1 .. SPDX-License-Identifier: BSD-3-Clause
2 Copyright(c) 2010-2016 Intel Corporation.
6 Testpmd Runtime Functions
7 =========================
9 Where the testpmd application is started in interactive mode, (``-i|--interactive``),
10 it displays a prompt that can be used to start and stop forwarding,
11 configure the application, display statistics (including the extended NIC
12 statistics aka xstats) , set the Flow Director and other tasks::
16 The testpmd prompt has some, limited, readline support.
17 Common bash command-line functions such as ``Ctrl+a`` and ``Ctrl+e`` to go to the start and end of the prompt line are supported
18 as well as access to the command history via the up-arrow.
20 There is also support for tab completion.
21 If you type a partial command and hit ``<TAB>`` you get a list of the available completions:
23 .. code-block:: console
25 testpmd> show port <TAB>
27 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
28 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
29 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
30 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
36 Some examples in this document are too long to fit on one line are are shown wrapped at `"\\"` for display purposes::
38 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
39 (pause_time) (send_xon) (port_id)
41 In the real ``testpmd>`` prompt these commands should be on a single line.
46 The testpmd has on-line help for the functions that are available at runtime.
47 These are divided into sections and can be accessed using help, help section or help all:
49 .. code-block:: console
53 help control : Start and stop forwarding.
54 help display : Displaying port, stats and config information.
55 help config : Configuration information.
56 help ports : Configuring ports.
57 help registers : Reading and setting port registers.
58 help filters : Filters configuration help.
59 help all : All of the above sections.
62 Command File Functions
63 ----------------------
65 To facilitate loading large number of commands or to avoid cutting and pasting where not
66 practical or possible testpmd supports alternative methods for executing commands.
68 * If started with the ``--cmdline-file=FILENAME`` command line argument testpmd
69 will execute all CLI commands contained within the file immediately before
70 starting packet forwarding or entering interactive mode.
72 .. code-block:: console
74 ./testpmd -n4 -r2 ... -- -i --cmdline-file=/home/ubuntu/flow-create-commands.txt
75 Interactive-mode selected
76 CLI commands to be read from /home/ubuntu/flow-create-commands.txt
77 Configuring Port 0 (socket 0)
78 Port 0: 7C:FE:90:CB:74:CE
79 Configuring Port 1 (socket 0)
80 Port 1: 7C:FE:90:CB:74:CA
81 Checking link statuses...
82 Port 0 Link Up - speed 10000 Mbps - full-duplex
83 Port 1 Link Up - speed 10000 Mbps - full-duplex
89 Flow rule #498 created
90 Flow rule #499 created
91 Read all CLI commands from /home/ubuntu/flow-create-commands.txt
95 * At run-time additional commands can be loaded in bulk by invoking the ``load FILENAME``
98 .. code-block:: console
100 testpmd> load /home/ubuntu/flow-create-commands.txt
105 Flow rule #498 created
106 Flow rule #499 created
107 Read all CLI commands from /home/ubuntu/flow-create-commands.txt
111 In all cases output from any included command will be displayed as standard output.
112 Execution will continue until the end of the file is reached regardless of
113 whether any errors occur. The end user must examine the output to determine if
114 any failures occurred.
123 Start packet forwarding with current configuration::
130 Start packet forwarding with current configuration after sending specified number of bursts of packets::
132 testpmd> start tx_first (""|burst_num)
134 The default burst number is 1 when ``burst_num`` not presented.
139 Stop packet forwarding, and display accumulated statistics::
154 The functions in the following sections are used to display information about the
155 testpmd configuration or the NIC status.
160 Display information for a given port or all ports::
162 testpmd> show port (info|summary|stats|xstats|fdir|stat_qmap|dcb_tc|cap) (port_id|all)
164 The available information categories are:
166 * ``info``: General port information such as MAC address.
168 * ``summary``: Brief port summary such as Device Name, Driver Name etc.
170 * ``stats``: RX/TX statistics.
172 * ``xstats``: RX/TX extended NIC statistics.
174 * ``fdir``: Flow Director information and statistics.
176 * ``stat_qmap``: Queue statistics mapping.
178 * ``dcb_tc``: DCB information such as TC mapping.
180 * ``cap``: Supported offload capabilities.
184 .. code-block:: console
186 testpmd> show port info 0
188 ********************* Infos for port 0 *********************
190 MAC address: XX:XX:XX:XX:XX:XX
192 memory allocation on the socket: 0
194 Link speed: 40000 Mbps
195 Link duplex: full-duplex
196 Promiscuous mode: enabled
197 Allmulticast mode: disabled
198 Maximum number of MAC addresses: 64
199 Maximum number of MAC addresses of hash filtering: 0
204 Redirection table size: 512
205 Supported flow types:
225 Display the rss redirection table entry indicated by masks on port X::
227 testpmd> show port (port_id) rss reta (size) (mask0, mask1...)
229 size is used to indicate the hardware supported reta size
234 Display the RSS hash functions and RSS hash key of a port::
236 testpmd> show port (port_id) rss-hash [key]
241 Clear the port statistics for a given port or for all ports::
243 testpmd> clear port (info|stats|xstats|fdir|stat_qmap) (port_id|all)
247 testpmd> clear port stats all
252 Display information for a given port's RX/TX queue::
254 testpmd> show (rxq|txq) info (port_id) (queue_id)
259 Displays the configuration of the application.
260 The configuration comes from the command-line, the runtime or the application defaults::
262 testpmd> show config (rxtx|cores|fwd|txpkts)
264 The available information categories are:
266 * ``rxtx``: RX/TX configuration items.
268 * ``cores``: List of forwarding cores.
270 * ``fwd``: Packet forwarding configuration.
272 * ``txpkts``: Packets to TX configuration.
276 .. code-block:: console
278 testpmd> show config rxtx
280 io packet forwarding - CRC stripping disabled - packets/burst=16
281 nb forwarding cores=2 - nb forwarding ports=1
282 RX queues=1 - RX desc=128 - RX free threshold=0
283 RX threshold registers: pthresh=8 hthresh=8 wthresh=4
284 TX queues=1 - TX desc=512 - TX free threshold=0
285 TX threshold registers: pthresh=36 hthresh=0 wthresh=0
286 TX RS bit threshold=0 - TXQ flags=0x0
291 Set the packet forwarding mode::
293 testpmd> set fwd (io|mac|macswap|flowgen| \
294 rxonly|txonly|csum|icmpecho) (""|retry)
296 ``retry`` can be specified for forwarding engines except ``rx_only``.
298 The available information categories are:
300 * ``io``: Forwards packets "as-is" in I/O mode.
301 This is the fastest possible forwarding operation as it does not access packets data.
302 This is the default mode.
304 * ``mac``: Changes the source and the destination Ethernet addresses of packets before forwarding them.
305 Default application behaviour is to set source Ethernet address to that of the transmitting interface, and destination
306 address to a dummy value (set during init). The user may specify a target destination Ethernet address via the 'eth-peer' or
307 'eth-peer-configfile' command-line options. It is not currently possible to specify a specific source Ethernet address.
309 * ``macswap``: MAC swap forwarding mode.
310 Swaps the source and the destination Ethernet addresses of packets before forwarding them.
312 * ``flowgen``: Multi-flow generation mode.
313 Originates a number of flows (with varying destination IP addresses), and terminate receive traffic.
315 * ``rxonly``: Receives packets but doesn't transmit them.
317 * ``txonly``: Generates and transmits packets without receiving any.
319 * ``csum``: Changes the checksum field with hardware or software methods depending on the offload flags on the packet.
321 * ``icmpecho``: Receives a burst of packets, lookup for IMCP echo requests and, if any, send back ICMP echo replies.
323 * ``ieee1588``: Demonstrate L2 IEEE1588 V2 PTP timestamping for RX and TX. Requires ``CONFIG_RTE_LIBRTE_IEEE1588=y``.
325 * ``softnic``: Demonstrates the softnic forwarding operation. In this mode, packet forwarding is
326 similar to I/O mode except for the fact that packets are loopback to the softnic ports only. Therefore, portmask parameter should be set to softnic port only. The various software based custom NIC pipelines specified through the softnic firmware (DPDK packet framework script) can be tested in this mode. Furthermore, it allows to build 5-level hierarchical QoS scheduler as a default option that can be enabled through CLI once testpmd application is initialised. The user can modify the default scheduler hierarchy or can specify the new QoS Scheduler hierarchy through CLI. Requires ``CONFIG_RTE_LIBRTE_PMD_SOFTNIC=y``.
330 testpmd> set fwd rxonly
332 Set rxonly packet forwarding mode
338 Display an RX descriptor for a port RX queue::
340 testpmd> read rxd (port_id) (queue_id) (rxd_id)
344 testpmd> read rxd 0 0 4
345 0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180
350 Display a TX descriptor for a port TX queue::
352 testpmd> read txd (port_id) (queue_id) (txd_id)
356 testpmd> read txd 0 0 4
357 0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C
362 Get loaded dynamic device personalization (DDP) package info list::
364 testpmd> ddp get list (port_id)
369 Display information about dynamic device personalization (DDP) profile::
371 testpmd> ddp get info (profile_path)
376 Display VF statistics::
378 testpmd> show vf stats (port_id) (vf_id)
383 Reset VF statistics::
385 testpmd> clear vf stats (port_id) (vf_id)
387 show port pctype mapping
388 ~~~~~~~~~~~~~~~~~~~~~~~~
390 List all items from the pctype mapping table::
392 testpmd> show port (port_id) pctype mapping
394 show rx offloading capabilities
395 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
397 List all per queue and per port Rx offloading capabilities of a port::
399 testpmd> show port (port_id) rx_offload capabilities
401 show rx offloading configuration
402 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
404 List port level and all queue level Rx offloading configuration::
406 testpmd> show port (port_id) rx_offload configuration
408 show tx offloading capabilities
409 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
411 List all per queue and per port Tx offloading capabilities of a port::
413 testpmd> show port (port_id) tx_offload capabilities
415 show tx offloading configuration
416 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
418 List port level and all queue level Tx offloading configuration::
420 testpmd> show port (port_id) tx_offload configuration
423 Configuration Functions
424 -----------------------
426 The testpmd application can be configured from the runtime as well as from the command-line.
428 This section details the available configuration functions that are available.
432 Configuration changes only become active when forwarding is started/restarted.
437 Reset forwarding to the default configuration::
444 Set the debug verbosity level::
446 testpmd> set verbose (level)
448 Currently the only available levels are 0 (silent except for error) and 1 (fully verbose).
453 Set the log level for a log type::
455 testpmd> set log global|(type) (level)
459 * ``type`` is the log name.
461 * ``level`` is the log level.
463 For example, to change the global log level::
464 testpmd> set log global (level)
466 Regexes can also be used for type. To change log level of user1, user2 and user3::
467 testpmd> set log user[1-3] (level)
472 Set the number of ports used by the application:
476 This is equivalent to the ``--nb-ports`` command-line option.
481 Set the number of cores used by the application::
483 testpmd> set nbcore (num)
485 This is equivalent to the ``--nb-cores`` command-line option.
489 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port.
494 Set the forwarding cores hexadecimal mask::
496 testpmd> set coremask (mask)
498 This is equivalent to the ``--coremask`` command-line option.
502 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding.
507 Set the forwarding ports hexadecimal mask::
509 testpmd> set portmask (mask)
511 This is equivalent to the ``--portmask`` command-line option.
516 Set number of packets per burst::
518 testpmd> set burst (num)
520 This is equivalent to the ``--burst command-line`` option.
522 When retry is enabled, the transmit delay time and number of retries can also be set::
524 testpmd> set burst tx delay (microseconds) retry (num)
529 Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode::
531 testpmd> set txpkts (x[,y]*)
533 Where x[,y]* represents a CSV list of values, without white space.
538 Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes::
540 testpmd> set txsplit (off|on|rand)
544 * ``off`` disable packet copy & split for CSUM mode.
546 * ``on`` split outgoing packet into multiple segments. Size of each segment
547 and number of segments per packet is determined by ``set txpkts`` command
550 * ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments.
555 Set the list of forwarding cores::
557 testpmd> set corelist (x[,y]*)
559 For example, to change the forwarding cores:
561 .. code-block:: console
563 testpmd> set corelist 3,1
564 testpmd> show config fwd
566 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled
567 Logical Core 3 (socket 0) forwards packets on 1 streams:
568 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01
569 Logical Core 1 (socket 0) forwards packets on 1 streams:
570 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
574 The cores are used in the same order as specified on the command line.
579 Set the list of forwarding ports::
581 testpmd> set portlist (x[,y]*)
583 For example, to change the port forwarding:
585 .. code-block:: console
587 testpmd> set portlist 0,2,1,3
588 testpmd> show config fwd
590 io packet forwarding - ports=4 - cores=1 - streams=4
591 Logical Core 3 (socket 0) forwards packets on 4 streams:
592 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01
593 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
594 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03
595 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02
600 Enable/disable tx loopback::
602 testpmd> set tx loopback (port_id) (on|off)
607 set drop enable bit for all queues::
609 testpmd> set all queues drop (port_id) (on|off)
611 set split drop enable (for VF)
612 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
614 set split drop enable bit for VF from PF::
616 testpmd> set vf split drop (port_id) (vf_id) (on|off)
618 set mac antispoof (for VF)
619 ~~~~~~~~~~~~~~~~~~~~~~~~~~
621 Set mac antispoof for a VF from the PF::
623 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off)
628 Enable/disable MACsec offload::
630 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off)
631 testpmd> set macsec offload (port_id) off
636 Configure MACsec secure connection (SC)::
638 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi)
642 The pi argument is ignored for tx.
643 Check the NIC Datasheet for hardware limits.
648 Configure MACsec secure association (SA)::
650 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key)
654 The IDX value must be 0 or 1.
655 Check the NIC Datasheet for hardware limits.
657 set broadcast mode (for VF)
658 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
660 Set broadcast mode for a VF from the PF::
662 testpmd> set vf broadcast (port_id) (vf_id) (on|off)
667 Set the VLAN strip on a port::
669 testpmd> vlan set strip (on|off) (port_id)
674 Set the VLAN strip for a queue on a port::
676 testpmd> vlan set stripq (on|off) (port_id,queue_id)
678 vlan set stripq (for VF)
679 ~~~~~~~~~~~~~~~~~~~~~~~~
681 Set VLAN strip for all queues in a pool for a VF from the PF::
683 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off)
685 vlan set insert (for VF)
686 ~~~~~~~~~~~~~~~~~~~~~~~~
688 Set VLAN insert for a VF from the PF::
690 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id)
692 vlan set tag (for VF)
693 ~~~~~~~~~~~~~~~~~~~~~
695 Set VLAN tag for a VF from the PF::
697 testpmd> set vf vlan tag (port_id) (vf_id) (on|off)
699 vlan set antispoof (for VF)
700 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
702 Set VLAN antispoof for a VF from the PF::
704 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off)
709 Set the VLAN filter on a port::
711 testpmd> vlan set filter (on|off) (port_id)
716 Set the VLAN QinQ (extended queue in queue) on for a port::
718 testpmd> vlan set qinq (on|off) (port_id)
723 Set the inner or outer VLAN TPID for packet filtering on a port::
725 testpmd> vlan set (inner|outer) tpid (value) (port_id)
729 TPID value must be a 16-bit number (value <= 65536).
734 Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID::
736 testpmd> rx_vlan add (vlan_id|all) (port_id)
740 VLAN filter must be set on that port. VLAN ID < 4096.
741 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries
742 in VFTA table. This is important if enabling all vlan_ids.
747 Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID::
749 testpmd> rx_vlan rm (vlan_id|all) (port_id)
754 Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID::
756 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask)
761 Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID::
763 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask)
768 Add a tunnel filter on a port::
770 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \
771 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
772 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
774 The available information categories are:
776 * ``vxlan``: Set tunnel type as VXLAN.
778 * ``nvgre``: Set tunnel type as NVGRE.
780 * ``ipingre``: Set tunnel type as IP-in-GRE.
782 * ``imac-ivlan``: Set filter type as Inner MAC and VLAN.
784 * ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID.
786 * ``imac-tenid``: Set filter type as Inner MAC and tenant ID.
788 * ``imac``: Set filter type as Inner MAC.
790 * ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID.
792 * ``oip``: Set filter type as Outer IP.
794 * ``iip``: Set filter type as Inner IP.
798 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \
799 192.168.2.2 0 ipingre oip 1 1
801 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP.
806 Remove a tunnel filter on a port::
808 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \
809 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
810 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
815 Add an UDP port for VXLAN packet filter on a port::
817 testpmd> rx_vxlan_port add (udp_port) (port_id)
822 Remove an UDP port for VXLAN packet filter on a port::
824 testpmd> rx_vxlan_port rm (udp_port) (port_id)
829 Set hardware insertion of VLAN IDs in packets sent on a port::
831 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer]
833 For example, set a single VLAN ID (5) insertion on port 0::
837 Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1::
845 Set port based hardware insertion of VLAN ID in packets sent on a port::
847 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off)
852 Disable hardware insertion of a VLAN header in packets sent on a port::
854 testpmd> tx_vlan reset (port_id)
859 Select hardware or software calculation of the checksum when
860 transmitting a packet using the ``csum`` forwarding engine::
862 testpmd> csum set (ip|udp|tcp|sctp|outer-ip|outer-udp) (hw|sw) (port_id)
866 * ``ip|udp|tcp|sctp`` always relate to the inner layer.
868 * ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized
869 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are
870 supported). See also the ``csum parse-tunnel`` command.
872 * ``outer-udp`` relates to the outer UDP layer in the case where the packet is recognized
873 as a tunnel packet by the forwarding engine (vxlan, vxlan-gpe are
874 supported). See also the ``csum parse-tunnel`` command.
878 Check the NIC Datasheet for hardware limits.
883 Set RSS queue region span on a port::
885 testpmd> set port (port_id) queue-region region_id (value) \
886 queue_start_index (value) queue_num (value)
888 Set flowtype mapping on a RSS queue region on a port::
890 testpmd> set port (port_id) queue-region region_id (value) flowtype (value)
894 * For the flowtype(pctype) of packet,the specific index for each type has
895 been defined in file i40e_type.h as enum i40e_filter_pctype.
897 Set user priority mapping on a RSS queue region on a port::
899 testpmd> set port (port_id) queue-region UP (value) region_id (value)
901 Flush all queue region related configuration on a port::
903 testpmd> set port (port_id) queue-region flush (on|off)
907 * "on"is just an enable function which server for other configuration,
908 it is for all configuration about queue region from up layer,
909 at first will only keep in DPDK softwarestored in driver,
910 only after "flush on", it commit all configuration to HW.
911 "off" is just clean all configuration about queue region just now,
912 and restore all to DPDK i40e driver default config when start up.
914 Show all queue region related configuration info on a port::
916 testpmd> show port (port_id) queue-region
920 Queue region only support on PF by now, so these command is
921 only for configuration of queue region on PF port.
926 Define how tunneled packets should be handled by the csum forward
929 testpmd> csum parse-tunnel (on|off) (tx_port_id)
931 If enabled, the csum forward engine will try to recognize supported
932 tunnel headers (vxlan, gre, ipip).
934 If disabled, treat tunnel packets as non-tunneled packets (a inner
935 header is handled as a packet payload).
939 The port argument is the TX port like in the ``csum set`` command.
943 Consider a packet in packet like the following::
945 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in
947 * If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
948 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the
949 ``outer-ip|outer-udp`` parameter relates to the outer headers (here ``ipv4_out`` and ``udp_out``).
951 * If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
952 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``.
957 Display tx checksum offload configuration::
959 testpmd> csum show (port_id)
964 Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine::
966 testpmd> tso set (segsize) (port_id)
970 Check the NIC datasheet for hardware limits.
975 Display the status of TCP Segmentation Offload::
977 testpmd> tso show (port_id)
982 Enable or disable GRO in ``csum`` forwarding engine::
984 testpmd> set port <port_id> gro on|off
986 If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4
987 packets received from the given port.
989 If disabled, packets received from the given port won't be performed
990 GRO. By default, GRO is disabled for all ports.
994 When enable GRO for a port, TCP/IPv4 packets received from the port
995 will be performed GRO. After GRO, all merged packets have bad
996 checksums, since the GRO library doesn't re-calculate checksums for
997 the merged packets. Therefore, if users want the merged packets to
998 have correct checksums, please select HW IP checksum calculation and
999 HW TCP checksum calculation for the port which the merged packets are
1005 Display GRO configuration for a given port::
1007 testpmd> show port <port_id> gro
1012 Set the cycle to flush the GROed packets from reassembly tables::
1014 testpmd> set gro flush <cycles>
1016 When enable GRO, the csum forwarding engine performs GRO on received
1017 packets, and the GROed packets are stored in reassembly tables. Users
1018 can use this command to determine when the GROed packets are flushed
1019 from the reassembly tables.
1021 The ``cycles`` is measured in GRO operation times. The csum forwarding
1022 engine flushes the GROed packets from the tables every ``cycles`` GRO
1025 By default, the value of ``cycles`` is 1, which means flush GROed packets
1026 from the reassembly tables as soon as one GRO operation finishes. The value
1027 of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``.
1029 Please note that the large value of ``cycles`` may cause the poor TCP/IP
1030 stack performance. Because the GROed packets are delayed to arrive the
1031 stack, thus causing more duplicated ACKs and TCP retransmissions.
1036 Toggle per-port GSO support in ``csum`` forwarding engine::
1038 testpmd> set port <port_id> gso on|off
1040 If enabled, the csum forwarding engine will perform GSO on supported IPv4
1041 packets, transmitted on the given port.
1043 If disabled, packets transmitted on the given port will not undergo GSO.
1044 By default, GSO is disabled for all ports.
1048 When GSO is enabled on a port, supported IPv4 packets transmitted on that
1049 port undergo GSO. Afterwards, the segmented packets are represented by
1050 multi-segment mbufs; however, the csum forwarding engine doesn't calculation
1051 of checksums for GSO'd segments in SW. As a result, if users want correct
1052 checksums in GSO segments, they should enable HW checksum calculation for
1055 For example, HW checksum calculation for VxLAN GSO'd packets may be enabled
1056 by setting the following options in the csum forwarding engine:
1058 testpmd> csum set outer_ip hw <port_id>
1060 testpmd> csum set ip hw <port_id>
1062 testpmd> csum set tcp hw <port_id>
1064 UDP GSO is the same as IP fragmentation, which treats the UDP header
1065 as the payload and does not modify it during segmentation. That is,
1066 after UDP GSO, only the first output fragment has the original UDP
1067 header. Therefore, users need to enable HW IP checksum calculation
1068 and SW UDP checksum calculation for GSO-enabled ports, if they want
1069 correct checksums for UDP/IPv4 packets.
1074 Set the maximum GSO segment size (measured in bytes), which includes the
1075 packet header and the packet payload for GSO-enabled ports (global)::
1077 testpmd> set gso segsz <length>
1082 Display the status of Generic Segmentation Offload for a given port::
1084 testpmd> show port <port_id> gso
1089 Add an alternative MAC address to a port::
1091 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX)
1096 Remove a MAC address from a port::
1098 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX)
1100 mac_addr add (for VF)
1101 ~~~~~~~~~~~~~~~~~~~~~
1103 Add an alternative MAC address for a VF to a port::
1105 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX)
1110 Set the default MAC address for a port::
1112 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX)
1114 mac_addr set (for VF)
1115 ~~~~~~~~~~~~~~~~~~~~~
1117 Set the MAC address for a VF from the PF::
1119 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX)
1124 Set the forwarding peer address for certain port::
1126 testpmd> set eth-peer (port_id) (perr_addr)
1128 This is equivalent to the ``--eth-peer`` command-line option.
1133 Set the unicast hash filter(s) on/off for a port::
1135 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
1140 Set the promiscuous mode on for a port or for all ports.
1141 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1143 testpmd> set promisc (port_id|all) (on|off)
1148 Set the allmulti mode for a port or for all ports::
1150 testpmd> set allmulti (port_id|all) (on|off)
1152 Same as the ifconfig (8) option. Controls how multicast packets are handled.
1154 set promisc (for VF)
1155 ~~~~~~~~~~~~~~~~~~~~
1157 Set the unicast promiscuous mode for a VF from PF.
1158 It's supported by Intel i40e NICs now.
1159 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1161 testpmd> set vf promisc (port_id) (vf_id) (on|off)
1163 set allmulticast (for VF)
1164 ~~~~~~~~~~~~~~~~~~~~~~~~~
1166 Set the multicast promiscuous mode for a VF from PF.
1167 It's supported by Intel i40e NICs now.
1168 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1170 testpmd> set vf allmulti (port_id) (vf_id) (on|off)
1172 set tx max bandwidth (for VF)
1173 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1175 Set TX max absolute bandwidth (Mbps) for a VF from PF::
1177 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth)
1179 set tc tx min bandwidth (for VF)
1180 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1182 Set all TCs' TX min relative bandwidth (%) for a VF from PF::
1184 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...)
1186 set tc tx max bandwidth (for VF)
1187 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1189 Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF::
1191 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth)
1193 set tc strict link priority mode
1194 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1196 Set some TCs' strict link priority mode on a physical port::
1198 testpmd> set tx strict-link-priority (port_id) (tc_bitmap)
1200 set tc tx min bandwidth
1201 ~~~~~~~~~~~~~~~~~~~~~~~
1203 Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs::
1205 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...)
1210 Set the link flow control parameter on a port::
1212 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1213 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
1214 autoneg (on|off) (port_id)
1218 * ``high_water`` (integer): High threshold value to trigger XOFF.
1220 * ``low_water`` (integer): Low threshold value to trigger XON.
1222 * ``pause_time`` (integer): Pause quota in the Pause frame.
1224 * ``send_xon`` (0/1): Send XON frame.
1226 * ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
1228 * ``autoneg``: Change the auto-negotiation parameter.
1233 Set the priority flow control parameter on a port::
1235 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1236 (pause_time) (priority) (port_id)
1240 * ``high_water`` (integer): High threshold value.
1242 * ``low_water`` (integer): Low threshold value.
1244 * ``pause_time`` (integer): Pause quota in the Pause frame.
1246 * ``priority`` (0-7): VLAN User Priority.
1251 Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
1253 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
1255 For example, to set rx queue 2 on port 0 to mapping 5::
1257 testpmd>set stat_qmap rx 0 2 5
1259 set xstats-hide-zero
1260 ~~~~~~~~~~~~~~~~~~~~
1262 Set the option to hide zero values for xstats display::
1264 testpmd> set xstats-hide-zero on|off
1268 By default, the zero values are displayed for xstats.
1270 set port - rx/tx (for VF)
1271 ~~~~~~~~~~~~~~~~~~~~~~~~~
1273 Set VF receive/transmit from a port::
1275 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
1277 set port - mac address filter (for VF)
1278 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1280 Add/Remove unicast or multicast MAC addr filter for a VF::
1282 testpmd> set port (port_id) vf (vf_id) (mac_addr) \
1283 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
1285 set port - rx mode(for VF)
1286 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1288 Set the VF receive mode of a port::
1290 testpmd> set port (port_id) vf (vf_id) \
1291 rxmode (AUPE|ROPE|BAM|MPE) (on|off)
1293 The available receive modes are:
1295 * ``AUPE``: Accepts untagged VLAN.
1297 * ``ROPE``: Accepts unicast hash.
1299 * ``BAM``: Accepts broadcast packets.
1301 * ``MPE``: Accepts all multicast packets.
1303 set port - tx_rate (for Queue)
1304 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1306 Set TX rate limitation for a queue on a port::
1308 testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1310 set port - tx_rate (for VF)
1311 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
1313 Set TX rate limitation for queues in VF on a port::
1315 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1317 set port - mirror rule
1318 ~~~~~~~~~~~~~~~~~~~~~~
1320 Set pool or vlan type mirror rule for a port::
1322 testpmd> set port (port_id) mirror-rule (rule_id) \
1323 (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1324 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1326 Set link mirror rule for a port::
1328 testpmd> set port (port_id) mirror-rule (rule_id) \
1329 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1331 For example to enable mirror traffic with vlan 0,1 to pool 0::
1333 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1335 reset port - mirror rule
1336 ~~~~~~~~~~~~~~~~~~~~~~~~
1338 Reset a mirror rule for a port::
1340 testpmd> reset port (port_id) mirror-rule (rule_id)
1345 Set the flush on RX streams before forwarding.
1346 The default is flush ``on``.
1347 Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1349 testpmd> set flush_rx off
1354 Set the bypass mode for the lowest port on bypass enabled NIC::
1356 testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1361 Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1363 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1364 mode (normal|bypass|isolate) (port_id)
1368 * ``timeout``: Enable bypass after watchdog timeout.
1370 * ``os_on``: Enable bypass when OS/board is powered on.
1372 * ``os_off``: Enable bypass when OS/board is powered off.
1374 * ``power_on``: Enable bypass when power supply is turned on.
1376 * ``power_off``: Enable bypass when power supply is turned off.
1382 Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1384 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1389 Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1391 testpmd> show bypass config (port_id)
1396 Set link up for a port::
1398 testpmd> set link-up port (port id)
1403 Set link down for a port::
1405 testpmd> set link-down port (port id)
1410 Enable E-tag insertion for a VF on a port::
1412 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1414 Disable E-tag insertion for a VF on a port::
1416 testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1418 Enable/disable E-tag stripping on a port::
1420 testpmd> E-tag set stripping (on|off) port (port_id)
1422 Enable/disable E-tag based forwarding on a port::
1424 testpmd> E-tag set forwarding (on|off) port (port_id)
1426 Add an E-tag forwarding filter on a port::
1428 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1430 Delete an E-tag forwarding filter on a port::
1431 testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1436 Load a dynamic device personalization (DDP) profile and store backup profile::
1438 testpmd> ddp add (port_id) (profile_path[,backup_profile_path])
1443 Delete a dynamic device personalization profile and restore backup profile::
1445 testpmd> ddp del (port_id) (backup_profile_path)
1450 List all items from the ptype mapping table::
1452 testpmd> ptype mapping get (port_id) (valid_only)
1456 * ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0).
1458 Replace a specific or a group of software defined ptype with a new one::
1460 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type)
1464 * ``target``: A specific software ptype or a mask to represent a group of software ptypes.
1466 * ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1).
1468 * ``pkt_type``: The new software ptype to replace the old ones.
1470 Update hardware defined ptype to software defined packet type mapping table::
1472 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype)
1476 * ``hw_ptype``: hardware ptype as the index of the ptype mapping table.
1478 * ``sw_ptype``: software ptype as the value of the ptype mapping table.
1480 Reset ptype mapping table::
1482 testpmd> ptype mapping reset (port_id)
1484 config per port Rx offloading
1485 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1487 Enable or disable a per port Rx offloading on all Rx queues of a port::
1489 testpmd> port config (port_id) rx_offload (offloading) on|off
1491 * ``offloading``: can be any of these offloading capability:
1492 vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro,
1493 qinq_strip, outer_ipv4_cksum, macsec_strip,
1494 header_split, vlan_filter, vlan_extend, jumbo_frame,
1495 crc_strip, scatter, timestamp, security, keep_crc
1497 This command should be run when the port is stopped, or else it will fail.
1499 config per queue Rx offloading
1500 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1502 Enable or disable a per queue Rx offloading only on a specific Rx queue::
1504 testpmd> port (port_id) rxq (queue_id) rx_offload (offloading) on|off
1506 * ``offloading``: can be any of these offloading capability:
1507 vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro,
1508 qinq_strip, outer_ipv4_cksum, macsec_strip,
1509 header_split, vlan_filter, vlan_extend, jumbo_frame,
1510 crc_strip, scatter, timestamp, security, keep_crc
1512 This command should be run when the port is stopped, or else it will fail.
1514 config per port Tx offloading
1515 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1517 Enable or disable a per port Tx offloading on all Tx queues of a port::
1519 testpmd> port config (port_id) tx_offload (offloading) on|off
1521 * ``offloading``: can be any of these offloading capability:
1522 vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum,
1523 sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum,
1524 qinq_insert, vxlan_tnl_tso, gre_tnl_tso,
1525 ipip_tnl_tso, geneve_tnl_tso, macsec_insert,
1526 mt_lockfree, multi_segs, mbuf_fast_free, security
1528 This command should be run when the port is stopped, or else it will fail.
1530 config per queue Tx offloading
1531 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1533 Enable or disable a per queue Tx offloading only on a specific Tx queue::
1535 testpmd> port (port_id) txq (queue_id) tx_offload (offloading) on|off
1537 * ``offloading``: can be any of these offloading capability:
1538 vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum,
1539 sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum,
1540 qinq_insert, vxlan_tnl_tso, gre_tnl_tso,
1541 ipip_tnl_tso, geneve_tnl_tso, macsec_insert,
1542 mt_lockfree, multi_segs, mbuf_fast_free, security
1544 This command should be run when the port is stopped, or else it will fail.
1546 Config VXLAN Encap outer layers
1547 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1549 Configure the outer layer to encapsulate a packet inside a VXLAN tunnel::
1551 set vxlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \
1552 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) eth-src (eth-src) \
1555 set vxlan-with-vlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \
1556 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) vlan-tci (vlan-tci) \
1557 eth-src (eth-src) eth-dst (eth-dst)
1559 Those command will set an internal configuration inside testpmd, any following
1560 flow rule using the action vxlan_encap will use the last configuration set.
1561 To have a different encapsulation header, one of those commands must be called
1562 before the flow rule creation.
1564 Config NVGRE Encap outer layers
1565 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1567 Configure the outer layer to encapsulate a packet inside a NVGRE tunnel::
1569 set nvgre ip-version (ipv4|ipv6) tni (tni) ip-src (ip-src) ip-dst (ip-dst) \
1570 eth-src (eth-src) eth-dst (eth-dst)
1571 set nvgre-with-vlan ip-version (ipv4|ipv6) tni (tni) ip-src (ip-src) \
1572 ip-dst (ip-dst) vlan-tci (vlan-tci) eth-src (eth-src) eth-dst (eth-dst)
1574 Those command will set an internal configuration inside testpmd, any following
1575 flow rule using the action nvgre_encap will use the last configuration set.
1576 To have a different encapsulation header, one of those commands must be called
1577 before the flow rule creation.
1582 The following sections show functions for configuring ports.
1586 Port configuration changes only become active when forwarding is started/restarted.
1591 Attach a port specified by pci address or virtual device args::
1593 testpmd> port attach (identifier)
1595 To attach a new pci device, the device should be recognized by kernel first.
1596 Then it should be moved under DPDK management.
1597 Finally the port can be attached to testpmd.
1599 For example, to move a pci device using ixgbe under DPDK management:
1601 .. code-block:: console
1603 # Check the status of the available devices.
1604 ./usertools/dpdk-devbind.py --status
1606 Network devices using DPDK-compatible driver
1607 ============================================
1610 Network devices using kernel driver
1611 ===================================
1612 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1615 # Bind the device to igb_uio.
1616 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1619 # Recheck the status of the devices.
1620 ./usertools/dpdk-devbind.py --status
1621 Network devices using DPDK-compatible driver
1622 ============================================
1623 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1625 To attach a port created by virtual device, above steps are not needed.
1627 For example, to attach a port whose pci address is 0000:0a:00.0.
1629 .. code-block:: console
1631 testpmd> port attach 0000:0a:00.0
1632 Attaching a new port...
1633 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1634 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1635 EAL: PCI memory mapped at 0x7f83bfa00000
1636 EAL: PCI memory mapped at 0x7f83bfa80000
1637 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1638 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1639 Port 0 is attached. Now total ports is 1
1642 For example, to attach a port created by pcap PMD.
1644 .. code-block:: console
1646 testpmd> port attach net_pcap0
1647 Attaching a new port...
1648 PMD: Initializing pmd_pcap for net_pcap0
1649 PMD: Creating pcap-backed ethdev on numa socket 0
1650 Port 0 is attached. Now total ports is 1
1653 In this case, identifier is ``net_pcap0``.
1654 This identifier format is the same as ``--vdev`` format of DPDK applications.
1656 For example, to re-attach a bonded port which has been previously detached,
1657 the mode and slave parameters must be given.
1659 .. code-block:: console
1661 testpmd> port attach net_bond_0,mode=0,slave=1
1662 Attaching a new port...
1663 EAL: Initializing pmd_bond for net_bond_0
1664 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1665 Port 0 is attached. Now total ports is 1
1672 Detach a specific port::
1674 testpmd> port detach (port_id)
1676 Before detaching a port, the port should be stopped and closed.
1678 For example, to detach a pci device port 0.
1680 .. code-block:: console
1682 testpmd> port stop 0
1685 testpmd> port close 0
1689 testpmd> port detach 0
1691 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1692 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1693 EAL: PCI memory unmapped at 0x7f83bfa00000
1694 EAL: PCI memory unmapped at 0x7f83bfa80000
1698 For example, to detach a virtual device port 0.
1700 .. code-block:: console
1702 testpmd> port stop 0
1705 testpmd> port close 0
1709 testpmd> port detach 0
1711 PMD: Closing pcap ethdev on numa socket 0
1712 Port 'net_pcap0' is detached. Now total ports is 0
1715 To remove a pci device completely from the system, first detach the port from testpmd.
1716 Then the device should be moved under kernel management.
1717 Finally the device can be removed using kernel pci hotplug functionality.
1719 For example, to move a pci device under kernel management:
1721 .. code-block:: console
1723 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1725 ./usertools/dpdk-devbind.py --status
1727 Network devices using DPDK-compatible driver
1728 ============================================
1731 Network devices using kernel driver
1732 ===================================
1733 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1735 To remove a port created by a virtual device, above steps are not needed.
1740 Start all ports or a specific port::
1742 testpmd> port start (port_id|all)
1747 Stop all ports or a specific port::
1749 testpmd> port stop (port_id|all)
1754 Close all ports or a specific port::
1756 testpmd> port close (port_id|all)
1758 port config - queue ring size
1759 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1761 Configure a rx/tx queue ring size::
1763 testpmd> port (port_id) (rxq|txq) (queue_id) ring_size (value)
1765 Only take effect after command that (re-)start the port or command that setup specific queue.
1767 port start/stop queue
1768 ~~~~~~~~~~~~~~~~~~~~~
1770 Start/stop a rx/tx queue on a specific port::
1772 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1774 port config - queue deferred start
1775 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1777 Switch on/off deferred start of a specific port queue::
1779 testpmd> port (port_id) (rxq|txq) (queue_id) deferred_start (on|off)
1782 ~~~~~~~~~~~~~~~~~~~~~
1784 Setup a rx/tx queue on a specific port::
1786 testpmd> port (port_id) (rxq|txq) (queue_id) setup
1788 Only take effect when port is started.
1793 Set the speed and duplex mode for all ports or a specific port::
1795 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1796 duplex (half|full|auto)
1798 port config - queues/descriptors
1799 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1801 Set number of queues/descriptors for rxq, txq, rxd and txd::
1803 testpmd> port config all (rxq|txq|rxd|txd) (value)
1805 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1807 port config - max-pkt-len
1808 ~~~~~~~~~~~~~~~~~~~~~~~~~
1810 Set the maximum packet length::
1812 testpmd> port config all max-pkt-len (value)
1814 This is equivalent to the ``--max-pkt-len`` command-line option.
1816 port config - CRC Strip
1817 ~~~~~~~~~~~~~~~~~~~~~~~
1819 Set hardware CRC stripping on or off for all ports::
1821 testpmd> port config all crc-strip (on|off)
1823 CRC stripping is on by default.
1825 The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option.
1827 port config - scatter
1828 ~~~~~~~~~~~~~~~~~~~~~~~
1830 Set RX scatter mode on or off for all ports::
1832 testpmd> port config all scatter (on|off)
1834 RX scatter mode is off by default.
1836 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1838 port config - RX Checksum
1839 ~~~~~~~~~~~~~~~~~~~~~~~~~
1841 Set hardware RX checksum offload to on or off for all ports::
1843 testpmd> port config all rx-cksum (on|off)
1845 Checksum offload is off by default.
1847 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1852 Set hardware VLAN on or off for all ports::
1854 testpmd> port config all hw-vlan (on|off)
1856 Hardware VLAN is off by default.
1858 The ``on`` option is equivalent to the ``--enable-hw-vlan`` command-line option.
1860 port config - VLAN filter
1861 ~~~~~~~~~~~~~~~~~~~~~~~~~
1863 Set hardware VLAN filter on or off for all ports::
1865 testpmd> port config all hw-vlan-filter (on|off)
1867 Hardware VLAN filter is off by default.
1869 The ``on`` option is equivalent to the ``--enable-hw-vlan-filter`` command-line option.
1871 port config - VLAN strip
1872 ~~~~~~~~~~~~~~~~~~~~~~~~
1874 Set hardware VLAN strip on or off for all ports::
1876 testpmd> port config all hw-vlan-strip (on|off)
1878 Hardware VLAN strip is off by default.
1880 The ``on`` option is equivalent to the ``--enable-hw-vlan-strip`` command-line option.
1882 port config - VLAN extend
1883 ~~~~~~~~~~~~~~~~~~~~~~~~~
1885 Set hardware VLAN extend on or off for all ports::
1887 testpmd> port config all hw-vlan-extend (on|off)
1889 Hardware VLAN extend is off by default.
1891 The ``on`` option is equivalent to the ``--enable-hw-vlan-extend`` command-line option.
1893 port config - Drop Packets
1894 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1896 Set packet drop for packets with no descriptors on or off for all ports::
1898 testpmd> port config all drop-en (on|off)
1900 Packet dropping for packets with no descriptors is off by default.
1902 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1907 Set the RSS (Receive Side Scaling) mode on or off::
1909 testpmd> port config all rss (all|default|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1911 RSS is on by default.
1913 The ``all`` option is equivalent to ip|tcp|udp|sctp|ether.
1914 The ``default`` option enables all supported RSS types reported by device info.
1915 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1917 port config - RSS Reta
1918 ~~~~~~~~~~~~~~~~~~~~~~
1920 Set the RSS (Receive Side Scaling) redirection table::
1922 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1927 Set the DCB mode for an individual port::
1929 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1931 The traffic class should be 4 or 8.
1936 Set the number of packets per burst::
1938 testpmd> port config all burst (value)
1940 This is equivalent to the ``--burst`` command-line option.
1942 port config - Threshold
1943 ~~~~~~~~~~~~~~~~~~~~~~~
1945 Set thresholds for TX/RX queues::
1947 testpmd> port config all (threshold) (value)
1949 Where the threshold type can be:
1951 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1953 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1955 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1957 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1959 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1961 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1963 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1965 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1967 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1969 These threshold options are also available from the command-line.
1974 Set the value of ether-type for E-tag::
1976 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1978 Enable/disable the E-tag support::
1980 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1982 port config pctype mapping
1983 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1985 Reset pctype mapping table::
1987 testpmd> port config (port_id) pctype mapping reset
1989 Update hardware defined pctype to software defined flow type mapping table::
1991 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id)
1995 * ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table.
1997 * ``flow_type_id``: software flow type id as the index of the pctype mapping table.
1999 port config input set
2000 ~~~~~~~~~~~~~~~~~~~~~
2002 Config RSS/FDIR/FDIR flexible payload input set for some pctype::
2003 testpmd> port config (port_id) pctype (pctype_id) \
2004 (hash_inset|fdir_inset|fdir_flx_inset) \
2005 (get|set|clear) field (field_idx)
2007 Clear RSS/FDIR/FDIR flexible payload input set for some pctype::
2008 testpmd> port config (port_id) pctype (pctype_id) \
2009 (hash_inset|fdir_inset|fdir_flx_inset) clear all
2013 * ``pctype_id``: hardware packet classification types.
2014 * ``field_idx``: hardware field index.
2016 port config udp_tunnel_port
2017 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
2019 Add/remove UDP tunnel port for VXLAN/GENEVE tunneling protocols::
2020 testpmd> port config (port_id) udp_tunnel_port add|rm vxlan|geneve (udp_port)
2022 Link Bonding Functions
2023 ----------------------
2025 The Link Bonding functions make it possible to dynamically create and
2026 manage link bonding devices from within testpmd interactive prompt.
2028 create bonded device
2029 ~~~~~~~~~~~~~~~~~~~~
2031 Create a new bonding device::
2033 testpmd> create bonded device (mode) (socket)
2035 For example, to create a bonded device in mode 1 on socket 0::
2037 testpmd> create bonded device 1 0
2038 created new bonded device (port X)
2043 Adds Ethernet device to a Link Bonding device::
2045 testpmd> add bonding slave (slave id) (port id)
2047 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
2049 testpmd> add bonding slave 6 10
2052 remove bonding slave
2053 ~~~~~~~~~~~~~~~~~~~~
2055 Removes an Ethernet slave device from a Link Bonding device::
2057 testpmd> remove bonding slave (slave id) (port id)
2059 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
2061 testpmd> remove bonding slave 6 10
2066 Set the Link Bonding mode of a Link Bonding device::
2068 testpmd> set bonding mode (value) (port id)
2070 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
2072 testpmd> set bonding mode 3 10
2077 Set an Ethernet slave device as the primary device on a Link Bonding device::
2079 testpmd> set bonding primary (slave id) (port id)
2081 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
2083 testpmd> set bonding primary 6 10
2088 Set the MAC address of a Link Bonding device::
2090 testpmd> set bonding mac (port id) (mac)
2092 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
2094 testpmd> set bonding mac 10 00:00:00:00:00:01
2096 set bonding xmit_balance_policy
2097 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2099 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
2101 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
2103 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
2105 testpmd> set bonding xmit_balance_policy 10 l34
2108 set bonding mon_period
2109 ~~~~~~~~~~~~~~~~~~~~~~
2111 Set the link status monitoring polling period in milliseconds for a bonding device.
2113 This adds support for PMD slave devices which do not support link status interrupts.
2114 When the mon_period is set to a value greater than 0 then all PMD's which do not support
2115 link status ISR will be queried every polling interval to check if their link status has changed::
2117 testpmd> set bonding mon_period (port_id) (value)
2119 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
2121 testpmd> set bonding mon_period 5 150
2124 set bonding lacp dedicated_queue
2125 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2127 Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic
2128 when in mode 4 (link-aggregration-802.3ad)::
2130 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable)
2133 set bonding agg_mode
2134 ~~~~~~~~~~~~~~~~~~~~
2136 Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad)::
2138 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable)
2144 Show the current configuration of a Link Bonding device::
2146 testpmd> show bonding config (port id)
2149 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
2150 in balance mode with a transmission policy of layer 2+3::
2152 testpmd> show bonding config 9
2154 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
2156 Active Slaves (3): [1 3 4]
2163 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
2164 This is mainly useful for debugging purposes.
2165 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
2166 and fields that can be accessed.
2171 Display the value of a port register::
2173 testpmd> read reg (port_id) (address)
2175 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
2177 testpmd> read reg 0 0xEE00
2178 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
2183 Display a port register bit field::
2185 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
2187 For example, reading the lowest two bits from the register in the example above::
2189 testpmd> read regfield 0 0xEE00 0 1
2190 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
2195 Display a single port register bit::
2197 testpmd> read regbit (port_id) (address) (bit_x)
2199 For example, reading the lowest bit from the register in the example above::
2201 testpmd> read regbit 0 0xEE00 0
2202 port 0 PCI register at offset 0xEE00: bit 0=1
2207 Set the value of a port register::
2209 testpmd> write reg (port_id) (address) (value)
2211 For example, to clear a register::
2213 testpmd> write reg 0 0xEE00 0x0
2214 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
2219 Set bit field of a port register::
2221 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
2223 For example, writing to the register cleared in the example above::
2225 testpmd> write regfield 0 0xEE00 0 1 2
2226 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
2231 Set single bit value of a port register::
2233 testpmd> write regbit (port_id) (address) (bit_x) (value)
2235 For example, to set the high bit in the register from the example above::
2237 testpmd> write regbit 0 0xEE00 31 1
2238 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
2240 Traffic Metering and Policing
2241 -----------------------------
2243 The following section shows functions for configuring traffic metering and
2244 policing on the ethernet device through the use of generic ethdev API.
2246 show port traffic management capability
2247 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2249 Show traffic metering and policing capability of the port::
2251 testpmd> show port meter cap (port_id)
2253 add port meter profile (srTCM rfc2967)
2254 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2256 Add meter profile (srTCM rfc2697) to the ethernet device::
2258 testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \
2263 * ``profile_id``: ID for the meter profile.
2264 * ``cir``: Committed Information Rate (CIR) (bytes/second).
2265 * ``cbs``: Committed Burst Size (CBS) (bytes).
2266 * ``ebs``: Excess Burst Size (EBS) (bytes).
2268 add port meter profile (trTCM rfc2968)
2269 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2271 Add meter profile (srTCM rfc2698) to the ethernet device::
2273 testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \
2274 (cir) (pir) (cbs) (pbs)
2278 * ``profile_id``: ID for the meter profile.
2279 * ``cir``: Committed information rate (bytes/second).
2280 * ``pir``: Peak information rate (bytes/second).
2281 * ``cbs``: Committed burst size (bytes).
2282 * ``pbs``: Peak burst size (bytes).
2284 add port meter profile (trTCM rfc4115)
2285 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2287 Add meter profile (trTCM rfc4115) to the ethernet device::
2289 testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \
2290 (cir) (eir) (cbs) (ebs)
2294 * ``profile_id``: ID for the meter profile.
2295 * ``cir``: Committed information rate (bytes/second).
2296 * ``eir``: Excess information rate (bytes/second).
2297 * ``cbs``: Committed burst size (bytes).
2298 * ``ebs``: Excess burst size (bytes).
2300 delete port meter profile
2301 ~~~~~~~~~~~~~~~~~~~~~~~~~
2303 Delete meter profile from the ethernet device::
2305 testpmd> del port meter profile (port_id) (profile_id)
2310 Create new meter object for the ethernet device::
2312 testpmd> create port meter (port_id) (mtr_id) (profile_id) \
2313 (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \
2314 (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\
2319 * ``mtr_id``: meter object ID.
2320 * ``profile_id``: ID for the meter profile.
2321 * ``meter_enable``: When this parameter has a non-zero value, the meter object
2322 gets enabled at the time of creation, otherwise remains disabled.
2323 * ``g_action``: Policer action for the packet with green color.
2324 * ``y_action``: Policer action for the packet with yellow color.
2325 * ``r_action``: Policer action for the packet with red color.
2326 * ``stats_mask``: Mask of statistics counter types to be enabled for the
2328 * ``shared``: When this parameter has a non-zero value, the meter object is
2329 shared by multiple flows. Otherwise, meter object is used by single flow.
2330 * ``use_pre_meter_color``: When this parameter has a non-zero value, the
2331 input color for the current meter object is determined by the latest meter
2332 object in the same flow. Otherwise, the current meter object uses the
2333 *dscp_table* to determine the input color.
2334 * ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input
2335 color, 0 <= x <= 63.
2340 Enable meter for the ethernet device::
2342 testpmd> enable port meter (port_id) (mtr_id)
2347 Disable meter for the ethernet device::
2349 testpmd> disable port meter (port_id) (mtr_id)
2354 Delete meter for the ethernet device::
2356 testpmd> del port meter (port_id) (mtr_id)
2358 Set port meter profile
2359 ~~~~~~~~~~~~~~~~~~~~~~
2361 Set meter profile for the ethernet device::
2363 testpmd> set port meter profile (port_id) (mtr_id) (profile_id)
2365 set port meter dscp table
2366 ~~~~~~~~~~~~~~~~~~~~~~~~~
2368 Set meter dscp table for the ethernet device::
2370 testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \
2371 (dscp_tbl_entry1)...(dscp_tbl_entry63)]
2373 set port meter policer action
2374 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2376 Set meter policer action for the ethernet device::
2378 testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \
2379 (action0) [(action1) (action1)]
2383 * ``action_mask``: Bit mask indicating which policer actions need to be
2384 updated. One or more policer actions can be updated in a single function
2385 invocation. To update the policer action associated with color C, bit
2386 (1 << C) needs to be set in *action_mask* and element at position C
2387 in the *actions* array needs to be valid.
2388 * ``actionx``: Policer action for the color x,
2389 RTE_MTR_GREEN <= x < RTE_MTR_COLORS
2391 set port meter stats mask
2392 ~~~~~~~~~~~~~~~~~~~~~~~~~
2394 Set meter stats mask for the ethernet device::
2396 testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask)
2400 * ``stats_mask``: Bit mask indicating statistics counter types to be enabled.
2402 show port meter stats
2403 ~~~~~~~~~~~~~~~~~~~~~
2405 Show meter stats of the ethernet device::
2407 testpmd> show port meter stats (port_id) (mtr_id) (clear)
2411 * ``clear``: Flag that indicates whether the statistics counters should
2412 be cleared (i.e. set to zero) immediately after they have been read or not.
2417 The following section shows functions for configuring traffic management on
2418 on the ethernet device through the use of generic TM API.
2420 show port traffic management capability
2421 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2423 Show traffic management capability of the port::
2425 testpmd> show port tm cap (port_id)
2427 show port traffic management capability (hierarchy level)
2428 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2430 Show traffic management hierarchy level capability of the port::
2432 testpmd> show port tm level cap (port_id) (level_id)
2434 show port traffic management capability (hierarchy node level)
2435 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2437 Show the traffic management hierarchy node capability of the port::
2439 testpmd> show port tm node cap (port_id) (node_id)
2441 show port traffic management hierarchy node type
2442 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2444 Show the port traffic management hierarchy node type::
2446 testpmd> show port tm node type (port_id) (node_id)
2448 show port traffic management hierarchy node stats
2449 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2451 Show the port traffic management hierarchy node statistics::
2453 testpmd> show port tm node stats (port_id) (node_id) (clear)
2457 * ``clear``: When this parameter has a non-zero value, the statistics counters
2458 are cleared (i.e. set to zero) immediately after they have been read,
2459 otherwise the statistics counters are left untouched.
2461 Add port traffic management private shaper profile
2462 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2464 Add the port traffic management private shaper profile::
2466 testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \
2467 (tb_rate) (tb_size) (packet_length_adjust)
2471 * ``shaper_profile id``: Shaper profile ID for the new profile.
2472 * ``tb_rate``: Token bucket rate (bytes per second).
2473 * ``tb_size``: Token bucket size (bytes).
2474 * ``packet_length_adjust``: The value (bytes) to be added to the length of
2475 each packet for the purpose of shaping. This parameter value can be used to
2476 correct the packet length with the framing overhead bytes that are consumed
2479 Delete port traffic management private shaper profile
2480 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2482 Delete the port traffic management private shaper::
2484 testpmd> del port tm node shaper profile (port_id) (shaper_profile_id)
2488 * ``shaper_profile id``: Shaper profile ID that needs to be deleted.
2490 Add port traffic management shared shaper
2491 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2493 Create the port traffic management shared shaper::
2495 testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \
2500 * ``shared_shaper_id``: Shared shaper ID to be created.
2501 * ``shaper_profile id``: Shaper profile ID for shared shaper.
2503 Set port traffic management shared shaper
2504 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2506 Update the port traffic management shared shaper::
2508 testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \
2513 * ``shared_shaper_id``: Shared shaper ID to be update.
2514 * ``shaper_profile id``: Shaper profile ID for shared shaper.
2516 Delete port traffic management shared shaper
2517 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2519 Delete the port traffic management shared shaper::
2521 testpmd> del port tm node shared shaper (port_id) (shared_shaper_id)
2525 * ``shared_shaper_id``: Shared shaper ID to be deleted.
2527 Set port traffic management hiearchy node private shaper
2528 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2530 set the port traffic management hierarchy node private shaper::
2532 testpmd> set port tm node shaper profile (port_id) (node_id) \
2537 * ``shaper_profile id``: Private shaper profile ID to be enabled on the
2540 Add port traffic management WRED profile
2541 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2543 Create a new WRED profile::
2545 testpmd> add port tm node wred profile (port_id) (wred_profile_id) \
2546 (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \
2547 (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \
2548 (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r)
2552 * ``wred_profile id``: Identifier for the newly create WRED profile
2553 * ``color_g``: Packet color (green)
2554 * ``min_th_g``: Minimum queue threshold for packet with green color
2555 * ``max_th_g``: Minimum queue threshold for packet with green color
2556 * ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp)
2557 * ``wq_log2_g``: Negated log2 of queue weight (wq)
2558 * ``color_y``: Packet color (yellow)
2559 * ``min_th_y``: Minimum queue threshold for packet with yellow color
2560 * ``max_th_y``: Minimum queue threshold for packet with yellow color
2561 * ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp)
2562 * ``wq_log2_y``: Negated log2 of queue weight (wq)
2563 * ``color_r``: Packet color (red)
2564 * ``min_th_r``: Minimum queue threshold for packet with yellow color
2565 * ``max_th_r``: Minimum queue threshold for packet with yellow color
2566 * ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp)
2567 * ``wq_log2_r``: Negated log2 of queue weight (wq)
2569 Delete port traffic management WRED profile
2570 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2572 Delete the WRED profile::
2574 testpmd> del port tm node wred profile (port_id) (wred_profile_id)
2576 Add port traffic management hierarchy nonleaf node
2577 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2579 Add nonleaf node to port traffic management hiearchy::
2581 testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \
2582 (priority) (weight) (level_id) (shaper_profile_id) \
2583 (n_sp_priorities) (stats_mask) (n_shared_shapers) \
2584 [(shared_shaper_0) (shared_shaper_1) ...] \
2588 * ``parent_node_id``: Node ID of the parent.
2589 * ``priority``: Node priority (highest node priority is zero). This is used by
2590 the SP algorithm running on the parent node for scheduling this node.
2591 * ``weight``: Node weight (lowest weight is one). The node weight is relative
2592 to the weight sum of all siblings that have the same priority. It is used by
2593 the WFQ algorithm running on the parent node for scheduling this node.
2594 * ``level_id``: Hiearchy level of the node.
2595 * ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2597 * ``n_sp_priorities``: Number of strict priorities.
2598 * ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2599 * ``n_shared_shapers``: Number of shared shapers.
2600 * ``shared_shaper_id``: Shared shaper id.
2602 Add port traffic management hierarchy leaf node
2603 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2605 Add leaf node to port traffic management hiearchy::
2607 testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \
2608 (priority) (weight) (level_id) (shaper_profile_id) \
2609 (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \
2610 [(shared_shaper_id) (shared_shaper_id) ...] \
2614 * ``parent_node_id``: Node ID of the parent.
2615 * ``priority``: Node priority (highest node priority is zero). This is used by
2616 the SP algorithm running on the parent node for scheduling this node.
2617 * ``weight``: Node weight (lowest weight is one). The node weight is relative
2618 to the weight sum of all siblings that have the same priority. It is used by
2619 the WFQ algorithm running on the parent node for scheduling this node.
2620 * ``level_id``: Hiearchy level of the node.
2621 * ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2623 * ``cman_mode``: Congestion management mode to be enabled for this node.
2624 * ``wred_profile_id``: WRED profile id to be enabled for this node.
2625 * ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2626 * ``n_shared_shapers``: Number of shared shapers.
2627 * ``shared_shaper_id``: Shared shaper id.
2629 Delete port traffic management hierarchy node
2630 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2632 Delete node from port traffic management hiearchy::
2634 testpmd> del port tm node (port_id) (node_id)
2636 Update port traffic management hierarchy parent node
2637 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2639 Update port traffic management hierarchy parent node::
2641 testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \
2644 This function can only be called after the hierarchy commit invocation. Its
2645 success depends on the port support for this operation, as advertised through
2646 the port capability set. This function is valid for all nodes of the traffic
2647 management hierarchy except root node.
2649 Suspend port traffic management hierarchy node
2650 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2652 testpmd> suspend port tm node (port_id) (node_id)
2654 Resume port traffic management hierarchy node
2655 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2657 testpmd> resume port tm node (port_id) (node_id)
2659 Commit port traffic management hierarchy
2660 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2662 Commit the traffic management hierarchy on the port::
2664 testpmd> port tm hierarchy commit (port_id) (clean_on_fail)
2668 * ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function
2669 call failure. On the other hand, hierarchy is preserved when this parameter
2672 Set port traffic management default hierarchy (softnic forwarding mode)
2673 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2675 set the traffic management default hierarchy on the port::
2677 testpmd> set port tm hierarchy default (port_id)
2682 This section details the available filter functions that are available.
2684 Note these functions interface the deprecated legacy filtering framework,
2685 superseded by *rte_flow*. See `Flow rules management`_.
2688 ~~~~~~~~~~~~~~~~~~~~
2690 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
2692 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
2693 ethertype (ether_type) (drop|fwd) queue (queue_id)
2695 The available information parameters are:
2697 * ``port_id``: The port which the Ethertype filter assigned on.
2699 * ``mac_addr``: Compare destination mac address.
2701 * ``mac_ignr``: Ignore destination mac address match.
2703 * ``mac_address``: Destination mac address to match.
2705 * ``ether_type``: The EtherType value want to match,
2706 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
2708 * ``queue_id``: The receive queue associated with this EtherType filter.
2709 It is meaningless when deleting or dropping.
2711 Example, to add/remove an ethertype filter rule::
2713 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
2714 ethertype 0x0806 fwd queue 3
2716 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
2717 ethertype 0x0806 fwd queue 3
2722 Add or delete a 2-tuple filter,
2723 which identifies packets by specific protocol and destination TCP/UDP port
2724 and forwards packets into one of the receive queues::
2726 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
2727 protocol (protocol_value) mask (mask_value) \
2728 tcp_flags (tcp_flags_value) priority (prio_value) \
2731 The available information parameters are:
2733 * ``port_id``: The port which the 2-tuple filter assigned on.
2735 * ``dst_port_value``: Destination port in L4.
2737 * ``protocol_value``: IP L4 protocol.
2739 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
2741 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
2743 * ``prio_value``: Priority of this filter.
2745 * ``queue_id``: The receive queue associated with this 2-tuple filter.
2747 Example, to add/remove an 2tuple filter rule::
2749 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
2750 tcp_flags 0x02 priority 3 queue 3
2752 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
2753 tcp_flags 0x02 priority 3 queue 3
2758 Add or delete a 5-tuple filter,
2759 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
2760 and routes packets into one of the receive queues::
2762 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
2763 (src_address) dst_port (dst_port_value) \
2764 src_port (src_port_value) protocol (protocol_value) \
2765 mask (mask_value) tcp_flags (tcp_flags_value) \
2766 priority (prio_value) queue (queue_id)
2768 The available information parameters are:
2770 * ``port_id``: The port which the 5-tuple filter assigned on.
2772 * ``dst_address``: Destination IP address.
2774 * ``src_address``: Source IP address.
2776 * ``dst_port_value``: TCP/UDP destination port.
2778 * ``src_port_value``: TCP/UDP source port.
2780 * ``protocol_value``: L4 protocol.
2782 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
2784 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
2786 * ``prio_value``: The priority of this filter.
2788 * ``queue_id``: The receive queue associated with this 5-tuple filter.
2790 Example, to add/remove an 5tuple filter rule::
2792 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2793 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2794 flags 0x0 priority 3 queue 3
2796 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2797 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2798 flags 0x0 priority 3 queue 3
2803 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
2805 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
2807 The available information parameters are:
2809 * ``port_id``: The port which the SYN filter assigned on.
2811 * ``high``: This SYN filter has higher priority than other filters.
2813 * ``low``: This SYN filter has lower priority than other filters.
2815 * ``queue_id``: The receive queue associated with this SYN filter
2819 testpmd> syn_filter 0 add priority high queue 3
2824 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
2825 and routed into one of the receive queues::
2827 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
2828 mask (mask_value) priority (prio_value) queue (queue_id)
2830 The available information parameters are:
2832 * ``port_id``: The port which the Flex filter is assigned on.
2834 * ``len_value``: Filter length in bytes, no greater than 128.
2836 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
2838 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
2840 * ``prio_value``: The priority of this filter.
2842 * ``queue_id``: The receive queue associated with this Flex filter.
2846 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
2847 mask 000C priority 3 queue 3
2849 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
2850 mask 000C priority 3 queue 3
2853 .. _testpmd_flow_director:
2855 flow_director_filter
2856 ~~~~~~~~~~~~~~~~~~~~
2858 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
2860 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
2861 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
2863 * Perfect match filters.
2864 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2865 The masked fields are for IP flow.
2867 * Signature filters.
2868 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
2870 * Perfect-mac-vlan match filters.
2871 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2872 The masked fields are for MAC VLAN flow.
2874 * Perfect-tunnel match filters.
2875 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2876 The masked fields are for tunnel flow.
2878 * Perfect-raw-flow-type match filters.
2879 The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet.
2880 The masked fields are specified by input sets.
2882 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
2883 per flow type and the flexible payload.
2885 The Flow Director can also mask out parts of all of these fields so that filters
2886 are only applied to certain fields or parts of the fields.
2888 Note that for raw flow type mode the source and destination fields in the
2889 raw packet buffer need to be presented in a reversed order with respect
2890 to the expected received packets.
2891 For example: IP source and destination addresses or TCP/UDP/SCTP
2892 source and destination ports
2894 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
2896 # Commands to add flow director filters of different flow types::
2898 flow_director_filter (port_id) mode IP (add|del|update) \
2899 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
2900 src (src_ip_address) dst (dst_ip_address) \
2901 tos (tos_value) proto (proto_value) ttl (ttl_value) \
2902 vlan (vlan_value) flexbytes (flexbytes_value) \
2903 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
2906 flow_director_filter (port_id) mode IP (add|del|update) \
2907 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
2908 src (src_ip_address) (src_port) \
2909 dst (dst_ip_address) (dst_port) \
2910 tos (tos_value) ttl (ttl_value) \
2911 vlan (vlan_value) flexbytes (flexbytes_value) \
2912 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
2915 flow_director_filter (port_id) mode IP (add|del|update) \
2916 flow (ipv4-sctp|ipv6-sctp) \
2917 src (src_ip_address) (src_port) \
2918 dst (dst_ip_address) (dst_port) \
2919 tos (tos_value) ttl (ttl_value) \
2920 tag (verification_tag) vlan (vlan_value) \
2921 flexbytes (flexbytes_value) (drop|fwd) \
2922 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
2924 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
2925 ether (ethertype) flexbytes (flexbytes_value) \
2926 (drop|fwd) pf|vf(vf_id) queue (queue_id)
2929 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
2930 mac (mac_address) vlan (vlan_value) \
2931 flexbytes (flexbytes_value) (drop|fwd) \
2932 queue (queue_id) fd_id (fd_id_value)
2934 flow_director_filter (port_id) mode Tunnel (add|del|update) \
2935 mac (mac_address) vlan (vlan_value) \
2936 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
2937 flexbytes (flexbytes_value) (drop|fwd) \
2938 queue (queue_id) fd_id (fd_id_value)
2940 flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \
2941 (drop|fwd) queue (queue_id) fd_id (fd_id_value) \
2942 packet (packet file name)
2944 For example, to add an ipv4-udp flow type filter::
2946 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
2947 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
2948 fwd pf queue 1 fd_id 1
2950 For example, add an ipv4-other flow type filter::
2952 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
2953 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
2954 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
2959 Flush all flow director filters on a device::
2961 testpmd> flush_flow_director (port_id)
2963 Example, to flush all flow director filter on port 0::
2965 testpmd> flush_flow_director 0
2970 Set flow director's input masks::
2972 flow_director_mask (port_id) mode IP vlan (vlan_value) \
2973 src_mask (ipv4_src) (ipv6_src) (src_port) \
2974 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2976 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2978 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2979 mac (mac_value) tunnel-type (tunnel_type_value) \
2980 tunnel-id (tunnel_id_value)
2982 Example, to set flow director mask on port 0::
2984 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2985 src_mask 255.255.255.255 \
2986 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2987 dst_mask 255.255.255.255 \
2988 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2990 flow_director_flex_mask
2991 ~~~~~~~~~~~~~~~~~~~~~~~
2993 set masks of flow director's flexible payload based on certain flow type::
2995 testpmd> flow_director_flex_mask (port_id) \
2996 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2997 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2998 l2_payload|all) (mask)
3000 Example, to set flow director's flex mask for all flow type on port 0::
3002 testpmd> flow_director_flex_mask 0 flow all \
3003 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
3006 flow_director_flex_payload
3007 ~~~~~~~~~~~~~~~~~~~~~~~~~~
3009 Configure flexible payload selection::
3011 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
3013 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
3015 testpmd> flow_director_flex_payload 0 l4 \
3016 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
3018 get_sym_hash_ena_per_port
3019 ~~~~~~~~~~~~~~~~~~~~~~~~~
3021 Get symmetric hash enable configuration per port::
3023 get_sym_hash_ena_per_port (port_id)
3025 For example, to get symmetric hash enable configuration of port 1::
3027 testpmd> get_sym_hash_ena_per_port 1
3029 set_sym_hash_ena_per_port
3030 ~~~~~~~~~~~~~~~~~~~~~~~~~
3032 Set symmetric hash enable configuration per port to enable or disable::
3034 set_sym_hash_ena_per_port (port_id) (enable|disable)
3036 For example, to set symmetric hash enable configuration of port 1 to enable::
3038 testpmd> set_sym_hash_ena_per_port 1 enable
3040 get_hash_global_config
3041 ~~~~~~~~~~~~~~~~~~~~~~
3043 Get the global configurations of hash filters::
3045 get_hash_global_config (port_id)
3047 For example, to get the global configurations of hash filters of port 1::
3049 testpmd> get_hash_global_config 1
3051 set_hash_global_config
3052 ~~~~~~~~~~~~~~~~~~~~~~
3054 Set the global configurations of hash filters::
3056 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
3057 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
3058 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \
3061 For example, to enable simple_xor for flow type of ipv6 on port 2::
3063 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
3068 Set the input set for hash::
3070 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
3071 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
3072 l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
3073 ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
3074 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
3075 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
3076 fld-8th|none) (select|add)
3078 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
3080 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
3085 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
3086 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
3088 Set the input set for flow director::
3090 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
3091 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
3092 l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
3093 ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
3094 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
3095 sctp-dst-port|sctp-veri-tag|none) (select|add)
3097 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
3099 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
3104 Set different GRE key length for input set::
3106 global_config (port_id) gre-key-len (number in bytes)
3108 For example to set GRE key length for input set to 4 bytes on port 0::
3110 testpmd> global_config 0 gre-key-len 4
3113 .. _testpmd_rte_flow:
3115 Flow rules management
3116 ---------------------
3118 Control of the generic flow API (*rte_flow*) is fully exposed through the
3119 ``flow`` command (validation, creation, destruction, queries and operation
3122 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
3123 features simultaneously may cause undefined side-effects and is therefore
3129 Because the ``flow`` command uses dynamic tokens to handle the large number
3130 of possible flow rules combinations, its behavior differs slightly from
3131 other commands, in particular:
3133 - Pressing *?* or the *<tab>* key displays contextual help for the current
3134 token, not that of the entire command.
3136 - Optional and repeated parameters are supported (provided they are listed
3137 in the contextual help).
3139 The first parameter stands for the operation mode. Possible operations and
3140 their general syntax are described below. They are covered in detail in the
3143 - Check whether a flow rule can be created::
3145 flow validate {port_id}
3146 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3147 pattern {item} [/ {item} [...]] / end
3148 actions {action} [/ {action} [...]] / end
3150 - Create a flow rule::
3152 flow create {port_id}
3153 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3154 pattern {item} [/ {item} [...]] / end
3155 actions {action} [/ {action} [...]] / end
3157 - Destroy specific flow rules::
3159 flow destroy {port_id} rule {rule_id} [...]
3161 - Destroy all flow rules::
3163 flow flush {port_id}
3165 - Query an existing flow rule::
3167 flow query {port_id} {rule_id} {action}
3169 - List existing flow rules sorted by priority, filtered by group
3172 flow list {port_id} [group {group_id}] [...]
3174 - Restrict ingress traffic to the defined flow rules::
3176 flow isolate {port_id} {boolean}
3178 Validating flow rules
3179 ~~~~~~~~~~~~~~~~~~~~~
3181 ``flow validate`` reports whether a flow rule would be accepted by the
3182 underlying device in its current state but stops short of creating it. It is
3183 bound to ``rte_flow_validate()``::
3185 flow validate {port_id}
3186 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3187 pattern {item} [/ {item} [...]] / end
3188 actions {action} [/ {action} [...]] / end
3190 If successful, it will show::
3194 Otherwise it will show an error message of the form::
3196 Caught error type [...] ([...]): [...]
3198 This command uses the same parameters as ``flow create``, their format is
3199 described in `Creating flow rules`_.
3201 Check whether redirecting any Ethernet packet received on port 0 to RX queue
3202 index 6 is supported::
3204 testpmd> flow validate 0 ingress pattern eth / end
3205 actions queue index 6 / end
3209 Port 0 does not support TCPv6 rules::
3211 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
3213 Caught error type 9 (specific pattern item): Invalid argument
3219 ``flow create`` validates and creates the specified flow rule. It is bound
3220 to ``rte_flow_create()``::
3222 flow create {port_id}
3223 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3224 pattern {item} [/ {item} [...]] / end
3225 actions {action} [/ {action} [...]] / end
3227 If successful, it will return a flow rule ID usable with other commands::
3229 Flow rule #[...] created
3231 Otherwise it will show an error message of the form::
3233 Caught error type [...] ([...]): [...]
3235 Parameters describe in the following order:
3237 - Attributes (*group*, *priority*, *ingress*, *egress*, *transfer* tokens).
3238 - A matching pattern, starting with the *pattern* token and terminated by an
3240 - Actions, starting with the *actions* token and terminated by an *end*
3243 These translate directly to *rte_flow* objects provided as-is to the
3244 underlying functions.
3246 The shortest valid definition only comprises mandatory tokens::
3248 testpmd> flow create 0 pattern end actions end
3250 Note that PMDs may refuse rules that essentially do nothing such as this
3253 **All unspecified object values are automatically initialized to 0.**
3258 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
3259 specified before the ``pattern`` token.
3261 - ``group {group id}``: priority group.
3262 - ``priority {level}``: priority level within group.
3263 - ``ingress``: rule applies to ingress traffic.
3264 - ``egress``: rule applies to egress traffic.
3265 - ``transfer``: apply rule directly to endpoints found in pattern.
3267 Each instance of an attribute specified several times overrides the previous
3268 value as shown below (group 4 is used)::
3270 testpmd> flow create 0 group 42 group 24 group 4 [...]
3272 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
3274 While not specifying a direction is an error, some rules may allow both
3277 Most rules affect RX therefore contain the ``ingress`` token::
3279 testpmd> flow create 0 ingress pattern [...]
3284 A matching pattern starts after the ``pattern`` token. It is made of pattern
3285 items and is terminated by a mandatory ``end`` item.
3287 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
3288 rte_flow_item_type``).
3290 The ``/`` token is used as a separator between pattern items as shown
3293 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
3295 Note that protocol items like these must be stacked from lowest to highest
3296 layer to make sense. For instance, the following rule is either invalid or
3297 unlikely to match any packet::
3299 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
3301 More information on these restrictions can be found in the *rte_flow*
3304 Several items support additional specification structures, for example
3305 ``ipv4`` allows specifying source and destination addresses as follows::
3307 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
3308 dst is 10.2.0.0 / end [...]
3310 This rule matches all IPv4 traffic with the specified properties.
3312 In this example, ``src`` and ``dst`` are field names of the underlying
3313 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
3314 in a similar fashion.
3316 The ``is`` token means that the subsequent value must be matched exactly,
3317 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
3318 accordingly. Possible assignment tokens are:
3320 - ``is``: match value perfectly (with full bit-mask).
3321 - ``spec``: match value according to configured bit-mask.
3322 - ``last``: specify upper bound to establish a range.
3323 - ``mask``: specify bit-mask with relevant bits set to one.
3324 - ``prefix``: generate bit-mask from a prefix length.
3326 These yield identical results::
3328 ipv4 src is 10.1.1.1
3332 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
3336 ipv4 src spec 10.1.1.1 src prefix 32
3340 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
3344 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
3346 Inclusive ranges can be defined with ``last``::
3348 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
3350 Note that ``mask`` affects both ``spec`` and ``last``::
3352 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
3353 # matches 10.1.0.0 to 10.2.255.255
3355 Properties can be modified multiple times::
3357 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
3361 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
3366 This section lists supported pattern items and their attributes, if any.
3368 - ``end``: end list of pattern items.
3370 - ``void``: no-op pattern item.
3372 - ``invert``: perform actions when pattern does not match.
3374 - ``any``: match any protocol for the current layer.
3376 - ``num {unsigned}``: number of layers covered.
3378 - ``pf``: match traffic from/to the physical function.
3380 - ``vf``: match traffic from/to a virtual function ID.
3382 - ``id {unsigned}``: VF ID.
3384 - ``phy_port``: match traffic from/to a specific physical port.
3386 - ``index {unsigned}``: physical port index.
3388 - ``port_id``: match traffic from/to a given DPDK port ID.
3390 - ``id {unsigned}``: DPDK port ID.
3392 - ``mark``: match value set in previously matched flow rule using the mark action.
3394 - ``id {unsigned}``: arbitrary integer value.
3396 - ``raw``: match an arbitrary byte string.
3398 - ``relative {boolean}``: look for pattern after the previous item.
3399 - ``search {boolean}``: search pattern from offset (see also limit).
3400 - ``offset {integer}``: absolute or relative offset for pattern.
3401 - ``limit {unsigned}``: search area limit for start of pattern.
3402 - ``pattern {string}``: byte string to look for.
3404 - ``eth``: match Ethernet header.
3406 - ``dst {MAC-48}``: destination MAC.
3407 - ``src {MAC-48}``: source MAC.
3408 - ``type {unsigned}``: EtherType or TPID.
3410 - ``vlan``: match 802.1Q/ad VLAN tag.
3412 - ``tci {unsigned}``: tag control information.
3413 - ``pcp {unsigned}``: priority code point.
3414 - ``dei {unsigned}``: drop eligible indicator.
3415 - ``vid {unsigned}``: VLAN identifier.
3416 - ``inner_type {unsigned}``: inner EtherType or TPID.
3418 - ``ipv4``: match IPv4 header.
3420 - ``tos {unsigned}``: type of service.
3421 - ``ttl {unsigned}``: time to live.
3422 - ``proto {unsigned}``: next protocol ID.
3423 - ``src {ipv4 address}``: source address.
3424 - ``dst {ipv4 address}``: destination address.
3426 - ``ipv6``: match IPv6 header.
3428 - ``tc {unsigned}``: traffic class.
3429 - ``flow {unsigned}``: flow label.
3430 - ``proto {unsigned}``: protocol (next header).
3431 - ``hop {unsigned}``: hop limit.
3432 - ``src {ipv6 address}``: source address.
3433 - ``dst {ipv6 address}``: destination address.
3435 - ``icmp``: match ICMP header.
3437 - ``type {unsigned}``: ICMP packet type.
3438 - ``code {unsigned}``: ICMP packet code.
3440 - ``udp``: match UDP header.
3442 - ``src {unsigned}``: UDP source port.
3443 - ``dst {unsigned}``: UDP destination port.
3445 - ``tcp``: match TCP header.
3447 - ``src {unsigned}``: TCP source port.
3448 - ``dst {unsigned}``: TCP destination port.
3450 - ``sctp``: match SCTP header.
3452 - ``src {unsigned}``: SCTP source port.
3453 - ``dst {unsigned}``: SCTP destination port.
3454 - ``tag {unsigned}``: validation tag.
3455 - ``cksum {unsigned}``: checksum.
3457 - ``vxlan``: match VXLAN header.
3459 - ``vni {unsigned}``: VXLAN identifier.
3461 - ``e_tag``: match IEEE 802.1BR E-Tag header.
3463 - ``grp_ecid_b {unsigned}``: GRP and E-CID base.
3465 - ``nvgre``: match NVGRE header.
3467 - ``tni {unsigned}``: virtual subnet ID.
3469 - ``mpls``: match MPLS header.
3471 - ``label {unsigned}``: MPLS label.
3473 - ``gre``: match GRE header.
3475 - ``protocol {unsigned}``: protocol type.
3477 - ``fuzzy``: fuzzy pattern match, expect faster than default.
3479 - ``thresh {unsigned}``: accuracy threshold.
3481 - ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header.
3483 - ``teid {unsigned}``: tunnel endpoint identifier.
3485 - ``geneve``: match GENEVE header.
3487 - ``vni {unsigned}``: virtual network identifier.
3488 - ``protocol {unsigned}``: protocol type.
3490 - ``vxlan-gpe``: match VXLAN-GPE header.
3492 - ``vni {unsigned}``: VXLAN-GPE identifier.
3494 - ``arp_eth_ipv4``: match ARP header for Ethernet/IPv4.
3496 - ``sha {MAC-48}``: sender hardware address.
3497 - ``spa {ipv4 address}``: sender IPv4 address.
3498 - ``tha {MAC-48}``: target hardware address.
3499 - ``tpa {ipv4 address}``: target IPv4 address.
3501 - ``ipv6_ext``: match presence of any IPv6 extension header.
3503 - ``next_hdr {unsigned}``: next header.
3505 - ``icmp6``: match any ICMPv6 header.
3507 - ``type {unsigned}``: ICMPv6 type.
3508 - ``code {unsigned}``: ICMPv6 code.
3510 - ``icmp6_nd_ns``: match ICMPv6 neighbor discovery solicitation.
3512 - ``target_addr {ipv6 address}``: target address.
3514 - ``icmp6_nd_na``: match ICMPv6 neighbor discovery advertisement.
3516 - ``target_addr {ipv6 address}``: target address.
3518 - ``icmp6_nd_opt``: match presence of any ICMPv6 neighbor discovery option.
3520 - ``type {unsigned}``: ND option type.
3522 - ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery source Ethernet
3523 link-layer address option.
3525 - ``sla {MAC-48}``: source Ethernet LLA.
3527 - ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery target Ethernet
3528 link-layer address option.
3530 - ``tla {MAC-48}``: target Ethernet LLA.
3535 A list of actions starts after the ``actions`` token in the same fashion as
3536 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
3537 terminated by a mandatory ``end`` action.
3539 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
3540 rte_flow_action_type``).
3542 Dropping all incoming UDPv4 packets can be expressed as follows::
3544 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3547 Several actions have configurable properties which must be specified when
3548 there is no valid default value. For example, ``queue`` requires a target
3551 This rule redirects incoming UDPv4 traffic to queue index 6::
3553 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3554 actions queue index 6 / end
3556 While this one could be rejected by PMDs (unspecified queue index)::
3558 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3561 As defined by *rte_flow*, the list is not ordered, all actions of a given
3562 rule are performed simultaneously. These are equivalent::
3564 queue index 6 / void / mark id 42 / end
3568 void / mark id 42 / queue index 6 / end
3570 All actions in a list should have different types, otherwise only the last
3571 action of a given type is taken into account::
3573 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
3577 drop / drop / drop / end # drop is performed only once
3581 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
3583 Considering they are performed simultaneously, opposite and overlapping
3584 actions can sometimes be combined when the end result is unambiguous::
3586 drop / queue index 6 / end # drop has no effect
3590 queue index 6 / rss queues 6 7 8 / end # queue has no effect
3594 drop / passthru / end # drop has no effect
3596 Note that PMDs may still refuse such combinations.
3601 This section lists supported actions and their attributes, if any.
3603 - ``end``: end list of actions.
3605 - ``void``: no-op action.
3607 - ``passthru``: let subsequent rule process matched packets.
3609 - ``jump``: redirect traffic to group on device.
3611 - ``group {unsigned}``: group to redirect to.
3613 - ``mark``: attach 32 bit value to packets.
3615 - ``id {unsigned}``: 32 bit value to return with packets.
3617 - ``flag``: flag packets.
3619 - ``queue``: assign packets to a given queue index.
3621 - ``index {unsigned}``: queue index to use.
3623 - ``drop``: drop packets (note: passthru has priority).
3625 - ``count``: enable counters for this rule.
3627 - ``rss``: spread packets among several queues.
3629 - ``func {hash function}``: RSS hash function to apply, allowed tokens are
3630 the same as `set_hash_global_config`_.
3632 - ``level {unsigned}``: encapsulation level for ``types``.
3634 - ``types [{RSS hash type} [...]] end``: specific RSS hash types, allowed
3635 tokens are the same as `set_hash_input_set`_, except that an empty list
3636 does not disable RSS but instead requests unspecified "best-effort"
3639 - ``key {string}``: RSS hash key, overrides ``key_len``.
3641 - ``key_len {unsigned}``: RSS hash key length in bytes, can be used in
3642 conjunction with ``key`` to pad or truncate it.
3644 - ``queues [{unsigned} [...]] end``: queue indices to use.
3646 - ``pf``: direct traffic to physical function.
3648 - ``vf``: direct traffic to a virtual function ID.
3650 - ``original {boolean}``: use original VF ID if possible.
3651 - ``id {unsigned}``: VF ID.
3653 - ``phy_port``: direct packets to physical port index.
3655 - ``original {boolean}``: use original port index if possible.
3656 - ``index {unsigned}``: physical port index.
3658 - ``port_id``: direct matching traffic to a given DPDK port ID.
3660 - ``original {boolean}``: use original DPDK port ID if possible.
3661 - ``id {unsigned}``: DPDK port ID.
3663 - ``of_set_mpls_ttl``: OpenFlow's ``OFPAT_SET_MPLS_TTL``.
3665 - ``mpls_ttl``: MPLS TTL.
3667 - ``of_dec_mpls_ttl``: OpenFlow's ``OFPAT_DEC_MPLS_TTL``.
3669 - ``of_set_nw_ttl``: OpenFlow's ``OFPAT_SET_NW_TTL``.
3671 - ``nw_ttl``: IP TTL.
3673 - ``of_dec_nw_ttl``: OpenFlow's ``OFPAT_DEC_NW_TTL``.
3675 - ``of_copy_ttl_out``: OpenFlow's ``OFPAT_COPY_TTL_OUT``.
3677 - ``of_copy_ttl_in``: OpenFlow's ``OFPAT_COPY_TTL_IN``.
3679 - ``of_pop_vlan``: OpenFlow's ``OFPAT_POP_VLAN``.
3681 - ``of_push_vlan``: OpenFlow's ``OFPAT_PUSH_VLAN``.
3683 - ``ethertype``: Ethertype.
3685 - ``of_set_vlan_vid``: OpenFlow's ``OFPAT_SET_VLAN_VID``.
3687 - ``vlan_vid``: VLAN id.
3689 - ``of_set_vlan_pcp``: OpenFlow's ``OFPAT_SET_VLAN_PCP``.
3691 - ``vlan_pcp``: VLAN priority.
3693 - ``of_pop_mpls``: OpenFlow's ``OFPAT_POP_MPLS``.
3695 - ``ethertype``: Ethertype.
3697 - ``of_push_mpls``: OpenFlow's ``OFPAT_PUSH_MPLS``.
3699 - ``ethertype``: Ethertype.
3701 - ``vxlan_encap``: Performs a VXLAN encapsulation, outer layer configuration
3702 is done through `Config VXLAN Encap outer layers`_.
3704 - ``vxlan_decap``: Performs a decapsulation action by stripping all headers of
3705 the VXLAN tunnel network overlay from the matched flow.
3707 - ``nvgre_encap``: Performs a NVGRE encapsulation, outer layer configuration
3708 is done through `Config NVGRE Encap outer layers`_.
3710 - ``nvgre_decap``: Performs a decapsulation action by stripping all headers of
3711 the NVGRE tunnel network overlay from the matched flow.
3713 - ``set_ipv4_src``: Set a new IPv4 source address in the outermost IPv4 header.
3715 - ``ipv4_addr``: New IPv4 source address.
3717 - ``set_ipv4_dst``: Set a new IPv4 destination address in the outermost IPv4
3720 - ``ipv4_addr``: New IPv4 destination address.
3722 - ``set_ipv6_src``: Set a new IPv6 source address in the outermost IPv6 header.
3724 - ``ipv6_addr``: New IPv6 source address.
3726 - ``set_ipv6_dst``: Set a new IPv6 destination address in the outermost IPv6
3729 - ``ipv6_addr``: New IPv6 destination address.
3731 - ``of_set_tp_src``: Set a new source port number in the outermost TCP/UDP
3734 - ``port``: New TCP/UDP source port number.
3736 - ``of_set_tp_dst``: Set a new destination port number in the outermost TCP/UDP
3739 - ``port``: New TCP/UDP destination port number.
3741 - ``mac_swap``: Swap the source and destination MAC addresses in the outermost
3744 Destroying flow rules
3745 ~~~~~~~~~~~~~~~~~~~~~
3747 ``flow destroy`` destroys one or more rules from their rule ID (as returned
3748 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
3749 times as necessary::
3751 flow destroy {port_id} rule {rule_id} [...]
3753 If successful, it will show::
3755 Flow rule #[...] destroyed
3757 It does not report anything for rule IDs that do not exist. The usual error
3758 message is shown when a rule cannot be destroyed::
3760 Caught error type [...] ([...]): [...]
3762 ``flow flush`` destroys all rules on a device and does not take extra
3763 arguments. It is bound to ``rte_flow_flush()``::
3765 flow flush {port_id}
3767 Any errors are reported as above.
3769 Creating several rules and destroying them::
3771 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3772 actions queue index 2 / end
3773 Flow rule #0 created
3774 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3775 actions queue index 3 / end
3776 Flow rule #1 created
3777 testpmd> flow destroy 0 rule 0 rule 1
3778 Flow rule #1 destroyed
3779 Flow rule #0 destroyed
3782 The same result can be achieved using ``flow flush``::
3784 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3785 actions queue index 2 / end
3786 Flow rule #0 created
3787 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3788 actions queue index 3 / end
3789 Flow rule #1 created
3790 testpmd> flow flush 0
3793 Non-existent rule IDs are ignored::
3795 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3796 actions queue index 2 / end
3797 Flow rule #0 created
3798 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3799 actions queue index 3 / end
3800 Flow rule #1 created
3801 testpmd> flow destroy 0 rule 42 rule 10 rule 2
3803 testpmd> flow destroy 0 rule 0
3804 Flow rule #0 destroyed
3810 ``flow query`` queries a specific action of a flow rule having that
3811 ability. Such actions collect information that can be reported using this
3812 command. It is bound to ``rte_flow_query()``::
3814 flow query {port_id} {rule_id} {action}
3816 If successful, it will display either the retrieved data for known actions
3817 or the following message::
3819 Cannot display result for action type [...] ([...])
3821 Otherwise, it will complain either that the rule does not exist or that some
3824 Flow rule #[...] not found
3828 Caught error type [...] ([...]): [...]
3830 Currently only the ``count`` action is supported. This action reports the
3831 number of packets that hit the flow rule and the total number of bytes. Its
3832 output has the following format::
3835 hits_set: [...] # whether "hits" contains a valid value
3836 bytes_set: [...] # whether "bytes" contains a valid value
3837 hits: [...] # number of packets
3838 bytes: [...] # number of bytes
3840 Querying counters for TCPv6 packets redirected to queue 6::
3842 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
3843 actions queue index 6 / count / end
3844 Flow rule #4 created
3845 testpmd> flow query 0 4 count
3856 ``flow list`` lists existing flow rules sorted by priority and optionally
3857 filtered by group identifiers::
3859 flow list {port_id} [group {group_id}] [...]
3861 This command only fails with the following message if the device does not
3866 Output consists of a header line followed by a short description of each
3867 flow rule, one per line. There is no output at all when no flow rules are
3868 configured on the device::
3870 ID Group Prio Attr Rule
3871 [...] [...] [...] [...] [...]
3873 ``Attr`` column flags:
3875 - ``i`` for ``ingress``.
3876 - ``e`` for ``egress``.
3878 Creating several flow rules and listing them::
3880 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3881 actions queue index 6 / end
3882 Flow rule #0 created
3883 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3884 actions queue index 2 / end
3885 Flow rule #1 created
3886 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
3887 actions rss queues 6 7 8 end / end
3888 Flow rule #2 created
3889 testpmd> flow list 0
3890 ID Group Prio Attr Rule
3891 0 0 0 i- ETH IPV4 => QUEUE
3892 1 0 0 i- ETH IPV6 => QUEUE
3893 2 0 5 i- ETH IPV4 UDP => RSS
3896 Rules are sorted by priority (i.e. group ID first, then priority level)::
3898 testpmd> flow list 1
3899 ID Group Prio Attr Rule
3900 0 0 0 i- ETH => COUNT
3901 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3902 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3903 1 24 0 i- ETH IPV4 UDP => QUEUE
3904 4 24 10 i- ETH IPV4 TCP => DROP
3905 3 24 20 i- ETH IPV4 => DROP
3906 2 24 42 i- ETH IPV4 UDP => QUEUE
3907 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3910 Output can be limited to specific groups::
3912 testpmd> flow list 1 group 0 group 63
3913 ID Group Prio Attr Rule
3914 0 0 0 i- ETH => COUNT
3915 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3916 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3917 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3920 Toggling isolated mode
3921 ~~~~~~~~~~~~~~~~~~~~~~
3923 ``flow isolate`` can be used to tell the underlying PMD that ingress traffic
3924 must only be injected from the defined flow rules; that no default traffic
3925 is expected outside those rules and the driver is free to assign more
3926 resources to handle them. It is bound to ``rte_flow_isolate()``::
3928 flow isolate {port_id} {boolean}
3930 If successful, enabling or disabling isolated mode shows either::
3932 Ingress traffic on port [...]
3933 is now restricted to the defined flow rules
3937 Ingress traffic on port [...]
3938 is not restricted anymore to the defined flow rules
3940 Otherwise, in case of error::
3942 Caught error type [...] ([...]): [...]
3944 Mainly due to its side effects, PMDs supporting this mode may not have the
3945 ability to toggle it more than once without reinitializing affected ports
3946 first (e.g. by exiting testpmd).
3948 Enabling isolated mode::
3950 testpmd> flow isolate 0 true
3951 Ingress traffic on port 0 is now restricted to the defined flow rules
3954 Disabling isolated mode::
3956 testpmd> flow isolate 0 false
3957 Ingress traffic on port 0 is not restricted anymore to the defined flow rules
3960 Sample QinQ flow rules
3961 ~~~~~~~~~~~~~~~~~~~~~~
3963 Before creating QinQ rule(s) the following commands should be issued to enable QinQ::
3965 testpmd> port stop 0
3966 testpmd> vlan set qinq on 0
3968 The above command sets the inner and outer TPID's to 0x8100.
3970 To change the TPID's the following commands should be used::
3972 testpmd> vlan set outer tpid 0xa100 0
3973 testpmd> vlan set inner tpid 0x9100 0
3974 testpmd> port start 0
3976 Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM.
3980 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 /
3981 vlan tci is 456 / end actions vf id 1 / queue index 0 / end
3982 Flow rule #0 validated
3984 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 /
3985 vlan tci is 456 / end actions vf id 123 / queue index 0 / end
3986 Flow rule #0 created
3988 testpmd> flow list 0
3989 ID Group Prio Attr Rule
3990 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3992 Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host.
3996 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 /
3997 vlan tci is 654 / end actions pf / queue index 0 / end
3998 Flow rule #1 validated
4000 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 /
4001 vlan tci is 654 / end actions pf / queue index 1 / end
4002 Flow rule #1 created
4004 testpmd> flow list 0
4005 ID Group Prio Attr Rule
4006 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
4007 1 0 0 i- ETH VLAN VLAN=>PF QUEUE
4009 Sample VXLAN encapsulation rule
4010 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
4012 VXLAN encapsulation outer layer has default value pre-configured in testpmd
4013 source code, those can be changed by using the following commands
4015 IPv4 VXLAN outer header::
4017 testpmd> set vxlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src 127.0.0.1
4018 ip-dst 128.0.0.1 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
4019 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
4022 testpmd> set vxlan-with-vlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src
4023 127.0.0.1 ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11
4024 eth-dst 22:22:22:22:22:22
4025 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
4028 IPv6 VXLAN outer header::
4030 testpmd> set vxlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4 ip-src ::1
4031 ip-dst ::2222 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
4032 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
4035 testpmd> set vxlan-with-vlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4
4036 ip-src ::1 ip-dst ::2222 vlan-tci 34 eth-src 11:11:11:11:11:11
4037 eth-dst 22:22:22:22:22:22
4038 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
4041 Sample NVGRE encapsulation rule
4042 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
4044 NVGRE encapsulation outer layer has default value pre-configured in testpmd
4045 source code, those can be changed by using the following commands
4047 IPv4 NVGRE outer header::
4049 testpmd> set nvgre ip-version ipv4 tni 4 ip-src 127.0.0.1 ip-dst 128.0.0.1
4050 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
4051 testpmd> flow create 0 ingress pattern end actions nvgre_encap /
4054 testpmd> set nvgre-with-vlan ip-version ipv4 tni 4 ip-src 127.0.0.1
4055 ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11
4056 eth-dst 22:22:22:22:22:22
4057 testpmd> flow create 0 ingress pattern end actions nvgre_encap /
4060 IPv6 NVGRE outer header::
4062 testpmd> set nvgre ip-version ipv6 tni 4 ip-src ::1 ip-dst ::2222
4063 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
4064 testpmd> flow create 0 ingress pattern end actions nvgre_encap /
4067 testpmd> set nvgre-with-vlan ip-version ipv6 tni 4 ip-src ::1 ip-dst ::2222
4068 vlan-tci 34 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
4069 testpmd> flow create 0 ingress pattern end actions nvgre_encap /
4075 The following sections show functions to load/unload eBPF based filters.
4080 Load an eBPF program as a callback for partciular RX/TX queue::
4082 testpmd> bpf-load rx|tx (portid) (queueid) (load-flags) (bpf-prog-filename)
4084 The available load-flags are:
4086 * ``J``: use JIT generated native code, otherwise BPF interpreter will be used.
4088 * ``M``: assume input parameter is a pointer to rte_mbuf, otherwise assume it is a pointer to first segment's data.
4094 You'll need clang v3.7 or above to build bpf program you'd like to load
4098 .. code-block:: console
4101 clang -O2 -target bpf -c t1.c
4103 Then to load (and JIT compile) t1.o at RX queue 0, port 1::
4105 .. code-block:: console
4107 testpmd> bpf-load rx 1 0 J ./dpdk.org/test/bpf/t1.o
4109 To load (not JITed) t1.o at TX queue 0, port 0::
4111 .. code-block:: console
4113 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o
4118 Unload previously loaded eBPF program for partciular RX/TX queue::
4120 testpmd> bpf-unload rx|tx (portid) (queueid)
4122 For example to unload BPF filter from TX queue 0, port 0:
4124 .. code-block:: console
4126 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o