2 Copyright(c) 2010-2016 Intel Corporation. All rights reserved.
5 Redistribution and use in source and binary forms, with or without
6 modification, are permitted provided that the following conditions
9 * Redistributions of source code must retain the above copyright
10 notice, this list of conditions and the following disclaimer.
11 * Redistributions in binary form must reproduce the above copyright
12 notice, this list of conditions and the following disclaimer in
13 the documentation and/or other materials provided with the
15 * Neither the name of Intel Corporation nor the names of its
16 contributors may be used to endorse or promote products derived
17 from this software without specific prior written permission.
19 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
20 "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
21 LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
22 A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
23 OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
24 SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
25 LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
26 DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
27 THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
28 (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
29 OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
33 Testpmd Runtime Functions
34 =========================
36 Where the testpmd application is started in interactive mode, (``-i|--interactive``),
37 it displays a prompt that can be used to start and stop forwarding,
38 configure the application, display statistics (including the extended NIC
39 statistics aka xstats) , set the Flow Director and other tasks::
43 The testpmd prompt has some, limited, readline support.
44 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
45 as well as access to the command history via the up-arrow.
47 There is also support for tab completion.
48 If you type a partial command and hit ``<TAB>`` you get a list of the available completions:
50 .. code-block:: console
52 testpmd> show port <TAB>
54 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
55 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
56 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
57 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
63 Some examples in this document are too long to fit on one line are are shown wrapped at `"\\"` for display purposes::
65 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
66 (pause_time) (send_xon) (port_id)
68 In the real ``testpmd>`` prompt these commands should be on a single line.
73 The testpmd has on-line help for the functions that are available at runtime.
74 These are divided into sections and can be accessed using help, help section or help all:
76 .. code-block:: console
80 help control : Start and stop forwarding.
81 help display : Displaying port, stats and config information.
82 help config : Configuration information.
83 help ports : Configuring ports.
84 help registers : Reading and setting port registers.
85 help filters : Filters configuration help.
86 help all : All of the above sections.
89 Command File Functions
90 ----------------------
92 To facilitate loading large number of commands or to avoid cutting and pasting where not
93 practical or possible testpmd supports alternative methods for executing commands.
95 * If started with the ``--cmdline-file=FILENAME`` command line argument testpmd
96 will execute all CLI commands contained within the file immediately before
97 starting packet forwarding or entering interactive mode.
99 .. code-block:: console
101 ./testpmd -n4 -r2 ... -- -i --cmdline-file=/home/ubuntu/flow-create-commands.txt
102 Interactive-mode selected
103 CLI commands to be read from /home/ubuntu/flow-create-commands.txt
104 Configuring Port 0 (socket 0)
105 Port 0: 7C:FE:90:CB:74:CE
106 Configuring Port 1 (socket 0)
107 Port 1: 7C:FE:90:CB:74:CA
108 Checking link statuses...
109 Port 0 Link Up - speed 10000 Mbps - full-duplex
110 Port 1 Link Up - speed 10000 Mbps - full-duplex
116 Flow rule #498 created
117 Flow rule #499 created
118 Read all CLI commands from /home/ubuntu/flow-create-commands.txt
122 * At run-time additional commands can be loaded in bulk by invoking the ``load FILENAME``
125 .. code-block:: console
127 testpmd> load /home/ubuntu/flow-create-commands.txt
132 Flow rule #498 created
133 Flow rule #499 created
134 Read all CLI commands from /home/ubuntu/flow-create-commands.txt
138 In all cases output from any included command will be displayed as standard output.
139 Execution will continue until the end of the file is reached regardless of
140 whether any errors occur. The end user must examine the output to determine if
141 any failures occurred.
150 Start packet forwarding with current configuration::
157 Start packet forwarding with current configuration after sending specified number of bursts of packets::
159 testpmd> start tx_first (""|burst_num)
161 The default burst number is 1 when ``burst_num`` not presented.
166 Stop packet forwarding, and display accumulated statistics::
181 The functions in the following sections are used to display information about the
182 testpmd configuration or the NIC status.
187 Display information for a given port or all ports::
189 testpmd> show port (info|stats|xstats|fdir|stat_qmap|dcb_tc|cap) (port_id|all)
191 The available information categories are:
193 * ``info``: General port information such as MAC address.
195 * ``stats``: RX/TX statistics.
197 * ``xstats``: RX/TX extended NIC statistics.
199 * ``fdir``: Flow Director information and statistics.
201 * ``stat_qmap``: Queue statistics mapping.
203 * ``dcb_tc``: DCB information such as TC mapping.
205 * ``cap``: Supported offload capabilities.
209 .. code-block:: console
211 testpmd> show port info 0
213 ********************* Infos for port 0 *********************
215 MAC address: XX:XX:XX:XX:XX:XX
217 memory allocation on the socket: 0
219 Link speed: 40000 Mbps
220 Link duplex: full-duplex
221 Promiscuous mode: enabled
222 Allmulticast mode: disabled
223 Maximum number of MAC addresses: 64
224 Maximum number of MAC addresses of hash filtering: 0
229 Redirection table size: 512
230 Supported flow types:
250 Display the rss redirection table entry indicated by masks on port X::
252 testpmd> show port (port_id) rss reta (size) (mask0, mask1...)
254 size is used to indicate the hardware supported reta size
259 Display the RSS hash functions and RSS hash key of a port::
261 testpmd> show port (port_id) rss-hash ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2-payload|ipv6-ex|ipv6-tcp-ex|ipv6-udp-ex [key]
266 Clear the port statistics for a given port or for all ports::
268 testpmd> clear port (info|stats|xstats|fdir|stat_qmap) (port_id|all)
272 testpmd> clear port stats all
277 Display information for a given port's RX/TX queue::
279 testpmd> show (rxq|txq) info (port_id) (queue_id)
284 Displays the configuration of the application.
285 The configuration comes from the command-line, the runtime or the application defaults::
287 testpmd> show config (rxtx|cores|fwd|txpkts)
289 The available information categories are:
291 * ``rxtx``: RX/TX configuration items.
293 * ``cores``: List of forwarding cores.
295 * ``fwd``: Packet forwarding configuration.
297 * ``txpkts``: Packets to TX configuration.
301 .. code-block:: console
303 testpmd> show config rxtx
305 io packet forwarding - CRC stripping disabled - packets/burst=16
306 nb forwarding cores=2 - nb forwarding ports=1
307 RX queues=1 - RX desc=128 - RX free threshold=0
308 RX threshold registers: pthresh=8 hthresh=8 wthresh=4
309 TX queues=1 - TX desc=512 - TX free threshold=0
310 TX threshold registers: pthresh=36 hthresh=0 wthresh=0
311 TX RS bit threshold=0 - TXQ flags=0x0
316 Set the packet forwarding mode::
318 testpmd> set fwd (io|mac|macswap|flowgen| \
319 rxonly|txonly|csum|icmpecho) (""|retry)
321 ``retry`` can be specified for forwarding engines except ``rx_only``.
323 The available information categories are:
325 * ``io``: Forwards packets "as-is" in I/O mode.
326 This is the fastest possible forwarding operation as it does not access packets data.
327 This is the default mode.
329 * ``mac``: Changes the source and the destination Ethernet addresses of packets before forwarding them.
330 Default application behaviour is to set source Ethernet address to that of the transmitting interface, and destination
331 address to a dummy value (set during init). The user may specify a target destination Ethernet address via the 'eth-peer' or
332 'eth-peer-configfile' command-line options. It is not currently possible to specify a specific source Ethernet address.
334 * ``macswap``: MAC swap forwarding mode.
335 Swaps the source and the destination Ethernet addresses of packets before forwarding them.
337 * ``flowgen``: Multi-flow generation mode.
338 Originates a number of flows (with varying destination IP addresses), and terminate receive traffic.
340 * ``rxonly``: Receives packets but doesn't transmit them.
342 * ``txonly``: Generates and transmits packets without receiving any.
344 * ``csum``: Changes the checksum field with hardware or software methods depending on the offload flags on the packet.
346 * ``icmpecho``: Receives a burst of packets, lookup for IMCP echo requests and, if any, send back ICMP echo replies.
348 * ``ieee1588``: Demonstrate L2 IEEE1588 V2 PTP timestamping for RX and TX. Requires ``CONFIG_RTE_LIBRTE_IEEE1588=y``.
350 * ``tm``: Traffic Management forwarding mode
351 Demonstrates the use of ethdev traffic management APIs and softnic PMD for
352 QoS traffic management. In this mode, 5-level hierarchical QoS scheduler is
353 available as an default option that can be enabled through CLI. The user can
354 also modify the default hierarchy or specify the new hierarchy through CLI for
355 implementing QoS scheduler. Requires ``CONFIG_RTE_LIBRTE_PMD_SOFTNIC=y`` ``CONFIG_RTE_LIBRTE_SCHED=y``.
359 testpmd> set fwd rxonly
361 Set rxonly packet forwarding mode
367 Display an RX descriptor for a port RX queue::
369 testpmd> read rxd (port_id) (queue_id) (rxd_id)
373 testpmd> read rxd 0 0 4
374 0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180
379 Display a TX descriptor for a port TX queue::
381 testpmd> read txd (port_id) (queue_id) (txd_id)
385 testpmd> read txd 0 0 4
386 0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C
391 Get loaded dynamic device personalization (DDP) package info list::
393 testpmd> ddp get list (port_id)
398 Display information about dynamic device personalization (DDP) profile::
400 testpmd> ddp get info (profile_path)
405 Display VF statistics::
407 testpmd> show vf stats (port_id) (vf_id)
412 Reset VF statistics::
414 testpmd> clear vf stats (port_id) (vf_id)
416 show port pctype mapping
417 ~~~~~~~~~~~~~~~~~~~~~~~~
419 List all items from the pctype mapping table::
421 testpmd> show port (port_id) pctype mapping
424 Configuration Functions
425 -----------------------
427 The testpmd application can be configured from the runtime as well as from the command-line.
429 This section details the available configuration functions that are available.
433 Configuration changes only become active when forwarding is started/restarted.
438 Reset forwarding to the default configuration::
445 Set the debug verbosity level::
447 testpmd> set verbose (level)
449 Currently the only available levels are 0 (silent except for error) and 1 (fully verbose).
454 Set the log level for a log type::
456 testpmd> set log global|(type) (level)
460 * ``type`` is the log name.
462 * ``level`` is the log level.
464 For example, to change the global log level::
465 testpmd> set log global (level)
467 Regexes can also be used for type. To change log level of user1, user2 and user3::
468 testpmd> set log user[1-3] (level)
473 Set the number of ports used by the application:
477 This is equivalent to the ``--nb-ports`` command-line option.
482 Set the number of cores used by the application::
484 testpmd> set nbcore (num)
486 This is equivalent to the ``--nb-cores`` command-line option.
490 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port.
495 Set the forwarding cores hexadecimal mask::
497 testpmd> set coremask (mask)
499 This is equivalent to the ``--coremask`` command-line option.
503 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding.
508 Set the forwarding ports hexadecimal mask::
510 testpmd> set portmask (mask)
512 This is equivalent to the ``--portmask`` command-line option.
517 Set number of packets per burst::
519 testpmd> set burst (num)
521 This is equivalent to the ``--burst command-line`` option.
523 When retry is enabled, the transmit delay time and number of retries can also be set::
525 testpmd> set burst tx delay (microseconds) retry (num)
530 Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode::
532 testpmd> set txpkts (x[,y]*)
534 Where x[,y]* represents a CSV list of values, without white space.
539 Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes::
541 testpmd> set txsplit (off|on|rand)
545 * ``off`` disable packet copy & split for CSUM mode.
547 * ``on`` split outgoing packet into multiple segments. Size of each segment
548 and number of segments per packet is determined by ``set txpkts`` command
551 * ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments.
556 Set the list of forwarding cores::
558 testpmd> set corelist (x[,y]*)
560 For example, to change the forwarding cores:
562 .. code-block:: console
564 testpmd> set corelist 3,1
565 testpmd> show config fwd
567 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled
568 Logical Core 3 (socket 0) forwards packets on 1 streams:
569 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01
570 Logical Core 1 (socket 0) forwards packets on 1 streams:
571 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
575 The cores are used in the same order as specified on the command line.
580 Set the list of forwarding ports::
582 testpmd> set portlist (x[,y]*)
584 For example, to change the port forwarding:
586 .. code-block:: console
588 testpmd> set portlist 0,2,1,3
589 testpmd> show config fwd
591 io packet forwarding - ports=4 - cores=1 - streams=4
592 Logical Core 3 (socket 0) forwards packets on 4 streams:
593 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01
594 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
595 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03
596 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02
601 Enable/disable tx loopback::
603 testpmd> set tx loopback (port_id) (on|off)
608 set drop enable bit for all queues::
610 testpmd> set all queues drop (port_id) (on|off)
612 set split drop enable (for VF)
613 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
615 set split drop enable bit for VF from PF::
617 testpmd> set vf split drop (port_id) (vf_id) (on|off)
619 set mac antispoof (for VF)
620 ~~~~~~~~~~~~~~~~~~~~~~~~~~
622 Set mac antispoof for a VF from the PF::
624 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off)
629 Enable/disable MACsec offload::
631 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off)
632 testpmd> set macsec offload (port_id) off
637 Configure MACsec secure connection (SC)::
639 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi)
643 The pi argument is ignored for tx.
644 Check the NIC Datasheet for hardware limits.
649 Configure MACsec secure association (SA)::
651 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key)
655 The IDX value must be 0 or 1.
656 Check the NIC Datasheet for hardware limits.
658 set broadcast mode (for VF)
659 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
661 Set broadcast mode for a VF from the PF::
663 testpmd> set vf broadcast (port_id) (vf_id) (on|off)
668 Set the VLAN strip on a port::
670 testpmd> vlan set strip (on|off) (port_id)
675 Set the VLAN strip for a queue on a port::
677 testpmd> vlan set stripq (on|off) (port_id,queue_id)
679 vlan set stripq (for VF)
680 ~~~~~~~~~~~~~~~~~~~~~~~~
682 Set VLAN strip for all queues in a pool for a VF from the PF::
684 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off)
686 vlan set insert (for VF)
687 ~~~~~~~~~~~~~~~~~~~~~~~~
689 Set VLAN insert for a VF from the PF::
691 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id)
693 vlan set tag (for VF)
694 ~~~~~~~~~~~~~~~~~~~~~
696 Set VLAN tag for a VF from the PF::
698 testpmd> set vf vlan tag (port_id) (vf_id) (on|off)
700 vlan set antispoof (for VF)
701 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
703 Set VLAN antispoof for a VF from the PF::
705 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off)
710 Set the VLAN filter on a port::
712 testpmd> vlan set filter (on|off) (port_id)
717 Set the VLAN QinQ (extended queue in queue) on for a port::
719 testpmd> vlan set qinq (on|off) (port_id)
724 Set the inner or outer VLAN TPID for packet filtering on a port::
726 testpmd> vlan set (inner|outer) tpid (value) (port_id)
730 TPID value must be a 16-bit number (value <= 65536).
735 Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID::
737 testpmd> rx_vlan add (vlan_id|all) (port_id)
741 VLAN filter must be set on that port. VLAN ID < 4096.
742 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries
743 in VFTA table. This is important if enabling all vlan_ids.
748 Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID::
750 testpmd> rx_vlan rm (vlan_id|all) (port_id)
755 Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID::
757 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask)
762 Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID::
764 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask)
769 Add a tunnel filter on a port::
771 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \
772 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
773 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
775 The available information categories are:
777 * ``vxlan``: Set tunnel type as VXLAN.
779 * ``nvgre``: Set tunnel type as NVGRE.
781 * ``ipingre``: Set tunnel type as IP-in-GRE.
783 * ``imac-ivlan``: Set filter type as Inner MAC and VLAN.
785 * ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID.
787 * ``imac-tenid``: Set filter type as Inner MAC and tenant ID.
789 * ``imac``: Set filter type as Inner MAC.
791 * ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID.
793 * ``oip``: Set filter type as Outer IP.
795 * ``iip``: Set filter type as Inner IP.
799 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \
800 192.168.2.2 0 ipingre oip 1 1
802 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP.
807 Remove a tunnel filter on a port::
809 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \
810 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
811 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
816 Add an UDP port for VXLAN packet filter on a port::
818 testpmd> rx_vxlan_port add (udp_port) (port_id)
823 Remove an UDP port for VXLAN packet filter on a port::
825 testpmd> rx_vxlan_port rm (udp_port) (port_id)
830 Set hardware insertion of VLAN IDs in packets sent on a port::
832 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer]
834 For example, set a single VLAN ID (5) insertion on port 0::
838 Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1::
846 Set port based hardware insertion of VLAN ID in packets sent on a port::
848 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off)
853 Disable hardware insertion of a VLAN header in packets sent on a port::
855 testpmd> tx_vlan reset (port_id)
860 Select hardware or software calculation of the checksum when
861 transmitting a packet using the ``csum`` forwarding engine::
863 testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id)
867 * ``ip|udp|tcp|sctp`` always relate to the inner layer.
869 * ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized
870 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are
871 supported). See also the ``csum parse-tunnel`` command.
875 Check the NIC Datasheet for hardware limits.
880 Set RSS queue region span on a port::
882 testpmd> set port (port_id) queue-region region_id (value) \
883 queue_start_index (value) queue_num (value)
885 Set flowtype mapping on a RSS queue region on a port::
887 testpmd> set port (port_id) queue-region region_id (value) flowtype (value)
891 * For the flowtype(pctype) of packet,the specific index for each type has
892 been defined in file i40e_type.h as enum i40e_filter_pctype.
894 Set user priority mapping on a RSS queue region on a port::
896 testpmd> set port (port_id) queue-region UP (value) region_id (value)
898 Flush all queue region related configuration on a port::
900 testpmd> set port (port_id) queue-region flush (on|off)
904 * "on"is just an enable function which server for other configuration,
905 it is for all configuration about queue region from up layer,
906 at first will only keep in DPDK softwarestored in driver,
907 only after "flush on", it commit all configuration to HW.
908 "off" is just clean all configuration about queue region just now,
909 and restore all to DPDK i40e driver default config when start up.
911 Show all queue region related configuration info on a port::
913 testpmd> show port (port_id) queue-region
917 Queue region only support on PF by now, so these command is
918 only for configuration of queue region on PF port.
923 Define how tunneled packets should be handled by the csum forward
926 testpmd> csum parse-tunnel (on|off) (tx_port_id)
928 If enabled, the csum forward engine will try to recognize supported
929 tunnel headers (vxlan, gre, ipip).
931 If disabled, treat tunnel packets as non-tunneled packets (a inner
932 header is handled as a packet payload).
936 The port argument is the TX port like in the ``csum set`` command.
940 Consider a packet in packet like the following::
942 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in
944 * If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
945 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the
946 ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``).
948 * If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
949 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``.
954 Display tx checksum offload configuration::
956 testpmd> csum show (port_id)
961 Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine::
963 testpmd> tso set (segsize) (port_id)
967 Check the NIC datasheet for hardware limits.
972 Display the status of TCP Segmentation Offload::
974 testpmd> tso show (port_id)
979 Enable or disable GRO in ``csum`` forwarding engine::
981 testpmd> set port <port_id> gro on|off
983 If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4
984 packets received from the given port.
986 If disabled, packets received from the given port won't be performed
987 GRO. By default, GRO is disabled for all ports.
991 When enable GRO for a port, TCP/IPv4 packets received from the port
992 will be performed GRO. After GRO, all merged packets have bad
993 checksums, since the GRO library doesn't re-calculate checksums for
994 the merged packets. Therefore, if users want the merged packets to
995 have correct checksums, please select HW IP checksum calculation and
996 HW TCP checksum calculation for the port which the merged packets are
1002 Display GRO configuration for a given port::
1004 testpmd> show port <port_id> gro
1009 Set the cycle to flush the GROed packets from reassembly tables::
1011 testpmd> set gro flush <cycles>
1013 When enable GRO, the csum forwarding engine performs GRO on received
1014 packets, and the GROed packets are stored in reassembly tables. Users
1015 can use this command to determine when the GROed packets are flushed
1016 from the reassembly tables.
1018 The ``cycles`` is measured in GRO operation times. The csum forwarding
1019 engine flushes the GROed packets from the tables every ``cycles`` GRO
1022 By default, the value of ``cycles`` is 1, which means flush GROed packets
1023 from the reassembly tables as soon as one GRO operation finishes. The value
1024 of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``.
1026 Please note that the large value of ``cycles`` may cause the poor TCP/IP
1027 stack performance. Because the GROed packets are delayed to arrive the
1028 stack, thus causing more duplicated ACKs and TCP retransmissions.
1033 Toggle per-port GSO support in ``csum`` forwarding engine::
1035 testpmd> set port <port_id> gso on|off
1037 If enabled, the csum forwarding engine will perform GSO on supported IPv4
1038 packets, transmitted on the given port.
1040 If disabled, packets transmitted on the given port will not undergo GSO.
1041 By default, GSO is disabled for all ports.
1045 When GSO is enabled on a port, supported IPv4 packets transmitted on that
1046 port undergo GSO. Afterwards, the segmented packets are represented by
1047 multi-segment mbufs; however, the csum forwarding engine doesn't calculation
1048 of checksums for GSO'd segments in SW. As a result, if users want correct
1049 checksums in GSO segments, they should enable HW checksum calculation for
1052 For example, HW checksum calculation for VxLAN GSO'd packets may be enabled
1053 by setting the following options in the csum forwarding engine:
1055 testpmd> csum set outer_ip hw <port_id>
1057 testpmd> csum set ip hw <port_id>
1059 testpmd> csum set tcp hw <port_id>
1064 Set the maximum GSO segment size (measured in bytes), which includes the
1065 packet header and the packet payload for GSO-enabled ports (global)::
1067 testpmd> set gso segsz <length>
1072 Display the status of Generic Segmentation Offload for a given port::
1074 testpmd> show port <port_id> gso
1079 Add an alternative MAC address to a port::
1081 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX)
1086 Remove a MAC address from a port::
1088 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX)
1090 mac_addr add (for VF)
1091 ~~~~~~~~~~~~~~~~~~~~~
1093 Add an alternative MAC address for a VF to a port::
1095 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX)
1100 Set the default MAC address for a port::
1102 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX)
1104 mac_addr set (for VF)
1105 ~~~~~~~~~~~~~~~~~~~~~
1107 Set the MAC address for a VF from the PF::
1109 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX)
1114 Set the forwarding peer address for certain port::
1116 testpmd> set eth-peer (port_id) (perr_addr)
1118 This is equivalent to the ``--eth-peer`` command-line option.
1123 Set the unicast hash filter(s) on/off for a port::
1125 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
1130 Set the promiscuous mode on for a port or for all ports.
1131 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1133 testpmd> set promisc (port_id|all) (on|off)
1138 Set the allmulti mode for a port or for all ports::
1140 testpmd> set allmulti (port_id|all) (on|off)
1142 Same as the ifconfig (8) option. Controls how multicast packets are handled.
1144 set promisc (for VF)
1145 ~~~~~~~~~~~~~~~~~~~~
1147 Set the unicast promiscuous mode for a VF from PF.
1148 It's supported by Intel i40e NICs now.
1149 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1151 testpmd> set vf promisc (port_id) (vf_id) (on|off)
1153 set allmulticast (for VF)
1154 ~~~~~~~~~~~~~~~~~~~~~~~~~
1156 Set the multicast promiscuous mode for a VF from PF.
1157 It's supported by Intel i40e NICs now.
1158 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1160 testpmd> set vf allmulti (port_id) (vf_id) (on|off)
1162 set tx max bandwidth (for VF)
1163 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1165 Set TX max absolute bandwidth (Mbps) for a VF from PF::
1167 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth)
1169 set tc tx min bandwidth (for VF)
1170 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1172 Set all TCs' TX min relative bandwidth (%) for a VF from PF::
1174 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...)
1176 set tc tx max bandwidth (for VF)
1177 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1179 Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF::
1181 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth)
1183 set tc strict link priority mode
1184 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1186 Set some TCs' strict link priority mode on a physical port::
1188 testpmd> set tx strict-link-priority (port_id) (tc_bitmap)
1190 set tc tx min bandwidth
1191 ~~~~~~~~~~~~~~~~~~~~~~~
1193 Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs::
1195 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...)
1200 Set the link flow control parameter on a port::
1202 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1203 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
1204 autoneg (on|off) (port_id)
1208 * ``high_water`` (integer): High threshold value to trigger XOFF.
1210 * ``low_water`` (integer): Low threshold value to trigger XON.
1212 * ``pause_time`` (integer): Pause quota in the Pause frame.
1214 * ``send_xon`` (0/1): Send XON frame.
1216 * ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
1218 * ``autoneg``: Change the auto-negotiation parameter.
1223 Set the priority flow control parameter on a port::
1225 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1226 (pause_time) (priority) (port_id)
1230 * ``high_water`` (integer): High threshold value.
1232 * ``low_water`` (integer): Low threshold value.
1234 * ``pause_time`` (integer): Pause quota in the Pause frame.
1236 * ``priority`` (0-7): VLAN User Priority.
1241 Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
1243 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
1245 For example, to set rx queue 2 on port 0 to mapping 5::
1247 testpmd>set stat_qmap rx 0 2 5
1249 set xstats-hide-zero
1250 ~~~~~~~~~~~~~~~~~~~~
1252 Set the option to hide zero values for xstats display::
1254 testpmd> set xstats-hide-zero on|off
1258 By default, the zero values are displayed for xstats.
1260 set port - rx/tx (for VF)
1261 ~~~~~~~~~~~~~~~~~~~~~~~~~
1263 Set VF receive/transmit from a port::
1265 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
1267 set port - mac address filter (for VF)
1268 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1270 Add/Remove unicast or multicast MAC addr filter for a VF::
1272 testpmd> set port (port_id) vf (vf_id) (mac_addr) \
1273 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
1275 set port - rx mode(for VF)
1276 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1278 Set the VF receive mode of a port::
1280 testpmd> set port (port_id) vf (vf_id) \
1281 rxmode (AUPE|ROPE|BAM|MPE) (on|off)
1283 The available receive modes are:
1285 * ``AUPE``: Accepts untagged VLAN.
1287 * ``ROPE``: Accepts unicast hash.
1289 * ``BAM``: Accepts broadcast packets.
1291 * ``MPE``: Accepts all multicast packets.
1293 set port - tx_rate (for Queue)
1294 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1296 Set TX rate limitation for a queue on a port::
1298 testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1300 set port - tx_rate (for VF)
1301 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
1303 Set TX rate limitation for queues in VF on a port::
1305 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1307 set port - mirror rule
1308 ~~~~~~~~~~~~~~~~~~~~~~
1310 Set pool or vlan type mirror rule for a port::
1312 testpmd> set port (port_id) mirror-rule (rule_id) \
1313 (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1314 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1316 Set link mirror rule for a port::
1318 testpmd> set port (port_id) mirror-rule (rule_id) \
1319 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1321 For example to enable mirror traffic with vlan 0,1 to pool 0::
1323 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1325 reset port - mirror rule
1326 ~~~~~~~~~~~~~~~~~~~~~~~~
1328 Reset a mirror rule for a port::
1330 testpmd> reset port (port_id) mirror-rule (rule_id)
1335 Set the flush on RX streams before forwarding.
1336 The default is flush ``on``.
1337 Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1339 testpmd> set flush_rx off
1344 Set the bypass mode for the lowest port on bypass enabled NIC::
1346 testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1351 Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1353 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1354 mode (normal|bypass|isolate) (port_id)
1358 * ``timeout``: Enable bypass after watchdog timeout.
1360 * ``os_on``: Enable bypass when OS/board is powered on.
1362 * ``os_off``: Enable bypass when OS/board is powered off.
1364 * ``power_on``: Enable bypass when power supply is turned on.
1366 * ``power_off``: Enable bypass when power supply is turned off.
1372 Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1374 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1379 Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1381 testpmd> show bypass config (port_id)
1386 Set link up for a port::
1388 testpmd> set link-up port (port id)
1393 Set link down for a port::
1395 testpmd> set link-down port (port id)
1400 Enable E-tag insertion for a VF on a port::
1402 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1404 Disable E-tag insertion for a VF on a port::
1406 testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1408 Enable/disable E-tag stripping on a port::
1410 testpmd> E-tag set stripping (on|off) port (port_id)
1412 Enable/disable E-tag based forwarding on a port::
1414 testpmd> E-tag set forwarding (on|off) port (port_id)
1416 Add an E-tag forwarding filter on a port::
1418 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1420 Delete an E-tag forwarding filter on a port::
1421 testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1426 Load a dynamic device personalization (DDP) profile and store backup profile::
1428 testpmd> ddp add (port_id) (profile_path[,backup_profile_path])
1433 Delete a dynamic device personalization profile and restore backup profile::
1435 testpmd> ddp del (port_id) (backup_profile_path)
1440 List all items from the ptype mapping table::
1442 testpmd> ptype mapping get (port_id) (valid_only)
1446 * ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0).
1448 Replace a specific or a group of software defined ptype with a new one::
1450 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type)
1454 * ``target``: A specific software ptype or a mask to represent a group of software ptypes.
1456 * ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1).
1458 * ``pkt_type``: The new software ptype to replace the old ones.
1460 Update hardware defined ptype to software defined packet type mapping table::
1462 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype)
1466 * ``hw_ptype``: hardware ptype as the index of the ptype mapping table.
1468 * ``sw_ptype``: software ptype as the value of the ptype mapping table.
1470 Reset ptype mapping table::
1472 testpmd> ptype mapping reset (port_id)
1477 The following sections show functions for configuring ports.
1481 Port configuration changes only become active when forwarding is started/restarted.
1486 Attach a port specified by pci address or virtual device args::
1488 testpmd> port attach (identifier)
1490 To attach a new pci device, the device should be recognized by kernel first.
1491 Then it should be moved under DPDK management.
1492 Finally the port can be attached to testpmd.
1494 For example, to move a pci device using ixgbe under DPDK management:
1496 .. code-block:: console
1498 # Check the status of the available devices.
1499 ./usertools/dpdk-devbind.py --status
1501 Network devices using DPDK-compatible driver
1502 ============================================
1505 Network devices using kernel driver
1506 ===================================
1507 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1510 # Bind the device to igb_uio.
1511 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1514 # Recheck the status of the devices.
1515 ./usertools/dpdk-devbind.py --status
1516 Network devices using DPDK-compatible driver
1517 ============================================
1518 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1520 To attach a port created by virtual device, above steps are not needed.
1522 For example, to attach a port whose pci address is 0000:0a:00.0.
1524 .. code-block:: console
1526 testpmd> port attach 0000:0a:00.0
1527 Attaching a new port...
1528 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1529 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1530 EAL: PCI memory mapped at 0x7f83bfa00000
1531 EAL: PCI memory mapped at 0x7f83bfa80000
1532 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1533 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1534 Port 0 is attached. Now total ports is 1
1537 For example, to attach a port created by pcap PMD.
1539 .. code-block:: console
1541 testpmd> port attach net_pcap0
1542 Attaching a new port...
1543 PMD: Initializing pmd_pcap for net_pcap0
1544 PMD: Creating pcap-backed ethdev on numa socket 0
1545 Port 0 is attached. Now total ports is 1
1548 In this case, identifier is ``net_pcap0``.
1549 This identifier format is the same as ``--vdev`` format of DPDK applications.
1551 For example, to re-attach a bonded port which has been previously detached,
1552 the mode and slave parameters must be given.
1554 .. code-block:: console
1556 testpmd> port attach net_bond_0,mode=0,slave=1
1557 Attaching a new port...
1558 EAL: Initializing pmd_bond for net_bond_0
1559 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1560 Port 0 is attached. Now total ports is 1
1567 Detach a specific port::
1569 testpmd> port detach (port_id)
1571 Before detaching a port, the port should be stopped and closed.
1573 For example, to detach a pci device port 0.
1575 .. code-block:: console
1577 testpmd> port stop 0
1580 testpmd> port close 0
1584 testpmd> port detach 0
1586 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1587 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1588 EAL: PCI memory unmapped at 0x7f83bfa00000
1589 EAL: PCI memory unmapped at 0x7f83bfa80000
1593 For example, to detach a virtual device port 0.
1595 .. code-block:: console
1597 testpmd> port stop 0
1600 testpmd> port close 0
1604 testpmd> port detach 0
1606 PMD: Closing pcap ethdev on numa socket 0
1607 Port 'net_pcap0' is detached. Now total ports is 0
1610 To remove a pci device completely from the system, first detach the port from testpmd.
1611 Then the device should be moved under kernel management.
1612 Finally the device can be removed using kernel pci hotplug functionality.
1614 For example, to move a pci device under kernel management:
1616 .. code-block:: console
1618 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1620 ./usertools/dpdk-devbind.py --status
1622 Network devices using DPDK-compatible driver
1623 ============================================
1626 Network devices using kernel driver
1627 ===================================
1628 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1630 To remove a port created by a virtual device, above steps are not needed.
1635 Start all ports or a specific port::
1637 testpmd> port start (port_id|all)
1642 Stop all ports or a specific port::
1644 testpmd> port stop (port_id|all)
1649 Close all ports or a specific port::
1651 testpmd> port close (port_id|all)
1653 port start/stop queue
1654 ~~~~~~~~~~~~~~~~~~~~~
1656 Start/stop a rx/tx queue on a specific port::
1658 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1660 Only take effect when port is started.
1665 Set the speed and duplex mode for all ports or a specific port::
1667 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1668 duplex (half|full|auto)
1670 port config - queues/descriptors
1671 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1673 Set number of queues/descriptors for rxq, txq, rxd and txd::
1675 testpmd> port config all (rxq|txq|rxd|txd) (value)
1677 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1679 port config - max-pkt-len
1680 ~~~~~~~~~~~~~~~~~~~~~~~~~
1682 Set the maximum packet length::
1684 testpmd> port config all max-pkt-len (value)
1686 This is equivalent to the ``--max-pkt-len`` command-line option.
1688 port config - CRC Strip
1689 ~~~~~~~~~~~~~~~~~~~~~~~
1691 Set hardware CRC stripping on or off for all ports::
1693 testpmd> port config all crc-strip (on|off)
1695 CRC stripping is on by default.
1697 The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option.
1699 port config - scatter
1700 ~~~~~~~~~~~~~~~~~~~~~~~
1702 Set RX scatter mode on or off for all ports::
1704 testpmd> port config all scatter (on|off)
1706 RX scatter mode is off by default.
1708 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1710 port config - RX Checksum
1711 ~~~~~~~~~~~~~~~~~~~~~~~~~
1713 Set hardware RX checksum offload to on or off for all ports::
1715 testpmd> port config all rx-cksum (on|off)
1717 Checksum offload is off by default.
1719 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1724 Set hardware VLAN on or off for all ports::
1726 testpmd> port config all hw-vlan (on|off)
1728 Hardware VLAN is off by default.
1730 The ``on`` option is equivalent to the ``--enable-hw-vlan`` command-line option.
1732 port config - VLAN filter
1733 ~~~~~~~~~~~~~~~~~~~~~~~~~
1735 Set hardware VLAN filter on or off for all ports::
1737 testpmd> port config all hw-vlan-filter (on|off)
1739 Hardware VLAN filter is off by default.
1741 The ``on`` option is equivalent to the ``--enable-hw-vlan-filter`` command-line option.
1743 port config - VLAN strip
1744 ~~~~~~~~~~~~~~~~~~~~~~~~
1746 Set hardware VLAN strip on or off for all ports::
1748 testpmd> port config all hw-vlan-strip (on|off)
1750 Hardware VLAN strip is off by default.
1752 The ``on`` option is equivalent to the ``--enable-hw-vlan-strip`` command-line option.
1754 port config - VLAN extend
1755 ~~~~~~~~~~~~~~~~~~~~~~~~~
1757 Set hardware VLAN extend on or off for all ports::
1759 testpmd> port config all hw-vlan-extend (on|off)
1761 Hardware VLAN extend is off by default.
1763 The ``on`` option is equivalent to the ``--enable-hw-vlan-extend`` command-line option.
1765 port config - Drop Packets
1766 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1768 Set packet drop for packets with no descriptors on or off for all ports::
1770 testpmd> port config all drop-en (on|off)
1772 Packet dropping for packets with no descriptors is off by default.
1774 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1779 Set the RSS (Receive Side Scaling) mode on or off::
1781 testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1783 RSS is on by default.
1785 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1787 port config - RSS Reta
1788 ~~~~~~~~~~~~~~~~~~~~~~
1790 Set the RSS (Receive Side Scaling) redirection table::
1792 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1797 Set the DCB mode for an individual port::
1799 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1801 The traffic class should be 4 or 8.
1806 Set the number of packets per burst::
1808 testpmd> port config all burst (value)
1810 This is equivalent to the ``--burst`` command-line option.
1812 port config - Threshold
1813 ~~~~~~~~~~~~~~~~~~~~~~~
1815 Set thresholds for TX/RX queues::
1817 testpmd> port config all (threshold) (value)
1819 Where the threshold type can be:
1821 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1823 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1825 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1827 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1829 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1831 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1833 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1835 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1837 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1839 These threshold options are also available from the command-line.
1844 Set the value of ether-type for E-tag::
1846 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1848 Enable/disable the E-tag support::
1850 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1852 port config pctype mapping
1853 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1855 Reset pctype mapping table::
1857 testpmd> port config (port_id) pctype mapping reset
1859 Update hardware defined pctype to software defined flow type mapping table::
1861 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id)
1865 * ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table.
1867 * ``flow_type_id``: software flow type id as the index of the pctype mapping table.
1869 port config input set
1870 ~~~~~~~~~~~~~~~~~~~~~
1872 Config RSS/FDIR/FDIR flexible payload input set for some pctype::
1873 testpmd> port config (port_id) pctype (pctype_id) \
1874 (hash_inset|fdir_inset|fdir_flx_inset) \
1875 (get|set|clear) field (field_idx)
1877 Clear RSS/FDIR/FDIR flexible payload input set for some pctype::
1878 testpmd> port config (port_id) pctype (pctype_id) \
1879 (hash_inset|fdir_inset|fdir_flx_inset) clear all
1883 * ``pctype_id``: hardware packet classification types.
1884 * ``field_idx``: hardware field index.
1886 Link Bonding Functions
1887 ----------------------
1889 The Link Bonding functions make it possible to dynamically create and
1890 manage link bonding devices from within testpmd interactive prompt.
1892 create bonded device
1893 ~~~~~~~~~~~~~~~~~~~~
1895 Create a new bonding device::
1897 testpmd> create bonded device (mode) (socket)
1899 For example, to create a bonded device in mode 1 on socket 0::
1901 testpmd> create bonded 1 0
1902 created new bonded device (port X)
1907 Adds Ethernet device to a Link Bonding device::
1909 testpmd> add bonding slave (slave id) (port id)
1911 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
1913 testpmd> add bonding slave 6 10
1916 remove bonding slave
1917 ~~~~~~~~~~~~~~~~~~~~
1919 Removes an Ethernet slave device from a Link Bonding device::
1921 testpmd> remove bonding slave (slave id) (port id)
1923 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
1925 testpmd> remove bonding slave 6 10
1930 Set the Link Bonding mode of a Link Bonding device::
1932 testpmd> set bonding mode (value) (port id)
1934 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
1936 testpmd> set bonding mode 3 10
1941 Set an Ethernet slave device as the primary device on a Link Bonding device::
1943 testpmd> set bonding primary (slave id) (port id)
1945 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
1947 testpmd> set bonding primary 6 10
1952 Set the MAC address of a Link Bonding device::
1954 testpmd> set bonding mac (port id) (mac)
1956 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
1958 testpmd> set bonding mac 10 00:00:00:00:00:01
1960 set bonding xmit_balance_policy
1961 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1963 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
1965 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
1967 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
1969 testpmd> set bonding xmit_balance_policy 10 l34
1972 set bonding mon_period
1973 ~~~~~~~~~~~~~~~~~~~~~~
1975 Set the link status monitoring polling period in milliseconds for a bonding device.
1977 This adds support for PMD slave devices which do not support link status interrupts.
1978 When the mon_period is set to a value greater than 0 then all PMD's which do not support
1979 link status ISR will be queried every polling interval to check if their link status has changed::
1981 testpmd> set bonding mon_period (port_id) (value)
1983 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
1985 testpmd> set bonding mon_period 5 150
1988 set bonding lacp dedicated_queue
1989 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1991 Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic
1992 when in mode 4 (link-aggregration-802.3ad)::
1994 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable)
1997 set bonding agg_mode
1998 ~~~~~~~~~~~~~~~~~~~~
2000 Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad)::
2002 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable)
2008 Show the current configuration of a Link Bonding device::
2010 testpmd> show bonding config (port id)
2013 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
2014 in balance mode with a transmission policy of layer 2+3::
2016 testpmd> show bonding config 9
2018 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
2020 Active Slaves (3): [1 3 4]
2027 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
2028 This is mainly useful for debugging purposes.
2029 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
2030 and fields that can be accessed.
2035 Display the value of a port register::
2037 testpmd> read reg (port_id) (address)
2039 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
2041 testpmd> read reg 0 0xEE00
2042 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
2047 Display a port register bit field::
2049 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
2051 For example, reading the lowest two bits from the register in the example above::
2053 testpmd> read regfield 0 0xEE00 0 1
2054 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
2059 Display a single port register bit::
2061 testpmd> read regbit (port_id) (address) (bit_x)
2063 For example, reading the lowest bit from the register in the example above::
2065 testpmd> read regbit 0 0xEE00 0
2066 port 0 PCI register at offset 0xEE00: bit 0=1
2071 Set the value of a port register::
2073 testpmd> write reg (port_id) (address) (value)
2075 For example, to clear a register::
2077 testpmd> write reg 0 0xEE00 0x0
2078 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
2083 Set bit field of a port register::
2085 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
2087 For example, writing to the register cleared in the example above::
2089 testpmd> write regfield 0 0xEE00 0 1 2
2090 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
2095 Set single bit value of a port register::
2097 testpmd> write regbit (port_id) (address) (bit_x) (value)
2099 For example, to set the high bit in the register from the example above::
2101 testpmd> write regbit 0 0xEE00 31 1
2102 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
2104 Traffic Metering and Policing
2105 -----------------------------
2107 The following section shows functions for configuring traffic metering and
2108 policing on the ethernet device through the use of generic ethdev API.
2110 show port traffic management capability
2111 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2113 Show traffic metering and policing capability of the port::
2115 testpmd> show port meter cap (port_id)
2117 add port meter profile (srTCM rfc2967)
2118 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2120 Add meter profile (srTCM rfc2697) to the ethernet device::
2122 testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \
2127 * ``profile_id``: ID for the meter profile.
2128 * ``cir``: Committed Information Rate (CIR) (bytes/second).
2129 * ``cbs``: Committed Burst Size (CBS) (bytes).
2130 * ``ebs``: Excess Burst Size (EBS) (bytes).
2132 add port meter profile (trTCM rfc2968)
2133 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2135 Add meter profile (srTCM rfc2698) to the ethernet device::
2137 testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \
2138 (cir) (pir) (cbs) (pbs)
2142 * ``profile_id``: ID for the meter profile.
2143 * ``cir``: Committed information rate (bytes/second).
2144 * ``pir``: Peak information rate (bytes/second).
2145 * ``cbs``: Committed burst size (bytes).
2146 * ``pbs``: Peak burst size (bytes).
2148 add port meter profile (trTCM rfc4115)
2149 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2151 Add meter profile (trTCM rfc4115) to the ethernet device::
2153 testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \
2154 (cir) (eir) (cbs) (ebs)
2158 * ``profile_id``: ID for the meter profile.
2159 * ``cir``: Committed information rate (bytes/second).
2160 * ``eir``: Excess information rate (bytes/second).
2161 * ``cbs``: Committed burst size (bytes).
2162 * ``ebs``: Excess burst size (bytes).
2164 delete port meter profile
2165 ~~~~~~~~~~~~~~~~~~~~~~~~~
2167 Delete meter profile from the ethernet device::
2169 testpmd> del port meter profile (port_id) (profile_id)
2174 Create new meter object for the ethernet device::
2176 testpmd> create port meter (port_id) (mtr_id) (profile_id) \
2177 (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \
2178 (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\
2183 * ``mtr_id``: meter object ID.
2184 * ``profile_id``: ID for the meter profile.
2185 * ``meter_enable``: When this parameter has a non-zero value, the meter object
2186 gets enabled at the time of creation, otherwise remains disabled.
2187 * ``g_action``: Policer action for the packet with green color.
2188 * ``y_action``: Policer action for the packet with yellow color.
2189 * ``r_action``: Policer action for the packet with red color.
2190 * ``stats_mask``: Mask of statistics counter types to be enabled for the
2192 * ``shared``: When this parameter has a non-zero value, the meter object is
2193 shared by multiple flows. Otherwise, meter object is used by single flow.
2194 * ``use_pre_meter_color``: When this parameter has a non-zero value, the
2195 input color for the current meter object is determined by the latest meter
2196 object in the same flow. Otherwise, the current meter object uses the
2197 *dscp_table* to determine the input color.
2198 * ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input
2199 color, 0 <= x <= 63.
2204 Enable meter for the ethernet device::
2206 testpmd> enable port meter (port_id) (mtr_id)
2211 Disable meter for the ethernet device::
2213 testpmd> disable port meter (port_id) (mtr_id)
2218 Delete meter for the ethernet device::
2220 testpmd> del port meter (port_id) (mtr_id)
2222 Set port meter profile
2223 ~~~~~~~~~~~~~~~~~~~~~~
2225 Set meter profile for the ethernet device::
2227 testpmd> set port meter profile (port_id) (mtr_id) (profile_id)
2229 set port meter dscp table
2230 ~~~~~~~~~~~~~~~~~~~~~~~~~
2232 Set meter dscp table for the ethernet device::
2234 testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \
2235 (dscp_tbl_entry1)...(dscp_tbl_entry63)]
2237 set port meter policer action
2238 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2240 Set meter policer action for the ethernet device::
2242 testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \
2243 (action0) [(action1) (action1)]
2247 * ``action_mask``: Bit mask indicating which policer actions need to be
2248 updated. One or more policer actions can be updated in a single function
2249 invocation. To update the policer action associated with color C, bit
2250 (1 << C) needs to be set in *action_mask* and element at position C
2251 in the *actions* array needs to be valid.
2252 * ``actionx``: Policer action for the color x,
2253 RTE_MTR_GREEN <= x < RTE_MTR_COLORS
2255 set port meter stats mask
2256 ~~~~~~~~~~~~~~~~~~~~~~~~~
2258 Set meter stats mask for the ethernet device::
2260 testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask)
2264 * ``stats_mask``: Bit mask indicating statistics counter types to be enabled.
2266 show port meter stats
2267 ~~~~~~~~~~~~~~~~~~~~~
2269 Show meter stats of the ethernet device::
2271 testpmd> show port meter stats (port_id) (mtr_id) (clear)
2275 * ``clear``: Flag that indicates whether the statistics counters should
2276 be cleared (i.e. set to zero) immediately after they have been read or not.
2281 The following section shows functions for configuring traffic management on
2282 on the ethernet device through the use of generic TM API.
2284 show port traffic management capability
2285 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2287 Show traffic management capability of the port::
2289 testpmd> show port tm cap (port_id)
2291 show port traffic management capability (hierarchy level)
2292 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2294 Show traffic management hierarchy level capability of the port::
2296 testpmd> show port tm level cap (port_id) (level_id)
2298 show port traffic management capability (hierarchy node level)
2299 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2301 Show the traffic management hierarchy node capability of the port::
2303 testpmd> show port tm node cap (port_id) (node_id)
2305 show port traffic management hierarchy node type
2306 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2308 Show the port traffic management hierarchy node type::
2310 testpmd> show port tm node type (port_id) (node_id)
2312 show port traffic management hierarchy node stats
2313 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2315 Show the port traffic management hierarchy node statistics::
2317 testpmd> show port tm node stats (port_id) (node_id) (clear)
2321 * ``clear``: When this parameter has a non-zero value, the statistics counters
2322 are cleared (i.e. set to zero) immediately after they have been read,
2323 otherwise the statistics counters are left untouched.
2325 Add port traffic management private shaper profile
2326 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2328 Add the port traffic management private shaper profile::
2330 testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \
2331 (tb_rate) (tb_size) (packet_length_adjust)
2335 * ``shaper_profile id``: Shaper profile ID for the new profile.
2336 * ``tb_rate``: Token bucket rate (bytes per second).
2337 * ``tb_size``: Token bucket size (bytes).
2338 * ``packet_length_adjust``: The value (bytes) to be added to the length of
2339 each packet for the purpose of shaping. This parameter value can be used to
2340 correct the packet length with the framing overhead bytes that are consumed
2343 Delete port traffic management private shaper profile
2344 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2346 Delete the port traffic management private shaper::
2348 testpmd> del port tm node shaper profile (port_id) (shaper_profile_id)
2352 * ``shaper_profile id``: Shaper profile ID that needs to be deleted.
2354 Add port traffic management shared shaper
2355 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2357 Create the port traffic management shared shaper::
2359 testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \
2364 * ``shared_shaper_id``: Shared shaper ID to be created.
2365 * ``shaper_profile id``: Shaper profile ID for shared shaper.
2367 Set port traffic management shared shaper
2368 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2370 Update the port traffic management shared shaper::
2372 testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \
2377 * ``shared_shaper_id``: Shared shaper ID to be update.
2378 * ``shaper_profile id``: Shaper profile ID for shared shaper.
2380 Delete port traffic management shared shaper
2381 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2383 Delete the port traffic management shared shaper::
2385 testpmd> del port tm node shared shaper (port_id) (shared_shaper_id)
2389 * ``shared_shaper_id``: Shared shaper ID to be deleted.
2391 Set port traffic management hiearchy node private shaper
2392 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2394 set the port traffic management hierarchy node private shaper::
2396 testpmd> set port tm node shaper profile (port_id) (node_id) \
2401 * ``shaper_profile id``: Private shaper profile ID to be enabled on the
2404 Add port traffic management WRED profile
2405 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2407 Create a new WRED profile::
2409 testpmd> add port tm node wred profile (port_id) (wred_profile_id) \
2410 (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \
2411 (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \
2412 (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r)
2416 * ``wred_profile id``: Identifier for the newly create WRED profile
2417 * ``color_g``: Packet color (green)
2418 * ``min_th_g``: Minimum queue threshold for packet with green color
2419 * ``max_th_g``: Minimum queue threshold for packet with green color
2420 * ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp)
2421 * ``wq_log2_g``: Negated log2 of queue weight (wq)
2422 * ``color_y``: Packet color (yellow)
2423 * ``min_th_y``: Minimum queue threshold for packet with yellow color
2424 * ``max_th_y``: Minimum queue threshold for packet with yellow color
2425 * ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp)
2426 * ``wq_log2_y``: Negated log2 of queue weight (wq)
2427 * ``color_r``: Packet color (red)
2428 * ``min_th_r``: Minimum queue threshold for packet with yellow color
2429 * ``max_th_r``: Minimum queue threshold for packet with yellow color
2430 * ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp)
2431 * ``wq_log2_r``: Negated log2 of queue weight (wq)
2433 Delete port traffic management WRED profile
2434 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2436 Delete the WRED profile::
2438 testpmd> del port tm node wred profile (port_id) (wred_profile_id)
2440 Add port traffic management hierarchy nonleaf node
2441 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2443 Add nonleaf node to port traffic management hiearchy::
2445 testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \
2446 (priority) (weight) (level_id) (shaper_profile_id) \
2447 (n_sp_priorities) (stats_mask) (n_shared_shapers) \
2448 [(shared_shaper_0) (shared_shaper_1) ...] \
2452 * ``parent_node_id``: Node ID of the parent.
2453 * ``priority``: Node priority (highest node priority is zero). This is used by
2454 the SP algorithm running on the parent node for scheduling this node.
2455 * ``weight``: Node weight (lowest weight is one). The node weight is relative
2456 to the weight sum of all siblings that have the same priority. It is used by
2457 the WFQ algorithm running on the parent node for scheduling this node.
2458 * ``level_id``: Hiearchy level of the node.
2459 * ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2461 * ``n_sp_priorities``: Number of strict priorities.
2462 * ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2463 * ``n_shared_shapers``: Number of shared shapers.
2464 * ``shared_shaper_id``: Shared shaper id.
2466 Add port traffic management hierarchy leaf node
2467 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2469 Add leaf node to port traffic management hiearchy::
2471 testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \
2472 (priority) (weight) (level_id) (shaper_profile_id) \
2473 (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \
2474 [(shared_shaper_id) (shared_shaper_id) ...] \
2478 * ``parent_node_id``: Node ID of the parent.
2479 * ``priority``: Node priority (highest node priority is zero). This is used by
2480 the SP algorithm running on the parent node for scheduling this node.
2481 * ``weight``: Node weight (lowest weight is one). The node weight is relative
2482 to the weight sum of all siblings that have the same priority. It is used by
2483 the WFQ algorithm running on the parent node for scheduling this node.
2484 * ``level_id``: Hiearchy level of the node.
2485 * ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2487 * ``cman_mode``: Congestion management mode to be enabled for this node.
2488 * ``wred_profile_id``: WRED profile id to be enabled for this node.
2489 * ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2490 * ``n_shared_shapers``: Number of shared shapers.
2491 * ``shared_shaper_id``: Shared shaper id.
2493 Delete port traffic management hierarchy node
2494 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2496 Delete node from port traffic management hiearchy::
2498 testpmd> del port tm node (port_id) (node_id)
2500 Update port traffic management hierarchy parent node
2501 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2503 Update port traffic management hierarchy parent node::
2505 testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \
2508 This function can only be called after the hierarchy commit invocation. Its
2509 success depends on the port support for this operation, as advertised through
2510 the port capability set. This function is valid for all nodes of the traffic
2511 management hierarchy except root node.
2513 Commit port traffic management hierarchy
2514 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2516 Commit the traffic management hierarchy on the port::
2518 testpmd> port tm hierarchy commit (port_id) (clean_on_fail)
2522 * ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function
2523 call failure. On the other hand, hierarchy is preserved when this parameter
2526 Set port traffic management default hierarchy (tm forwarding mode)
2527 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2529 set the traffic management default hierarchy on the port::
2531 testpmd> set port tm hierarchy default (port_id)
2536 This section details the available filter functions that are available.
2538 Note these functions interface the deprecated legacy filtering framework,
2539 superseded by *rte_flow*. See `Flow rules management`_.
2542 ~~~~~~~~~~~~~~~~~~~~
2544 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
2546 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
2547 ethertype (ether_type) (drop|fwd) queue (queue_id)
2549 The available information parameters are:
2551 * ``port_id``: The port which the Ethertype filter assigned on.
2553 * ``mac_addr``: Compare destination mac address.
2555 * ``mac_ignr``: Ignore destination mac address match.
2557 * ``mac_address``: Destination mac address to match.
2559 * ``ether_type``: The EtherType value want to match,
2560 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
2562 * ``queue_id``: The receive queue associated with this EtherType filter.
2563 It is meaningless when deleting or dropping.
2565 Example, to add/remove an ethertype filter rule::
2567 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
2568 ethertype 0x0806 fwd queue 3
2570 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
2571 ethertype 0x0806 fwd queue 3
2576 Add or delete a 2-tuple filter,
2577 which identifies packets by specific protocol and destination TCP/UDP port
2578 and forwards packets into one of the receive queues::
2580 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
2581 protocol (protocol_value) mask (mask_value) \
2582 tcp_flags (tcp_flags_value) priority (prio_value) \
2585 The available information parameters are:
2587 * ``port_id``: The port which the 2-tuple filter assigned on.
2589 * ``dst_port_value``: Destination port in L4.
2591 * ``protocol_value``: IP L4 protocol.
2593 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
2595 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
2597 * ``prio_value``: Priority of this filter.
2599 * ``queue_id``: The receive queue associated with this 2-tuple filter.
2601 Example, to add/remove an 2tuple filter rule::
2603 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
2604 tcp_flags 0x02 priority 3 queue 3
2606 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
2607 tcp_flags 0x02 priority 3 queue 3
2612 Add or delete a 5-tuple filter,
2613 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
2614 and routes packets into one of the receive queues::
2616 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
2617 (src_address) dst_port (dst_port_value) \
2618 src_port (src_port_value) protocol (protocol_value) \
2619 mask (mask_value) tcp_flags (tcp_flags_value) \
2620 priority (prio_value) queue (queue_id)
2622 The available information parameters are:
2624 * ``port_id``: The port which the 5-tuple filter assigned on.
2626 * ``dst_address``: Destination IP address.
2628 * ``src_address``: Source IP address.
2630 * ``dst_port_value``: TCP/UDP destination port.
2632 * ``src_port_value``: TCP/UDP source port.
2634 * ``protocol_value``: L4 protocol.
2636 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
2638 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
2640 * ``prio_value``: The priority of this filter.
2642 * ``queue_id``: The receive queue associated with this 5-tuple filter.
2644 Example, to add/remove an 5tuple filter rule::
2646 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2647 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2648 flags 0x0 priority 3 queue 3
2650 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2651 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2652 flags 0x0 priority 3 queue 3
2657 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
2659 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
2661 The available information parameters are:
2663 * ``port_id``: The port which the SYN filter assigned on.
2665 * ``high``: This SYN filter has higher priority than other filters.
2667 * ``low``: This SYN filter has lower priority than other filters.
2669 * ``queue_id``: The receive queue associated with this SYN filter
2673 testpmd> syn_filter 0 add priority high queue 3
2678 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
2679 and routed into one of the receive queues::
2681 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
2682 mask (mask_value) priority (prio_value) queue (queue_id)
2684 The available information parameters are:
2686 * ``port_id``: The port which the Flex filter is assigned on.
2688 * ``len_value``: Filter length in bytes, no greater than 128.
2690 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
2692 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
2694 * ``prio_value``: The priority of this filter.
2696 * ``queue_id``: The receive queue associated with this Flex filter.
2700 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
2701 mask 000C priority 3 queue 3
2703 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
2704 mask 000C priority 3 queue 3
2707 .. _testpmd_flow_director:
2709 flow_director_filter
2710 ~~~~~~~~~~~~~~~~~~~~
2712 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
2714 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
2715 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
2717 * Perfect match filters.
2718 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2719 The masked fields are for IP flow.
2721 * Signature filters.
2722 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
2724 * Perfect-mac-vlan match filters.
2725 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2726 The masked fields are for MAC VLAN flow.
2728 * Perfect-tunnel match filters.
2729 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2730 The masked fields are for tunnel flow.
2732 * Perfect-raw-flow-type match filters.
2733 The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet.
2734 The masked fields are specified by input sets.
2736 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
2737 per flow type and the flexible payload.
2739 The Flow Director can also mask out parts of all of these fields so that filters
2740 are only applied to certain fields or parts of the fields.
2742 Note that for raw flow type mode the source and destination fields in the
2743 raw packet buffer need to be presented in a reversed order with respect
2744 to the expected received packets.
2745 For example: IP source and destination addresses or TCP/UDP/SCTP
2746 source and destination ports
2748 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
2750 # Commands to add flow director filters of different flow types::
2752 flow_director_filter (port_id) mode IP (add|del|update) \
2753 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
2754 src (src_ip_address) dst (dst_ip_address) \
2755 tos (tos_value) proto (proto_value) ttl (ttl_value) \
2756 vlan (vlan_value) flexbytes (flexbytes_value) \
2757 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
2760 flow_director_filter (port_id) mode IP (add|del|update) \
2761 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
2762 src (src_ip_address) (src_port) \
2763 dst (dst_ip_address) (dst_port) \
2764 tos (tos_value) ttl (ttl_value) \
2765 vlan (vlan_value) flexbytes (flexbytes_value) \
2766 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
2769 flow_director_filter (port_id) mode IP (add|del|update) \
2770 flow (ipv4-sctp|ipv6-sctp) \
2771 src (src_ip_address) (src_port) \
2772 dst (dst_ip_address) (dst_port) \
2773 tos (tos_value) ttl (ttl_value) \
2774 tag (verification_tag) vlan (vlan_value) \
2775 flexbytes (flexbytes_value) (drop|fwd) \
2776 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
2778 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
2779 ether (ethertype) flexbytes (flexbytes_value) \
2780 (drop|fwd) pf|vf(vf_id) queue (queue_id)
2783 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
2784 mac (mac_address) vlan (vlan_value) \
2785 flexbytes (flexbytes_value) (drop|fwd) \
2786 queue (queue_id) fd_id (fd_id_value)
2788 flow_director_filter (port_id) mode Tunnel (add|del|update) \
2789 mac (mac_address) vlan (vlan_value) \
2790 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
2791 flexbytes (flexbytes_value) (drop|fwd) \
2792 queue (queue_id) fd_id (fd_id_value)
2794 flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \
2795 (drop|fwd) queue (queue_id) fd_id (fd_id_value) \
2796 packet (packet file name)
2798 For example, to add an ipv4-udp flow type filter::
2800 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
2801 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
2802 fwd pf queue 1 fd_id 1
2804 For example, add an ipv4-other flow type filter::
2806 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
2807 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
2808 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
2813 Flush all flow director filters on a device::
2815 testpmd> flush_flow_director (port_id)
2817 Example, to flush all flow director filter on port 0::
2819 testpmd> flush_flow_director 0
2824 Set flow director's input masks::
2826 flow_director_mask (port_id) mode IP vlan (vlan_value) \
2827 src_mask (ipv4_src) (ipv6_src) (src_port) \
2828 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2830 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2832 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2833 mac (mac_value) tunnel-type (tunnel_type_value) \
2834 tunnel-id (tunnel_id_value)
2836 Example, to set flow director mask on port 0::
2838 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2839 src_mask 255.255.255.255 \
2840 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2841 dst_mask 255.255.255.255 \
2842 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2844 flow_director_flex_mask
2845 ~~~~~~~~~~~~~~~~~~~~~~~
2847 set masks of flow director's flexible payload based on certain flow type::
2849 testpmd> flow_director_flex_mask (port_id) \
2850 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2851 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2852 l2_payload|all) (mask)
2854 Example, to set flow director's flex mask for all flow type on port 0::
2856 testpmd> flow_director_flex_mask 0 flow all \
2857 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2860 flow_director_flex_payload
2861 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2863 Configure flexible payload selection::
2865 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2867 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2869 testpmd> flow_director_flex_payload 0 l4 \
2870 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2872 get_sym_hash_ena_per_port
2873 ~~~~~~~~~~~~~~~~~~~~~~~~~
2875 Get symmetric hash enable configuration per port::
2877 get_sym_hash_ena_per_port (port_id)
2879 For example, to get symmetric hash enable configuration of port 1::
2881 testpmd> get_sym_hash_ena_per_port 1
2883 set_sym_hash_ena_per_port
2884 ~~~~~~~~~~~~~~~~~~~~~~~~~
2886 Set symmetric hash enable configuration per port to enable or disable::
2888 set_sym_hash_ena_per_port (port_id) (enable|disable)
2890 For example, to set symmetric hash enable configuration of port 1 to enable::
2892 testpmd> set_sym_hash_ena_per_port 1 enable
2894 get_hash_global_config
2895 ~~~~~~~~~~~~~~~~~~~~~~
2897 Get the global configurations of hash filters::
2899 get_hash_global_config (port_id)
2901 For example, to get the global configurations of hash filters of port 1::
2903 testpmd> get_hash_global_config 1
2905 set_hash_global_config
2906 ~~~~~~~~~~~~~~~~~~~~~~
2908 Set the global configurations of hash filters::
2910 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2911 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2912 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \
2915 For example, to enable simple_xor for flow type of ipv6 on port 2::
2917 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2922 Set the input set for hash::
2924 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2925 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2926 l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
2927 ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2928 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2929 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2930 fld-8th|none) (select|add)
2932 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2934 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2939 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2940 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2942 Set the input set for flow director::
2944 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2945 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2946 l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
2947 ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2948 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2949 sctp-dst-port|sctp-veri-tag|none) (select|add)
2951 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2953 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2958 Set different GRE key length for input set::
2960 global_config (port_id) gre-key-len (number in bytes)
2962 For example to set GRE key length for input set to 4 bytes on port 0::
2964 testpmd> global_config 0 gre-key-len 4
2967 .. _testpmd_rte_flow:
2969 Flow rules management
2970 ---------------------
2972 Control of the generic flow API (*rte_flow*) is fully exposed through the
2973 ``flow`` command (validation, creation, destruction, queries and operation
2976 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2977 features simultaneously may cause undefined side-effects and is therefore
2983 Because the ``flow`` command uses dynamic tokens to handle the large number
2984 of possible flow rules combinations, its behavior differs slightly from
2985 other commands, in particular:
2987 - Pressing *?* or the *<tab>* key displays contextual help for the current
2988 token, not that of the entire command.
2990 - Optional and repeated parameters are supported (provided they are listed
2991 in the contextual help).
2993 The first parameter stands for the operation mode. Possible operations and
2994 their general syntax are described below. They are covered in detail in the
2997 - Check whether a flow rule can be created::
2999 flow validate {port_id}
3000 [group {group_id}] [priority {level}] [ingress] [egress]
3001 pattern {item} [/ {item} [...]] / end
3002 actions {action} [/ {action} [...]] / end
3004 - Create a flow rule::
3006 flow create {port_id}
3007 [group {group_id}] [priority {level}] [ingress] [egress]
3008 pattern {item} [/ {item} [...]] / end
3009 actions {action} [/ {action} [...]] / end
3011 - Destroy specific flow rules::
3013 flow destroy {port_id} rule {rule_id} [...]
3015 - Destroy all flow rules::
3017 flow flush {port_id}
3019 - Query an existing flow rule::
3021 flow query {port_id} {rule_id} {action}
3023 - List existing flow rules sorted by priority, filtered by group
3026 flow list {port_id} [group {group_id}] [...]
3028 - Restrict ingress traffic to the defined flow rules::
3030 flow isolate {port_id} {boolean}
3032 Validating flow rules
3033 ~~~~~~~~~~~~~~~~~~~~~
3035 ``flow validate`` reports whether a flow rule would be accepted by the
3036 underlying device in its current state but stops short of creating it. It is
3037 bound to ``rte_flow_validate()``::
3039 flow validate {port_id}
3040 [group {group_id}] [priority {level}] [ingress] [egress]
3041 pattern {item} [/ {item} [...]] / end
3042 actions {action} [/ {action} [...]] / end
3044 If successful, it will show::
3048 Otherwise it will show an error message of the form::
3050 Caught error type [...] ([...]): [...]
3052 This command uses the same parameters as ``flow create``, their format is
3053 described in `Creating flow rules`_.
3055 Check whether redirecting any Ethernet packet received on port 0 to RX queue
3056 index 6 is supported::
3058 testpmd> flow validate 0 ingress pattern eth / end
3059 actions queue index 6 / end
3063 Port 0 does not support TCPv6 rules::
3065 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
3067 Caught error type 9 (specific pattern item): Invalid argument
3073 ``flow create`` validates and creates the specified flow rule. It is bound
3074 to ``rte_flow_create()``::
3076 flow create {port_id}
3077 [group {group_id}] [priority {level}] [ingress] [egress]
3078 pattern {item} [/ {item} [...]] / end
3079 actions {action} [/ {action} [...]] / end
3081 If successful, it will return a flow rule ID usable with other commands::
3083 Flow rule #[...] created
3085 Otherwise it will show an error message of the form::
3087 Caught error type [...] ([...]): [...]
3089 Parameters describe in the following order:
3091 - Attributes (*group*, *priority*, *ingress*, *egress* tokens).
3092 - A matching pattern, starting with the *pattern* token and terminated by an
3094 - Actions, starting with the *actions* token and terminated by an *end*
3097 These translate directly to *rte_flow* objects provided as-is to the
3098 underlying functions.
3100 The shortest valid definition only comprises mandatory tokens::
3102 testpmd> flow create 0 pattern end actions end
3104 Note that PMDs may refuse rules that essentially do nothing such as this
3107 **All unspecified object values are automatically initialized to 0.**
3112 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
3113 specified before the ``pattern`` token.
3115 - ``group {group id}``: priority group.
3116 - ``priority {level}``: priority level within group.
3117 - ``ingress``: rule applies to ingress traffic.
3118 - ``egress``: rule applies to egress traffic.
3120 Each instance of an attribute specified several times overrides the previous
3121 value as shown below (group 4 is used)::
3123 testpmd> flow create 0 group 42 group 24 group 4 [...]
3125 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
3127 While not specifying a direction is an error, some rules may allow both
3130 Most rules affect RX therefore contain the ``ingress`` token::
3132 testpmd> flow create 0 ingress pattern [...]
3137 A matching pattern starts after the ``pattern`` token. It is made of pattern
3138 items and is terminated by a mandatory ``end`` item.
3140 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
3141 rte_flow_item_type``).
3143 The ``/`` token is used as a separator between pattern items as shown
3146 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
3148 Note that protocol items like these must be stacked from lowest to highest
3149 layer to make sense. For instance, the following rule is either invalid or
3150 unlikely to match any packet::
3152 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
3154 More information on these restrictions can be found in the *rte_flow*
3157 Several items support additional specification structures, for example
3158 ``ipv4`` allows specifying source and destination addresses as follows::
3160 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
3161 dst is 10.2.0.0 / end [...]
3163 This rule matches all IPv4 traffic with the specified properties.
3165 In this example, ``src`` and ``dst`` are field names of the underlying
3166 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
3167 in a similar fashion.
3169 The ``is`` token means that the subsequent value must be matched exactly,
3170 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
3171 accordingly. Possible assignment tokens are:
3173 - ``is``: match value perfectly (with full bit-mask).
3174 - ``spec``: match value according to configured bit-mask.
3175 - ``last``: specify upper bound to establish a range.
3176 - ``mask``: specify bit-mask with relevant bits set to one.
3177 - ``prefix``: generate bit-mask from a prefix length.
3179 These yield identical results::
3181 ipv4 src is 10.1.1.1
3185 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
3189 ipv4 src spec 10.1.1.1 src prefix 32
3193 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
3197 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
3199 Inclusive ranges can be defined with ``last``::
3201 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
3203 Note that ``mask`` affects both ``spec`` and ``last``::
3205 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
3206 # matches 10.1.0.0 to 10.2.255.255
3208 Properties can be modified multiple times::
3210 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
3214 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
3219 This section lists supported pattern items and their attributes, if any.
3221 - ``end``: end list of pattern items.
3223 - ``void``: no-op pattern item.
3225 - ``invert``: perform actions when pattern does not match.
3227 - ``any``: match any protocol for the current layer.
3229 - ``num {unsigned}``: number of layers covered.
3231 - ``pf``: match packets addressed to the physical function.
3233 - ``vf``: match packets addressed to a virtual function ID.
3235 - ``id {unsigned}``: destination VF ID.
3237 - ``port``: device-specific physical port index to use.
3239 - ``index {unsigned}``: physical port index.
3241 - ``raw``: match an arbitrary byte string.
3243 - ``relative {boolean}``: look for pattern after the previous item.
3244 - ``search {boolean}``: search pattern from offset (see also limit).
3245 - ``offset {integer}``: absolute or relative offset for pattern.
3246 - ``limit {unsigned}``: search area limit for start of pattern.
3247 - ``pattern {string}``: byte string to look for.
3249 - ``eth``: match Ethernet header.
3251 - ``dst {MAC-48}``: destination MAC.
3252 - ``src {MAC-48}``: source MAC.
3253 - ``type {unsigned}``: EtherType.
3255 - ``vlan``: match 802.1Q/ad VLAN tag.
3257 - ``tpid {unsigned}``: tag protocol identifier.
3258 - ``tci {unsigned}``: tag control information.
3259 - ``pcp {unsigned}``: priority code point.
3260 - ``dei {unsigned}``: drop eligible indicator.
3261 - ``vid {unsigned}``: VLAN identifier.
3263 - ``ipv4``: match IPv4 header.
3265 - ``tos {unsigned}``: type of service.
3266 - ``ttl {unsigned}``: time to live.
3267 - ``proto {unsigned}``: next protocol ID.
3268 - ``src {ipv4 address}``: source address.
3269 - ``dst {ipv4 address}``: destination address.
3271 - ``ipv6``: match IPv6 header.
3273 - ``tc {unsigned}``: traffic class.
3274 - ``flow {unsigned}``: flow label.
3275 - ``proto {unsigned}``: protocol (next header).
3276 - ``hop {unsigned}``: hop limit.
3277 - ``src {ipv6 address}``: source address.
3278 - ``dst {ipv6 address}``: destination address.
3280 - ``icmp``: match ICMP header.
3282 - ``type {unsigned}``: ICMP packet type.
3283 - ``code {unsigned}``: ICMP packet code.
3285 - ``udp``: match UDP header.
3287 - ``src {unsigned}``: UDP source port.
3288 - ``dst {unsigned}``: UDP destination port.
3290 - ``tcp``: match TCP header.
3292 - ``src {unsigned}``: TCP source port.
3293 - ``dst {unsigned}``: TCP destination port.
3295 - ``sctp``: match SCTP header.
3297 - ``src {unsigned}``: SCTP source port.
3298 - ``dst {unsigned}``: SCTP destination port.
3299 - ``tag {unsigned}``: validation tag.
3300 - ``cksum {unsigned}``: checksum.
3302 - ``vxlan``: match VXLAN header.
3304 - ``vni {unsigned}``: VXLAN identifier.
3306 - ``e_tag``: match IEEE 802.1BR E-Tag header.
3308 - ``grp_ecid_b {unsigned}``: GRP and E-CID base.
3310 - ``nvgre``: match NVGRE header.
3312 - ``tni {unsigned}``: virtual subnet ID.
3314 - ``mpls``: match MPLS header.
3316 - ``label {unsigned}``: MPLS label.
3318 - ``gre``: match GRE header.
3320 - ``protocol {unsigned}``: protocol type.
3322 - ``fuzzy``: fuzzy pattern match, expect faster than default.
3324 - ``thresh {unsigned}``: accuracy threshold.
3326 - ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header.
3328 - ``teid {unsigned}``: tunnel endpoint identifier.
3330 - ``geneve``: match GENEVE header.
3332 - ``vni {unsigned}``: virtual network identifier.
3333 - ``protocol {unsigned}``: protocol type.
3338 A list of actions starts after the ``actions`` token in the same fashion as
3339 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
3340 terminated by a mandatory ``end`` action.
3342 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
3343 rte_flow_action_type``).
3345 Dropping all incoming UDPv4 packets can be expressed as follows::
3347 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3350 Several actions have configurable properties which must be specified when
3351 there is no valid default value. For example, ``queue`` requires a target
3354 This rule redirects incoming UDPv4 traffic to queue index 6::
3356 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3357 actions queue index 6 / end
3359 While this one could be rejected by PMDs (unspecified queue index)::
3361 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3364 As defined by *rte_flow*, the list is not ordered, all actions of a given
3365 rule are performed simultaneously. These are equivalent::
3367 queue index 6 / void / mark id 42 / end
3371 void / mark id 42 / queue index 6 / end
3373 All actions in a list should have different types, otherwise only the last
3374 action of a given type is taken into account::
3376 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
3380 drop / drop / drop / end # drop is performed only once
3384 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
3386 Considering they are performed simultaneously, opposite and overlapping
3387 actions can sometimes be combined when the end result is unambiguous::
3389 drop / queue index 6 / end # drop has no effect
3393 drop / dup index 6 / end # same as above
3397 queue index 6 / rss queues 6 7 8 / end # queue has no effect
3401 drop / passthru / end # drop has no effect
3403 Note that PMDs may still refuse such combinations.
3408 This section lists supported actions and their attributes, if any.
3410 - ``end``: end list of actions.
3412 - ``void``: no-op action.
3414 - ``passthru``: let subsequent rule process matched packets.
3416 - ``mark``: attach 32 bit value to packets.
3418 - ``id {unsigned}``: 32 bit value to return with packets.
3420 - ``flag``: flag packets.
3422 - ``queue``: assign packets to a given queue index.
3424 - ``index {unsigned}``: queue index to use.
3426 - ``drop``: drop packets (note: passthru has priority).
3428 - ``count``: enable counters for this rule.
3430 - ``dup``: duplicate packets to a given queue index.
3432 - ``index {unsigned}``: queue index to duplicate packets to.
3434 - ``rss``: spread packets among several queues.
3436 - ``queues [{unsigned} [...]] end``: queue indices to use.
3438 - ``pf``: redirect packets to physical device function.
3440 - ``vf``: redirect packets to virtual device function.
3442 - ``original {boolean}``: use original VF ID if possible.
3443 - ``id {unsigned}``: VF ID to redirect packets to.
3445 Destroying flow rules
3446 ~~~~~~~~~~~~~~~~~~~~~
3448 ``flow destroy`` destroys one or more rules from their rule ID (as returned
3449 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
3450 times as necessary::
3452 flow destroy {port_id} rule {rule_id} [...]
3454 If successful, it will show::
3456 Flow rule #[...] destroyed
3458 It does not report anything for rule IDs that do not exist. The usual error
3459 message is shown when a rule cannot be destroyed::
3461 Caught error type [...] ([...]): [...]
3463 ``flow flush`` destroys all rules on a device and does not take extra
3464 arguments. It is bound to ``rte_flow_flush()``::
3466 flow flush {port_id}
3468 Any errors are reported as above.
3470 Creating several rules and destroying them::
3472 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3473 actions queue index 2 / end
3474 Flow rule #0 created
3475 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3476 actions queue index 3 / end
3477 Flow rule #1 created
3478 testpmd> flow destroy 0 rule 0 rule 1
3479 Flow rule #1 destroyed
3480 Flow rule #0 destroyed
3483 The same result can be achieved using ``flow flush``::
3485 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3486 actions queue index 2 / end
3487 Flow rule #0 created
3488 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3489 actions queue index 3 / end
3490 Flow rule #1 created
3491 testpmd> flow flush 0
3494 Non-existent rule IDs are ignored::
3496 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3497 actions queue index 2 / end
3498 Flow rule #0 created
3499 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3500 actions queue index 3 / end
3501 Flow rule #1 created
3502 testpmd> flow destroy 0 rule 42 rule 10 rule 2
3504 testpmd> flow destroy 0 rule 0
3505 Flow rule #0 destroyed
3511 ``flow query`` queries a specific action of a flow rule having that
3512 ability. Such actions collect information that can be reported using this
3513 command. It is bound to ``rte_flow_query()``::
3515 flow query {port_id} {rule_id} {action}
3517 If successful, it will display either the retrieved data for known actions
3518 or the following message::
3520 Cannot display result for action type [...] ([...])
3522 Otherwise, it will complain either that the rule does not exist or that some
3525 Flow rule #[...] not found
3529 Caught error type [...] ([...]): [...]
3531 Currently only the ``count`` action is supported. This action reports the
3532 number of packets that hit the flow rule and the total number of bytes. Its
3533 output has the following format::
3536 hits_set: [...] # whether "hits" contains a valid value
3537 bytes_set: [...] # whether "bytes" contains a valid value
3538 hits: [...] # number of packets
3539 bytes: [...] # number of bytes
3541 Querying counters for TCPv6 packets redirected to queue 6::
3543 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
3544 actions queue index 6 / count / end
3545 Flow rule #4 created
3546 testpmd> flow query 0 4 count
3557 ``flow list`` lists existing flow rules sorted by priority and optionally
3558 filtered by group identifiers::
3560 flow list {port_id} [group {group_id}] [...]
3562 This command only fails with the following message if the device does not
3567 Output consists of a header line followed by a short description of each
3568 flow rule, one per line. There is no output at all when no flow rules are
3569 configured on the device::
3571 ID Group Prio Attr Rule
3572 [...] [...] [...] [...] [...]
3574 ``Attr`` column flags:
3576 - ``i`` for ``ingress``.
3577 - ``e`` for ``egress``.
3579 Creating several flow rules and listing them::
3581 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3582 actions queue index 6 / end
3583 Flow rule #0 created
3584 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3585 actions queue index 2 / end
3586 Flow rule #1 created
3587 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
3588 actions rss queues 6 7 8 end / end
3589 Flow rule #2 created
3590 testpmd> flow list 0
3591 ID Group Prio Attr Rule
3592 0 0 0 i- ETH IPV4 => QUEUE
3593 1 0 0 i- ETH IPV6 => QUEUE
3594 2 0 5 i- ETH IPV4 UDP => RSS
3597 Rules are sorted by priority (i.e. group ID first, then priority level)::
3599 testpmd> flow list 1
3600 ID Group Prio Attr Rule
3601 0 0 0 i- ETH => COUNT
3602 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3603 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3604 1 24 0 i- ETH IPV4 UDP => QUEUE
3605 4 24 10 i- ETH IPV4 TCP => DROP
3606 3 24 20 i- ETH IPV4 => DROP
3607 2 24 42 i- ETH IPV4 UDP => QUEUE
3608 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3611 Output can be limited to specific groups::
3613 testpmd> flow list 1 group 0 group 63
3614 ID Group Prio Attr Rule
3615 0 0 0 i- ETH => COUNT
3616 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3617 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3618 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3621 Toggling isolated mode
3622 ~~~~~~~~~~~~~~~~~~~~~~
3624 ``flow isolate`` can be used to tell the underlying PMD that ingress traffic
3625 must only be injected from the defined flow rules; that no default traffic
3626 is expected outside those rules and the driver is free to assign more
3627 resources to handle them. It is bound to ``rte_flow_isolate()``::
3629 flow isolate {port_id} {boolean}
3631 If successful, enabling or disabling isolated mode shows either::
3633 Ingress traffic on port [...]
3634 is now restricted to the defined flow rules
3638 Ingress traffic on port [...]
3639 is not restricted anymore to the defined flow rules
3641 Otherwise, in case of error::
3643 Caught error type [...] ([...]): [...]
3645 Mainly due to its side effects, PMDs supporting this mode may not have the
3646 ability to toggle it more than once without reinitializing affected ports
3647 first (e.g. by exiting testpmd).
3649 Enabling isolated mode::
3651 testpmd> flow isolate 0 true
3652 Ingress traffic on port 0 is now restricted to the defined flow rules
3655 Disabling isolated mode::
3657 testpmd> flow isolate 0 false
3658 Ingress traffic on port 0 is not restricted anymore to the defined flow rules
3661 Sample QinQ flow rules
3662 ~~~~~~~~~~~~~~~~~~~~~~
3664 Before creating QinQ rule(s) the following commands should be issued to enable QinQ::
3666 testpmd> port stop 0
3667 testpmd> vlan set qinq on 0
3669 The above command sets the inner and outer TPID's to 0x8100.
3671 To change the TPID's the following commands should be used::
3673 testpmd> vlan set outer tpid 0xa100 0
3674 testpmd> vlan set inner tpid 0x9100 0
3675 testpmd> port start 0
3677 Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM.
3681 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 /
3682 vlan tci is 456 / end actions vf id 1 / queue index 0 / end
3683 Flow rule #0 validated
3685 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 /
3686 vlan tci is 456 / end actions vf id 123 / queue index 0 / end
3687 Flow rule #0 created
3689 testpmd> flow list 0
3690 ID Group Prio Attr Rule
3691 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3693 Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host.
3697 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 /
3698 vlan tci is 654 / end actions pf / queue index 0 / end
3699 Flow rule #1 validated
3701 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 /
3702 vlan tci is 654 / end actions pf / queue index 1 / end
3703 Flow rule #1 created
3705 testpmd> flow list 0
3706 ID Group Prio Attr Rule
3707 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3708 1 0 0 i- ETH VLAN VLAN=>PF QUEUE