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 Commit port traffic management hierarchy
2511 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2513 Commit the traffic management hierarchy on the port::
2515 testpmd> port tm hierarchy commit (port_id) (clean_on_fail)
2519 * ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function
2520 call failure. On the other hand, hierarchy is preserved when this parameter
2523 Set port traffic management default hierarchy (tm forwarding mode)
2524 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2526 set the traffic management default hierarchy on the port::
2528 testpmd> set port tm hierarchy default (port_id)
2533 This section details the available filter functions that are available.
2535 Note these functions interface the deprecated legacy filtering framework,
2536 superseded by *rte_flow*. See `Flow rules management`_.
2539 ~~~~~~~~~~~~~~~~~~~~
2541 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
2543 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
2544 ethertype (ether_type) (drop|fwd) queue (queue_id)
2546 The available information parameters are:
2548 * ``port_id``: The port which the Ethertype filter assigned on.
2550 * ``mac_addr``: Compare destination mac address.
2552 * ``mac_ignr``: Ignore destination mac address match.
2554 * ``mac_address``: Destination mac address to match.
2556 * ``ether_type``: The EtherType value want to match,
2557 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
2559 * ``queue_id``: The receive queue associated with this EtherType filter.
2560 It is meaningless when deleting or dropping.
2562 Example, to add/remove an ethertype filter rule::
2564 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
2565 ethertype 0x0806 fwd queue 3
2567 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
2568 ethertype 0x0806 fwd queue 3
2573 Add or delete a 2-tuple filter,
2574 which identifies packets by specific protocol and destination TCP/UDP port
2575 and forwards packets into one of the receive queues::
2577 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
2578 protocol (protocol_value) mask (mask_value) \
2579 tcp_flags (tcp_flags_value) priority (prio_value) \
2582 The available information parameters are:
2584 * ``port_id``: The port which the 2-tuple filter assigned on.
2586 * ``dst_port_value``: Destination port in L4.
2588 * ``protocol_value``: IP L4 protocol.
2590 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
2592 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
2594 * ``prio_value``: Priority of this filter.
2596 * ``queue_id``: The receive queue associated with this 2-tuple filter.
2598 Example, to add/remove an 2tuple filter rule::
2600 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
2601 tcp_flags 0x02 priority 3 queue 3
2603 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
2604 tcp_flags 0x02 priority 3 queue 3
2609 Add or delete a 5-tuple filter,
2610 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
2611 and routes packets into one of the receive queues::
2613 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
2614 (src_address) dst_port (dst_port_value) \
2615 src_port (src_port_value) protocol (protocol_value) \
2616 mask (mask_value) tcp_flags (tcp_flags_value) \
2617 priority (prio_value) queue (queue_id)
2619 The available information parameters are:
2621 * ``port_id``: The port which the 5-tuple filter assigned on.
2623 * ``dst_address``: Destination IP address.
2625 * ``src_address``: Source IP address.
2627 * ``dst_port_value``: TCP/UDP destination port.
2629 * ``src_port_value``: TCP/UDP source port.
2631 * ``protocol_value``: L4 protocol.
2633 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
2635 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
2637 * ``prio_value``: The priority of this filter.
2639 * ``queue_id``: The receive queue associated with this 5-tuple filter.
2641 Example, to add/remove an 5tuple filter rule::
2643 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2644 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2645 flags 0x0 priority 3 queue 3
2647 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2648 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2649 flags 0x0 priority 3 queue 3
2654 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
2656 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
2658 The available information parameters are:
2660 * ``port_id``: The port which the SYN filter assigned on.
2662 * ``high``: This SYN filter has higher priority than other filters.
2664 * ``low``: This SYN filter has lower priority than other filters.
2666 * ``queue_id``: The receive queue associated with this SYN filter
2670 testpmd> syn_filter 0 add priority high queue 3
2675 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
2676 and routed into one of the receive queues::
2678 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
2679 mask (mask_value) priority (prio_value) queue (queue_id)
2681 The available information parameters are:
2683 * ``port_id``: The port which the Flex filter is assigned on.
2685 * ``len_value``: Filter length in bytes, no greater than 128.
2687 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
2689 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
2691 * ``prio_value``: The priority of this filter.
2693 * ``queue_id``: The receive queue associated with this Flex filter.
2697 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
2698 mask 000C priority 3 queue 3
2700 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
2701 mask 000C priority 3 queue 3
2704 .. _testpmd_flow_director:
2706 flow_director_filter
2707 ~~~~~~~~~~~~~~~~~~~~
2709 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
2711 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
2712 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
2714 * Perfect match filters.
2715 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2716 The masked fields are for IP flow.
2718 * Signature filters.
2719 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
2721 * Perfect-mac-vlan match filters.
2722 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2723 The masked fields are for MAC VLAN flow.
2725 * Perfect-tunnel match filters.
2726 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2727 The masked fields are for tunnel flow.
2729 * Perfect-raw-flow-type match filters.
2730 The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet.
2731 The masked fields are specified by input sets.
2733 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
2734 per flow type and the flexible payload.
2736 The Flow Director can also mask out parts of all of these fields so that filters
2737 are only applied to certain fields or parts of the fields.
2739 Note that for raw flow type mode the source and destination fields in the
2740 raw packet buffer need to be presented in a reversed order with respect
2741 to the expected received packets.
2742 For example: IP source and destination addresses or TCP/UDP/SCTP
2743 source and destination ports
2745 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
2747 # Commands to add flow director filters of different flow types::
2749 flow_director_filter (port_id) mode IP (add|del|update) \
2750 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
2751 src (src_ip_address) dst (dst_ip_address) \
2752 tos (tos_value) proto (proto_value) ttl (ttl_value) \
2753 vlan (vlan_value) flexbytes (flexbytes_value) \
2754 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
2757 flow_director_filter (port_id) mode IP (add|del|update) \
2758 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
2759 src (src_ip_address) (src_port) \
2760 dst (dst_ip_address) (dst_port) \
2761 tos (tos_value) ttl (ttl_value) \
2762 vlan (vlan_value) flexbytes (flexbytes_value) \
2763 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
2766 flow_director_filter (port_id) mode IP (add|del|update) \
2767 flow (ipv4-sctp|ipv6-sctp) \
2768 src (src_ip_address) (src_port) \
2769 dst (dst_ip_address) (dst_port) \
2770 tos (tos_value) ttl (ttl_value) \
2771 tag (verification_tag) vlan (vlan_value) \
2772 flexbytes (flexbytes_value) (drop|fwd) \
2773 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
2775 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
2776 ether (ethertype) flexbytes (flexbytes_value) \
2777 (drop|fwd) pf|vf(vf_id) queue (queue_id)
2780 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
2781 mac (mac_address) vlan (vlan_value) \
2782 flexbytes (flexbytes_value) (drop|fwd) \
2783 queue (queue_id) fd_id (fd_id_value)
2785 flow_director_filter (port_id) mode Tunnel (add|del|update) \
2786 mac (mac_address) vlan (vlan_value) \
2787 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
2788 flexbytes (flexbytes_value) (drop|fwd) \
2789 queue (queue_id) fd_id (fd_id_value)
2791 flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \
2792 (drop|fwd) queue (queue_id) fd_id (fd_id_value) \
2793 packet (packet file name)
2795 For example, to add an ipv4-udp flow type filter::
2797 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
2798 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
2799 fwd pf queue 1 fd_id 1
2801 For example, add an ipv4-other flow type filter::
2803 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
2804 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
2805 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
2810 Flush all flow director filters on a device::
2812 testpmd> flush_flow_director (port_id)
2814 Example, to flush all flow director filter on port 0::
2816 testpmd> flush_flow_director 0
2821 Set flow director's input masks::
2823 flow_director_mask (port_id) mode IP vlan (vlan_value) \
2824 src_mask (ipv4_src) (ipv6_src) (src_port) \
2825 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2827 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2829 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2830 mac (mac_value) tunnel-type (tunnel_type_value) \
2831 tunnel-id (tunnel_id_value)
2833 Example, to set flow director mask on port 0::
2835 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2836 src_mask 255.255.255.255 \
2837 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2838 dst_mask 255.255.255.255 \
2839 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2841 flow_director_flex_mask
2842 ~~~~~~~~~~~~~~~~~~~~~~~
2844 set masks of flow director's flexible payload based on certain flow type::
2846 testpmd> flow_director_flex_mask (port_id) \
2847 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2848 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2849 l2_payload|all) (mask)
2851 Example, to set flow director's flex mask for all flow type on port 0::
2853 testpmd> flow_director_flex_mask 0 flow all \
2854 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2857 flow_director_flex_payload
2858 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2860 Configure flexible payload selection::
2862 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2864 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2866 testpmd> flow_director_flex_payload 0 l4 \
2867 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2869 get_sym_hash_ena_per_port
2870 ~~~~~~~~~~~~~~~~~~~~~~~~~
2872 Get symmetric hash enable configuration per port::
2874 get_sym_hash_ena_per_port (port_id)
2876 For example, to get symmetric hash enable configuration of port 1::
2878 testpmd> get_sym_hash_ena_per_port 1
2880 set_sym_hash_ena_per_port
2881 ~~~~~~~~~~~~~~~~~~~~~~~~~
2883 Set symmetric hash enable configuration per port to enable or disable::
2885 set_sym_hash_ena_per_port (port_id) (enable|disable)
2887 For example, to set symmetric hash enable configuration of port 1 to enable::
2889 testpmd> set_sym_hash_ena_per_port 1 enable
2891 get_hash_global_config
2892 ~~~~~~~~~~~~~~~~~~~~~~
2894 Get the global configurations of hash filters::
2896 get_hash_global_config (port_id)
2898 For example, to get the global configurations of hash filters of port 1::
2900 testpmd> get_hash_global_config 1
2902 set_hash_global_config
2903 ~~~~~~~~~~~~~~~~~~~~~~
2905 Set the global configurations of hash filters::
2907 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2908 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2909 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \
2912 For example, to enable simple_xor for flow type of ipv6 on port 2::
2914 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2919 Set the input set for hash::
2921 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2922 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2923 l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
2924 ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2925 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2926 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2927 fld-8th|none) (select|add)
2929 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2931 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2936 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2937 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2939 Set the input set for flow director::
2941 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2942 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2943 l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
2944 ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2945 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2946 sctp-dst-port|sctp-veri-tag|none) (select|add)
2948 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2950 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2955 Set different GRE key length for input set::
2957 global_config (port_id) gre-key-len (number in bytes)
2959 For example to set GRE key length for input set to 4 bytes on port 0::
2961 testpmd> global_config 0 gre-key-len 4
2964 .. _testpmd_rte_flow:
2966 Flow rules management
2967 ---------------------
2969 Control of the generic flow API (*rte_flow*) is fully exposed through the
2970 ``flow`` command (validation, creation, destruction, queries and operation
2973 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2974 features simultaneously may cause undefined side-effects and is therefore
2980 Because the ``flow`` command uses dynamic tokens to handle the large number
2981 of possible flow rules combinations, its behavior differs slightly from
2982 other commands, in particular:
2984 - Pressing *?* or the *<tab>* key displays contextual help for the current
2985 token, not that of the entire command.
2987 - Optional and repeated parameters are supported (provided they are listed
2988 in the contextual help).
2990 The first parameter stands for the operation mode. Possible operations and
2991 their general syntax are described below. They are covered in detail in the
2994 - Check whether a flow rule can be created::
2996 flow validate {port_id}
2997 [group {group_id}] [priority {level}] [ingress] [egress]
2998 pattern {item} [/ {item} [...]] / end
2999 actions {action} [/ {action} [...]] / end
3001 - Create a flow rule::
3003 flow create {port_id}
3004 [group {group_id}] [priority {level}] [ingress] [egress]
3005 pattern {item} [/ {item} [...]] / end
3006 actions {action} [/ {action} [...]] / end
3008 - Destroy specific flow rules::
3010 flow destroy {port_id} rule {rule_id} [...]
3012 - Destroy all flow rules::
3014 flow flush {port_id}
3016 - Query an existing flow rule::
3018 flow query {port_id} {rule_id} {action}
3020 - List existing flow rules sorted by priority, filtered by group
3023 flow list {port_id} [group {group_id}] [...]
3025 - Restrict ingress traffic to the defined flow rules::
3027 flow isolate {port_id} {boolean}
3029 Validating flow rules
3030 ~~~~~~~~~~~~~~~~~~~~~
3032 ``flow validate`` reports whether a flow rule would be accepted by the
3033 underlying device in its current state but stops short of creating it. It is
3034 bound to ``rte_flow_validate()``::
3036 flow validate {port_id}
3037 [group {group_id}] [priority {level}] [ingress] [egress]
3038 pattern {item} [/ {item} [...]] / end
3039 actions {action} [/ {action} [...]] / end
3041 If successful, it will show::
3045 Otherwise it will show an error message of the form::
3047 Caught error type [...] ([...]): [...]
3049 This command uses the same parameters as ``flow create``, their format is
3050 described in `Creating flow rules`_.
3052 Check whether redirecting any Ethernet packet received on port 0 to RX queue
3053 index 6 is supported::
3055 testpmd> flow validate 0 ingress pattern eth / end
3056 actions queue index 6 / end
3060 Port 0 does not support TCPv6 rules::
3062 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
3064 Caught error type 9 (specific pattern item): Invalid argument
3070 ``flow create`` validates and creates the specified flow rule. It is bound
3071 to ``rte_flow_create()``::
3073 flow create {port_id}
3074 [group {group_id}] [priority {level}] [ingress] [egress]
3075 pattern {item} [/ {item} [...]] / end
3076 actions {action} [/ {action} [...]] / end
3078 If successful, it will return a flow rule ID usable with other commands::
3080 Flow rule #[...] created
3082 Otherwise it will show an error message of the form::
3084 Caught error type [...] ([...]): [...]
3086 Parameters describe in the following order:
3088 - Attributes (*group*, *priority*, *ingress*, *egress* tokens).
3089 - A matching pattern, starting with the *pattern* token and terminated by an
3091 - Actions, starting with the *actions* token and terminated by an *end*
3094 These translate directly to *rte_flow* objects provided as-is to the
3095 underlying functions.
3097 The shortest valid definition only comprises mandatory tokens::
3099 testpmd> flow create 0 pattern end actions end
3101 Note that PMDs may refuse rules that essentially do nothing such as this
3104 **All unspecified object values are automatically initialized to 0.**
3109 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
3110 specified before the ``pattern`` token.
3112 - ``group {group id}``: priority group.
3113 - ``priority {level}``: priority level within group.
3114 - ``ingress``: rule applies to ingress traffic.
3115 - ``egress``: rule applies to egress traffic.
3117 Each instance of an attribute specified several times overrides the previous
3118 value as shown below (group 4 is used)::
3120 testpmd> flow create 0 group 42 group 24 group 4 [...]
3122 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
3124 While not specifying a direction is an error, some rules may allow both
3127 Most rules affect RX therefore contain the ``ingress`` token::
3129 testpmd> flow create 0 ingress pattern [...]
3134 A matching pattern starts after the ``pattern`` token. It is made of pattern
3135 items and is terminated by a mandatory ``end`` item.
3137 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
3138 rte_flow_item_type``).
3140 The ``/`` token is used as a separator between pattern items as shown
3143 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
3145 Note that protocol items like these must be stacked from lowest to highest
3146 layer to make sense. For instance, the following rule is either invalid or
3147 unlikely to match any packet::
3149 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
3151 More information on these restrictions can be found in the *rte_flow*
3154 Several items support additional specification structures, for example
3155 ``ipv4`` allows specifying source and destination addresses as follows::
3157 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
3158 dst is 10.2.0.0 / end [...]
3160 This rule matches all IPv4 traffic with the specified properties.
3162 In this example, ``src`` and ``dst`` are field names of the underlying
3163 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
3164 in a similar fashion.
3166 The ``is`` token means that the subsequent value must be matched exactly,
3167 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
3168 accordingly. Possible assignment tokens are:
3170 - ``is``: match value perfectly (with full bit-mask).
3171 - ``spec``: match value according to configured bit-mask.
3172 - ``last``: specify upper bound to establish a range.
3173 - ``mask``: specify bit-mask with relevant bits set to one.
3174 - ``prefix``: generate bit-mask from a prefix length.
3176 These yield identical results::
3178 ipv4 src is 10.1.1.1
3182 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
3186 ipv4 src spec 10.1.1.1 src prefix 32
3190 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
3194 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
3196 Inclusive ranges can be defined with ``last``::
3198 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
3200 Note that ``mask`` affects both ``spec`` and ``last``::
3202 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
3203 # matches 10.1.0.0 to 10.2.255.255
3205 Properties can be modified multiple times::
3207 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
3211 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
3216 This section lists supported pattern items and their attributes, if any.
3218 - ``end``: end list of pattern items.
3220 - ``void``: no-op pattern item.
3222 - ``invert``: perform actions when pattern does not match.
3224 - ``any``: match any protocol for the current layer.
3226 - ``num {unsigned}``: number of layers covered.
3228 - ``pf``: match packets addressed to the physical function.
3230 - ``vf``: match packets addressed to a virtual function ID.
3232 - ``id {unsigned}``: destination VF ID.
3234 - ``port``: device-specific physical port index to use.
3236 - ``index {unsigned}``: physical port index.
3238 - ``raw``: match an arbitrary byte string.
3240 - ``relative {boolean}``: look for pattern after the previous item.
3241 - ``search {boolean}``: search pattern from offset (see also limit).
3242 - ``offset {integer}``: absolute or relative offset for pattern.
3243 - ``limit {unsigned}``: search area limit for start of pattern.
3244 - ``pattern {string}``: byte string to look for.
3246 - ``eth``: match Ethernet header.
3248 - ``dst {MAC-48}``: destination MAC.
3249 - ``src {MAC-48}``: source MAC.
3250 - ``type {unsigned}``: EtherType.
3252 - ``vlan``: match 802.1Q/ad VLAN tag.
3254 - ``tpid {unsigned}``: tag protocol identifier.
3255 - ``tci {unsigned}``: tag control information.
3256 - ``pcp {unsigned}``: priority code point.
3257 - ``dei {unsigned}``: drop eligible indicator.
3258 - ``vid {unsigned}``: VLAN identifier.
3260 - ``ipv4``: match IPv4 header.
3262 - ``tos {unsigned}``: type of service.
3263 - ``ttl {unsigned}``: time to live.
3264 - ``proto {unsigned}``: next protocol ID.
3265 - ``src {ipv4 address}``: source address.
3266 - ``dst {ipv4 address}``: destination address.
3268 - ``ipv6``: match IPv6 header.
3270 - ``tc {unsigned}``: traffic class.
3271 - ``flow {unsigned}``: flow label.
3272 - ``proto {unsigned}``: protocol (next header).
3273 - ``hop {unsigned}``: hop limit.
3274 - ``src {ipv6 address}``: source address.
3275 - ``dst {ipv6 address}``: destination address.
3277 - ``icmp``: match ICMP header.
3279 - ``type {unsigned}``: ICMP packet type.
3280 - ``code {unsigned}``: ICMP packet code.
3282 - ``udp``: match UDP header.
3284 - ``src {unsigned}``: UDP source port.
3285 - ``dst {unsigned}``: UDP destination port.
3287 - ``tcp``: match TCP header.
3289 - ``src {unsigned}``: TCP source port.
3290 - ``dst {unsigned}``: TCP destination port.
3292 - ``sctp``: match SCTP header.
3294 - ``src {unsigned}``: SCTP source port.
3295 - ``dst {unsigned}``: SCTP destination port.
3296 - ``tag {unsigned}``: validation tag.
3297 - ``cksum {unsigned}``: checksum.
3299 - ``vxlan``: match VXLAN header.
3301 - ``vni {unsigned}``: VXLAN identifier.
3303 - ``e_tag``: match IEEE 802.1BR E-Tag header.
3305 - ``grp_ecid_b {unsigned}``: GRP and E-CID base.
3307 - ``nvgre``: match NVGRE header.
3309 - ``tni {unsigned}``: virtual subnet ID.
3311 - ``mpls``: match MPLS header.
3313 - ``label {unsigned}``: MPLS label.
3315 - ``gre``: match GRE header.
3317 - ``protocol {unsigned}``: protocol type.
3319 - ``fuzzy``: fuzzy pattern match, expect faster than default.
3321 - ``thresh {unsigned}``: accuracy threshold.
3323 - ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header.
3325 - ``teid {unsigned}``: tunnel endpoint identifier.
3327 - ``geneve``: match GENEVE header.
3329 - ``vni {unsigned}``: virtual network identifier.
3330 - ``protocol {unsigned}``: protocol type.
3335 A list of actions starts after the ``actions`` token in the same fashion as
3336 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
3337 terminated by a mandatory ``end`` action.
3339 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
3340 rte_flow_action_type``).
3342 Dropping all incoming UDPv4 packets can be expressed as follows::
3344 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3347 Several actions have configurable properties which must be specified when
3348 there is no valid default value. For example, ``queue`` requires a target
3351 This rule redirects incoming UDPv4 traffic to queue index 6::
3353 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3354 actions queue index 6 / end
3356 While this one could be rejected by PMDs (unspecified queue index)::
3358 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3361 As defined by *rte_flow*, the list is not ordered, all actions of a given
3362 rule are performed simultaneously. These are equivalent::
3364 queue index 6 / void / mark id 42 / end
3368 void / mark id 42 / queue index 6 / end
3370 All actions in a list should have different types, otherwise only the last
3371 action of a given type is taken into account::
3373 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
3377 drop / drop / drop / end # drop is performed only once
3381 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
3383 Considering they are performed simultaneously, opposite and overlapping
3384 actions can sometimes be combined when the end result is unambiguous::
3386 drop / queue index 6 / end # drop has no effect
3390 queue index 6 / rss queues 6 7 8 / end # queue has no effect
3394 drop / passthru / end # drop has no effect
3396 Note that PMDs may still refuse such combinations.
3401 This section lists supported actions and their attributes, if any.
3403 - ``end``: end list of actions.
3405 - ``void``: no-op action.
3407 - ``passthru``: let subsequent rule process matched packets.
3409 - ``mark``: attach 32 bit value to packets.
3411 - ``id {unsigned}``: 32 bit value to return with packets.
3413 - ``flag``: flag packets.
3415 - ``queue``: assign packets to a given queue index.
3417 - ``index {unsigned}``: queue index to use.
3419 - ``drop``: drop packets (note: passthru has priority).
3421 - ``count``: enable counters for this rule.
3423 - ``rss``: spread packets among several queues.
3425 - ``types [{RSS hash type} [...]] end``: specific RSS hash types, allowed
3426 tokens are the same as `set_hash_input_set`_, except that an empty list
3427 does not disable RSS but instead requests unspecified "best-effort"
3430 - ``key {string}``: RSS hash key, overrides ``key_len``.
3432 - ``key_len {unsigned}``: RSS hash key length in bytes, can be used in
3433 conjunction with ``key`` to pad or truncate it.
3435 - ``queues [{unsigned} [...]] end``: queue indices to use.
3437 - ``pf``: redirect packets to physical device function.
3439 - ``vf``: redirect packets to virtual device function.
3441 - ``original {boolean}``: use original VF ID if possible.
3442 - ``id {unsigned}``: VF ID to redirect packets to.
3444 Destroying flow rules
3445 ~~~~~~~~~~~~~~~~~~~~~
3447 ``flow destroy`` destroys one or more rules from their rule ID (as returned
3448 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
3449 times as necessary::
3451 flow destroy {port_id} rule {rule_id} [...]
3453 If successful, it will show::
3455 Flow rule #[...] destroyed
3457 It does not report anything for rule IDs that do not exist. The usual error
3458 message is shown when a rule cannot be destroyed::
3460 Caught error type [...] ([...]): [...]
3462 ``flow flush`` destroys all rules on a device and does not take extra
3463 arguments. It is bound to ``rte_flow_flush()``::
3465 flow flush {port_id}
3467 Any errors are reported as above.
3469 Creating several rules and destroying them::
3471 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3472 actions queue index 2 / end
3473 Flow rule #0 created
3474 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3475 actions queue index 3 / end
3476 Flow rule #1 created
3477 testpmd> flow destroy 0 rule 0 rule 1
3478 Flow rule #1 destroyed
3479 Flow rule #0 destroyed
3482 The same result can be achieved using ``flow flush``::
3484 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3485 actions queue index 2 / end
3486 Flow rule #0 created
3487 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3488 actions queue index 3 / end
3489 Flow rule #1 created
3490 testpmd> flow flush 0
3493 Non-existent rule IDs are ignored::
3495 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3496 actions queue index 2 / end
3497 Flow rule #0 created
3498 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3499 actions queue index 3 / end
3500 Flow rule #1 created
3501 testpmd> flow destroy 0 rule 42 rule 10 rule 2
3503 testpmd> flow destroy 0 rule 0
3504 Flow rule #0 destroyed
3510 ``flow query`` queries a specific action of a flow rule having that
3511 ability. Such actions collect information that can be reported using this
3512 command. It is bound to ``rte_flow_query()``::
3514 flow query {port_id} {rule_id} {action}
3516 If successful, it will display either the retrieved data for known actions
3517 or the following message::
3519 Cannot display result for action type [...] ([...])
3521 Otherwise, it will complain either that the rule does not exist or that some
3524 Flow rule #[...] not found
3528 Caught error type [...] ([...]): [...]
3530 Currently only the ``count`` action is supported. This action reports the
3531 number of packets that hit the flow rule and the total number of bytes. Its
3532 output has the following format::
3535 hits_set: [...] # whether "hits" contains a valid value
3536 bytes_set: [...] # whether "bytes" contains a valid value
3537 hits: [...] # number of packets
3538 bytes: [...] # number of bytes
3540 Querying counters for TCPv6 packets redirected to queue 6::
3542 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
3543 actions queue index 6 / count / end
3544 Flow rule #4 created
3545 testpmd> flow query 0 4 count
3556 ``flow list`` lists existing flow rules sorted by priority and optionally
3557 filtered by group identifiers::
3559 flow list {port_id} [group {group_id}] [...]
3561 This command only fails with the following message if the device does not
3566 Output consists of a header line followed by a short description of each
3567 flow rule, one per line. There is no output at all when no flow rules are
3568 configured on the device::
3570 ID Group Prio Attr Rule
3571 [...] [...] [...] [...] [...]
3573 ``Attr`` column flags:
3575 - ``i`` for ``ingress``.
3576 - ``e`` for ``egress``.
3578 Creating several flow rules and listing them::
3580 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3581 actions queue index 6 / end
3582 Flow rule #0 created
3583 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3584 actions queue index 2 / end
3585 Flow rule #1 created
3586 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
3587 actions rss queues 6 7 8 end / end
3588 Flow rule #2 created
3589 testpmd> flow list 0
3590 ID Group Prio Attr Rule
3591 0 0 0 i- ETH IPV4 => QUEUE
3592 1 0 0 i- ETH IPV6 => QUEUE
3593 2 0 5 i- ETH IPV4 UDP => RSS
3596 Rules are sorted by priority (i.e. group ID first, then priority level)::
3598 testpmd> flow list 1
3599 ID Group Prio Attr Rule
3600 0 0 0 i- ETH => COUNT
3601 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3602 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3603 1 24 0 i- ETH IPV4 UDP => QUEUE
3604 4 24 10 i- ETH IPV4 TCP => DROP
3605 3 24 20 i- ETH IPV4 => DROP
3606 2 24 42 i- ETH IPV4 UDP => QUEUE
3607 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3610 Output can be limited to specific groups::
3612 testpmd> flow list 1 group 0 group 63
3613 ID Group Prio Attr Rule
3614 0 0 0 i- ETH => COUNT
3615 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3616 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3617 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3620 Toggling isolated mode
3621 ~~~~~~~~~~~~~~~~~~~~~~
3623 ``flow isolate`` can be used to tell the underlying PMD that ingress traffic
3624 must only be injected from the defined flow rules; that no default traffic
3625 is expected outside those rules and the driver is free to assign more
3626 resources to handle them. It is bound to ``rte_flow_isolate()``::
3628 flow isolate {port_id} {boolean}
3630 If successful, enabling or disabling isolated mode shows either::
3632 Ingress traffic on port [...]
3633 is now restricted to the defined flow rules
3637 Ingress traffic on port [...]
3638 is not restricted anymore to the defined flow rules
3640 Otherwise, in case of error::
3642 Caught error type [...] ([...]): [...]
3644 Mainly due to its side effects, PMDs supporting this mode may not have the
3645 ability to toggle it more than once without reinitializing affected ports
3646 first (e.g. by exiting testpmd).
3648 Enabling isolated mode::
3650 testpmd> flow isolate 0 true
3651 Ingress traffic on port 0 is now restricted to the defined flow rules
3654 Disabling isolated mode::
3656 testpmd> flow isolate 0 false
3657 Ingress traffic on port 0 is not restricted anymore to the defined flow rules
3660 Sample QinQ flow rules
3661 ~~~~~~~~~~~~~~~~~~~~~~
3663 Before creating QinQ rule(s) the following commands should be issued to enable QinQ::
3665 testpmd> port stop 0
3666 testpmd> vlan set qinq on 0
3668 The above command sets the inner and outer TPID's to 0x8100.
3670 To change the TPID's the following commands should be used::
3672 testpmd> vlan set outer tpid 0xa100 0
3673 testpmd> vlan set inner tpid 0x9100 0
3674 testpmd> port start 0
3676 Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM.
3680 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 /
3681 vlan tci is 456 / end actions vf id 1 / queue index 0 / end
3682 Flow rule #0 validated
3684 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 /
3685 vlan tci is 456 / end actions vf id 123 / queue index 0 / end
3686 Flow rule #0 created
3688 testpmd> flow list 0
3689 ID Group Prio Attr Rule
3690 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3692 Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host.
3696 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 /
3697 vlan tci is 654 / end actions pf / queue index 0 / end
3698 Flow rule #1 validated
3700 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 /
3701 vlan tci is 654 / end actions pf / queue index 1 / end
3702 Flow rule #1 created
3704 testpmd> flow list 0
3705 ID Group Prio Attr Rule
3706 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3707 1 0 0 i- ETH VLAN VLAN=>PF QUEUE