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|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 * ``stats``: RX/TX statistics.
170 * ``xstats``: RX/TX extended NIC statistics.
172 * ``fdir``: Flow Director information and statistics.
174 * ``stat_qmap``: Queue statistics mapping.
176 * ``dcb_tc``: DCB information such as TC mapping.
178 * ``cap``: Supported offload capabilities.
182 .. code-block:: console
184 testpmd> show port info 0
186 ********************* Infos for port 0 *********************
188 MAC address: XX:XX:XX:XX:XX:XX
190 memory allocation on the socket: 0
192 Link speed: 40000 Mbps
193 Link duplex: full-duplex
194 Promiscuous mode: enabled
195 Allmulticast mode: disabled
196 Maximum number of MAC addresses: 64
197 Maximum number of MAC addresses of hash filtering: 0
202 Redirection table size: 512
203 Supported flow types:
223 Display the rss redirection table entry indicated by masks on port X::
225 testpmd> show port (port_id) rss reta (size) (mask0, mask1...)
227 size is used to indicate the hardware supported reta size
232 Display the RSS hash functions and RSS hash key of a port::
234 testpmd> show port (port_id) rss-hash ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2-payload|ipv6-ex|ipv6-tcp-ex|ipv6-udp-ex [key]
239 Clear the port statistics for a given port or for all ports::
241 testpmd> clear port (info|stats|xstats|fdir|stat_qmap) (port_id|all)
245 testpmd> clear port stats all
250 Display information for a given port's RX/TX queue::
252 testpmd> show (rxq|txq) info (port_id) (queue_id)
257 Displays the configuration of the application.
258 The configuration comes from the command-line, the runtime or the application defaults::
260 testpmd> show config (rxtx|cores|fwd|txpkts)
262 The available information categories are:
264 * ``rxtx``: RX/TX configuration items.
266 * ``cores``: List of forwarding cores.
268 * ``fwd``: Packet forwarding configuration.
270 * ``txpkts``: Packets to TX configuration.
274 .. code-block:: console
276 testpmd> show config rxtx
278 io packet forwarding - CRC stripping disabled - packets/burst=16
279 nb forwarding cores=2 - nb forwarding ports=1
280 RX queues=1 - RX desc=128 - RX free threshold=0
281 RX threshold registers: pthresh=8 hthresh=8 wthresh=4
282 TX queues=1 - TX desc=512 - TX free threshold=0
283 TX threshold registers: pthresh=36 hthresh=0 wthresh=0
284 TX RS bit threshold=0 - TXQ flags=0x0
289 Set the packet forwarding mode::
291 testpmd> set fwd (io|mac|macswap|flowgen| \
292 rxonly|txonly|csum|icmpecho) (""|retry)
294 ``retry`` can be specified for forwarding engines except ``rx_only``.
296 The available information categories are:
298 * ``io``: Forwards packets "as-is" in I/O mode.
299 This is the fastest possible forwarding operation as it does not access packets data.
300 This is the default mode.
302 * ``mac``: Changes the source and the destination Ethernet addresses of packets before forwarding them.
303 Default application behaviour is to set source Ethernet address to that of the transmitting interface, and destination
304 address to a dummy value (set during init). The user may specify a target destination Ethernet address via the 'eth-peer' or
305 'eth-peer-configfile' command-line options. It is not currently possible to specify a specific source Ethernet address.
307 * ``macswap``: MAC swap forwarding mode.
308 Swaps the source and the destination Ethernet addresses of packets before forwarding them.
310 * ``flowgen``: Multi-flow generation mode.
311 Originates a number of flows (with varying destination IP addresses), and terminate receive traffic.
313 * ``rxonly``: Receives packets but doesn't transmit them.
315 * ``txonly``: Generates and transmits packets without receiving any.
317 * ``csum``: Changes the checksum field with hardware or software methods depending on the offload flags on the packet.
319 * ``icmpecho``: Receives a burst of packets, lookup for IMCP echo requests and, if any, send back ICMP echo replies.
321 * ``ieee1588``: Demonstrate L2 IEEE1588 V2 PTP timestamping for RX and TX. Requires ``CONFIG_RTE_LIBRTE_IEEE1588=y``.
323 * ``tm``: Traffic Management forwarding mode
324 Demonstrates the use of ethdev traffic management APIs and softnic PMD for
325 QoS traffic management. In this mode, 5-level hierarchical QoS scheduler is
326 available as an default option that can be enabled through CLI. The user can
327 also modify the default hierarchy or specify the new hierarchy through CLI for
328 implementing QoS scheduler. Requires ``CONFIG_RTE_LIBRTE_PMD_SOFTNIC=y`` ``CONFIG_RTE_LIBRTE_SCHED=y``.
332 testpmd> set fwd rxonly
334 Set rxonly packet forwarding mode
340 Display an RX descriptor for a port RX queue::
342 testpmd> read rxd (port_id) (queue_id) (rxd_id)
346 testpmd> read rxd 0 0 4
347 0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180
352 Display a TX descriptor for a port TX queue::
354 testpmd> read txd (port_id) (queue_id) (txd_id)
358 testpmd> read txd 0 0 4
359 0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C
364 Get loaded dynamic device personalization (DDP) package info list::
366 testpmd> ddp get list (port_id)
371 Display information about dynamic device personalization (DDP) profile::
373 testpmd> ddp get info (profile_path)
378 Display VF statistics::
380 testpmd> show vf stats (port_id) (vf_id)
385 Reset VF statistics::
387 testpmd> clear vf stats (port_id) (vf_id)
389 show port pctype mapping
390 ~~~~~~~~~~~~~~~~~~~~~~~~
392 List all items from the pctype mapping table::
394 testpmd> show port (port_id) pctype mapping
397 Configuration Functions
398 -----------------------
400 The testpmd application can be configured from the runtime as well as from the command-line.
402 This section details the available configuration functions that are available.
406 Configuration changes only become active when forwarding is started/restarted.
411 Reset forwarding to the default configuration::
418 Set the debug verbosity level::
420 testpmd> set verbose (level)
422 Currently the only available levels are 0 (silent except for error) and 1 (fully verbose).
427 Set the log level for a log type::
429 testpmd> set log global|(type) (level)
433 * ``type`` is the log name.
435 * ``level`` is the log level.
437 For example, to change the global log level::
438 testpmd> set log global (level)
440 Regexes can also be used for type. To change log level of user1, user2 and user3::
441 testpmd> set log user[1-3] (level)
446 Set the number of ports used by the application:
450 This is equivalent to the ``--nb-ports`` command-line option.
455 Set the number of cores used by the application::
457 testpmd> set nbcore (num)
459 This is equivalent to the ``--nb-cores`` command-line option.
463 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port.
468 Set the forwarding cores hexadecimal mask::
470 testpmd> set coremask (mask)
472 This is equivalent to the ``--coremask`` command-line option.
476 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding.
481 Set the forwarding ports hexadecimal mask::
483 testpmd> set portmask (mask)
485 This is equivalent to the ``--portmask`` command-line option.
490 Set number of packets per burst::
492 testpmd> set burst (num)
494 This is equivalent to the ``--burst command-line`` option.
496 When retry is enabled, the transmit delay time and number of retries can also be set::
498 testpmd> set burst tx delay (microseconds) retry (num)
503 Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode::
505 testpmd> set txpkts (x[,y]*)
507 Where x[,y]* represents a CSV list of values, without white space.
512 Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes::
514 testpmd> set txsplit (off|on|rand)
518 * ``off`` disable packet copy & split for CSUM mode.
520 * ``on`` split outgoing packet into multiple segments. Size of each segment
521 and number of segments per packet is determined by ``set txpkts`` command
524 * ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments.
529 Set the list of forwarding cores::
531 testpmd> set corelist (x[,y]*)
533 For example, to change the forwarding cores:
535 .. code-block:: console
537 testpmd> set corelist 3,1
538 testpmd> show config fwd
540 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled
541 Logical Core 3 (socket 0) forwards packets on 1 streams:
542 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01
543 Logical Core 1 (socket 0) forwards packets on 1 streams:
544 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
548 The cores are used in the same order as specified on the command line.
553 Set the list of forwarding ports::
555 testpmd> set portlist (x[,y]*)
557 For example, to change the port forwarding:
559 .. code-block:: console
561 testpmd> set portlist 0,2,1,3
562 testpmd> show config fwd
564 io packet forwarding - ports=4 - cores=1 - streams=4
565 Logical Core 3 (socket 0) forwards packets on 4 streams:
566 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01
567 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
568 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03
569 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02
574 Enable/disable tx loopback::
576 testpmd> set tx loopback (port_id) (on|off)
581 set drop enable bit for all queues::
583 testpmd> set all queues drop (port_id) (on|off)
585 set split drop enable (for VF)
586 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
588 set split drop enable bit for VF from PF::
590 testpmd> set vf split drop (port_id) (vf_id) (on|off)
592 set mac antispoof (for VF)
593 ~~~~~~~~~~~~~~~~~~~~~~~~~~
595 Set mac antispoof for a VF from the PF::
597 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off)
602 Enable/disable MACsec offload::
604 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off)
605 testpmd> set macsec offload (port_id) off
610 Configure MACsec secure connection (SC)::
612 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi)
616 The pi argument is ignored for tx.
617 Check the NIC Datasheet for hardware limits.
622 Configure MACsec secure association (SA)::
624 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key)
628 The IDX value must be 0 or 1.
629 Check the NIC Datasheet for hardware limits.
631 set broadcast mode (for VF)
632 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
634 Set broadcast mode for a VF from the PF::
636 testpmd> set vf broadcast (port_id) (vf_id) (on|off)
641 Set the VLAN strip on a port::
643 testpmd> vlan set strip (on|off) (port_id)
648 Set the VLAN strip for a queue on a port::
650 testpmd> vlan set stripq (on|off) (port_id,queue_id)
652 vlan set stripq (for VF)
653 ~~~~~~~~~~~~~~~~~~~~~~~~
655 Set VLAN strip for all queues in a pool for a VF from the PF::
657 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off)
659 vlan set insert (for VF)
660 ~~~~~~~~~~~~~~~~~~~~~~~~
662 Set VLAN insert for a VF from the PF::
664 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id)
666 vlan set tag (for VF)
667 ~~~~~~~~~~~~~~~~~~~~~
669 Set VLAN tag for a VF from the PF::
671 testpmd> set vf vlan tag (port_id) (vf_id) (on|off)
673 vlan set antispoof (for VF)
674 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
676 Set VLAN antispoof for a VF from the PF::
678 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off)
683 Set the VLAN filter on a port::
685 testpmd> vlan set filter (on|off) (port_id)
690 Set the VLAN QinQ (extended queue in queue) on for a port::
692 testpmd> vlan set qinq (on|off) (port_id)
697 Set the inner or outer VLAN TPID for packet filtering on a port::
699 testpmd> vlan set (inner|outer) tpid (value) (port_id)
703 TPID value must be a 16-bit number (value <= 65536).
708 Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID::
710 testpmd> rx_vlan add (vlan_id|all) (port_id)
714 VLAN filter must be set on that port. VLAN ID < 4096.
715 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries
716 in VFTA table. This is important if enabling all vlan_ids.
721 Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID::
723 testpmd> rx_vlan rm (vlan_id|all) (port_id)
728 Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID::
730 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask)
735 Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID::
737 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask)
742 Add a tunnel filter on a port::
744 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \
745 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
746 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
748 The available information categories are:
750 * ``vxlan``: Set tunnel type as VXLAN.
752 * ``nvgre``: Set tunnel type as NVGRE.
754 * ``ipingre``: Set tunnel type as IP-in-GRE.
756 * ``imac-ivlan``: Set filter type as Inner MAC and VLAN.
758 * ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID.
760 * ``imac-tenid``: Set filter type as Inner MAC and tenant ID.
762 * ``imac``: Set filter type as Inner MAC.
764 * ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID.
766 * ``oip``: Set filter type as Outer IP.
768 * ``iip``: Set filter type as Inner IP.
772 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \
773 192.168.2.2 0 ipingre oip 1 1
775 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP.
780 Remove a tunnel filter on a port::
782 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \
783 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
784 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
789 Add an UDP port for VXLAN packet filter on a port::
791 testpmd> rx_vxlan_port add (udp_port) (port_id)
796 Remove an UDP port for VXLAN packet filter on a port::
798 testpmd> rx_vxlan_port rm (udp_port) (port_id)
803 Set hardware insertion of VLAN IDs in packets sent on a port::
805 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer]
807 For example, set a single VLAN ID (5) insertion on port 0::
811 Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1::
819 Set port based hardware insertion of VLAN ID in packets sent on a port::
821 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off)
826 Disable hardware insertion of a VLAN header in packets sent on a port::
828 testpmd> tx_vlan reset (port_id)
833 Select hardware or software calculation of the checksum when
834 transmitting a packet using the ``csum`` forwarding engine::
836 testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id)
840 * ``ip|udp|tcp|sctp`` always relate to the inner layer.
842 * ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized
843 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are
844 supported). See also the ``csum parse-tunnel`` command.
848 Check the NIC Datasheet for hardware limits.
853 Set RSS queue region span on a port::
855 testpmd> set port (port_id) queue-region region_id (value) \
856 queue_start_index (value) queue_num (value)
858 Set flowtype mapping on a RSS queue region on a port::
860 testpmd> set port (port_id) queue-region region_id (value) flowtype (value)
864 * For the flowtype(pctype) of packet,the specific index for each type has
865 been defined in file i40e_type.h as enum i40e_filter_pctype.
867 Set user priority mapping on a RSS queue region on a port::
869 testpmd> set port (port_id) queue-region UP (value) region_id (value)
871 Flush all queue region related configuration on a port::
873 testpmd> set port (port_id) queue-region flush (on|off)
877 * "on"is just an enable function which server for other configuration,
878 it is for all configuration about queue region from up layer,
879 at first will only keep in DPDK softwarestored in driver,
880 only after "flush on", it commit all configuration to HW.
881 "off" is just clean all configuration about queue region just now,
882 and restore all to DPDK i40e driver default config when start up.
884 Show all queue region related configuration info on a port::
886 testpmd> show port (port_id) queue-region
890 Queue region only support on PF by now, so these command is
891 only for configuration of queue region on PF port.
896 Define how tunneled packets should be handled by the csum forward
899 testpmd> csum parse-tunnel (on|off) (tx_port_id)
901 If enabled, the csum forward engine will try to recognize supported
902 tunnel headers (vxlan, gre, ipip).
904 If disabled, treat tunnel packets as non-tunneled packets (a inner
905 header is handled as a packet payload).
909 The port argument is the TX port like in the ``csum set`` command.
913 Consider a packet in packet like the following::
915 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in
917 * If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
918 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the
919 ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``).
921 * If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
922 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``.
927 Display tx checksum offload configuration::
929 testpmd> csum show (port_id)
934 Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine::
936 testpmd> tso set (segsize) (port_id)
940 Check the NIC datasheet for hardware limits.
945 Display the status of TCP Segmentation Offload::
947 testpmd> tso show (port_id)
952 Enable or disable GRO in ``csum`` forwarding engine::
954 testpmd> set port <port_id> gro on|off
956 If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4
957 packets received from the given port.
959 If disabled, packets received from the given port won't be performed
960 GRO. By default, GRO is disabled for all ports.
964 When enable GRO for a port, TCP/IPv4 packets received from the port
965 will be performed GRO. After GRO, all merged packets have bad
966 checksums, since the GRO library doesn't re-calculate checksums for
967 the merged packets. Therefore, if users want the merged packets to
968 have correct checksums, please select HW IP checksum calculation and
969 HW TCP checksum calculation for the port which the merged packets are
975 Display GRO configuration for a given port::
977 testpmd> show port <port_id> gro
982 Set the cycle to flush the GROed packets from reassembly tables::
984 testpmd> set gro flush <cycles>
986 When enable GRO, the csum forwarding engine performs GRO on received
987 packets, and the GROed packets are stored in reassembly tables. Users
988 can use this command to determine when the GROed packets are flushed
989 from the reassembly tables.
991 The ``cycles`` is measured in GRO operation times. The csum forwarding
992 engine flushes the GROed packets from the tables every ``cycles`` GRO
995 By default, the value of ``cycles`` is 1, which means flush GROed packets
996 from the reassembly tables as soon as one GRO operation finishes. The value
997 of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``.
999 Please note that the large value of ``cycles`` may cause the poor TCP/IP
1000 stack performance. Because the GROed packets are delayed to arrive the
1001 stack, thus causing more duplicated ACKs and TCP retransmissions.
1006 Toggle per-port GSO support in ``csum`` forwarding engine::
1008 testpmd> set port <port_id> gso on|off
1010 If enabled, the csum forwarding engine will perform GSO on supported IPv4
1011 packets, transmitted on the given port.
1013 If disabled, packets transmitted on the given port will not undergo GSO.
1014 By default, GSO is disabled for all ports.
1018 When GSO is enabled on a port, supported IPv4 packets transmitted on that
1019 port undergo GSO. Afterwards, the segmented packets are represented by
1020 multi-segment mbufs; however, the csum forwarding engine doesn't calculation
1021 of checksums for GSO'd segments in SW. As a result, if users want correct
1022 checksums in GSO segments, they should enable HW checksum calculation for
1025 For example, HW checksum calculation for VxLAN GSO'd packets may be enabled
1026 by setting the following options in the csum forwarding engine:
1028 testpmd> csum set outer_ip hw <port_id>
1030 testpmd> csum set ip hw <port_id>
1032 testpmd> csum set tcp hw <port_id>
1037 Set the maximum GSO segment size (measured in bytes), which includes the
1038 packet header and the packet payload for GSO-enabled ports (global)::
1040 testpmd> set gso segsz <length>
1045 Display the status of Generic Segmentation Offload for a given port::
1047 testpmd> show port <port_id> gso
1052 Add an alternative MAC address to a port::
1054 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX)
1059 Remove a MAC address from a port::
1061 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX)
1063 mac_addr add (for VF)
1064 ~~~~~~~~~~~~~~~~~~~~~
1066 Add an alternative MAC address for a VF to a port::
1068 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX)
1073 Set the default MAC address for a port::
1075 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX)
1077 mac_addr set (for VF)
1078 ~~~~~~~~~~~~~~~~~~~~~
1080 Set the MAC address for a VF from the PF::
1082 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX)
1087 Set the forwarding peer address for certain port::
1089 testpmd> set eth-peer (port_id) (perr_addr)
1091 This is equivalent to the ``--eth-peer`` command-line option.
1096 Set the unicast hash filter(s) on/off for a port::
1098 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
1103 Set the promiscuous mode on for a port or for all ports.
1104 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1106 testpmd> set promisc (port_id|all) (on|off)
1111 Set the allmulti mode for a port or for all ports::
1113 testpmd> set allmulti (port_id|all) (on|off)
1115 Same as the ifconfig (8) option. Controls how multicast packets are handled.
1117 set promisc (for VF)
1118 ~~~~~~~~~~~~~~~~~~~~
1120 Set the unicast promiscuous mode for a VF from PF.
1121 It's supported by Intel i40e NICs now.
1122 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1124 testpmd> set vf promisc (port_id) (vf_id) (on|off)
1126 set allmulticast (for VF)
1127 ~~~~~~~~~~~~~~~~~~~~~~~~~
1129 Set the multicast promiscuous mode for a VF from PF.
1130 It's supported by Intel i40e NICs now.
1131 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1133 testpmd> set vf allmulti (port_id) (vf_id) (on|off)
1135 set tx max bandwidth (for VF)
1136 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1138 Set TX max absolute bandwidth (Mbps) for a VF from PF::
1140 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth)
1142 set tc tx min bandwidth (for VF)
1143 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1145 Set all TCs' TX min relative bandwidth (%) for a VF from PF::
1147 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...)
1149 set tc tx max bandwidth (for VF)
1150 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1152 Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF::
1154 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth)
1156 set tc strict link priority mode
1157 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1159 Set some TCs' strict link priority mode on a physical port::
1161 testpmd> set tx strict-link-priority (port_id) (tc_bitmap)
1163 set tc tx min bandwidth
1164 ~~~~~~~~~~~~~~~~~~~~~~~
1166 Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs::
1168 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...)
1173 Set the link flow control parameter on a port::
1175 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1176 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
1177 autoneg (on|off) (port_id)
1181 * ``high_water`` (integer): High threshold value to trigger XOFF.
1183 * ``low_water`` (integer): Low threshold value to trigger XON.
1185 * ``pause_time`` (integer): Pause quota in the Pause frame.
1187 * ``send_xon`` (0/1): Send XON frame.
1189 * ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
1191 * ``autoneg``: Change the auto-negotiation parameter.
1196 Set the priority flow control parameter on a port::
1198 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1199 (pause_time) (priority) (port_id)
1203 * ``high_water`` (integer): High threshold value.
1205 * ``low_water`` (integer): Low threshold value.
1207 * ``pause_time`` (integer): Pause quota in the Pause frame.
1209 * ``priority`` (0-7): VLAN User Priority.
1214 Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
1216 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
1218 For example, to set rx queue 2 on port 0 to mapping 5::
1220 testpmd>set stat_qmap rx 0 2 5
1222 set xstats-hide-zero
1223 ~~~~~~~~~~~~~~~~~~~~
1225 Set the option to hide zero values for xstats display::
1227 testpmd> set xstats-hide-zero on|off
1231 By default, the zero values are displayed for xstats.
1233 set port - rx/tx (for VF)
1234 ~~~~~~~~~~~~~~~~~~~~~~~~~
1236 Set VF receive/transmit from a port::
1238 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
1240 set port - mac address filter (for VF)
1241 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1243 Add/Remove unicast or multicast MAC addr filter for a VF::
1245 testpmd> set port (port_id) vf (vf_id) (mac_addr) \
1246 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
1248 set port - rx mode(for VF)
1249 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1251 Set the VF receive mode of a port::
1253 testpmd> set port (port_id) vf (vf_id) \
1254 rxmode (AUPE|ROPE|BAM|MPE) (on|off)
1256 The available receive modes are:
1258 * ``AUPE``: Accepts untagged VLAN.
1260 * ``ROPE``: Accepts unicast hash.
1262 * ``BAM``: Accepts broadcast packets.
1264 * ``MPE``: Accepts all multicast packets.
1266 set port - tx_rate (for Queue)
1267 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1269 Set TX rate limitation for a queue on a port::
1271 testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1273 set port - tx_rate (for VF)
1274 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
1276 Set TX rate limitation for queues in VF on a port::
1278 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1280 set port - mirror rule
1281 ~~~~~~~~~~~~~~~~~~~~~~
1283 Set pool or vlan type mirror rule for a port::
1285 testpmd> set port (port_id) mirror-rule (rule_id) \
1286 (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1287 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1289 Set link mirror rule for a port::
1291 testpmd> set port (port_id) mirror-rule (rule_id) \
1292 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1294 For example to enable mirror traffic with vlan 0,1 to pool 0::
1296 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1298 reset port - mirror rule
1299 ~~~~~~~~~~~~~~~~~~~~~~~~
1301 Reset a mirror rule for a port::
1303 testpmd> reset port (port_id) mirror-rule (rule_id)
1308 Set the flush on RX streams before forwarding.
1309 The default is flush ``on``.
1310 Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1312 testpmd> set flush_rx off
1317 Set the bypass mode for the lowest port on bypass enabled NIC::
1319 testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1324 Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1326 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1327 mode (normal|bypass|isolate) (port_id)
1331 * ``timeout``: Enable bypass after watchdog timeout.
1333 * ``os_on``: Enable bypass when OS/board is powered on.
1335 * ``os_off``: Enable bypass when OS/board is powered off.
1337 * ``power_on``: Enable bypass when power supply is turned on.
1339 * ``power_off``: Enable bypass when power supply is turned off.
1345 Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1347 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1352 Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1354 testpmd> show bypass config (port_id)
1359 Set link up for a port::
1361 testpmd> set link-up port (port id)
1366 Set link down for a port::
1368 testpmd> set link-down port (port id)
1373 Enable E-tag insertion for a VF on a port::
1375 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1377 Disable E-tag insertion for a VF on a port::
1379 testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1381 Enable/disable E-tag stripping on a port::
1383 testpmd> E-tag set stripping (on|off) port (port_id)
1385 Enable/disable E-tag based forwarding on a port::
1387 testpmd> E-tag set forwarding (on|off) port (port_id)
1389 Add an E-tag forwarding filter on a port::
1391 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1393 Delete an E-tag forwarding filter on a port::
1394 testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1399 Load a dynamic device personalization (DDP) profile and store backup profile::
1401 testpmd> ddp add (port_id) (profile_path[,backup_profile_path])
1406 Delete a dynamic device personalization profile and restore backup profile::
1408 testpmd> ddp del (port_id) (backup_profile_path)
1413 List all items from the ptype mapping table::
1415 testpmd> ptype mapping get (port_id) (valid_only)
1419 * ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0).
1421 Replace a specific or a group of software defined ptype with a new one::
1423 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type)
1427 * ``target``: A specific software ptype or a mask to represent a group of software ptypes.
1429 * ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1).
1431 * ``pkt_type``: The new software ptype to replace the old ones.
1433 Update hardware defined ptype to software defined packet type mapping table::
1435 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype)
1439 * ``hw_ptype``: hardware ptype as the index of the ptype mapping table.
1441 * ``sw_ptype``: software ptype as the value of the ptype mapping table.
1443 Reset ptype mapping table::
1445 testpmd> ptype mapping reset (port_id)
1450 The following sections show functions for configuring ports.
1454 Port configuration changes only become active when forwarding is started/restarted.
1459 Attach a port specified by pci address or virtual device args::
1461 testpmd> port attach (identifier)
1463 To attach a new pci device, the device should be recognized by kernel first.
1464 Then it should be moved under DPDK management.
1465 Finally the port can be attached to testpmd.
1467 For example, to move a pci device using ixgbe under DPDK management:
1469 .. code-block:: console
1471 # Check the status of the available devices.
1472 ./usertools/dpdk-devbind.py --status
1474 Network devices using DPDK-compatible driver
1475 ============================================
1478 Network devices using kernel driver
1479 ===================================
1480 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1483 # Bind the device to igb_uio.
1484 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1487 # Recheck the status of the devices.
1488 ./usertools/dpdk-devbind.py --status
1489 Network devices using DPDK-compatible driver
1490 ============================================
1491 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1493 To attach a port created by virtual device, above steps are not needed.
1495 For example, to attach a port whose pci address is 0000:0a:00.0.
1497 .. code-block:: console
1499 testpmd> port attach 0000:0a:00.0
1500 Attaching a new port...
1501 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1502 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1503 EAL: PCI memory mapped at 0x7f83bfa00000
1504 EAL: PCI memory mapped at 0x7f83bfa80000
1505 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1506 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1507 Port 0 is attached. Now total ports is 1
1510 For example, to attach a port created by pcap PMD.
1512 .. code-block:: console
1514 testpmd> port attach net_pcap0
1515 Attaching a new port...
1516 PMD: Initializing pmd_pcap for net_pcap0
1517 PMD: Creating pcap-backed ethdev on numa socket 0
1518 Port 0 is attached. Now total ports is 1
1521 In this case, identifier is ``net_pcap0``.
1522 This identifier format is the same as ``--vdev`` format of DPDK applications.
1524 For example, to re-attach a bonded port which has been previously detached,
1525 the mode and slave parameters must be given.
1527 .. code-block:: console
1529 testpmd> port attach net_bond_0,mode=0,slave=1
1530 Attaching a new port...
1531 EAL: Initializing pmd_bond for net_bond_0
1532 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1533 Port 0 is attached. Now total ports is 1
1540 Detach a specific port::
1542 testpmd> port detach (port_id)
1544 Before detaching a port, the port should be stopped and closed.
1546 For example, to detach a pci device port 0.
1548 .. code-block:: console
1550 testpmd> port stop 0
1553 testpmd> port close 0
1557 testpmd> port detach 0
1559 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1560 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1561 EAL: PCI memory unmapped at 0x7f83bfa00000
1562 EAL: PCI memory unmapped at 0x7f83bfa80000
1566 For example, to detach a virtual device port 0.
1568 .. code-block:: console
1570 testpmd> port stop 0
1573 testpmd> port close 0
1577 testpmd> port detach 0
1579 PMD: Closing pcap ethdev on numa socket 0
1580 Port 'net_pcap0' is detached. Now total ports is 0
1583 To remove a pci device completely from the system, first detach the port from testpmd.
1584 Then the device should be moved under kernel management.
1585 Finally the device can be removed using kernel pci hotplug functionality.
1587 For example, to move a pci device under kernel management:
1589 .. code-block:: console
1591 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1593 ./usertools/dpdk-devbind.py --status
1595 Network devices using DPDK-compatible driver
1596 ============================================
1599 Network devices using kernel driver
1600 ===================================
1601 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1603 To remove a port created by a virtual device, above steps are not needed.
1608 Start all ports or a specific port::
1610 testpmd> port start (port_id|all)
1615 Stop all ports or a specific port::
1617 testpmd> port stop (port_id|all)
1622 Close all ports or a specific port::
1624 testpmd> port close (port_id|all)
1626 port config - queue ring size
1627 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1629 Configure a rx/tx queue ring size::
1631 testpmd> port (port_id) (rxq|txq) (queue_id) ring_size (value)
1633 Only take effect after command that (re-)start the port or command that setup specific queue.
1635 port start/stop queue
1636 ~~~~~~~~~~~~~~~~~~~~~
1638 Start/stop a rx/tx queue on a specific port::
1640 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1643 ~~~~~~~~~~~~~~~~~~~~~
1645 Setup a rx/tx queue on a specific port::
1647 testpmd> port (port_id) (rxq|txq) (queue_id) setup
1649 Only take effect when port is started.
1654 Set the speed and duplex mode for all ports or a specific port::
1656 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1657 duplex (half|full|auto)
1659 port config - queues/descriptors
1660 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1662 Set number of queues/descriptors for rxq, txq, rxd and txd::
1664 testpmd> port config all (rxq|txq|rxd|txd) (value)
1666 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1668 port config - max-pkt-len
1669 ~~~~~~~~~~~~~~~~~~~~~~~~~
1671 Set the maximum packet length::
1673 testpmd> port config all max-pkt-len (value)
1675 This is equivalent to the ``--max-pkt-len`` command-line option.
1677 port config - CRC Strip
1678 ~~~~~~~~~~~~~~~~~~~~~~~
1680 Set hardware CRC stripping on or off for all ports::
1682 testpmd> port config all crc-strip (on|off)
1684 CRC stripping is on by default.
1686 The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option.
1688 port config - scatter
1689 ~~~~~~~~~~~~~~~~~~~~~~~
1691 Set RX scatter mode on or off for all ports::
1693 testpmd> port config all scatter (on|off)
1695 RX scatter mode is off by default.
1697 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1699 port config - RX Checksum
1700 ~~~~~~~~~~~~~~~~~~~~~~~~~
1702 Set hardware RX checksum offload to on or off for all ports::
1704 testpmd> port config all rx-cksum (on|off)
1706 Checksum offload is off by default.
1708 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1713 Set hardware VLAN on or off for all ports::
1715 testpmd> port config all hw-vlan (on|off)
1717 Hardware VLAN is off by default.
1719 The ``on`` option is equivalent to the ``--enable-hw-vlan`` command-line option.
1721 port config - VLAN filter
1722 ~~~~~~~~~~~~~~~~~~~~~~~~~
1724 Set hardware VLAN filter on or off for all ports::
1726 testpmd> port config all hw-vlan-filter (on|off)
1728 Hardware VLAN filter is off by default.
1730 The ``on`` option is equivalent to the ``--enable-hw-vlan-filter`` command-line option.
1732 port config - VLAN strip
1733 ~~~~~~~~~~~~~~~~~~~~~~~~
1735 Set hardware VLAN strip on or off for all ports::
1737 testpmd> port config all hw-vlan-strip (on|off)
1739 Hardware VLAN strip is off by default.
1741 The ``on`` option is equivalent to the ``--enable-hw-vlan-strip`` command-line option.
1743 port config - VLAN extend
1744 ~~~~~~~~~~~~~~~~~~~~~~~~~
1746 Set hardware VLAN extend on or off for all ports::
1748 testpmd> port config all hw-vlan-extend (on|off)
1750 Hardware VLAN extend is off by default.
1752 The ``on`` option is equivalent to the ``--enable-hw-vlan-extend`` command-line option.
1754 port config - Drop Packets
1755 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1757 Set packet drop for packets with no descriptors on or off for all ports::
1759 testpmd> port config all drop-en (on|off)
1761 Packet dropping for packets with no descriptors is off by default.
1763 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1768 Set the RSS (Receive Side Scaling) mode on or off::
1770 testpmd> port config all rss (all|default|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1772 RSS is on by default.
1774 The ``all`` option is equivalent to ip|tcp|udp|sctp|ether.
1775 The ``default`` option enables all supported RSS types reported by device info.
1776 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1778 port config - RSS Reta
1779 ~~~~~~~~~~~~~~~~~~~~~~
1781 Set the RSS (Receive Side Scaling) redirection table::
1783 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1788 Set the DCB mode for an individual port::
1790 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1792 The traffic class should be 4 or 8.
1797 Set the number of packets per burst::
1799 testpmd> port config all burst (value)
1801 This is equivalent to the ``--burst`` command-line option.
1803 port config - Threshold
1804 ~~~~~~~~~~~~~~~~~~~~~~~
1806 Set thresholds for TX/RX queues::
1808 testpmd> port config all (threshold) (value)
1810 Where the threshold type can be:
1812 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1814 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1816 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1818 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1820 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1822 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1824 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1826 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1828 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1830 These threshold options are also available from the command-line.
1835 Set the value of ether-type for E-tag::
1837 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1839 Enable/disable the E-tag support::
1841 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1843 port config pctype mapping
1844 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1846 Reset pctype mapping table::
1848 testpmd> port config (port_id) pctype mapping reset
1850 Update hardware defined pctype to software defined flow type mapping table::
1852 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id)
1856 * ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table.
1858 * ``flow_type_id``: software flow type id as the index of the pctype mapping table.
1860 port config input set
1861 ~~~~~~~~~~~~~~~~~~~~~
1863 Config RSS/FDIR/FDIR flexible payload input set for some pctype::
1864 testpmd> port config (port_id) pctype (pctype_id) \
1865 (hash_inset|fdir_inset|fdir_flx_inset) \
1866 (get|set|clear) field (field_idx)
1868 Clear RSS/FDIR/FDIR flexible payload input set for some pctype::
1869 testpmd> port config (port_id) pctype (pctype_id) \
1870 (hash_inset|fdir_inset|fdir_flx_inset) clear all
1874 * ``pctype_id``: hardware packet classification types.
1875 * ``field_idx``: hardware field index.
1877 port config udp_tunnel_port
1878 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
1880 Add/remove UDP tunnel port for VXLAN/GENEVE tunneling protocols::
1881 testpmd> port config (port_id) udp_tunnel_port add|rm vxlan|geneve (udp_port)
1883 Link Bonding Functions
1884 ----------------------
1886 The Link Bonding functions make it possible to dynamically create and
1887 manage link bonding devices from within testpmd interactive prompt.
1889 create bonded device
1890 ~~~~~~~~~~~~~~~~~~~~
1892 Create a new bonding device::
1894 testpmd> create bonded device (mode) (socket)
1896 For example, to create a bonded device in mode 1 on socket 0::
1898 testpmd> create bonded 1 0
1899 created new bonded device (port X)
1904 Adds Ethernet device to a Link Bonding device::
1906 testpmd> add bonding slave (slave id) (port id)
1908 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
1910 testpmd> add bonding slave 6 10
1913 remove bonding slave
1914 ~~~~~~~~~~~~~~~~~~~~
1916 Removes an Ethernet slave device from a Link Bonding device::
1918 testpmd> remove bonding slave (slave id) (port id)
1920 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
1922 testpmd> remove bonding slave 6 10
1927 Set the Link Bonding mode of a Link Bonding device::
1929 testpmd> set bonding mode (value) (port id)
1931 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
1933 testpmd> set bonding mode 3 10
1938 Set an Ethernet slave device as the primary device on a Link Bonding device::
1940 testpmd> set bonding primary (slave id) (port id)
1942 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
1944 testpmd> set bonding primary 6 10
1949 Set the MAC address of a Link Bonding device::
1951 testpmd> set bonding mac (port id) (mac)
1953 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
1955 testpmd> set bonding mac 10 00:00:00:00:00:01
1957 set bonding xmit_balance_policy
1958 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1960 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
1962 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
1964 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
1966 testpmd> set bonding xmit_balance_policy 10 l34
1969 set bonding mon_period
1970 ~~~~~~~~~~~~~~~~~~~~~~
1972 Set the link status monitoring polling period in milliseconds for a bonding device.
1974 This adds support for PMD slave devices which do not support link status interrupts.
1975 When the mon_period is set to a value greater than 0 then all PMD's which do not support
1976 link status ISR will be queried every polling interval to check if their link status has changed::
1978 testpmd> set bonding mon_period (port_id) (value)
1980 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
1982 testpmd> set bonding mon_period 5 150
1985 set bonding lacp dedicated_queue
1986 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1988 Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic
1989 when in mode 4 (link-aggregration-802.3ad)::
1991 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable)
1994 set bonding agg_mode
1995 ~~~~~~~~~~~~~~~~~~~~
1997 Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad)::
1999 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable)
2005 Show the current configuration of a Link Bonding device::
2007 testpmd> show bonding config (port id)
2010 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
2011 in balance mode with a transmission policy of layer 2+3::
2013 testpmd> show bonding config 9
2015 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
2017 Active Slaves (3): [1 3 4]
2024 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
2025 This is mainly useful for debugging purposes.
2026 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
2027 and fields that can be accessed.
2032 Display the value of a port register::
2034 testpmd> read reg (port_id) (address)
2036 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
2038 testpmd> read reg 0 0xEE00
2039 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
2044 Display a port register bit field::
2046 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
2048 For example, reading the lowest two bits from the register in the example above::
2050 testpmd> read regfield 0 0xEE00 0 1
2051 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
2056 Display a single port register bit::
2058 testpmd> read regbit (port_id) (address) (bit_x)
2060 For example, reading the lowest bit from the register in the example above::
2062 testpmd> read regbit 0 0xEE00 0
2063 port 0 PCI register at offset 0xEE00: bit 0=1
2068 Set the value of a port register::
2070 testpmd> write reg (port_id) (address) (value)
2072 For example, to clear a register::
2074 testpmd> write reg 0 0xEE00 0x0
2075 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
2080 Set bit field of a port register::
2082 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
2084 For example, writing to the register cleared in the example above::
2086 testpmd> write regfield 0 0xEE00 0 1 2
2087 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
2092 Set single bit value of a port register::
2094 testpmd> write regbit (port_id) (address) (bit_x) (value)
2096 For example, to set the high bit in the register from the example above::
2098 testpmd> write regbit 0 0xEE00 31 1
2099 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
2101 Traffic Metering and Policing
2102 -----------------------------
2104 The following section shows functions for configuring traffic metering and
2105 policing on the ethernet device through the use of generic ethdev API.
2107 show port traffic management capability
2108 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2110 Show traffic metering and policing capability of the port::
2112 testpmd> show port meter cap (port_id)
2114 add port meter profile (srTCM rfc2967)
2115 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2117 Add meter profile (srTCM rfc2697) to the ethernet device::
2119 testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \
2124 * ``profile_id``: ID for the meter profile.
2125 * ``cir``: Committed Information Rate (CIR) (bytes/second).
2126 * ``cbs``: Committed Burst Size (CBS) (bytes).
2127 * ``ebs``: Excess Burst Size (EBS) (bytes).
2129 add port meter profile (trTCM rfc2968)
2130 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2132 Add meter profile (srTCM rfc2698) to the ethernet device::
2134 testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \
2135 (cir) (pir) (cbs) (pbs)
2139 * ``profile_id``: ID for the meter profile.
2140 * ``cir``: Committed information rate (bytes/second).
2141 * ``pir``: Peak information rate (bytes/second).
2142 * ``cbs``: Committed burst size (bytes).
2143 * ``pbs``: Peak burst size (bytes).
2145 add port meter profile (trTCM rfc4115)
2146 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2148 Add meter profile (trTCM rfc4115) to the ethernet device::
2150 testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \
2151 (cir) (eir) (cbs) (ebs)
2155 * ``profile_id``: ID for the meter profile.
2156 * ``cir``: Committed information rate (bytes/second).
2157 * ``eir``: Excess information rate (bytes/second).
2158 * ``cbs``: Committed burst size (bytes).
2159 * ``ebs``: Excess burst size (bytes).
2161 delete port meter profile
2162 ~~~~~~~~~~~~~~~~~~~~~~~~~
2164 Delete meter profile from the ethernet device::
2166 testpmd> del port meter profile (port_id) (profile_id)
2171 Create new meter object for the ethernet device::
2173 testpmd> create port meter (port_id) (mtr_id) (profile_id) \
2174 (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \
2175 (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\
2180 * ``mtr_id``: meter object ID.
2181 * ``profile_id``: ID for the meter profile.
2182 * ``meter_enable``: When this parameter has a non-zero value, the meter object
2183 gets enabled at the time of creation, otherwise remains disabled.
2184 * ``g_action``: Policer action for the packet with green color.
2185 * ``y_action``: Policer action for the packet with yellow color.
2186 * ``r_action``: Policer action for the packet with red color.
2187 * ``stats_mask``: Mask of statistics counter types to be enabled for the
2189 * ``shared``: When this parameter has a non-zero value, the meter object is
2190 shared by multiple flows. Otherwise, meter object is used by single flow.
2191 * ``use_pre_meter_color``: When this parameter has a non-zero value, the
2192 input color for the current meter object is determined by the latest meter
2193 object in the same flow. Otherwise, the current meter object uses the
2194 *dscp_table* to determine the input color.
2195 * ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input
2196 color, 0 <= x <= 63.
2201 Enable meter for the ethernet device::
2203 testpmd> enable port meter (port_id) (mtr_id)
2208 Disable meter for the ethernet device::
2210 testpmd> disable port meter (port_id) (mtr_id)
2215 Delete meter for the ethernet device::
2217 testpmd> del port meter (port_id) (mtr_id)
2219 Set port meter profile
2220 ~~~~~~~~~~~~~~~~~~~~~~
2222 Set meter profile for the ethernet device::
2224 testpmd> set port meter profile (port_id) (mtr_id) (profile_id)
2226 set port meter dscp table
2227 ~~~~~~~~~~~~~~~~~~~~~~~~~
2229 Set meter dscp table for the ethernet device::
2231 testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \
2232 (dscp_tbl_entry1)...(dscp_tbl_entry63)]
2234 set port meter policer action
2235 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2237 Set meter policer action for the ethernet device::
2239 testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \
2240 (action0) [(action1) (action1)]
2244 * ``action_mask``: Bit mask indicating which policer actions need to be
2245 updated. One or more policer actions can be updated in a single function
2246 invocation. To update the policer action associated with color C, bit
2247 (1 << C) needs to be set in *action_mask* and element at position C
2248 in the *actions* array needs to be valid.
2249 * ``actionx``: Policer action for the color x,
2250 RTE_MTR_GREEN <= x < RTE_MTR_COLORS
2252 set port meter stats mask
2253 ~~~~~~~~~~~~~~~~~~~~~~~~~
2255 Set meter stats mask for the ethernet device::
2257 testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask)
2261 * ``stats_mask``: Bit mask indicating statistics counter types to be enabled.
2263 show port meter stats
2264 ~~~~~~~~~~~~~~~~~~~~~
2266 Show meter stats of the ethernet device::
2268 testpmd> show port meter stats (port_id) (mtr_id) (clear)
2272 * ``clear``: Flag that indicates whether the statistics counters should
2273 be cleared (i.e. set to zero) immediately after they have been read or not.
2278 The following section shows functions for configuring traffic management on
2279 on the ethernet device through the use of generic TM API.
2281 show port traffic management capability
2282 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2284 Show traffic management capability of the port::
2286 testpmd> show port tm cap (port_id)
2288 show port traffic management capability (hierarchy level)
2289 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2291 Show traffic management hierarchy level capability of the port::
2293 testpmd> show port tm level cap (port_id) (level_id)
2295 show port traffic management capability (hierarchy node level)
2296 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2298 Show the traffic management hierarchy node capability of the port::
2300 testpmd> show port tm node cap (port_id) (node_id)
2302 show port traffic management hierarchy node type
2303 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2305 Show the port traffic management hierarchy node type::
2307 testpmd> show port tm node type (port_id) (node_id)
2309 show port traffic management hierarchy node stats
2310 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2312 Show the port traffic management hierarchy node statistics::
2314 testpmd> show port tm node stats (port_id) (node_id) (clear)
2318 * ``clear``: When this parameter has a non-zero value, the statistics counters
2319 are cleared (i.e. set to zero) immediately after they have been read,
2320 otherwise the statistics counters are left untouched.
2322 Add port traffic management private shaper profile
2323 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2325 Add the port traffic management private shaper profile::
2327 testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \
2328 (tb_rate) (tb_size) (packet_length_adjust)
2332 * ``shaper_profile id``: Shaper profile ID for the new profile.
2333 * ``tb_rate``: Token bucket rate (bytes per second).
2334 * ``tb_size``: Token bucket size (bytes).
2335 * ``packet_length_adjust``: The value (bytes) to be added to the length of
2336 each packet for the purpose of shaping. This parameter value can be used to
2337 correct the packet length with the framing overhead bytes that are consumed
2340 Delete port traffic management private shaper profile
2341 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2343 Delete the port traffic management private shaper::
2345 testpmd> del port tm node shaper profile (port_id) (shaper_profile_id)
2349 * ``shaper_profile id``: Shaper profile ID that needs to be deleted.
2351 Add port traffic management shared shaper
2352 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2354 Create the port traffic management shared shaper::
2356 testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \
2361 * ``shared_shaper_id``: Shared shaper ID to be created.
2362 * ``shaper_profile id``: Shaper profile ID for shared shaper.
2364 Set port traffic management shared shaper
2365 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2367 Update the port traffic management shared shaper::
2369 testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \
2374 * ``shared_shaper_id``: Shared shaper ID to be update.
2375 * ``shaper_profile id``: Shaper profile ID for shared shaper.
2377 Delete port traffic management shared shaper
2378 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2380 Delete the port traffic management shared shaper::
2382 testpmd> del port tm node shared shaper (port_id) (shared_shaper_id)
2386 * ``shared_shaper_id``: Shared shaper ID to be deleted.
2388 Set port traffic management hiearchy node private shaper
2389 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2391 set the port traffic management hierarchy node private shaper::
2393 testpmd> set port tm node shaper profile (port_id) (node_id) \
2398 * ``shaper_profile id``: Private shaper profile ID to be enabled on the
2401 Add port traffic management WRED profile
2402 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2404 Create a new WRED profile::
2406 testpmd> add port tm node wred profile (port_id) (wred_profile_id) \
2407 (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \
2408 (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \
2409 (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r)
2413 * ``wred_profile id``: Identifier for the newly create WRED profile
2414 * ``color_g``: Packet color (green)
2415 * ``min_th_g``: Minimum queue threshold for packet with green color
2416 * ``max_th_g``: Minimum queue threshold for packet with green color
2417 * ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp)
2418 * ``wq_log2_g``: Negated log2 of queue weight (wq)
2419 * ``color_y``: Packet color (yellow)
2420 * ``min_th_y``: Minimum queue threshold for packet with yellow color
2421 * ``max_th_y``: Minimum queue threshold for packet with yellow color
2422 * ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp)
2423 * ``wq_log2_y``: Negated log2 of queue weight (wq)
2424 * ``color_r``: Packet color (red)
2425 * ``min_th_r``: Minimum queue threshold for packet with yellow color
2426 * ``max_th_r``: Minimum queue threshold for packet with yellow color
2427 * ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp)
2428 * ``wq_log2_r``: Negated log2 of queue weight (wq)
2430 Delete port traffic management WRED profile
2431 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2433 Delete the WRED profile::
2435 testpmd> del port tm node wred profile (port_id) (wred_profile_id)
2437 Add port traffic management hierarchy nonleaf node
2438 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2440 Add nonleaf node to port traffic management hiearchy::
2442 testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \
2443 (priority) (weight) (level_id) (shaper_profile_id) \
2444 (n_sp_priorities) (stats_mask) (n_shared_shapers) \
2445 [(shared_shaper_0) (shared_shaper_1) ...] \
2449 * ``parent_node_id``: Node ID of the parent.
2450 * ``priority``: Node priority (highest node priority is zero). This is used by
2451 the SP algorithm running on the parent node for scheduling this node.
2452 * ``weight``: Node weight (lowest weight is one). The node weight is relative
2453 to the weight sum of all siblings that have the same priority. It is used by
2454 the WFQ algorithm running on the parent node for scheduling this node.
2455 * ``level_id``: Hiearchy level of the node.
2456 * ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2458 * ``n_sp_priorities``: Number of strict priorities.
2459 * ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2460 * ``n_shared_shapers``: Number of shared shapers.
2461 * ``shared_shaper_id``: Shared shaper id.
2463 Add port traffic management hierarchy leaf node
2464 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2466 Add leaf node to port traffic management hiearchy::
2468 testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \
2469 (priority) (weight) (level_id) (shaper_profile_id) \
2470 (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \
2471 [(shared_shaper_id) (shared_shaper_id) ...] \
2475 * ``parent_node_id``: Node ID of the parent.
2476 * ``priority``: Node priority (highest node priority is zero). This is used by
2477 the SP algorithm running on the parent node for scheduling this node.
2478 * ``weight``: Node weight (lowest weight is one). The node weight is relative
2479 to the weight sum of all siblings that have the same priority. It is used by
2480 the WFQ algorithm running on the parent node for scheduling this node.
2481 * ``level_id``: Hiearchy level of the node.
2482 * ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2484 * ``cman_mode``: Congestion management mode to be enabled for this node.
2485 * ``wred_profile_id``: WRED profile id to be enabled for this node.
2486 * ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2487 * ``n_shared_shapers``: Number of shared shapers.
2488 * ``shared_shaper_id``: Shared shaper id.
2490 Delete port traffic management hierarchy node
2491 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2493 Delete node from port traffic management hiearchy::
2495 testpmd> del port tm node (port_id) (node_id)
2497 Update port traffic management hierarchy parent node
2498 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2500 Update port traffic management hierarchy parent node::
2502 testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \
2505 This function can only be called after the hierarchy commit invocation. Its
2506 success depends on the port support for this operation, as advertised through
2507 the port capability set. This function is valid for all nodes of the traffic
2508 management hierarchy except root node.
2510 Suspend port traffic management hierarchy node
2511 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2513 testpmd> suspend port tm node (port_id) (node_id)
2515 Commit port traffic management hierarchy
2516 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2518 Commit the traffic management hierarchy on the port::
2520 testpmd> port tm hierarchy commit (port_id) (clean_on_fail)
2524 * ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function
2525 call failure. On the other hand, hierarchy is preserved when this parameter
2528 Set port traffic management default hierarchy (tm forwarding mode)
2529 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2531 set the traffic management default hierarchy on the port::
2533 testpmd> set port tm hierarchy default (port_id)
2538 This section details the available filter functions that are available.
2540 Note these functions interface the deprecated legacy filtering framework,
2541 superseded by *rte_flow*. See `Flow rules management`_.
2544 ~~~~~~~~~~~~~~~~~~~~
2546 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
2548 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
2549 ethertype (ether_type) (drop|fwd) queue (queue_id)
2551 The available information parameters are:
2553 * ``port_id``: The port which the Ethertype filter assigned on.
2555 * ``mac_addr``: Compare destination mac address.
2557 * ``mac_ignr``: Ignore destination mac address match.
2559 * ``mac_address``: Destination mac address to match.
2561 * ``ether_type``: The EtherType value want to match,
2562 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
2564 * ``queue_id``: The receive queue associated with this EtherType filter.
2565 It is meaningless when deleting or dropping.
2567 Example, to add/remove an ethertype filter rule::
2569 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
2570 ethertype 0x0806 fwd queue 3
2572 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
2573 ethertype 0x0806 fwd queue 3
2578 Add or delete a 2-tuple filter,
2579 which identifies packets by specific protocol and destination TCP/UDP port
2580 and forwards packets into one of the receive queues::
2582 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
2583 protocol (protocol_value) mask (mask_value) \
2584 tcp_flags (tcp_flags_value) priority (prio_value) \
2587 The available information parameters are:
2589 * ``port_id``: The port which the 2-tuple filter assigned on.
2591 * ``dst_port_value``: Destination port in L4.
2593 * ``protocol_value``: IP L4 protocol.
2595 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
2597 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
2599 * ``prio_value``: Priority of this filter.
2601 * ``queue_id``: The receive queue associated with this 2-tuple filter.
2603 Example, to add/remove an 2tuple filter rule::
2605 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
2606 tcp_flags 0x02 priority 3 queue 3
2608 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
2609 tcp_flags 0x02 priority 3 queue 3
2614 Add or delete a 5-tuple filter,
2615 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
2616 and routes packets into one of the receive queues::
2618 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
2619 (src_address) dst_port (dst_port_value) \
2620 src_port (src_port_value) protocol (protocol_value) \
2621 mask (mask_value) tcp_flags (tcp_flags_value) \
2622 priority (prio_value) queue (queue_id)
2624 The available information parameters are:
2626 * ``port_id``: The port which the 5-tuple filter assigned on.
2628 * ``dst_address``: Destination IP address.
2630 * ``src_address``: Source IP address.
2632 * ``dst_port_value``: TCP/UDP destination port.
2634 * ``src_port_value``: TCP/UDP source port.
2636 * ``protocol_value``: L4 protocol.
2638 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
2640 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
2642 * ``prio_value``: The priority of this filter.
2644 * ``queue_id``: The receive queue associated with this 5-tuple filter.
2646 Example, to add/remove an 5tuple filter rule::
2648 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2649 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2650 flags 0x0 priority 3 queue 3
2652 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2653 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2654 flags 0x0 priority 3 queue 3
2659 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
2661 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
2663 The available information parameters are:
2665 * ``port_id``: The port which the SYN filter assigned on.
2667 * ``high``: This SYN filter has higher priority than other filters.
2669 * ``low``: This SYN filter has lower priority than other filters.
2671 * ``queue_id``: The receive queue associated with this SYN filter
2675 testpmd> syn_filter 0 add priority high queue 3
2680 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
2681 and routed into one of the receive queues::
2683 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
2684 mask (mask_value) priority (prio_value) queue (queue_id)
2686 The available information parameters are:
2688 * ``port_id``: The port which the Flex filter is assigned on.
2690 * ``len_value``: Filter length in bytes, no greater than 128.
2692 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
2694 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
2696 * ``prio_value``: The priority of this filter.
2698 * ``queue_id``: The receive queue associated with this Flex filter.
2702 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
2703 mask 000C priority 3 queue 3
2705 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
2706 mask 000C priority 3 queue 3
2709 .. _testpmd_flow_director:
2711 flow_director_filter
2712 ~~~~~~~~~~~~~~~~~~~~
2714 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
2716 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
2717 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
2719 * Perfect match filters.
2720 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2721 The masked fields are for IP flow.
2723 * Signature filters.
2724 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
2726 * Perfect-mac-vlan match filters.
2727 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2728 The masked fields are for MAC VLAN flow.
2730 * Perfect-tunnel match filters.
2731 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2732 The masked fields are for tunnel flow.
2734 * Perfect-raw-flow-type match filters.
2735 The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet.
2736 The masked fields are specified by input sets.
2738 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
2739 per flow type and the flexible payload.
2741 The Flow Director can also mask out parts of all of these fields so that filters
2742 are only applied to certain fields or parts of the fields.
2744 Note that for raw flow type mode the source and destination fields in the
2745 raw packet buffer need to be presented in a reversed order with respect
2746 to the expected received packets.
2747 For example: IP source and destination addresses or TCP/UDP/SCTP
2748 source and destination ports
2750 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
2752 # Commands to add flow director filters of different flow types::
2754 flow_director_filter (port_id) mode IP (add|del|update) \
2755 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
2756 src (src_ip_address) dst (dst_ip_address) \
2757 tos (tos_value) proto (proto_value) ttl (ttl_value) \
2758 vlan (vlan_value) flexbytes (flexbytes_value) \
2759 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
2762 flow_director_filter (port_id) mode IP (add|del|update) \
2763 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
2764 src (src_ip_address) (src_port) \
2765 dst (dst_ip_address) (dst_port) \
2766 tos (tos_value) ttl (ttl_value) \
2767 vlan (vlan_value) flexbytes (flexbytes_value) \
2768 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
2771 flow_director_filter (port_id) mode IP (add|del|update) \
2772 flow (ipv4-sctp|ipv6-sctp) \
2773 src (src_ip_address) (src_port) \
2774 dst (dst_ip_address) (dst_port) \
2775 tos (tos_value) ttl (ttl_value) \
2776 tag (verification_tag) vlan (vlan_value) \
2777 flexbytes (flexbytes_value) (drop|fwd) \
2778 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
2780 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
2781 ether (ethertype) flexbytes (flexbytes_value) \
2782 (drop|fwd) pf|vf(vf_id) queue (queue_id)
2785 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
2786 mac (mac_address) vlan (vlan_value) \
2787 flexbytes (flexbytes_value) (drop|fwd) \
2788 queue (queue_id) fd_id (fd_id_value)
2790 flow_director_filter (port_id) mode Tunnel (add|del|update) \
2791 mac (mac_address) vlan (vlan_value) \
2792 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
2793 flexbytes (flexbytes_value) (drop|fwd) \
2794 queue (queue_id) fd_id (fd_id_value)
2796 flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \
2797 (drop|fwd) queue (queue_id) fd_id (fd_id_value) \
2798 packet (packet file name)
2800 For example, to add an ipv4-udp flow type filter::
2802 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
2803 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
2804 fwd pf queue 1 fd_id 1
2806 For example, add an ipv4-other flow type filter::
2808 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
2809 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
2810 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
2815 Flush all flow director filters on a device::
2817 testpmd> flush_flow_director (port_id)
2819 Example, to flush all flow director filter on port 0::
2821 testpmd> flush_flow_director 0
2826 Set flow director's input masks::
2828 flow_director_mask (port_id) mode IP vlan (vlan_value) \
2829 src_mask (ipv4_src) (ipv6_src) (src_port) \
2830 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2832 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2834 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2835 mac (mac_value) tunnel-type (tunnel_type_value) \
2836 tunnel-id (tunnel_id_value)
2838 Example, to set flow director mask on port 0::
2840 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2841 src_mask 255.255.255.255 \
2842 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2843 dst_mask 255.255.255.255 \
2844 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2846 flow_director_flex_mask
2847 ~~~~~~~~~~~~~~~~~~~~~~~
2849 set masks of flow director's flexible payload based on certain flow type::
2851 testpmd> flow_director_flex_mask (port_id) \
2852 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2853 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2854 l2_payload|all) (mask)
2856 Example, to set flow director's flex mask for all flow type on port 0::
2858 testpmd> flow_director_flex_mask 0 flow all \
2859 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2862 flow_director_flex_payload
2863 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2865 Configure flexible payload selection::
2867 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2869 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2871 testpmd> flow_director_flex_payload 0 l4 \
2872 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2874 get_sym_hash_ena_per_port
2875 ~~~~~~~~~~~~~~~~~~~~~~~~~
2877 Get symmetric hash enable configuration per port::
2879 get_sym_hash_ena_per_port (port_id)
2881 For example, to get symmetric hash enable configuration of port 1::
2883 testpmd> get_sym_hash_ena_per_port 1
2885 set_sym_hash_ena_per_port
2886 ~~~~~~~~~~~~~~~~~~~~~~~~~
2888 Set symmetric hash enable configuration per port to enable or disable::
2890 set_sym_hash_ena_per_port (port_id) (enable|disable)
2892 For example, to set symmetric hash enable configuration of port 1 to enable::
2894 testpmd> set_sym_hash_ena_per_port 1 enable
2896 get_hash_global_config
2897 ~~~~~~~~~~~~~~~~~~~~~~
2899 Get the global configurations of hash filters::
2901 get_hash_global_config (port_id)
2903 For example, to get the global configurations of hash filters of port 1::
2905 testpmd> get_hash_global_config 1
2907 set_hash_global_config
2908 ~~~~~~~~~~~~~~~~~~~~~~
2910 Set the global configurations of hash filters::
2912 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2913 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2914 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \
2917 For example, to enable simple_xor for flow type of ipv6 on port 2::
2919 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2924 Set the input set for hash::
2926 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2927 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2928 l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
2929 ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2930 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2931 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2932 fld-8th|none) (select|add)
2934 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2936 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2941 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2942 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2944 Set the input set for flow director::
2946 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2947 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2948 l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
2949 ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2950 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2951 sctp-dst-port|sctp-veri-tag|none) (select|add)
2953 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2955 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2960 Set different GRE key length for input set::
2962 global_config (port_id) gre-key-len (number in bytes)
2964 For example to set GRE key length for input set to 4 bytes on port 0::
2966 testpmd> global_config 0 gre-key-len 4
2969 .. _testpmd_rte_flow:
2971 Flow rules management
2972 ---------------------
2974 Control of the generic flow API (*rte_flow*) is fully exposed through the
2975 ``flow`` command (validation, creation, destruction, queries and operation
2978 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2979 features simultaneously may cause undefined side-effects and is therefore
2985 Because the ``flow`` command uses dynamic tokens to handle the large number
2986 of possible flow rules combinations, its behavior differs slightly from
2987 other commands, in particular:
2989 - Pressing *?* or the *<tab>* key displays contextual help for the current
2990 token, not that of the entire command.
2992 - Optional and repeated parameters are supported (provided they are listed
2993 in the contextual help).
2995 The first parameter stands for the operation mode. Possible operations and
2996 their general syntax are described below. They are covered in detail in the
2999 - Check whether a flow rule can be created::
3001 flow validate {port_id}
3002 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3003 pattern {item} [/ {item} [...]] / end
3004 actions {action} [/ {action} [...]] / end
3006 - Create a flow rule::
3008 flow create {port_id}
3009 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3010 pattern {item} [/ {item} [...]] / end
3011 actions {action} [/ {action} [...]] / end
3013 - Destroy specific flow rules::
3015 flow destroy {port_id} rule {rule_id} [...]
3017 - Destroy all flow rules::
3019 flow flush {port_id}
3021 - Query an existing flow rule::
3023 flow query {port_id} {rule_id} {action}
3025 - List existing flow rules sorted by priority, filtered by group
3028 flow list {port_id} [group {group_id}] [...]
3030 - Restrict ingress traffic to the defined flow rules::
3032 flow isolate {port_id} {boolean}
3034 Validating flow rules
3035 ~~~~~~~~~~~~~~~~~~~~~
3037 ``flow validate`` reports whether a flow rule would be accepted by the
3038 underlying device in its current state but stops short of creating it. It is
3039 bound to ``rte_flow_validate()``::
3041 flow validate {port_id}
3042 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3043 pattern {item} [/ {item} [...]] / end
3044 actions {action} [/ {action} [...]] / end
3046 If successful, it will show::
3050 Otherwise it will show an error message of the form::
3052 Caught error type [...] ([...]): [...]
3054 This command uses the same parameters as ``flow create``, their format is
3055 described in `Creating flow rules`_.
3057 Check whether redirecting any Ethernet packet received on port 0 to RX queue
3058 index 6 is supported::
3060 testpmd> flow validate 0 ingress pattern eth / end
3061 actions queue index 6 / end
3065 Port 0 does not support TCPv6 rules::
3067 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
3069 Caught error type 9 (specific pattern item): Invalid argument
3075 ``flow create`` validates and creates the specified flow rule. It is bound
3076 to ``rte_flow_create()``::
3078 flow create {port_id}
3079 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3080 pattern {item} [/ {item} [...]] / end
3081 actions {action} [/ {action} [...]] / end
3083 If successful, it will return a flow rule ID usable with other commands::
3085 Flow rule #[...] created
3087 Otherwise it will show an error message of the form::
3089 Caught error type [...] ([...]): [...]
3091 Parameters describe in the following order:
3093 - Attributes (*group*, *priority*, *ingress*, *egress*, *transfer* tokens).
3094 - A matching pattern, starting with the *pattern* token and terminated by an
3096 - Actions, starting with the *actions* token and terminated by an *end*
3099 These translate directly to *rte_flow* objects provided as-is to the
3100 underlying functions.
3102 The shortest valid definition only comprises mandatory tokens::
3104 testpmd> flow create 0 pattern end actions end
3106 Note that PMDs may refuse rules that essentially do nothing such as this
3109 **All unspecified object values are automatically initialized to 0.**
3114 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
3115 specified before the ``pattern`` token.
3117 - ``group {group id}``: priority group.
3118 - ``priority {level}``: priority level within group.
3119 - ``ingress``: rule applies to ingress traffic.
3120 - ``egress``: rule applies to egress traffic.
3121 - ``transfer``: apply rule directly to endpoints found in pattern.
3123 Each instance of an attribute specified several times overrides the previous
3124 value as shown below (group 4 is used)::
3126 testpmd> flow create 0 group 42 group 24 group 4 [...]
3128 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
3130 While not specifying a direction is an error, some rules may allow both
3133 Most rules affect RX therefore contain the ``ingress`` token::
3135 testpmd> flow create 0 ingress pattern [...]
3140 A matching pattern starts after the ``pattern`` token. It is made of pattern
3141 items and is terminated by a mandatory ``end`` item.
3143 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
3144 rte_flow_item_type``).
3146 The ``/`` token is used as a separator between pattern items as shown
3149 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
3151 Note that protocol items like these must be stacked from lowest to highest
3152 layer to make sense. For instance, the following rule is either invalid or
3153 unlikely to match any packet::
3155 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
3157 More information on these restrictions can be found in the *rte_flow*
3160 Several items support additional specification structures, for example
3161 ``ipv4`` allows specifying source and destination addresses as follows::
3163 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
3164 dst is 10.2.0.0 / end [...]
3166 This rule matches all IPv4 traffic with the specified properties.
3168 In this example, ``src`` and ``dst`` are field names of the underlying
3169 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
3170 in a similar fashion.
3172 The ``is`` token means that the subsequent value must be matched exactly,
3173 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
3174 accordingly. Possible assignment tokens are:
3176 - ``is``: match value perfectly (with full bit-mask).
3177 - ``spec``: match value according to configured bit-mask.
3178 - ``last``: specify upper bound to establish a range.
3179 - ``mask``: specify bit-mask with relevant bits set to one.
3180 - ``prefix``: generate bit-mask from a prefix length.
3182 These yield identical results::
3184 ipv4 src is 10.1.1.1
3188 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
3192 ipv4 src spec 10.1.1.1 src prefix 32
3196 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
3200 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
3202 Inclusive ranges can be defined with ``last``::
3204 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
3206 Note that ``mask`` affects both ``spec`` and ``last``::
3208 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
3209 # matches 10.1.0.0 to 10.2.255.255
3211 Properties can be modified multiple times::
3213 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
3217 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
3222 This section lists supported pattern items and their attributes, if any.
3224 - ``end``: end list of pattern items.
3226 - ``void``: no-op pattern item.
3228 - ``invert``: perform actions when pattern does not match.
3230 - ``any``: match any protocol for the current layer.
3232 - ``num {unsigned}``: number of layers covered.
3234 - ``pf``: match traffic from/to the physical function.
3236 - ``vf``: match traffic from/to a virtual function ID.
3238 - ``id {unsigned}``: VF ID.
3240 - ``phy_port``: match traffic from/to a specific physical port.
3242 - ``index {unsigned}``: physical port index.
3244 - ``port_id``: match traffic from/to a given DPDK port ID.
3246 - ``id {unsigned}``: DPDK port ID.
3248 - ``mark``: match value set in previously matched flow rule using the mark action.
3250 - ``id {unsigned}``: arbitrary integer value.
3252 - ``raw``: match an arbitrary byte string.
3254 - ``relative {boolean}``: look for pattern after the previous item.
3255 - ``search {boolean}``: search pattern from offset (see also limit).
3256 - ``offset {integer}``: absolute or relative offset for pattern.
3257 - ``limit {unsigned}``: search area limit for start of pattern.
3258 - ``pattern {string}``: byte string to look for.
3260 - ``eth``: match Ethernet header.
3262 - ``dst {MAC-48}``: destination MAC.
3263 - ``src {MAC-48}``: source MAC.
3264 - ``type {unsigned}``: EtherType or TPID.
3266 - ``vlan``: match 802.1Q/ad VLAN tag.
3268 - ``tci {unsigned}``: tag control information.
3269 - ``pcp {unsigned}``: priority code point.
3270 - ``dei {unsigned}``: drop eligible indicator.
3271 - ``vid {unsigned}``: VLAN identifier.
3272 - ``inner_type {unsigned}``: inner EtherType or TPID.
3274 - ``ipv4``: match IPv4 header.
3276 - ``tos {unsigned}``: type of service.
3277 - ``ttl {unsigned}``: time to live.
3278 - ``proto {unsigned}``: next protocol ID.
3279 - ``src {ipv4 address}``: source address.
3280 - ``dst {ipv4 address}``: destination address.
3282 - ``ipv6``: match IPv6 header.
3284 - ``tc {unsigned}``: traffic class.
3285 - ``flow {unsigned}``: flow label.
3286 - ``proto {unsigned}``: protocol (next header).
3287 - ``hop {unsigned}``: hop limit.
3288 - ``src {ipv6 address}``: source address.
3289 - ``dst {ipv6 address}``: destination address.
3291 - ``icmp``: match ICMP header.
3293 - ``type {unsigned}``: ICMP packet type.
3294 - ``code {unsigned}``: ICMP packet code.
3296 - ``udp``: match UDP header.
3298 - ``src {unsigned}``: UDP source port.
3299 - ``dst {unsigned}``: UDP destination port.
3301 - ``tcp``: match TCP header.
3303 - ``src {unsigned}``: TCP source port.
3304 - ``dst {unsigned}``: TCP destination port.
3306 - ``sctp``: match SCTP header.
3308 - ``src {unsigned}``: SCTP source port.
3309 - ``dst {unsigned}``: SCTP destination port.
3310 - ``tag {unsigned}``: validation tag.
3311 - ``cksum {unsigned}``: checksum.
3313 - ``vxlan``: match VXLAN header.
3315 - ``vni {unsigned}``: VXLAN identifier.
3317 - ``e_tag``: match IEEE 802.1BR E-Tag header.
3319 - ``grp_ecid_b {unsigned}``: GRP and E-CID base.
3321 - ``nvgre``: match NVGRE header.
3323 - ``tni {unsigned}``: virtual subnet ID.
3325 - ``mpls``: match MPLS header.
3327 - ``label {unsigned}``: MPLS label.
3329 - ``gre``: match GRE header.
3331 - ``protocol {unsigned}``: protocol type.
3333 - ``fuzzy``: fuzzy pattern match, expect faster than default.
3335 - ``thresh {unsigned}``: accuracy threshold.
3337 - ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header.
3339 - ``teid {unsigned}``: tunnel endpoint identifier.
3341 - ``geneve``: match GENEVE header.
3343 - ``vni {unsigned}``: virtual network identifier.
3344 - ``protocol {unsigned}``: protocol type.
3346 - ``vxlan-gpe``: match VXLAN-GPE header.
3348 - ``vni {unsigned}``: VXLAN-GPE identifier.
3350 - ``arp_eth_ipv4``: match ARP header for Ethernet/IPv4.
3352 - ``sha {MAC-48}``: sender hardware address.
3353 - ``spa {ipv4 address}``: sender IPv4 address.
3354 - ``tha {MAC-48}``: target hardware address.
3355 - ``tpa {ipv4 address}``: target IPv4 address.
3357 - ``ipv6_ext``: match presence of any IPv6 extension header.
3359 - ``next_hdr {unsigned}``: next header.
3361 - ``icmp6``: match any ICMPv6 header.
3363 - ``type {unsigned}``: ICMPv6 type.
3364 - ``code {unsigned}``: ICMPv6 code.
3366 - ``icmp6_nd_ns``: match ICMPv6 neighbor discovery solicitation.
3368 - ``target_addr {ipv6 address}``: target address.
3370 - ``icmp6_nd_na``: match ICMPv6 neighbor discovery advertisement.
3372 - ``target_addr {ipv6 address}``: target address.
3374 - ``icmp6_nd_opt``: match presence of any ICMPv6 neighbor discovery option.
3376 - ``type {unsigned}``: ND option type.
3378 - ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery source Ethernet
3379 link-layer address option.
3381 - ``sla {MAC-48}``: source Ethernet LLA.
3383 - ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery target Ethernet
3384 link-layer address option.
3386 - ``tla {MAC-48}``: target Ethernet LLA.
3391 A list of actions starts after the ``actions`` token in the same fashion as
3392 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
3393 terminated by a mandatory ``end`` action.
3395 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
3396 rte_flow_action_type``).
3398 Dropping all incoming UDPv4 packets can be expressed as follows::
3400 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3403 Several actions have configurable properties which must be specified when
3404 there is no valid default value. For example, ``queue`` requires a target
3407 This rule redirects incoming UDPv4 traffic to queue index 6::
3409 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3410 actions queue index 6 / end
3412 While this one could be rejected by PMDs (unspecified queue index)::
3414 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3417 As defined by *rte_flow*, the list is not ordered, all actions of a given
3418 rule are performed simultaneously. These are equivalent::
3420 queue index 6 / void / mark id 42 / end
3424 void / mark id 42 / queue index 6 / end
3426 All actions in a list should have different types, otherwise only the last
3427 action of a given type is taken into account::
3429 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
3433 drop / drop / drop / end # drop is performed only once
3437 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
3439 Considering they are performed simultaneously, opposite and overlapping
3440 actions can sometimes be combined when the end result is unambiguous::
3442 drop / queue index 6 / end # drop has no effect
3446 queue index 6 / rss queues 6 7 8 / end # queue has no effect
3450 drop / passthru / end # drop has no effect
3452 Note that PMDs may still refuse such combinations.
3457 This section lists supported actions and their attributes, if any.
3459 - ``end``: end list of actions.
3461 - ``void``: no-op action.
3463 - ``passthru``: let subsequent rule process matched packets.
3465 - ``jump``: redirect traffic to group on device.
3467 - ``group {unsigned}``: group to redirect to.
3469 - ``mark``: attach 32 bit value to packets.
3471 - ``id {unsigned}``: 32 bit value to return with packets.
3473 - ``flag``: flag packets.
3475 - ``queue``: assign packets to a given queue index.
3477 - ``index {unsigned}``: queue index to use.
3479 - ``drop``: drop packets (note: passthru has priority).
3481 - ``count``: enable counters for this rule.
3483 - ``rss``: spread packets among several queues.
3485 - ``func {hash function}``: RSS hash function to apply, allowed tokens are
3486 the same as `set_hash_global_config`_.
3488 - ``level {unsigned}``: encapsulation level for ``types``.
3490 - ``types [{RSS hash type} [...]] end``: specific RSS hash types, allowed
3491 tokens are the same as `set_hash_input_set`_, except that an empty list
3492 does not disable RSS but instead requests unspecified "best-effort"
3495 - ``key {string}``: RSS hash key, overrides ``key_len``.
3497 - ``key_len {unsigned}``: RSS hash key length in bytes, can be used in
3498 conjunction with ``key`` to pad or truncate it.
3500 - ``queues [{unsigned} [...]] end``: queue indices to use.
3502 - ``pf``: direct traffic to physical function.
3504 - ``vf``: direct traffic to a virtual function ID.
3506 - ``original {boolean}``: use original VF ID if possible.
3507 - ``id {unsigned}``: VF ID.
3509 - ``phy_port``: direct packets to physical port index.
3511 - ``original {boolean}``: use original port index if possible.
3512 - ``index {unsigned}``: physical port index.
3514 - ``port_id``: direct matching traffic to a given DPDK port ID.
3516 - ``original {boolean}``: use original DPDK port ID if possible.
3517 - ``id {unsigned}``: DPDK port ID.
3519 - ``of_set_mpls_ttl``: OpenFlow's ``OFPAT_SET_MPLS_TTL``.
3521 - ``mpls_ttl``: MPLS TTL.
3523 - ``of_dec_mpls_ttl``: OpenFlow's ``OFPAT_DEC_MPLS_TTL``.
3525 - ``of_set_nw_ttl``: OpenFlow's ``OFPAT_SET_NW_TTL``.
3527 - ``nw_ttl``: IP TTL.
3529 - ``of_dec_nw_ttl``: OpenFlow's ``OFPAT_DEC_NW_TTL``.
3531 - ``of_copy_ttl_out``: OpenFlow's ``OFPAT_COPY_TTL_OUT``.
3533 - ``of_copy_ttl_in``: OpenFlow's ``OFPAT_COPY_TTL_IN``.
3535 - ``of_pop_vlan``: OpenFlow's ``OFPAT_POP_VLAN``.
3537 - ``of_push_vlan``: OpenFlow's ``OFPAT_PUSH_VLAN``.
3539 - ``ethertype``: Ethertype.
3541 - ``of_set_vlan_vid``: OpenFlow's ``OFPAT_SET_VLAN_VID``.
3543 - ``vlan_vid``: VLAN id.
3545 - ``of_set_vlan_pcp``: OpenFlow's ``OFPAT_SET_VLAN_PCP``.
3547 - ``vlan_pcp``: VLAN priority.
3549 - ``of_pop_mpls``: OpenFlow's ``OFPAT_POP_MPLS``.
3551 - ``ethertype``: Ethertype.
3553 - ``of_push_mpls``: OpenFlow's ``OFPAT_PUSH_MPLS``.
3555 - ``ethertype``: Ethertype.
3557 Destroying flow rules
3558 ~~~~~~~~~~~~~~~~~~~~~
3560 ``flow destroy`` destroys one or more rules from their rule ID (as returned
3561 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
3562 times as necessary::
3564 flow destroy {port_id} rule {rule_id} [...]
3566 If successful, it will show::
3568 Flow rule #[...] destroyed
3570 It does not report anything for rule IDs that do not exist. The usual error
3571 message is shown when a rule cannot be destroyed::
3573 Caught error type [...] ([...]): [...]
3575 ``flow flush`` destroys all rules on a device and does not take extra
3576 arguments. It is bound to ``rte_flow_flush()``::
3578 flow flush {port_id}
3580 Any errors are reported as above.
3582 Creating several rules and destroying them::
3584 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3585 actions queue index 2 / end
3586 Flow rule #0 created
3587 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3588 actions queue index 3 / end
3589 Flow rule #1 created
3590 testpmd> flow destroy 0 rule 0 rule 1
3591 Flow rule #1 destroyed
3592 Flow rule #0 destroyed
3595 The same result can be achieved using ``flow flush``::
3597 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3598 actions queue index 2 / end
3599 Flow rule #0 created
3600 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3601 actions queue index 3 / end
3602 Flow rule #1 created
3603 testpmd> flow flush 0
3606 Non-existent rule IDs are ignored::
3608 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3609 actions queue index 2 / end
3610 Flow rule #0 created
3611 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3612 actions queue index 3 / end
3613 Flow rule #1 created
3614 testpmd> flow destroy 0 rule 42 rule 10 rule 2
3616 testpmd> flow destroy 0 rule 0
3617 Flow rule #0 destroyed
3623 ``flow query`` queries a specific action of a flow rule having that
3624 ability. Such actions collect information that can be reported using this
3625 command. It is bound to ``rte_flow_query()``::
3627 flow query {port_id} {rule_id} {action}
3629 If successful, it will display either the retrieved data for known actions
3630 or the following message::
3632 Cannot display result for action type [...] ([...])
3634 Otherwise, it will complain either that the rule does not exist or that some
3637 Flow rule #[...] not found
3641 Caught error type [...] ([...]): [...]
3643 Currently only the ``count`` action is supported. This action reports the
3644 number of packets that hit the flow rule and the total number of bytes. Its
3645 output has the following format::
3648 hits_set: [...] # whether "hits" contains a valid value
3649 bytes_set: [...] # whether "bytes" contains a valid value
3650 hits: [...] # number of packets
3651 bytes: [...] # number of bytes
3653 Querying counters for TCPv6 packets redirected to queue 6::
3655 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
3656 actions queue index 6 / count / end
3657 Flow rule #4 created
3658 testpmd> flow query 0 4 count
3669 ``flow list`` lists existing flow rules sorted by priority and optionally
3670 filtered by group identifiers::
3672 flow list {port_id} [group {group_id}] [...]
3674 This command only fails with the following message if the device does not
3679 Output consists of a header line followed by a short description of each
3680 flow rule, one per line. There is no output at all when no flow rules are
3681 configured on the device::
3683 ID Group Prio Attr Rule
3684 [...] [...] [...] [...] [...]
3686 ``Attr`` column flags:
3688 - ``i`` for ``ingress``.
3689 - ``e`` for ``egress``.
3691 Creating several flow rules and listing them::
3693 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3694 actions queue index 6 / end
3695 Flow rule #0 created
3696 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3697 actions queue index 2 / end
3698 Flow rule #1 created
3699 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
3700 actions rss queues 6 7 8 end / end
3701 Flow rule #2 created
3702 testpmd> flow list 0
3703 ID Group Prio Attr Rule
3704 0 0 0 i- ETH IPV4 => QUEUE
3705 1 0 0 i- ETH IPV6 => QUEUE
3706 2 0 5 i- ETH IPV4 UDP => RSS
3709 Rules are sorted by priority (i.e. group ID first, then priority level)::
3711 testpmd> flow list 1
3712 ID Group Prio Attr Rule
3713 0 0 0 i- ETH => COUNT
3714 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3715 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3716 1 24 0 i- ETH IPV4 UDP => QUEUE
3717 4 24 10 i- ETH IPV4 TCP => DROP
3718 3 24 20 i- ETH IPV4 => DROP
3719 2 24 42 i- ETH IPV4 UDP => QUEUE
3720 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3723 Output can be limited to specific groups::
3725 testpmd> flow list 1 group 0 group 63
3726 ID Group Prio Attr Rule
3727 0 0 0 i- ETH => COUNT
3728 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3729 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3730 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3733 Toggling isolated mode
3734 ~~~~~~~~~~~~~~~~~~~~~~
3736 ``flow isolate`` can be used to tell the underlying PMD that ingress traffic
3737 must only be injected from the defined flow rules; that no default traffic
3738 is expected outside those rules and the driver is free to assign more
3739 resources to handle them. It is bound to ``rte_flow_isolate()``::
3741 flow isolate {port_id} {boolean}
3743 If successful, enabling or disabling isolated mode shows either::
3745 Ingress traffic on port [...]
3746 is now restricted to the defined flow rules
3750 Ingress traffic on port [...]
3751 is not restricted anymore to the defined flow rules
3753 Otherwise, in case of error::
3755 Caught error type [...] ([...]): [...]
3757 Mainly due to its side effects, PMDs supporting this mode may not have the
3758 ability to toggle it more than once without reinitializing affected ports
3759 first (e.g. by exiting testpmd).
3761 Enabling isolated mode::
3763 testpmd> flow isolate 0 true
3764 Ingress traffic on port 0 is now restricted to the defined flow rules
3767 Disabling isolated mode::
3769 testpmd> flow isolate 0 false
3770 Ingress traffic on port 0 is not restricted anymore to the defined flow rules
3773 Sample QinQ flow rules
3774 ~~~~~~~~~~~~~~~~~~~~~~
3776 Before creating QinQ rule(s) the following commands should be issued to enable QinQ::
3778 testpmd> port stop 0
3779 testpmd> vlan set qinq on 0
3781 The above command sets the inner and outer TPID's to 0x8100.
3783 To change the TPID's the following commands should be used::
3785 testpmd> vlan set outer tpid 0xa100 0
3786 testpmd> vlan set inner tpid 0x9100 0
3787 testpmd> port start 0
3789 Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM.
3793 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 /
3794 vlan tci is 456 / end actions vf id 1 / queue index 0 / end
3795 Flow rule #0 validated
3797 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 /
3798 vlan tci is 456 / end actions vf id 123 / queue index 0 / end
3799 Flow rule #0 created
3801 testpmd> flow list 0
3802 ID Group Prio Attr Rule
3803 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3805 Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host.
3809 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 /
3810 vlan tci is 654 / end actions pf / queue index 0 / end
3811 Flow rule #1 validated
3813 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 /
3814 vlan tci is 654 / end actions pf / queue index 1 / end
3815 Flow rule #1 created
3817 testpmd> flow list 0
3818 ID Group Prio Attr Rule
3819 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3820 1 0 0 i- ETH VLAN VLAN=>PF QUEUE