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> set port <port_id> gro on|off
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, all merged packets have bad
926 checksums, since the GRO library doesn't re-calculate checksums for
927 the merged packets. Therefore, if users want the merged packets to
928 have correct checksums, please select HW IP checksum calculation and
929 HW TCP checksum calculation for the port which the merged packets are
935 Display GRO configuration for a given port::
937 testpmd> show port <port_id> gro
942 Set the cycle to flush the GROed packets from reassembly tables::
944 testpmd> set gro flush <cycles>
946 When enable GRO, the csum forwarding engine performs GRO on received
947 packets, and the GROed packets are stored in reassembly tables. Users
948 can use this command to determine when the GROed packets are flushed
949 from the reassembly tables.
951 The ``cycles`` is measured in GRO operation times. The csum forwarding
952 engine flushes the GROed packets from the tables every ``cycles`` GRO
955 By default, the value of ``cycles`` is 1, which means flush GROed packets
956 from the reassembly tables as soon as one GRO operation finishes. The value
957 of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``.
959 Please note that the large value of ``cycles`` may cause the poor TCP/IP
960 stack performance. Because the GROed packets are delayed to arrive the
961 stack, thus causing more duplicated ACKs and TCP retransmissions.
966 Toggle per-port GSO support in ``csum`` forwarding engine::
968 testpmd> set port <port_id> gso on|off
970 If enabled, the csum forwarding engine will perform GSO on supported IPv4
971 packets, transmitted on the given port.
973 If disabled, packets transmitted on the given port will not undergo GSO.
974 By default, GSO is disabled for all ports.
978 When GSO is enabled on a port, supported IPv4 packets transmitted on that
979 port undergo GSO. Afterwards, the segmented packets are represented by
980 multi-segment mbufs; however, the csum forwarding engine doesn't calculation
981 of checksums for GSO'd segments in SW. As a result, if users want correct
982 checksums in GSO segments, they should enable HW checksum calculation for
985 For example, HW checksum calculation for VxLAN GSO'd packets may be enabled
986 by setting the following options in the csum forwarding engine:
988 testpmd> csum set outer_ip hw <port_id>
990 testpmd> csum set ip hw <port_id>
992 testpmd> csum set tcp hw <port_id>
997 Set the maximum GSO segment size (measured in bytes), which includes the
998 packet header and the packet payload for GSO-enabled ports (global)::
1000 testpmd> set gso segsz <length>
1005 Display the status of Generic Segmentation Offload for a given port::
1007 testpmd> show port <port_id> gso
1012 Add an alternative MAC address to a port::
1014 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX)
1019 Remove a MAC address from a port::
1021 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX)
1023 mac_addr add (for VF)
1024 ~~~~~~~~~~~~~~~~~~~~~
1026 Add an alternative MAC address for a VF to a port::
1028 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX)
1033 Set the default MAC address for a port::
1035 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX)
1037 mac_addr set (for VF)
1038 ~~~~~~~~~~~~~~~~~~~~~
1040 Set the MAC address for a VF from the PF::
1042 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX)
1047 Set the unicast hash filter(s) on/off for a port::
1049 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
1054 Set the promiscuous mode on for a port or for all ports.
1055 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1057 testpmd> set promisc (port_id|all) (on|off)
1062 Set the allmulti mode for a port or for all ports::
1064 testpmd> set allmulti (port_id|all) (on|off)
1066 Same as the ifconfig (8) option. Controls how multicast packets are handled.
1068 set promisc (for VF)
1069 ~~~~~~~~~~~~~~~~~~~~
1071 Set the unicast promiscuous mode for a VF from PF.
1072 It's supported by Intel i40e NICs now.
1073 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1075 testpmd> set vf promisc (port_id) (vf_id) (on|off)
1077 set allmulticast (for VF)
1078 ~~~~~~~~~~~~~~~~~~~~~~~~~
1080 Set the multicast promiscuous mode for a VF from PF.
1081 It's supported by Intel i40e NICs now.
1082 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1084 testpmd> set vf allmulti (port_id) (vf_id) (on|off)
1086 set tx max bandwidth (for VF)
1087 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1089 Set TX max absolute bandwidth (Mbps) for a VF from PF::
1091 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth)
1093 set tc tx min bandwidth (for VF)
1094 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1096 Set all TCs' TX min relative bandwidth (%) for a VF from PF::
1098 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...)
1100 set tc tx max bandwidth (for VF)
1101 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1103 Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF::
1105 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth)
1107 set tc strict link priority mode
1108 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1110 Set some TCs' strict link priority mode on a physical port::
1112 testpmd> set tx strict-link-priority (port_id) (tc_bitmap)
1114 set tc tx min bandwidth
1115 ~~~~~~~~~~~~~~~~~~~~~~~
1117 Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs::
1119 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...)
1124 Set the link flow control parameter on a port::
1126 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1127 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
1128 autoneg (on|off) (port_id)
1132 * ``high_water`` (integer): High threshold value to trigger XOFF.
1134 * ``low_water`` (integer): Low threshold value to trigger XON.
1136 * ``pause_time`` (integer): Pause quota in the Pause frame.
1138 * ``send_xon`` (0/1): Send XON frame.
1140 * ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
1142 * ``autoneg``: Change the auto-negotiation parameter.
1147 Set the priority flow control parameter on a port::
1149 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1150 (pause_time) (priority) (port_id)
1154 * ``high_water`` (integer): High threshold value.
1156 * ``low_water`` (integer): Low threshold value.
1158 * ``pause_time`` (integer): Pause quota in the Pause frame.
1160 * ``priority`` (0-7): VLAN User Priority.
1165 Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
1167 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
1169 For example, to set rx queue 2 on port 0 to mapping 5::
1171 testpmd>set stat_qmap rx 0 2 5
1173 set port - rx/tx (for VF)
1174 ~~~~~~~~~~~~~~~~~~~~~~~~~
1176 Set VF receive/transmit from a port::
1178 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
1180 set port - mac address filter (for VF)
1181 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1183 Add/Remove unicast or multicast MAC addr filter for a VF::
1185 testpmd> set port (port_id) vf (vf_id) (mac_addr) \
1186 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
1188 set port - rx mode(for VF)
1189 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1191 Set the VF receive mode of a port::
1193 testpmd> set port (port_id) vf (vf_id) \
1194 rxmode (AUPE|ROPE|BAM|MPE) (on|off)
1196 The available receive modes are:
1198 * ``AUPE``: Accepts untagged VLAN.
1200 * ``ROPE``: Accepts unicast hash.
1202 * ``BAM``: Accepts broadcast packets.
1204 * ``MPE``: Accepts all multicast packets.
1206 set port - tx_rate (for Queue)
1207 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1209 Set TX rate limitation for a queue on a port::
1211 testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1213 set port - tx_rate (for VF)
1214 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
1216 Set TX rate limitation for queues in VF on a port::
1218 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1220 set port - mirror rule
1221 ~~~~~~~~~~~~~~~~~~~~~~
1223 Set pool or vlan type mirror rule for a port::
1225 testpmd> set port (port_id) mirror-rule (rule_id) \
1226 (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1227 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1229 Set link mirror rule for a port::
1231 testpmd> set port (port_id) mirror-rule (rule_id) \
1232 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1234 For example to enable mirror traffic with vlan 0,1 to pool 0::
1236 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1238 reset port - mirror rule
1239 ~~~~~~~~~~~~~~~~~~~~~~~~
1241 Reset a mirror rule for a port::
1243 testpmd> reset port (port_id) mirror-rule (rule_id)
1248 Set the flush on RX streams before forwarding.
1249 The default is flush ``on``.
1250 Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1252 testpmd> set flush_rx off
1257 Set the bypass mode for the lowest port on bypass enabled NIC::
1259 testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1264 Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1266 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1267 mode (normal|bypass|isolate) (port_id)
1271 * ``timeout``: Enable bypass after watchdog timeout.
1273 * ``os_on``: Enable bypass when OS/board is powered on.
1275 * ``os_off``: Enable bypass when OS/board is powered off.
1277 * ``power_on``: Enable bypass when power supply is turned on.
1279 * ``power_off``: Enable bypass when power supply is turned off.
1285 Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1287 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1292 Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1294 testpmd> show bypass config (port_id)
1299 Set link up for a port::
1301 testpmd> set link-up port (port id)
1306 Set link down for a port::
1308 testpmd> set link-down port (port id)
1313 Enable E-tag insertion for a VF on a port::
1315 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1317 Disable E-tag insertion for a VF on a port::
1319 testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1321 Enable/disable E-tag stripping on a port::
1323 testpmd> E-tag set stripping (on|off) port (port_id)
1325 Enable/disable E-tag based forwarding on a port::
1327 testpmd> E-tag set forwarding (on|off) port (port_id)
1329 Add an E-tag forwarding filter on a port::
1331 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1333 Delete an E-tag forwarding filter on a port::
1334 testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1339 Load a dynamic device personalization (DDP) package::
1341 testpmd> ddp add (port_id) (package_path[,output_path])
1346 Delete a dynamic device personalization package::
1348 testpmd> ddp del (port_id) (package_path)
1353 List all items from the ptype mapping table::
1355 testpmd> ptype mapping get (port_id) (valid_only)
1359 * ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0).
1361 Replace a specific or a group of software defined ptype with a new one::
1363 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type)
1367 * ``target``: A specific software ptype or a mask to represent a group of software ptypes.
1369 * ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1).
1371 * ``pkt_type``: The new software ptype to replace the old ones.
1373 Update hardware defined ptype to software defined packet type mapping table::
1375 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype)
1379 * ``hw_ptype``: hardware ptype as the index of the ptype mapping table.
1381 * ``sw_ptype``: software ptype as the value of the ptype mapping table.
1383 Reset ptype mapping table::
1385 testpmd> ptype mapping reset (port_id)
1390 The following sections show functions for configuring ports.
1394 Port configuration changes only become active when forwarding is started/restarted.
1399 Attach a port specified by pci address or virtual device args::
1401 testpmd> port attach (identifier)
1403 To attach a new pci device, the device should be recognized by kernel first.
1404 Then it should be moved under DPDK management.
1405 Finally the port can be attached to testpmd.
1407 For example, to move a pci device using ixgbe under DPDK management:
1409 .. code-block:: console
1411 # Check the status of the available devices.
1412 ./usertools/dpdk-devbind.py --status
1414 Network devices using DPDK-compatible driver
1415 ============================================
1418 Network devices using kernel driver
1419 ===================================
1420 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1423 # Bind the device to igb_uio.
1424 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1427 # Recheck the status of the devices.
1428 ./usertools/dpdk-devbind.py --status
1429 Network devices using DPDK-compatible driver
1430 ============================================
1431 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1433 To attach a port created by virtual device, above steps are not needed.
1435 For example, to attach a port whose pci address is 0000:0a:00.0.
1437 .. code-block:: console
1439 testpmd> port attach 0000:0a:00.0
1440 Attaching a new port...
1441 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1442 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1443 EAL: PCI memory mapped at 0x7f83bfa00000
1444 EAL: PCI memory mapped at 0x7f83bfa80000
1445 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1446 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1447 Port 0 is attached. Now total ports is 1
1450 For example, to attach a port created by pcap PMD.
1452 .. code-block:: console
1454 testpmd> port attach net_pcap0
1455 Attaching a new port...
1456 PMD: Initializing pmd_pcap for net_pcap0
1457 PMD: Creating pcap-backed ethdev on numa socket 0
1458 Port 0 is attached. Now total ports is 1
1461 In this case, identifier is ``net_pcap0``.
1462 This identifier format is the same as ``--vdev`` format of DPDK applications.
1464 For example, to re-attach a bonded port which has been previously detached,
1465 the mode and slave parameters must be given.
1467 .. code-block:: console
1469 testpmd> port attach net_bond_0,mode=0,slave=1
1470 Attaching a new port...
1471 EAL: Initializing pmd_bond for net_bond_0
1472 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1473 Port 0 is attached. Now total ports is 1
1480 Detach a specific port::
1482 testpmd> port detach (port_id)
1484 Before detaching a port, the port should be stopped and closed.
1486 For example, to detach a pci device port 0.
1488 .. code-block:: console
1490 testpmd> port stop 0
1493 testpmd> port close 0
1497 testpmd> port detach 0
1499 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1500 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1501 EAL: PCI memory unmapped at 0x7f83bfa00000
1502 EAL: PCI memory unmapped at 0x7f83bfa80000
1506 For example, to detach a virtual device port 0.
1508 .. code-block:: console
1510 testpmd> port stop 0
1513 testpmd> port close 0
1517 testpmd> port detach 0
1519 PMD: Closing pcap ethdev on numa socket 0
1520 Port 'net_pcap0' is detached. Now total ports is 0
1523 To remove a pci device completely from the system, first detach the port from testpmd.
1524 Then the device should be moved under kernel management.
1525 Finally the device can be removed using kernel pci hotplug functionality.
1527 For example, to move a pci device under kernel management:
1529 .. code-block:: console
1531 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1533 ./usertools/dpdk-devbind.py --status
1535 Network devices using DPDK-compatible driver
1536 ============================================
1539 Network devices using kernel driver
1540 ===================================
1541 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1543 To remove a port created by a virtual device, above steps are not needed.
1548 Start all ports or a specific port::
1550 testpmd> port start (port_id|all)
1555 Stop all ports or a specific port::
1557 testpmd> port stop (port_id|all)
1562 Close all ports or a specific port::
1564 testpmd> port close (port_id|all)
1566 port start/stop queue
1567 ~~~~~~~~~~~~~~~~~~~~~
1569 Start/stop a rx/tx queue on a specific port::
1571 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1573 Only take effect when port is started.
1578 Set the speed and duplex mode for all ports or a specific port::
1580 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1581 duplex (half|full|auto)
1583 port config - queues/descriptors
1584 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1586 Set number of queues/descriptors for rxq, txq, rxd and txd::
1588 testpmd> port config all (rxq|txq|rxd|txd) (value)
1590 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1592 port config - max-pkt-len
1593 ~~~~~~~~~~~~~~~~~~~~~~~~~
1595 Set the maximum packet length::
1597 testpmd> port config all max-pkt-len (value)
1599 This is equivalent to the ``--max-pkt-len`` command-line option.
1601 port config - CRC Strip
1602 ~~~~~~~~~~~~~~~~~~~~~~~
1604 Set hardware CRC stripping on or off for all ports::
1606 testpmd> port config all crc-strip (on|off)
1608 CRC stripping is on by default.
1610 The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option.
1612 port config - scatter
1613 ~~~~~~~~~~~~~~~~~~~~~~~
1615 Set RX scatter mode on or off for all ports::
1617 testpmd> port config all scatter (on|off)
1619 RX scatter mode is off by default.
1621 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1623 port config - TX queue flags
1624 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1626 Set a hexadecimal bitmap of TX queue flags for all ports::
1628 testpmd> port config all txqflags value
1630 This command is equivalent to the ``--txqflags`` command-line option.
1632 port config - RX Checksum
1633 ~~~~~~~~~~~~~~~~~~~~~~~~~
1635 Set hardware RX checksum offload to on or off for all ports::
1637 testpmd> port config all rx-cksum (on|off)
1639 Checksum offload is off by default.
1641 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1646 Set hardware VLAN on or off for all ports::
1648 testpmd> port config all hw-vlan (on|off)
1650 Hardware VLAN is on by default.
1652 The ``off`` option is equivalent to the ``--disable-hw-vlan`` command-line option.
1654 port config - VLAN filter
1655 ~~~~~~~~~~~~~~~~~~~~~~~~~
1657 Set hardware VLAN filter on or off for all ports::
1659 testpmd> port config all hw-vlan-filter (on|off)
1661 Hardware VLAN filter is on by default.
1663 The ``off`` option is equivalent to the ``--disable-hw-vlan-filter`` command-line option.
1665 port config - VLAN strip
1666 ~~~~~~~~~~~~~~~~~~~~~~~~
1668 Set hardware VLAN strip on or off for all ports::
1670 testpmd> port config all hw-vlan-strip (on|off)
1672 Hardware VLAN strip is on by default.
1674 The ``off`` option is equivalent to the ``--disable-hw-vlan-strip`` command-line option.
1676 port config - VLAN extend
1677 ~~~~~~~~~~~~~~~~~~~~~~~~~
1679 Set hardware VLAN extend on or off for all ports::
1681 testpmd> port config all hw-vlan-extend (on|off)
1683 Hardware VLAN extend is off by default.
1685 The ``off`` option is equivalent to the ``--disable-hw-vlan-extend`` command-line option.
1687 port config - Drop Packets
1688 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1690 Set packet drop for packets with no descriptors on or off for all ports::
1692 testpmd> port config all drop-en (on|off)
1694 Packet dropping for packets with no descriptors is off by default.
1696 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1701 Set the RSS (Receive Side Scaling) mode on or off::
1703 testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1705 RSS is on by default.
1707 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1709 port config - RSS Reta
1710 ~~~~~~~~~~~~~~~~~~~~~~
1712 Set the RSS (Receive Side Scaling) redirection table::
1714 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1719 Set the DCB mode for an individual port::
1721 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1723 The traffic class should be 4 or 8.
1728 Set the number of packets per burst::
1730 testpmd> port config all burst (value)
1732 This is equivalent to the ``--burst`` command-line option.
1734 port config - Threshold
1735 ~~~~~~~~~~~~~~~~~~~~~~~
1737 Set thresholds for TX/RX queues::
1739 testpmd> port config all (threshold) (value)
1741 Where the threshold type can be:
1743 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1745 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1747 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1749 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1751 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1753 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1755 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1757 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1759 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1761 These threshold options are also available from the command-line.
1766 Set the value of ether-type for E-tag::
1768 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1770 Enable/disable the E-tag support::
1772 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1774 port config pctype mapping
1775 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1777 Reset pctype mapping table::
1779 testpmd> port config (port_id) pctype mapping reset
1781 Update hardware defined pctype to software defined flow type mapping table::
1783 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id)
1787 * ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table.
1789 * ``flow_type_id``: software flow type id as the index of the pctype mapping table.
1792 Link Bonding Functions
1793 ----------------------
1795 The Link Bonding functions make it possible to dynamically create and
1796 manage link bonding devices from within testpmd interactive prompt.
1798 create bonded device
1799 ~~~~~~~~~~~~~~~~~~~~
1801 Create a new bonding device::
1803 testpmd> create bonded device (mode) (socket)
1805 For example, to create a bonded device in mode 1 on socket 0::
1807 testpmd> create bonded 1 0
1808 created new bonded device (port X)
1813 Adds Ethernet device to a Link Bonding device::
1815 testpmd> add bonding slave (slave id) (port id)
1817 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
1819 testpmd> add bonding slave 6 10
1822 remove bonding slave
1823 ~~~~~~~~~~~~~~~~~~~~
1825 Removes an Ethernet slave device from a Link Bonding device::
1827 testpmd> remove bonding slave (slave id) (port id)
1829 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
1831 testpmd> remove bonding slave 6 10
1836 Set the Link Bonding mode of a Link Bonding device::
1838 testpmd> set bonding mode (value) (port id)
1840 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
1842 testpmd> set bonding mode 3 10
1847 Set an Ethernet slave device as the primary device on a Link Bonding device::
1849 testpmd> set bonding primary (slave id) (port id)
1851 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
1853 testpmd> set bonding primary 6 10
1858 Set the MAC address of a Link Bonding device::
1860 testpmd> set bonding mac (port id) (mac)
1862 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
1864 testpmd> set bonding mac 10 00:00:00:00:00:01
1866 set bonding xmit_balance_policy
1867 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1869 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
1871 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
1873 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
1875 testpmd> set bonding xmit_balance_policy 10 l34
1878 set bonding mon_period
1879 ~~~~~~~~~~~~~~~~~~~~~~
1881 Set the link status monitoring polling period in milliseconds for a bonding device.
1883 This adds support for PMD slave devices which do not support link status interrupts.
1884 When the mon_period is set to a value greater than 0 then all PMD's which do not support
1885 link status ISR will be queried every polling interval to check if their link status has changed::
1887 testpmd> set bonding mon_period (port_id) (value)
1889 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
1891 testpmd> set bonding mon_period 5 150
1894 set bonding lacp dedicated_queue
1895 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1897 Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic
1898 when in mode 4 (link-aggregration-802.3ad)::
1900 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable)
1903 set bonding agg_mode
1904 ~~~~~~~~~~~~~~~~~~~~
1906 Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad)::
1908 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable)
1914 Show the current configuration of a Link Bonding device::
1916 testpmd> show bonding config (port id)
1919 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
1920 in balance mode with a transmission policy of layer 2+3::
1922 testpmd> show bonding config 9
1924 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
1926 Active Slaves (3): [1 3 4]
1933 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
1934 This is mainly useful for debugging purposes.
1935 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
1936 and fields that can be accessed.
1941 Display the value of a port register::
1943 testpmd> read reg (port_id) (address)
1945 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
1947 testpmd> read reg 0 0xEE00
1948 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
1953 Display a port register bit field::
1955 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
1957 For example, reading the lowest two bits from the register in the example above::
1959 testpmd> read regfield 0 0xEE00 0 1
1960 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
1965 Display a single port register bit::
1967 testpmd> read regbit (port_id) (address) (bit_x)
1969 For example, reading the lowest bit from the register in the example above::
1971 testpmd> read regbit 0 0xEE00 0
1972 port 0 PCI register at offset 0xEE00: bit 0=1
1977 Set the value of a port register::
1979 testpmd> write reg (port_id) (address) (value)
1981 For example, to clear a register::
1983 testpmd> write reg 0 0xEE00 0x0
1984 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
1989 Set bit field of a port register::
1991 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
1993 For example, writing to the register cleared in the example above::
1995 testpmd> write regfield 0 0xEE00 0 1 2
1996 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
2001 Set single bit value of a port register::
2003 testpmd> write regbit (port_id) (address) (bit_x) (value)
2005 For example, to set the high bit in the register from the example above::
2007 testpmd> write regbit 0 0xEE00 31 1
2008 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
2014 This section details the available filter functions that are available.
2016 Note these functions interface the deprecated legacy filtering framework,
2017 superseded by *rte_flow*. See `Flow rules management`_.
2020 ~~~~~~~~~~~~~~~~~~~~
2022 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
2024 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
2025 ethertype (ether_type) (drop|fwd) queue (queue_id)
2027 The available information parameters are:
2029 * ``port_id``: The port which the Ethertype filter assigned on.
2031 * ``mac_addr``: Compare destination mac address.
2033 * ``mac_ignr``: Ignore destination mac address match.
2035 * ``mac_address``: Destination mac address to match.
2037 * ``ether_type``: The EtherType value want to match,
2038 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
2040 * ``queue_id``: The receive queue associated with this EtherType filter.
2041 It is meaningless when deleting or dropping.
2043 Example, to add/remove an ethertype filter rule::
2045 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
2046 ethertype 0x0806 fwd queue 3
2048 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
2049 ethertype 0x0806 fwd queue 3
2054 Add or delete a 2-tuple filter,
2055 which identifies packets by specific protocol and destination TCP/UDP port
2056 and forwards packets into one of the receive queues::
2058 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
2059 protocol (protocol_value) mask (mask_value) \
2060 tcp_flags (tcp_flags_value) priority (prio_value) \
2063 The available information parameters are:
2065 * ``port_id``: The port which the 2-tuple filter assigned on.
2067 * ``dst_port_value``: Destination port in L4.
2069 * ``protocol_value``: IP L4 protocol.
2071 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
2073 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
2075 * ``prio_value``: Priority of this filter.
2077 * ``queue_id``: The receive queue associated with this 2-tuple filter.
2079 Example, to add/remove an 2tuple filter rule::
2081 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
2082 tcp_flags 0x02 priority 3 queue 3
2084 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
2085 tcp_flags 0x02 priority 3 queue 3
2090 Add or delete a 5-tuple filter,
2091 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
2092 and routes packets into one of the receive queues::
2094 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
2095 (src_address) dst_port (dst_port_value) \
2096 src_port (src_port_value) protocol (protocol_value) \
2097 mask (mask_value) tcp_flags (tcp_flags_value) \
2098 priority (prio_value) queue (queue_id)
2100 The available information parameters are:
2102 * ``port_id``: The port which the 5-tuple filter assigned on.
2104 * ``dst_address``: Destination IP address.
2106 * ``src_address``: Source IP address.
2108 * ``dst_port_value``: TCP/UDP destination port.
2110 * ``src_port_value``: TCP/UDP source port.
2112 * ``protocol_value``: L4 protocol.
2114 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
2116 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
2118 * ``prio_value``: The priority of this filter.
2120 * ``queue_id``: The receive queue associated with this 5-tuple filter.
2122 Example, to add/remove an 5tuple filter rule::
2124 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2125 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2126 flags 0x0 priority 3 queue 3
2128 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2129 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2130 flags 0x0 priority 3 queue 3
2135 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
2137 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
2139 The available information parameters are:
2141 * ``port_id``: The port which the SYN filter assigned on.
2143 * ``high``: This SYN filter has higher priority than other filters.
2145 * ``low``: This SYN filter has lower priority than other filters.
2147 * ``queue_id``: The receive queue associated with this SYN filter
2151 testpmd> syn_filter 0 add priority high queue 3
2156 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
2157 and routed into one of the receive queues::
2159 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
2160 mask (mask_value) priority (prio_value) queue (queue_id)
2162 The available information parameters are:
2164 * ``port_id``: The port which the Flex filter is assigned on.
2166 * ``len_value``: Filter length in bytes, no greater than 128.
2168 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
2170 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
2172 * ``prio_value``: The priority of this filter.
2174 * ``queue_id``: The receive queue associated with this Flex filter.
2178 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
2179 mask 000C priority 3 queue 3
2181 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
2182 mask 000C priority 3 queue 3
2185 .. _testpmd_flow_director:
2187 flow_director_filter
2188 ~~~~~~~~~~~~~~~~~~~~
2190 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
2192 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
2193 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
2195 * Perfect match filters.
2196 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2197 The masked fields are for IP flow.
2199 * Signature filters.
2200 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
2202 * Perfect-mac-vlan match filters.
2203 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2204 The masked fields are for MAC VLAN flow.
2206 * Perfect-tunnel match filters.
2207 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2208 The masked fields are for tunnel flow.
2210 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
2211 per flow type and the flexible payload.
2213 The Flow Director can also mask out parts of all of these fields so that filters
2214 are only applied to certain fields or parts of the fields.
2216 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
2218 # Commands to add flow director filters of different flow types::
2220 flow_director_filter (port_id) mode IP (add|del|update) \
2221 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
2222 src (src_ip_address) dst (dst_ip_address) \
2223 tos (tos_value) proto (proto_value) ttl (ttl_value) \
2224 vlan (vlan_value) flexbytes (flexbytes_value) \
2225 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
2228 flow_director_filter (port_id) mode IP (add|del|update) \
2229 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
2230 src (src_ip_address) (src_port) \
2231 dst (dst_ip_address) (dst_port) \
2232 tos (tos_value) ttl (ttl_value) \
2233 vlan (vlan_value) flexbytes (flexbytes_value) \
2234 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
2237 flow_director_filter (port_id) mode IP (add|del|update) \
2238 flow (ipv4-sctp|ipv6-sctp) \
2239 src (src_ip_address) (src_port) \
2240 dst (dst_ip_address) (dst_port) \
2241 tos (tos_value) ttl (ttl_value) \
2242 tag (verification_tag) vlan (vlan_value) \
2243 flexbytes (flexbytes_value) (drop|fwd) \
2244 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
2246 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
2247 ether (ethertype) flexbytes (flexbytes_value) \
2248 (drop|fwd) pf|vf(vf_id) queue (queue_id)
2251 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
2252 mac (mac_address) vlan (vlan_value) \
2253 flexbytes (flexbytes_value) (drop|fwd) \
2254 queue (queue_id) fd_id (fd_id_value)
2256 flow_director_filter (port_id) mode Tunnel (add|del|update) \
2257 mac (mac_address) vlan (vlan_value) \
2258 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
2259 flexbytes (flexbytes_value) (drop|fwd) \
2260 queue (queue_id) fd_id (fd_id_value)
2262 For example, to add an ipv4-udp flow type filter::
2264 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
2265 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
2266 fwd pf queue 1 fd_id 1
2268 For example, add an ipv4-other flow type filter::
2270 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
2271 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
2272 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
2277 Flush all flow director filters on a device::
2279 testpmd> flush_flow_director (port_id)
2281 Example, to flush all flow director filter on port 0::
2283 testpmd> flush_flow_director 0
2288 Set flow director's input masks::
2290 flow_director_mask (port_id) mode IP vlan (vlan_value) \
2291 src_mask (ipv4_src) (ipv6_src) (src_port) \
2292 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2294 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2296 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2297 mac (mac_value) tunnel-type (tunnel_type_value) \
2298 tunnel-id (tunnel_id_value)
2300 Example, to set flow director mask on port 0::
2302 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2303 src_mask 255.255.255.255 \
2304 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2305 dst_mask 255.255.255.255 \
2306 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2308 flow_director_flex_mask
2309 ~~~~~~~~~~~~~~~~~~~~~~~
2311 set masks of flow director's flexible payload based on certain flow type::
2313 testpmd> flow_director_flex_mask (port_id) \
2314 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2315 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2316 l2_payload|all) (mask)
2318 Example, to set flow director's flex mask for all flow type on port 0::
2320 testpmd> flow_director_flex_mask 0 flow all \
2321 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2324 flow_director_flex_payload
2325 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2327 Configure flexible payload selection::
2329 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2331 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2333 testpmd> flow_director_flex_payload 0 l4 \
2334 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2336 get_sym_hash_ena_per_port
2337 ~~~~~~~~~~~~~~~~~~~~~~~~~
2339 Get symmetric hash enable configuration per port::
2341 get_sym_hash_ena_per_port (port_id)
2343 For example, to get symmetric hash enable configuration of port 1::
2345 testpmd> get_sym_hash_ena_per_port 1
2347 set_sym_hash_ena_per_port
2348 ~~~~~~~~~~~~~~~~~~~~~~~~~
2350 Set symmetric hash enable configuration per port to enable or disable::
2352 set_sym_hash_ena_per_port (port_id) (enable|disable)
2354 For example, to set symmetric hash enable configuration of port 1 to enable::
2356 testpmd> set_sym_hash_ena_per_port 1 enable
2358 get_hash_global_config
2359 ~~~~~~~~~~~~~~~~~~~~~~
2361 Get the global configurations of hash filters::
2363 get_hash_global_config (port_id)
2365 For example, to get the global configurations of hash filters of port 1::
2367 testpmd> get_hash_global_config 1
2369 set_hash_global_config
2370 ~~~~~~~~~~~~~~~~~~~~~~
2372 Set the global configurations of hash filters::
2374 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2375 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2376 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload) \
2379 For example, to enable simple_xor for flow type of ipv6 on port 2::
2381 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2386 Set the input set for hash::
2388 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2389 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2390 l2_payload) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2391 ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2392 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2393 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2394 fld-8th|none) (select|add)
2396 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2398 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2403 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2404 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2406 Set the input set for flow director::
2408 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2409 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2410 l2_payload) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2411 ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2412 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2413 sctp-dst-port|sctp-veri-tag|none) (select|add)
2415 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2417 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2422 Set different GRE key length for input set::
2424 global_config (port_id) gre-key-len (number in bytes)
2426 For example to set GRE key length for input set to 4 bytes on port 0::
2428 testpmd> global_config 0 gre-key-len 4
2431 .. _testpmd_rte_flow:
2433 Flow rules management
2434 ---------------------
2436 Control of the generic flow API (*rte_flow*) is fully exposed through the
2437 ``flow`` command (validation, creation, destruction, queries and operation
2440 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2441 features simultaneously may cause undefined side-effects and is therefore
2447 Because the ``flow`` command uses dynamic tokens to handle the large number
2448 of possible flow rules combinations, its behavior differs slightly from
2449 other commands, in particular:
2451 - Pressing *?* or the *<tab>* key displays contextual help for the current
2452 token, not that of the entire command.
2454 - Optional and repeated parameters are supported (provided they are listed
2455 in the contextual help).
2457 The first parameter stands for the operation mode. Possible operations and
2458 their general syntax are described below. They are covered in detail in the
2461 - Check whether a flow rule can be created::
2463 flow validate {port_id}
2464 [group {group_id}] [priority {level}] [ingress] [egress]
2465 pattern {item} [/ {item} [...]] / end
2466 actions {action} [/ {action} [...]] / end
2468 - Create a flow rule::
2470 flow create {port_id}
2471 [group {group_id}] [priority {level}] [ingress] [egress]
2472 pattern {item} [/ {item} [...]] / end
2473 actions {action} [/ {action} [...]] / end
2475 - Destroy specific flow rules::
2477 flow destroy {port_id} rule {rule_id} [...]
2479 - Destroy all flow rules::
2481 flow flush {port_id}
2483 - Query an existing flow rule::
2485 flow query {port_id} {rule_id} {action}
2487 - List existing flow rules sorted by priority, filtered by group
2490 flow list {port_id} [group {group_id}] [...]
2492 - Restrict ingress traffic to the defined flow rules::
2494 flow isolate {port_id} {boolean}
2496 Validating flow rules
2497 ~~~~~~~~~~~~~~~~~~~~~
2499 ``flow validate`` reports whether a flow rule would be accepted by the
2500 underlying device in its current state but stops short of creating it. It is
2501 bound to ``rte_flow_validate()``::
2503 flow validate {port_id}
2504 [group {group_id}] [priority {level}] [ingress] [egress]
2505 pattern {item} [/ {item} [...]] / end
2506 actions {action} [/ {action} [...]] / end
2508 If successful, it will show::
2512 Otherwise it will show an error message of the form::
2514 Caught error type [...] ([...]): [...]
2516 This command uses the same parameters as ``flow create``, their format is
2517 described in `Creating flow rules`_.
2519 Check whether redirecting any Ethernet packet received on port 0 to RX queue
2520 index 6 is supported::
2522 testpmd> flow validate 0 ingress pattern eth / end
2523 actions queue index 6 / end
2527 Port 0 does not support TCPv6 rules::
2529 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
2531 Caught error type 9 (specific pattern item): Invalid argument
2537 ``flow create`` validates and creates the specified flow rule. It is bound
2538 to ``rte_flow_create()``::
2540 flow create {port_id}
2541 [group {group_id}] [priority {level}] [ingress] [egress]
2542 pattern {item} [/ {item} [...]] / end
2543 actions {action} [/ {action} [...]] / end
2545 If successful, it will return a flow rule ID usable with other commands::
2547 Flow rule #[...] created
2549 Otherwise it will show an error message of the form::
2551 Caught error type [...] ([...]): [...]
2553 Parameters describe in the following order:
2555 - Attributes (*group*, *priority*, *ingress*, *egress* tokens).
2556 - A matching pattern, starting with the *pattern* token and terminated by an
2558 - Actions, starting with the *actions* token and terminated by an *end*
2561 These translate directly to *rte_flow* objects provided as-is to the
2562 underlying functions.
2564 The shortest valid definition only comprises mandatory tokens::
2566 testpmd> flow create 0 pattern end actions end
2568 Note that PMDs may refuse rules that essentially do nothing such as this
2571 **All unspecified object values are automatically initialized to 0.**
2576 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
2577 specified before the ``pattern`` token.
2579 - ``group {group id}``: priority group.
2580 - ``priority {level}``: priority level within group.
2581 - ``ingress``: rule applies to ingress traffic.
2582 - ``egress``: rule applies to egress traffic.
2584 Each instance of an attribute specified several times overrides the previous
2585 value as shown below (group 4 is used)::
2587 testpmd> flow create 0 group 42 group 24 group 4 [...]
2589 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
2591 While not specifying a direction is an error, some rules may allow both
2594 Most rules affect RX therefore contain the ``ingress`` token::
2596 testpmd> flow create 0 ingress pattern [...]
2601 A matching pattern starts after the ``pattern`` token. It is made of pattern
2602 items and is terminated by a mandatory ``end`` item.
2604 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
2605 rte_flow_item_type``).
2607 The ``/`` token is used as a separator between pattern items as shown
2610 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
2612 Note that protocol items like these must be stacked from lowest to highest
2613 layer to make sense. For instance, the following rule is either invalid or
2614 unlikely to match any packet::
2616 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
2618 More information on these restrictions can be found in the *rte_flow*
2621 Several items support additional specification structures, for example
2622 ``ipv4`` allows specifying source and destination addresses as follows::
2624 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
2625 dst is 10.2.0.0 / end [...]
2627 This rule matches all IPv4 traffic with the specified properties.
2629 In this example, ``src`` and ``dst`` are field names of the underlying
2630 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
2631 in a similar fashion.
2633 The ``is`` token means that the subsequent value must be matched exactly,
2634 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
2635 accordingly. Possible assignment tokens are:
2637 - ``is``: match value perfectly (with full bit-mask).
2638 - ``spec``: match value according to configured bit-mask.
2639 - ``last``: specify upper bound to establish a range.
2640 - ``mask``: specify bit-mask with relevant bits set to one.
2641 - ``prefix``: generate bit-mask from a prefix length.
2643 These yield identical results::
2645 ipv4 src is 10.1.1.1
2649 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
2653 ipv4 src spec 10.1.1.1 src prefix 32
2657 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
2661 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
2663 Inclusive ranges can be defined with ``last``::
2665 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
2667 Note that ``mask`` affects both ``spec`` and ``last``::
2669 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
2670 # matches 10.1.0.0 to 10.2.255.255
2672 Properties can be modified multiple times::
2674 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
2678 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
2683 This section lists supported pattern items and their attributes, if any.
2685 - ``end``: end list of pattern items.
2687 - ``void``: no-op pattern item.
2689 - ``invert``: perform actions when pattern does not match.
2691 - ``any``: match any protocol for the current layer.
2693 - ``num {unsigned}``: number of layers covered.
2695 - ``pf``: match packets addressed to the physical function.
2697 - ``vf``: match packets addressed to a virtual function ID.
2699 - ``id {unsigned}``: destination VF ID.
2701 - ``port``: device-specific physical port index to use.
2703 - ``index {unsigned}``: physical port index.
2705 - ``raw``: match an arbitrary byte string.
2707 - ``relative {boolean}``: look for pattern after the previous item.
2708 - ``search {boolean}``: search pattern from offset (see also limit).
2709 - ``offset {integer}``: absolute or relative offset for pattern.
2710 - ``limit {unsigned}``: search area limit for start of pattern.
2711 - ``pattern {string}``: byte string to look for.
2713 - ``eth``: match Ethernet header.
2715 - ``dst {MAC-48}``: destination MAC.
2716 - ``src {MAC-48}``: source MAC.
2717 - ``type {unsigned}``: EtherType.
2719 - ``vlan``: match 802.1Q/ad VLAN tag.
2721 - ``tpid {unsigned}``: tag protocol identifier.
2722 - ``tci {unsigned}``: tag control information.
2723 - ``pcp {unsigned}``: priority code point.
2724 - ``dei {unsigned}``: drop eligible indicator.
2725 - ``vid {unsigned}``: VLAN identifier.
2727 - ``ipv4``: match IPv4 header.
2729 - ``tos {unsigned}``: type of service.
2730 - ``ttl {unsigned}``: time to live.
2731 - ``proto {unsigned}``: next protocol ID.
2732 - ``src {ipv4 address}``: source address.
2733 - ``dst {ipv4 address}``: destination address.
2735 - ``ipv6``: match IPv6 header.
2737 - ``tc {unsigned}``: traffic class.
2738 - ``flow {unsigned}``: flow label.
2739 - ``proto {unsigned}``: protocol (next header).
2740 - ``hop {unsigned}``: hop limit.
2741 - ``src {ipv6 address}``: source address.
2742 - ``dst {ipv6 address}``: destination address.
2744 - ``icmp``: match ICMP header.
2746 - ``type {unsigned}``: ICMP packet type.
2747 - ``code {unsigned}``: ICMP packet code.
2749 - ``udp``: match UDP header.
2751 - ``src {unsigned}``: UDP source port.
2752 - ``dst {unsigned}``: UDP destination port.
2754 - ``tcp``: match TCP header.
2756 - ``src {unsigned}``: TCP source port.
2757 - ``dst {unsigned}``: TCP destination port.
2759 - ``sctp``: match SCTP header.
2761 - ``src {unsigned}``: SCTP source port.
2762 - ``dst {unsigned}``: SCTP destination port.
2763 - ``tag {unsigned}``: validation tag.
2764 - ``cksum {unsigned}``: checksum.
2766 - ``vxlan``: match VXLAN header.
2768 - ``vni {unsigned}``: VXLAN identifier.
2770 - ``e_tag``: match IEEE 802.1BR E-Tag header.
2772 - ``grp_ecid_b {unsigned}``: GRP and E-CID base.
2774 - ``nvgre``: match NVGRE header.
2776 - ``tni {unsigned}``: virtual subnet ID.
2778 - ``mpls``: match MPLS header.
2780 - ``label {unsigned}``: MPLS label.
2782 - ``gre``: match GRE header.
2784 - ``protocol {unsigned}``: protocol type.
2786 - ``fuzzy``: fuzzy pattern match, expect faster than default.
2788 - ``thresh {unsigned}``: accuracy threshold.
2790 - ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header.
2792 - ``teid {unsigned}``: tunnel endpoint identifier.
2797 A list of actions starts after the ``actions`` token in the same fashion as
2798 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
2799 terminated by a mandatory ``end`` action.
2801 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
2802 rte_flow_action_type``).
2804 Dropping all incoming UDPv4 packets can be expressed as follows::
2806 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2809 Several actions have configurable properties which must be specified when
2810 there is no valid default value. For example, ``queue`` requires a target
2813 This rule redirects incoming UDPv4 traffic to queue index 6::
2815 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2816 actions queue index 6 / end
2818 While this one could be rejected by PMDs (unspecified queue index)::
2820 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2823 As defined by *rte_flow*, the list is not ordered, all actions of a given
2824 rule are performed simultaneously. These are equivalent::
2826 queue index 6 / void / mark id 42 / end
2830 void / mark id 42 / queue index 6 / end
2832 All actions in a list should have different types, otherwise only the last
2833 action of a given type is taken into account::
2835 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
2839 drop / drop / drop / end # drop is performed only once
2843 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
2845 Considering they are performed simultaneously, opposite and overlapping
2846 actions can sometimes be combined when the end result is unambiguous::
2848 drop / queue index 6 / end # drop has no effect
2852 drop / dup index 6 / end # same as above
2856 queue index 6 / rss queues 6 7 8 / end # queue has no effect
2860 drop / passthru / end # drop has no effect
2862 Note that PMDs may still refuse such combinations.
2867 This section lists supported actions and their attributes, if any.
2869 - ``end``: end list of actions.
2871 - ``void``: no-op action.
2873 - ``passthru``: let subsequent rule process matched packets.
2875 - ``mark``: attach 32 bit value to packets.
2877 - ``id {unsigned}``: 32 bit value to return with packets.
2879 - ``flag``: flag packets.
2881 - ``queue``: assign packets to a given queue index.
2883 - ``index {unsigned}``: queue index to use.
2885 - ``drop``: drop packets (note: passthru has priority).
2887 - ``count``: enable counters for this rule.
2889 - ``dup``: duplicate packets to a given queue index.
2891 - ``index {unsigned}``: queue index to duplicate packets to.
2893 - ``rss``: spread packets among several queues.
2895 - ``queues [{unsigned} [...]] end``: queue indices to use.
2897 - ``pf``: redirect packets to physical device function.
2899 - ``vf``: redirect packets to virtual device function.
2901 - ``original {boolean}``: use original VF ID if possible.
2902 - ``id {unsigned}``: VF ID to redirect packets to.
2904 Destroying flow rules
2905 ~~~~~~~~~~~~~~~~~~~~~
2907 ``flow destroy`` destroys one or more rules from their rule ID (as returned
2908 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
2909 times as necessary::
2911 flow destroy {port_id} rule {rule_id} [...]
2913 If successful, it will show::
2915 Flow rule #[...] destroyed
2917 It does not report anything for rule IDs that do not exist. The usual error
2918 message is shown when a rule cannot be destroyed::
2920 Caught error type [...] ([...]): [...]
2922 ``flow flush`` destroys all rules on a device and does not take extra
2923 arguments. It is bound to ``rte_flow_flush()``::
2925 flow flush {port_id}
2927 Any errors are reported as above.
2929 Creating several rules and destroying them::
2931 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2932 actions queue index 2 / end
2933 Flow rule #0 created
2934 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2935 actions queue index 3 / end
2936 Flow rule #1 created
2937 testpmd> flow destroy 0 rule 0 rule 1
2938 Flow rule #1 destroyed
2939 Flow rule #0 destroyed
2942 The same result can be achieved using ``flow flush``::
2944 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2945 actions queue index 2 / end
2946 Flow rule #0 created
2947 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2948 actions queue index 3 / end
2949 Flow rule #1 created
2950 testpmd> flow flush 0
2953 Non-existent rule IDs are ignored::
2955 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2956 actions queue index 2 / end
2957 Flow rule #0 created
2958 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2959 actions queue index 3 / end
2960 Flow rule #1 created
2961 testpmd> flow destroy 0 rule 42 rule 10 rule 2
2963 testpmd> flow destroy 0 rule 0
2964 Flow rule #0 destroyed
2970 ``flow query`` queries a specific action of a flow rule having that
2971 ability. Such actions collect information that can be reported using this
2972 command. It is bound to ``rte_flow_query()``::
2974 flow query {port_id} {rule_id} {action}
2976 If successful, it will display either the retrieved data for known actions
2977 or the following message::
2979 Cannot display result for action type [...] ([...])
2981 Otherwise, it will complain either that the rule does not exist or that some
2984 Flow rule #[...] not found
2988 Caught error type [...] ([...]): [...]
2990 Currently only the ``count`` action is supported. This action reports the
2991 number of packets that hit the flow rule and the total number of bytes. Its
2992 output has the following format::
2995 hits_set: [...] # whether "hits" contains a valid value
2996 bytes_set: [...] # whether "bytes" contains a valid value
2997 hits: [...] # number of packets
2998 bytes: [...] # number of bytes
3000 Querying counters for TCPv6 packets redirected to queue 6::
3002 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
3003 actions queue index 6 / count / end
3004 Flow rule #4 created
3005 testpmd> flow query 0 4 count
3016 ``flow list`` lists existing flow rules sorted by priority and optionally
3017 filtered by group identifiers::
3019 flow list {port_id} [group {group_id}] [...]
3021 This command only fails with the following message if the device does not
3026 Output consists of a header line followed by a short description of each
3027 flow rule, one per line. There is no output at all when no flow rules are
3028 configured on the device::
3030 ID Group Prio Attr Rule
3031 [...] [...] [...] [...] [...]
3033 ``Attr`` column flags:
3035 - ``i`` for ``ingress``.
3036 - ``e`` for ``egress``.
3038 Creating several flow rules and listing them::
3040 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3041 actions queue index 6 / end
3042 Flow rule #0 created
3043 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3044 actions queue index 2 / end
3045 Flow rule #1 created
3046 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
3047 actions rss queues 6 7 8 end / end
3048 Flow rule #2 created
3049 testpmd> flow list 0
3050 ID Group Prio Attr Rule
3051 0 0 0 i- ETH IPV4 => QUEUE
3052 1 0 0 i- ETH IPV6 => QUEUE
3053 2 0 5 i- ETH IPV4 UDP => RSS
3056 Rules are sorted by priority (i.e. group ID first, then priority level)::
3058 testpmd> flow list 1
3059 ID Group Prio Attr Rule
3060 0 0 0 i- ETH => COUNT
3061 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3062 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3063 1 24 0 i- ETH IPV4 UDP => QUEUE
3064 4 24 10 i- ETH IPV4 TCP => DROP
3065 3 24 20 i- ETH IPV4 => DROP
3066 2 24 42 i- ETH IPV4 UDP => QUEUE
3067 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3070 Output can be limited to specific groups::
3072 testpmd> flow list 1 group 0 group 63
3073 ID Group Prio Attr Rule
3074 0 0 0 i- ETH => COUNT
3075 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3076 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3077 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3080 Toggling isolated mode
3081 ~~~~~~~~~~~~~~~~~~~~~~
3083 ``flow isolate`` can be used to tell the underlying PMD that ingress traffic
3084 must only be injected from the defined flow rules; that no default traffic
3085 is expected outside those rules and the driver is free to assign more
3086 resources to handle them. It is bound to ``rte_flow_isolate()``::
3088 flow isolate {port_id} {boolean}
3090 If successful, enabling or disabling isolated mode shows either::
3092 Ingress traffic on port [...]
3093 is now restricted to the defined flow rules
3097 Ingress traffic on port [...]
3098 is not restricted anymore to the defined flow rules
3100 Otherwise, in case of error::
3102 Caught error type [...] ([...]): [...]
3104 Mainly due to its side effects, PMDs supporting this mode may not have the
3105 ability to toggle it more than once without reinitializing affected ports
3106 first (e.g. by exiting testpmd).
3108 Enabling isolated mode::
3110 testpmd> flow isolate 0 true
3111 Ingress traffic on port 0 is now restricted to the defined flow rules
3114 Disabling isolated mode::
3116 testpmd> flow isolate 0 false
3117 Ingress traffic on port 0 is not restricted anymore to the defined flow rules
3120 Sample QinQ flow rules
3121 ~~~~~~~~~~~~~~~~~~~~~~
3123 Before creating QinQ rule(s) the following commands should be issued to enable QinQ::
3125 testpmd> port stop 0
3126 testpmd> vlan set qinq on 0
3128 The above command sets the inner and outer TPID's to 0x8100.
3130 To change the TPID's the following commands should be used::
3132 testpmd> vlan set outer tpid 0xa100 0
3133 testpmd> vlan set inner tpid 0x9100 0
3134 testpmd> port start 0
3136 Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM.
3140 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 /
3141 vlan tci is 456 / end actions vf id 1 / queue index 0 / end
3142 Flow rule #0 validated
3144 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 /
3145 vlan tci is 456 / end actions vf id 123 / queue index 0 / end
3146 Flow rule #0 created
3148 testpmd> flow list 0
3149 ID Group Prio Attr Rule
3150 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3152 Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host.
3156 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 /
3157 vlan tci is 654 / end actions pf / queue index 0 / end
3158 Flow rule #1 validated
3160 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 /
3161 vlan tci is 654 / end actions pf / queue index 1 / end
3162 Flow rule #1 created
3164 testpmd> flow list 0
3165 ID Group Prio Attr Rule
3166 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3167 1 0 0 i- ETH VLAN VLAN=>PF QUEUE