan-tci 34 eth-src 11:11:11:11:11:11
4116 eth-dst 22:22:22:22:22:22
4117 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
4120 Sample NVGRE encapsulation rule
4121 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
4123 NVGRE encapsulation outer layer has default value pre-configured in testpmd
4124 source code, those can be changed by using the following commands
4126 IPv4 NVGRE outer header::
4128 testpmd> set nvgre ip-version ipv4 tni 4 ip-src 127.0.0.1 ip-dst 128.0.0.1
4129 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 nvgre_encap /
4133 testpmd> set nvgre-with-vlan ip-version ipv4 tni 4 ip-src 127.0.0.1
4134 ip-dst 128.0.0.1 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 nvgre_encap /
4139 IPv6 NVGRE outer header::
4141 testpmd> set nvgre ip-version ipv6 tni 4 ip-src ::1 ip-dst ::2222
4142 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
4143 testpmd> flow create 0 ingress pattern end actions nvgre_encap /
4146 testpmd> set nvgre-with-vlan ip-version ipv6 tni 4 ip-src ::1 ip-dst ::2222
4147 vlan-tci 34 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
4148 testpmd> flow create 0 ingress pattern end actions nvgre_encap /
4154 The following sections show functions to load/unload eBPF based filters.
4159 Load an eBPF program as a callback for partciular RX/TX queue::
4161 testpmd> bpf-load rx|tx (portid) (queueid) (load-flags) (bpf-prog-filename)
4163 The available load-flags are:
4165 * ``J``: use JIT generated native code, otherwise BPF interpreter will be used.
4167 * ``M``: assume input parameter is a pointer to rte_mbuf, otherwise assume it is a pointer to first segment's data.
4173 You'll need clang v3.7 or above to build bpf program you'd like to load
4177 .. code-block:: console
4180 clang -O2 -target bpf -c t1.c
4182 Then to load (and JIT compile) t1.o at RX queue 0, port 1::
4184 .. code-block:: console
4186 testpmd> bpf-load rx 1 0 J ./dpdk.org/test/bpf/t1.o
4188 To load (not JITed) t1.o at TX queue 0, port 0::
4190 .. code-block:: console
4192 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o
4197 Unload previously loaded eBPF program for partciular RX/TX queue::
4199 testpmd> bpf-unload rx|tx (portid) (queueid)
4201 For example to unload BPF filter from TX queue 0, port 0:
4203 .. code-block:: console
4205 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o