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 Note: TX timestamping is only available in the "Full Featured" TX path. To force ``testpmd`` into this mode set ``--txqflags=0``.
354 testpmd> set fwd rxonly
356 Set rxonly packet forwarding mode
362 Display an RX descriptor for a port RX queue::
364 testpmd> read rxd (port_id) (queue_id) (rxd_id)
368 testpmd> read rxd 0 0 4
369 0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180
374 Display a TX descriptor for a port TX queue::
376 testpmd> read txd (port_id) (queue_id) (txd_id)
380 testpmd> read txd 0 0 4
381 0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C
386 Get loaded dynamic device personalization (DDP) package info list::
388 testpmd> ddp get list (port_id)
393 Display information about dynamic device personalization (DDP) profile::
395 testpmd> ddp get info (profile_path)
400 Display VF statistics::
402 testpmd> show vf stats (port_id) (vf_id)
407 Reset VF statistics::
409 testpmd> clear vf stats (port_id) (vf_id)
411 show port pctype mapping
412 ~~~~~~~~~~~~~~~~~~~~~~~~
414 List all items from the pctype mapping table::
416 testpmd> show port (port_id) pctype mapping
419 Configuration Functions
420 -----------------------
422 The testpmd application can be configured from the runtime as well as from the command-line.
424 This section details the available configuration functions that are available.
428 Configuration changes only become active when forwarding is started/restarted.
433 Reset forwarding to the default configuration::
440 Set the debug verbosity level::
442 testpmd> set verbose (level)
444 Currently the only available levels are 0 (silent except for error) and 1 (fully verbose).
449 Set the number of ports used by the application:
453 This is equivalent to the ``--nb-ports`` command-line option.
458 Set the number of cores used by the application::
460 testpmd> set nbcore (num)
462 This is equivalent to the ``--nb-cores`` command-line option.
466 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port.
471 Set the forwarding cores hexadecimal mask::
473 testpmd> set coremask (mask)
475 This is equivalent to the ``--coremask`` command-line option.
479 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding.
484 Set the forwarding ports hexadecimal mask::
486 testpmd> set portmask (mask)
488 This is equivalent to the ``--portmask`` command-line option.
493 Set number of packets per burst::
495 testpmd> set burst (num)
497 This is equivalent to the ``--burst command-line`` option.
499 When retry is enabled, the transmit delay time and number of retries can also be set::
501 testpmd> set burst tx delay (microseconds) retry (num)
506 Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode::
508 testpmd> set txpkts (x[,y]*)
510 Where x[,y]* represents a CSV list of values, without white space.
515 Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes::
517 testpmd> set txsplit (off|on|rand)
521 * ``off`` disable packet copy & split for CSUM mode.
523 * ``on`` split outgoing packet into multiple segments. Size of each segment
524 and number of segments per packet is determined by ``set txpkts`` command
527 * ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments.
532 Set the list of forwarding cores::
534 testpmd> set corelist (x[,y]*)
536 For example, to change the forwarding cores:
538 .. code-block:: console
540 testpmd> set corelist 3,1
541 testpmd> show config fwd
543 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled
544 Logical Core 3 (socket 0) forwards packets on 1 streams:
545 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01
546 Logical Core 1 (socket 0) forwards packets on 1 streams:
547 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
551 The cores are used in the same order as specified on the command line.
556 Set the list of forwarding ports::
558 testpmd> set portlist (x[,y]*)
560 For example, to change the port forwarding:
562 .. code-block:: console
564 testpmd> set portlist 0,2,1,3
565 testpmd> show config fwd
567 io packet forwarding - ports=4 - cores=1 - streams=4
568 Logical Core 3 (socket 0) forwards packets on 4 streams:
569 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01
570 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
571 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03
572 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02
577 Enable/disable tx loopback::
579 testpmd> set tx loopback (port_id) (on|off)
584 set drop enable bit for all queues::
586 testpmd> set all queues drop (port_id) (on|off)
588 set split drop enable (for VF)
589 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
591 set split drop enable bit for VF from PF::
593 testpmd> set vf split drop (port_id) (vf_id) (on|off)
595 set mac antispoof (for VF)
596 ~~~~~~~~~~~~~~~~~~~~~~~~~~
598 Set mac antispoof for a VF from the PF::
600 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off)
605 Enable/disable MACsec offload::
607 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off)
608 testpmd> set macsec offload (port_id) off
613 Configure MACsec secure connection (SC)::
615 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi)
619 The pi argument is ignored for tx.
620 Check the NIC Datasheet for hardware limits.
625 Configure MACsec secure association (SA)::
627 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key)
631 The IDX value must be 0 or 1.
632 Check the NIC Datasheet for hardware limits.
634 set broadcast mode (for VF)
635 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
637 Set broadcast mode for a VF from the PF::
639 testpmd> set vf broadcast (port_id) (vf_id) (on|off)
644 Set the VLAN strip on a port::
646 testpmd> vlan set strip (on|off) (port_id)
651 Set the VLAN strip for a queue on a port::
653 testpmd> vlan set stripq (on|off) (port_id,queue_id)
655 vlan set stripq (for VF)
656 ~~~~~~~~~~~~~~~~~~~~~~~~
658 Set VLAN strip for all queues in a pool for a VF from the PF::
660 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off)
662 vlan set insert (for VF)
663 ~~~~~~~~~~~~~~~~~~~~~~~~
665 Set VLAN insert for a VF from the PF::
667 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id)
669 vlan set tag (for VF)
670 ~~~~~~~~~~~~~~~~~~~~~
672 Set VLAN tag for a VF from the PF::
674 testpmd> set vf vlan tag (port_id) (vf_id) (on|off)
676 vlan set antispoof (for VF)
677 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
679 Set VLAN antispoof for a VF from the PF::
681 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off)
686 Set the VLAN filter on a port::
688 testpmd> vlan set filter (on|off) (port_id)
693 Set the VLAN QinQ (extended queue in queue) on for a port::
695 testpmd> vlan set qinq (on|off) (port_id)
700 Set the inner or outer VLAN TPID for packet filtering on a port::
702 testpmd> vlan set (inner|outer) tpid (value) (port_id)
706 TPID value must be a 16-bit number (value <= 65536).
711 Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID::
713 testpmd> rx_vlan add (vlan_id|all) (port_id)
717 VLAN filter must be set on that port. VLAN ID < 4096.
718 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries
719 in VFTA table. This is important if enabling all vlan_ids.
724 Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID::
726 testpmd> rx_vlan rm (vlan_id|all) (port_id)
731 Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID::
733 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask)
738 Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID::
740 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask)
745 Add a tunnel filter on a port::
747 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \
748 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
749 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
751 The available information categories are:
753 * ``vxlan``: Set tunnel type as VXLAN.
755 * ``nvgre``: Set tunnel type as NVGRE.
757 * ``ipingre``: Set tunnel type as IP-in-GRE.
759 * ``imac-ivlan``: Set filter type as Inner MAC and VLAN.
761 * ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID.
763 * ``imac-tenid``: Set filter type as Inner MAC and tenant ID.
765 * ``imac``: Set filter type as Inner MAC.
767 * ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID.
769 * ``oip``: Set filter type as Outer IP.
771 * ``iip``: Set filter type as Inner IP.
775 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \
776 192.168.2.2 0 ipingre oip 1 1
778 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP.
783 Remove a tunnel filter on a port::
785 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \
786 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
787 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
792 Add an UDP port for VXLAN packet filter on a port::
794 testpmd> rx_vxlan_port add (udp_port) (port_id)
799 Remove an UDP port for VXLAN packet filter on a port::
801 testpmd> rx_vxlan_port rm (udp_port) (port_id)
806 Set hardware insertion of VLAN IDs in packets sent on a port::
808 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer]
810 For example, set a single VLAN ID (5) insertion on port 0::
814 Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1::
822 Set port based hardware insertion of VLAN ID in packets sent on a port::
824 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off)
829 Disable hardware insertion of a VLAN header in packets sent on a port::
831 testpmd> tx_vlan reset (port_id)
836 Select hardware or software calculation of the checksum when
837 transmitting a packet using the ``csum`` forwarding engine::
839 testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id)
843 * ``ip|udp|tcp|sctp`` always relate to the inner layer.
845 * ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized
846 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are
847 supported). See also the ``csum parse-tunnel`` command.
851 Check the NIC Datasheet for hardware limits.
856 Define how tunneled packets should be handled by the csum forward
859 testpmd> csum parse-tunnel (on|off) (tx_port_id)
861 If enabled, the csum forward engine will try to recognize supported
862 tunnel headers (vxlan, gre, ipip).
864 If disabled, treat tunnel packets as non-tunneled packets (a inner
865 header is handled as a packet payload).
869 The port argument is the TX port like in the ``csum set`` command.
873 Consider a packet in packet like the following::
875 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in
877 * If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
878 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the
879 ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``).
881 * If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
882 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``.
887 Display tx checksum offload configuration::
889 testpmd> csum show (port_id)
894 Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine::
896 testpmd> tso set (segsize) (port_id)
900 Check the NIC datasheet for hardware limits.
905 Display the status of TCP Segmentation Offload::
907 testpmd> tso show (port_id)
912 Enable or disable GRO in ``csum`` forwarding engine::
914 testpmd> gro (on|off) (port_id)
916 If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4
917 packets received from the given port.
919 If disabled, packets received from the given port won't be performed
920 GRO. By default, GRO is disabled for all ports.
924 When enable GRO for a port, TCP/IPv4 packets received from the port
925 will be performed GRO. After GRO, the merged packets are multi-segments.
926 But csum forwarding engine doesn't support to calculate TCP checksum
927 for multi-segment packets in SW. So please select TCP HW checksum
928 calculation for the port which GROed packets are transmitted to.
933 Set max flow number and max packet number per-flow for GRO::
935 testpmd> gro set (max_flow_num) (max_item_num_per_flow) (port_id)
937 The product of ``max_flow_num`` and ``max_item_num_per_flow`` is the max
938 number of packets a GRO table can store.
940 If current packet number is greater than or equal to the max value, GRO
941 will stop processing incoming packets.
946 Add an alternative MAC address to a port::
948 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX)
953 Remove a MAC address from a port::
955 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX)
957 mac_addr add (for VF)
958 ~~~~~~~~~~~~~~~~~~~~~
960 Add an alternative MAC address for a VF to a port::
962 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX)
967 Set the default MAC address for a port::
969 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX)
971 mac_addr set (for VF)
972 ~~~~~~~~~~~~~~~~~~~~~
974 Set the MAC address for a VF from the PF::
976 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX)
981 Set the unicast hash filter(s) on/off for a port::
983 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
988 Set the promiscuous mode on for a port or for all ports.
989 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
991 testpmd> set promisc (port_id|all) (on|off)
996 Set the allmulti mode for a port or for all ports::
998 testpmd> set allmulti (port_id|all) (on|off)
1000 Same as the ifconfig (8) option. Controls how multicast packets are handled.
1002 set promisc (for VF)
1003 ~~~~~~~~~~~~~~~~~~~~
1005 Set the unicast promiscuous mode for a VF from PF.
1006 It's supported by Intel i40e NICs now.
1007 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1009 testpmd> set vf promisc (port_id) (vf_id) (on|off)
1011 set allmulticast (for VF)
1012 ~~~~~~~~~~~~~~~~~~~~~~~~~
1014 Set the multicast promiscuous mode for a VF from PF.
1015 It's supported by Intel i40e NICs now.
1016 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1018 testpmd> set vf allmulti (port_id) (vf_id) (on|off)
1020 set tx max bandwidth (for VF)
1021 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1023 Set TX max absolute bandwidth (Mbps) for a VF from PF::
1025 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth)
1027 set tc tx min bandwidth (for VF)
1028 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1030 Set all TCs' TX min relative bandwidth (%) for a VF from PF::
1032 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...)
1034 set tc tx max bandwidth (for VF)
1035 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1037 Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF::
1039 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth)
1041 set tc strict link priority mode
1042 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1044 Set some TCs' strict link priority mode on a physical port::
1046 testpmd> set tx strict-link-priority (port_id) (tc_bitmap)
1048 set tc tx min bandwidth
1049 ~~~~~~~~~~~~~~~~~~~~~~~
1051 Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs::
1053 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...)
1058 Set the link flow control parameter on a port::
1060 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1061 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
1062 autoneg (on|off) (port_id)
1066 * ``high_water`` (integer): High threshold value to trigger XOFF.
1068 * ``low_water`` (integer): Low threshold value to trigger XON.
1070 * ``pause_time`` (integer): Pause quota in the Pause frame.
1072 * ``send_xon`` (0/1): Send XON frame.
1074 * ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
1076 * ``autoneg``: Change the auto-negotiation parameter.
1081 Set the priority flow control parameter on a port::
1083 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1084 (pause_time) (priority) (port_id)
1088 * ``high_water`` (integer): High threshold value.
1090 * ``low_water`` (integer): Low threshold value.
1092 * ``pause_time`` (integer): Pause quota in the Pause frame.
1094 * ``priority`` (0-7): VLAN User Priority.
1099 Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
1101 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
1103 For example, to set rx queue 2 on port 0 to mapping 5::
1105 testpmd>set stat_qmap rx 0 2 5
1107 set port - rx/tx (for VF)
1108 ~~~~~~~~~~~~~~~~~~~~~~~~~
1110 Set VF receive/transmit from a port::
1112 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
1114 set port - mac address filter (for VF)
1115 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1117 Add/Remove unicast or multicast MAC addr filter for a VF::
1119 testpmd> set port (port_id) vf (vf_id) (mac_addr) \
1120 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
1122 set port - rx mode(for VF)
1123 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1125 Set the VF receive mode of a port::
1127 testpmd> set port (port_id) vf (vf_id) \
1128 rxmode (AUPE|ROPE|BAM|MPE) (on|off)
1130 The available receive modes are:
1132 * ``AUPE``: Accepts untagged VLAN.
1134 * ``ROPE``: Accepts unicast hash.
1136 * ``BAM``: Accepts broadcast packets.
1138 * ``MPE``: Accepts all multicast packets.
1140 set port - tx_rate (for Queue)
1141 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1143 Set TX rate limitation for a queue on a port::
1145 testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1147 set port - tx_rate (for VF)
1148 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
1150 Set TX rate limitation for queues in VF on a port::
1152 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1154 set port - mirror rule
1155 ~~~~~~~~~~~~~~~~~~~~~~
1157 Set pool or vlan type mirror rule for a port::
1159 testpmd> set port (port_id) mirror-rule (rule_id) \
1160 (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1161 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1163 Set link mirror rule for a port::
1165 testpmd> set port (port_id) mirror-rule (rule_id) \
1166 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1168 For example to enable mirror traffic with vlan 0,1 to pool 0::
1170 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1172 reset port - mirror rule
1173 ~~~~~~~~~~~~~~~~~~~~~~~~
1175 Reset a mirror rule for a port::
1177 testpmd> reset port (port_id) mirror-rule (rule_id)
1182 Set the flush on RX streams before forwarding.
1183 The default is flush ``on``.
1184 Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1186 testpmd> set flush_rx off
1191 Set the bypass mode for the lowest port on bypass enabled NIC::
1193 testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1198 Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1200 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1201 mode (normal|bypass|isolate) (port_id)
1205 * ``timeout``: Enable bypass after watchdog timeout.
1207 * ``os_on``: Enable bypass when OS/board is powered on.
1209 * ``os_off``: Enable bypass when OS/board is powered off.
1211 * ``power_on``: Enable bypass when power supply is turned on.
1213 * ``power_off``: Enable bypass when power supply is turned off.
1219 Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1221 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1226 Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1228 testpmd> show bypass config (port_id)
1233 Set link up for a port::
1235 testpmd> set link-up port (port id)
1240 Set link down for a port::
1242 testpmd> set link-down port (port id)
1247 Enable E-tag insertion for a VF on a port::
1249 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1251 Disable E-tag insertion for a VF on a port::
1253 testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1255 Enable/disable E-tag stripping on a port::
1257 testpmd> E-tag set stripping (on|off) port (port_id)
1259 Enable/disable E-tag based forwarding on a port::
1261 testpmd> E-tag set forwarding (on|off) port (port_id)
1263 Add an E-tag forwarding filter on a port::
1265 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1267 Delete an E-tag forwarding filter on a port::
1268 testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1273 Load a dynamic device personalization (DDP) package::
1275 testpmd> ddp add (port_id) (package_path[,output_path])
1280 Delete a dynamic device personalization package::
1282 testpmd> ddp del (port_id) (package_path)
1287 List all items from the ptype mapping table::
1289 testpmd> ptype mapping get (port_id) (valid_only)
1293 * ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0).
1295 Replace a specific or a group of software defined ptype with a new one::
1297 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type)
1301 * ``target``: A specific software ptype or a mask to represent a group of software ptypes.
1303 * ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1).
1305 * ``pkt_type``: The new software ptype to replace the old ones.
1307 Update hardware defined ptype to software defined packet type mapping table::
1309 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype)
1313 * ``hw_ptype``: hardware ptype as the index of the ptype mapping table.
1315 * ``sw_ptype``: software ptype as the value of the ptype mapping table.
1317 Reset ptype mapping table::
1319 testpmd> ptype mapping reset (port_id)
1324 The following sections show functions for configuring ports.
1328 Port configuration changes only become active when forwarding is started/restarted.
1333 Attach a port specified by pci address or virtual device args::
1335 testpmd> port attach (identifier)
1337 To attach a new pci device, the device should be recognized by kernel first.
1338 Then it should be moved under DPDK management.
1339 Finally the port can be attached to testpmd.
1341 For example, to move a pci device using ixgbe under DPDK management:
1343 .. code-block:: console
1345 # Check the status of the available devices.
1346 ./usertools/dpdk-devbind.py --status
1348 Network devices using DPDK-compatible driver
1349 ============================================
1352 Network devices using kernel driver
1353 ===================================
1354 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1357 # Bind the device to igb_uio.
1358 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1361 # Recheck the status of the devices.
1362 ./usertools/dpdk-devbind.py --status
1363 Network devices using DPDK-compatible driver
1364 ============================================
1365 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1367 To attach a port created by virtual device, above steps are not needed.
1369 For example, to attach a port whose pci address is 0000:0a:00.0.
1371 .. code-block:: console
1373 testpmd> port attach 0000:0a:00.0
1374 Attaching a new port...
1375 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1376 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1377 EAL: PCI memory mapped at 0x7f83bfa00000
1378 EAL: PCI memory mapped at 0x7f83bfa80000
1379 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1380 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1381 Port 0 is attached. Now total ports is 1
1384 For example, to attach a port created by pcap PMD.
1386 .. code-block:: console
1388 testpmd> port attach net_pcap0
1389 Attaching a new port...
1390 PMD: Initializing pmd_pcap for net_pcap0
1391 PMD: Creating pcap-backed ethdev on numa socket 0
1392 Port 0 is attached. Now total ports is 1
1395 In this case, identifier is ``net_pcap0``.
1396 This identifier format is the same as ``--vdev`` format of DPDK applications.
1398 For example, to re-attach a bonded port which has been previously detached,
1399 the mode and slave parameters must be given.
1401 .. code-block:: console
1403 testpmd> port attach net_bond_0,mode=0,slave=1
1404 Attaching a new port...
1405 EAL: Initializing pmd_bond for net_bond_0
1406 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1407 Port 0 is attached. Now total ports is 1
1414 Detach a specific port::
1416 testpmd> port detach (port_id)
1418 Before detaching a port, the port should be stopped and closed.
1420 For example, to detach a pci device port 0.
1422 .. code-block:: console
1424 testpmd> port stop 0
1427 testpmd> port close 0
1431 testpmd> port detach 0
1433 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1434 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1435 EAL: PCI memory unmapped at 0x7f83bfa00000
1436 EAL: PCI memory unmapped at 0x7f83bfa80000
1440 For example, to detach a virtual device port 0.
1442 .. code-block:: console
1444 testpmd> port stop 0
1447 testpmd> port close 0
1451 testpmd> port detach 0
1453 PMD: Closing pcap ethdev on numa socket 0
1454 Port 'net_pcap0' is detached. Now total ports is 0
1457 To remove a pci device completely from the system, first detach the port from testpmd.
1458 Then the device should be moved under kernel management.
1459 Finally the device can be removed using kernel pci hotplug functionality.
1461 For example, to move a pci device under kernel management:
1463 .. code-block:: console
1465 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1467 ./usertools/dpdk-devbind.py --status
1469 Network devices using DPDK-compatible driver
1470 ============================================
1473 Network devices using kernel driver
1474 ===================================
1475 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1477 To remove a port created by a virtual device, above steps are not needed.
1482 Start all ports or a specific port::
1484 testpmd> port start (port_id|all)
1489 Stop all ports or a specific port::
1491 testpmd> port stop (port_id|all)
1496 Close all ports or a specific port::
1498 testpmd> port close (port_id|all)
1500 port start/stop queue
1501 ~~~~~~~~~~~~~~~~~~~~~
1503 Start/stop a rx/tx queue on a specific port::
1505 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1507 Only take effect when port is started.
1512 Set the speed and duplex mode for all ports or a specific port::
1514 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1515 duplex (half|full|auto)
1517 port config - queues/descriptors
1518 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1520 Set number of queues/descriptors for rxq, txq, rxd and txd::
1522 testpmd> port config all (rxq|txq|rxd|txd) (value)
1524 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1526 port config - max-pkt-len
1527 ~~~~~~~~~~~~~~~~~~~~~~~~~
1529 Set the maximum packet length::
1531 testpmd> port config all max-pkt-len (value)
1533 This is equivalent to the ``--max-pkt-len`` command-line option.
1535 port config - CRC Strip
1536 ~~~~~~~~~~~~~~~~~~~~~~~
1538 Set hardware CRC stripping on or off for all ports::
1540 testpmd> port config all crc-strip (on|off)
1542 CRC stripping is on by default.
1544 The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option.
1546 port config - scatter
1547 ~~~~~~~~~~~~~~~~~~~~~~~
1549 Set RX scatter mode on or off for all ports::
1551 testpmd> port config all scatter (on|off)
1553 RX scatter mode is off by default.
1555 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1557 port config - TX queue flags
1558 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1560 Set a hexadecimal bitmap of TX queue flags for all ports::
1562 testpmd> port config all txqflags value
1564 This command is equivalent to the ``--txqflags`` command-line option.
1566 port config - RX Checksum
1567 ~~~~~~~~~~~~~~~~~~~~~~~~~
1569 Set hardware RX checksum offload to on or off for all ports::
1571 testpmd> port config all rx-cksum (on|off)
1573 Checksum offload is off by default.
1575 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1580 Set hardware VLAN on or off for all ports::
1582 testpmd> port config all hw-vlan (on|off)
1584 Hardware VLAN is on by default.
1586 The ``off`` option is equivalent to the ``--disable-hw-vlan`` command-line option.
1588 port config - VLAN filter
1589 ~~~~~~~~~~~~~~~~~~~~~~~~~
1591 Set hardware VLAN filter on or off for all ports::
1593 testpmd> port config all hw-vlan-filter (on|off)
1595 Hardware VLAN filter is on by default.
1597 The ``off`` option is equivalent to the ``--disable-hw-vlan-filter`` command-line option.
1599 port config - VLAN strip
1600 ~~~~~~~~~~~~~~~~~~~~~~~~
1602 Set hardware VLAN strip on or off for all ports::
1604 testpmd> port config all hw-vlan-strip (on|off)
1606 Hardware VLAN strip is on by default.
1608 The ``off`` option is equivalent to the ``--disable-hw-vlan-strip`` command-line option.
1610 port config - VLAN extend
1611 ~~~~~~~~~~~~~~~~~~~~~~~~~
1613 Set hardware VLAN extend on or off for all ports::
1615 testpmd> port config all hw-vlan-extend (on|off)
1617 Hardware VLAN extend is off by default.
1619 The ``off`` option is equivalent to the ``--disable-hw-vlan-extend`` command-line option.
1621 port config - Drop Packets
1622 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1624 Set packet drop for packets with no descriptors on or off for all ports::
1626 testpmd> port config all drop-en (on|off)
1628 Packet dropping for packets with no descriptors is off by default.
1630 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1635 Set the RSS (Receive Side Scaling) mode on or off::
1637 testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1639 RSS is on by default.
1641 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1643 port config - RSS Reta
1644 ~~~~~~~~~~~~~~~~~~~~~~
1646 Set the RSS (Receive Side Scaling) redirection table::
1648 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1653 Set the DCB mode for an individual port::
1655 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1657 The traffic class should be 4 or 8.
1662 Set the number of packets per burst::
1664 testpmd> port config all burst (value)
1666 This is equivalent to the ``--burst`` command-line option.
1668 port config - Threshold
1669 ~~~~~~~~~~~~~~~~~~~~~~~
1671 Set thresholds for TX/RX queues::
1673 testpmd> port config all (threshold) (value)
1675 Where the threshold type can be:
1677 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1679 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1681 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1683 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1685 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1687 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1689 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1691 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1693 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1695 These threshold options are also available from the command-line.
1700 Set the value of ether-type for E-tag::
1702 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1704 Enable/disable the E-tag support::
1706 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1708 port config pctype mapping
1709 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1711 Reset pctype mapping table::
1713 testpmd> port config (port_id) pctype mapping reset
1715 Update hardware defined pctype to software defined flow type mapping table::
1717 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id)
1721 * ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table.
1723 * ``flow_type_id``: software flow type id as the index of the pctype mapping table.
1726 Link Bonding Functions
1727 ----------------------
1729 The Link Bonding functions make it possible to dynamically create and
1730 manage link bonding devices from within testpmd interactive prompt.
1732 create bonded device
1733 ~~~~~~~~~~~~~~~~~~~~
1735 Create a new bonding device::
1737 testpmd> create bonded device (mode) (socket)
1739 For example, to create a bonded device in mode 1 on socket 0::
1741 testpmd> create bonded 1 0
1742 created new bonded device (port X)
1747 Adds Ethernet device to a Link Bonding device::
1749 testpmd> add bonding slave (slave id) (port id)
1751 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
1753 testpmd> add bonding slave 6 10
1756 remove bonding slave
1757 ~~~~~~~~~~~~~~~~~~~~
1759 Removes an Ethernet slave device from a Link Bonding device::
1761 testpmd> remove bonding slave (slave id) (port id)
1763 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
1765 testpmd> remove bonding slave 6 10
1770 Set the Link Bonding mode of a Link Bonding device::
1772 testpmd> set bonding mode (value) (port id)
1774 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
1776 testpmd> set bonding mode 3 10
1781 Set an Ethernet slave device as the primary device on a Link Bonding device::
1783 testpmd> set bonding primary (slave id) (port id)
1785 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
1787 testpmd> set bonding primary 6 10
1792 Set the MAC address of a Link Bonding device::
1794 testpmd> set bonding mac (port id) (mac)
1796 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
1798 testpmd> set bonding mac 10 00:00:00:00:00:01
1800 set bonding xmit_balance_policy
1801 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1803 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
1805 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
1807 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
1809 testpmd> set bonding xmit_balance_policy 10 l34
1812 set bonding mon_period
1813 ~~~~~~~~~~~~~~~~~~~~~~
1815 Set the link status monitoring polling period in milliseconds for a bonding device.
1817 This adds support for PMD slave devices which do not support link status interrupts.
1818 When the mon_period is set to a value greater than 0 then all PMD's which do not support
1819 link status ISR will be queried every polling interval to check if their link status has changed::
1821 testpmd> set bonding mon_period (port_id) (value)
1823 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
1825 testpmd> set bonding mon_period 5 150
1828 set bonding lacp dedicated_queue
1829 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1831 Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic
1832 when in mode 4 (link-aggregration-802.3ad)::
1834 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable)
1837 set bonding agg_mode
1838 ~~~~~~~~~~~~~~~~~~~~
1840 Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad)::
1842 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable)
1848 Show the current configuration of a Link Bonding device::
1850 testpmd> show bonding config (port id)
1853 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
1854 in balance mode with a transmission policy of layer 2+3::
1856 testpmd> show bonding config 9
1858 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
1860 Active Slaves (3): [1 3 4]
1867 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
1868 This is mainly useful for debugging purposes.
1869 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
1870 and fields that can be accessed.
1875 Display the value of a port register::
1877 testpmd> read reg (port_id) (address)
1879 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
1881 testpmd> read reg 0 0xEE00
1882 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
1887 Display a port register bit field::
1889 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
1891 For example, reading the lowest two bits from the register in the example above::
1893 testpmd> read regfield 0 0xEE00 0 1
1894 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
1899 Display a single port register bit::
1901 testpmd> read regbit (port_id) (address) (bit_x)
1903 For example, reading the lowest bit from the register in the example above::
1905 testpmd> read regbit 0 0xEE00 0
1906 port 0 PCI register at offset 0xEE00: bit 0=1
1911 Set the value of a port register::
1913 testpmd> write reg (port_id) (address) (value)
1915 For example, to clear a register::
1917 testpmd> write reg 0 0xEE00 0x0
1918 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
1923 Set bit field of a port register::
1925 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
1927 For example, writing to the register cleared in the example above::
1929 testpmd> write regfield 0 0xEE00 0 1 2
1930 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
1935 Set single bit value of a port register::
1937 testpmd> write regbit (port_id) (address) (bit_x) (value)
1939 For example, to set the high bit in the register from the example above::
1941 testpmd> write regbit 0 0xEE00 31 1
1942 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
1948 This section details the available filter functions that are available.
1950 Note these functions interface the deprecated legacy filtering framework,
1951 superseded by *rte_flow*. See `Flow rules management`_.
1954 ~~~~~~~~~~~~~~~~~~~~
1956 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
1958 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
1959 ethertype (ether_type) (drop|fwd) queue (queue_id)
1961 The available information parameters are:
1963 * ``port_id``: The port which the Ethertype filter assigned on.
1965 * ``mac_addr``: Compare destination mac address.
1967 * ``mac_ignr``: Ignore destination mac address match.
1969 * ``mac_address``: Destination mac address to match.
1971 * ``ether_type``: The EtherType value want to match,
1972 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
1974 * ``queue_id``: The receive queue associated with this EtherType filter.
1975 It is meaningless when deleting or dropping.
1977 Example, to add/remove an ethertype filter rule::
1979 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
1980 ethertype 0x0806 fwd queue 3
1982 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
1983 ethertype 0x0806 fwd queue 3
1988 Add or delete a 2-tuple filter,
1989 which identifies packets by specific protocol and destination TCP/UDP port
1990 and forwards packets into one of the receive queues::
1992 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
1993 protocol (protocol_value) mask (mask_value) \
1994 tcp_flags (tcp_flags_value) priority (prio_value) \
1997 The available information parameters are:
1999 * ``port_id``: The port which the 2-tuple filter assigned on.
2001 * ``dst_port_value``: Destination port in L4.
2003 * ``protocol_value``: IP L4 protocol.
2005 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
2007 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
2009 * ``prio_value``: Priority of this filter.
2011 * ``queue_id``: The receive queue associated with this 2-tuple filter.
2013 Example, to add/remove an 2tuple filter rule::
2015 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
2016 tcp_flags 0x02 priority 3 queue 3
2018 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
2019 tcp_flags 0x02 priority 3 queue 3
2024 Add or delete a 5-tuple filter,
2025 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
2026 and routes packets into one of the receive queues::
2028 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
2029 (src_address) dst_port (dst_port_value) \
2030 src_port (src_port_value) protocol (protocol_value) \
2031 mask (mask_value) tcp_flags (tcp_flags_value) \
2032 priority (prio_value) queue (queue_id)
2034 The available information parameters are:
2036 * ``port_id``: The port which the 5-tuple filter assigned on.
2038 * ``dst_address``: Destination IP address.
2040 * ``src_address``: Source IP address.
2042 * ``dst_port_value``: TCP/UDP destination port.
2044 * ``src_port_value``: TCP/UDP source port.
2046 * ``protocol_value``: L4 protocol.
2048 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
2050 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
2052 * ``prio_value``: The priority of this filter.
2054 * ``queue_id``: The receive queue associated with this 5-tuple filter.
2056 Example, to add/remove an 5tuple filter rule::
2058 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2059 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2060 flags 0x0 priority 3 queue 3
2062 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2063 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2064 flags 0x0 priority 3 queue 3
2069 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
2071 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
2073 The available information parameters are:
2075 * ``port_id``: The port which the SYN filter assigned on.
2077 * ``high``: This SYN filter has higher priority than other filters.
2079 * ``low``: This SYN filter has lower priority than other filters.
2081 * ``queue_id``: The receive queue associated with this SYN filter
2085 testpmd> syn_filter 0 add priority high queue 3
2090 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
2091 and routed into one of the receive queues::
2093 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
2094 mask (mask_value) priority (prio_value) queue (queue_id)
2096 The available information parameters are:
2098 * ``port_id``: The port which the Flex filter is assigned on.
2100 * ``len_value``: Filter length in bytes, no greater than 128.
2102 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
2104 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
2106 * ``prio_value``: The priority of this filter.
2108 * ``queue_id``: The receive queue associated with this Flex filter.
2112 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
2113 mask 000C priority 3 queue 3
2115 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
2116 mask 000C priority 3 queue 3
2119 .. _testpmd_flow_director:
2121 flow_director_filter
2122 ~~~~~~~~~~~~~~~~~~~~
2124 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
2126 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
2127 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
2129 * Perfect match filters.
2130 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2131 The masked fields are for IP flow.
2133 * Signature filters.
2134 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
2136 * Perfect-mac-vlan match filters.
2137 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2138 The masked fields are for MAC VLAN flow.
2140 * Perfect-tunnel match filters.
2141 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2142 The masked fields are for tunnel flow.
2144 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
2145 per flow type and the flexible payload.
2147 The Flow Director can also mask out parts of all of these fields so that filters
2148 are only applied to certain fields or parts of the fields.
2150 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
2152 # Commands to add flow director filters of different flow types::
2154 flow_director_filter (port_id) mode IP (add|del|update) \
2155 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
2156 src (src_ip_address) dst (dst_ip_address) \
2157 tos (tos_value) proto (proto_value) ttl (ttl_value) \
2158 vlan (vlan_value) flexbytes (flexbytes_value) \
2159 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
2162 flow_director_filter (port_id) mode IP (add|del|update) \
2163 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
2164 src (src_ip_address) (src_port) \
2165 dst (dst_ip_address) (dst_port) \
2166 tos (tos_value) ttl (ttl_value) \
2167 vlan (vlan_value) flexbytes (flexbytes_value) \
2168 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
2171 flow_director_filter (port_id) mode IP (add|del|update) \
2172 flow (ipv4-sctp|ipv6-sctp) \
2173 src (src_ip_address) (src_port) \
2174 dst (dst_ip_address) (dst_port) \
2175 tos (tos_value) ttl (ttl_value) \
2176 tag (verification_tag) vlan (vlan_value) \
2177 flexbytes (flexbytes_value) (drop|fwd) \
2178 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
2180 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
2181 ether (ethertype) flexbytes (flexbytes_value) \
2182 (drop|fwd) pf|vf(vf_id) queue (queue_id)
2185 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
2186 mac (mac_address) vlan (vlan_value) \
2187 flexbytes (flexbytes_value) (drop|fwd) \
2188 queue (queue_id) fd_id (fd_id_value)
2190 flow_director_filter (port_id) mode Tunnel (add|del|update) \
2191 mac (mac_address) vlan (vlan_value) \
2192 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
2193 flexbytes (flexbytes_value) (drop|fwd) \
2194 queue (queue_id) fd_id (fd_id_value)
2196 For example, to add an ipv4-udp flow type filter::
2198 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
2199 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
2200 fwd pf queue 1 fd_id 1
2202 For example, add an ipv4-other flow type filter::
2204 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
2205 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
2206 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
2211 Flush all flow director filters on a device::
2213 testpmd> flush_flow_director (port_id)
2215 Example, to flush all flow director filter on port 0::
2217 testpmd> flush_flow_director 0
2222 Set flow director's input masks::
2224 flow_director_mask (port_id) mode IP vlan (vlan_value) \
2225 src_mask (ipv4_src) (ipv6_src) (src_port) \
2226 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2228 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2230 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2231 mac (mac_value) tunnel-type (tunnel_type_value) \
2232 tunnel-id (tunnel_id_value)
2234 Example, to set flow director mask on port 0::
2236 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2237 src_mask 255.255.255.255 \
2238 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2239 dst_mask 255.255.255.255 \
2240 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2242 flow_director_flex_mask
2243 ~~~~~~~~~~~~~~~~~~~~~~~
2245 set masks of flow director's flexible payload based on certain flow type::
2247 testpmd> flow_director_flex_mask (port_id) \
2248 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2249 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2250 l2_payload|all) (mask)
2252 Example, to set flow director's flex mask for all flow type on port 0::
2254 testpmd> flow_director_flex_mask 0 flow all \
2255 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2258 flow_director_flex_payload
2259 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2261 Configure flexible payload selection::
2263 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2265 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2267 testpmd> flow_director_flex_payload 0 l4 \
2268 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2270 get_sym_hash_ena_per_port
2271 ~~~~~~~~~~~~~~~~~~~~~~~~~
2273 Get symmetric hash enable configuration per port::
2275 get_sym_hash_ena_per_port (port_id)
2277 For example, to get symmetric hash enable configuration of port 1::
2279 testpmd> get_sym_hash_ena_per_port 1
2281 set_sym_hash_ena_per_port
2282 ~~~~~~~~~~~~~~~~~~~~~~~~~
2284 Set symmetric hash enable configuration per port to enable or disable::
2286 set_sym_hash_ena_per_port (port_id) (enable|disable)
2288 For example, to set symmetric hash enable configuration of port 1 to enable::
2290 testpmd> set_sym_hash_ena_per_port 1 enable
2292 get_hash_global_config
2293 ~~~~~~~~~~~~~~~~~~~~~~
2295 Get the global configurations of hash filters::
2297 get_hash_global_config (port_id)
2299 For example, to get the global configurations of hash filters of port 1::
2301 testpmd> get_hash_global_config 1
2303 set_hash_global_config
2304 ~~~~~~~~~~~~~~~~~~~~~~
2306 Set the global configurations of hash filters::
2308 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2309 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2310 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload) \
2313 For example, to enable simple_xor for flow type of ipv6 on port 2::
2315 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2320 Set the input set for hash::
2322 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2323 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2324 l2_payload) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2325 ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2326 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2327 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2328 fld-8th|none) (select|add)
2330 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2332 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2337 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2338 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2340 Set the input set for flow director::
2342 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2343 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2344 l2_payload) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2345 ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2346 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2347 sctp-dst-port|sctp-veri-tag|none) (select|add)
2349 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2351 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2356 Set different GRE key length for input set::
2358 global_config (port_id) gre-key-len (number in bytes)
2360 For example to set GRE key length for input set to 4 bytes on port 0::
2362 testpmd> global_config 0 gre-key-len 4
2365 .. _testpmd_rte_flow:
2367 Flow rules management
2368 ---------------------
2370 Control of the generic flow API (*rte_flow*) is fully exposed through the
2371 ``flow`` command (validation, creation, destruction, queries and operation
2374 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2375 features simultaneously may cause undefined side-effects and is therefore
2381 Because the ``flow`` command uses dynamic tokens to handle the large number
2382 of possible flow rules combinations, its behavior differs slightly from
2383 other commands, in particular:
2385 - Pressing *?* or the *<tab>* key displays contextual help for the current
2386 token, not that of the entire command.
2388 - Optional and repeated parameters are supported (provided they are listed
2389 in the contextual help).
2391 The first parameter stands for the operation mode. Possible operations and
2392 their general syntax are described below. They are covered in detail in the
2395 - Check whether a flow rule can be created::
2397 flow validate {port_id}
2398 [group {group_id}] [priority {level}] [ingress] [egress]
2399 pattern {item} [/ {item} [...]] / end
2400 actions {action} [/ {action} [...]] / end
2402 - Create a flow rule::
2404 flow create {port_id}
2405 [group {group_id}] [priority {level}] [ingress] [egress]
2406 pattern {item} [/ {item} [...]] / end
2407 actions {action} [/ {action} [...]] / end
2409 - Destroy specific flow rules::
2411 flow destroy {port_id} rule {rule_id} [...]
2413 - Destroy all flow rules::
2415 flow flush {port_id}
2417 - Query an existing flow rule::
2419 flow query {port_id} {rule_id} {action}
2421 - List existing flow rules sorted by priority, filtered by group
2424 flow list {port_id} [group {group_id}] [...]
2426 - Restrict ingress traffic to the defined flow rules::
2428 flow isolate {port_id} {boolean}
2430 Validating flow rules
2431 ~~~~~~~~~~~~~~~~~~~~~
2433 ``flow validate`` reports whether a flow rule would be accepted by the
2434 underlying device in its current state but stops short of creating it. It is
2435 bound to ``rte_flow_validate()``::
2437 flow validate {port_id}
2438 [group {group_id}] [priority {level}] [ingress] [egress]
2439 pattern {item} [/ {item} [...]] / end
2440 actions {action} [/ {action} [...]] / end
2442 If successful, it will show::
2446 Otherwise it will show an error message of the form::
2448 Caught error type [...] ([...]): [...]
2450 This command uses the same parameters as ``flow create``, their format is
2451 described in `Creating flow rules`_.
2453 Check whether redirecting any Ethernet packet received on port 0 to RX queue
2454 index 6 is supported::
2456 testpmd> flow validate 0 ingress pattern eth / end
2457 actions queue index 6 / end
2461 Port 0 does not support TCPv6 rules::
2463 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
2465 Caught error type 9 (specific pattern item): Invalid argument
2471 ``flow create`` validates and creates the specified flow rule. It is bound
2472 to ``rte_flow_create()``::
2474 flow create {port_id}
2475 [group {group_id}] [priority {level}] [ingress] [egress]
2476 pattern {item} [/ {item} [...]] / end
2477 actions {action} [/ {action} [...]] / end
2479 If successful, it will return a flow rule ID usable with other commands::
2481 Flow rule #[...] created
2483 Otherwise it will show an error message of the form::
2485 Caught error type [...] ([...]): [...]
2487 Parameters describe in the following order:
2489 - Attributes (*group*, *priority*, *ingress*, *egress* tokens).
2490 - A matching pattern, starting with the *pattern* token and terminated by an
2492 - Actions, starting with the *actions* token and terminated by an *end*
2495 These translate directly to *rte_flow* objects provided as-is to the
2496 underlying functions.
2498 The shortest valid definition only comprises mandatory tokens::
2500 testpmd> flow create 0 pattern end actions end
2502 Note that PMDs may refuse rules that essentially do nothing such as this
2505 **All unspecified object values are automatically initialized to 0.**
2510 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
2511 specified before the ``pattern`` token.
2513 - ``group {group id}``: priority group.
2514 - ``priority {level}``: priority level within group.
2515 - ``ingress``: rule applies to ingress traffic.
2516 - ``egress``: rule applies to egress traffic.
2518 Each instance of an attribute specified several times overrides the previous
2519 value as shown below (group 4 is used)::
2521 testpmd> flow create 0 group 42 group 24 group 4 [...]
2523 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
2525 While not specifying a direction is an error, some rules may allow both
2528 Most rules affect RX therefore contain the ``ingress`` token::
2530 testpmd> flow create 0 ingress pattern [...]
2535 A matching pattern starts after the ``pattern`` token. It is made of pattern
2536 items and is terminated by a mandatory ``end`` item.
2538 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
2539 rte_flow_item_type``).
2541 The ``/`` token is used as a separator between pattern items as shown
2544 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
2546 Note that protocol items like these must be stacked from lowest to highest
2547 layer to make sense. For instance, the following rule is either invalid or
2548 unlikely to match any packet::
2550 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
2552 More information on these restrictions can be found in the *rte_flow*
2555 Several items support additional specification structures, for example
2556 ``ipv4`` allows specifying source and destination addresses as follows::
2558 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
2559 dst is 10.2.0.0 / end [...]
2561 This rule matches all IPv4 traffic with the specified properties.
2563 In this example, ``src`` and ``dst`` are field names of the underlying
2564 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
2565 in a similar fashion.
2567 The ``is`` token means that the subsequent value must be matched exactly,
2568 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
2569 accordingly. Possible assignment tokens are:
2571 - ``is``: match value perfectly (with full bit-mask).
2572 - ``spec``: match value according to configured bit-mask.
2573 - ``last``: specify upper bound to establish a range.
2574 - ``mask``: specify bit-mask with relevant bits set to one.
2575 - ``prefix``: generate bit-mask from a prefix length.
2577 These yield identical results::
2579 ipv4 src is 10.1.1.1
2583 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
2587 ipv4 src spec 10.1.1.1 src prefix 32
2591 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
2595 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
2597 Inclusive ranges can be defined with ``last``::
2599 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
2601 Note that ``mask`` affects both ``spec`` and ``last``::
2603 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
2604 # matches 10.1.0.0 to 10.2.255.255
2606 Properties can be modified multiple times::
2608 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
2612 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
2617 This section lists supported pattern items and their attributes, if any.
2619 - ``end``: end list of pattern items.
2621 - ``void``: no-op pattern item.
2623 - ``invert``: perform actions when pattern does not match.
2625 - ``any``: match any protocol for the current layer.
2627 - ``num {unsigned}``: number of layers covered.
2629 - ``pf``: match packets addressed to the physical function.
2631 - ``vf``: match packets addressed to a virtual function ID.
2633 - ``id {unsigned}``: destination VF ID.
2635 - ``port``: device-specific physical port index to use.
2637 - ``index {unsigned}``: physical port index.
2639 - ``raw``: match an arbitrary byte string.
2641 - ``relative {boolean}``: look for pattern after the previous item.
2642 - ``search {boolean}``: search pattern from offset (see also limit).
2643 - ``offset {integer}``: absolute or relative offset for pattern.
2644 - ``limit {unsigned}``: search area limit for start of pattern.
2645 - ``pattern {string}``: byte string to look for.
2647 - ``eth``: match Ethernet header.
2649 - ``dst {MAC-48}``: destination MAC.
2650 - ``src {MAC-48}``: source MAC.
2651 - ``type {unsigned}``: EtherType.
2653 - ``vlan``: match 802.1Q/ad VLAN tag.
2655 - ``tpid {unsigned}``: tag protocol identifier.
2656 - ``tci {unsigned}``: tag control information.
2657 - ``pcp {unsigned}``: priority code point.
2658 - ``dei {unsigned}``: drop eligible indicator.
2659 - ``vid {unsigned}``: VLAN identifier.
2661 - ``ipv4``: match IPv4 header.
2663 - ``tos {unsigned}``: type of service.
2664 - ``ttl {unsigned}``: time to live.
2665 - ``proto {unsigned}``: next protocol ID.
2666 - ``src {ipv4 address}``: source address.
2667 - ``dst {ipv4 address}``: destination address.
2669 - ``ipv6``: match IPv6 header.
2671 - ``tc {unsigned}``: traffic class.
2672 - ``flow {unsigned}``: flow label.
2673 - ``proto {unsigned}``: protocol (next header).
2674 - ``hop {unsigned}``: hop limit.
2675 - ``src {ipv6 address}``: source address.
2676 - ``dst {ipv6 address}``: destination address.
2678 - ``icmp``: match ICMP header.
2680 - ``type {unsigned}``: ICMP packet type.
2681 - ``code {unsigned}``: ICMP packet code.
2683 - ``udp``: match UDP header.
2685 - ``src {unsigned}``: UDP source port.
2686 - ``dst {unsigned}``: UDP destination port.
2688 - ``tcp``: match TCP header.
2690 - ``src {unsigned}``: TCP source port.
2691 - ``dst {unsigned}``: TCP destination port.
2693 - ``sctp``: match SCTP header.
2695 - ``src {unsigned}``: SCTP source port.
2696 - ``dst {unsigned}``: SCTP destination port.
2697 - ``tag {unsigned}``: validation tag.
2698 - ``cksum {unsigned}``: checksum.
2700 - ``vxlan``: match VXLAN header.
2702 - ``vni {unsigned}``: VXLAN identifier.
2704 - ``e_tag``: match IEEE 802.1BR E-Tag header.
2706 - ``grp_ecid_b {unsigned}``: GRP and E-CID base.
2708 - ``nvgre``: match NVGRE header.
2710 - ``tni {unsigned}``: virtual subnet ID.
2712 - ``mpls``: match MPLS header.
2714 - ``label {unsigned}``: MPLS label.
2716 - ``gre``: match GRE header.
2718 - ``protocol {unsigned}``: protocol type.
2720 - ``fuzzy``: fuzzy pattern match, expect faster than default.
2722 - ``thresh {unsigned}``: accuracy threshold.
2724 - ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header.
2726 - ``teid {unsigned}``: tunnel endpoint identifier.
2731 A list of actions starts after the ``actions`` token in the same fashion as
2732 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
2733 terminated by a mandatory ``end`` action.
2735 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
2736 rte_flow_action_type``).
2738 Dropping all incoming UDPv4 packets can be expressed as follows::
2740 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2743 Several actions have configurable properties which must be specified when
2744 there is no valid default value. For example, ``queue`` requires a target
2747 This rule redirects incoming UDPv4 traffic to queue index 6::
2749 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2750 actions queue index 6 / end
2752 While this one could be rejected by PMDs (unspecified queue index)::
2754 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2757 As defined by *rte_flow*, the list is not ordered, all actions of a given
2758 rule are performed simultaneously. These are equivalent::
2760 queue index 6 / void / mark id 42 / end
2764 void / mark id 42 / queue index 6 / end
2766 All actions in a list should have different types, otherwise only the last
2767 action of a given type is taken into account::
2769 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
2773 drop / drop / drop / end # drop is performed only once
2777 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
2779 Considering they are performed simultaneously, opposite and overlapping
2780 actions can sometimes be combined when the end result is unambiguous::
2782 drop / queue index 6 / end # drop has no effect
2786 drop / dup index 6 / end # same as above
2790 queue index 6 / rss queues 6 7 8 / end # queue has no effect
2794 drop / passthru / end # drop has no effect
2796 Note that PMDs may still refuse such combinations.
2801 This section lists supported actions and their attributes, if any.
2803 - ``end``: end list of actions.
2805 - ``void``: no-op action.
2807 - ``passthru``: let subsequent rule process matched packets.
2809 - ``mark``: attach 32 bit value to packets.
2811 - ``id {unsigned}``: 32 bit value to return with packets.
2813 - ``flag``: flag packets.
2815 - ``queue``: assign packets to a given queue index.
2817 - ``index {unsigned}``: queue index to use.
2819 - ``drop``: drop packets (note: passthru has priority).
2821 - ``count``: enable counters for this rule.
2823 - ``dup``: duplicate packets to a given queue index.
2825 - ``index {unsigned}``: queue index to duplicate packets to.
2827 - ``rss``: spread packets among several queues.
2829 - ``queues [{unsigned} [...]] end``: queue indices to use.
2831 - ``pf``: redirect packets to physical device function.
2833 - ``vf``: redirect packets to virtual device function.
2835 - ``original {boolean}``: use original VF ID if possible.
2836 - ``id {unsigned}``: VF ID to redirect packets to.
2838 Destroying flow rules
2839 ~~~~~~~~~~~~~~~~~~~~~
2841 ``flow destroy`` destroys one or more rules from their rule ID (as returned
2842 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
2843 times as necessary::
2845 flow destroy {port_id} rule {rule_id} [...]
2847 If successful, it will show::
2849 Flow rule #[...] destroyed
2851 It does not report anything for rule IDs that do not exist. The usual error
2852 message is shown when a rule cannot be destroyed::
2854 Caught error type [...] ([...]): [...]
2856 ``flow flush`` destroys all rules on a device and does not take extra
2857 arguments. It is bound to ``rte_flow_flush()``::
2859 flow flush {port_id}
2861 Any errors are reported as above.
2863 Creating several rules and destroying them::
2865 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2866 actions queue index 2 / end
2867 Flow rule #0 created
2868 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2869 actions queue index 3 / end
2870 Flow rule #1 created
2871 testpmd> flow destroy 0 rule 0 rule 1
2872 Flow rule #1 destroyed
2873 Flow rule #0 destroyed
2876 The same result can be achieved using ``flow flush``::
2878 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2879 actions queue index 2 / end
2880 Flow rule #0 created
2881 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2882 actions queue index 3 / end
2883 Flow rule #1 created
2884 testpmd> flow flush 0
2887 Non-existent rule IDs are ignored::
2889 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2890 actions queue index 2 / end
2891 Flow rule #0 created
2892 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2893 actions queue index 3 / end
2894 Flow rule #1 created
2895 testpmd> flow destroy 0 rule 42 rule 10 rule 2
2897 testpmd> flow destroy 0 rule 0
2898 Flow rule #0 destroyed
2904 ``flow query`` queries a specific action of a flow rule having that
2905 ability. Such actions collect information that can be reported using this
2906 command. It is bound to ``rte_flow_query()``::
2908 flow query {port_id} {rule_id} {action}
2910 If successful, it will display either the retrieved data for known actions
2911 or the following message::
2913 Cannot display result for action type [...] ([...])
2915 Otherwise, it will complain either that the rule does not exist or that some
2918 Flow rule #[...] not found
2922 Caught error type [...] ([...]): [...]
2924 Currently only the ``count`` action is supported. This action reports the
2925 number of packets that hit the flow rule and the total number of bytes. Its
2926 output has the following format::
2929 hits_set: [...] # whether "hits" contains a valid value
2930 bytes_set: [...] # whether "bytes" contains a valid value
2931 hits: [...] # number of packets
2932 bytes: [...] # number of bytes
2934 Querying counters for TCPv6 packets redirected to queue 6::
2936 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
2937 actions queue index 6 / count / end
2938 Flow rule #4 created
2939 testpmd> flow query 0 4 count
2950 ``flow list`` lists existing flow rules sorted by priority and optionally
2951 filtered by group identifiers::
2953 flow list {port_id} [group {group_id}] [...]
2955 This command only fails with the following message if the device does not
2960 Output consists of a header line followed by a short description of each
2961 flow rule, one per line. There is no output at all when no flow rules are
2962 configured on the device::
2964 ID Group Prio Attr Rule
2965 [...] [...] [...] [...] [...]
2967 ``Attr`` column flags:
2969 - ``i`` for ``ingress``.
2970 - ``e`` for ``egress``.
2972 Creating several flow rules and listing them::
2974 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2975 actions queue index 6 / end
2976 Flow rule #0 created
2977 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2978 actions queue index 2 / end
2979 Flow rule #1 created
2980 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
2981 actions rss queues 6 7 8 end / end
2982 Flow rule #2 created
2983 testpmd> flow list 0
2984 ID Group Prio Attr Rule
2985 0 0 0 i- ETH IPV4 => QUEUE
2986 1 0 0 i- ETH IPV6 => QUEUE
2987 2 0 5 i- ETH IPV4 UDP => RSS
2990 Rules are sorted by priority (i.e. group ID first, then priority level)::
2992 testpmd> flow list 1
2993 ID Group Prio Attr Rule
2994 0 0 0 i- ETH => COUNT
2995 6 0 500 i- ETH IPV6 TCP => DROP COUNT
2996 5 0 1000 i- ETH IPV6 ICMP => QUEUE
2997 1 24 0 i- ETH IPV4 UDP => QUEUE
2998 4 24 10 i- ETH IPV4 TCP => DROP
2999 3 24 20 i- ETH IPV4 => DROP
3000 2 24 42 i- ETH IPV4 UDP => QUEUE
3001 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3004 Output can be limited to specific groups::
3006 testpmd> flow list 1 group 0 group 63
3007 ID Group Prio Attr Rule
3008 0 0 0 i- ETH => COUNT
3009 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3010 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3011 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3014 Toggling isolated mode
3015 ~~~~~~~~~~~~~~~~~~~~~~
3017 ``flow isolate`` can be used to tell the underlying PMD that ingress traffic
3018 must only be injected from the defined flow rules; that no default traffic
3019 is expected outside those rules and the driver is free to assign more
3020 resources to handle them. It is bound to ``rte_flow_isolate()``::
3022 flow isolate {port_id} {boolean}
3024 If successful, enabling or disabling isolated mode shows either::
3026 Ingress traffic on port [...]
3027 is now restricted to the defined flow rules
3031 Ingress traffic on port [...]
3032 is not restricted anymore to the defined flow rules
3034 Otherwise, in case of error::
3036 Caught error type [...] ([...]): [...]
3038 Mainly due to its side effects, PMDs supporting this mode may not have the
3039 ability to toggle it more than once without reinitializing affected ports
3040 first (e.g. by exiting testpmd).
3042 Enabling isolated mode::
3044 testpmd> flow isolate 0 true
3045 Ingress traffic on port 0 is now restricted to the defined flow rules
3048 Disabling isolated mode::
3050 testpmd> flow isolate 0 false
3051 Ingress traffic on port 0 is not restricted anymore to the defined flow rules
3054 Sample QinQ flow rules
3055 ~~~~~~~~~~~~~~~~~~~~~~
3057 Before creating QinQ rule(s) the following commands should be issued to enable QinQ::
3059 testpmd> port stop 0
3060 testpmd> vlan set qinq on 0
3062 The above command sets the inner and outer TPID's to 0x8100.
3064 To change the TPID's the following commands should be used::
3066 testpmd> vlan set outer tpid 0xa100 0
3067 testpmd> vlan set inner tpid 0x9100 0
3068 testpmd> port start 0
3070 Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM.
3074 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 /
3075 vlan tci is 456 / end actions vf id 1 / queue index 0 / end
3076 Flow rule #0 validated
3078 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 /
3079 vlan tci is 456 / end actions vf id 123 / queue index 0 / end
3080 Flow rule #0 created
3082 testpmd> flow list 0
3083 ID Group Prio Attr Rule
3084 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3086 Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host.
3090 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 /
3091 vlan tci is 654 / end actions pf / queue index 0 / end
3092 Flow rule #1 validated
3094 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 /
3095 vlan tci is 654 / end actions pf / queue index 1 / end
3096 Flow rule #1 created
3098 testpmd> flow list 0
3099 ID Group Prio Attr Rule
3100 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3101 1 0 0 i- ETH VLAN VLAN=>PF QUEUE