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 [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|outer-udp) (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.
870 * ``outer-udp`` relates to the outer UDP layer in the case where the packet is recognized
871 as a tunnel packet by the forwarding engine (vxlan, vxlan-gpe are
872 supported). See also the ``csum parse-tunnel`` command.
876 Check the NIC Datasheet for hardware limits.
881 Set RSS queue region span on a port::
883 testpmd> set port (port_id) queue-region region_id (value) \
884 queue_start_index (value) queue_num (value)
886 Set flowtype mapping on a RSS queue region on a port::
888 testpmd> set port (port_id) queue-region region_id (value) flowtype (value)
892 * For the flowtype(pctype) of packet,the specific index for each type has
893 been defined in file i40e_type.h as enum i40e_filter_pctype.
895 Set user priority mapping on a RSS queue region on a port::
897 testpmd> set port (port_id) queue-region UP (value) region_id (value)
899 Flush all queue region related configuration on a port::
901 testpmd> set port (port_id) queue-region flush (on|off)
905 * "on"is just an enable function which server for other configuration,
906 it is for all configuration about queue region from up layer,
907 at first will only keep in DPDK softwarestored in driver,
908 only after "flush on", it commit all configuration to HW.
909 "off" is just clean all configuration about queue region just now,
910 and restore all to DPDK i40e driver default config when start up.
912 Show all queue region related configuration info on a port::
914 testpmd> show port (port_id) queue-region
918 Queue region only support on PF by now, so these command is
919 only for configuration of queue region on PF port.
924 Define how tunneled packets should be handled by the csum forward
927 testpmd> csum parse-tunnel (on|off) (tx_port_id)
929 If enabled, the csum forward engine will try to recognize supported
930 tunnel headers (vxlan, gre, ipip).
932 If disabled, treat tunnel packets as non-tunneled packets (a inner
933 header is handled as a packet payload).
937 The port argument is the TX port like in the ``csum set`` command.
941 Consider a packet in packet like the following::
943 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in
945 * If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
946 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the
947 ``outer-ip|outer-udp`` parameter relates to the outer headers (here ``ipv4_out`` and ``udp_out``).
949 * If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
950 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``.
955 Display tx checksum offload configuration::
957 testpmd> csum show (port_id)
962 Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine::
964 testpmd> tso set (segsize) (port_id)
968 Check the NIC datasheet for hardware limits.
973 Display the status of TCP Segmentation Offload::
975 testpmd> tso show (port_id)
980 Enable or disable GRO in ``csum`` forwarding engine::
982 testpmd> set port <port_id> gro on|off
984 If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4
985 packets received from the given port.
987 If disabled, packets received from the given port won't be performed
988 GRO. By default, GRO is disabled for all ports.
992 When enable GRO for a port, TCP/IPv4 packets received from the port
993 will be performed GRO. After GRO, all merged packets have bad
994 checksums, since the GRO library doesn't re-calculate checksums for
995 the merged packets. Therefore, if users want the merged packets to
996 have correct checksums, please select HW IP checksum calculation and
997 HW TCP checksum calculation for the port which the merged packets are
1003 Display GRO configuration for a given port::
1005 testpmd> show port <port_id> gro
1010 Set the cycle to flush the GROed packets from reassembly tables::
1012 testpmd> set gro flush <cycles>
1014 When enable GRO, the csum forwarding engine performs GRO on received
1015 packets, and the GROed packets are stored in reassembly tables. Users
1016 can use this command to determine when the GROed packets are flushed
1017 from the reassembly tables.
1019 The ``cycles`` is measured in GRO operation times. The csum forwarding
1020 engine flushes the GROed packets from the tables every ``cycles`` GRO
1023 By default, the value of ``cycles`` is 1, which means flush GROed packets
1024 from the reassembly tables as soon as one GRO operation finishes. The value
1025 of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``.
1027 Please note that the large value of ``cycles`` may cause the poor TCP/IP
1028 stack performance. Because the GROed packets are delayed to arrive the
1029 stack, thus causing more duplicated ACKs and TCP retransmissions.
1034 Toggle per-port GSO support in ``csum`` forwarding engine::
1036 testpmd> set port <port_id> gso on|off
1038 If enabled, the csum forwarding engine will perform GSO on supported IPv4
1039 packets, transmitted on the given port.
1041 If disabled, packets transmitted on the given port will not undergo GSO.
1042 By default, GSO is disabled for all ports.
1046 When GSO is enabled on a port, supported IPv4 packets transmitted on that
1047 port undergo GSO. Afterwards, the segmented packets are represented by
1048 multi-segment mbufs; however, the csum forwarding engine doesn't calculation
1049 of checksums for GSO'd segments in SW. As a result, if users want correct
1050 checksums in GSO segments, they should enable HW checksum calculation for
1053 For example, HW checksum calculation for VxLAN GSO'd packets may be enabled
1054 by setting the following options in the csum forwarding engine:
1056 testpmd> csum set outer_ip hw <port_id>
1058 testpmd> csum set ip hw <port_id>
1060 testpmd> csum set tcp hw <port_id>
1062 UDP GSO is the same as IP fragmentation, which treats the UDP header
1063 as the payload and does not modify it during segmentation. That is,
1064 after UDP GSO, only the first output fragment has the original UDP
1065 header. Therefore, users need to enable HW IP checksum calculation
1066 and SW UDP checksum calculation for GSO-enabled ports, if they want
1067 correct checksums for UDP/IPv4 packets.
1072 Set the maximum GSO segment size (measured in bytes), which includes the
1073 packet header and the packet payload for GSO-enabled ports (global)::
1075 testpmd> set gso segsz <length>
1080 Display the status of Generic Segmentation Offload for a given port::
1082 testpmd> show port <port_id> gso
1087 Add an alternative MAC address to a port::
1089 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX)
1094 Remove a MAC address from a port::
1096 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX)
1098 mac_addr add (for VF)
1099 ~~~~~~~~~~~~~~~~~~~~~
1101 Add an alternative MAC address for a VF to a port::
1103 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX)
1108 Set the default MAC address for a port::
1110 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX)
1112 mac_addr set (for VF)
1113 ~~~~~~~~~~~~~~~~~~~~~
1115 Set the MAC address for a VF from the PF::
1117 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX)
1122 Set the forwarding peer address for certain port::
1124 testpmd> set eth-peer (port_id) (perr_addr)
1126 This is equivalent to the ``--eth-peer`` command-line option.
1131 Set the unicast hash filter(s) on/off for a port::
1133 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
1138 Set the promiscuous mode on for a port or for all ports.
1139 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1141 testpmd> set promisc (port_id|all) (on|off)
1146 Set the allmulti mode for a port or for all ports::
1148 testpmd> set allmulti (port_id|all) (on|off)
1150 Same as the ifconfig (8) option. Controls how multicast packets are handled.
1152 set promisc (for VF)
1153 ~~~~~~~~~~~~~~~~~~~~
1155 Set the unicast promiscuous mode for a VF from PF.
1156 It's supported by Intel i40e NICs now.
1157 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1159 testpmd> set vf promisc (port_id) (vf_id) (on|off)
1161 set allmulticast (for VF)
1162 ~~~~~~~~~~~~~~~~~~~~~~~~~
1164 Set the multicast promiscuous mode for a VF from PF.
1165 It's supported by Intel i40e NICs now.
1166 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1168 testpmd> set vf allmulti (port_id) (vf_id) (on|off)
1170 set tx max bandwidth (for VF)
1171 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1173 Set TX max absolute bandwidth (Mbps) for a VF from PF::
1175 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth)
1177 set tc tx min bandwidth (for VF)
1178 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1180 Set all TCs' TX min relative bandwidth (%) for a VF from PF::
1182 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...)
1184 set tc tx max bandwidth (for VF)
1185 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1187 Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF::
1189 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth)
1191 set tc strict link priority mode
1192 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1194 Set some TCs' strict link priority mode on a physical port::
1196 testpmd> set tx strict-link-priority (port_id) (tc_bitmap)
1198 set tc tx min bandwidth
1199 ~~~~~~~~~~~~~~~~~~~~~~~
1201 Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs::
1203 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...)
1208 Set the link flow control parameter on a port::
1210 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1211 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
1212 autoneg (on|off) (port_id)
1216 * ``high_water`` (integer): High threshold value to trigger XOFF.
1218 * ``low_water`` (integer): Low threshold value to trigger XON.
1220 * ``pause_time`` (integer): Pause quota in the Pause frame.
1222 * ``send_xon`` (0/1): Send XON frame.
1224 * ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
1226 * ``autoneg``: Change the auto-negotiation parameter.
1231 Set the priority flow control parameter on a port::
1233 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1234 (pause_time) (priority) (port_id)
1238 * ``high_water`` (integer): High threshold value.
1240 * ``low_water`` (integer): Low threshold value.
1242 * ``pause_time`` (integer): Pause quota in the Pause frame.
1244 * ``priority`` (0-7): VLAN User Priority.
1249 Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
1251 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
1253 For example, to set rx queue 2 on port 0 to mapping 5::
1255 testpmd>set stat_qmap rx 0 2 5
1257 set xstats-hide-zero
1258 ~~~~~~~~~~~~~~~~~~~~
1260 Set the option to hide zero values for xstats display::
1262 testpmd> set xstats-hide-zero on|off
1266 By default, the zero values are displayed for xstats.
1268 set port - rx/tx (for VF)
1269 ~~~~~~~~~~~~~~~~~~~~~~~~~
1271 Set VF receive/transmit from a port::
1273 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
1275 set port - mac address filter (for VF)
1276 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1278 Add/Remove unicast or multicast MAC addr filter for a VF::
1280 testpmd> set port (port_id) vf (vf_id) (mac_addr) \
1281 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
1283 set port - rx mode(for VF)
1284 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1286 Set the VF receive mode of a port::
1288 testpmd> set port (port_id) vf (vf_id) \
1289 rxmode (AUPE|ROPE|BAM|MPE) (on|off)
1291 The available receive modes are:
1293 * ``AUPE``: Accepts untagged VLAN.
1295 * ``ROPE``: Accepts unicast hash.
1297 * ``BAM``: Accepts broadcast packets.
1299 * ``MPE``: Accepts all multicast packets.
1301 set port - tx_rate (for Queue)
1302 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1304 Set TX rate limitation for a queue on a port::
1306 testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1308 set port - tx_rate (for VF)
1309 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
1311 Set TX rate limitation for queues in VF on a port::
1313 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1315 set port - mirror rule
1316 ~~~~~~~~~~~~~~~~~~~~~~
1318 Set pool or vlan type mirror rule for a port::
1320 testpmd> set port (port_id) mirror-rule (rule_id) \
1321 (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1322 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1324 Set link mirror rule for a port::
1326 testpmd> set port (port_id) mirror-rule (rule_id) \
1327 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1329 For example to enable mirror traffic with vlan 0,1 to pool 0::
1331 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1333 reset port - mirror rule
1334 ~~~~~~~~~~~~~~~~~~~~~~~~
1336 Reset a mirror rule for a port::
1338 testpmd> reset port (port_id) mirror-rule (rule_id)
1343 Set the flush on RX streams before forwarding.
1344 The default is flush ``on``.
1345 Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1347 testpmd> set flush_rx off
1352 Set the bypass mode for the lowest port on bypass enabled NIC::
1354 testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1359 Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1361 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1362 mode (normal|bypass|isolate) (port_id)
1366 * ``timeout``: Enable bypass after watchdog timeout.
1368 * ``os_on``: Enable bypass when OS/board is powered on.
1370 * ``os_off``: Enable bypass when OS/board is powered off.
1372 * ``power_on``: Enable bypass when power supply is turned on.
1374 * ``power_off``: Enable bypass when power supply is turned off.
1380 Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1382 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1387 Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1389 testpmd> show bypass config (port_id)
1394 Set link up for a port::
1396 testpmd> set link-up port (port id)
1401 Set link down for a port::
1403 testpmd> set link-down port (port id)
1408 Enable E-tag insertion for a VF on a port::
1410 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1412 Disable E-tag insertion for a VF on a port::
1414 testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1416 Enable/disable E-tag stripping on a port::
1418 testpmd> E-tag set stripping (on|off) port (port_id)
1420 Enable/disable E-tag based forwarding on a port::
1422 testpmd> E-tag set forwarding (on|off) port (port_id)
1424 Add an E-tag forwarding filter on a port::
1426 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1428 Delete an E-tag forwarding filter on a port::
1429 testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1434 Load a dynamic device personalization (DDP) profile and store backup profile::
1436 testpmd> ddp add (port_id) (profile_path[,backup_profile_path])
1441 Delete a dynamic device personalization profile and restore backup profile::
1443 testpmd> ddp del (port_id) (backup_profile_path)
1448 List all items from the ptype mapping table::
1450 testpmd> ptype mapping get (port_id) (valid_only)
1454 * ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0).
1456 Replace a specific or a group of software defined ptype with a new one::
1458 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type)
1462 * ``target``: A specific software ptype or a mask to represent a group of software ptypes.
1464 * ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1).
1466 * ``pkt_type``: The new software ptype to replace the old ones.
1468 Update hardware defined ptype to software defined packet type mapping table::
1470 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype)
1474 * ``hw_ptype``: hardware ptype as the index of the ptype mapping table.
1476 * ``sw_ptype``: software ptype as the value of the ptype mapping table.
1478 Reset ptype mapping table::
1480 testpmd> ptype mapping reset (port_id)
1482 config per port Rx offloading
1483 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1485 Enable or disable a per port Rx offloading on all Rx queues of a port::
1487 testpmd> port config (port_id) rx_offload (offloading) on|off
1489 * ``offloading``: can be any of these offloading capability:
1490 vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro,
1491 qinq_strip, outer_ipv4_cksum, macsec_strip,
1492 header_split, vlan_filter, vlan_extend, jumbo_frame,
1493 crc_strip, scatter, timestamp, security, keep_crc
1495 This command should be run when the port is stopped, or else it will fail.
1497 config per queue Rx offloading
1498 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1500 Enable or disable a per queue Rx offloading only on a specific Rx queue::
1502 testpmd> port (port_id) rxq (queue_id) rx_offload (offloading) on|off
1504 * ``offloading``: can be any of these offloading capability:
1505 vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro,
1506 qinq_strip, outer_ipv4_cksum, macsec_strip,
1507 header_split, vlan_filter, vlan_extend, jumbo_frame,
1508 crc_strip, scatter, timestamp, security, keep_crc
1510 This command should be run when the port is stopped, or else it will fail.
1512 config per port Tx offloading
1513 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1515 Enable or disable a per port Tx offloading on all Tx queues of a port::
1517 testpmd> port config (port_id) tx_offload (offloading) on|off
1519 * ``offloading``: can be any of these offloading capability:
1520 vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum,
1521 sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum,
1522 qinq_insert, vxlan_tnl_tso, gre_tnl_tso,
1523 ipip_tnl_tso, geneve_tnl_tso, macsec_insert,
1524 mt_lockfree, multi_segs, mbuf_fast_free, security
1526 This command should be run when the port is stopped, or else it will fail.
1528 config per queue Tx offloading
1529 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1531 Enable or disable a per queue Tx offloading only on a specific Tx queue::
1533 testpmd> port (port_id) txq (queue_id) tx_offload (offloading) on|off
1535 * ``offloading``: can be any of these offloading capability:
1536 vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum,
1537 sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum,
1538 qinq_insert, vxlan_tnl_tso, gre_tnl_tso,
1539 ipip_tnl_tso, geneve_tnl_tso, macsec_insert,
1540 mt_lockfree, multi_segs, mbuf_fast_free, security
1542 This command should be run when the port is stopped, or else it will fail.
1544 Config VXLAN Encap outer layers
1545 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1547 Configure the outer layer to encapsulate a packet inside a VXLAN tunnel::
1549 set vxlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \
1550 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) eth-src (eth-src) \
1553 set vxlan-with-vlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \
1554 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) vlan-tci (vlan-tci) \
1555 eth-src (eth-src) eth-dst (eth-dst)
1557 Those command will set an internal configuration inside testpmd, any following
1558 flow rule using the action vxlan_encap will use the last configuration set.
1559 To have a different encapsulation header, one of those commands must be called
1560 before the flow rule creation.
1562 Config NVGRE Encap outer layers
1563 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1565 Configure the outer layer to encapsulate a packet inside a NVGRE tunnel::
1567 set nvgre ip-version (ipv4|ipv6) tni (tni) ip-src (ip-src) ip-dst (ip-dst) \
1568 eth-src (eth-src) eth-dst (eth-dst)
1569 set nvgre-with-vlan ip-version (ipv4|ipv6) tni (tni) ip-src (ip-src) \
1570 ip-dst (ip-dst) vlan-tci (vlan-tci) eth-src (eth-src) eth-dst (eth-dst)
1572 Those command will set an internal configuration inside testpmd, any following
1573 flow rule using the action nvgre_encap will use the last configuration set.
1574 To have a different encapsulation header, one of those commands must be called
1575 before the flow rule creation.
1580 The following sections show functions for configuring ports.
1584 Port configuration changes only become active when forwarding is started/restarted.
1589 Attach a port specified by pci address or virtual device args::
1591 testpmd> port attach (identifier)
1593 To attach a new pci device, the device should be recognized by kernel first.
1594 Then it should be moved under DPDK management.
1595 Finally the port can be attached to testpmd.
1597 For example, to move a pci device using ixgbe under DPDK management:
1599 .. code-block:: console
1601 # Check the status of the available devices.
1602 ./usertools/dpdk-devbind.py --status
1604 Network devices using DPDK-compatible driver
1605 ============================================
1608 Network devices using kernel driver
1609 ===================================
1610 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1613 # Bind the device to igb_uio.
1614 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1617 # Recheck the status of the devices.
1618 ./usertools/dpdk-devbind.py --status
1619 Network devices using DPDK-compatible driver
1620 ============================================
1621 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1623 To attach a port created by virtual device, above steps are not needed.
1625 For example, to attach a port whose pci address is 0000:0a:00.0.
1627 .. code-block:: console
1629 testpmd> port attach 0000:0a:00.0
1630 Attaching a new port...
1631 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1632 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1633 EAL: PCI memory mapped at 0x7f83bfa00000
1634 EAL: PCI memory mapped at 0x7f83bfa80000
1635 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1636 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1637 Port 0 is attached. Now total ports is 1
1640 For example, to attach a port created by pcap PMD.
1642 .. code-block:: console
1644 testpmd> port attach net_pcap0
1645 Attaching a new port...
1646 PMD: Initializing pmd_pcap for net_pcap0
1647 PMD: Creating pcap-backed ethdev on numa socket 0
1648 Port 0 is attached. Now total ports is 1
1651 In this case, identifier is ``net_pcap0``.
1652 This identifier format is the same as ``--vdev`` format of DPDK applications.
1654 For example, to re-attach a bonded port which has been previously detached,
1655 the mode and slave parameters must be given.
1657 .. code-block:: console
1659 testpmd> port attach net_bond_0,mode=0,slave=1
1660 Attaching a new port...
1661 EAL: Initializing pmd_bond for net_bond_0
1662 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1663 Port 0 is attached. Now total ports is 1
1670 Detach a specific port::
1672 testpmd> port detach (port_id)
1674 Before detaching a port, the port should be stopped and closed.
1676 For example, to detach a pci device port 0.
1678 .. code-block:: console
1680 testpmd> port stop 0
1683 testpmd> port close 0
1687 testpmd> port detach 0
1689 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1690 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1691 EAL: PCI memory unmapped at 0x7f83bfa00000
1692 EAL: PCI memory unmapped at 0x7f83bfa80000
1696 For example, to detach a virtual device port 0.
1698 .. code-block:: console
1700 testpmd> port stop 0
1703 testpmd> port close 0
1707 testpmd> port detach 0
1709 PMD: Closing pcap ethdev on numa socket 0
1710 Port 'net_pcap0' is detached. Now total ports is 0
1713 To remove a pci device completely from the system, first detach the port from testpmd.
1714 Then the device should be moved under kernel management.
1715 Finally the device can be removed using kernel pci hotplug functionality.
1717 For example, to move a pci device under kernel management:
1719 .. code-block:: console
1721 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1723 ./usertools/dpdk-devbind.py --status
1725 Network devices using DPDK-compatible driver
1726 ============================================
1729 Network devices using kernel driver
1730 ===================================
1731 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1733 To remove a port created by a virtual device, above steps are not needed.
1738 Start all ports or a specific port::
1740 testpmd> port start (port_id|all)
1745 Stop all ports or a specific port::
1747 testpmd> port stop (port_id|all)
1752 Close all ports or a specific port::
1754 testpmd> port close (port_id|all)
1756 port config - queue ring size
1757 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1759 Configure a rx/tx queue ring size::
1761 testpmd> port (port_id) (rxq|txq) (queue_id) ring_size (value)
1763 Only take effect after command that (re-)start the port or command that setup specific queue.
1765 port start/stop queue
1766 ~~~~~~~~~~~~~~~~~~~~~
1768 Start/stop a rx/tx queue on a specific port::
1770 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1772 port config - queue deferred start
1773 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1775 Switch on/off deferred start of a specific port queue::
1777 testpmd> port (port_id) (rxq|txq) (queue_id) deferred_start (on|off)
1780 ~~~~~~~~~~~~~~~~~~~~~
1782 Setup a rx/tx queue on a specific port::
1784 testpmd> port (port_id) (rxq|txq) (queue_id) setup
1786 Only take effect when port is started.
1791 Set the speed and duplex mode for all ports or a specific port::
1793 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1794 duplex (half|full|auto)
1796 port config - queues/descriptors
1797 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1799 Set number of queues/descriptors for rxq, txq, rxd and txd::
1801 testpmd> port config all (rxq|txq|rxd|txd) (value)
1803 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1805 port config - max-pkt-len
1806 ~~~~~~~~~~~~~~~~~~~~~~~~~
1808 Set the maximum packet length::
1810 testpmd> port config all max-pkt-len (value)
1812 This is equivalent to the ``--max-pkt-len`` command-line option.
1814 port config - CRC Strip
1815 ~~~~~~~~~~~~~~~~~~~~~~~
1817 Set hardware CRC stripping on or off for all ports::
1819 testpmd> port config all crc-strip (on|off)
1821 CRC stripping is on by default.
1823 The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option.
1825 port config - scatter
1826 ~~~~~~~~~~~~~~~~~~~~~~~
1828 Set RX scatter mode on or off for all ports::
1830 testpmd> port config all scatter (on|off)
1832 RX scatter mode is off by default.
1834 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1836 port config - RX Checksum
1837 ~~~~~~~~~~~~~~~~~~~~~~~~~
1839 Set hardware RX checksum offload to on or off for all ports::
1841 testpmd> port config all rx-cksum (on|off)
1843 Checksum offload is off by default.
1845 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1850 Set hardware VLAN on or off for all ports::
1852 testpmd> port config all hw-vlan (on|off)
1854 Hardware VLAN is off by default.
1856 The ``on`` option is equivalent to the ``--enable-hw-vlan`` command-line option.
1858 port config - VLAN filter
1859 ~~~~~~~~~~~~~~~~~~~~~~~~~
1861 Set hardware VLAN filter on or off for all ports::
1863 testpmd> port config all hw-vlan-filter (on|off)
1865 Hardware VLAN filter is off by default.
1867 The ``on`` option is equivalent to the ``--enable-hw-vlan-filter`` command-line option.
1869 port config - VLAN strip
1870 ~~~~~~~~~~~~~~~~~~~~~~~~
1872 Set hardware VLAN strip on or off for all ports::
1874 testpmd> port config all hw-vlan-strip (on|off)
1876 Hardware VLAN strip is off by default.
1878 The ``on`` option is equivalent to the ``--enable-hw-vlan-strip`` command-line option.
1880 port config - VLAN extend
1881 ~~~~~~~~~~~~~~~~~~~~~~~~~
1883 Set hardware VLAN extend on or off for all ports::
1885 testpmd> port config all hw-vlan-extend (on|off)
1887 Hardware VLAN extend is off by default.
1889 The ``on`` option is equivalent to the ``--enable-hw-vlan-extend`` command-line option.
1891 port config - Drop Packets
1892 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1894 Set packet drop for packets with no descriptors on or off for all ports::
1896 testpmd> port config all drop-en (on|off)
1898 Packet dropping for packets with no descriptors is off by default.
1900 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1905 Set the RSS (Receive Side Scaling) mode on or off::
1907 testpmd> port config all rss (all|default|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1909 RSS is on by default.
1911 The ``all`` option is equivalent to ip|tcp|udp|sctp|ether.
1912 The ``default`` option enables all supported RSS types reported by device info.
1913 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1915 port config - RSS Reta
1916 ~~~~~~~~~~~~~~~~~~~~~~
1918 Set the RSS (Receive Side Scaling) redirection table::
1920 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1925 Set the DCB mode for an individual port::
1927 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1929 The traffic class should be 4 or 8.
1934 Set the number of packets per burst::
1936 testpmd> port config all burst (value)
1938 This is equivalent to the ``--burst`` command-line option.
1940 port config - Threshold
1941 ~~~~~~~~~~~~~~~~~~~~~~~
1943 Set thresholds for TX/RX queues::
1945 testpmd> port config all (threshold) (value)
1947 Where the threshold type can be:
1949 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1951 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1953 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1955 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1957 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1959 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1961 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1963 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1965 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1967 These threshold options are also available from the command-line.
1972 Set the value of ether-type for E-tag::
1974 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1976 Enable/disable the E-tag support::
1978 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1980 port config pctype mapping
1981 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1983 Reset pctype mapping table::
1985 testpmd> port config (port_id) pctype mapping reset
1987 Update hardware defined pctype to software defined flow type mapping table::
1989 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id)
1993 * ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table.
1995 * ``flow_type_id``: software flow type id as the index of the pctype mapping table.
1997 port config input set
1998 ~~~~~~~~~~~~~~~~~~~~~
2000 Config RSS/FDIR/FDIR flexible payload input set for some pctype::
2001 testpmd> port config (port_id) pctype (pctype_id) \
2002 (hash_inset|fdir_inset|fdir_flx_inset) \
2003 (get|set|clear) field (field_idx)
2005 Clear RSS/FDIR/FDIR flexible payload input set for some pctype::
2006 testpmd> port config (port_id) pctype (pctype_id) \
2007 (hash_inset|fdir_inset|fdir_flx_inset) clear all
2011 * ``pctype_id``: hardware packet classification types.
2012 * ``field_idx``: hardware field index.
2014 port config udp_tunnel_port
2015 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
2017 Add/remove UDP tunnel port for VXLAN/GENEVE tunneling protocols::
2018 testpmd> port config (port_id) udp_tunnel_port add|rm vxlan|geneve (udp_port)
2020 Link Bonding Functions
2021 ----------------------
2023 The Link Bonding functions make it possible to dynamically create and
2024 manage link bonding devices from within testpmd interactive prompt.
2026 create bonded device
2027 ~~~~~~~~~~~~~~~~~~~~
2029 Create a new bonding device::
2031 testpmd> create bonded device (mode) (socket)
2033 For example, to create a bonded device in mode 1 on socket 0::
2035 testpmd> create bonded device 1 0
2036 created new bonded device (port X)
2041 Adds Ethernet device to a Link Bonding device::
2043 testpmd> add bonding slave (slave id) (port id)
2045 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
2047 testpmd> add bonding slave 6 10
2050 remove bonding slave
2051 ~~~~~~~~~~~~~~~~~~~~
2053 Removes an Ethernet slave device from a Link Bonding device::
2055 testpmd> remove bonding slave (slave id) (port id)
2057 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
2059 testpmd> remove bonding slave 6 10
2064 Set the Link Bonding mode of a Link Bonding device::
2066 testpmd> set bonding mode (value) (port id)
2068 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
2070 testpmd> set bonding mode 3 10
2075 Set an Ethernet slave device as the primary device on a Link Bonding device::
2077 testpmd> set bonding primary (slave id) (port id)
2079 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
2081 testpmd> set bonding primary 6 10
2086 Set the MAC address of a Link Bonding device::
2088 testpmd> set bonding mac (port id) (mac)
2090 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
2092 testpmd> set bonding mac 10 00:00:00:00:00:01
2094 set bonding xmit_balance_policy
2095 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2097 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
2099 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
2101 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
2103 testpmd> set bonding xmit_balance_policy 10 l34
2106 set bonding mon_period
2107 ~~~~~~~~~~~~~~~~~~~~~~
2109 Set the link status monitoring polling period in milliseconds for a bonding device.
2111 This adds support for PMD slave devices which do not support link status interrupts.
2112 When the mon_period is set to a value greater than 0 then all PMD's which do not support
2113 link status ISR will be queried every polling interval to check if their link status has changed::
2115 testpmd> set bonding mon_period (port_id) (value)
2117 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
2119 testpmd> set bonding mon_period 5 150
2122 set bonding lacp dedicated_queue
2123 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2125 Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic
2126 when in mode 4 (link-aggregration-802.3ad)::
2128 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable)
2131 set bonding agg_mode
2132 ~~~~~~~~~~~~~~~~~~~~
2134 Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad)::
2136 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable)
2142 Show the current configuration of a Link Bonding device::
2144 testpmd> show bonding config (port id)
2147 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
2148 in balance mode with a transmission policy of layer 2+3::
2150 testpmd> show bonding config 9
2152 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
2154 Active Slaves (3): [1 3 4]
2161 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
2162 This is mainly useful for debugging purposes.
2163 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
2164 and fields that can be accessed.
2169 Display the value of a port register::
2171 testpmd> read reg (port_id) (address)
2173 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
2175 testpmd> read reg 0 0xEE00
2176 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
2181 Display a port register bit field::
2183 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
2185 For example, reading the lowest two bits from the register in the example above::
2187 testpmd> read regfield 0 0xEE00 0 1
2188 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
2193 Display a single port register bit::
2195 testpmd> read regbit (port_id) (address) (bit_x)
2197 For example, reading the lowest bit from the register in the example above::
2199 testpmd> read regbit 0 0xEE00 0
2200 port 0 PCI register at offset 0xEE00: bit 0=1
2205 Set the value of a port register::
2207 testpmd> write reg (port_id) (address) (value)
2209 For example, to clear a register::
2211 testpmd> write reg 0 0xEE00 0x0
2212 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
2217 Set bit field of a port register::
2219 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
2221 For example, writing to the register cleared in the example above::
2223 testpmd> write regfield 0 0xEE00 0 1 2
2224 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
2229 Set single bit value of a port register::
2231 testpmd> write regbit (port_id) (address) (bit_x) (value)
2233 For example, to set the high bit in the register from the example above::
2235 testpmd> write regbit 0 0xEE00 31 1
2236 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
2238 Traffic Metering and Policing
2239 -----------------------------
2241 The following section shows functions for configuring traffic metering and
2242 policing on the ethernet device through the use of generic ethdev API.
2244 show port traffic management capability
2245 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2247 Show traffic metering and policing capability of the port::
2249 testpmd> show port meter cap (port_id)
2251 add port meter profile (srTCM rfc2967)
2252 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2254 Add meter profile (srTCM rfc2697) to the ethernet device::
2256 testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \
2261 * ``profile_id``: ID for the meter profile.
2262 * ``cir``: Committed Information Rate (CIR) (bytes/second).
2263 * ``cbs``: Committed Burst Size (CBS) (bytes).
2264 * ``ebs``: Excess Burst Size (EBS) (bytes).
2266 add port meter profile (trTCM rfc2968)
2267 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2269 Add meter profile (srTCM rfc2698) to the ethernet device::
2271 testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \
2272 (cir) (pir) (cbs) (pbs)
2276 * ``profile_id``: ID for the meter profile.
2277 * ``cir``: Committed information rate (bytes/second).
2278 * ``pir``: Peak information rate (bytes/second).
2279 * ``cbs``: Committed burst size (bytes).
2280 * ``pbs``: Peak burst size (bytes).
2282 add port meter profile (trTCM rfc4115)
2283 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2285 Add meter profile (trTCM rfc4115) to the ethernet device::
2287 testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \
2288 (cir) (eir) (cbs) (ebs)
2292 * ``profile_id``: ID for the meter profile.
2293 * ``cir``: Committed information rate (bytes/second).
2294 * ``eir``: Excess information rate (bytes/second).
2295 * ``cbs``: Committed burst size (bytes).
2296 * ``ebs``: Excess burst size (bytes).
2298 delete port meter profile
2299 ~~~~~~~~~~~~~~~~~~~~~~~~~
2301 Delete meter profile from the ethernet device::
2303 testpmd> del port meter profile (port_id) (profile_id)
2308 Create new meter object for the ethernet device::
2310 testpmd> create port meter (port_id) (mtr_id) (profile_id) \
2311 (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \
2312 (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\
2317 * ``mtr_id``: meter object ID.
2318 * ``profile_id``: ID for the meter profile.
2319 * ``meter_enable``: When this parameter has a non-zero value, the meter object
2320 gets enabled at the time of creation, otherwise remains disabled.
2321 * ``g_action``: Policer action for the packet with green color.
2322 * ``y_action``: Policer action for the packet with yellow color.
2323 * ``r_action``: Policer action for the packet with red color.
2324 * ``stats_mask``: Mask of statistics counter types to be enabled for the
2326 * ``shared``: When this parameter has a non-zero value, the meter object is
2327 shared by multiple flows. Otherwise, meter object is used by single flow.
2328 * ``use_pre_meter_color``: When this parameter has a non-zero value, the
2329 input color for the current meter object is determined by the latest meter
2330 object in the same flow. Otherwise, the current meter object uses the
2331 *dscp_table* to determine the input color.
2332 * ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input
2333 color, 0 <= x <= 63.
2338 Enable meter for the ethernet device::
2340 testpmd> enable port meter (port_id) (mtr_id)
2345 Disable meter for the ethernet device::
2347 testpmd> disable port meter (port_id) (mtr_id)
2352 Delete meter for the ethernet device::
2354 testpmd> del port meter (port_id) (mtr_id)
2356 Set port meter profile
2357 ~~~~~~~~~~~~~~~~~~~~~~
2359 Set meter profile for the ethernet device::
2361 testpmd> set port meter profile (port_id) (mtr_id) (profile_id)
2363 set port meter dscp table
2364 ~~~~~~~~~~~~~~~~~~~~~~~~~
2366 Set meter dscp table for the ethernet device::
2368 testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \
2369 (dscp_tbl_entry1)...(dscp_tbl_entry63)]
2371 set port meter policer action
2372 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2374 Set meter policer action for the ethernet device::
2376 testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \
2377 (action0) [(action1) (action1)]
2381 * ``action_mask``: Bit mask indicating which policer actions need to be
2382 updated. One or more policer actions can be updated in a single function
2383 invocation. To update the policer action associated with color C, bit
2384 (1 << C) needs to be set in *action_mask* and element at position C
2385 in the *actions* array needs to be valid.
2386 * ``actionx``: Policer action for the color x,
2387 RTE_MTR_GREEN <= x < RTE_MTR_COLORS
2389 set port meter stats mask
2390 ~~~~~~~~~~~~~~~~~~~~~~~~~
2392 Set meter stats mask for the ethernet device::
2394 testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask)
2398 * ``stats_mask``: Bit mask indicating statistics counter types to be enabled.
2400 show port meter stats
2401 ~~~~~~~~~~~~~~~~~~~~~
2403 Show meter stats of the ethernet device::
2405 testpmd> show port meter stats (port_id) (mtr_id) (clear)
2409 * ``clear``: Flag that indicates whether the statistics counters should
2410 be cleared (i.e. set to zero) immediately after they have been read or not.
2415 The following section shows functions for configuring traffic management on
2416 on the ethernet device through the use of generic TM API.
2418 show port traffic management capability
2419 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2421 Show traffic management capability of the port::
2423 testpmd> show port tm cap (port_id)
2425 show port traffic management capability (hierarchy level)
2426 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2428 Show traffic management hierarchy level capability of the port::
2430 testpmd> show port tm level cap (port_id) (level_id)
2432 show port traffic management capability (hierarchy node level)
2433 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2435 Show the traffic management hierarchy node capability of the port::
2437 testpmd> show port tm node cap (port_id) (node_id)
2439 show port traffic management hierarchy node type
2440 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2442 Show the port traffic management hierarchy node type::
2444 testpmd> show port tm node type (port_id) (node_id)
2446 show port traffic management hierarchy node stats
2447 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2449 Show the port traffic management hierarchy node statistics::
2451 testpmd> show port tm node stats (port_id) (node_id) (clear)
2455 * ``clear``: When this parameter has a non-zero value, the statistics counters
2456 are cleared (i.e. set to zero) immediately after they have been read,
2457 otherwise the statistics counters are left untouched.
2459 Add port traffic management private shaper profile
2460 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2462 Add the port traffic management private shaper profile::
2464 testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \
2465 (tb_rate) (tb_size) (packet_length_adjust)
2469 * ``shaper_profile id``: Shaper profile ID for the new profile.
2470 * ``tb_rate``: Token bucket rate (bytes per second).
2471 * ``tb_size``: Token bucket size (bytes).
2472 * ``packet_length_adjust``: The value (bytes) to be added to the length of
2473 each packet for the purpose of shaping. This parameter value can be used to
2474 correct the packet length with the framing overhead bytes that are consumed
2477 Delete port traffic management private shaper profile
2478 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2480 Delete the port traffic management private shaper::
2482 testpmd> del port tm node shaper profile (port_id) (shaper_profile_id)
2486 * ``shaper_profile id``: Shaper profile ID that needs to be deleted.
2488 Add port traffic management shared shaper
2489 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2491 Create the port traffic management shared shaper::
2493 testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \
2498 * ``shared_shaper_id``: Shared shaper ID to be created.
2499 * ``shaper_profile id``: Shaper profile ID for shared shaper.
2501 Set port traffic management shared shaper
2502 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2504 Update the port traffic management shared shaper::
2506 testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \
2511 * ``shared_shaper_id``: Shared shaper ID to be update.
2512 * ``shaper_profile id``: Shaper profile ID for shared shaper.
2514 Delete port traffic management shared shaper
2515 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2517 Delete the port traffic management shared shaper::
2519 testpmd> del port tm node shared shaper (port_id) (shared_shaper_id)
2523 * ``shared_shaper_id``: Shared shaper ID to be deleted.
2525 Set port traffic management hiearchy node private shaper
2526 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2528 set the port traffic management hierarchy node private shaper::
2530 testpmd> set port tm node shaper profile (port_id) (node_id) \
2535 * ``shaper_profile id``: Private shaper profile ID to be enabled on the
2538 Add port traffic management WRED profile
2539 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2541 Create a new WRED profile::
2543 testpmd> add port tm node wred profile (port_id) (wred_profile_id) \
2544 (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \
2545 (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \
2546 (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r)
2550 * ``wred_profile id``: Identifier for the newly create WRED profile
2551 * ``color_g``: Packet color (green)
2552 * ``min_th_g``: Minimum queue threshold for packet with green color
2553 * ``max_th_g``: Minimum queue threshold for packet with green color
2554 * ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp)
2555 * ``wq_log2_g``: Negated log2 of queue weight (wq)
2556 * ``color_y``: Packet color (yellow)
2557 * ``min_th_y``: Minimum queue threshold for packet with yellow color
2558 * ``max_th_y``: Minimum queue threshold for packet with yellow color
2559 * ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp)
2560 * ``wq_log2_y``: Negated log2 of queue weight (wq)
2561 * ``color_r``: Packet color (red)
2562 * ``min_th_r``: Minimum queue threshold for packet with yellow color
2563 * ``max_th_r``: Minimum queue threshold for packet with yellow color
2564 * ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp)
2565 * ``wq_log2_r``: Negated log2 of queue weight (wq)
2567 Delete port traffic management WRED profile
2568 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2570 Delete the WRED profile::
2572 testpmd> del port tm node wred profile (port_id) (wred_profile_id)
2574 Add port traffic management hierarchy nonleaf node
2575 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2577 Add nonleaf node to port traffic management hiearchy::
2579 testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \
2580 (priority) (weight) (level_id) (shaper_profile_id) \
2581 (n_sp_priorities) (stats_mask) (n_shared_shapers) \
2582 [(shared_shaper_0) (shared_shaper_1) ...] \
2586 * ``parent_node_id``: Node ID of the parent.
2587 * ``priority``: Node priority (highest node priority is zero). This is used by
2588 the SP algorithm running on the parent node for scheduling this node.
2589 * ``weight``: Node weight (lowest weight is one). The node weight is relative
2590 to the weight sum of all siblings that have the same priority. It is used by
2591 the WFQ algorithm running on the parent node for scheduling this node.
2592 * ``level_id``: Hiearchy level of the node.
2593 * ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2595 * ``n_sp_priorities``: Number of strict priorities.
2596 * ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2597 * ``n_shared_shapers``: Number of shared shapers.
2598 * ``shared_shaper_id``: Shared shaper id.
2600 Add port traffic management hierarchy leaf node
2601 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2603 Add leaf node to port traffic management hiearchy::
2605 testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \
2606 (priority) (weight) (level_id) (shaper_profile_id) \
2607 (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \
2608 [(shared_shaper_id) (shared_shaper_id) ...] \
2612 * ``parent_node_id``: Node ID of the parent.
2613 * ``priority``: Node priority (highest node priority is zero). This is used by
2614 the SP algorithm running on the parent node for scheduling this node.
2615 * ``weight``: Node weight (lowest weight is one). The node weight is relative
2616 to the weight sum of all siblings that have the same priority. It is used by
2617 the WFQ algorithm running on the parent node for scheduling this node.
2618 * ``level_id``: Hiearchy level of the node.
2619 * ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2621 * ``cman_mode``: Congestion management mode to be enabled for this node.
2622 * ``wred_profile_id``: WRED profile id to be enabled for this node.
2623 * ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2624 * ``n_shared_shapers``: Number of shared shapers.
2625 * ``shared_shaper_id``: Shared shaper id.
2627 Delete port traffic management hierarchy node
2628 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2630 Delete node from port traffic management hiearchy::
2632 testpmd> del port tm node (port_id) (node_id)
2634 Update port traffic management hierarchy parent node
2635 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2637 Update port traffic management hierarchy parent node::
2639 testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \
2642 This function can only be called after the hierarchy commit invocation. Its
2643 success depends on the port support for this operation, as advertised through
2644 the port capability set. This function is valid for all nodes of the traffic
2645 management hierarchy except root node.
2647 Suspend port traffic management hierarchy node
2648 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2650 testpmd> suspend port tm node (port_id) (node_id)
2652 Resume port traffic management hierarchy node
2653 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2655 testpmd> resume port tm node (port_id) (node_id)
2657 Commit port traffic management hierarchy
2658 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2660 Commit the traffic management hierarchy on the port::
2662 testpmd> port tm hierarchy commit (port_id) (clean_on_fail)
2666 * ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function
2667 call failure. On the other hand, hierarchy is preserved when this parameter
2670 Set port traffic management default hierarchy (softnic forwarding mode)
2671 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2673 set the traffic management default hierarchy on the port::
2675 testpmd> set port tm hierarchy default (port_id)
2680 This section details the available filter functions that are available.
2682 Note these functions interface the deprecated legacy filtering framework,
2683 superseded by *rte_flow*. See `Flow rules management`_.
2686 ~~~~~~~~~~~~~~~~~~~~
2688 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
2690 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
2691 ethertype (ether_type) (drop|fwd) queue (queue_id)
2693 The available information parameters are:
2695 * ``port_id``: The port which the Ethertype filter assigned on.
2697 * ``mac_addr``: Compare destination mac address.
2699 * ``mac_ignr``: Ignore destination mac address match.
2701 * ``mac_address``: Destination mac address to match.
2703 * ``ether_type``: The EtherType value want to match,
2704 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
2706 * ``queue_id``: The receive queue associated with this EtherType filter.
2707 It is meaningless when deleting or dropping.
2709 Example, to add/remove an ethertype filter rule::
2711 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
2712 ethertype 0x0806 fwd queue 3
2714 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
2715 ethertype 0x0806 fwd queue 3
2720 Add or delete a 2-tuple filter,
2721 which identifies packets by specific protocol and destination TCP/UDP port
2722 and forwards packets into one of the receive queues::
2724 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
2725 protocol (protocol_value) mask (mask_value) \
2726 tcp_flags (tcp_flags_value) priority (prio_value) \
2729 The available information parameters are:
2731 * ``port_id``: The port which the 2-tuple filter assigned on.
2733 * ``dst_port_value``: Destination port in L4.
2735 * ``protocol_value``: IP L4 protocol.
2737 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
2739 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
2741 * ``prio_value``: Priority of this filter.
2743 * ``queue_id``: The receive queue associated with this 2-tuple filter.
2745 Example, to add/remove an 2tuple filter rule::
2747 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
2748 tcp_flags 0x02 priority 3 queue 3
2750 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
2751 tcp_flags 0x02 priority 3 queue 3
2756 Add or delete a 5-tuple filter,
2757 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
2758 and routes packets into one of the receive queues::
2760 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
2761 (src_address) dst_port (dst_port_value) \
2762 src_port (src_port_value) protocol (protocol_value) \
2763 mask (mask_value) tcp_flags (tcp_flags_value) \
2764 priority (prio_value) queue (queue_id)
2766 The available information parameters are:
2768 * ``port_id``: The port which the 5-tuple filter assigned on.
2770 * ``dst_address``: Destination IP address.
2772 * ``src_address``: Source IP address.
2774 * ``dst_port_value``: TCP/UDP destination port.
2776 * ``src_port_value``: TCP/UDP source port.
2778 * ``protocol_value``: L4 protocol.
2780 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
2782 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
2784 * ``prio_value``: The priority of this filter.
2786 * ``queue_id``: The receive queue associated with this 5-tuple filter.
2788 Example, to add/remove an 5tuple filter rule::
2790 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2791 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2792 flags 0x0 priority 3 queue 3
2794 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2795 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2796 flags 0x0 priority 3 queue 3
2801 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
2803 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
2805 The available information parameters are:
2807 * ``port_id``: The port which the SYN filter assigned on.
2809 * ``high``: This SYN filter has higher priority than other filters.
2811 * ``low``: This SYN filter has lower priority than other filters.
2813 * ``queue_id``: The receive queue associated with this SYN filter
2817 testpmd> syn_filter 0 add priority high queue 3
2822 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
2823 and routed into one of the receive queues::
2825 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
2826 mask (mask_value) priority (prio_value) queue (queue_id)
2828 The available information parameters are:
2830 * ``port_id``: The port which the Flex filter is assigned on.
2832 * ``len_value``: Filter length in bytes, no greater than 128.
2834 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
2836 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
2838 * ``prio_value``: The priority of this filter.
2840 * ``queue_id``: The receive queue associated with this Flex filter.
2844 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
2845 mask 000C priority 3 queue 3
2847 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
2848 mask 000C priority 3 queue 3
2851 .. _testpmd_flow_director:
2853 flow_director_filter
2854 ~~~~~~~~~~~~~~~~~~~~
2856 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
2858 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
2859 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
2861 * Perfect 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 IP flow.
2865 * Signature filters.
2866 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
2868 * Perfect-mac-vlan match filters.
2869 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2870 The masked fields are for MAC VLAN flow.
2872 * Perfect-tunnel match filters.
2873 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2874 The masked fields are for tunnel flow.
2876 * Perfect-raw-flow-type match filters.
2877 The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet.
2878 The masked fields are specified by input sets.
2880 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
2881 per flow type and the flexible payload.
2883 The Flow Director can also mask out parts of all of these fields so that filters
2884 are only applied to certain fields or parts of the fields.
2886 Note that for raw flow type mode the source and destination fields in the
2887 raw packet buffer need to be presented in a reversed order with respect
2888 to the expected received packets.
2889 For example: IP source and destination addresses or TCP/UDP/SCTP
2890 source and destination ports
2892 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
2894 # Commands to add flow director filters of different flow types::
2896 flow_director_filter (port_id) mode IP (add|del|update) \
2897 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
2898 src (src_ip_address) dst (dst_ip_address) \
2899 tos (tos_value) proto (proto_value) ttl (ttl_value) \
2900 vlan (vlan_value) flexbytes (flexbytes_value) \
2901 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
2904 flow_director_filter (port_id) mode IP (add|del|update) \
2905 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
2906 src (src_ip_address) (src_port) \
2907 dst (dst_ip_address) (dst_port) \
2908 tos (tos_value) ttl (ttl_value) \
2909 vlan (vlan_value) flexbytes (flexbytes_value) \
2910 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
2913 flow_director_filter (port_id) mode IP (add|del|update) \
2914 flow (ipv4-sctp|ipv6-sctp) \
2915 src (src_ip_address) (src_port) \
2916 dst (dst_ip_address) (dst_port) \
2917 tos (tos_value) ttl (ttl_value) \
2918 tag (verification_tag) vlan (vlan_value) \
2919 flexbytes (flexbytes_value) (drop|fwd) \
2920 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
2922 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
2923 ether (ethertype) flexbytes (flexbytes_value) \
2924 (drop|fwd) pf|vf(vf_id) queue (queue_id)
2927 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
2928 mac (mac_address) vlan (vlan_value) \
2929 flexbytes (flexbytes_value) (drop|fwd) \
2930 queue (queue_id) fd_id (fd_id_value)
2932 flow_director_filter (port_id) mode Tunnel (add|del|update) \
2933 mac (mac_address) vlan (vlan_value) \
2934 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
2935 flexbytes (flexbytes_value) (drop|fwd) \
2936 queue (queue_id) fd_id (fd_id_value)
2938 flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \
2939 (drop|fwd) queue (queue_id) fd_id (fd_id_value) \
2940 packet (packet file name)
2942 For example, to add an ipv4-udp flow type filter::
2944 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
2945 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
2946 fwd pf queue 1 fd_id 1
2948 For example, add an ipv4-other flow type filter::
2950 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
2951 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
2952 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
2957 Flush all flow director filters on a device::
2959 testpmd> flush_flow_director (port_id)
2961 Example, to flush all flow director filter on port 0::
2963 testpmd> flush_flow_director 0
2968 Set flow director's input masks::
2970 flow_director_mask (port_id) mode IP vlan (vlan_value) \
2971 src_mask (ipv4_src) (ipv6_src) (src_port) \
2972 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2974 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2976 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2977 mac (mac_value) tunnel-type (tunnel_type_value) \
2978 tunnel-id (tunnel_id_value)
2980 Example, to set flow director mask on port 0::
2982 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2983 src_mask 255.255.255.255 \
2984 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2985 dst_mask 255.255.255.255 \
2986 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2988 flow_director_flex_mask
2989 ~~~~~~~~~~~~~~~~~~~~~~~
2991 set masks of flow director's flexible payload based on certain flow type::
2993 testpmd> flow_director_flex_mask (port_id) \
2994 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2995 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2996 l2_payload|all) (mask)
2998 Example, to set flow director's flex mask for all flow type on port 0::
3000 testpmd> flow_director_flex_mask 0 flow all \
3001 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
3004 flow_director_flex_payload
3005 ~~~~~~~~~~~~~~~~~~~~~~~~~~
3007 Configure flexible payload selection::
3009 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
3011 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
3013 testpmd> flow_director_flex_payload 0 l4 \
3014 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
3016 get_sym_hash_ena_per_port
3017 ~~~~~~~~~~~~~~~~~~~~~~~~~
3019 Get symmetric hash enable configuration per port::
3021 get_sym_hash_ena_per_port (port_id)
3023 For example, to get symmetric hash enable configuration of port 1::
3025 testpmd> get_sym_hash_ena_per_port 1
3027 set_sym_hash_ena_per_port
3028 ~~~~~~~~~~~~~~~~~~~~~~~~~
3030 Set symmetric hash enable configuration per port to enable or disable::
3032 set_sym_hash_ena_per_port (port_id) (enable|disable)
3034 For example, to set symmetric hash enable configuration of port 1 to enable::
3036 testpmd> set_sym_hash_ena_per_port 1 enable
3038 get_hash_global_config
3039 ~~~~~~~~~~~~~~~~~~~~~~
3041 Get the global configurations of hash filters::
3043 get_hash_global_config (port_id)
3045 For example, to get the global configurations of hash filters of port 1::
3047 testpmd> get_hash_global_config 1
3049 set_hash_global_config
3050 ~~~~~~~~~~~~~~~~~~~~~~
3052 Set the global configurations of hash filters::
3054 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
3055 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
3056 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \
3059 For example, to enable simple_xor for flow type of ipv6 on port 2::
3061 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
3066 Set the input set for hash::
3068 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
3069 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
3070 l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
3071 ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
3072 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
3073 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
3074 fld-8th|none) (select|add)
3076 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
3078 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
3083 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
3084 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
3086 Set the input set for flow director::
3088 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
3089 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
3090 l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
3091 ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
3092 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
3093 sctp-dst-port|sctp-veri-tag|none) (select|add)
3095 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
3097 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
3102 Set different GRE key length for input set::
3104 global_config (port_id) gre-key-len (number in bytes)
3106 For example to set GRE key length for input set to 4 bytes on port 0::
3108 testpmd> global_config 0 gre-key-len 4
3111 .. _testpmd_rte_flow:
3113 Flow rules management
3114 ---------------------
3116 Control of the generic flow API (*rte_flow*) is fully exposed through the
3117 ``flow`` command (validation, creation, destruction, queries and operation
3120 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
3121 features simultaneously may cause undefined side-effects and is therefore
3127 Because the ``flow`` command uses dynamic tokens to handle the large number
3128 of possible flow rules combinations, its behavior differs slightly from
3129 other commands, in particular:
3131 - Pressing *?* or the *<tab>* key displays contextual help for the current
3132 token, not that of the entire command.
3134 - Optional and repeated parameters are supported (provided they are listed
3135 in the contextual help).
3137 The first parameter stands for the operation mode. Possible operations and
3138 their general syntax are described below. They are covered in detail in the
3141 - Check whether a flow rule can be created::
3143 flow validate {port_id}
3144 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3145 pattern {item} [/ {item} [...]] / end
3146 actions {action} [/ {action} [...]] / end
3148 - Create a flow rule::
3150 flow create {port_id}
3151 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3152 pattern {item} [/ {item} [...]] / end
3153 actions {action} [/ {action} [...]] / end
3155 - Destroy specific flow rules::
3157 flow destroy {port_id} rule {rule_id} [...]
3159 - Destroy all flow rules::
3161 flow flush {port_id}
3163 - Query an existing flow rule::
3165 flow query {port_id} {rule_id} {action}
3167 - List existing flow rules sorted by priority, filtered by group
3170 flow list {port_id} [group {group_id}] [...]
3172 - Restrict ingress traffic to the defined flow rules::
3174 flow isolate {port_id} {boolean}
3176 Validating flow rules
3177 ~~~~~~~~~~~~~~~~~~~~~
3179 ``flow validate`` reports whether a flow rule would be accepted by the
3180 underlying device in its current state but stops short of creating it. It is
3181 bound to ``rte_flow_validate()``::
3183 flow validate {port_id}
3184 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3185 pattern {item} [/ {item} [...]] / end
3186 actions {action} [/ {action} [...]] / end
3188 If successful, it will show::
3192 Otherwise it will show an error message of the form::
3194 Caught error type [...] ([...]): [...]
3196 This command uses the same parameters as ``flow create``, their format is
3197 described in `Creating flow rules`_.
3199 Check whether redirecting any Ethernet packet received on port 0 to RX queue
3200 index 6 is supported::
3202 testpmd> flow validate 0 ingress pattern eth / end
3203 actions queue index 6 / end
3207 Port 0 does not support TCPv6 rules::
3209 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
3211 Caught error type 9 (specific pattern item): Invalid argument
3217 ``flow create`` validates and creates the specified flow rule. It is bound
3218 to ``rte_flow_create()``::
3220 flow create {port_id}
3221 [group {group_id}] [priority {level}] [ingress] [egress] [transfer]
3222 pattern {item} [/ {item} [...]] / end
3223 actions {action} [/ {action} [...]] / end
3225 If successful, it will return a flow rule ID usable with other commands::
3227 Flow rule #[...] created
3229 Otherwise it will show an error message of the form::
3231 Caught error type [...] ([...]): [...]
3233 Parameters describe in the following order:
3235 - Attributes (*group*, *priority*, *ingress*, *egress*, *transfer* tokens).
3236 - A matching pattern, starting with the *pattern* token and terminated by an
3238 - Actions, starting with the *actions* token and terminated by an *end*
3241 These translate directly to *rte_flow* objects provided as-is to the
3242 underlying functions.
3244 The shortest valid definition only comprises mandatory tokens::
3246 testpmd> flow create 0 pattern end actions end
3248 Note that PMDs may refuse rules that essentially do nothing such as this
3251 **All unspecified object values are automatically initialized to 0.**
3256 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
3257 specified before the ``pattern`` token.
3259 - ``group {group id}``: priority group.
3260 - ``priority {level}``: priority level within group.
3261 - ``ingress``: rule applies to ingress traffic.
3262 - ``egress``: rule applies to egress traffic.
3263 - ``transfer``: apply rule directly to endpoints found in pattern.
3265 Each instance of an attribute specified several times overrides the previous
3266 value as shown below (group 4 is used)::
3268 testpmd> flow create 0 group 42 group 24 group 4 [...]
3270 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
3272 While not specifying a direction is an error, some rules may allow both
3275 Most rules affect RX therefore contain the ``ingress`` token::
3277 testpmd> flow create 0 ingress pattern [...]
3282 A matching pattern starts after the ``pattern`` token. It is made of pattern
3283 items and is terminated by a mandatory ``end`` item.
3285 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
3286 rte_flow_item_type``).
3288 The ``/`` token is used as a separator between pattern items as shown
3291 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
3293 Note that protocol items like these must be stacked from lowest to highest
3294 layer to make sense. For instance, the following rule is either invalid or
3295 unlikely to match any packet::
3297 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
3299 More information on these restrictions can be found in the *rte_flow*
3302 Several items support additional specification structures, for example
3303 ``ipv4`` allows specifying source and destination addresses as follows::
3305 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
3306 dst is 10.2.0.0 / end [...]
3308 This rule matches all IPv4 traffic with the specified properties.
3310 In this example, ``src`` and ``dst`` are field names of the underlying
3311 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
3312 in a similar fashion.
3314 The ``is`` token means that the subsequent value must be matched exactly,
3315 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
3316 accordingly. Possible assignment tokens are:
3318 - ``is``: match value perfectly (with full bit-mask).
3319 - ``spec``: match value according to configured bit-mask.
3320 - ``last``: specify upper bound to establish a range.
3321 - ``mask``: specify bit-mask with relevant bits set to one.
3322 - ``prefix``: generate bit-mask from a prefix length.
3324 These yield identical results::
3326 ipv4 src is 10.1.1.1
3330 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
3334 ipv4 src spec 10.1.1.1 src prefix 32
3338 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
3342 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
3344 Inclusive ranges can be defined with ``last``::
3346 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
3348 Note that ``mask`` affects both ``spec`` and ``last``::
3350 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
3351 # matches 10.1.0.0 to 10.2.255.255
3353 Properties can be modified multiple times::
3355 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
3359 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
3364 This section lists supported pattern items and their attributes, if any.
3366 - ``end``: end list of pattern items.
3368 - ``void``: no-op pattern item.
3370 - ``invert``: perform actions when pattern does not match.
3372 - ``any``: match any protocol for the current layer.
3374 - ``num {unsigned}``: number of layers covered.
3376 - ``pf``: match traffic from/to the physical function.
3378 - ``vf``: match traffic from/to a virtual function ID.
3380 - ``id {unsigned}``: VF ID.
3382 - ``phy_port``: match traffic from/to a specific physical port.
3384 - ``index {unsigned}``: physical port index.
3386 - ``port_id``: match traffic from/to a given DPDK port ID.
3388 - ``id {unsigned}``: DPDK port ID.
3390 - ``mark``: match value set in previously matched flow rule using the mark action.
3392 - ``id {unsigned}``: arbitrary integer value.
3394 - ``raw``: match an arbitrary byte string.
3396 - ``relative {boolean}``: look for pattern after the previous item.
3397 - ``search {boolean}``: search pattern from offset (see also limit).
3398 - ``offset {integer}``: absolute or relative offset for pattern.
3399 - ``limit {unsigned}``: search area limit for start of pattern.
3400 - ``pattern {string}``: byte string to look for.
3402 - ``eth``: match Ethernet header.
3404 - ``dst {MAC-48}``: destination MAC.
3405 - ``src {MAC-48}``: source MAC.
3406 - ``type {unsigned}``: EtherType or TPID.
3408 - ``vlan``: match 802.1Q/ad VLAN tag.
3410 - ``tci {unsigned}``: tag control information.
3411 - ``pcp {unsigned}``: priority code point.
3412 - ``dei {unsigned}``: drop eligible indicator.
3413 - ``vid {unsigned}``: VLAN identifier.
3414 - ``inner_type {unsigned}``: inner EtherType or TPID.
3416 - ``ipv4``: match IPv4 header.
3418 - ``tos {unsigned}``: type of service.
3419 - ``ttl {unsigned}``: time to live.
3420 - ``proto {unsigned}``: next protocol ID.
3421 - ``src {ipv4 address}``: source address.
3422 - ``dst {ipv4 address}``: destination address.
3424 - ``ipv6``: match IPv6 header.
3426 - ``tc {unsigned}``: traffic class.
3427 - ``flow {unsigned}``: flow label.
3428 - ``proto {unsigned}``: protocol (next header).
3429 - ``hop {unsigned}``: hop limit.
3430 - ``src {ipv6 address}``: source address.
3431 - ``dst {ipv6 address}``: destination address.
3433 - ``icmp``: match ICMP header.
3435 - ``type {unsigned}``: ICMP packet type.
3436 - ``code {unsigned}``: ICMP packet code.
3438 - ``udp``: match UDP header.
3440 - ``src {unsigned}``: UDP source port.
3441 - ``dst {unsigned}``: UDP destination port.
3443 - ``tcp``: match TCP header.
3445 - ``src {unsigned}``: TCP source port.
3446 - ``dst {unsigned}``: TCP destination port.
3448 - ``sctp``: match SCTP header.
3450 - ``src {unsigned}``: SCTP source port.
3451 - ``dst {unsigned}``: SCTP destination port.
3452 - ``tag {unsigned}``: validation tag.
3453 - ``cksum {unsigned}``: checksum.
3455 - ``vxlan``: match VXLAN header.
3457 - ``vni {unsigned}``: VXLAN identifier.
3459 - ``e_tag``: match IEEE 802.1BR E-Tag header.
3461 - ``grp_ecid_b {unsigned}``: GRP and E-CID base.
3463 - ``nvgre``: match NVGRE header.
3465 - ``tni {unsigned}``: virtual subnet ID.
3467 - ``mpls``: match MPLS header.
3469 - ``label {unsigned}``: MPLS label.
3471 - ``gre``: match GRE header.
3473 - ``protocol {unsigned}``: protocol type.
3475 - ``fuzzy``: fuzzy pattern match, expect faster than default.
3477 - ``thresh {unsigned}``: accuracy threshold.
3479 - ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header.
3481 - ``teid {unsigned}``: tunnel endpoint identifier.
3483 - ``geneve``: match GENEVE header.
3485 - ``vni {unsigned}``: virtual network identifier.
3486 - ``protocol {unsigned}``: protocol type.
3488 - ``vxlan-gpe``: match VXLAN-GPE header.
3490 - ``vni {unsigned}``: VXLAN-GPE identifier.
3492 - ``arp_eth_ipv4``: match ARP header for Ethernet/IPv4.
3494 - ``sha {MAC-48}``: sender hardware address.
3495 - ``spa {ipv4 address}``: sender IPv4 address.
3496 - ``tha {MAC-48}``: target hardware address.
3497 - ``tpa {ipv4 address}``: target IPv4 address.
3499 - ``ipv6_ext``: match presence of any IPv6 extension header.
3501 - ``next_hdr {unsigned}``: next header.
3503 - ``icmp6``: match any ICMPv6 header.
3505 - ``type {unsigned}``: ICMPv6 type.
3506 - ``code {unsigned}``: ICMPv6 code.
3508 - ``icmp6_nd_ns``: match ICMPv6 neighbor discovery solicitation.
3510 - ``target_addr {ipv6 address}``: target address.
3512 - ``icmp6_nd_na``: match ICMPv6 neighbor discovery advertisement.
3514 - ``target_addr {ipv6 address}``: target address.
3516 - ``icmp6_nd_opt``: match presence of any ICMPv6 neighbor discovery option.
3518 - ``type {unsigned}``: ND option type.
3520 - ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery source Ethernet
3521 link-layer address option.
3523 - ``sla {MAC-48}``: source Ethernet LLA.
3525 - ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery target Ethernet
3526 link-layer address option.
3528 - ``tla {MAC-48}``: target Ethernet LLA.
3533 A list of actions starts after the ``actions`` token in the same fashion as
3534 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
3535 terminated by a mandatory ``end`` action.
3537 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
3538 rte_flow_action_type``).
3540 Dropping all incoming UDPv4 packets can be expressed as follows::
3542 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3545 Several actions have configurable properties which must be specified when
3546 there is no valid default value. For example, ``queue`` requires a target
3549 This rule redirects incoming UDPv4 traffic to queue index 6::
3551 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3552 actions queue index 6 / end
3554 While this one could be rejected by PMDs (unspecified queue index)::
3556 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3559 As defined by *rte_flow*, the list is not ordered, all actions of a given
3560 rule are performed simultaneously. These are equivalent::
3562 queue index 6 / void / mark id 42 / end
3566 void / mark id 42 / queue index 6 / end
3568 All actions in a list should have different types, otherwise only the last
3569 action of a given type is taken into account::
3571 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
3575 drop / drop / drop / end # drop is performed only once
3579 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
3581 Considering they are performed simultaneously, opposite and overlapping
3582 actions can sometimes be combined when the end result is unambiguous::
3584 drop / queue index 6 / end # drop has no effect
3588 queue index 6 / rss queues 6 7 8 / end # queue has no effect
3592 drop / passthru / end # drop has no effect
3594 Note that PMDs may still refuse such combinations.
3599 This section lists supported actions and their attributes, if any.
3601 - ``end``: end list of actions.
3603 - ``void``: no-op action.
3605 - ``passthru``: let subsequent rule process matched packets.
3607 - ``jump``: redirect traffic to group on device.
3609 - ``group {unsigned}``: group to redirect to.
3611 - ``mark``: attach 32 bit value to packets.
3613 - ``id {unsigned}``: 32 bit value to return with packets.
3615 - ``flag``: flag packets.
3617 - ``queue``: assign packets to a given queue index.
3619 - ``index {unsigned}``: queue index to use.
3621 - ``drop``: drop packets (note: passthru has priority).
3623 - ``count``: enable counters for this rule.
3625 - ``rss``: spread packets among several queues.
3627 - ``func {hash function}``: RSS hash function to apply, allowed tokens are
3628 the same as `set_hash_global_config`_.
3630 - ``level {unsigned}``: encapsulation level for ``types``.
3632 - ``types [{RSS hash type} [...]] end``: specific RSS hash types, allowed
3633 tokens are the same as `set_hash_input_set`_, except that an empty list
3634 does not disable RSS but instead requests unspecified "best-effort"
3637 - ``key {string}``: RSS hash key, overrides ``key_len``.
3639 - ``key_len {unsigned}``: RSS hash key length in bytes, can be used in
3640 conjunction with ``key`` to pad or truncate it.
3642 - ``queues [{unsigned} [...]] end``: queue indices to use.
3644 - ``pf``: direct traffic to physical function.
3646 - ``vf``: direct traffic to a virtual function ID.
3648 - ``original {boolean}``: use original VF ID if possible.
3649 - ``id {unsigned}``: VF ID.
3651 - ``phy_port``: direct packets to physical port index.
3653 - ``original {boolean}``: use original port index if possible.
3654 - ``index {unsigned}``: physical port index.
3656 - ``port_id``: direct matching traffic to a given DPDK port ID.
3658 - ``original {boolean}``: use original DPDK port ID if possible.
3659 - ``id {unsigned}``: DPDK port ID.
3661 - ``of_set_mpls_ttl``: OpenFlow's ``OFPAT_SET_MPLS_TTL``.
3663 - ``mpls_ttl``: MPLS TTL.
3665 - ``of_dec_mpls_ttl``: OpenFlow's ``OFPAT_DEC_MPLS_TTL``.
3667 - ``of_set_nw_ttl``: OpenFlow's ``OFPAT_SET_NW_TTL``.
3669 - ``nw_ttl``: IP TTL.
3671 - ``of_dec_nw_ttl``: OpenFlow's ``OFPAT_DEC_NW_TTL``.
3673 - ``of_copy_ttl_out``: OpenFlow's ``OFPAT_COPY_TTL_OUT``.
3675 - ``of_copy_ttl_in``: OpenFlow's ``OFPAT_COPY_TTL_IN``.
3677 - ``of_pop_vlan``: OpenFlow's ``OFPAT_POP_VLAN``.
3679 - ``of_push_vlan``: OpenFlow's ``OFPAT_PUSH_VLAN``.
3681 - ``ethertype``: Ethertype.
3683 - ``of_set_vlan_vid``: OpenFlow's ``OFPAT_SET_VLAN_VID``.
3685 - ``vlan_vid``: VLAN id.
3687 - ``of_set_vlan_pcp``: OpenFlow's ``OFPAT_SET_VLAN_PCP``.
3689 - ``vlan_pcp``: VLAN priority.
3691 - ``of_pop_mpls``: OpenFlow's ``OFPAT_POP_MPLS``.
3693 - ``ethertype``: Ethertype.
3695 - ``of_push_mpls``: OpenFlow's ``OFPAT_PUSH_MPLS``.
3697 - ``ethertype``: Ethertype.
3699 - ``vxlan_encap``: Performs a VXLAN encapsulation, outer layer configuration
3700 is done through `Config VXLAN Encap outer layers`_.
3702 - ``vxlan_decap``: Performs a decapsulation action by stripping all headers of
3703 the VXLAN tunnel network overlay from the matched flow.
3705 - ``nvgre_encap``: Performs a NVGRE encapsulation, outer layer configuration
3706 is done through `Config NVGRE Encap outer layers`_.
3708 - ``nvgre_decap``: Performs a decapsulation action by stripping all headers of
3709 the NVGRE tunnel network overlay from the matched flow.
3711 - ``set_ipv4_src``: Set a new IPv4 source address in the outermost IPv4 header.
3713 - ``ipv4_addr``: New IPv4 source address.
3715 - ``set_ipv4_dst``: Set a new IPv4 destination address in the outermost IPv4
3718 - ``ipv4_addr``: New IPv4 destination address.
3720 - ``set_ipv6_src``: Set a new IPv6 source address in the outermost IPv6 header.
3722 - ``ipv6_addr``: New IPv6 source address.
3724 - ``set_ipv6_dst``: Set a new IPv6 destination address in the outermost IPv6
3727 - ``ipv6_addr``: New IPv6 destination address.
3729 - ``of_set_tp_src``: Set a new source port number in the outermost TCP/UDP
3732 - ``port``: New TCP/UDP source port number.
3734 - ``of_set_tp_dst``: Set a new destination port number in the outermost TCP/UDP
3737 - ``port``: New TCP/UDP destination port number.
3739 - ``mac_swap``: Swap the source and destination MAC addresses in the outermost
3742 Destroying flow rules
3743 ~~~~~~~~~~~~~~~~~~~~~
3745 ``flow destroy`` destroys one or more rules from their rule ID (as returned
3746 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
3747 times as necessary::
3749 flow destroy {port_id} rule {rule_id} [...]
3751 If successful, it will show::
3753 Flow rule #[...] destroyed
3755 It does not report anything for rule IDs that do not exist. The usual error
3756 message is shown when a rule cannot be destroyed::
3758 Caught error type [...] ([...]): [...]
3760 ``flow flush`` destroys all rules on a device and does not take extra
3761 arguments. It is bound to ``rte_flow_flush()``::
3763 flow flush {port_id}
3765 Any errors are reported as above.
3767 Creating several rules and destroying them::
3769 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3770 actions queue index 2 / end
3771 Flow rule #0 created
3772 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3773 actions queue index 3 / end
3774 Flow rule #1 created
3775 testpmd> flow destroy 0 rule 0 rule 1
3776 Flow rule #1 destroyed
3777 Flow rule #0 destroyed
3780 The same result can be achieved using ``flow flush``::
3782 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3783 actions queue index 2 / end
3784 Flow rule #0 created
3785 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3786 actions queue index 3 / end
3787 Flow rule #1 created
3788 testpmd> flow flush 0
3791 Non-existent rule IDs are ignored::
3793 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3794 actions queue index 2 / end
3795 Flow rule #0 created
3796 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3797 actions queue index 3 / end
3798 Flow rule #1 created
3799 testpmd> flow destroy 0 rule 42 rule 10 rule 2
3801 testpmd> flow destroy 0 rule 0
3802 Flow rule #0 destroyed
3808 ``flow query`` queries a specific action of a flow rule having that
3809 ability. Such actions collect information that can be reported using this
3810 command. It is bound to ``rte_flow_query()``::
3812 flow query {port_id} {rule_id} {action}
3814 If successful, it will display either the retrieved data for known actions
3815 or the following message::
3817 Cannot display result for action type [...] ([...])
3819 Otherwise, it will complain either that the rule does not exist or that some
3822 Flow rule #[...] not found
3826 Caught error type [...] ([...]): [...]
3828 Currently only the ``count`` action is supported. This action reports the
3829 number of packets that hit the flow rule and the total number of bytes. Its
3830 output has the following format::
3833 hits_set: [...] # whether "hits" contains a valid value
3834 bytes_set: [...] # whether "bytes" contains a valid value
3835 hits: [...] # number of packets
3836 bytes: [...] # number of bytes
3838 Querying counters for TCPv6 packets redirected to queue 6::
3840 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
3841 actions queue index 6 / count / end
3842 Flow rule #4 created
3843 testpmd> flow query 0 4 count
3854 ``flow list`` lists existing flow rules sorted by priority and optionally
3855 filtered by group identifiers::
3857 flow list {port_id} [group {group_id}] [...]
3859 This command only fails with the following message if the device does not
3864 Output consists of a header line followed by a short description of each
3865 flow rule, one per line. There is no output at all when no flow rules are
3866 configured on the device::
3868 ID Group Prio Attr Rule
3869 [...] [...] [...] [...] [...]
3871 ``Attr`` column flags:
3873 - ``i`` for ``ingress``.
3874 - ``e`` for ``egress``.
3876 Creating several flow rules and listing them::
3878 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3879 actions queue index 6 / end
3880 Flow rule #0 created
3881 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3882 actions queue index 2 / end
3883 Flow rule #1 created
3884 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
3885 actions rss queues 6 7 8 end / end
3886 Flow rule #2 created
3887 testpmd> flow list 0
3888 ID Group Prio Attr Rule
3889 0 0 0 i- ETH IPV4 => QUEUE
3890 1 0 0 i- ETH IPV6 => QUEUE
3891 2 0 5 i- ETH IPV4 UDP => RSS
3894 Rules are sorted by priority (i.e. group ID first, then priority level)::
3896 testpmd> flow list 1
3897 ID Group Prio Attr Rule
3898 0 0 0 i- ETH => COUNT
3899 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3900 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3901 1 24 0 i- ETH IPV4 UDP => QUEUE
3902 4 24 10 i- ETH IPV4 TCP => DROP
3903 3 24 20 i- ETH IPV4 => DROP
3904 2 24 42 i- ETH IPV4 UDP => QUEUE
3905 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3908 Output can be limited to specific groups::
3910 testpmd> flow list 1 group 0 group 63
3911 ID Group Prio Attr Rule
3912 0 0 0 i- ETH => COUNT
3913 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3914 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3915 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3918 Toggling isolated mode
3919 ~~~~~~~~~~~~~~~~~~~~~~
3921 ``flow isolate`` can be used to tell the underlying PMD that ingress traffic
3922 must only be injected from the defined flow rules; that no default traffic
3923 is expected outside those rules and the driver is free to assign more
3924 resources to handle them. It is bound to ``rte_flow_isolate()``::
3926 flow isolate {port_id} {boolean}
3928 If successful, enabling or disabling isolated mode shows either::
3930 Ingress traffic on port [...]
3931 is now restricted to the defined flow rules
3935 Ingress traffic on port [...]
3936 is not restricted anymore to the defined flow rules
3938 Otherwise, in case of error::
3940 Caught error type [...] ([...]): [...]
3942 Mainly due to its side effects, PMDs supporting this mode may not have the
3943 ability to toggle it more than once without reinitializing affected ports
3944 first (e.g. by exiting testpmd).
3946 Enabling isolated mode::
3948 testpmd> flow isolate 0 true
3949 Ingress traffic on port 0 is now restricted to the defined flow rules
3952 Disabling isolated mode::
3954 testpmd> flow isolate 0 false
3955 Ingress traffic on port 0 is not restricted anymore to the defined flow rules
3958 Sample QinQ flow rules
3959 ~~~~~~~~~~~~~~~~~~~~~~
3961 Before creating QinQ rule(s) the following commands should be issued to enable QinQ::
3963 testpmd> port stop 0
3964 testpmd> vlan set qinq on 0
3966 The above command sets the inner and outer TPID's to 0x8100.
3968 To change the TPID's the following commands should be used::
3970 testpmd> vlan set outer tpid 0xa100 0
3971 testpmd> vlan set inner tpid 0x9100 0
3972 testpmd> port start 0
3974 Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM.
3978 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 /
3979 vlan tci is 456 / end actions vf id 1 / queue index 0 / end
3980 Flow rule #0 validated
3982 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 /
3983 vlan tci is 456 / end actions vf id 123 / queue index 0 / end
3984 Flow rule #0 created
3986 testpmd> flow list 0
3987 ID Group Prio Attr Rule
3988 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3990 Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host.
3994 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 /
3995 vlan tci is 654 / end actions pf / queue index 0 / end
3996 Flow rule #1 validated
3998 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 /
3999 vlan tci is 654 / end actions pf / queue index 1 / end
4000 Flow rule #1 created
4002 testpmd> flow list 0
4003 ID Group Prio Attr Rule
4004 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
4005 1 0 0 i- ETH VLAN VLAN=>PF QUEUE
4007 Sample VXLAN encapsulation rule
4008 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
4010 VXLAN encapsulation outer layer has default value pre-configured in testpmd
4011 source code, those can be changed by using the following commands
4013 IPv4 VXLAN outer header::
4015 testpmd> set vxlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src 127.0.0.1
4016 ip-dst 128.0.0.1 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
4017 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
4020 testpmd> set vxlan-with-vlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src
4021 127.0.0.1 ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11
4022 eth-dst 22:22:22:22:22:22
4023 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
4026 IPv6 VXLAN outer header::
4028 testpmd> set vxlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4 ip-src ::1
4029 ip-dst ::2222 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
4030 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
4033 testpmd> set vxlan-with-vlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4
4034 ip-src ::1 ip-dst ::2222 vlan-tci 34 eth-src 11:11:11:11:11:11
4035 eth-dst 22:22:22:22:22:22
4036 testpmd> flow create 0 ingress pattern end actions vxlan_encap /
4039 Sample NVGRE encapsulation rule
4040 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
4042 NVGRE encapsulation outer layer has default value pre-configured in testpmd
4043 source code, those can be changed by using the following commands
4045 IPv4 NVGRE outer header::
4047 testpmd> set nvgre ip-version ipv4 tni 4 ip-src 127.0.0.1 ip-dst 128.0.0.1
4048 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
4049 testpmd> flow create 0 ingress pattern end actions nvgre_encap /
4052 testpmd> set nvgre-with-vlan ip-version ipv4 tni 4 ip-src 127.0.0.1
4053 ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11
4054 eth-dst 22:22:22:22:22:22
4055 testpmd> flow create 0 ingress pattern end actions nvgre_encap /
4058 IPv6 NVGRE outer header::
4060 testpmd> set nvgre ip-version ipv6 tni 4 ip-src ::1 ip-dst ::2222
4061 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
4062 testpmd> flow create 0 ingress pattern end actions nvgre_encap /
4065 testpmd> set nvgre-with-vlan ip-version ipv6 tni 4 ip-src ::1 ip-dst ::2222
4066 vlan-tci 34 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22
4067 testpmd> flow create 0 ingress pattern end actions nvgre_encap /
4073 The following sections show functions to load/unload eBPF based filters.
4078 Load an eBPF program as a callback for partciular RX/TX queue::
4080 testpmd> bpf-load rx|tx (portid) (queueid) (load-flags) (bpf-prog-filename)
4082 The available load-flags are:
4084 * ``J``: use JIT generated native code, otherwise BPF interpreter will be used.
4086 * ``M``: assume input parameter is a pointer to rte_mbuf, otherwise assume it is a pointer to first segment's data.
4092 You'll need clang v3.7 or above to build bpf program you'd like to load
4096 .. code-block:: console
4099 clang -O2 -target bpf -c t1.c
4101 Then to load (and JIT compile) t1.o at RX queue 0, port 1::
4103 .. code-block:: console
4105 testpmd> bpf-load rx 1 0 J ./dpdk.org/test/bpf/t1.o
4107 To load (not JITed) t1.o at TX queue 0, port 0::
4109 .. code-block:: console
4111 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o
4116 Unload previously loaded eBPF program for partciular RX/TX queue::
4118 testpmd> bpf-unload rx|tx (portid) (queueid)
4120 For example to unload BPF filter from TX queue 0, port 0:
4122 .. code-block:: console
4124 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o