1 .. SPDX-License-Identifier: BSD-3-Clause
2 Copyright(c) 2010-2016 Intel Corporation.
6 Testpmd Runtime Functions
7 =========================
9 Where the testpmd application is started in interactive mode, (``-i|--interactive``),
10 it displays a prompt that can be used to start and stop forwarding,
11 configure the application, display statistics (including the extended NIC
12 statistics aka xstats) , set the Flow Director and other tasks::
16 The testpmd prompt has some, limited, readline support.
17 Common bash command-line functions such as ``Ctrl+a`` and ``Ctrl+e`` to go to the start and end of the prompt line are supported
18 as well as access to the command history via the up-arrow.
20 There is also support for tab completion.
21 If you type a partial command and hit ``<TAB>`` you get a list of the available completions:
23 .. code-block:: console
25 testpmd> show port <TAB>
27 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
28 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
29 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
30 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
36 Some examples in this document are too long to fit on one line are are shown wrapped at `"\\"` for display purposes::
38 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
39 (pause_time) (send_xon) (port_id)
41 In the real ``testpmd>`` prompt these commands should be on a single line.
46 The testpmd has on-line help for the functions that are available at runtime.
47 These are divided into sections and can be accessed using help, help section or help all:
49 .. code-block:: console
53 help control : Start and stop forwarding.
54 help display : Displaying port, stats and config information.
55 help config : Configuration information.
56 help ports : Configuring ports.
57 help registers : Reading and setting port registers.
58 help filters : Filters configuration help.
59 help all : All of the above sections.
62 Command File Functions
63 ----------------------
65 To facilitate loading large number of commands or to avoid cutting and pasting where not
66 practical or possible testpmd supports alternative methods for executing commands.
68 * If started with the ``--cmdline-file=FILENAME`` command line argument testpmd
69 will execute all CLI commands contained within the file immediately before
70 starting packet forwarding or entering interactive mode.
72 .. code-block:: console
74 ./testpmd -n4 -r2 ... -- -i --cmdline-file=/home/ubuntu/flow-create-commands.txt
75 Interactive-mode selected
76 CLI commands to be read from /home/ubuntu/flow-create-commands.txt
77 Configuring Port 0 (socket 0)
78 Port 0: 7C:FE:90:CB:74:CE
79 Configuring Port 1 (socket 0)
80 Port 1: 7C:FE:90:CB:74:CA
81 Checking link statuses...
82 Port 0 Link Up - speed 10000 Mbps - full-duplex
83 Port 1 Link Up - speed 10000 Mbps - full-duplex
89 Flow rule #498 created
90 Flow rule #499 created
91 Read all CLI commands from /home/ubuntu/flow-create-commands.txt
95 * At run-time additional commands can be loaded in bulk by invoking the ``load FILENAME``
98 .. code-block:: console
100 testpmd> load /home/ubuntu/flow-create-commands.txt
105 Flow rule #498 created
106 Flow rule #499 created
107 Read all CLI commands from /home/ubuntu/flow-create-commands.txt
111 In all cases output from any included command will be displayed as standard output.
112 Execution will continue until the end of the file is reached regardless of
113 whether any errors occur. The end user must examine the output to determine if
114 any failures occurred.
123 Start packet forwarding with current configuration::
130 Start packet forwarding with current configuration after sending specified number of bursts of packets::
132 testpmd> start tx_first (""|burst_num)
134 The default burst number is 1 when ``burst_num`` not presented.
139 Stop packet forwarding, and display accumulated statistics::
154 The functions in the following sections are used to display information about the
155 testpmd configuration or the NIC status.
160 Display information for a given port or all ports::
162 testpmd> show port (info|stats|xstats|fdir|stat_qmap|dcb_tc|cap) (port_id|all)
164 The available information categories are:
166 * ``info``: General port information such as MAC address.
168 * ``stats``: RX/TX statistics.
170 * ``xstats``: RX/TX extended NIC statistics.
172 * ``fdir``: Flow Director information and statistics.
174 * ``stat_qmap``: Queue statistics mapping.
176 * ``dcb_tc``: DCB information such as TC mapping.
178 * ``cap``: Supported offload capabilities.
182 .. code-block:: console
184 testpmd> show port info 0
186 ********************* Infos for port 0 *********************
188 MAC address: XX:XX:XX:XX:XX:XX
190 memory allocation on the socket: 0
192 Link speed: 40000 Mbps
193 Link duplex: full-duplex
194 Promiscuous mode: enabled
195 Allmulticast mode: disabled
196 Maximum number of MAC addresses: 64
197 Maximum number of MAC addresses of hash filtering: 0
202 Redirection table size: 512
203 Supported flow types:
223 Display the rss redirection table entry indicated by masks on port X::
225 testpmd> show port (port_id) rss reta (size) (mask0, mask1...)
227 size is used to indicate the hardware supported reta size
232 Display the RSS hash functions and RSS hash key of a port::
234 testpmd> show port (port_id) rss-hash ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2-payload|ipv6-ex|ipv6-tcp-ex|ipv6-udp-ex [key]
239 Clear the port statistics for a given port or for all ports::
241 testpmd> clear port (info|stats|xstats|fdir|stat_qmap) (port_id|all)
245 testpmd> clear port stats all
250 Display information for a given port's RX/TX queue::
252 testpmd> show (rxq|txq) info (port_id) (queue_id)
257 Displays the configuration of the application.
258 The configuration comes from the command-line, the runtime or the application defaults::
260 testpmd> show config (rxtx|cores|fwd|txpkts)
262 The available information categories are:
264 * ``rxtx``: RX/TX configuration items.
266 * ``cores``: List of forwarding cores.
268 * ``fwd``: Packet forwarding configuration.
270 * ``txpkts``: Packets to TX configuration.
274 .. code-block:: console
276 testpmd> show config rxtx
278 io packet forwarding - CRC stripping disabled - packets/burst=16
279 nb forwarding cores=2 - nb forwarding ports=1
280 RX queues=1 - RX desc=128 - RX free threshold=0
281 RX threshold registers: pthresh=8 hthresh=8 wthresh=4
282 TX queues=1 - TX desc=512 - TX free threshold=0
283 TX threshold registers: pthresh=36 hthresh=0 wthresh=0
284 TX RS bit threshold=0 - TXQ flags=0x0
289 Set the packet forwarding mode::
291 testpmd> set fwd (io|mac|macswap|flowgen| \
292 rxonly|txonly|csum|icmpecho) (""|retry)
294 ``retry`` can be specified for forwarding engines except ``rx_only``.
296 The available information categories are:
298 * ``io``: Forwards packets "as-is" in I/O mode.
299 This is the fastest possible forwarding operation as it does not access packets data.
300 This is the default mode.
302 * ``mac``: Changes the source and the destination Ethernet addresses of packets before forwarding them.
303 Default application behaviour is to set source Ethernet address to that of the transmitting interface, and destination
304 address to a dummy value (set during init). The user may specify a target destination Ethernet address via the 'eth-peer' or
305 'eth-peer-configfile' command-line options. It is not currently possible to specify a specific source Ethernet address.
307 * ``macswap``: MAC swap forwarding mode.
308 Swaps the source and the destination Ethernet addresses of packets before forwarding them.
310 * ``flowgen``: Multi-flow generation mode.
311 Originates a number of flows (with varying destination IP addresses), and terminate receive traffic.
313 * ``rxonly``: Receives packets but doesn't transmit them.
315 * ``txonly``: Generates and transmits packets without receiving any.
317 * ``csum``: Changes the checksum field with hardware or software methods depending on the offload flags on the packet.
319 * ``icmpecho``: Receives a burst of packets, lookup for IMCP echo requests and, if any, send back ICMP echo replies.
321 * ``ieee1588``: Demonstrate L2 IEEE1588 V2 PTP timestamping for RX and TX. Requires ``CONFIG_RTE_LIBRTE_IEEE1588=y``.
323 * ``softnic``: Demonstrates the softnic forwarding operation. In this mode, packet forwarding is
324 similar to I/O mode except for the fact that packets are loopback to the softnic ports only. Therefore, portmask parameter should be set to softnic port only. The various software based custom NIC pipelines specified through the softnic firmware (DPDK packet framework script) can be tested in this mode. Furthermore, it allows to build 5-level hierarchical QoS scheduler as a default option that can be enabled through CLI once testpmd application is initialised. The user can modify the default scheduler hierarchy or can specify the new QoS Scheduler hierarchy through CLI. Requires ``CONFIG_RTE_LIBRTE_PMD_SOFTNIC=y``.
328 testpmd> set fwd rxonly
330 Set rxonly packet forwarding mode
336 Display an RX descriptor for a port RX queue::
338 testpmd> read rxd (port_id) (queue_id) (rxd_id)
342 testpmd> read rxd 0 0 4
343 0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180
348 Display a TX descriptor for a port TX queue::
350 testpmd> read txd (port_id) (queue_id) (txd_id)
354 testpmd> read txd 0 0 4
355 0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C
360 Get loaded dynamic device personalization (DDP) package info list::
362 testpmd> ddp get list (port_id)
367 Display information about dynamic device personalization (DDP) profile::
369 testpmd> ddp get info (profile_path)
374 Display VF statistics::
376 testpmd> show vf stats (port_id) (vf_id)
381 Reset VF statistics::
383 testpmd> clear vf stats (port_id) (vf_id)
385 show port pctype mapping
386 ~~~~~~~~~~~~~~~~~~~~~~~~
388 List all items from the pctype mapping table::
390 testpmd> show port (port_id) pctype mapping
392 show rx offloading capabilities
393 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
395 List all per queue and per port Rx offloading capabilities of a port::
397 testpmd> show port (port_id) rx_offload capabilities
399 show rx offloading configuration
400 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
402 List port level and all queue level Rx offloading configuration::
404 testpmd> show port (port_id) rx_offload configuration
406 show tx offloading capabilities
407 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
409 List all per queue and per port Tx offloading capabilities of a port::
411 testpmd> show port (port_id) tx_offload capabilities
413 show tx offloading configuration
414 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
416 List port level and all queue level Tx offloading configuration::
418 testpmd> show port (port_id) tx_offload configuration
421 Configuration Functions
422 -----------------------
424 The testpmd application can be configured from the runtime as well as from the command-line.
426 This section details the available configuration functions that are available.
430 Configuration changes only become active when forwarding is started/restarted.
435 Reset forwarding to the default configuration::
442 Set the debug verbosity level::
444 testpmd> set verbose (level)
446 Currently the only available levels are 0 (silent except for error) and 1 (fully verbose).
451 Set the log level for a log type::
453 testpmd> set log global|(type) (level)
457 * ``type`` is the log name.
459 * ``level`` is the log level.
461 For example, to change the global log level::
462 testpmd> set log global (level)
464 Regexes can also be used for type. To change log level of user1, user2 and user3::
465 testpmd> set log user[1-3] (level)
470 Set the number of ports used by the application:
474 This is equivalent to the ``--nb-ports`` command-line option.
479 Set the number of cores used by the application::
481 testpmd> set nbcore (num)
483 This is equivalent to the ``--nb-cores`` command-line option.
487 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port.
492 Set the forwarding cores hexadecimal mask::
494 testpmd> set coremask (mask)
496 This is equivalent to the ``--coremask`` command-line option.
500 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding.
505 Set the forwarding ports hexadecimal mask::
507 testpmd> set portmask (mask)
509 This is equivalent to the ``--portmask`` command-line option.
514 Set number of packets per burst::
516 testpmd> set burst (num)
518 This is equivalent to the ``--burst command-line`` option.
520 When retry is enabled, the transmit delay time and number of retries can also be set::
522 testpmd> set burst tx delay (microseconds) retry (num)
527 Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode::
529 testpmd> set txpkts (x[,y]*)
531 Where x[,y]* represents a CSV list of values, without white space.
536 Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes::
538 testpmd> set txsplit (off|on|rand)
542 * ``off`` disable packet copy & split for CSUM mode.
544 * ``on`` split outgoing packet into multiple segments. Size of each segment
545 and number of segments per packet is determined by ``set txpkts`` command
548 * ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments.
553 Set the list of forwarding cores::
555 testpmd> set corelist (x[,y]*)
557 For example, to change the forwarding cores:
559 .. code-block:: console
561 testpmd> set corelist 3,1
562 testpmd> show config fwd
564 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled
565 Logical Core 3 (socket 0) forwards packets on 1 streams:
566 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01
567 Logical Core 1 (socket 0) forwards packets on 1 streams:
568 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
572 The cores are used in the same order as specified on the command line.
577 Set the list of forwarding ports::
579 testpmd> set portlist (x[,y]*)
581 For example, to change the port forwarding:
583 .. code-block:: console
585 testpmd> set portlist 0,2,1,3
586 testpmd> show config fwd
588 io packet forwarding - ports=4 - cores=1 - streams=4
589 Logical Core 3 (socket 0) forwards packets on 4 streams:
590 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01
591 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
592 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03
593 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02
598 Enable/disable tx loopback::
600 testpmd> set tx loopback (port_id) (on|off)
605 set drop enable bit for all queues::
607 testpmd> set all queues drop (port_id) (on|off)
609 set split drop enable (for VF)
610 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
612 set split drop enable bit for VF from PF::
614 testpmd> set vf split drop (port_id) (vf_id) (on|off)
616 set mac antispoof (for VF)
617 ~~~~~~~~~~~~~~~~~~~~~~~~~~
619 Set mac antispoof for a VF from the PF::
621 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off)
626 Enable/disable MACsec offload::
628 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off)
629 testpmd> set macsec offload (port_id) off
634 Configure MACsec secure connection (SC)::
636 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi)
640 The pi argument is ignored for tx.
641 Check the NIC Datasheet for hardware limits.
646 Configure MACsec secure association (SA)::
648 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key)
652 The IDX value must be 0 or 1.
653 Check the NIC Datasheet for hardware limits.
655 set broadcast mode (for VF)
656 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
658 Set broadcast mode for a VF from the PF::
660 testpmd> set vf broadcast (port_id) (vf_id) (on|off)
665 Set the VLAN strip on a port::
667 testpmd> vlan set strip (on|off) (port_id)
672 Set the VLAN strip for a queue on a port::
674 testpmd> vlan set stripq (on|off) (port_id,queue_id)
676 vlan set stripq (for VF)
677 ~~~~~~~~~~~~~~~~~~~~~~~~
679 Set VLAN strip for all queues in a pool for a VF from the PF::
681 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off)
683 vlan set insert (for VF)
684 ~~~~~~~~~~~~~~~~~~~~~~~~
686 Set VLAN insert for a VF from the PF::
688 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id)
690 vlan set tag (for VF)
691 ~~~~~~~~~~~~~~~~~~~~~
693 Set VLAN tag for a VF from the PF::
695 testpmd> set vf vlan tag (port_id) (vf_id) (on|off)
697 vlan set antispoof (for VF)
698 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
700 Set VLAN antispoof for a VF from the PF::
702 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off)
707 Set the VLAN filter on a port::
709 testpmd> vlan set filter (on|off) (port_id)
714 Set the VLAN QinQ (extended queue in queue) on for a port::
716 testpmd> vlan set qinq (on|off) (port_id)
721 Set the inner or outer VLAN TPID for packet filtering on a port::
723 testpmd> vlan set (inner|outer) tpid (value) (port_id)
727 TPID value must be a 16-bit number (value <= 65536).
732 Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID::
734 testpmd> rx_vlan add (vlan_id|all) (port_id)
738 VLAN filter must be set on that port. VLAN ID < 4096.
739 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries
740 in VFTA table. This is important if enabling all vlan_ids.
745 Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID::
747 testpmd> rx_vlan rm (vlan_id|all) (port_id)
752 Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID::
754 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask)
759 Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID::
761 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask)
766 Add a tunnel filter on a port::
768 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \
769 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
770 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
772 The available information categories are:
774 * ``vxlan``: Set tunnel type as VXLAN.
776 * ``nvgre``: Set tunnel type as NVGRE.
778 * ``ipingre``: Set tunnel type as IP-in-GRE.
780 * ``imac-ivlan``: Set filter type as Inner MAC and VLAN.
782 * ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID.
784 * ``imac-tenid``: Set filter type as Inner MAC and tenant ID.
786 * ``imac``: Set filter type as Inner MAC.
788 * ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID.
790 * ``oip``: Set filter type as Outer IP.
792 * ``iip``: Set filter type as Inner IP.
796 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \
797 192.168.2.2 0 ipingre oip 1 1
799 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP.
804 Remove a tunnel filter on a port::
806 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \
807 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
808 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
813 Add an UDP port for VXLAN packet filter on a port::
815 testpmd> rx_vxlan_port add (udp_port) (port_id)
820 Remove an UDP port for VXLAN packet filter on a port::
822 testpmd> rx_vxlan_port rm (udp_port) (port_id)
827 Set hardware insertion of VLAN IDs in packets sent on a port::
829 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer]
831 For example, set a single VLAN ID (5) insertion on port 0::
835 Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1::
843 Set port based hardware insertion of VLAN ID in packets sent on a port::
845 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off)
850 Disable hardware insertion of a VLAN header in packets sent on a port::
852 testpmd> tx_vlan reset (port_id)
857 Select hardware or software calculation of the checksum when
858 transmitting a packet using the ``csum`` forwarding engine::
860 testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id)
864 * ``ip|udp|tcp|sctp`` always relate to the inner layer.
866 * ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized
867 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are
868 supported). See also the ``csum parse-tunnel`` command.
872 Check the NIC Datasheet for hardware limits.
877 Set RSS queue region span on a port::
879 testpmd> set port (port_id) queue-region region_id (value) \
880 queue_start_index (value) queue_num (value)
882 Set flowtype mapping on a RSS queue region on a port::
884 testpmd> set port (port_id) queue-region region_id (value) flowtype (value)
888 * For the flowtype(pctype) of packet,the specific index for each type has
889 been defined in file i40e_type.h as enum i40e_filter_pctype.
891 Set user priority mapping on a RSS queue region on a port::
893 testpmd> set port (port_id) queue-region UP (value) region_id (value)
895 Flush all queue region related configuration on a port::
897 testpmd> set port (port_id) queue-region flush (on|off)
901 * "on"is just an enable function which server for other configuration,
902 it is for all configuration about queue region from up layer,
903 at first will only keep in DPDK softwarestored in driver,
904 only after "flush on", it commit all configuration to HW.
905 "off" is just clean all configuration about queue region just now,
906 and restore all to DPDK i40e driver default config when start up.
908 Show all queue region related configuration info on a port::
910 testpmd> show port (port_id) queue-region
914 Queue region only support on PF by now, so these command is
915 only for configuration of queue region on PF port.
920 Define how tunneled packets should be handled by the csum forward
923 testpmd> csum parse-tunnel (on|off) (tx_port_id)
925 If enabled, the csum forward engine will try to recognize supported
926 tunnel headers (vxlan, gre, ipip).
928 If disabled, treat tunnel packets as non-tunneled packets (a inner
929 header is handled as a packet payload).
933 The port argument is the TX port like in the ``csum set`` command.
937 Consider a packet in packet like the following::
939 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in
941 * If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
942 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the
943 ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``).
945 * If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
946 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``.
951 Display tx checksum offload configuration::
953 testpmd> csum show (port_id)
958 Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine::
960 testpmd> tso set (segsize) (port_id)
964 Check the NIC datasheet for hardware limits.
969 Display the status of TCP Segmentation Offload::
971 testpmd> tso show (port_id)
976 Enable or disable GRO in ``csum`` forwarding engine::
978 testpmd> set port <port_id> gro on|off
980 If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4
981 packets received from the given port.
983 If disabled, packets received from the given port won't be performed
984 GRO. By default, GRO is disabled for all ports.
988 When enable GRO for a port, TCP/IPv4 packets received from the port
989 will be performed GRO. After GRO, all merged packets have bad
990 checksums, since the GRO library doesn't re-calculate checksums for
991 the merged packets. Therefore, if users want the merged packets to
992 have correct checksums, please select HW IP checksum calculation and
993 HW TCP checksum calculation for the port which the merged packets are
999 Display GRO configuration for a given port::
1001 testpmd> show port <port_id> gro
1006 Set the cycle to flush the GROed packets from reassembly tables::
1008 testpmd> set gro flush <cycles>
1010 When enable GRO, the csum forwarding engine performs GRO on received
1011 packets, and the GROed packets are stored in reassembly tables. Users
1012 can use this command to determine when the GROed packets are flushed
1013 from the reassembly tables.
1015 The ``cycles`` is measured in GRO operation times. The csum forwarding
1016 engine flushes the GROed packets from the tables every ``cycles`` GRO
1019 By default, the value of ``cycles`` is 1, which means flush GROed packets
1020 from the reassembly tables as soon as one GRO operation finishes. The value
1021 of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``.
1023 Please note that the large value of ``cycles`` may cause the poor TCP/IP
1024 stack performance. Because the GROed packets are delayed to arrive the
1025 stack, thus causing more duplicated ACKs and TCP retransmissions.
1030 Toggle per-port GSO support in ``csum`` forwarding engine::
1032 testpmd> set port <port_id> gso on|off
1034 If enabled, the csum forwarding engine will perform GSO on supported IPv4
1035 packets, transmitted on the given port.
1037 If disabled, packets transmitted on the given port will not undergo GSO.
1038 By default, GSO is disabled for all ports.
1042 When GSO is enabled on a port, supported IPv4 packets transmitted on that
1043 port undergo GSO. Afterwards, the segmented packets are represented by
1044 multi-segment mbufs; however, the csum forwarding engine doesn't calculation
1045 of checksums for GSO'd segments in SW. As a result, if users want correct
1046 checksums in GSO segments, they should enable HW checksum calculation for
1049 For example, HW checksum calculation for VxLAN GSO'd packets may be enabled
1050 by setting the following options in the csum forwarding engine:
1052 testpmd> csum set outer_ip hw <port_id>
1054 testpmd> csum set ip hw <port_id>
1056 testpmd> csum set tcp hw <port_id>
1058 UDP GSO is the same as IP fragmentation, which treats the UDP header
1059 as the payload and does not modify it during segmentation. That is,
1060 after UDP GSO, only the first output fragment has the original UDP
1061 header. Therefore, users need to enable HW IP checksum calculation
1062 and SW UDP checksum calculation for GSO-enabled ports, if they want
1063 correct checksums for UDP/IPv4 packets.
1068 Set the maximum GSO segment size (measured in bytes), which includes the
1069 packet header and the packet payload for GSO-enabled ports (global)::
1071 testpmd> set gso segsz <length>
1076 Display the status of Generic Segmentation Offload for a given port::
1078 testpmd> show port <port_id> gso
1083 Add an alternative MAC address to a port::
1085 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX)
1090 Remove a MAC address from a port::
1092 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX)
1094 mac_addr add (for VF)
1095 ~~~~~~~~~~~~~~~~~~~~~
1097 Add an alternative MAC address for a VF to a port::
1099 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX)
1104 Set the default MAC address for a port::
1106 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX)
1108 mac_addr set (for VF)
1109 ~~~~~~~~~~~~~~~~~~~~~
1111 Set the MAC address for a VF from the PF::
1113 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX)
1118 Set the forwarding peer address for certain port::
1120 testpmd> set eth-peer (port_id) (perr_addr)
1122 This is equivalent to the ``--eth-peer`` command-line option.
1127 Set the unicast hash filter(s) on/off for a port::
1129 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
1134 Set the promiscuous mode on for a port or for all ports.
1135 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1137 testpmd> set promisc (port_id|all) (on|off)
1142 Set the allmulti mode for a port or for all ports::
1144 testpmd> set allmulti (port_id|all) (on|off)
1146 Same as the ifconfig (8) option. Controls how multicast packets are handled.
1148 set promisc (for VF)
1149 ~~~~~~~~~~~~~~~~~~~~
1151 Set the unicast promiscuous mode for a VF from PF.
1152 It's supported by Intel i40e NICs now.
1153 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1155 testpmd> set vf promisc (port_id) (vf_id) (on|off)
1157 set allmulticast (for VF)
1158 ~~~~~~~~~~~~~~~~~~~~~~~~~
1160 Set the multicast promiscuous mode for a VF from PF.
1161 It's supported by Intel i40e NICs now.
1162 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1164 testpmd> set vf allmulti (port_id) (vf_id) (on|off)
1166 set tx max bandwidth (for VF)
1167 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1169 Set TX max absolute bandwidth (Mbps) for a VF from PF::
1171 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth)
1173 set tc tx min bandwidth (for VF)
1174 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1176 Set all TCs' TX min relative bandwidth (%) for a VF from PF::
1178 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...)
1180 set tc tx max bandwidth (for VF)
1181 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1183 Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF::
1185 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth)
1187 set tc strict link priority mode
1188 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1190 Set some TCs' strict link priority mode on a physical port::
1192 testpmd> set tx strict-link-priority (port_id) (tc_bitmap)
1194 set tc tx min bandwidth
1195 ~~~~~~~~~~~~~~~~~~~~~~~
1197 Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs::
1199 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...)
1204 Set the link flow control parameter on a port::
1206 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1207 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
1208 autoneg (on|off) (port_id)
1212 * ``high_water`` (integer): High threshold value to trigger XOFF.
1214 * ``low_water`` (integer): Low threshold value to trigger XON.
1216 * ``pause_time`` (integer): Pause quota in the Pause frame.
1218 * ``send_xon`` (0/1): Send XON frame.
1220 * ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
1222 * ``autoneg``: Change the auto-negotiation parameter.
1227 Set the priority flow control parameter on a port::
1229 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1230 (pause_time) (priority) (port_id)
1234 * ``high_water`` (integer): High threshold value.
1236 * ``low_water`` (integer): Low threshold value.
1238 * ``pause_time`` (integer): Pause quota in the Pause frame.
1240 * ``priority`` (0-7): VLAN User Priority.
1245 Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
1247 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
1249 For example, to set rx queue 2 on port 0 to mapping 5::
1251 testpmd>set stat_qmap rx 0 2 5
1253 set xstats-hide-zero
1254 ~~~~~~~~~~~~~~~~~~~~
1256 Set the option to hide zero values for xstats display::
1258 testpmd> set xstats-hide-zero on|off
1262 By default, the zero values are displayed for xstats.
1264 set port - rx/tx (for VF)
1265 ~~~~~~~~~~~~~~~~~~~~~~~~~
1267 Set VF receive/transmit from a port::
1269 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
1271 set port - mac address filter (for VF)
1272 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1274 Add/Remove unicast or multicast MAC addr filter for a VF::
1276 testpmd> set port (port_id) vf (vf_id) (mac_addr) \
1277 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
1279 set port - rx mode(for VF)
1280 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1282 Set the VF receive mode of a port::
1284 testpmd> set port (port_id) vf (vf_id) \
1285 rxmode (AUPE|ROPE|BAM|MPE) (on|off)
1287 The available receive modes are:
1289 * ``AUPE``: Accepts untagged VLAN.
1291 * ``ROPE``: Accepts unicast hash.
1293 * ``BAM``: Accepts broadcast packets.
1295 * ``MPE``: Accepts all multicast packets.
1297 set port - tx_rate (for Queue)
1298 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1300 Set TX rate limitation for a queue on a port::
1302 testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1304 set port - tx_rate (for VF)
1305 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
1307 Set TX rate limitation for queues in VF on a port::
1309 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1311 set port - mirror rule
1312 ~~~~~~~~~~~~~~~~~~~~~~
1314 Set pool or vlan type mirror rule for a port::
1316 testpmd> set port (port_id) mirror-rule (rule_id) \
1317 (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1318 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1320 Set link mirror rule for a port::
1322 testpmd> set port (port_id) mirror-rule (rule_id) \
1323 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1325 For example to enable mirror traffic with vlan 0,1 to pool 0::
1327 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1329 reset port - mirror rule
1330 ~~~~~~~~~~~~~~~~~~~~~~~~
1332 Reset a mirror rule for a port::
1334 testpmd> reset port (port_id) mirror-rule (rule_id)
1339 Set the flush on RX streams before forwarding.
1340 The default is flush ``on``.
1341 Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1343 testpmd> set flush_rx off
1348 Set the bypass mode for the lowest port on bypass enabled NIC::
1350 testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1355 Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1357 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1358 mode (normal|bypass|isolate) (port_id)
1362 * ``timeout``: Enable bypass after watchdog timeout.
1364 * ``os_on``: Enable bypass when OS/board is powered on.
1366 * ``os_off``: Enable bypass when OS/board is powered off.
1368 * ``power_on``: Enable bypass when power supply is turned on.
1370 * ``power_off``: Enable bypass when power supply is turned off.
1376 Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1378 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1383 Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1385 testpmd> show bypass config (port_id)
1390 Set link up for a port::
1392 testpmd> set link-up port (port id)
1397 Set link down for a port::
1399 testpmd> set link-down port (port id)
1404 Enable E-tag insertion for a VF on a port::
1406 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1408 Disable E-tag insertion for a VF on a port::
1410 testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1412 Enable/disable E-tag stripping on a port::
1414 testpmd> E-tag set stripping (on|off) port (port_id)
1416 Enable/disable E-tag based forwarding on a port::
1418 testpmd> E-tag set forwarding (on|off) port (port_id)
1420 Add an E-tag forwarding filter on a port::
1422 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1424 Delete an E-tag forwarding filter on a port::
1425 testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1430 Load a dynamic device personalization (DDP) profile and store backup profile::
1432 testpmd> ddp add (port_id) (profile_path[,backup_profile_path])
1437 Delete a dynamic device personalization profile and restore backup profile::
1439 testpmd> ddp del (port_id) (backup_profile_path)
1444 List all items from the ptype mapping table::
1446 testpmd> ptype mapping get (port_id) (valid_only)
1450 * ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0).
1452 Replace a specific or a group of software defined ptype with a new one::
1454 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type)
1458 * ``target``: A specific software ptype or a mask to represent a group of software ptypes.
1460 * ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1).
1462 * ``pkt_type``: The new software ptype to replace the old ones.
1464 Update hardware defined ptype to software defined packet type mapping table::
1466 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype)
1470 * ``hw_ptype``: hardware ptype as the index of the ptype mapping table.
1472 * ``sw_ptype``: software ptype as the value of the ptype mapping table.
1474 Reset ptype mapping table::
1476 testpmd> ptype mapping reset (port_id)
1478 config per port Rx offloading
1479 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1481 Enable or disable a per port Rx offloading on all Rx queues of a port::
1483 testpmd> port config (port_id) rx_offload (offloading) on|off
1485 * ``offloading``: can be any of these offloading capability:
1486 vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro,
1487 qinq_strip, outer_ipv4_cksum, macsec_strip,
1488 header_split, vlan_filter, vlan_extend, jumbo_frame,
1489 crc_strip, scatter, timestamp, security, keep_crc
1491 This command should be run when the port is stopped, or else it will fail.
1493 config per queue Rx offloading
1494 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1496 Enable or disable a per queue Rx offloading only on a specific Rx queue::
1498 testpmd> port (port_id) rxq (queue_id) rx_offload (offloading) on|off
1500 * ``offloading``: can be any of these offloading capability:
1501 vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro,
1502 qinq_strip, outer_ipv4_cksum, macsec_strip,
1503 header_split, vlan_filter, vlan_extend, jumbo_frame,
1504 crc_strip, scatter, timestamp, security, keep_crc
1506 This command should be run when the port is stopped, or else it will fail.
1508 config per port Tx offloading
1509 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1511 Enable or disable a per port Tx offloading on all Tx queues of a port::
1513 testpmd> port config (port_id) tx_offload (offloading) on|off
1515 * ``offloading``: can be any of these offloading capability:
1516 vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum,
1517 sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum,
1518 qinq_insert, vxlan_tnl_tso, gre_tnl_tso,
1519 ipip_tnl_tso, geneve_tnl_tso, macsec_insert,
1520 mt_lockfree, multi_segs, mbuf_fast_free, security
1522 This command should be run when the port is stopped, or else it will fail.
1524 config per queue Tx offloading
1525 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1527 Enable or disable a per queue Tx offloading only on a specific Tx queue::
1529 testpmd> port (port_id) txq (queue_id) tx_offload (offloading) on|off
1531 * ``offloading``: can be any of these offloading capability:
1532 vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum,
1533 sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum,
1534 qinq_insert, vxlan_tnl_tso, gre_tnl_tso,
1535 ipip_tnl_tso, geneve_tnl_tso, macsec_insert,
1536 mt_lockfree, multi_segs, mbuf_fast_free, security
1538 This command should be run when the port is stopped, or else it will fail.
1540 Config VXLAN Encap outer layers
1541 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1543 Configure the outer layer to encapsulate a packet inside a VXLAN tunnel::
1545 set vxlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \
1546 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) eth-src (eth-src) \
1549 set vxlan-with-vlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \
1550 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) vlan-tci (vlan-tci) \
1551 eth-src (eth-src) eth-dst (eth-dst)
1553 Those command will set an internal configuration inside testpmd, any following
1554 flow rule using the action vxlan_encap will use the last configuration set.
1555 To have a different encapsulation header, one of those commands must be called
1556 before the flow rule creation.
1558 Config NVGRE Encap outer layers
1559 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1561 Configure the outer layer to encapsulate a packet inside a NVGRE tunnel::
1563 set nvgre ip-version (ipv4|ipv6) tni (tni) ip-src (ip-src) ip-dst (ip-dst) \
1564 eth-src (eth-src) eth-dst (eth-dst)
1565 set nvgre-with-vlan ip-version (ipv4|ipv6) tni (tni) ip-src (ip-src) \
1566 ip-dst (ip-dst) vlan-tci (vlan-tci) eth-src (eth-src) eth-dst (eth-dst)
1568 Those command will set an internal configuration inside testpmd, any following
1569 flow rule using the action nvgre_encap will use the last configuration set.
1570 To have a different encapsulation header, one of those commands must be called
1571 before the flow rule creation.
1576 The following sections show functions for configuring ports.
1580 Port configuration changes only become active when forwarding is started/restarted.
1585 Attach a port specified by pci address or virtual device args::
1587 testpmd> port attach (identifier)
1589 To attach a new pci device, the device should be recognized by kernel first.
1590 Then it should be moved under DPDK management.
1591 Finally the port can be attached to testpmd.
1593 For example, to move a pci device using ixgbe under DPDK management:
1595 .. code-block:: console
1597 # Check the status of the available devices.
1598 ./usertools/dpdk-devbind.py --status
1600 Network devices using DPDK-compatible driver
1601 ============================================
1604 Network devices using kernel driver
1605 ===================================
1606 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1609 # Bind the device to igb_uio.
1610 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1613 # Recheck the status of the devices.
1614 ./usertools/dpdk-devbind.py --status
1615 Network devices using DPDK-compatible driver
1616 ============================================
1617 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1619 To attach a port created by virtual device, above steps are not needed.
1621 For example, to attach a port whose pci address is 0000:0a:00.0.
1623 .. code-block:: console
1625 testpmd> port attach 0000:0a:00.0
1626 Attaching a new port...
1627 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1628 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1629 EAL: PCI memory mapped at 0x7f83bfa00000
1630 EAL: PCI memory mapped at 0x7f83bfa80000
1631 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1632 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1633 Port 0 is attached. Now total ports is 1
1636 For example, to attach a port created by pcap PMD.
1638 .. code-block:: console
1640 testpmd> port attach net_pcap0
1641 Attaching a new port...
1642 PMD: Initializing pmd_pcap for net_pcap0
1643 PMD: Creating pcap-backed ethdev on numa socket 0
1644 Port 0 is attached. Now total ports is 1
1647 In this case, identifier is ``net_pcap0``.
1648 This identifier format is the same as ``--vdev`` format of DPDK applications.
1650 For example, to re-attach a bonded port which has been previously detached,
1651 the mode and slave parameters must be given.
1653 .. code-block:: console
1655 testpmd> port attach net_bond_0,mode=0,slave=1
1656 Attaching a new port...
1657 EAL: Initializing pmd_bond for net_bond_0
1658 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1659 Port 0 is attached. Now total ports is 1
1666 Detach a specific port::
1668 testpmd> port detach (port_id)
1670 Before detaching a port, the port should be stopped and closed.
1672 For example, to detach a pci device port 0.
1674 .. code-block:: console
1676 testpmd> port stop 0
1679 testpmd> port close 0
1683 testpmd> port detach 0
1685 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1686 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1687 EAL: PCI memory unmapped at 0x7f83bfa00000
1688 EAL: PCI memory unmapped at 0x7f83bfa80000
1692 For example, to detach a virtual device port 0.
1694 .. code-block:: console
1696 testpmd> port stop 0
1699 testpmd> port close 0
1703 testpmd> port detach 0
1705 PMD: Closing pcap ethdev on numa socket 0
1706 Port 'net_pcap0' is detached. Now total ports is 0
1709 To remove a pci device completely from the system, first detach the port from testpmd.
1710 Then the device should be moved under kernel management.
1711 Finally the device can be removed using kernel pci hotplug functionality.
1713 For example, to move a pci device under kernel management:
1715 .. code-block:: console
1717 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1719 ./usertools/dpdk-devbind.py --status
1721 Network devices using DPDK-compatible driver
1722 ============================================
1725 Network devices using kernel driver
1726 ===================================
1727 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1729 To remove a port created by a virtual device, above steps are not needed.
1734 Start all ports or a specific port::
1736 testpmd> port start (port_id|all)
1741 Stop all ports or a specific port::
1743 testpmd> port stop (port_id|all)
1748 Close all ports or a specific port::
1750 testpmd> port close (port_id|all)
1752 port config - queue ring size
1753 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1755 Configure a rx/tx queue ring size::
1757 testpmd> port (port_id) (rxq|txq) (queue_id) ring_size (value)
1759 Only take effect after command that (re-)start the port or command that setup specific queue.
1761 port start/stop queue
1762 ~~~~~~~~~~~~~~~~~~~~~
1764 Start/stop a rx/tx queue on a specific port::
1766 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1769 ~~~~~~~~~~~~~~~~~~~~~
1771 Setup a rx/tx queue on a specific port::
1773 testpmd> port (port_id) (rxq|txq) (queue_id) setup
1775 Only take effect when port is started.
1780 Set the speed and duplex mode for all ports or a specific port::
1782 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1783 duplex (half|full|auto)
1785 port config - queues/descriptors
1786 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1788 Set number of queues/descriptors for rxq, txq, rxd and txd::
1790 testpmd> port config all (rxq|txq|rxd|txd) (value)
1792 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1794 port config - max-pkt-len
1795 ~~~~~~~~~~~~~~~~~~~~~~~~~
1797 Set the maximum packet length::
1799 testpmd> port config all max-pkt-len (value)
1801 This is equivalent to the ``--max-pkt-len`` command-line option.
1803 port config - CRC Strip
1804 ~~~~~~~~~~~~~~~~~~~~~~~
1806 Set hardware CRC stripping on or off for all ports::
1808 testpmd> port config all crc-strip (on|off)
1810 CRC stripping is on by default.
1812 The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option.
1814 port config - scatter
1815 ~~~~~~~~~~~~~~~~~~~~~~~
1817 Set RX scatter mode on or off for all ports::
1819 testpmd> port config all scatter (on|off)
1821 RX scatter mode is off by default.
1823 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1825 port config - RX Checksum
1826 ~~~~~~~~~~~~~~~~~~~~~~~~~
1828 Set hardware RX checksum offload to on or off for all ports::
1830 testpmd> port config all rx-cksum (on|off)
1832 Checksum offload is off by default.
1834 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1839 Set hardware VLAN on or off for all ports::
1841 testpmd> port config all hw-vlan (on|off)
1843 Hardware VLAN is off by default.
1845 The ``on`` option is equivalent to the ``--enable-hw-vlan`` command-line option.
1847 port config - VLAN filter
1848 ~~~~~~~~~~~~~~~~~~~~~~~~~
1850 Set hardware VLAN filter on or off for all ports::
1852 testpmd> port config all hw-vlan-filter (on|off)
1854 Hardware VLAN filter is off by default.
1856 The ``on`` option is equivalent to the ``--enable-hw-vlan-filter`` command-line option.
1858 port config - VLAN strip
1859 ~~~~~~~~~~~~~~~~~~~~~~~~
1861 Set hardware VLAN strip on or off for all ports::
1863 testpmd> port config all hw-vlan-strip (on|off)
1865 Hardware VLAN strip is off by default.
1867 The ``on`` option is equivalent to the ``--enable-hw-vlan-strip`` command-line option.
1869 port config - VLAN extend
1870 ~~~~~~~~~~~~~~~~~~~~~~~~~
1872 Set hardware VLAN extend on or off for all ports::
1874 testpmd> port config all hw-vlan-extend (on|off)
1876 Hardware VLAN extend is off by default.
1878 The ``on`` option is equivalent to the ``--enable-hw-vlan-extend`` command-line option.
1880 port config - Drop Packets
1881 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1883 Set packet drop for packets with no descriptors on or off for all ports::
1885 testpmd> port config all drop-en (on|off)
1887 Packet dropping for packets with no descriptors is off by default.
1889 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1894 Set the RSS (Receive Side Scaling) mode on or off::
1896 testpmd> port config all rss (all|default|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1898 RSS is on by default.
1900 The ``all`` option is equivalent to ip|tcp|udp|sctp|ether.
1901 The ``default`` option enables all supported RSS types reported by device info.
1902 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1904 port config - RSS Reta
1905 ~~~~~~~~~~~~~~~~~~~~~~
1907 Set the RSS (Receive Side Scaling) redirection table::
1909 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1914 Set the DCB mode for an individual port::
1916 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1918 The traffic class should be 4 or 8.
1923 Set the number of packets per burst::
1925 testpmd> port config all burst (value)
1927 This is equivalent to the ``--burst`` command-line option.
1929 port config - Threshold
1930 ~~~~~~~~~~~~~~~~~~~~~~~
1932 Set thresholds for TX/RX queues::
1934 testpmd> port config all (threshold) (value)
1936 Where the threshold type can be:
1938 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1940 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1942 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1944 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1946 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1948 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1950 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1952 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1954 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1956 These threshold options are also available from the command-line.
1961 Set the value of ether-type for E-tag::
1963 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1965 Enable/disable the E-tag support::
1967 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1969 port config pctype mapping
1970 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1972 Reset pctype mapping table::
1974 testpmd> port config (port_id) pctype mapping reset
1976 Update hardware defined pctype to software defined flow type mapping table::
1978 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id)
1982 * ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table.
1984 * ``flow_type_id``: software flow type id as the index of the pctype mapping table.
1986 port config input set
1987 ~~~~~~~~~~~~~~~~~~~~~
1989 Config RSS/FDIR/FDIR flexible payload input set for some pctype::
1990 testpmd> port config (port_id) pctype (pctype_id) \
1991 (hash_inset|fdir_inset|fdir_flx_inset) \
1992 (get|set|clear) field (field_idx)
1994 Clear RSS/FDIR/FDIR flexible payload input set for some pctype::
1995 testpmd> port config (port_id) pctype (pctype_id) \
1996 (hash_inset|fdir_inset|fdir_flx_inset) clear all
2000 * ``pctype_id``: hardware packet classification types.
2001 * ``field_idx``: hardware field index.
2003 port config udp_tunnel_port
2004 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
2006 Add/remove UDP tunnel port for VXLAN/GENEVE tunneling protocols::
2007 testpmd> port config (port_id) udp_tunnel_port add|rm vxlan|geneve (udp_port)
2009 Link Bonding Functions
2010 ----------------------
2012 The Link Bonding functions make it possible to dynamically create and
2013 manage link bonding devices from within testpmd interactive prompt.
2015 create bonded device
2016 ~~~~~~~~~~~~~~~~~~~~
2018 Create a new bonding device::
2020 testpmd> create bonded device (mode) (socket)
2022 For example, to create a bonded device in mode 1 on socket 0::
2024 testpmd> create bonded device 1 0
2025 created new bonded device (port X)
2030 Adds Ethernet device to a Link Bonding device::
2032 testpmd> add bonding slave (slave id) (port id)
2034 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
2036 testpmd> add bonding slave 6 10
2039 remove bonding slave
2040 ~~~~~~~~~~~~~~~~~~~~
2042 Removes an Ethernet slave device from a Link Bonding device::
2044 testpmd> remove bonding slave (slave id) (port id)
2046 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
2048 testpmd> remove bonding slave 6 10
2053 Set the Link Bonding mode of a Link Bonding device::
2055 testpmd> set bonding mode (value) (port id)
2057 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
2059 testpmd> set bonding mode 3 10
2064 Set an Ethernet slave device as the primary device on a Link Bonding device::
2066 testpmd> set bonding primary (slave id) (port id)
2068 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
2070 testpmd> set bonding primary 6 10
2075 Set the MAC address of a Link Bonding device::
2077 testpmd> set bonding mac (port id) (mac)
2079 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
2081 testpmd> set bonding mac 10 00:00:00:00:00:01
2083 set bonding xmit_balance_policy
2084 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2086 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
2088 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
2090 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
2092 testpmd> set bonding xmit_balance_policy 10 l34
2095 set bonding mon_period
2096 ~~~~~~~~~~~~~~~~~~~~~~
2098 Set the link status monitoring polling period in milliseconds for a bonding device.
2100 This adds support for PMD slave devices which do not support link status interrupts.
2101 When the mon_period is set to a value greater than 0 then all PMD's which do not support
2102 link status ISR will be queried every polling interval to check if their link status has changed::
2104 testpmd> set bonding mon_period (port_id) (value)
2106 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
2108 testpmd> set bonding mon_period 5 150
2111 set bonding lacp dedicated_queue
2112 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2114 Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic
2115 when in mode 4 (link-aggregration-802.3ad)::
2117 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable)
2120 set bonding agg_mode
2121 ~~~~~~~~~~~~~~~~~~~~
2123 Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad)::
2125 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable)
2131 Show the current configuration of a Link Bonding device::
2133 testpmd> show bonding config (port id)
2136 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
2137 in balance mode with a transmission policy of layer 2+3::
2139 testpmd> show bonding config 9
2141 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
2143 Active Slaves (3): [1 3 4]
2150 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
2151 This is mainly useful for debugging purposes.
2152 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
2153 and fields that can be accessed.
2158 Display the value of a port register::
2160 testpmd> read reg (port_id) (address)
2162 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
2164 testpmd> read reg 0 0xEE00
2165 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
2170 Display a port register bit field::
2172 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
2174 For example, reading the lowest two bits from the register in the example above::
2176 testpmd> read regfield 0 0xEE00 0 1
2177 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
2182 Display a single port register bit::
2184 testpmd> read regbit (port_id) (address) (bit_x)
2186 For example, reading the lowest bit from the register in the example above::
2188 testpmd> read regbit 0 0xEE00 0
2189 port 0 PCI register at offset 0xEE00: bit 0=1
2194 Set the value of a port register::
2196 testpmd> write reg (port_id) (address) (value)
2198 For example, to clear a register::
2200 testpmd> write reg 0 0xEE00 0x0
2201 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
2206 Set bit field of a port register::
2208 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
2210 For example, writing to the register cleared in the example above::
2212 testpmd> write regfield 0 0xEE00 0 1 2
2213 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
2218 Set single bit value of a port register::
2220 testpmd> write regbit (port_id) (address) (bit_x) (value)
2222 For example, to set the high bit in the register from the example above::
2224 testpmd> write regbit 0 0xEE00 31 1
2225 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
2227 Traffic Metering and Policing
2228 -----------------------------
2230 The following section shows functions for configuring traffic metering and
2231 policing on the ethernet device through the use of generic ethdev API.
2233 show port traffic management capability
2234 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2236 Show traffic metering and policing capability of the port::
2238 testpmd> show port meter cap (port_id)
2240 add port meter profile (srTCM rfc2967)
2241 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2243 Add meter profile (srTCM rfc2697) to the ethernet device::
2245 testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \
2250 * ``profile_id``: ID for the meter profile.
2251 * ``cir``: Committed Information Rate (CIR) (bytes/second).
2252 * ``cbs``: Committed Burst Size (CBS) (bytes).
2253 * ``ebs``: Excess Burst Size (EBS) (bytes).
2255 add port meter profile (trTCM rfc2968)
2256 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2258 Add meter profile (srTCM rfc2698) to the ethernet device::
2260 testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \
2261 (cir) (pir) (cbs) (pbs)
2265 * ``profile_id``: ID for the meter profile.
2266 * ``cir``: Committed information rate (bytes/second).
2267 * ``pir``: Peak information rate (bytes/second).
2268 * ``cbs``: Committed burst size (bytes).
2269 * ``pbs``: Peak burst size (bytes).
2271 add port meter profile (trTCM rfc4115)
2272 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2274 Add meter profile (trTCM rfc4115) to the ethernet device::
2276 testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \
2277 (cir) (eir) (cbs) (ebs)
2281 * ``profile_id``: ID for the meter profile.
2282 * ``cir``: Committed information rate (bytes/second).
2283 * ``eir``: Excess information rate (bytes/second).
2284 * ``cbs``: Committed burst size (bytes).
2285 * ``ebs``: Excess burst size (bytes).
2287 delete port meter profile
2288 ~~~~~~~~~~~~~~~~~~~~~~~~~
2290 Delete meter profile from the ethernet device::
2292 testpmd> del port meter profile (port_id) (profile_id)
2297 Create new meter object for the ethernet device::
2299 testpmd> create port meter (port_id) (mtr_id) (profile_id) \
2300 (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \
2301 (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\
2306 * ``mtr_id``: meter object ID.
2307 * ``profile_id``: ID for the meter profile.
2308 * ``meter_enable``: When this parameter has a non-zero value, the meter object
2309 gets enabled at the time of creation, otherwise remains disabled.
2310 * ``g_action``: Policer action for the packet with green color.
2311 * ``y_action``: Policer action for the packet with yellow color.
2312 * ``r_action``: Policer action for the packet with red color.
2313 * ``stats_mask``: Mask of statistics counter types to be enabled for the
2315 * ``shared``: When this parameter has a non-zero value, the meter object is
2316 shared by multiple flows. Otherwise, meter object is used by single flow.
2317 * ``use_pre_meter_color``: When this parameter has a non-zero value, the
2318 input color for the current meter object is determined by the latest meter
2319 object in the same flow. Otherwise, the current meter object uses the
2320 *dscp_table* to determine the input color.
2321 * ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input
2322 color, 0 <= x <= 63.
2327 Enable meter for the ethernet device::
2329 testpmd> enable port meter (port_id) (mtr_id)
2334 Disable meter for the ethernet device::
2336 testpmd> disable port meter (port_id) (mtr_id)
2341 Delete meter for the ethernet device::
2343 testpmd> del port meter (port_id) (mtr_id)
2345 Set port meter profile
2346 ~~~~~~~~~~~~~~~~~~~~~~
2348 Set meter profile for the ethernet device::
2350 testpmd> set port meter profile (port_id) (mtr_id) (profile_id)
2352 set port meter dscp table
2353 ~~~~~~~~~~~~~~~~~~~~~~~~~
2355 Set meter dscp table for the ethernet device::
2357 testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \
2358 (dscp_tbl_entry1)...(dscp_tbl_entry63)]
2360 set port meter policer action
2361 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2363 Set meter policer action for the ethernet device::
2365 testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \
2366 (action0) [(action1) (action1)]
2370 * ``action_mask``: Bit mask indicating which policer actions need to be
2371 updated. One or more policer actions can be updated in a single function
2372 invocation. To update the policer action associated with color C, bit
2373 (1 << C) needs to be set in *action_mask* and element at position C
2374 in the *actions* array needs to be valid.
2375 * ``actionx``: Policer action for the color x,
2376 RTE_MTR_GREEN <= x < RTE_MTR_COLORS
2378 set port meter stats mask
2379 ~~~~~~~~~~~~~~~~~~~~~~~~~
2381 Set meter stats mask for the ethernet device::
2383 testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask)
2387 * ``stats_mask``: Bit mask indicating statistics counter types to be enabled.
2389 show port meter stats
2390 ~~~~~~~~~~~~~~~~~~~~~
2392 Show meter stats of the ethernet device::
2394 testpmd> show port meter stats (port_id) (mtr_id) (clear)
2398 * ``clear``: Flag that indicates whether the statistics counters should
2399 be cleared (i.e. set to zero) immediately after they have been read or not.
2404 The following section shows functions for configuring traffic management on
2405 on the ethernet device through the use of generic TM API.
2407 show port traffic management capability
2408 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2410 Show traffic management capability of the port::
2412 testpmd> show port tm cap (port_id)
2414 show port traffic management capability (hierarchy level)
2415 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2417 Show traffic management hierarchy level capability of the port::
2419 testpmd> show port tm level cap (port_id) (level_id)
2421 show port traffic management capability (hierarchy node level)
2422 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2424 Show the traffic management hierarchy node capability of the port::
2426 testpmd> show port tm node cap (port_id) (node_id)
2428 show port traffic management hierarchy node type
2429 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2431 Show the port traffic management hierarchy node type::
2433 testpmd> show port tm node type (port_id) (node_id)
2435 show port traffic management hierarchy node stats
2436 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2438 Show the port traffic management hierarchy node statistics::
2440 testpmd> show port tm node stats (port_id) (node_id) (clear)
2444 * ``clear``: When this parameter has a non-zero value, the statistics counters
2445 are cleared (i.e. set to zero) immediately after they have been read,
2446 otherwise the statistics counters are left untouched.
2448 Add port traffic management private shaper profile
2449 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2451 Add the port traffic management private shaper profile::
2453 testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \
2454 (tb_rate) (tb_size) (packet_length_adjust)
2458 * ``shaper_profile id``: Shaper profile ID for the new profile.
2459 * ``tb_rate``: Token bucket rate (bytes per second).
2460 * ``tb_size``: Token bucket size (bytes).
2461 * ``packet_length_adjust``: The value (bytes) to be added to the length of
2462 each packet for the purpose of shaping. This parameter value can be used to
2463 correct the packet length with the framing overhead bytes that are consumed
2466 Delete port traffic management private shaper profile
2467 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2469 Delete the port traffic management private shaper::
2471 testpmd> del port tm node shaper profile (port_id) (shaper_profile_id)
2475 * ``shaper_profile id``: Shaper profile ID that needs to be deleted.
2477 Add port traffic management shared shaper
2478 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2480 Create the port traffic management shared shaper::
2482 testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \
2487 * ``shared_shaper_id``: Shared shaper ID to be created.
2488 * ``shaper_profile id``: Shaper profile ID for shared shaper.
2490 Set port traffic management shared shaper
2491 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2493 Update the port traffic management shared shaper::
2495 testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \
2500 * ``shared_shaper_id``: Shared shaper ID to be update.
2501 * ``shaper_profile id``: Shaper profile ID for shared shaper.
2503 Delete port traffic management shared shaper
2504 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2506 Delete the port traffic management shared shaper::
2508 testpmd> del port tm node shared shaper (port_id) (shared_shaper_id)
2512 * ``shared_shaper_id``: Shared shaper ID to be deleted.
2514 Set port traffic management hiearchy node private shaper
2515 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2517 set the port traffic management hierarchy node private shaper::
2519 testpmd> set port tm node shaper profile (port_id) (node_id) \
2524 * ``shaper_profile id``: Private shaper profile ID to be enabled on the
2527 Add port traffic management WRED profile
2528 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2530 Create a new WRED profile::
2532 testpmd> add port tm node wred profile (port_id) (wred_profile_id) \
2533 (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \
2534 (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \
2535 (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r)
2539 * ``wred_profile id``: Identifier for the newly create WRED profile
2540 * ``color_g``: Packet color (green)
2541 * ``min_th_g``: Minimum queue threshold for packet with green color
2542 * ``max_th_g``: Minimum queue threshold for packet with green color
2543 * ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp)
2544 * ``wq_log2_g``: Negated log2 of queue weight (wq)
2545 * ``color_y``: Packet color (yellow)
2546 * ``min_th_y``: Minimum queue threshold for packet with yellow color
2547 * ``max_th_y``: Minimum queue threshold for packet with yellow color
2548 * ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp)
2549 * ``wq_log2_y``: Negated log2 of queue weight (wq)
2550 * ``color_r``: Packet color (red)
2551 * ``min_th_r``: Minimum queue threshold for packet with yellow color
2552 * ``max_th_r``: Minimum queue threshold for packet with yellow color
2553 * ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp)
2554 * ``wq_log2_r``: Negated log2 of queue weight (wq)
2556 Delete port traffic management WRED profile
2557 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2559 Delete the WRED profile::
2561 testpmd> del port tm node wred profile (port_id) (wred_profile_id)
2563 Add port traffic management hierarchy nonleaf node
2564 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2566 Add nonleaf node to port traffic management hiearchy::
2568 testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \
2569 (priority) (weight) (level_id) (shaper_profile_id) \
2570 (n_sp_priorities) (stats_mask) (n_shared_shapers) \
2571 [(shared_shaper_0) (shared_shaper_1) ...] \
2575 * ``parent_node_id``: Node ID of the parent.
2576 * ``priority``: Node priority (highest node priority is zero). This is used by
2577 the SP algorithm running on the parent node for scheduling this node.
2578 * ``weight``: Node weight (lowest weight is one). The node weight is relative
2579 to the weight sum of all siblings that have the same priority. It is used by
2580 the WFQ algorithm running on the parent node for scheduling this node.
2581 * ``level_id``: Hiearchy level of the node.
2582 * ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2584 * ``n_sp_priorities``: Number of strict priorities.
2585 * ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2586 * ``n_shared_shapers``: Number of shared shapers.
2587 * ``shared_shaper_id``: Shared shaper id.
2589 Add port traffic management hierarchy leaf node
2590 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2592 Add leaf node to port traffic management hiearchy::
2594 testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \
2595 (priority) (weight) (level_id) (shaper_profile_id) \
2596 (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \
2597 [(shared_shaper_id) (shared_shaper_id) ...] \
2601 * ``parent_node_id``: Node ID of the parent.
2602 * ``priority``: Node priority (highest node priority is zero). This is used by
2603 the SP algorithm running on the parent node for scheduling this node.
2604 * ``weight``: Node weight (lowest weight is one). The node weight is relative
2605 to the weight sum of all siblings that have the same priority. It is used by
2606 the WFQ algorithm running on the parent node for scheduling this node.
2607 * ``level_id``: Hiearchy level of the node.
2608 * ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2610 * ``cman_mode``: Congestion management mode to be enabled for this node.
2611 * ``wred_profile_id``: WRED profile id to be enabled for this node.
2612 * ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2613 * ``n_shared_shapers``: Number of shared shapers.
2614 * ``shared_shaper_id``: Shared shaper id.
2616 Delete port traffic management hierarchy node
2617 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2619 Delete node from port traffic management hiearchy::
2621 testpmd> del port tm node (port_id) (node_id)
2623 Update port traffic management hierarchy parent node
2624 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2626 Update port traffic management hierarchy parent node::
2628 testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \
2631 This function can only be called after the hierarchy commit invocation. Its
2632 success depends on the port support for this operation, as advertised through
2633 the port capability set. This function is valid for all nodes of the traffic
2634 management hierarchy except root node.
2636 Suspend port traffic management hierarchy node
2637 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2639 testpmd> suspend port tm node (port_id) (node_id)
2641 Resume port traffic management hierarchy node
2642 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2644 testpmd> resume port tm node (port_id) (node_id)
2646 Commit port traffic management hierarchy
2647 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2649 Commit the traffic management hierarchy on the port::
2651 testpmd> port tm hierarchy commit (port_id) (clean_on_fail)
2655 * ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function
2656 call failure. On the other hand, hierarchy is preserved when this parameter
2659 Set port traffic management default hierarchy (softnic forwarding mode)
2660 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2662 set the traffic management default hierarchy on the port::
2664 testpmd> set port tm hierarchy default (port_id)
2669 This section details the available filter functions that are available.
2671 Note these functions interface the deprecated legacy filtering framework,
2672 superseded by *rte_flow*. See `Flow rules management`_.
2675 ~~~~~~~~~~~~~~~~~~~~
2677 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
2679 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
2680 ethertype (ether_type) (drop|fwd) queue (queue_id)
2682 The available information parameters are:
2684 * ``port_id``: The port which the Ethertype filter assigned on.
2686 * ``mac_addr``: Compare destination mac address.
2688 * ``mac_ignr``: Ignore destination mac address match.
2690 * ``mac_address``: Destination mac address to match.
2692 * ``ether_type``: The EtherType value want to match,
2693 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
2695 * ``queue_id``: The receive queue associated with this EtherType filter.
2696 It is meaningless when deleting or dropping.
2698 Example, to add/remove an ethertype filter rule::
2700 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
2701 ethertype 0x0806 fwd queue 3
2703 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
2704 ethertype 0x0806 fwd queue 3
2709 Add or delete a 2-tuple filter,
2710 which identifies packets by specific protocol and destination TCP/UDP port
2711 and forwards packets into one of the receive queues::
2713 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
2714 protocol (protocol_value) mask (mask_value) \
2715 tcp_flags (tcp_flags_value) priority (prio_value) \
2718 The available information parameters are:
2720 * ``port_id``: The port which the 2-tuple filter assigned on.
2722 * ``dst_port_value``: Destination port in L4.
2724 * ``protocol_value``: IP L4 protocol.
2726 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
2728 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
2730 * ``prio_value``: Priority of this filter.
2732 * ``queue_id``: The receive queue associated with this 2-tuple filter.
2734 Example, to add/remove an 2tuple filter rule::
2736 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
2737 tcp_flags 0x02 priority 3 queue 3
2739 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
2740 tcp_flags 0x02 priority 3 queue 3
2745 Add or delete a 5-tuple filter,
2746 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
2747 and routes packets into one of the receive queues::
2749 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
2750 (src_address) dst_port (dst_port_value) \
2751 src_port (src_port_value) protocol (protocol_value) \
2752 mask (mask_value) tcp_flags (tcp_flags_value) \
2753 priority (prio_value) queue (queue_id)
2755 The available information parameters are:
2757 * ``port_id``: The port which the 5-tuple filter assigned on.
2759 * ``dst_address``: Destination IP address.
2761 * ``src_address``: Source IP address.
2763 * ``dst_port_value``: TCP/UDP destination port.
2765 * ``src_port_value``: TCP/UDP source port.
2767 * ``protocol_value``: L4 protocol.
2769 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
2771 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
2773 * ``prio_value``: The priority of this filter.
2775 * ``queue_id``: The receive queue associated with this 5-tuple filter.
2777 Example, to add/remove an 5tuple filter rule::
2779 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2780 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2781 flags 0x0 priority 3 queue 3
2783 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2784 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2785 flags 0x0 priority 3 queue 3
2790 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
2792 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
2794 The available information parameters are:
2796 * ``port_id``: The port which the SYN filter assigned on.
2798 * ``high``: This SYN filter has higher priority than other filters.
2800 * ``low``: This SYN filter has lower priority than other filters.
2802 * ``queue_id``: The receive queue associated with this SYN filter
2806 testpmd> syn_filter 0 add priority high queue 3
2811 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
2812 and routed into one of the receive queues::
2814 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
2815 mask (mask_value) priority (prio_value) queue (queue_id)
2817 The available information parameters are:
2819 * ``port_id``: The port which the Flex filter is assigned on.
2821 * ``len_value``: Filter length in bytes, no greater than 128.
2823 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
2825 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
2827 * ``prio_value``: The priority of this filter.
2829 * ``queue_id``: The receive queue associated with this Flex filter.
2833 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
2834 mask 000C priority 3 queue 3
2836 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
2837 mask 000C priority 3 queue 3
2840 .. _testpmd_flow_director:
2842 flow_director_filter
2843 ~~~~~~~~~~~~~~~~~~~~
2845 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
2847 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
2848 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
2850 * Perfect match filters.
2851 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2852 The masked fields are for IP flow.
2854 * Signature filters.
2855 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
2857 * Perfect-mac-vlan match filters.
2858 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2859 The masked fields are for MAC VLAN flow.
2861 * Perfect-tunnel match filters.
2862 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2863 The masked fields are for tunnel flow.
2865 * Perfect-raw-flow-type match filters.
2866 The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet.
2867 The masked fields are specified by input sets.
2869 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
2870 per flow type and the flexible payload.
2872 The Flow Director can also mask out parts of all of these fields so that filters
2873 are only applied to certain fields or parts of the fields.
2875 Note that for raw flow type mode the source and destination fields in the
2876 raw packet buffer need to be presented in a reversed order with respect
2877 to the expected received packets.
2878 For example: IP source and destination addresses or TCP/UDP/SCTP
2879 source and destination ports
2881 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
2883 # Commands to add flow director filters of different flow types::
2885 flow_director_filter (port_id) mode IP (add|del|update) \
2886 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
2887 src (src_ip_address) dst (dst_ip_address) \
2888 tos (tos_value) proto (proto_value) ttl (ttl_value) \
2889 vlan (vlan_value) flexbytes (flexbytes_value) \
2890 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
2893 flow_director_filter (port_id) mode IP (add|del|update) \
2894 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
2895 src (src_ip_address) (src_port) \
2896 dst (dst_ip_address) (dst_port) \
2897 tos (tos_value) ttl (ttl_value) \
2898 vlan (vlan_value) flexbytes (flexbytes_value) \
2899 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
2902 flow_director_filter (port_id) mode IP (add|del|update) \
2903 flow (ipv4-sctp|ipv6-sctp) \
2904 src (src_ip_address) (src_port) \
2905 dst (dst_ip_address) (dst_port) \
2906 tos (tos_value) ttl (ttl_value) \
2907 tag (verification_tag) vlan (vlan_value) \
2908 flexbytes (flexbytes_value) (drop|fwd) \
2909 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
2911 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
2912 ether (ethertype) flexbytes (flexbytes_value) \
2913 (drop|fwd) pf|vf(vf_id) queue (queue_id)
2916 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
2917 mac (mac_address) vlan (vlan_value) \
2918 flexbytes (flexbytes_value) (drop|fwd) \
2919 queue (queue_id) fd_id (fd_id_value)
2921 flow_director_filter (port_id) mode Tunnel (add|del|update) \
2922 mac (mac_address) vlan (vlan_value) \
2923 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
2924 flexbytes (flexbytes_value) (drop|fwd) \
2925 queue (queue_id) fd_id (fd_id_value)
2927 flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \
2928 (drop|fwd) queue (queue_id) fd_id (fd_id_value) \
2929 packet (packet file name)
2931 For example, to add an ipv4-udp flow type filter::
2933 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
2934 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
2935 fwd pf queue 1 fd_id 1
2937 For example, add an ipv4-other flow type filter::
2939 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
2940 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
2941 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
2946 Flush all flow director filters on a device::
2948 testpmd> flush_flow_director (port_id)
2950 Example, to flush all flow director filter on port 0::
2952 testpmd> flush_flow_director 0
2957 Set flow director's input masks::
2959 flow_director_mask (port_id) mode IP vlan (vlan_value) \
2960 src_mask (ipv4_src) (ipv6_src) (src_port) \
2961 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2963 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2965 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2966 mac (mac_value) tunnel-type (tunnel_type_value) \
2967 tunnel-id (tunnel_id_value)
2969 Example, to set flow director mask on port 0::
2971 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2972 src_mask 255.255.255.255 \
2973 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2974 dst_mask 255.255.255.255 \
2975 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2977 flow_director_flex_mask
2978 ~~~~~~~~~~~~~~~~~~~~~~~
2980 set masks of flow director's flexible payload based on certain flow type::
2982 testpmd> flow_director_flex_mask (port_id) \
2983 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2984 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2985 l2_payload|all) (mask)
2987 Example, to set flow director's flex mask for all flow type on port 0::
2989 testpmd> flow_director_flex_mask 0 flow all \
2990 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2993 flow_director_flex_payload
2994 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2996 Configure flexible payload selection::
2998 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
3000 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
3002 testpmd> flow_director_flex_payload 0 l4 \
3003 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
3005 get_sym_hash_ena_per_port
3006 ~~~~~~~~~~~~~~~~~~~~~~~~~
3008 Get symmetric hash enable configuration per port::
3010 get_sym_hash_ena_per_port (port_id)
3012 For example, to get symmetric hash enable configuration of port 1::
3014 testpmd> get_sym_hash_ena_per_port 1
3016 set_sym_hash_ena_per_port
3017 ~~~~~~~~~~~~~~~~~~~~~~~~~
3019 Set symmetric hash enable configuration per port to enable or disable::
3021 set_sym_hash_ena_per_port (port_id) (enable|disable)
3023 For example, to set symmetric hash enable configuration of port 1 to enable::
3025 testpmd> set_sym_hash_ena_per_port 1 enable
3027 get_hash_global_config
3028 ~~~~~~~~~~~~~~~~~~~~~~
3030 Get the global configurations of hash filters::
3032 get_hash_global_config (port_id)
3034 For example, to get the global configurations of hash filters of port 1::
3036 testpmd> get_hash_global_config 1
3038 set_hash_global_config
3039 ~~~~~~~~~~~~~~~~~~~~~~
3041 Set the global configurations of hash filters::
3043 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
3044 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
3045 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \
3048 For example, to enable simple_xor for flow type of ipv6 on port 2::
3050 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
3055 Set the input set for hash::
3057 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
3058 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
3059 l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
3060 ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
3061 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
3062 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
3063 fld-8th|none) (select|add)
3065 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
3067 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
3072 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
3073 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
3075 Set the input set for flow director::
3077 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
3078 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
3079 l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
3080 ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
3081 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
3082 sctp-dst-port|sctp-veri-tag|none) (select|add)
3084 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
3086 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
3091 Set different GRE key length for input set::
3093 global_config (port_id) gre-key-len (number in bytes)
3095 For example to set GRE key length for input set to 4 bytes on port 0::
3097 testpmd> global_config 0 gre-key-len 4
3100 .. _testpmd_rte_flow:
3102 Flow rules management
3103 ---------------------
3105 Control of the generic flow API (*rte_flow*) is fully exposed through the
3106 ``flow`` command (validation, creation, destruction, queries and operation
3109 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
3110 features simultaneously may cause undefined side-effects and is therefore
3116 Because the ``flow`` command uses dynamic tokens to handle the large number
3117 of possible flow rules combinations, its behavior differs slightly from
3118 other commands, in particular:
3120 - Pressing *?* or the *<tab>* key displays contextual help for the current
3121 token, not that of the entire command.
3123 - Optional and repeated parameters are supported (provided they are listed
3124 in the contextual help).
3126 The first parameter stands for the operation mode. Possible operations and
3127 their general syntax are described below. They are covered in detail in the
3130 - Check whether a flow rule can be created::
3132 flow validate {port_id}
3133 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3134 pattern {item} [/ {item} [...]] / end
3135 actions {action} [/ {action} [...]] / end
3137 - Create a flow rule::
3139 flow create {port_id}
3140 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3141 pattern {item} [/ {item} [...]] / end
3142 actions {action} [/ {action} [...]] / end
3144 - Destroy specific flow rules::
3146 flow destroy {port_id} rule {rule_id} [...]
3148 - Destroy all flow rules::
3150 flow flush {port_id}
3152 - Query an existing flow rule::
3154 flow query {port_id} {rule_id} {action}
3156 - List existing flow rules sorted by priority, filtered by group
3159 flow list {port_id} [group {group_id}] [...]
3161 - Restrict ingress traffic to the defined flow rules::
3163 flow isolate {port_id} {boolean}
3165 Validating flow rules
3166 ~~~~~~~~~~~~~~~~~~~~~
3168 ``flow validate`` reports whether a flow rule would be accepted by the
3169 underlying device in its current state but stops short of creating it. It is
3170 bound to ``rte_flow_validate()``::
3172 flow validate {port_id}
3173 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3174 pattern {item} [/ {item} [...]] / end
3175 actions {action} [/ {action} [...]] / end
3177 If successful, it will show::
3181 Otherwise it will show an error message of the form::
3183 Caught error type [...] ([...]): [...]
3185 This command uses the same parameters as ``flow create``, their format is
3186 described in `Creating flow rules`_.
3188 Check whether redirecting any Ethernet packet received on port 0 to RX queue
3189 index 6 is supported::
3191 testpmd> flow validate 0 ingress pattern eth / end
3192 actions queue index 6 / end
3196 Port 0 does not support TCPv6 rules::
3198 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
3200 Caught error type 9 (specific pattern item): Invalid argument
3206 ``flow create`` validates and creates the specified flow rule. It is bound
3207 to ``rte_flow_create()``::
3209 flow create {port_id}
3210 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3211 pattern {item} [/ {item} [...]] / end
3212 actions {action} [/ {action} [...]] / end
3214 If successful, it will return a flow rule ID usable with other commands::
3216 Flow rule #[...] created
3218 Otherwise it will show an error message of the form::
3220 Caught error type [...] ([...]): [...]
3222 Parameters describe in the following order:
3224 - Attributes (*group*, *priority*, *ingress*, *egress*, *transfer* tokens).
3225 - A matching pattern, starting with the *pattern* token and terminated by an
3227 - Actions, starting with the *actions* token and terminated by an *end*
3230 These translate directly to *rte_flow* objects provided as-is to the
3231 underlying functions.
3233 The shortest valid definition only comprises mandatory tokens::
3235 testpmd> flow create 0 pattern end actions end
3237 Note that PMDs may refuse rules that essentially do nothing such as this
3240 **All unspecified object values are automatically initialized to 0.**
3245 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
3246 specified before the ``pattern`` token.
3248 - ``group {group id}``: priority group.
3249 - ``priority {level}``: priority level within group.
3250 - ``ingress``: rule applies to ingress traffic.
3251 - ``egress``: rule applies to egress traffic.
3252 - ``transfer``: apply rule directly to endpoints found in pattern.
3254 Each instance of an attribute specified several times overrides the previous
3255 value as shown below (group 4 is used)::
3257 testpmd> flow create 0 group 42 group 24 group 4 [...]
3259 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
3261 While not specifying a direction is an error, some rules may allow both
3264 Most rules affect RX therefore contain the ``ingress`` token::
3266 testpmd> flow create 0 ingress pattern [...]
3271 A matching pattern starts after the ``pattern`` token. It is made of pattern
3272 items and is terminated by a mandatory ``end`` item.
3274 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
3275 rte_flow_item_type``).
3277 The ``/`` token is used as a separator between pattern items as shown
3280 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
3282 Note that protocol items like these must be stacked from lowest to highest
3283 layer to make sense. For instance, the following rule is either invalid or
3284 unlikely to match any packet::
3286 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
3288 More information on these restrictions can be found in the *rte_flow*
3291 Several items support additional specification structures, for example
3292 ``ipv4`` allows specifying source and destination addresses as follows::
3294 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
3295 dst is 10.2.0.0 / end [...]
3297 This rule matches all IPv4 traffic with the specified properties.
3299 In this example, ``src`` and ``dst`` are field names of the underlying
3300 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
3301 in a similar fashion.
3303 The ``is`` token means that the subsequent value must be matched exactly,
3304 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
3305 accordingly. Possible assignment tokens are:
3307 - ``is``: match value perfectly (with full bit-mask).
3308 - ``spec``: match value according to configured bit-mask.
3309 - ``last``: specify upper bound to establish a range.
3310 - ``mask``: specify bit-mask with relevant bits set to one.
3311 - ``prefix``: generate bit-mask from a prefix length.
3313 These yield identical results::
3315 ipv4 src is 10.1.1.1
3319 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
3323 ipv4 src spec 10.1.1.1 src prefix 32
3327 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
3331 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
3333 Inclusive ranges can be defined with ``last``::
3335 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
3337 Note that ``mask`` affects both ``spec`` and ``last``::
3339 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
3340 # matches 10.1.0.0 to 10.2.255.255
3342 Properties can be modified multiple times::
3344 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
3348 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
3353 This section lists supported pattern items and their attributes, if any.
3355 - ``end``: end list of pattern items.
3357 - ``void``: no-op pattern item.
3359 - ``invert``: perform actions when pattern does not match.
3361 - ``any``: match any protocol for the current layer.
3363 - ``num {unsigned}``: number of layers covered.
3365 - ``pf``: match traffic from/to the physical function.
3367 - ``vf``: match traffic from/to a virtual function ID.
3369 - ``id {unsigned}``: VF ID.
3371 - ``phy_port``: match traffic from/to a specific physical port.
3373 - ``index {unsigned}``: physical port index.
3375 - ``port_id``: match traffic from/to a given DPDK port ID.
3377 - ``id {unsigned}``: DPDK port ID.
3379 - ``mark``: match value set in previously matched flow rule using the mark action.
3381 - ``id {unsigned}``: arbitrary integer value.
3383 - ``raw``: match an arbitrary byte string.
3385 - ``relative {boolean}``: look for pattern after the previous item.
3386 - ``search {boolean}``: search pattern from offset (see also limit).
3387 - ``offset {integer}``: absolute or relative offset for pattern.
3388 - ``limit {unsigned}``: search area limit for start of pattern.
3389 - ``pattern {string}``: byte string to look for.
3391 - ``eth``: match Ethernet header.
3393 - ``dst {MAC-48}``: destination MAC.
3394 - ``src {MAC-48}``: source MAC.
3395 - ``type {unsigned}``: EtherType or TPID.
3397 - ``vlan``: match 802.1Q/ad VLAN tag.
3399 - ``tci {unsigned}``: tag control information.
3400 - ``pcp {unsigned}``: priority code point.
3401 - ``dei {unsigned}``: drop eligible indicator.
3402 - ``vid {unsigned}``: VLAN identifier.
3403 - ``inner_type {unsigned}``: inner EtherType or TPID.
3405 - ``ipv4``: match IPv4 header.
3407 - ``tos {unsigned}``: type of service.
3408 - ``ttl {unsigned}``: time to live.
3409 - ``proto {unsigned}``: next protocol ID.
3410 - ``src {ipv4 address}``: source address.
3411 - ``dst {ipv4 address}``: destination address.
3413 - ``ipv6``: match IPv6 header.
3415 - ``tc {unsigned}``: traffic class.
3416 - ``flow {unsigned}``: flow label.
3417 - ``proto {unsigned}``: protocol (next header).
3418 - ``hop {unsigned}``: hop limit.
3419 - ``src {ipv6 address}``: source address.
3420 - ``dst {ipv6 address}``: destination address.
3422 - ``icmp``: match ICMP header.
3424 - ``type {unsigned}``: ICMP packet type.
3425 - ``code {unsigned}``: ICMP packet code.
3427 - ``udp``: match UDP header.
3429 - ``src {unsigned}``: UDP source port.
3430 - ``dst {unsigned}``: UDP destination port.
3432 - ``tcp``: match TCP header.
3434 - ``src {unsigned}``: TCP source port.
3435 - ``dst {unsigned}``: TCP destination port.
3437 - ``sctp``: match SCTP header.
3439 - ``src {unsigned}``: SCTP source port.
3440 - ``dst {unsigned}``: SCTP destination port.
3441 - ``tag {unsigned}``: validation tag.
3442 - ``cksum {unsigned}``: checksum.
3444 - ``vxlan``: match VXLAN header.
3446 - ``vni {unsigned}``: VXLAN identifier.
3448 - ``e_tag``: match IEEE 802.1BR E-Tag header.
3450 - ``grp_ecid_b {unsigned}``: GRP and E-CID base.
3452 - ``nvgre``: match NVGRE header.
3454 - ``tni {unsigned}``: virtual subnet ID.
3456 - ``mpls``: match MPLS header.
3458 - ``label {unsigned}``: MPLS label.
3460 - ``gre``: match GRE header.
3462 - ``protocol {unsigned}``: protocol type.
3464 - ``fuzzy``: fuzzy pattern match, expect faster than default.
3466 - ``thresh {unsigned}``: accuracy threshold.
3468 - ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header.
3470 - ``teid {unsigned}``: tunnel endpoint identifier.
3472 - ``geneve``: match GENEVE header.
3474 - ``vni {unsigned}``: virtual network identifier.
3475 - ``protocol {unsigned}``: protocol type.
3477 - ``vxlan-gpe``: match VXLAN-GPE header.
3479 - ``vni {unsigned}``: VXLAN-GPE identifier.
3481 - ``arp_eth_ipv4``: match ARP header for Ethernet/IPv4.
3483 - ``sha {MAC-48}``: sender hardware address.
3484 - ``spa {ipv4 address}``: sender IPv4 address.
3485 - ``tha {MAC-48}``: target hardware address.
3486 - ``tpa {ipv4 address}``: target IPv4 address.
3488 - ``ipv6_ext``: match presence of any IPv6 extension header.
3490 - ``next_hdr {unsigned}``: next header.
3492 - ``icmp6``: match any ICMPv6 header.
3494 - ``type {unsigned}``: ICMPv6 type.
3495 - ``code {unsigned}``: ICMPv6 code.
3497 - ``icmp6_nd_ns``: match ICMPv6 neighbor discovery solicitation.
3499 - ``target_addr {ipv6 address}``: target address.
3501 - ``icmp6_nd_na``: match ICMPv6 neighbor discovery advertisement.
3503 - ``target_addr {ipv6 address}``: target address.
3505 - ``icmp6_nd_opt``: match presence of any ICMPv6 neighbor discovery option.
3507 - ``type {unsigned}``: ND option type.
3509 - ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery source Ethernet
3510 link-layer address option.
3512 - ``sla {MAC-48}``: source Ethernet LLA.
3514 - ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery target Ethernet
3515 link-layer address option.
3517 - ``tla {MAC-48}``: target Ethernet LLA.
3522 A list of actions starts after the ``actions`` token in the same fashion as
3523 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
3524 terminated by a mandatory ``end`` action.
3526 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
3527 rte_flow_action_type``).
3529 Dropping all incoming UDPv4 packets can be expressed as follows::
3531 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3534 Several actions have configurable properties which must be specified when
3535 there is no valid default value. For example, ``queue`` requires a target
3538 This rule redirects incoming UDPv4 traffic to queue index 6::
3540 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3541 actions queue index 6 / end
3543 While this one could be rejected by PMDs (unspecified queue index)::
3545 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3548 As defined by *rte_flow*, the list is not ordered, all actions of a given
3549 rule are performed simultaneously. These are equivalent::
3551 queue index 6 / void / mark id 42 / end
3555 void / mark id 42 / queue index 6 / end
3557 All actions in a list should have different types, otherwise only the last
3558 action of a given type is taken into account::
3560 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
3564 drop / drop / drop / end # drop is performed only once
3568 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
3570 Considering they are performed simultaneously, opposite and overlapping
3571 actions can sometimes be combined when the end result is unambiguous::
3573 drop / queue index 6 / end # drop has no effect
3577 queue index 6 / rss queues 6 7 8 / end # queue has no effect
3581 drop / passthru / end # drop has no effect
3583 Note that PMDs may still refuse such combinations.
3588 This section lists supported actions and their attributes, if any.
3590 - ``end``: end list of actions.
3592 - ``void``: no-op action.
3594 - ``passthru``: let subsequent rule process matched packets.
3596 - ``jump``: redirect traffic to group on device.
3598 - ``group {unsigned}``: group to redirect to.
3600 - ``mark``: attach 32 bit value to packets.
3602 - ``id {unsigned}``: 32 bit value to return with packets.
3604 - ``flag``: flag packets.
3606 - ``queue``: assign packets to a given queue index.
3608 - ``index {unsigned}``: queue index to use.
3610 - ``drop``: drop packets (note: passthru has priority).
3612 - ``count``: enable counters for this rule.
3614 - ``rss``: spread packets among several queues.
3616 - ``func {hash function}``: RSS hash function to apply, allowed tokens are
3617 the same as `set_hash_global_config`_.
3619 - ``level {unsigned}``: encapsulation level for ``types``.
3621 - ``types [{RSS hash type} [...]] end``: specific RSS hash types, allowed
3622 tokens are the same as `set_hash_input_set`_, except that an empty list
3623 does not disable RSS but instead requests unspecified "best-effort"
3626 - ``key {string}``: RSS hash key, overrides ``key_len``.
3628 - ``key_len {unsigned}``: RSS hash key length in bytes, can be used in
3629 conjunction with ``key`` to pad or truncate it.
3631 - ``queues [{unsigned} [...]] end``: queue indices to use.
3633 - ``pf``: direct traffic to physical function.
3635 - ``vf``: direct traffic to a virtual function ID.
3637 - ``original {boolean}``: use original VF ID if possible.
3638 - ``id {unsigned}``: VF ID.
3640 - ``phy_port``: direct packets to physical port index.
3642 - ``original {boolean}``: use original port index if possible.
3643 - ``index {unsigned}``: physical port index.
3645 - ``port_id``: direct matching traffic to a given DPDK port ID.
3647 - ``original {boolean}``: use original DPDK port ID if possible.
3648 - ``id {unsigned}``: DPDK port ID.
3650 - ``of_set_mpls_ttl``: OpenFlow's ``OFPAT_SET_MPLS_TTL``.
3652 - ``mpls_ttl``: MPLS TTL.
3654 - ``of_dec_mpls_ttl``: OpenFlow's ``OFPAT_DEC_MPLS_TTL``.
3656 - ``of_set_nw_ttl``: OpenFlow's ``OFPAT_SET_NW_TTL``.
3658 - ``nw_ttl``: IP TTL.
3660 - ``of_dec_nw_ttl``: OpenFlow's ``OFPAT_DEC_NW_TTL``.
3662 - ``of_copy_ttl_out``: OpenFlow's ``OFPAT_COPY_TTL_OUT``.
3664 - ``of_copy_ttl_in``: OpenFlow's ``OFPAT_COPY_TTL_IN``.
3666 - ``of_pop_vlan``: OpenFlow's ``OFPAT_POP_VLAN``.
3668 - ``of_push_vlan``: OpenFlow's ``OFPAT_PUSH_VLAN``.
3670 - ``ethertype``: Ethertype.
3672 - ``of_set_vlan_vid``: OpenFlow's ``OFPAT_SET_VLAN_VID``.
3674 - ``vlan_vid``: VLAN id.
3676 - ``of_set_vlan_pcp``: OpenFlow's ``OFPAT_SET_VLAN_PCP``.
3678 - ``vlan_pcp``: VLAN priority.
3680 - ``of_pop_mpls``: OpenFlow's ``OFPAT_POP_MPLS``.
3682 - ``ethertype``: Ethertype.
3684 - ``of_push_mpls``: OpenFlow's ``OFPAT_PUSH_MPLS``.
3686 - ``ethertype``: Ethertype.
3688 - ``vxlan_encap``: Performs a VXLAN encapsulation, outer layer configuration
3689 is done through `Config VXLAN Encap outer layers`_.
3691 - ``vxlan_decap``: Performs a decapsulation action by stripping all headers of
3692 the VXLAN tunnel network overlay from the matched flow.
3694 - ``nvgre_encap``: Performs a NVGRE encapsulation, outer layer configuration
3695 is done through `Config NVGRE Encap outer layers`_.
3697 - ``nvgre_decap``: Performs a decapsulation action by stripping all headers of
3698 the NVGRE tunnel network overlay from the matched flow.
3700 Destroying flow rules
3701 ~~~~~~~~~~~~~~~~~~~~~
3703 ``flow destroy`` destroys one or more rules from their rule ID (as returned
3704 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
3705 times as necessary::
3707 flow destroy {port_id} rule {rule_id} [...]
3709 If successful, it will show::
3711 Flow rule #[...] destroyed
3713 It does not report anything for rule IDs that do not exist. The usual error
3714 message is shown when a rule cannot be destroyed::
3716 Caught error type [...] ([...]): [...]
3718 ``flow flush`` destroys all rules on a device and does not take extra
3719 arguments. It is bound to ``rte_flow_flush()``::
3721 flow flush {port_id}
3723 Any errors are reported as above.
3725 Creating several rules and destroying them::
3727 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3728 actions queue index 2 / end
3729 Flow rule #0 created
3730 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3731 actions queue index 3 / end
3732 Flow rule #1 created
3733 testpmd> flow destroy 0 rule 0 rule 1
3734 Flow rule #1 destroyed
3735 Flow rule #0 destroyed
3738 The same result can be achieved using ``flow flush``::
3740 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3741 actions queue index 2 / end
3742 Flow rule #0 created
3743 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3744 actions queue index 3 / end
3745 Flow rule #1 created
3746 testpmd> flow flush 0
3749 Non-existent rule IDs are ignored::
3751 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3752 actions queue index 2 / end
3753 Flow rule #0 created
3754 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3755 actions queue index 3 / end
3756 Flow rule #1 created
3757 testpmd> flow destroy 0 rule 42 rule 10 rule 2
3759 testpmd> flow destroy 0 rule 0
3760 Flow rule #0 destroyed
3766 ``flow query`` queries a specific action of a flow rule having that
3767 ability. Such actions collect information that can be reported using this
3768 command. It is bound to ``rte_flow_query()``::
3770 flow query {port_id} {rule_id} {action}
3772 If successful, it will display either the retrieved data for known actions
3773 or the following message::
3775 Cannot display result for action type [...] ([...])
3777 Otherwise, it will complain either that the rule does not exist or that some
3780 Flow rule #[...] not found
3784 Caught error type [...] ([...]): [...]
3786 Currently only the ``count`` action is supported. This action reports the
3787 number of packets that hit the flow rule and the total number of bytes. Its
3788 output has the following format::
3791 hits_set: [...] # whether "hits" contains a valid value
3792 bytes_set: [...] # whether "bytes" contains a valid value
3793 hits: [...] # number of packets
3794 bytes: [...] # number of bytes
3796 Querying counters for TCPv6 packets redirected to queue 6::
3798 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
3799 actions queue index 6 / count / end
3800 Flow rule #4 created
3801 testpmd> flow query 0 4 count
3812 ``flow list`` lists existing flow rules sorted by priority and optionally
3813 filtered by group identifiers::
3815 flow list {port_id} [group {group_id}] [...]
3817 This command only fails with the following message if the device does not
3822 Output consists of a header line followed by a short description of each
3823 flow rule, one per line. There is no output at all when no flow rules are
3824 configured on the device::
3826 ID Group Prio Attr Rule
3827 [...] [...] [...] [...] [...]
3829 ``Attr`` column flags:
3831 - ``i`` for ``ingress``.
3832 - ``e`` for ``egress``.
3834 Creating several flow rules and listing them::
3836 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3837 actions queue index 6 / end
3838 Flow rule #0 created
3839 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3840 actions queue index 2 / end
3841 Flow rule #1 created
3842 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
3843 actions rss queues 6 7 8 end / end
3844 Flow rule #2 created
3845 testpmd> flow list 0
3846 ID Group Prio Attr Rule
3847 0 0 0 i- ETH IPV4 => QUEUE
3848 1 0 0 i- ETH IPV6 => QUEUE
3849 2 0 5 i- ETH IPV4 UDP => RSS
3852 Rules are sorted by priority (i.e. group ID first, then priority level)::
3854 testpmd> flow list 1
3855 ID Group Prio Attr Rule
3856 0 0 0 i- ETH => COUNT
3857 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3858 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3859 1 24 0 i- ETH IPV4 UDP => QUEUE
3860 4 24 10 i- ETH IPV4 TCP => DROP
3861 3 24 20 i- ETH IPV4 => DROP
3862 2 24 42 i- ETH IPV4 UDP => QUEUE
3863 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3866 Output can be limited to specific groups::
3868 testpmd> flow list 1 group 0 group 63
3869 ID Group Prio Attr Rule
3870 0 0 0 i- ETH => COUNT
3871 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3872 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3873 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3876 Toggling isolated mode
3877 ~~~~~~~~~~~~~~~~~~~~~~
3879 ``flow isolate`` can be used to tell the underlying PMD that ingress traffic
3880 must only be injected from the defined flow rules; that no default traffic
3881 is expected outside those rules and the driver is free to assign more
3882 resources to handle them. It is bound to ``rte_flow_isolate()``::
3884 flow isolate {port_id} {boolean}
3886 If successful, enabling or disabling isolated mode shows either::
3888 Ingress traffic on port [...]
3889 is now restricted to the defined flow rules
3893 Ingress traffic on port [...]
3894 is not restricted anymore to the defined flow rules
3896 Otherwise, in case of error::
3898 Caught error type [...] ([...]): [...]
3900 Mainly due to its side effects, PMDs supporting this mode may not have the
3901 ability to toggle it more than once without reinitializing affected ports
3902 first (e.g. by exiting testpmd).
3904 Enabling isolated mode::
3906 testpmd> flow isolate 0 true
3907 Ingress traffic on port 0 is now restricted to the defined flow rules
3910 Disabling isolated mode::
3912 testpmd> flow isolate 0 false
3913 Ingress traffic on port 0 is not restricted anymore to the defined flow rules
3916 Sample QinQ flow rules
3917 ~~~~~~~~~~~~~~~~~~~~~~
3919 Before creating QinQ rule(s) the following commands should be issued to enable QinQ::
3921 testpmd> port stop 0
3922 testpmd> vlan set qinq on 0
3924 The above command sets the inner and outer TPID's to 0x8100.
3926 To change the TPID's the following commands should be used::
3928 testpmd> vlan set outer tpid 0xa100 0
3929 testpmd> vlan set inner tpid 0x9100 0
3930 testpmd> port start 0
3932 Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM.
3936 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 /
3937 vlan tci is 456 / end actions vf id 1 / queue index 0 / end
3938 Flow rule #0 validated
3940 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 /
3941 vlan tci is 456 / end actions vf id 123 / queue index 0 / end
3942 Flow rule #0 created
3944 testpmd> flow list 0
3945 ID Group Prio Attr Rule
3946 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3948 Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host.
3952 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 /
3953 vlan tci is 654 / end actions pf / queue index 0 / end
3954 Flow rule #1 validated
3956 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 /
3957 vlan tci is 654 / end actions pf / queue index 1 / end
3958 Flow rule #1 created
3960 testpmd> flow list 0
3961 ID Group Prio Attr Rule
3962 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3963 1 0 0 i- ETH VLAN VLAN=>PF QUEUE
3965 Sample VXLAN encapsulation rule
3966 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
3968 VXLAN encapsulation outer layer has default value pre-configured in testpmd
3969 source code, those can be changed by using the following commands
3971 IPv4 VXLAN outer header::
3973 testpmd> set vxlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src 127.0.0.1
3974 ip-dst 128.0.0.1 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
3975 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
3978 testpmd> set vxlan-with-vlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src
3979 127.0.0.1 ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11
3980 eth-dst 22:22:22:22:22:22
3981 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
3984 IPv6 VXLAN outer header::
3986 testpmd> set vxlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4 ip-src ::1
3987 ip-dst ::2222 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
3988 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
3991 testpmd> set vxlan-with-vlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4
3992 ip-src ::1 ip-dst ::2222 vlan-tci 34 eth-src 11:11:11:11:11:11
3993 eth-dst 22:22:22:22:22:22
3994 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
3997 Sample NVGRE encapsulation rule
3998 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
4000 NVGRE encapsulation outer layer has default value pre-configured in testpmd
4001 source code, those can be changed by using the following commands
4003 IPv4 NVGRE outer header::
4005 testpmd> set nvgre ip-version ipv4 tni 4 ip-src 127.0.0.1 ip-dst 128.0.0.1
4006 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
4007 testpmd> flow create 0 ingress pattern end actions nvgre_encap /
4010 testpmd> set nvgre-with-vlan ip-version ipv4 tni 4 ip-src 127.0.0.1
4011 ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11
4012 eth-dst 22:22:22:22:22:22
4013 testpmd> flow create 0 ingress pattern end actions nvgre_encap /
4016 IPv6 NVGRE outer header::
4018 testpmd> set nvgre ip-version ipv6 tni 4 ip-src ::1 ip-dst ::2222
4019 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
4020 testpmd> flow create 0 ingress pattern end actions nvgre_encap /
4023 testpmd> set nvgre-with-vlan ip-version ipv6 tni 4 ip-src ::1 ip-dst ::2222
4024 vlan-tci 34 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
4025 testpmd> flow create 0 ingress pattern end actions nvgre_encap /
4031 The following sections show functions to load/unload eBPF based filters.
4036 Load an eBPF program as a callback for partciular RX/TX queue::
4038 testpmd> bpf-load rx|tx (portid) (queueid) (load-flags) (bpf-prog-filename)
4040 The available load-flags are:
4042 * ``J``: use JIT generated native code, otherwise BPF interpreter will be used.
4044 * ``M``: assume input parameter is a pointer to rte_mbuf, otherwise assume it is a pointer to first segment's data.
4050 You'll need clang v3.7 or above to build bpf program you'd like to load
4054 .. code-block:: console
4057 clang -O2 -target bpf -c t1.c
4059 Then to load (and JIT compile) t1.o at RX queue 0, port 1::
4061 .. code-block:: console
4063 testpmd> bpf-load rx 1 0 J ./dpdk.org/test/bpf/t1.o
4065 To load (not JITed) t1.o at TX queue 0, port 0::
4067 .. code-block:: console
4069 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o
4074 Unload previously loaded eBPF program for partciular RX/TX queue::
4076 testpmd> bpf-unload rx|tx (portid) (queueid)
4078 For example to unload BPF filter from TX queue 0, port 0:
4080 .. code-block:: console
4082 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o