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 Currently the only available levels are 0 (silent except for error) and 1 (fully verbose).
457 Set the log level for a log type::
459 testpmd> set log global|(type) (level)
463 * ``type`` is the log name.
465 * ``level`` is the log level.
467 For example, to change the global log level::
468 testpmd> set log global (level)
470 Regexes can also be used for type. To change log level of user1, user2 and user3::
471 testpmd> set log user[1-3] (level)
476 Set the number of ports used by the application:
480 This is equivalent to the ``--nb-ports`` command-line option.
485 Set the number of cores used by the application::
487 testpmd> set nbcore (num)
489 This is equivalent to the ``--nb-cores`` command-line option.
493 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port.
498 Set the forwarding cores hexadecimal mask::
500 testpmd> set coremask (mask)
502 This is equivalent to the ``--coremask`` command-line option.
506 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding.
511 Set the forwarding ports hexadecimal mask::
513 testpmd> set portmask (mask)
515 This is equivalent to the ``--portmask`` command-line option.
520 Set number of packets per burst::
522 testpmd> set burst (num)
524 This is equivalent to the ``--burst command-line`` option.
526 When retry is enabled, the transmit delay time and number of retries can also be set::
528 testpmd> set burst tx delay (microseconds) retry (num)
533 Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode::
535 testpmd> set txpkts (x[,y]*)
537 Where x[,y]* represents a CSV list of values, without white space.
542 Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes::
544 testpmd> set txsplit (off|on|rand)
548 * ``off`` disable packet copy & split for CSUM mode.
550 * ``on`` split outgoing packet into multiple segments. Size of each segment
551 and number of segments per packet is determined by ``set txpkts`` command
554 * ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments.
559 Set the list of forwarding cores::
561 testpmd> set corelist (x[,y]*)
563 For example, to change the forwarding cores:
565 .. code-block:: console
567 testpmd> set corelist 3,1
568 testpmd> show config fwd
570 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled
571 Logical Core 3 (socket 0) forwards packets on 1 streams:
572 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01
573 Logical Core 1 (socket 0) forwards packets on 1 streams:
574 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
578 The cores are used in the same order as specified on the command line.
583 Set the list of forwarding ports::
585 testpmd> set portlist (x[,y]*)
587 For example, to change the port forwarding:
589 .. code-block:: console
591 testpmd> set portlist 0,2,1,3
592 testpmd> show config fwd
594 io packet forwarding - ports=4 - cores=1 - streams=4
595 Logical Core 3 (socket 0) forwards packets on 4 streams:
596 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01
597 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
598 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03
599 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02
604 Enable/disable tx loopback::
606 testpmd> set tx loopback (port_id) (on|off)
611 set drop enable bit for all queues::
613 testpmd> set all queues drop (port_id) (on|off)
615 set split drop enable (for VF)
616 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
618 set split drop enable bit for VF from PF::
620 testpmd> set vf split drop (port_id) (vf_id) (on|off)
622 set mac antispoof (for VF)
623 ~~~~~~~~~~~~~~~~~~~~~~~~~~
625 Set mac antispoof for a VF from the PF::
627 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off)
632 Enable/disable MACsec offload::
634 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off)
635 testpmd> set macsec offload (port_id) off
640 Configure MACsec secure connection (SC)::
642 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi)
646 The pi argument is ignored for tx.
647 Check the NIC Datasheet for hardware limits.
652 Configure MACsec secure association (SA)::
654 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key)
658 The IDX value must be 0 or 1.
659 Check the NIC Datasheet for hardware limits.
661 set broadcast mode (for VF)
662 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
664 Set broadcast mode for a VF from the PF::
666 testpmd> set vf broadcast (port_id) (vf_id) (on|off)
671 Set the VLAN strip on a port::
673 testpmd> vlan set strip (on|off) (port_id)
678 Set the VLAN strip for a queue on a port::
680 testpmd> vlan set stripq (on|off) (port_id,queue_id)
682 vlan set stripq (for VF)
683 ~~~~~~~~~~~~~~~~~~~~~~~~
685 Set VLAN strip for all queues in a pool for a VF from the PF::
687 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off)
689 vlan set insert (for VF)
690 ~~~~~~~~~~~~~~~~~~~~~~~~
692 Set VLAN insert for a VF from the PF::
694 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id)
696 vlan set tag (for VF)
697 ~~~~~~~~~~~~~~~~~~~~~
699 Set VLAN tag for a VF from the PF::
701 testpmd> set vf vlan tag (port_id) (vf_id) (on|off)
703 vlan set antispoof (for VF)
704 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
706 Set VLAN antispoof for a VF from the PF::
708 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off)
713 Set the VLAN filter on a port::
715 testpmd> vlan set filter (on|off) (port_id)
720 Set the VLAN QinQ (extended queue in queue) on for a port::
722 testpmd> vlan set qinq (on|off) (port_id)
727 Set the inner or outer VLAN TPID for packet filtering on a port::
729 testpmd> vlan set (inner|outer) tpid (value) (port_id)
733 TPID value must be a 16-bit number (value <= 65536).
738 Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID::
740 testpmd> rx_vlan add (vlan_id|all) (port_id)
744 VLAN filter must be set on that port. VLAN ID < 4096.
745 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries
746 in VFTA table. This is important if enabling all vlan_ids.
751 Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID::
753 testpmd> rx_vlan rm (vlan_id|all) (port_id)
758 Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID::
760 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask)
765 Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID::
767 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask)
772 Add a tunnel filter on a port::
774 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \
775 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
776 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
778 The available information categories are:
780 * ``vxlan``: Set tunnel type as VXLAN.
782 * ``nvgre``: Set tunnel type as NVGRE.
784 * ``ipingre``: Set tunnel type as IP-in-GRE.
786 * ``imac-ivlan``: Set filter type as Inner MAC and VLAN.
788 * ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID.
790 * ``imac-tenid``: Set filter type as Inner MAC and tenant ID.
792 * ``imac``: Set filter type as Inner MAC.
794 * ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID.
796 * ``oip``: Set filter type as Outer IP.
798 * ``iip``: Set filter type as Inner IP.
802 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \
803 192.168.2.2 0 ipingre oip 1 1
805 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP.
810 Remove a tunnel filter on a port::
812 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \
813 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
814 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
819 Add an UDP port for VXLAN packet filter on a port::
821 testpmd> rx_vxlan_port add (udp_port) (port_id)
826 Remove an UDP port for VXLAN packet filter on a port::
828 testpmd> rx_vxlan_port rm (udp_port) (port_id)
833 Set hardware insertion of VLAN IDs in packets sent on a port::
835 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer]
837 For example, set a single VLAN ID (5) insertion on port 0::
841 Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1::
849 Set port based hardware insertion of VLAN ID in packets sent on a port::
851 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off)
856 Disable hardware insertion of a VLAN header in packets sent on a port::
858 testpmd> tx_vlan reset (port_id)
863 Select hardware or software calculation of the checksum when
864 transmitting a packet using the ``csum`` forwarding engine::
866 testpmd> csum set (ip|udp|tcp|sctp|outer-ip|outer-udp) (hw|sw) (port_id)
870 * ``ip|udp|tcp|sctp`` always relate to the inner layer.
872 * ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized
873 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are
874 supported). See also the ``csum parse-tunnel`` command.
876 * ``outer-udp`` relates to the outer UDP layer in the case where the packet is recognized
877 as a tunnel packet by the forwarding engine (vxlan, vxlan-gpe are
878 supported). See also the ``csum parse-tunnel`` command.
882 Check the NIC Datasheet for hardware limits.
887 Set RSS queue region span on a port::
889 testpmd> set port (port_id) queue-region region_id (value) \
890 queue_start_index (value) queue_num (value)
892 Set flowtype mapping on a RSS queue region on a port::
894 testpmd> set port (port_id) queue-region region_id (value) flowtype (value)
898 * For the flowtype(pctype) of packet,the specific index for each type has
899 been defined in file i40e_type.h as enum i40e_filter_pctype.
901 Set user priority mapping on a RSS queue region on a port::
903 testpmd> set port (port_id) queue-region UP (value) region_id (value)
905 Flush all queue region related configuration on a port::
907 testpmd> set port (port_id) queue-region flush (on|off)
911 * "on"is just an enable function which server for other configuration,
912 it is for all configuration about queue region from up layer,
913 at first will only keep in DPDK softwarestored in driver,
914 only after "flush on", it commit all configuration to HW.
915 "off" is just clean all configuration about queue region just now,
916 and restore all to DPDK i40e driver default config when start up.
918 Show all queue region related configuration info on a port::
920 testpmd> show port (port_id) queue-region
924 Queue region only support on PF by now, so these command is
925 only for configuration of queue region on PF port.
930 Define how tunneled packets should be handled by the csum forward
933 testpmd> csum parse-tunnel (on|off) (tx_port_id)
935 If enabled, the csum forward engine will try to recognize supported
936 tunnel headers (vxlan, gre, ipip).
938 If disabled, treat tunnel packets as non-tunneled packets (a inner
939 header is handled as a packet payload).
943 The port argument is the TX port like in the ``csum set`` command.
947 Consider a packet in packet like the following::
949 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in
951 * If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
952 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the
953 ``outer-ip|outer-udp`` parameter relates to the outer headers (here ``ipv4_out`` and ``udp_out``).
955 * If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
956 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``.
961 Display tx checksum offload configuration::
963 testpmd> csum show (port_id)
968 Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine::
970 testpmd> tso set (segsize) (port_id)
974 Check the NIC datasheet for hardware limits.
979 Display the status of TCP Segmentation Offload::
981 testpmd> tso show (port_id)
986 Enable or disable GRO in ``csum`` forwarding engine::
988 testpmd> set port <port_id> gro on|off
990 If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4
991 packets received from the given port.
993 If disabled, packets received from the given port won't be performed
994 GRO. By default, GRO is disabled for all ports.
998 When enable GRO for a port, TCP/IPv4 packets received from the port
999 will be performed GRO. After GRO, all merged packets have bad
1000 checksums, since the GRO library doesn't re-calculate checksums for
1001 the merged packets. Therefore, if users want the merged packets to
1002 have correct checksums, please select HW IP checksum calculation and
1003 HW TCP checksum calculation for the port which the merged packets are
1009 Display GRO configuration for a given port::
1011 testpmd> show port <port_id> gro
1016 Set the cycle to flush the GROed packets from reassembly tables::
1018 testpmd> set gro flush <cycles>
1020 When enable GRO, the csum forwarding engine performs GRO on received
1021 packets, and the GROed packets are stored in reassembly tables. Users
1022 can use this command to determine when the GROed packets are flushed
1023 from the reassembly tables.
1025 The ``cycles`` is measured in GRO operation times. The csum forwarding
1026 engine flushes the GROed packets from the tables every ``cycles`` GRO
1029 By default, the value of ``cycles`` is 1, which means flush GROed packets
1030 from the reassembly tables as soon as one GRO operation finishes. The value
1031 of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``.
1033 Please note that the large value of ``cycles`` may cause the poor TCP/IP
1034 stack performance. Because the GROed packets are delayed to arrive the
1035 stack, thus causing more duplicated ACKs and TCP retransmissions.
1040 Toggle per-port GSO support in ``csum`` forwarding engine::
1042 testpmd> set port <port_id> gso on|off
1044 If enabled, the csum forwarding engine will perform GSO on supported IPv4
1045 packets, transmitted on the given port.
1047 If disabled, packets transmitted on the given port will not undergo GSO.
1048 By default, GSO is disabled for all ports.
1052 When GSO is enabled on a port, supported IPv4 packets transmitted on that
1053 port undergo GSO. Afterwards, the segmented packets are represented by
1054 multi-segment mbufs; however, the csum forwarding engine doesn't calculation
1055 of checksums for GSO'd segments in SW. As a result, if users want correct
1056 checksums in GSO segments, they should enable HW checksum calculation for
1059 For example, HW checksum calculation for VxLAN GSO'd packets may be enabled
1060 by setting the following options in the csum forwarding engine:
1062 testpmd> csum set outer_ip hw <port_id>
1064 testpmd> csum set ip hw <port_id>
1066 testpmd> csum set tcp hw <port_id>
1068 UDP GSO is the same as IP fragmentation, which treats the UDP header
1069 as the payload and does not modify it during segmentation. That is,
1070 after UDP GSO, only the first output fragment has the original UDP
1071 header. Therefore, users need to enable HW IP checksum calculation
1072 and SW UDP checksum calculation for GSO-enabled ports, if they want
1073 correct checksums for UDP/IPv4 packets.
1078 Set the maximum GSO segment size (measured in bytes), which includes the
1079 packet header and the packet payload for GSO-enabled ports (global)::
1081 testpmd> set gso segsz <length>
1086 Display the status of Generic Segmentation Offload for a given port::
1088 testpmd> show port <port_id> gso
1093 Add an alternative MAC address to a port::
1095 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX)
1100 Remove a MAC address from a port::
1102 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX)
1104 mac_addr add (for VF)
1105 ~~~~~~~~~~~~~~~~~~~~~
1107 Add an alternative MAC address for a VF to a port::
1109 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX)
1114 Set the default MAC address for a port::
1116 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX)
1118 mac_addr set (for VF)
1119 ~~~~~~~~~~~~~~~~~~~~~
1121 Set the MAC address for a VF from the PF::
1123 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX)
1128 Set the forwarding peer address for certain port::
1130 testpmd> set eth-peer (port_id) (perr_addr)
1132 This is equivalent to the ``--eth-peer`` command-line option.
1137 Set the unicast hash filter(s) on/off for a port::
1139 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
1144 Set the promiscuous mode on for a port or for all ports.
1145 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1147 testpmd> set promisc (port_id|all) (on|off)
1152 Set the allmulti mode for a port or for all ports::
1154 testpmd> set allmulti (port_id|all) (on|off)
1156 Same as the ifconfig (8) option. Controls how multicast packets are handled.
1158 set promisc (for VF)
1159 ~~~~~~~~~~~~~~~~~~~~
1161 Set the unicast promiscuous mode for a VF from PF.
1162 It's supported by Intel i40e NICs now.
1163 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1165 testpmd> set vf promisc (port_id) (vf_id) (on|off)
1167 set allmulticast (for VF)
1168 ~~~~~~~~~~~~~~~~~~~~~~~~~
1170 Set the multicast promiscuous mode for a VF from PF.
1171 It's supported by Intel i40e NICs now.
1172 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1174 testpmd> set vf allmulti (port_id) (vf_id) (on|off)
1176 set tx max bandwidth (for VF)
1177 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1179 Set TX max absolute bandwidth (Mbps) for a VF from PF::
1181 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth)
1183 set tc tx min bandwidth (for VF)
1184 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1186 Set all TCs' TX min relative bandwidth (%) for a VF from PF::
1188 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...)
1190 set tc tx max bandwidth (for VF)
1191 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1193 Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF::
1195 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth)
1197 set tc strict link priority mode
1198 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1200 Set some TCs' strict link priority mode on a physical port::
1202 testpmd> set tx strict-link-priority (port_id) (tc_bitmap)
1204 set tc tx min bandwidth
1205 ~~~~~~~~~~~~~~~~~~~~~~~
1207 Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs::
1209 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...)
1214 Set the link flow control parameter on a port::
1216 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1217 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
1218 autoneg (on|off) (port_id)
1222 * ``high_water`` (integer): High threshold value to trigger XOFF.
1224 * ``low_water`` (integer): Low threshold value to trigger XON.
1226 * ``pause_time`` (integer): Pause quota in the Pause frame.
1228 * ``send_xon`` (0/1): Send XON frame.
1230 * ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
1232 * ``autoneg``: Change the auto-negotiation parameter.
1237 Set the priority flow control parameter on a port::
1239 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1240 (pause_time) (priority) (port_id)
1244 * ``high_water`` (integer): High threshold value.
1246 * ``low_water`` (integer): Low threshold value.
1248 * ``pause_time`` (integer): Pause quota in the Pause frame.
1250 * ``priority`` (0-7): VLAN User Priority.
1255 Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
1257 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
1259 For example, to set rx queue 2 on port 0 to mapping 5::
1261 testpmd>set stat_qmap rx 0 2 5
1263 set xstats-hide-zero
1264 ~~~~~~~~~~~~~~~~~~~~
1266 Set the option to hide zero values for xstats display::
1268 testpmd> set xstats-hide-zero on|off
1272 By default, the zero values are displayed for xstats.
1274 set port - rx/tx (for VF)
1275 ~~~~~~~~~~~~~~~~~~~~~~~~~
1277 Set VF receive/transmit from a port::
1279 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
1281 set port - mac address filter (for VF)
1282 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1284 Add/Remove unicast or multicast MAC addr filter for a VF::
1286 testpmd> set port (port_id) vf (vf_id) (mac_addr) \
1287 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
1289 set port - rx mode(for VF)
1290 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1292 Set the VF receive mode of a port::
1294 testpmd> set port (port_id) vf (vf_id) \
1295 rxmode (AUPE|ROPE|BAM|MPE) (on|off)
1297 The available receive modes are:
1299 * ``AUPE``: Accepts untagged VLAN.
1301 * ``ROPE``: Accepts unicast hash.
1303 * ``BAM``: Accepts broadcast packets.
1305 * ``MPE``: Accepts all multicast packets.
1307 set port - tx_rate (for Queue)
1308 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1310 Set TX rate limitation for a queue on a port::
1312 testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1314 set port - tx_rate (for VF)
1315 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
1317 Set TX rate limitation for queues in VF on a port::
1319 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1321 set port - mirror rule
1322 ~~~~~~~~~~~~~~~~~~~~~~
1324 Set pool or vlan type mirror rule for a port::
1326 testpmd> set port (port_id) mirror-rule (rule_id) \
1327 (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1328 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1330 Set link mirror rule for a port::
1332 testpmd> set port (port_id) mirror-rule (rule_id) \
1333 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1335 For example to enable mirror traffic with vlan 0,1 to pool 0::
1337 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1339 reset port - mirror rule
1340 ~~~~~~~~~~~~~~~~~~~~~~~~
1342 Reset a mirror rule for a port::
1344 testpmd> reset port (port_id) mirror-rule (rule_id)
1349 Set the flush on RX streams before forwarding.
1350 The default is flush ``on``.
1351 Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1353 testpmd> set flush_rx off
1358 Set the bypass mode for the lowest port on bypass enabled NIC::
1360 testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1365 Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1367 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1368 mode (normal|bypass|isolate) (port_id)
1372 * ``timeout``: Enable bypass after watchdog timeout.
1374 * ``os_on``: Enable bypass when OS/board is powered on.
1376 * ``os_off``: Enable bypass when OS/board is powered off.
1378 * ``power_on``: Enable bypass when power supply is turned on.
1380 * ``power_off``: Enable bypass when power supply is turned off.
1386 Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1388 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1393 Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1395 testpmd> show bypass config (port_id)
1400 Set link up for a port::
1402 testpmd> set link-up port (port id)
1407 Set link down for a port::
1409 testpmd> set link-down port (port id)
1414 Enable E-tag insertion for a VF on a port::
1416 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1418 Disable E-tag insertion for a VF on a port::
1420 testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1422 Enable/disable E-tag stripping on a port::
1424 testpmd> E-tag set stripping (on|off) port (port_id)
1426 Enable/disable E-tag based forwarding on a port::
1428 testpmd> E-tag set forwarding (on|off) port (port_id)
1430 Add an E-tag forwarding filter on a port::
1432 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1434 Delete an E-tag forwarding filter on a port::
1435 testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1440 Load a dynamic device personalization (DDP) profile and store backup profile::
1442 testpmd> ddp add (port_id) (profile_path[,backup_profile_path])
1447 Delete a dynamic device personalization profile and restore backup profile::
1449 testpmd> ddp del (port_id) (backup_profile_path)
1454 List all items from the ptype mapping table::
1456 testpmd> ptype mapping get (port_id) (valid_only)
1460 * ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0).
1462 Replace a specific or a group of software defined ptype with a new one::
1464 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type)
1468 * ``target``: A specific software ptype or a mask to represent a group of software ptypes.
1470 * ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1).
1472 * ``pkt_type``: The new software ptype to replace the old ones.
1474 Update hardware defined ptype to software defined packet type mapping table::
1476 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype)
1480 * ``hw_ptype``: hardware ptype as the index of the ptype mapping table.
1482 * ``sw_ptype``: software ptype as the value of the ptype mapping table.
1484 Reset ptype mapping table::
1486 testpmd> ptype mapping reset (port_id)
1488 config per port Rx offloading
1489 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1491 Enable or disable a per port Rx offloading on all Rx queues of a port::
1493 testpmd> port config (port_id) rx_offload (offloading) on|off
1495 * ``offloading``: can be any of these offloading capability:
1496 vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro,
1497 qinq_strip, outer_ipv4_cksum, macsec_strip,
1498 header_split, vlan_filter, vlan_extend, jumbo_frame,
1499 crc_strip, scatter, timestamp, security, keep_crc
1501 This command should be run when the port is stopped, or else it will fail.
1503 config per queue Rx offloading
1504 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1506 Enable or disable a per queue Rx offloading only on a specific Rx queue::
1508 testpmd> port (port_id) rxq (queue_id) rx_offload (offloading) on|off
1510 * ``offloading``: can be any of these offloading capability:
1511 vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro,
1512 qinq_strip, outer_ipv4_cksum, macsec_strip,
1513 header_split, vlan_filter, vlan_extend, jumbo_frame,
1514 crc_strip, scatter, timestamp, security, keep_crc
1516 This command should be run when the port is stopped, or else it will fail.
1518 config per port Tx offloading
1519 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1521 Enable or disable a per port Tx offloading on all Tx queues of a port::
1523 testpmd> port config (port_id) tx_offload (offloading) on|off
1525 * ``offloading``: can be any of these offloading capability:
1526 vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum,
1527 sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum,
1528 qinq_insert, vxlan_tnl_tso, gre_tnl_tso,
1529 ipip_tnl_tso, geneve_tnl_tso, macsec_insert,
1530 mt_lockfree, multi_segs, mbuf_fast_free, security
1532 This command should be run when the port is stopped, or else it will fail.
1534 config per queue Tx offloading
1535 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1537 Enable or disable a per queue Tx offloading only on a specific Tx queue::
1539 testpmd> port (port_id) txq (queue_id) tx_offload (offloading) on|off
1541 * ``offloading``: can be any of these offloading capability:
1542 vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum,
1543 sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum,
1544 qinq_insert, vxlan_tnl_tso, gre_tnl_tso,
1545 ipip_tnl_tso, geneve_tnl_tso, macsec_insert,
1546 mt_lockfree, multi_segs, mbuf_fast_free, security
1548 This command should be run when the port is stopped, or else it will fail.
1550 Config VXLAN Encap outer layers
1551 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1553 Configure the outer layer to encapsulate a packet inside a VXLAN tunnel::
1555 set vxlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \
1556 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) eth-src (eth-src) \
1559 set vxlan-with-vlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \
1560 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) vlan-tci (vlan-tci) \
1561 eth-src (eth-src) eth-dst (eth-dst)
1563 Those command will set an internal configuration inside testpmd, any following
1564 flow rule using the action vxlan_encap will use the last configuration set.
1565 To have a different encapsulation header, one of those commands must be called
1566 before the flow rule creation.
1568 Config NVGRE Encap outer layers
1569 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1571 Configure the outer layer to encapsulate a packet inside a NVGRE tunnel::
1573 set nvgre ip-version (ipv4|ipv6) tni (tni) ip-src (ip-src) ip-dst (ip-dst) \
1574 eth-src (eth-src) eth-dst (eth-dst)
1575 set nvgre-with-vlan ip-version (ipv4|ipv6) tni (tni) ip-src (ip-src) \
1576 ip-dst (ip-dst) vlan-tci (vlan-tci) eth-src (eth-src) eth-dst (eth-dst)
1578 Those command will set an internal configuration inside testpmd, any following
1579 flow rule using the action nvgre_encap will use the last configuration set.
1580 To have a different encapsulation header, one of those commands must be called
1581 before the flow rule creation.
1586 The following sections show functions for configuring ports.
1590 Port configuration changes only become active when forwarding is started/restarted.
1595 Attach a port specified by pci address or virtual device args::
1597 testpmd> port attach (identifier)
1599 To attach a new pci device, the device should be recognized by kernel first.
1600 Then it should be moved under DPDK management.
1601 Finally the port can be attached to testpmd.
1603 For example, to move a pci device using ixgbe under DPDK management:
1605 .. code-block:: console
1607 # Check the status of the available devices.
1608 ./usertools/dpdk-devbind.py --status
1610 Network devices using DPDK-compatible driver
1611 ============================================
1614 Network devices using kernel driver
1615 ===================================
1616 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1619 # Bind the device to igb_uio.
1620 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1623 # Recheck the status of the devices.
1624 ./usertools/dpdk-devbind.py --status
1625 Network devices using DPDK-compatible driver
1626 ============================================
1627 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1629 To attach a port created by virtual device, above steps are not needed.
1631 For example, to attach a port whose pci address is 0000:0a:00.0.
1633 .. code-block:: console
1635 testpmd> port attach 0000:0a:00.0
1636 Attaching a new port...
1637 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1638 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1639 EAL: PCI memory mapped at 0x7f83bfa00000
1640 EAL: PCI memory mapped at 0x7f83bfa80000
1641 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1642 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1643 Port 0 is attached. Now total ports is 1
1646 For example, to attach a port created by pcap PMD.
1648 .. code-block:: console
1650 testpmd> port attach net_pcap0
1651 Attaching a new port...
1652 PMD: Initializing pmd_pcap for net_pcap0
1653 PMD: Creating pcap-backed ethdev on numa socket 0
1654 Port 0 is attached. Now total ports is 1
1657 In this case, identifier is ``net_pcap0``.
1658 This identifier format is the same as ``--vdev`` format of DPDK applications.
1660 For example, to re-attach a bonded port which has been previously detached,
1661 the mode and slave parameters must be given.
1663 .. code-block:: console
1665 testpmd> port attach net_bond_0,mode=0,slave=1
1666 Attaching a new port...
1667 EAL: Initializing pmd_bond for net_bond_0
1668 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1669 Port 0 is attached. Now total ports is 1
1676 Detach a specific port::
1678 testpmd> port detach (port_id)
1680 Before detaching a port, the port should be stopped and closed.
1682 For example, to detach a pci device port 0.
1684 .. code-block:: console
1686 testpmd> port stop 0
1689 testpmd> port close 0
1693 testpmd> port detach 0
1695 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1696 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1697 EAL: PCI memory unmapped at 0x7f83bfa00000
1698 EAL: PCI memory unmapped at 0x7f83bfa80000
1702 For example, to detach a virtual device port 0.
1704 .. code-block:: console
1706 testpmd> port stop 0
1709 testpmd> port close 0
1713 testpmd> port detach 0
1715 PMD: Closing pcap ethdev on numa socket 0
1716 Port 'net_pcap0' is detached. Now total ports is 0
1719 To remove a pci device completely from the system, first detach the port from testpmd.
1720 Then the device should be moved under kernel management.
1721 Finally the device can be removed using kernel pci hotplug functionality.
1723 For example, to move a pci device under kernel management:
1725 .. code-block:: console
1727 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1729 ./usertools/dpdk-devbind.py --status
1731 Network devices using DPDK-compatible driver
1732 ============================================
1735 Network devices using kernel driver
1736 ===================================
1737 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1739 To remove a port created by a virtual device, above steps are not needed.
1744 Start all ports or a specific port::
1746 testpmd> port start (port_id|all)
1751 Stop all ports or a specific port::
1753 testpmd> port stop (port_id|all)
1758 Close all ports or a specific port::
1760 testpmd> port close (port_id|all)
1762 port config - queue ring size
1763 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1765 Configure a rx/tx queue ring size::
1767 testpmd> port (port_id) (rxq|txq) (queue_id) ring_size (value)
1769 Only take effect after command that (re-)start the port or command that setup specific queue.
1771 port start/stop queue
1772 ~~~~~~~~~~~~~~~~~~~~~
1774 Start/stop a rx/tx queue on a specific port::
1776 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1778 port config - queue deferred start
1779 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1781 Switch on/off deferred start of a specific port queue::
1783 testpmd> port (port_id) (rxq|txq) (queue_id) deferred_start (on|off)
1786 ~~~~~~~~~~~~~~~~~~~~~
1788 Setup a rx/tx queue on a specific port::
1790 testpmd> port (port_id) (rxq|txq) (queue_id) setup
1792 Only take effect when port is started.
1797 Set the speed and duplex mode for all ports or a specific port::
1799 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1800 duplex (half|full|auto)
1802 port config - queues/descriptors
1803 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1805 Set number of queues/descriptors for rxq, txq, rxd and txd::
1807 testpmd> port config all (rxq|txq|rxd|txd) (value)
1809 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1811 port config - max-pkt-len
1812 ~~~~~~~~~~~~~~~~~~~~~~~~~
1814 Set the maximum packet length::
1816 testpmd> port config all max-pkt-len (value)
1818 This is equivalent to the ``--max-pkt-len`` command-line option.
1820 port config - CRC Strip
1821 ~~~~~~~~~~~~~~~~~~~~~~~
1823 Set hardware CRC stripping on or off for all ports::
1825 testpmd> port config all crc-strip (on|off)
1827 CRC stripping is on by default.
1829 The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option.
1831 port config - scatter
1832 ~~~~~~~~~~~~~~~~~~~~~~~
1834 Set RX scatter mode on or off for all ports::
1836 testpmd> port config all scatter (on|off)
1838 RX scatter mode is off by default.
1840 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1842 port config - RX Checksum
1843 ~~~~~~~~~~~~~~~~~~~~~~~~~
1845 Set hardware RX checksum offload to on or off for all ports::
1847 testpmd> port config all rx-cksum (on|off)
1849 Checksum offload is off by default.
1851 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1856 Set hardware VLAN on or off for all ports::
1858 testpmd> port config all hw-vlan (on|off)
1860 Hardware VLAN is off by default.
1862 The ``on`` option is equivalent to the ``--enable-hw-vlan`` command-line option.
1864 port config - VLAN filter
1865 ~~~~~~~~~~~~~~~~~~~~~~~~~
1867 Set hardware VLAN filter on or off for all ports::
1869 testpmd> port config all hw-vlan-filter (on|off)
1871 Hardware VLAN filter is off by default.
1873 The ``on`` option is equivalent to the ``--enable-hw-vlan-filter`` command-line option.
1875 port config - VLAN strip
1876 ~~~~~~~~~~~~~~~~~~~~~~~~
1878 Set hardware VLAN strip on or off for all ports::
1880 testpmd> port config all hw-vlan-strip (on|off)
1882 Hardware VLAN strip is off by default.
1884 The ``on`` option is equivalent to the ``--enable-hw-vlan-strip`` command-line option.
1886 port config - VLAN extend
1887 ~~~~~~~~~~~~~~~~~~~~~~~~~
1889 Set hardware VLAN extend on or off for all ports::
1891 testpmd> port config all hw-vlan-extend (on|off)
1893 Hardware VLAN extend is off by default.
1895 The ``on`` option is equivalent to the ``--enable-hw-vlan-extend`` command-line option.
1897 port config - Drop Packets
1898 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1900 Set packet drop for packets with no descriptors on or off for all ports::
1902 testpmd> port config all drop-en (on|off)
1904 Packet dropping for packets with no descriptors is off by default.
1906 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1911 Set the RSS (Receive Side Scaling) mode on or off::
1913 testpmd> port config all rss (all|default|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1915 RSS is on by default.
1917 The ``all`` option is equivalent to ip|tcp|udp|sctp|ether.
1918 The ``default`` option enables all supported RSS types reported by device info.
1919 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1921 port config - RSS Reta
1922 ~~~~~~~~~~~~~~~~~~~~~~
1924 Set the RSS (Receive Side Scaling) redirection table::
1926 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1931 Set the DCB mode for an individual port::
1933 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1935 The traffic class should be 4 or 8.
1940 Set the number of packets per burst::
1942 testpmd> port config all burst (value)
1944 This is equivalent to the ``--burst`` command-line option.
1946 port config - Threshold
1947 ~~~~~~~~~~~~~~~~~~~~~~~
1949 Set thresholds for TX/RX queues::
1951 testpmd> port config all (threshold) (value)
1953 Where the threshold type can be:
1955 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1957 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1959 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1961 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1963 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1965 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1967 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1969 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1971 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1973 These threshold options are also available from the command-line.
1978 Set the value of ether-type for E-tag::
1980 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1982 Enable/disable the E-tag support::
1984 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1986 port config pctype mapping
1987 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1989 Reset pctype mapping table::
1991 testpmd> port config (port_id) pctype mapping reset
1993 Update hardware defined pctype to software defined flow type mapping table::
1995 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id)
1999 * ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table.
2001 * ``flow_type_id``: software flow type id as the index of the pctype mapping table.
2003 port config input set
2004 ~~~~~~~~~~~~~~~~~~~~~
2006 Config RSS/FDIR/FDIR flexible payload input set for some pctype::
2007 testpmd> port config (port_id) pctype (pctype_id) \
2008 (hash_inset|fdir_inset|fdir_flx_inset) \
2009 (get|set|clear) field (field_idx)
2011 Clear 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) clear all
2017 * ``pctype_id``: hardware packet classification types.
2018 * ``field_idx``: hardware field index.
2020 port config udp_tunnel_port
2021 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
2023 Add/remove UDP tunnel port for VXLAN/GENEVE tunneling protocols::
2024 testpmd> port config (port_id) udp_tunnel_port add|rm vxlan|geneve (udp_port)
2026 Link Bonding Functions
2027 ----------------------
2029 The Link Bonding functions make it possible to dynamically create and
2030 manage link bonding devices from within testpmd interactive prompt.
2032 create bonded device
2033 ~~~~~~~~~~~~~~~~~~~~
2035 Create a new bonding device::
2037 testpmd> create bonded device (mode) (socket)
2039 For example, to create a bonded device in mode 1 on socket 0::
2041 testpmd> create bonded device 1 0
2042 created new bonded device (port X)
2047 Adds Ethernet device to a Link Bonding device::
2049 testpmd> add bonding slave (slave id) (port id)
2051 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
2053 testpmd> add bonding slave 6 10
2056 remove bonding slave
2057 ~~~~~~~~~~~~~~~~~~~~
2059 Removes an Ethernet slave device from a Link Bonding device::
2061 testpmd> remove bonding slave (slave id) (port id)
2063 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
2065 testpmd> remove bonding slave 6 10
2070 Set the Link Bonding mode of a Link Bonding device::
2072 testpmd> set bonding mode (value) (port id)
2074 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
2076 testpmd> set bonding mode 3 10
2081 Set an Ethernet slave device as the primary device on a Link Bonding device::
2083 testpmd> set bonding primary (slave id) (port id)
2085 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
2087 testpmd> set bonding primary 6 10
2092 Set the MAC address of a Link Bonding device::
2094 testpmd> set bonding mac (port id) (mac)
2096 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
2098 testpmd> set bonding mac 10 00:00:00:00:00:01
2100 set bonding xmit_balance_policy
2101 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2103 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
2105 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
2107 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
2109 testpmd> set bonding xmit_balance_policy 10 l34
2112 set bonding mon_period
2113 ~~~~~~~~~~~~~~~~~~~~~~
2115 Set the link status monitoring polling period in milliseconds for a bonding device.
2117 This adds support for PMD slave devices which do not support link status interrupts.
2118 When the mon_period is set to a value greater than 0 then all PMD's which do not support
2119 link status ISR will be queried every polling interval to check if their link status has changed::
2121 testpmd> set bonding mon_period (port_id) (value)
2123 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
2125 testpmd> set bonding mon_period 5 150
2128 set bonding lacp dedicated_queue
2129 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2131 Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic
2132 when in mode 4 (link-aggregration-802.3ad)::
2134 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable)
2137 set bonding agg_mode
2138 ~~~~~~~~~~~~~~~~~~~~
2140 Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad)::
2142 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable)
2148 Show the current configuration of a Link Bonding device::
2150 testpmd> show bonding config (port id)
2153 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
2154 in balance mode with a transmission policy of layer 2+3::
2156 testpmd> show bonding config 9
2158 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
2160 Active Slaves (3): [1 3 4]
2167 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
2168 This is mainly useful for debugging purposes.
2169 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
2170 and fields that can be accessed.
2175 Display the value of a port register::
2177 testpmd> read reg (port_id) (address)
2179 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
2181 testpmd> read reg 0 0xEE00
2182 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
2187 Display a port register bit field::
2189 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
2191 For example, reading the lowest two bits from the register in the example above::
2193 testpmd> read regfield 0 0xEE00 0 1
2194 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
2199 Display a single port register bit::
2201 testpmd> read regbit (port_id) (address) (bit_x)
2203 For example, reading the lowest bit from the register in the example above::
2205 testpmd> read regbit 0 0xEE00 0
2206 port 0 PCI register at offset 0xEE00: bit 0=1
2211 Set the value of a port register::
2213 testpmd> write reg (port_id) (address) (value)
2215 For example, to clear a register::
2217 testpmd> write reg 0 0xEE00 0x0
2218 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
2223 Set bit field of a port register::
2225 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
2227 For example, writing to the register cleared in the example above::
2229 testpmd> write regfield 0 0xEE00 0 1 2
2230 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
2235 Set single bit value of a port register::
2237 testpmd> write regbit (port_id) (address) (bit_x) (value)
2239 For example, to set the high bit in the register from the example above::
2241 testpmd> write regbit 0 0xEE00 31 1
2242 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
2244 Traffic Metering and Policing
2245 -----------------------------
2247 The following section shows functions for configuring traffic metering and
2248 policing on the ethernet device through the use of generic ethdev API.
2250 show port traffic management capability
2251 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2253 Show traffic metering and policing capability of the port::
2255 testpmd> show port meter cap (port_id)
2257 add port meter profile (srTCM rfc2967)
2258 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2260 Add meter profile (srTCM rfc2697) to the ethernet device::
2262 testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \
2267 * ``profile_id``: ID for the meter profile.
2268 * ``cir``: Committed Information Rate (CIR) (bytes/second).
2269 * ``cbs``: Committed Burst Size (CBS) (bytes).
2270 * ``ebs``: Excess Burst Size (EBS) (bytes).
2272 add port meter profile (trTCM rfc2968)
2273 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2275 Add meter profile (srTCM rfc2698) to the ethernet device::
2277 testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \
2278 (cir) (pir) (cbs) (pbs)
2282 * ``profile_id``: ID for the meter profile.
2283 * ``cir``: Committed information rate (bytes/second).
2284 * ``pir``: Peak information rate (bytes/second).
2285 * ``cbs``: Committed burst size (bytes).
2286 * ``pbs``: Peak burst size (bytes).
2288 add port meter profile (trTCM rfc4115)
2289 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2291 Add meter profile (trTCM rfc4115) to the ethernet device::
2293 testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \
2294 (cir) (eir) (cbs) (ebs)
2298 * ``profile_id``: ID for the meter profile.
2299 * ``cir``: Committed information rate (bytes/second).
2300 * ``eir``: Excess information rate (bytes/second).
2301 * ``cbs``: Committed burst size (bytes).
2302 * ``ebs``: Excess burst size (bytes).
2304 delete port meter profile
2305 ~~~~~~~~~~~~~~~~~~~~~~~~~
2307 Delete meter profile from the ethernet device::
2309 testpmd> del port meter profile (port_id) (profile_id)
2314 Create new meter object for the ethernet device::
2316 testpmd> create port meter (port_id) (mtr_id) (profile_id) \
2317 (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \
2318 (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\
2323 * ``mtr_id``: meter object ID.
2324 * ``profile_id``: ID for the meter profile.
2325 * ``meter_enable``: When this parameter has a non-zero value, the meter object
2326 gets enabled at the time of creation, otherwise remains disabled.
2327 * ``g_action``: Policer action for the packet with green color.
2328 * ``y_action``: Policer action for the packet with yellow color.
2329 * ``r_action``: Policer action for the packet with red color.
2330 * ``stats_mask``: Mask of statistics counter types to be enabled for the
2332 * ``shared``: When this parameter has a non-zero value, the meter object is
2333 shared by multiple flows. Otherwise, meter object is used by single flow.
2334 * ``use_pre_meter_color``: When this parameter has a non-zero value, the
2335 input color for the current meter object is determined by the latest meter
2336 object in the same flow. Otherwise, the current meter object uses the
2337 *dscp_table* to determine the input color.
2338 * ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input
2339 color, 0 <= x <= 63.
2344 Enable meter for the ethernet device::
2346 testpmd> enable port meter (port_id) (mtr_id)
2351 Disable meter for the ethernet device::
2353 testpmd> disable port meter (port_id) (mtr_id)
2358 Delete meter for the ethernet device::
2360 testpmd> del port meter (port_id) (mtr_id)
2362 Set port meter profile
2363 ~~~~~~~~~~~~~~~~~~~~~~
2365 Set meter profile for the ethernet device::
2367 testpmd> set port meter profile (port_id) (mtr_id) (profile_id)
2369 set port meter dscp table
2370 ~~~~~~~~~~~~~~~~~~~~~~~~~
2372 Set meter dscp table for the ethernet device::
2374 testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \
2375 (dscp_tbl_entry1)...(dscp_tbl_entry63)]
2377 set port meter policer action
2378 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2380 Set meter policer action for the ethernet device::
2382 testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \
2383 (action0) [(action1) (action1)]
2387 * ``action_mask``: Bit mask indicating which policer actions need to be
2388 updated. One or more policer actions can be updated in a single function
2389 invocation. To update the policer action associated with color C, bit
2390 (1 << C) needs to be set in *action_mask* and element at position C
2391 in the *actions* array needs to be valid.
2392 * ``actionx``: Policer action for the color x,
2393 RTE_MTR_GREEN <= x < RTE_MTR_COLORS
2395 set port meter stats mask
2396 ~~~~~~~~~~~~~~~~~~~~~~~~~
2398 Set meter stats mask for the ethernet device::
2400 testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask)
2404 * ``stats_mask``: Bit mask indicating statistics counter types to be enabled.
2406 show port meter stats
2407 ~~~~~~~~~~~~~~~~~~~~~
2409 Show meter stats of the ethernet device::
2411 testpmd> show port meter stats (port_id) (mtr_id) (clear)
2415 * ``clear``: Flag that indicates whether the statistics counters should
2416 be cleared (i.e. set to zero) immediately after they have been read or not.
2421 The following section shows functions for configuring traffic management on
2422 on the ethernet device through the use of generic TM API.
2424 show port traffic management capability
2425 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2427 Show traffic management capability of the port::
2429 testpmd> show port tm cap (port_id)
2431 show port traffic management capability (hierarchy level)
2432 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2434 Show traffic management hierarchy level capability of the port::
2436 testpmd> show port tm level cap (port_id) (level_id)
2438 show port traffic management capability (hierarchy node level)
2439 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2441 Show the traffic management hierarchy node capability of the port::
2443 testpmd> show port tm node cap (port_id) (node_id)
2445 show port traffic management hierarchy node type
2446 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2448 Show the port traffic management hierarchy node type::
2450 testpmd> show port tm node type (port_id) (node_id)
2452 show port traffic management hierarchy node stats
2453 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2455 Show the port traffic management hierarchy node statistics::
2457 testpmd> show port tm node stats (port_id) (node_id) (clear)
2461 * ``clear``: When this parameter has a non-zero value, the statistics counters
2462 are cleared (i.e. set to zero) immediately after they have been read,
2463 otherwise the statistics counters are left untouched.
2465 Add port traffic management private shaper profile
2466 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2468 Add the port traffic management private shaper profile::
2470 testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \
2471 (tb_rate) (tb_size) (packet_length_adjust)
2475 * ``shaper_profile id``: Shaper profile ID for the new profile.
2476 * ``tb_rate``: Token bucket rate (bytes per second).
2477 * ``tb_size``: Token bucket size (bytes).
2478 * ``packet_length_adjust``: The value (bytes) to be added to the length of
2479 each packet for the purpose of shaping. This parameter value can be used to
2480 correct the packet length with the framing overhead bytes that are consumed
2483 Delete port traffic management private shaper profile
2484 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2486 Delete the port traffic management private shaper::
2488 testpmd> del port tm node shaper profile (port_id) (shaper_profile_id)
2492 * ``shaper_profile id``: Shaper profile ID that needs to be deleted.
2494 Add port traffic management shared shaper
2495 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2497 Create the port traffic management shared shaper::
2499 testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \
2504 * ``shared_shaper_id``: Shared shaper ID to be created.
2505 * ``shaper_profile id``: Shaper profile ID for shared shaper.
2507 Set port traffic management shared shaper
2508 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2510 Update the port traffic management shared shaper::
2512 testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \
2517 * ``shared_shaper_id``: Shared shaper ID to be update.
2518 * ``shaper_profile id``: Shaper profile ID for shared shaper.
2520 Delete port traffic management shared shaper
2521 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2523 Delete the port traffic management shared shaper::
2525 testpmd> del port tm node shared shaper (port_id) (shared_shaper_id)
2529 * ``shared_shaper_id``: Shared shaper ID to be deleted.
2531 Set port traffic management hiearchy node private shaper
2532 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2534 set the port traffic management hierarchy node private shaper::
2536 testpmd> set port tm node shaper profile (port_id) (node_id) \
2541 * ``shaper_profile id``: Private shaper profile ID to be enabled on the
2544 Add port traffic management WRED profile
2545 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2547 Create a new WRED profile::
2549 testpmd> add port tm node wred profile (port_id) (wred_profile_id) \
2550 (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \
2551 (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \
2552 (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r)
2556 * ``wred_profile id``: Identifier for the newly create WRED profile
2557 * ``color_g``: Packet color (green)
2558 * ``min_th_g``: Minimum queue threshold for packet with green color
2559 * ``max_th_g``: Minimum queue threshold for packet with green color
2560 * ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp)
2561 * ``wq_log2_g``: Negated log2 of queue weight (wq)
2562 * ``color_y``: Packet color (yellow)
2563 * ``min_th_y``: Minimum queue threshold for packet with yellow color
2564 * ``max_th_y``: Minimum queue threshold for packet with yellow color
2565 * ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp)
2566 * ``wq_log2_y``: Negated log2 of queue weight (wq)
2567 * ``color_r``: Packet color (red)
2568 * ``min_th_r``: Minimum queue threshold for packet with yellow color
2569 * ``max_th_r``: Minimum queue threshold for packet with yellow color
2570 * ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp)
2571 * ``wq_log2_r``: Negated log2 of queue weight (wq)
2573 Delete port traffic management WRED profile
2574 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2576 Delete the WRED profile::
2578 testpmd> del port tm node wred profile (port_id) (wred_profile_id)
2580 Add port traffic management hierarchy nonleaf node
2581 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2583 Add nonleaf node to port traffic management hiearchy::
2585 testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \
2586 (priority) (weight) (level_id) (shaper_profile_id) \
2587 (n_sp_priorities) (stats_mask) (n_shared_shapers) \
2588 [(shared_shaper_0) (shared_shaper_1) ...] \
2592 * ``parent_node_id``: Node ID of the parent.
2593 * ``priority``: Node priority (highest node priority is zero). This is used by
2594 the SP algorithm running on the parent node for scheduling this node.
2595 * ``weight``: Node weight (lowest weight is one). The node weight is relative
2596 to the weight sum of all siblings that have the same priority. It is used by
2597 the WFQ algorithm running on the parent node for scheduling this node.
2598 * ``level_id``: Hiearchy level of the node.
2599 * ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2601 * ``n_sp_priorities``: Number of strict priorities.
2602 * ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2603 * ``n_shared_shapers``: Number of shared shapers.
2604 * ``shared_shaper_id``: Shared shaper id.
2606 Add port traffic management hierarchy leaf node
2607 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2609 Add leaf node to port traffic management hiearchy::
2611 testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \
2612 (priority) (weight) (level_id) (shaper_profile_id) \
2613 (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \
2614 [(shared_shaper_id) (shared_shaper_id) ...] \
2618 * ``parent_node_id``: Node ID of the parent.
2619 * ``priority``: Node priority (highest node priority is zero). This is used by
2620 the SP algorithm running on the parent node for scheduling this node.
2621 * ``weight``: Node weight (lowest weight is one). The node weight is relative
2622 to the weight sum of all siblings that have the same priority. It is used by
2623 the WFQ algorithm running on the parent node for scheduling this node.
2624 * ``level_id``: Hiearchy level of the node.
2625 * ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2627 * ``cman_mode``: Congestion management mode to be enabled for this node.
2628 * ``wred_profile_id``: WRED profile id to be enabled for this node.
2629 * ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2630 * ``n_shared_shapers``: Number of shared shapers.
2631 * ``shared_shaper_id``: Shared shaper id.
2633 Delete port traffic management hierarchy node
2634 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2636 Delete node from port traffic management hiearchy::
2638 testpmd> del port tm node (port_id) (node_id)
2640 Update port traffic management hierarchy parent node
2641 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2643 Update port traffic management hierarchy parent node::
2645 testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \
2648 This function can only be called after the hierarchy commit invocation. Its
2649 success depends on the port support for this operation, as advertised through
2650 the port capability set. This function is valid for all nodes of the traffic
2651 management hierarchy except root node.
2653 Suspend port traffic management hierarchy node
2654 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2656 testpmd> suspend port tm node (port_id) (node_id)
2658 Resume port traffic management hierarchy node
2659 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2661 testpmd> resume port tm node (port_id) (node_id)
2663 Commit port traffic management hierarchy
2664 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2666 Commit the traffic management hierarchy on the port::
2668 testpmd> port tm hierarchy commit (port_id) (clean_on_fail)
2672 * ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function
2673 call failure. On the other hand, hierarchy is preserved when this parameter
2676 Set port traffic management mark VLAN dei
2677 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2679 Enables/Disables the traffic management marking on the port for VLAN packets::
2681 testpmd> set port tm mark vlan_dei <port_id> <green> <yellow> <red>
2685 * ``port_id``: The port which on which VLAN packets marked as ``green`` or
2686 ``yellow`` or ``red`` will have dei bit enabled
2688 * ``green`` enable 1, disable 0 marking for dei bit of VLAN packets marked as green
2690 * ``yellow`` enable 1, disable 0 marking for dei bit of VLAN packets marked as yellow
2692 * ``red`` enable 1, disable 0 marking for dei bit of VLAN packets marked as red
2694 Set port traffic management mark IP dscp
2695 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2697 Enables/Disables the traffic management marking on the port for IP dscp packets::
2699 testpmd> set port tm mark ip_dscp <port_id> <green> <yellow> <red>
2703 * ``port_id``: The port which on which IP packets marked as ``green`` or
2704 ``yellow`` or ``red`` will have IP dscp bits updated
2706 * ``green`` enable 1, disable 0 marking IP dscp to low drop precedence for green packets
2708 * ``yellow`` enable 1, disable 0 marking IP dscp to medium drop precedence for yellow packets
2710 * ``red`` enable 1, disable 0 marking IP dscp to high drop precedence for red packets
2712 Set port traffic management mark IP ecn
2713 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2715 Enables/Disables the traffic management marking on the port for IP ecn packets::
2717 testpmd> set port tm mark ip_ecn <port_id> <green> <yellow> <red>
2721 * ``port_id``: The port which on which IP packets marked as ``green`` or
2722 ``yellow`` or ``red`` will have IP ecn bits updated
2724 * ``green`` enable 1, disable 0 marking IP ecn for green marked packets with ecn of 2'b01 or 2'b10
2725 to ecn of 2'b11 when IP is caring TCP or SCTP
2727 * ``yellow`` enable 1, disable 0 marking IP ecn for yellow marked packets with ecn of 2'b01 or 2'b10
2728 to ecn of 2'b11 when IP is caring TCP or SCTP
2730 * ``red`` enable 1, disable 0 marking IP ecn for yellow marked packets with ecn of 2'b01 or 2'b10
2731 to ecn of 2'b11 when IP is caring TCP or SCTP
2733 Set port traffic management default hierarchy (softnic forwarding mode)
2734 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2736 set the traffic management default hierarchy on the port::
2738 testpmd> set port tm hierarchy default (port_id)
2743 This section details the available filter functions that are available.
2745 Note these functions interface the deprecated legacy filtering framework,
2746 superseded by *rte_flow*. See `Flow rules management`_.
2749 ~~~~~~~~~~~~~~~~~~~~
2751 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
2753 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
2754 ethertype (ether_type) (drop|fwd) queue (queue_id)
2756 The available information parameters are:
2758 * ``port_id``: The port which the Ethertype filter assigned on.
2760 * ``mac_addr``: Compare destination mac address.
2762 * ``mac_ignr``: Ignore destination mac address match.
2764 * ``mac_address``: Destination mac address to match.
2766 * ``ether_type``: The EtherType value want to match,
2767 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
2769 * ``queue_id``: The receive queue associated with this EtherType filter.
2770 It is meaningless when deleting or dropping.
2772 Example, to add/remove an ethertype filter rule::
2774 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
2775 ethertype 0x0806 fwd queue 3
2777 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
2778 ethertype 0x0806 fwd queue 3
2783 Add or delete a 2-tuple filter,
2784 which identifies packets by specific protocol and destination TCP/UDP port
2785 and forwards packets into one of the receive queues::
2787 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
2788 protocol (protocol_value) mask (mask_value) \
2789 tcp_flags (tcp_flags_value) priority (prio_value) \
2792 The available information parameters are:
2794 * ``port_id``: The port which the 2-tuple filter assigned on.
2796 * ``dst_port_value``: Destination port in L4.
2798 * ``protocol_value``: IP L4 protocol.
2800 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
2802 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
2804 * ``prio_value``: Priority of this filter.
2806 * ``queue_id``: The receive queue associated with this 2-tuple filter.
2808 Example, to add/remove an 2tuple filter rule::
2810 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
2811 tcp_flags 0x02 priority 3 queue 3
2813 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
2814 tcp_flags 0x02 priority 3 queue 3
2819 Add or delete a 5-tuple filter,
2820 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
2821 and routes packets into one of the receive queues::
2823 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
2824 (src_address) dst_port (dst_port_value) \
2825 src_port (src_port_value) protocol (protocol_value) \
2826 mask (mask_value) tcp_flags (tcp_flags_value) \
2827 priority (prio_value) queue (queue_id)
2829 The available information parameters are:
2831 * ``port_id``: The port which the 5-tuple filter assigned on.
2833 * ``dst_address``: Destination IP address.
2835 * ``src_address``: Source IP address.
2837 * ``dst_port_value``: TCP/UDP destination port.
2839 * ``src_port_value``: TCP/UDP source port.
2841 * ``protocol_value``: L4 protocol.
2843 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
2845 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
2847 * ``prio_value``: The priority of this filter.
2849 * ``queue_id``: The receive queue associated with this 5-tuple filter.
2851 Example, to add/remove an 5tuple filter rule::
2853 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2854 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2855 flags 0x0 priority 3 queue 3
2857 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2858 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2859 flags 0x0 priority 3 queue 3
2864 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
2866 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
2868 The available information parameters are:
2870 * ``port_id``: The port which the SYN filter assigned on.
2872 * ``high``: This SYN filter has higher priority than other filters.
2874 * ``low``: This SYN filter has lower priority than other filters.
2876 * ``queue_id``: The receive queue associated with this SYN filter
2880 testpmd> syn_filter 0 add priority high queue 3
2885 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
2886 and routed into one of the receive queues::
2888 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
2889 mask (mask_value) priority (prio_value) queue (queue_id)
2891 The available information parameters are:
2893 * ``port_id``: The port which the Flex filter is assigned on.
2895 * ``len_value``: Filter length in bytes, no greater than 128.
2897 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
2899 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
2901 * ``prio_value``: The priority of this filter.
2903 * ``queue_id``: The receive queue associated with this Flex filter.
2907 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
2908 mask 000C priority 3 queue 3
2910 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
2911 mask 000C priority 3 queue 3
2914 .. _testpmd_flow_director:
2916 flow_director_filter
2917 ~~~~~~~~~~~~~~~~~~~~
2919 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
2921 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
2922 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
2924 * Perfect match filters.
2925 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2926 The masked fields are for IP flow.
2928 * Signature filters.
2929 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
2931 * Perfect-mac-vlan match filters.
2932 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2933 The masked fields are for MAC VLAN flow.
2935 * Perfect-tunnel match filters.
2936 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2937 The masked fields are for tunnel flow.
2939 * Perfect-raw-flow-type match filters.
2940 The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet.
2941 The masked fields are specified by input sets.
2943 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
2944 per flow type and the flexible payload.
2946 The Flow Director can also mask out parts of all of these fields so that filters
2947 are only applied to certain fields or parts of the fields.
2949 Note that for raw flow type mode the source and destination fields in the
2950 raw packet buffer need to be presented in a reversed order with respect
2951 to the expected received packets.
2952 For example: IP source and destination addresses or TCP/UDP/SCTP
2953 source and destination ports
2955 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
2957 # Commands to add flow director filters of different flow types::
2959 flow_director_filter (port_id) mode IP (add|del|update) \
2960 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
2961 src (src_ip_address) dst (dst_ip_address) \
2962 tos (tos_value) proto (proto_value) ttl (ttl_value) \
2963 vlan (vlan_value) flexbytes (flexbytes_value) \
2964 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
2967 flow_director_filter (port_id) mode IP (add|del|update) \
2968 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
2969 src (src_ip_address) (src_port) \
2970 dst (dst_ip_address) (dst_port) \
2971 tos (tos_value) ttl (ttl_value) \
2972 vlan (vlan_value) flexbytes (flexbytes_value) \
2973 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
2976 flow_director_filter (port_id) mode IP (add|del|update) \
2977 flow (ipv4-sctp|ipv6-sctp) \
2978 src (src_ip_address) (src_port) \
2979 dst (dst_ip_address) (dst_port) \
2980 tos (tos_value) ttl (ttl_value) \
2981 tag (verification_tag) vlan (vlan_value) \
2982 flexbytes (flexbytes_value) (drop|fwd) \
2983 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
2985 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
2986 ether (ethertype) flexbytes (flexbytes_value) \
2987 (drop|fwd) pf|vf(vf_id) queue (queue_id)
2990 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
2991 mac (mac_address) vlan (vlan_value) \
2992 flexbytes (flexbytes_value) (drop|fwd) \
2993 queue (queue_id) fd_id (fd_id_value)
2995 flow_director_filter (port_id) mode Tunnel (add|del|update) \
2996 mac (mac_address) vlan (vlan_value) \
2997 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
2998 flexbytes (flexbytes_value) (drop|fwd) \
2999 queue (queue_id) fd_id (fd_id_value)
3001 flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \
3002 (drop|fwd) queue (queue_id) fd_id (fd_id_value) \
3003 packet (packet file name)
3005 For example, to add an ipv4-udp flow type filter::
3007 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
3008 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
3009 fwd pf queue 1 fd_id 1
3011 For example, add an ipv4-other flow type filter::
3013 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
3014 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
3015 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
3020 Flush all flow director filters on a device::
3022 testpmd> flush_flow_director (port_id)
3024 Example, to flush all flow director filter on port 0::
3026 testpmd> flush_flow_director 0
3031 Set flow director's input masks::
3033 flow_director_mask (port_id) mode IP vlan (vlan_value) \
3034 src_mask (ipv4_src) (ipv6_src) (src_port) \
3035 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
3037 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
3039 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
3040 mac (mac_value) tunnel-type (tunnel_type_value) \
3041 tunnel-id (tunnel_id_value)
3043 Example, to set flow director mask on port 0::
3045 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
3046 src_mask 255.255.255.255 \
3047 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
3048 dst_mask 255.255.255.255 \
3049 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
3051 flow_director_flex_mask
3052 ~~~~~~~~~~~~~~~~~~~~~~~
3054 set masks of flow director's flexible payload based on certain flow type::
3056 testpmd> flow_director_flex_mask (port_id) \
3057 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
3058 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
3059 l2_payload|all) (mask)
3061 Example, to set flow director's flex mask for all flow type on port 0::
3063 testpmd> flow_director_flex_mask 0 flow all \
3064 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
3067 flow_director_flex_payload
3068 ~~~~~~~~~~~~~~~~~~~~~~~~~~
3070 Configure flexible payload selection::
3072 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
3074 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
3076 testpmd> flow_director_flex_payload 0 l4 \
3077 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
3079 get_sym_hash_ena_per_port
3080 ~~~~~~~~~~~~~~~~~~~~~~~~~
3082 Get symmetric hash enable configuration per port::
3084 get_sym_hash_ena_per_port (port_id)
3086 For example, to get symmetric hash enable configuration of port 1::
3088 testpmd> get_sym_hash_ena_per_port 1
3090 set_sym_hash_ena_per_port
3091 ~~~~~~~~~~~~~~~~~~~~~~~~~
3093 Set symmetric hash enable configuration per port to enable or disable::
3095 set_sym_hash_ena_per_port (port_id) (enable|disable)
3097 For example, to set symmetric hash enable configuration of port 1 to enable::
3099 testpmd> set_sym_hash_ena_per_port 1 enable
3101 get_hash_global_config
3102 ~~~~~~~~~~~~~~~~~~~~~~
3104 Get the global configurations of hash filters::
3106 get_hash_global_config (port_id)
3108 For example, to get the global configurations of hash filters of port 1::
3110 testpmd> get_hash_global_config 1
3112 set_hash_global_config
3113 ~~~~~~~~~~~~~~~~~~~~~~
3115 Set the global configurations of hash filters::
3117 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
3118 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
3119 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \
3122 For example, to enable simple_xor for flow type of ipv6 on port 2::
3124 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
3129 Set the input set for hash::
3131 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
3132 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
3133 l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
3134 ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
3135 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
3136 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
3137 fld-8th|none) (select|add)
3139 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
3141 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
3146 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
3147 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
3149 Set the input set for flow director::
3151 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
3152 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
3153 l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
3154 ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
3155 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
3156 sctp-dst-port|sctp-veri-tag|none) (select|add)
3158 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
3160 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
3165 Set different GRE key length for input set::
3167 global_config (port_id) gre-key-len (number in bytes)
3169 For example to set GRE key length for input set to 4 bytes on port 0::
3171 testpmd> global_config 0 gre-key-len 4
3174 .. _testpmd_rte_flow:
3176 Flow rules management
3177 ---------------------
3179 Control of the generic flow API (*rte_flow*) is fully exposed through the
3180 ``flow`` command (validation, creation, destruction, queries and operation
3183 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
3184 features simultaneously may cause undefined side-effects and is therefore
3190 Because the ``flow`` command uses dynamic tokens to handle the large number
3191 of possible flow rules combinations, its behavior differs slightly from
3192 other commands, in particular:
3194 - Pressing *?* or the *<tab>* key displays contextual help for the current
3195 token, not that of the entire command.
3197 - Optional and repeated parameters are supported (provided they are listed
3198 in the contextual help).
3200 The first parameter stands for the operation mode. Possible operations and
3201 their general syntax are described below. They are covered in detail in the
3204 - Check whether a flow rule can be created::
3206 flow validate {port_id}
3207 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3208 pattern {item} [/ {item} [...]] / end
3209 actions {action} [/ {action} [...]] / end
3211 - Create a flow rule::
3213 flow create {port_id}
3214 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3215 pattern {item} [/ {item} [...]] / end
3216 actions {action} [/ {action} [...]] / end
3218 - Destroy specific flow rules::
3220 flow destroy {port_id} rule {rule_id} [...]
3222 - Destroy all flow rules::
3224 flow flush {port_id}
3226 - Query an existing flow rule::
3228 flow query {port_id} {rule_id} {action}
3230 - List existing flow rules sorted by priority, filtered by group
3233 flow list {port_id} [group {group_id}] [...]
3235 - Restrict ingress traffic to the defined flow rules::
3237 flow isolate {port_id} {boolean}
3239 Validating flow rules
3240 ~~~~~~~~~~~~~~~~~~~~~
3242 ``flow validate`` reports whether a flow rule would be accepted by the
3243 underlying device in its current state but stops short of creating it. It is
3244 bound to ``rte_flow_validate()``::
3246 flow validate {port_id}
3247 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3248 pattern {item} [/ {item} [...]] / end
3249 actions {action} [/ {action} [...]] / end
3251 If successful, it will show::
3255 Otherwise it will show an error message of the form::
3257 Caught error type [...] ([...]): [...]
3259 This command uses the same parameters as ``flow create``, their format is
3260 described in `Creating flow rules`_.
3262 Check whether redirecting any Ethernet packet received on port 0 to RX queue
3263 index 6 is supported::
3265 testpmd> flow validate 0 ingress pattern eth / end
3266 actions queue index 6 / end
3270 Port 0 does not support TCPv6 rules::
3272 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
3274 Caught error type 9 (specific pattern item): Invalid argument
3280 ``flow create`` validates and creates the specified flow rule. It is bound
3281 to ``rte_flow_create()``::
3283 flow create {port_id}
3284 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3285 pattern {item} [/ {item} [...]] / end
3286 actions {action} [/ {action} [...]] / end
3288 If successful, it will return a flow rule ID usable with other commands::
3290 Flow rule #[...] created
3292 Otherwise it will show an error message of the form::
3294 Caught error type [...] ([...]): [...]
3296 Parameters describe in the following order:
3298 - Attributes (*group*, *priority*, *ingress*, *egress*, *transfer* tokens).
3299 - A matching pattern, starting with the *pattern* token and terminated by an
3301 - Actions, starting with the *actions* token and terminated by an *end*
3304 These translate directly to *rte_flow* objects provided as-is to the
3305 underlying functions.
3307 The shortest valid definition only comprises mandatory tokens::
3309 testpmd> flow create 0 pattern end actions end
3311 Note that PMDs may refuse rules that essentially do nothing such as this
3314 **All unspecified object values are automatically initialized to 0.**
3319 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
3320 specified before the ``pattern`` token.
3322 - ``group {group id}``: priority group.
3323 - ``priority {level}``: priority level within group.
3324 - ``ingress``: rule applies to ingress traffic.
3325 - ``egress``: rule applies to egress traffic.
3326 - ``transfer``: apply rule directly to endpoints found in pattern.
3328 Each instance of an attribute specified several times overrides the previous
3329 value as shown below (group 4 is used)::
3331 testpmd> flow create 0 group 42 group 24 group 4 [...]
3333 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
3335 While not specifying a direction is an error, some rules may allow both
3338 Most rules affect RX therefore contain the ``ingress`` token::
3340 testpmd> flow create 0 ingress pattern [...]
3345 A matching pattern starts after the ``pattern`` token. It is made of pattern
3346 items and is terminated by a mandatory ``end`` item.
3348 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
3349 rte_flow_item_type``).
3351 The ``/`` token is used as a separator between pattern items as shown
3354 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
3356 Note that protocol items like these must be stacked from lowest to highest
3357 layer to make sense. For instance, the following rule is either invalid or
3358 unlikely to match any packet::
3360 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
3362 More information on these restrictions can be found in the *rte_flow*
3365 Several items support additional specification structures, for example
3366 ``ipv4`` allows specifying source and destination addresses as follows::
3368 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
3369 dst is 10.2.0.0 / end [...]
3371 This rule matches all IPv4 traffic with the specified properties.
3373 In this example, ``src`` and ``dst`` are field names of the underlying
3374 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
3375 in a similar fashion.
3377 The ``is`` token means that the subsequent value must be matched exactly,
3378 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
3379 accordingly. Possible assignment tokens are:
3381 - ``is``: match value perfectly (with full bit-mask).
3382 - ``spec``: match value according to configured bit-mask.
3383 - ``last``: specify upper bound to establish a range.
3384 - ``mask``: specify bit-mask with relevant bits set to one.
3385 - ``prefix``: generate bit-mask from a prefix length.
3387 These yield identical results::
3389 ipv4 src is 10.1.1.1
3393 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
3397 ipv4 src spec 10.1.1.1 src prefix 32
3401 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
3405 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
3407 Inclusive ranges can be defined with ``last``::
3409 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
3411 Note that ``mask`` affects both ``spec`` and ``last``::
3413 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
3414 # matches 10.1.0.0 to 10.2.255.255
3416 Properties can be modified multiple times::
3418 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
3422 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
3427 This section lists supported pattern items and their attributes, if any.
3429 - ``end``: end list of pattern items.
3431 - ``void``: no-op pattern item.
3433 - ``invert``: perform actions when pattern does not match.
3435 - ``any``: match any protocol for the current layer.
3437 - ``num {unsigned}``: number of layers covered.
3439 - ``pf``: match traffic from/to the physical function.
3441 - ``vf``: match traffic from/to a virtual function ID.
3443 - ``id {unsigned}``: VF ID.
3445 - ``phy_port``: match traffic from/to a specific physical port.
3447 - ``index {unsigned}``: physical port index.
3449 - ``port_id``: match traffic from/to a given DPDK port ID.
3451 - ``id {unsigned}``: DPDK port ID.
3453 - ``mark``: match value set in previously matched flow rule using the mark action.
3455 - ``id {unsigned}``: arbitrary integer value.
3457 - ``raw``: match an arbitrary byte string.
3459 - ``relative {boolean}``: look for pattern after the previous item.
3460 - ``search {boolean}``: search pattern from offset (see also limit).
3461 - ``offset {integer}``: absolute or relative offset for pattern.
3462 - ``limit {unsigned}``: search area limit for start of pattern.
3463 - ``pattern {string}``: byte string to look for.
3465 - ``eth``: match Ethernet header.
3467 - ``dst {MAC-48}``: destination MAC.
3468 - ``src {MAC-48}``: source MAC.
3469 - ``type {unsigned}``: EtherType or TPID.
3471 - ``vlan``: match 802.1Q/ad VLAN tag.
3473 - ``tci {unsigned}``: tag control information.
3474 - ``pcp {unsigned}``: priority code point.
3475 - ``dei {unsigned}``: drop eligible indicator.
3476 - ``vid {unsigned}``: VLAN identifier.
3477 - ``inner_type {unsigned}``: inner EtherType or TPID.
3479 - ``ipv4``: match IPv4 header.
3481 - ``tos {unsigned}``: type of service.
3482 - ``ttl {unsigned}``: time to live.
3483 - ``proto {unsigned}``: next protocol ID.
3484 - ``src {ipv4 address}``: source address.
3485 - ``dst {ipv4 address}``: destination address.
3487 - ``ipv6``: match IPv6 header.
3489 - ``tc {unsigned}``: traffic class.
3490 - ``flow {unsigned}``: flow label.
3491 - ``proto {unsigned}``: protocol (next header).
3492 - ``hop {unsigned}``: hop limit.
3493 - ``src {ipv6 address}``: source address.
3494 - ``dst {ipv6 address}``: destination address.
3496 - ``icmp``: match ICMP header.
3498 - ``type {unsigned}``: ICMP packet type.
3499 - ``code {unsigned}``: ICMP packet code.
3501 - ``udp``: match UDP header.
3503 - ``src {unsigned}``: UDP source port.
3504 - ``dst {unsigned}``: UDP destination port.
3506 - ``tcp``: match TCP header.
3508 - ``src {unsigned}``: TCP source port.
3509 - ``dst {unsigned}``: TCP destination port.
3511 - ``sctp``: match SCTP header.
3513 - ``src {unsigned}``: SCTP source port.
3514 - ``dst {unsigned}``: SCTP destination port.
3515 - ``tag {unsigned}``: validation tag.
3516 - ``cksum {unsigned}``: checksum.
3518 - ``vxlan``: match VXLAN header.
3520 - ``vni {unsigned}``: VXLAN identifier.
3522 - ``e_tag``: match IEEE 802.1BR E-Tag header.
3524 - ``grp_ecid_b {unsigned}``: GRP and E-CID base.
3526 - ``nvgre``: match NVGRE header.
3528 - ``tni {unsigned}``: virtual subnet ID.
3530 - ``mpls``: match MPLS header.
3532 - ``label {unsigned}``: MPLS label.
3534 - ``gre``: match GRE header.
3536 - ``protocol {unsigned}``: protocol type.
3538 - ``fuzzy``: fuzzy pattern match, expect faster than default.
3540 - ``thresh {unsigned}``: accuracy threshold.
3542 - ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header.
3544 - ``teid {unsigned}``: tunnel endpoint identifier.
3546 - ``geneve``: match GENEVE header.
3548 - ``vni {unsigned}``: virtual network identifier.
3549 - ``protocol {unsigned}``: protocol type.
3551 - ``vxlan-gpe``: match VXLAN-GPE header.
3553 - ``vni {unsigned}``: VXLAN-GPE identifier.
3555 - ``arp_eth_ipv4``: match ARP header for Ethernet/IPv4.
3557 - ``sha {MAC-48}``: sender hardware address.
3558 - ``spa {ipv4 address}``: sender IPv4 address.
3559 - ``tha {MAC-48}``: target hardware address.
3560 - ``tpa {ipv4 address}``: target IPv4 address.
3562 - ``ipv6_ext``: match presence of any IPv6 extension header.
3564 - ``next_hdr {unsigned}``: next header.
3566 - ``icmp6``: match any ICMPv6 header.
3568 - ``type {unsigned}``: ICMPv6 type.
3569 - ``code {unsigned}``: ICMPv6 code.
3571 - ``icmp6_nd_ns``: match ICMPv6 neighbor discovery solicitation.
3573 - ``target_addr {ipv6 address}``: target address.
3575 - ``icmp6_nd_na``: match ICMPv6 neighbor discovery advertisement.
3577 - ``target_addr {ipv6 address}``: target address.
3579 - ``icmp6_nd_opt``: match presence of any ICMPv6 neighbor discovery option.
3581 - ``type {unsigned}``: ND option type.
3583 - ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery source Ethernet
3584 link-layer address option.
3586 - ``sla {MAC-48}``: source Ethernet LLA.
3588 - ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery target Ethernet
3589 link-layer address option.
3591 - ``tla {MAC-48}``: target Ethernet LLA.
3596 A list of actions starts after the ``actions`` token in the same fashion as
3597 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
3598 terminated by a mandatory ``end`` action.
3600 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
3601 rte_flow_action_type``).
3603 Dropping all incoming UDPv4 packets can be expressed as follows::
3605 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3608 Several actions have configurable properties which must be specified when
3609 there is no valid default value. For example, ``queue`` requires a target
3612 This rule redirects incoming UDPv4 traffic to queue index 6::
3614 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3615 actions queue index 6 / end
3617 While this one could be rejected by PMDs (unspecified queue index)::
3619 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3622 As defined by *rte_flow*, the list is not ordered, all actions of a given
3623 rule are performed simultaneously. These are equivalent::
3625 queue index 6 / void / mark id 42 / end
3629 void / mark id 42 / queue index 6 / end
3631 All actions in a list should have different types, otherwise only the last
3632 action of a given type is taken into account::
3634 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
3638 drop / drop / drop / end # drop is performed only once
3642 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
3644 Considering they are performed simultaneously, opposite and overlapping
3645 actions can sometimes be combined when the end result is unambiguous::
3647 drop / queue index 6 / end # drop has no effect
3651 queue index 6 / rss queues 6 7 8 / end # queue has no effect
3655 drop / passthru / end # drop has no effect
3657 Note that PMDs may still refuse such combinations.
3662 This section lists supported actions and their attributes, if any.
3664 - ``end``: end list of actions.
3666 - ``void``: no-op action.
3668 - ``passthru``: let subsequent rule process matched packets.
3670 - ``jump``: redirect traffic to group on device.
3672 - ``group {unsigned}``: group to redirect to.
3674 - ``mark``: attach 32 bit value to packets.
3676 - ``id {unsigned}``: 32 bit value to return with packets.
3678 - ``flag``: flag packets.
3680 - ``queue``: assign packets to a given queue index.
3682 - ``index {unsigned}``: queue index to use.
3684 - ``drop``: drop packets (note: passthru has priority).
3686 - ``count``: enable counters for this rule.
3688 - ``rss``: spread packets among several queues.
3690 - ``func {hash function}``: RSS hash function to apply, allowed tokens are
3691 the same as `set_hash_global_config`_.
3693 - ``level {unsigned}``: encapsulation level for ``types``.
3695 - ``types [{RSS hash type} [...]] end``: specific RSS hash types, allowed
3696 tokens are the same as `set_hash_input_set`_, except that an empty list
3697 does not disable RSS but instead requests unspecified "best-effort"
3700 - ``key {string}``: RSS hash key, overrides ``key_len``.
3702 - ``key_len {unsigned}``: RSS hash key length in bytes, can be used in
3703 conjunction with ``key`` to pad or truncate it.
3705 - ``queues [{unsigned} [...]] end``: queue indices to use.
3707 - ``pf``: direct traffic to physical function.
3709 - ``vf``: direct traffic to a virtual function ID.
3711 - ``original {boolean}``: use original VF ID if possible.
3712 - ``id {unsigned}``: VF ID.
3714 - ``phy_port``: direct packets to physical port index.
3716 - ``original {boolean}``: use original port index if possible.
3717 - ``index {unsigned}``: physical port index.
3719 - ``port_id``: direct matching traffic to a given DPDK port ID.
3721 - ``original {boolean}``: use original DPDK port ID if possible.
3722 - ``id {unsigned}``: DPDK port ID.
3724 - ``of_set_mpls_ttl``: OpenFlow's ``OFPAT_SET_MPLS_TTL``.
3726 - ``mpls_ttl``: MPLS TTL.
3728 - ``of_dec_mpls_ttl``: OpenFlow's ``OFPAT_DEC_MPLS_TTL``.
3730 - ``of_set_nw_ttl``: OpenFlow's ``OFPAT_SET_NW_TTL``.
3732 - ``nw_ttl``: IP TTL.
3734 - ``of_dec_nw_ttl``: OpenFlow's ``OFPAT_DEC_NW_TTL``.
3736 - ``of_copy_ttl_out``: OpenFlow's ``OFPAT_COPY_TTL_OUT``.
3738 - ``of_copy_ttl_in``: OpenFlow's ``OFPAT_COPY_TTL_IN``.
3740 - ``of_pop_vlan``: OpenFlow's ``OFPAT_POP_VLAN``.
3742 - ``of_push_vlan``: OpenFlow's ``OFPAT_PUSH_VLAN``.
3744 - ``ethertype``: Ethertype.
3746 - ``of_set_vlan_vid``: OpenFlow's ``OFPAT_SET_VLAN_VID``.
3748 - ``vlan_vid``: VLAN id.
3750 - ``of_set_vlan_pcp``: OpenFlow's ``OFPAT_SET_VLAN_PCP``.
3752 - ``vlan_pcp``: VLAN priority.
3754 - ``of_pop_mpls``: OpenFlow's ``OFPAT_POP_MPLS``.
3756 - ``ethertype``: Ethertype.
3758 - ``of_push_mpls``: OpenFlow's ``OFPAT_PUSH_MPLS``.
3760 - ``ethertype``: Ethertype.
3762 - ``vxlan_encap``: Performs a VXLAN encapsulation, outer layer configuration
3763 is done through `Config VXLAN Encap outer layers`_.
3765 - ``vxlan_decap``: Performs a decapsulation action by stripping all headers of
3766 the VXLAN tunnel network overlay from the matched flow.
3768 - ``nvgre_encap``: Performs a NVGRE encapsulation, outer layer configuration
3769 is done through `Config NVGRE Encap outer layers`_.
3771 - ``nvgre_decap``: Performs a decapsulation action by stripping all headers of
3772 the NVGRE tunnel network overlay from the matched flow.
3774 - ``set_ipv4_src``: Set a new IPv4 source address in the outermost IPv4 header.
3776 - ``ipv4_addr``: New IPv4 source address.
3778 - ``set_ipv4_dst``: Set a new IPv4 destination address in the outermost IPv4
3781 - ``ipv4_addr``: New IPv4 destination address.
3783 - ``set_ipv6_src``: Set a new IPv6 source address in the outermost IPv6 header.
3785 - ``ipv6_addr``: New IPv6 source address.
3787 - ``set_ipv6_dst``: Set a new IPv6 destination address in the outermost IPv6
3790 - ``ipv6_addr``: New IPv6 destination address.
3792 - ``of_set_tp_src``: Set a new source port number in the outermost TCP/UDP
3795 - ``port``: New TCP/UDP source port number.
3797 - ``of_set_tp_dst``: Set a new destination port number in the outermost TCP/UDP
3800 - ``port``: New TCP/UDP destination port number.
3802 - ``mac_swap``: Swap the source and destination MAC addresses in the outermost
3805 - ``dec_ttl``: Performs a decrease TTL value action
3807 - ``set_ttl``: Set TTL value with specificed value
3808 - ``ttl_value {unsigned}``: The new TTL value to be set
3810 Destroying flow rules
3811 ~~~~~~~~~~~~~~~~~~~~~
3813 ``flow destroy`` destroys one or more rules from their rule ID (as returned
3814 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
3815 times as necessary::
3817 flow destroy {port_id} rule {rule_id} [...]
3819 If successful, it will show::
3821 Flow rule #[...] destroyed
3823 It does not report anything for rule IDs that do not exist. The usual error
3824 message is shown when a rule cannot be destroyed::
3826 Caught error type [...] ([...]): [...]
3828 ``flow flush`` destroys all rules on a device and does not take extra
3829 arguments. It is bound to ``rte_flow_flush()``::
3831 flow flush {port_id}
3833 Any errors are reported as above.
3835 Creating several rules and destroying them::
3837 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3838 actions queue index 2 / end
3839 Flow rule #0 created
3840 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3841 actions queue index 3 / end
3842 Flow rule #1 created
3843 testpmd> flow destroy 0 rule 0 rule 1
3844 Flow rule #1 destroyed
3845 Flow rule #0 destroyed
3848 The same result can be achieved using ``flow flush``::
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 flush 0
3859 Non-existent rule IDs are ignored::
3861 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3862 actions queue index 2 / end
3863 Flow rule #0 created
3864 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3865 actions queue index 3 / end
3866 Flow rule #1 created
3867 testpmd> flow destroy 0 rule 42 rule 10 rule 2
3869 testpmd> flow destroy 0 rule 0
3870 Flow rule #0 destroyed
3876 ``flow query`` queries a specific action of a flow rule having that
3877 ability. Such actions collect information that can be reported using this
3878 command. It is bound to ``rte_flow_query()``::
3880 flow query {port_id} {rule_id} {action}
3882 If successful, it will display either the retrieved data for known actions
3883 or the following message::
3885 Cannot display result for action type [...] ([...])
3887 Otherwise, it will complain either that the rule does not exist or that some
3890 Flow rule #[...] not found
3894 Caught error type [...] ([...]): [...]
3896 Currently only the ``count`` action is supported. This action reports the
3897 number of packets that hit the flow rule and the total number of bytes. Its
3898 output has the following format::
3901 hits_set: [...] # whether "hits" contains a valid value
3902 bytes_set: [...] # whether "bytes" contains a valid value
3903 hits: [...] # number of packets
3904 bytes: [...] # number of bytes
3906 Querying counters for TCPv6 packets redirected to queue 6::
3908 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
3909 actions queue index 6 / count / end
3910 Flow rule #4 created
3911 testpmd> flow query 0 4 count
3922 ``flow list`` lists existing flow rules sorted by priority and optionally
3923 filtered by group identifiers::
3925 flow list {port_id} [group {group_id}] [...]
3927 This command only fails with the following message if the device does not
3932 Output consists of a header line followed by a short description of each
3933 flow rule, one per line. There is no output at all when no flow rules are
3934 configured on the device::
3936 ID Group Prio Attr Rule
3937 [...] [...] [...] [...] [...]
3939 ``Attr`` column flags:
3941 - ``i`` for ``ingress``.
3942 - ``e`` for ``egress``.
3944 Creating several flow rules and listing them::
3946 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3947 actions queue index 6 / end
3948 Flow rule #0 created
3949 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3950 actions queue index 2 / end
3951 Flow rule #1 created
3952 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
3953 actions rss queues 6 7 8 end / end
3954 Flow rule #2 created
3955 testpmd> flow list 0
3956 ID Group Prio Attr Rule
3957 0 0 0 i- ETH IPV4 => QUEUE
3958 1 0 0 i- ETH IPV6 => QUEUE
3959 2 0 5 i- ETH IPV4 UDP => RSS
3962 Rules are sorted by priority (i.e. group ID first, then priority level)::
3964 testpmd> flow list 1
3965 ID Group Prio Attr Rule
3966 0 0 0 i- ETH => COUNT
3967 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3968 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3969 1 24 0 i- ETH IPV4 UDP => QUEUE
3970 4 24 10 i- ETH IPV4 TCP => DROP
3971 3 24 20 i- ETH IPV4 => DROP
3972 2 24 42 i- ETH IPV4 UDP => QUEUE
3973 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3976 Output can be limited to specific groups::
3978 testpmd> flow list 1 group 0 group 63
3979 ID Group Prio Attr Rule
3980 0 0 0 i- ETH => COUNT
3981 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3982 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3983 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3986 Toggling isolated mode
3987 ~~~~~~~~~~~~~~~~~~~~~~
3989 ``flow isolate`` can be used to tell the underlying PMD that ingress traffic
3990 must only be injected from the defined flow rules; that no default traffic
3991 is expected outside those rules and the driver is free to assign more
3992 resources to handle them. It is bound to ``rte_flow_isolate()``::
3994 flow isolate {port_id} {boolean}
3996 If successful, enabling or disabling isolated mode shows either::
3998 Ingress traffic on port [...]
3999 is now restricted to the defined flow rules
4003 Ingress traffic on port [...]
4004 is not restricted anymore to the defined flow rules
4006 Otherwise, in case of error::
4008 Caught error type [...] ([...]): [...]
4010 Mainly due to its side effects, PMDs supporting this mode may not have the
4011 ability to toggle it more than once without reinitializing affected ports
4012 first (e.g. by exiting testpmd).
4014 Enabling isolated mode::
4016 testpmd> flow isolate 0 true
4017 Ingress traffic on port 0 is now restricted to the defined flow rules
4020 Disabling isolated mode::
4022 testpmd> flow isolate 0 false
4023 Ingress traffic on port 0 is not restricted anymore to the defined flow rules
4026 Sample QinQ flow rules
4027 ~~~~~~~~~~~~~~~~~~~~~~
4029 Before creating QinQ rule(s) the following commands should be issued to enable QinQ::
4031 testpmd> port stop 0
4032 testpmd> vlan set qinq on 0
4034 The above command sets the inner and outer TPID's to 0x8100.
4036 To change the TPID's the following commands should be used::
4038 testpmd> vlan set outer tpid 0xa100 0
4039 testpmd> vlan set inner tpid 0x9100 0
4040 testpmd> port start 0
4042 Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM.
4046 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 /
4047 vlan tci is 456 / end actions vf id 1 / queue index 0 / end
4048 Flow rule #0 validated
4050 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 /
4051 vlan tci is 456 / end actions vf id 123 / queue index 0 / end
4052 Flow rule #0 created
4054 testpmd> flow list 0
4055 ID Group Prio Attr Rule
4056 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
4058 Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host.
4062 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 /
4063 vlan tci is 654 / end actions pf / queue index 0 / end
4064 Flow rule #1 validated
4066 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 /
4067 vlan tci is 654 / end actions pf / queue index 1 / end
4068 Flow rule #1 created
4070 testpmd> flow list 0
4071 ID Group Prio Attr Rule
4072 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
4073 1 0 0 i- ETH VLAN VLAN=>PF QUEUE
4075 Sample VXLAN encapsulation rule
4076 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
4078 VXLAN encapsulation outer layer has default value pre-configured in testpmd
4079 source code, those can be changed by using the following commands
4081 IPv4 VXLAN outer header::
4083 testpmd> set vxlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src 127.0.0.1
4084 ip-dst 128.0.0.1 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
4085 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
4088 testpmd> set vxlan-with-vlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src
4089 127.0.0.1 ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11
4090 eth-dst 22:22:22:22:22:22
4091 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
4094 IPv6 VXLAN outer header::
4096 testpmd> set vxlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4 ip-src ::1
4097 ip-dst ::2222 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 ipv6 vni 4 udp-src 4 udp-dst 4
4102 ip-src ::1 ip-dst ::2222 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 Sample NVGRE encapsulation rule
4108 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
4110 NVGRE encapsulation outer layer has default value pre-configured in testpmd
4111 source code, those can be changed by using the following commands
4113 IPv4 NVGRE outer header::
4115 testpmd> set nvgre ip-version ipv4 tni 4 ip-src 127.0.0.1 ip-dst 128.0.0.1
4116 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
4117 testpmd> flow create 0 ingress pattern end actions nvgre_encap /
4120 testpmd> set nvgre-with-vlan ip-version ipv4 tni 4 ip-src 127.0.0.1
4121 ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11
4122 eth-dst 22:22:22:22:22:22
4123 testpmd> flow create 0 ingress pattern end actions nvgre_encap /
4126 IPv6 NVGRE outer header::
4128 testpmd> set nvgre ip-version ipv6 tni 4 ip-src ::1 ip-dst ::2222
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 ipv6 tni 4 ip-src ::1 ip-dst ::2222
4134 vlan-tci 34 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
4135 testpmd> flow create 0 ingress pattern end actions nvgre_encap /
4141 The following sections show functions to load/unload eBPF based filters.
4146 Load an eBPF program as a callback for partciular RX/TX queue::
4148 testpmd> bpf-load rx|tx (portid) (queueid) (load-flags) (bpf-prog-filename)
4150 The available load-flags are:
4152 * ``J``: use JIT generated native code, otherwise BPF interpreter will be used.
4154 * ``M``: assume input parameter is a pointer to rte_mbuf, otherwise assume it is a pointer to first segment's data.
4160 You'll need clang v3.7 or above to build bpf program you'd like to load
4164 .. code-block:: console
4167 clang -O2 -target bpf -c t1.c
4169 Then to load (and JIT compile) t1.o at RX queue 0, port 1::
4171 .. code-block:: console
4173 testpmd> bpf-load rx 1 0 J ./dpdk.org/test/bpf/t1.o
4175 To load (not JITed) t1.o at TX queue 0, port 0::
4177 .. code-block:: console
4179 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o
4184 Unload previously loaded eBPF program for partciular RX/TX queue::
4186 testpmd> bpf-unload rx|tx (portid) (queueid)
4188 For example to unload BPF filter from TX queue 0, port 0:
4190 .. code-block:: console
4192 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o