2 Copyright(c) 2010-2016 Intel Corporation. All rights reserved.
5 Redistribution and use in source and binary forms, with or without
6 modification, are permitted provided that the following conditions
9 * Redistributions of source code must retain the above copyright
10 notice, this list of conditions and the following disclaimer.
11 * Redistributions in binary form must reproduce the above copyright
12 notice, this list of conditions and the following disclaimer in
13 the documentation and/or other materials provided with the
15 * Neither the name of Intel Corporation nor the names of its
16 contributors may be used to endorse or promote products derived
17 from this software without specific prior written permission.
19 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
20 "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
21 LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
22 A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
23 OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
24 SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
25 LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
26 DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
27 THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
28 (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
29 OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
33 Testpmd Runtime Functions
34 =========================
36 Where the testpmd application is started in interactive mode, (``-i|--interactive``),
37 it displays a prompt that can be used to start and stop forwarding,
38 configure the application, display statistics (including the extended NIC
39 statistics aka xstats) , set the Flow Director and other tasks::
43 The testpmd prompt has some, limited, readline support.
44 Common bash command-line functions such as ``Ctrl+a`` and ``Ctrl+e`` to go to the start and end of the prompt line are supported
45 as well as access to the command history via the up-arrow.
47 There is also support for tab completion.
48 If you type a partial command and hit ``<TAB>`` you get a list of the available completions:
50 .. code-block:: console
52 testpmd> show port <TAB>
54 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
55 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
56 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
57 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
63 Some examples in this document are too long to fit on one line are are shown wrapped at `"\\"` for display purposes::
65 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
66 (pause_time) (send_xon) (port_id)
68 In the real ``testpmd>`` prompt these commands should be on a single line.
73 The testpmd has on-line help for the functions that are available at runtime.
74 These are divided into sections and can be accessed using help, help section or help all:
76 .. code-block:: console
80 help control : Start and stop forwarding.
81 help display : Displaying port, stats and config information.
82 help config : Configuration information.
83 help ports : Configuring ports.
84 help registers : Reading and setting port registers.
85 help filters : Filters configuration help.
86 help all : All of the above sections.
89 Command File Functions
90 ----------------------
92 To facilitate loading large number of commands or to avoid cutting and pasting where not
93 practical or possible testpmd supports alternative methods for executing commands.
95 * If started with the ``--cmdline-file=FILENAME`` command line argument testpmd
96 will execute all CLI commands contained within the file immediately before
97 starting packet forwarding or entering interactive mode.
99 .. code-block:: console
101 ./testpmd -n4 -r2 ... -- -i --cmdline-file=/home/ubuntu/flow-create-commands.txt
102 Interactive-mode selected
103 CLI commands to be read from /home/ubuntu/flow-create-commands.txt
104 Configuring Port 0 (socket 0)
105 Port 0: 7C:FE:90:CB:74:CE
106 Configuring Port 1 (socket 0)
107 Port 1: 7C:FE:90:CB:74:CA
108 Checking link statuses...
109 Port 0 Link Up - speed 10000 Mbps - full-duplex
110 Port 1 Link Up - speed 10000 Mbps - full-duplex
116 Flow rule #498 created
117 Flow rule #499 created
118 Read all CLI commands from /home/ubuntu/flow-create-commands.txt
122 * At run-time additional commands can be loaded in bulk by invoking the ``load FILENAME``
125 .. code-block:: console
127 testpmd> load /home/ubuntu/flow-create-commands.txt
132 Flow rule #498 created
133 Flow rule #499 created
134 Read all CLI commands from /home/ubuntu/flow-create-commands.txt
138 In all cases output from any included command will be displayed as standard output.
139 Execution will continue until the end of the file is reached regardless of
140 whether any errors occur. The end user must examine the output to determine if
141 any failures occurred.
150 Start packet forwarding with current configuration::
157 Start packet forwarding with current configuration after sending specified number of bursts of packets::
159 testpmd> start tx_first (""|burst_num)
161 The default burst number is 1 when ``burst_num`` not presented.
166 Stop packet forwarding, and display accumulated statistics::
181 The functions in the following sections are used to display information about the
182 testpmd configuration or the NIC status.
187 Display information for a given port or all ports::
189 testpmd> show port (info|stats|xstats|fdir|stat_qmap|dcb_tc|cap) (port_id|all)
191 The available information categories are:
193 * ``info``: General port information such as MAC address.
195 * ``stats``: RX/TX statistics.
197 * ``xstats``: RX/TX extended NIC statistics.
199 * ``fdir``: Flow Director information and statistics.
201 * ``stat_qmap``: Queue statistics mapping.
203 * ``dcb_tc``: DCB information such as TC mapping.
205 * ``cap``: Supported offload capabilities.
209 .. code-block:: console
211 testpmd> show port info 0
213 ********************* Infos for port 0 *********************
215 MAC address: XX:XX:XX:XX:XX:XX
217 memory allocation on the socket: 0
219 Link speed: 40000 Mbps
220 Link duplex: full-duplex
221 Promiscuous mode: enabled
222 Allmulticast mode: disabled
223 Maximum number of MAC addresses: 64
224 Maximum number of MAC addresses of hash filtering: 0
229 Redirection table size: 512
230 Supported flow types:
250 Display the rss redirection table entry indicated by masks on port X::
252 testpmd> show port (port_id) rss reta (size) (mask0, mask1...)
254 size is used to indicate the hardware supported reta size
259 Display the RSS hash functions and RSS hash key of a port::
261 testpmd> show port (port_id) rss-hash ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2-payload|ipv6-ex|ipv6-tcp-ex|ipv6-udp-ex [key]
266 Clear the port statistics for a given port or for all ports::
268 testpmd> clear port (info|stats|xstats|fdir|stat_qmap) (port_id|all)
272 testpmd> clear port stats all
277 Display information for a given port's RX/TX queue::
279 testpmd> show (rxq|txq) info (port_id) (queue_id)
284 Displays the configuration of the application.
285 The configuration comes from the command-line, the runtime or the application defaults::
287 testpmd> show config (rxtx|cores|fwd|txpkts)
289 The available information categories are:
291 * ``rxtx``: RX/TX configuration items.
293 * ``cores``: List of forwarding cores.
295 * ``fwd``: Packet forwarding configuration.
297 * ``txpkts``: Packets to TX configuration.
301 .. code-block:: console
303 testpmd> show config rxtx
305 io packet forwarding - CRC stripping disabled - packets/burst=16
306 nb forwarding cores=2 - nb forwarding ports=1
307 RX queues=1 - RX desc=128 - RX free threshold=0
308 RX threshold registers: pthresh=8 hthresh=8 wthresh=4
309 TX queues=1 - TX desc=512 - TX free threshold=0
310 TX threshold registers: pthresh=36 hthresh=0 wthresh=0
311 TX RS bit threshold=0 - TXQ flags=0x0
316 Set the packet forwarding mode::
318 testpmd> set fwd (io|mac|macswap|flowgen| \
319 rxonly|txonly|csum|icmpecho) (""|retry)
321 ``retry`` can be specified for forwarding engines except ``rx_only``.
323 The available information categories are:
325 * ``io``: Forwards packets "as-is" in I/O mode.
326 This is the fastest possible forwarding operation as it does not access packets data.
327 This is the default mode.
329 * ``mac``: Changes the source and the destination Ethernet addresses of packets before forwarding them.
330 Default application behaviour is to set source Ethernet address to that of the transmitting interface, and destination
331 address to a dummy value (set during init). The user may specify a target destination Ethernet address via the 'eth-peer' or
332 'eth-peer-configfile' command-line options. It is not currently possible to specify a specific source Ethernet address.
334 * ``macswap``: MAC swap forwarding mode.
335 Swaps the source and the destination Ethernet addresses of packets before forwarding them.
337 * ``flowgen``: Multi-flow generation mode.
338 Originates a number of flows (with varying destination IP addresses), and terminate receive traffic.
340 * ``rxonly``: Receives packets but doesn't transmit them.
342 * ``txonly``: Generates and transmits packets without receiving any.
344 * ``csum``: Changes the checksum field with hardware or software methods depending on the offload flags on the packet.
346 * ``icmpecho``: Receives a burst of packets, lookup for IMCP echo requests and, if any, send back ICMP echo replies.
348 * ``ieee1588``: Demonstrate L2 IEEE1588 V2 PTP timestamping for RX and TX. Requires ``CONFIG_RTE_LIBRTE_IEEE1588=y``.
350 * ``tm``: Traffic Management forwarding mode
351 Demonstrates the use of ethdev traffic management APIs and softnic PMD for
352 QoS traffic management. In this mode, 5-level hierarchical QoS scheduler is
353 available as an default option that can be enabled through CLI. The user can
354 also modify the default hierarchy or specify the new hierarchy through CLI for
355 implementing QoS scheduler. Requires ``CONFIG_RTE_LIBRTE_PMD_SOFTNIC=y`` ``CONFIG_RTE_LIBRTE_SCHED=y``.
357 Note: TX timestamping is only available in the "Full Featured" TX path. To force ``testpmd`` into this mode set ``--txqflags=0``.
361 testpmd> set fwd rxonly
363 Set rxonly packet forwarding mode
369 Display an RX descriptor for a port RX queue::
371 testpmd> read rxd (port_id) (queue_id) (rxd_id)
375 testpmd> read rxd 0 0 4
376 0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180
381 Display a TX descriptor for a port TX queue::
383 testpmd> read txd (port_id) (queue_id) (txd_id)
387 testpmd> read txd 0 0 4
388 0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C
393 Get loaded dynamic device personalization (DDP) package info list::
395 testpmd> ddp get list (port_id)
400 Display information about dynamic device personalization (DDP) profile::
402 testpmd> ddp get info (profile_path)
407 Display VF statistics::
409 testpmd> show vf stats (port_id) (vf_id)
414 Reset VF statistics::
416 testpmd> clear vf stats (port_id) (vf_id)
418 show port pctype mapping
419 ~~~~~~~~~~~~~~~~~~~~~~~~
421 List all items from the pctype mapping table::
423 testpmd> show port (port_id) pctype mapping
426 Configuration Functions
427 -----------------------
429 The testpmd application can be configured from the runtime as well as from the command-line.
431 This section details the available configuration functions that are available.
435 Configuration changes only become active when forwarding is started/restarted.
440 Reset forwarding to the default configuration::
447 Set the debug verbosity level::
449 testpmd> set verbose (level)
451 Currently the only available levels are 0 (silent except for error) and 1 (fully verbose).
456 Set the number of ports used by the application:
460 This is equivalent to the ``--nb-ports`` command-line option.
465 Set the number of cores used by the application::
467 testpmd> set nbcore (num)
469 This is equivalent to the ``--nb-cores`` command-line option.
473 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port.
478 Set the forwarding cores hexadecimal mask::
480 testpmd> set coremask (mask)
482 This is equivalent to the ``--coremask`` command-line option.
486 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding.
491 Set the forwarding ports hexadecimal mask::
493 testpmd> set portmask (mask)
495 This is equivalent to the ``--portmask`` command-line option.
500 Set number of packets per burst::
502 testpmd> set burst (num)
504 This is equivalent to the ``--burst command-line`` option.
506 When retry is enabled, the transmit delay time and number of retries can also be set::
508 testpmd> set burst tx delay (microseconds) retry (num)
513 Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode::
515 testpmd> set txpkts (x[,y]*)
517 Where x[,y]* represents a CSV list of values, without white space.
522 Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes::
524 testpmd> set txsplit (off|on|rand)
528 * ``off`` disable packet copy & split for CSUM mode.
530 * ``on`` split outgoing packet into multiple segments. Size of each segment
531 and number of segments per packet is determined by ``set txpkts`` command
534 * ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments.
539 Set the list of forwarding cores::
541 testpmd> set corelist (x[,y]*)
543 For example, to change the forwarding cores:
545 .. code-block:: console
547 testpmd> set corelist 3,1
548 testpmd> show config fwd
550 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled
551 Logical Core 3 (socket 0) forwards packets on 1 streams:
552 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01
553 Logical Core 1 (socket 0) forwards packets on 1 streams:
554 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
558 The cores are used in the same order as specified on the command line.
563 Set the list of forwarding ports::
565 testpmd> set portlist (x[,y]*)
567 For example, to change the port forwarding:
569 .. code-block:: console
571 testpmd> set portlist 0,2,1,3
572 testpmd> show config fwd
574 io packet forwarding - ports=4 - cores=1 - streams=4
575 Logical Core 3 (socket 0) forwards packets on 4 streams:
576 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01
577 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
578 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03
579 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02
584 Enable/disable tx loopback::
586 testpmd> set tx loopback (port_id) (on|off)
591 set drop enable bit for all queues::
593 testpmd> set all queues drop (port_id) (on|off)
595 set split drop enable (for VF)
596 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
598 set split drop enable bit for VF from PF::
600 testpmd> set vf split drop (port_id) (vf_id) (on|off)
602 set mac antispoof (for VF)
603 ~~~~~~~~~~~~~~~~~~~~~~~~~~
605 Set mac antispoof for a VF from the PF::
607 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off)
612 Enable/disable MACsec offload::
614 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off)
615 testpmd> set macsec offload (port_id) off
620 Configure MACsec secure connection (SC)::
622 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi)
626 The pi argument is ignored for tx.
627 Check the NIC Datasheet for hardware limits.
632 Configure MACsec secure association (SA)::
634 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key)
638 The IDX value must be 0 or 1.
639 Check the NIC Datasheet for hardware limits.
641 set broadcast mode (for VF)
642 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
644 Set broadcast mode for a VF from the PF::
646 testpmd> set vf broadcast (port_id) (vf_id) (on|off)
651 Set the VLAN strip on a port::
653 testpmd> vlan set strip (on|off) (port_id)
658 Set the VLAN strip for a queue on a port::
660 testpmd> vlan set stripq (on|off) (port_id,queue_id)
662 vlan set stripq (for VF)
663 ~~~~~~~~~~~~~~~~~~~~~~~~
665 Set VLAN strip for all queues in a pool for a VF from the PF::
667 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off)
669 vlan set insert (for VF)
670 ~~~~~~~~~~~~~~~~~~~~~~~~
672 Set VLAN insert for a VF from the PF::
674 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id)
676 vlan set tag (for VF)
677 ~~~~~~~~~~~~~~~~~~~~~
679 Set VLAN tag for a VF from the PF::
681 testpmd> set vf vlan tag (port_id) (vf_id) (on|off)
683 vlan set antispoof (for VF)
684 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
686 Set VLAN antispoof for a VF from the PF::
688 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off)
693 Set the VLAN filter on a port::
695 testpmd> vlan set filter (on|off) (port_id)
700 Set the VLAN QinQ (extended queue in queue) on for a port::
702 testpmd> vlan set qinq (on|off) (port_id)
707 Set the inner or outer VLAN TPID for packet filtering on a port::
709 testpmd> vlan set (inner|outer) tpid (value) (port_id)
713 TPID value must be a 16-bit number (value <= 65536).
718 Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID::
720 testpmd> rx_vlan add (vlan_id|all) (port_id)
724 VLAN filter must be set on that port. VLAN ID < 4096.
725 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries
726 in VFTA table. This is important if enabling all vlan_ids.
731 Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID::
733 testpmd> rx_vlan rm (vlan_id|all) (port_id)
738 Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID::
740 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask)
745 Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID::
747 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask)
752 Add a tunnel filter on a port::
754 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \
755 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
756 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
758 The available information categories are:
760 * ``vxlan``: Set tunnel type as VXLAN.
762 * ``nvgre``: Set tunnel type as NVGRE.
764 * ``ipingre``: Set tunnel type as IP-in-GRE.
766 * ``imac-ivlan``: Set filter type as Inner MAC and VLAN.
768 * ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID.
770 * ``imac-tenid``: Set filter type as Inner MAC and tenant ID.
772 * ``imac``: Set filter type as Inner MAC.
774 * ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID.
776 * ``oip``: Set filter type as Outer IP.
778 * ``iip``: Set filter type as Inner IP.
782 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \
783 192.168.2.2 0 ipingre oip 1 1
785 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP.
790 Remove a tunnel filter on a port::
792 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \
793 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
794 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
799 Add an UDP port for VXLAN packet filter on a port::
801 testpmd> rx_vxlan_port add (udp_port) (port_id)
806 Remove an UDP port for VXLAN packet filter on a port::
808 testpmd> rx_vxlan_port rm (udp_port) (port_id)
813 Set hardware insertion of VLAN IDs in packets sent on a port::
815 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer]
817 For example, set a single VLAN ID (5) insertion on port 0::
821 Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1::
829 Set port based hardware insertion of VLAN ID in packets sent on a port::
831 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off)
836 Disable hardware insertion of a VLAN header in packets sent on a port::
838 testpmd> tx_vlan reset (port_id)
843 Select hardware or software calculation of the checksum when
844 transmitting a packet using the ``csum`` forwarding engine::
846 testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id)
850 * ``ip|udp|tcp|sctp`` always relate to the inner layer.
852 * ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized
853 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are
854 supported). See also the ``csum parse-tunnel`` command.
858 Check the NIC Datasheet for hardware limits.
863 Set RSS queue region span on a port::
865 testpmd> set port (port_id) queue-region region_id (value) \
866 queue_start_index (value) queue_num (value)
868 Set flowtype mapping on a RSS queue region on a port::
870 testpmd> set port (port_id) queue-region region_id (value) flowtype (value)
874 * For the flowtype(pctype) of packet,the specific index for each type has
875 been defined in file i40e_type.h as enum i40e_filter_pctype.
877 Set user priority mapping on a RSS queue region on a port::
879 testpmd> set port (port_id) queue-region UP (value) region_id (value)
881 Flush all queue region related configuration on a port::
883 testpmd> set port (port_id) queue-region flush (on|off)
887 * "on"is just an enable function which server for other configuration,
888 it is for all configuration about queue region from up layer,
889 at first will only keep in DPDK softwarestored in driver,
890 only after "flush on", it commit all configuration to HW.
891 "off" is just clean all configuration about queue region just now,
892 and restore all to DPDK i40e driver default config when start up.
894 Show all queue region related configuration info on a port::
896 testpmd> show port (port_id) queue-region
900 Queue region only support on PF by now, so these command is
901 only for configuration of queue region on PF port.
906 Define how tunneled packets should be handled by the csum forward
909 testpmd> csum parse-tunnel (on|off) (tx_port_id)
911 If enabled, the csum forward engine will try to recognize supported
912 tunnel headers (vxlan, gre, ipip).
914 If disabled, treat tunnel packets as non-tunneled packets (a inner
915 header is handled as a packet payload).
919 The port argument is the TX port like in the ``csum set`` command.
923 Consider a packet in packet like the following::
925 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in
927 * If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
928 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the
929 ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``).
931 * If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
932 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``.
937 Display tx checksum offload configuration::
939 testpmd> csum show (port_id)
944 Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine::
946 testpmd> tso set (segsize) (port_id)
950 Check the NIC datasheet for hardware limits.
955 Display the status of TCP Segmentation Offload::
957 testpmd> tso show (port_id)
962 Enable or disable GRO in ``csum`` forwarding engine::
964 testpmd> set port <port_id> gro on|off
966 If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4
967 packets received from the given port.
969 If disabled, packets received from the given port won't be performed
970 GRO. By default, GRO is disabled for all ports.
974 When enable GRO for a port, TCP/IPv4 packets received from the port
975 will be performed GRO. After GRO, all merged packets have bad
976 checksums, since the GRO library doesn't re-calculate checksums for
977 the merged packets. Therefore, if users want the merged packets to
978 have correct checksums, please select HW IP checksum calculation and
979 HW TCP checksum calculation for the port which the merged packets are
985 Display GRO configuration for a given port::
987 testpmd> show port <port_id> gro
992 Set the cycle to flush the GROed packets from reassembly tables::
994 testpmd> set gro flush <cycles>
996 When enable GRO, the csum forwarding engine performs GRO on received
997 packets, and the GROed packets are stored in reassembly tables. Users
998 can use this command to determine when the GROed packets are flushed
999 from the reassembly tables.
1001 The ``cycles`` is measured in GRO operation times. The csum forwarding
1002 engine flushes the GROed packets from the tables every ``cycles`` GRO
1005 By default, the value of ``cycles`` is 1, which means flush GROed packets
1006 from the reassembly tables as soon as one GRO operation finishes. The value
1007 of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``.
1009 Please note that the large value of ``cycles`` may cause the poor TCP/IP
1010 stack performance. Because the GROed packets are delayed to arrive the
1011 stack, thus causing more duplicated ACKs and TCP retransmissions.
1016 Toggle per-port GSO support in ``csum`` forwarding engine::
1018 testpmd> set port <port_id> gso on|off
1020 If enabled, the csum forwarding engine will perform GSO on supported IPv4
1021 packets, transmitted on the given port.
1023 If disabled, packets transmitted on the given port will not undergo GSO.
1024 By default, GSO is disabled for all ports.
1028 When GSO is enabled on a port, supported IPv4 packets transmitted on that
1029 port undergo GSO. Afterwards, the segmented packets are represented by
1030 multi-segment mbufs; however, the csum forwarding engine doesn't calculation
1031 of checksums for GSO'd segments in SW. As a result, if users want correct
1032 checksums in GSO segments, they should enable HW checksum calculation for
1035 For example, HW checksum calculation for VxLAN GSO'd packets may be enabled
1036 by setting the following options in the csum forwarding engine:
1038 testpmd> csum set outer_ip hw <port_id>
1040 testpmd> csum set ip hw <port_id>
1042 testpmd> csum set tcp hw <port_id>
1047 Set the maximum GSO segment size (measured in bytes), which includes the
1048 packet header and the packet payload for GSO-enabled ports (global)::
1050 testpmd> set gso segsz <length>
1055 Display the status of Generic Segmentation Offload for a given port::
1057 testpmd> show port <port_id> gso
1062 Add an alternative MAC address to a port::
1064 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX)
1069 Remove a MAC address from a port::
1071 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX)
1073 mac_addr add (for VF)
1074 ~~~~~~~~~~~~~~~~~~~~~
1076 Add an alternative MAC address for a VF to a port::
1078 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX)
1083 Set the default MAC address for a port::
1085 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX)
1087 mac_addr set (for VF)
1088 ~~~~~~~~~~~~~~~~~~~~~
1090 Set the MAC address for a VF from the PF::
1092 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX)
1097 Set the unicast hash filter(s) on/off for a port::
1099 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
1104 Set the promiscuous mode on for a port or for all ports.
1105 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1107 testpmd> set promisc (port_id|all) (on|off)
1112 Set the allmulti mode for a port or for all ports::
1114 testpmd> set allmulti (port_id|all) (on|off)
1116 Same as the ifconfig (8) option. Controls how multicast packets are handled.
1118 set promisc (for VF)
1119 ~~~~~~~~~~~~~~~~~~~~
1121 Set the unicast promiscuous mode for a VF from PF.
1122 It's supported by Intel i40e NICs now.
1123 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1125 testpmd> set vf promisc (port_id) (vf_id) (on|off)
1127 set allmulticast (for VF)
1128 ~~~~~~~~~~~~~~~~~~~~~~~~~
1130 Set the multicast promiscuous mode for a VF from PF.
1131 It's supported by Intel i40e NICs now.
1132 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1134 testpmd> set vf allmulti (port_id) (vf_id) (on|off)
1136 set tx max bandwidth (for VF)
1137 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1139 Set TX max absolute bandwidth (Mbps) for a VF from PF::
1141 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth)
1143 set tc tx min bandwidth (for VF)
1144 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1146 Set all TCs' TX min relative bandwidth (%) for a VF from PF::
1148 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...)
1150 set tc tx max bandwidth (for VF)
1151 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1153 Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF::
1155 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth)
1157 set tc strict link priority mode
1158 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1160 Set some TCs' strict link priority mode on a physical port::
1162 testpmd> set tx strict-link-priority (port_id) (tc_bitmap)
1164 set tc tx min bandwidth
1165 ~~~~~~~~~~~~~~~~~~~~~~~
1167 Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs::
1169 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...)
1174 Set the link flow control parameter on a port::
1176 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1177 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
1178 autoneg (on|off) (port_id)
1182 * ``high_water`` (integer): High threshold value to trigger XOFF.
1184 * ``low_water`` (integer): Low threshold value to trigger XON.
1186 * ``pause_time`` (integer): Pause quota in the Pause frame.
1188 * ``send_xon`` (0/1): Send XON frame.
1190 * ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
1192 * ``autoneg``: Change the auto-negotiation parameter.
1197 Set the priority flow control parameter on a port::
1199 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1200 (pause_time) (priority) (port_id)
1204 * ``high_water`` (integer): High threshold value.
1206 * ``low_water`` (integer): Low threshold value.
1208 * ``pause_time`` (integer): Pause quota in the Pause frame.
1210 * ``priority`` (0-7): VLAN User Priority.
1215 Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
1217 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
1219 For example, to set rx queue 2 on port 0 to mapping 5::
1221 testpmd>set stat_qmap rx 0 2 5
1223 set port - rx/tx (for VF)
1224 ~~~~~~~~~~~~~~~~~~~~~~~~~
1226 Set VF receive/transmit from a port::
1228 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
1230 set port - mac address filter (for VF)
1231 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1233 Add/Remove unicast or multicast MAC addr filter for a VF::
1235 testpmd> set port (port_id) vf (vf_id) (mac_addr) \
1236 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
1238 set port - rx mode(for VF)
1239 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1241 Set the VF receive mode of a port::
1243 testpmd> set port (port_id) vf (vf_id) \
1244 rxmode (AUPE|ROPE|BAM|MPE) (on|off)
1246 The available receive modes are:
1248 * ``AUPE``: Accepts untagged VLAN.
1250 * ``ROPE``: Accepts unicast hash.
1252 * ``BAM``: Accepts broadcast packets.
1254 * ``MPE``: Accepts all multicast packets.
1256 set port - tx_rate (for Queue)
1257 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1259 Set TX rate limitation for a queue on a port::
1261 testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1263 set port - tx_rate (for VF)
1264 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
1266 Set TX rate limitation for queues in VF on a port::
1268 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1270 set port - mirror rule
1271 ~~~~~~~~~~~~~~~~~~~~~~
1273 Set pool or vlan type mirror rule for a port::
1275 testpmd> set port (port_id) mirror-rule (rule_id) \
1276 (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1277 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1279 Set link mirror rule for a port::
1281 testpmd> set port (port_id) mirror-rule (rule_id) \
1282 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1284 For example to enable mirror traffic with vlan 0,1 to pool 0::
1286 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1288 reset port - mirror rule
1289 ~~~~~~~~~~~~~~~~~~~~~~~~
1291 Reset a mirror rule for a port::
1293 testpmd> reset port (port_id) mirror-rule (rule_id)
1298 Set the flush on RX streams before forwarding.
1299 The default is flush ``on``.
1300 Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1302 testpmd> set flush_rx off
1307 Set the bypass mode for the lowest port on bypass enabled NIC::
1309 testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1314 Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1316 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1317 mode (normal|bypass|isolate) (port_id)
1321 * ``timeout``: Enable bypass after watchdog timeout.
1323 * ``os_on``: Enable bypass when OS/board is powered on.
1325 * ``os_off``: Enable bypass when OS/board is powered off.
1327 * ``power_on``: Enable bypass when power supply is turned on.
1329 * ``power_off``: Enable bypass when power supply is turned off.
1335 Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1337 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1342 Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1344 testpmd> show bypass config (port_id)
1349 Set link up for a port::
1351 testpmd> set link-up port (port id)
1356 Set link down for a port::
1358 testpmd> set link-down port (port id)
1363 Enable E-tag insertion for a VF on a port::
1365 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1367 Disable E-tag insertion for a VF on a port::
1369 testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1371 Enable/disable E-tag stripping on a port::
1373 testpmd> E-tag set stripping (on|off) port (port_id)
1375 Enable/disable E-tag based forwarding on a port::
1377 testpmd> E-tag set forwarding (on|off) port (port_id)
1379 Add an E-tag forwarding filter on a port::
1381 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1383 Delete an E-tag forwarding filter on a port::
1384 testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1389 Load a dynamic device personalization (DDP) package::
1391 testpmd> ddp add (port_id) (package_path[,output_path])
1396 Delete a dynamic device personalization package::
1398 testpmd> ddp del (port_id) (package_path)
1403 List all items from the ptype mapping table::
1405 testpmd> ptype mapping get (port_id) (valid_only)
1409 * ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0).
1411 Replace a specific or a group of software defined ptype with a new one::
1413 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type)
1417 * ``target``: A specific software ptype or a mask to represent a group of software ptypes.
1419 * ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1).
1421 * ``pkt_type``: The new software ptype to replace the old ones.
1423 Update hardware defined ptype to software defined packet type mapping table::
1425 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype)
1429 * ``hw_ptype``: hardware ptype as the index of the ptype mapping table.
1431 * ``sw_ptype``: software ptype as the value of the ptype mapping table.
1433 Reset ptype mapping table::
1435 testpmd> ptype mapping reset (port_id)
1440 The following sections show functions for configuring ports.
1444 Port configuration changes only become active when forwarding is started/restarted.
1449 Attach a port specified by pci address or virtual device args::
1451 testpmd> port attach (identifier)
1453 To attach a new pci device, the device should be recognized by kernel first.
1454 Then it should be moved under DPDK management.
1455 Finally the port can be attached to testpmd.
1457 For example, to move a pci device using ixgbe under DPDK management:
1459 .. code-block:: console
1461 # Check the status of the available devices.
1462 ./usertools/dpdk-devbind.py --status
1464 Network devices using DPDK-compatible driver
1465 ============================================
1468 Network devices using kernel driver
1469 ===================================
1470 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1473 # Bind the device to igb_uio.
1474 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1477 # Recheck the status of the devices.
1478 ./usertools/dpdk-devbind.py --status
1479 Network devices using DPDK-compatible driver
1480 ============================================
1481 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1483 To attach a port created by virtual device, above steps are not needed.
1485 For example, to attach a port whose pci address is 0000:0a:00.0.
1487 .. code-block:: console
1489 testpmd> port attach 0000:0a:00.0
1490 Attaching a new port...
1491 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1492 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1493 EAL: PCI memory mapped at 0x7f83bfa00000
1494 EAL: PCI memory mapped at 0x7f83bfa80000
1495 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1496 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1497 Port 0 is attached. Now total ports is 1
1500 For example, to attach a port created by pcap PMD.
1502 .. code-block:: console
1504 testpmd> port attach net_pcap0
1505 Attaching a new port...
1506 PMD: Initializing pmd_pcap for net_pcap0
1507 PMD: Creating pcap-backed ethdev on numa socket 0
1508 Port 0 is attached. Now total ports is 1
1511 In this case, identifier is ``net_pcap0``.
1512 This identifier format is the same as ``--vdev`` format of DPDK applications.
1514 For example, to re-attach a bonded port which has been previously detached,
1515 the mode and slave parameters must be given.
1517 .. code-block:: console
1519 testpmd> port attach net_bond_0,mode=0,slave=1
1520 Attaching a new port...
1521 EAL: Initializing pmd_bond for net_bond_0
1522 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1523 Port 0 is attached. Now total ports is 1
1530 Detach a specific port::
1532 testpmd> port detach (port_id)
1534 Before detaching a port, the port should be stopped and closed.
1536 For example, to detach a pci device port 0.
1538 .. code-block:: console
1540 testpmd> port stop 0
1543 testpmd> port close 0
1547 testpmd> port detach 0
1549 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1550 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1551 EAL: PCI memory unmapped at 0x7f83bfa00000
1552 EAL: PCI memory unmapped at 0x7f83bfa80000
1556 For example, to detach a virtual device port 0.
1558 .. code-block:: console
1560 testpmd> port stop 0
1563 testpmd> port close 0
1567 testpmd> port detach 0
1569 PMD: Closing pcap ethdev on numa socket 0
1570 Port 'net_pcap0' is detached. Now total ports is 0
1573 To remove a pci device completely from the system, first detach the port from testpmd.
1574 Then the device should be moved under kernel management.
1575 Finally the device can be removed using kernel pci hotplug functionality.
1577 For example, to move a pci device under kernel management:
1579 .. code-block:: console
1581 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1583 ./usertools/dpdk-devbind.py --status
1585 Network devices using DPDK-compatible driver
1586 ============================================
1589 Network devices using kernel driver
1590 ===================================
1591 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1593 To remove a port created by a virtual device, above steps are not needed.
1598 Start all ports or a specific port::
1600 testpmd> port start (port_id|all)
1605 Stop all ports or a specific port::
1607 testpmd> port stop (port_id|all)
1612 Close all ports or a specific port::
1614 testpmd> port close (port_id|all)
1616 port start/stop queue
1617 ~~~~~~~~~~~~~~~~~~~~~
1619 Start/stop a rx/tx queue on a specific port::
1621 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1623 Only take effect when port is started.
1628 Set the speed and duplex mode for all ports or a specific port::
1630 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1631 duplex (half|full|auto)
1633 port config - queues/descriptors
1634 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1636 Set number of queues/descriptors for rxq, txq, rxd and txd::
1638 testpmd> port config all (rxq|txq|rxd|txd) (value)
1640 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1642 port config - max-pkt-len
1643 ~~~~~~~~~~~~~~~~~~~~~~~~~
1645 Set the maximum packet length::
1647 testpmd> port config all max-pkt-len (value)
1649 This is equivalent to the ``--max-pkt-len`` command-line option.
1651 port config - CRC Strip
1652 ~~~~~~~~~~~~~~~~~~~~~~~
1654 Set hardware CRC stripping on or off for all ports::
1656 testpmd> port config all crc-strip (on|off)
1658 CRC stripping is on by default.
1660 The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option.
1662 port config - scatter
1663 ~~~~~~~~~~~~~~~~~~~~~~~
1665 Set RX scatter mode on or off for all ports::
1667 testpmd> port config all scatter (on|off)
1669 RX scatter mode is off by default.
1671 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1673 port config - TX queue flags
1674 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1676 Set a hexadecimal bitmap of TX queue flags for all ports::
1678 testpmd> port config all txqflags value
1680 This command is equivalent to the ``--txqflags`` command-line option.
1682 port config - RX Checksum
1683 ~~~~~~~~~~~~~~~~~~~~~~~~~
1685 Set hardware RX checksum offload to on or off for all ports::
1687 testpmd> port config all rx-cksum (on|off)
1689 Checksum offload is off by default.
1691 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1696 Set hardware VLAN on or off for all ports::
1698 testpmd> port config all hw-vlan (on|off)
1700 Hardware VLAN is on by default.
1702 The ``off`` option is equivalent to the ``--disable-hw-vlan`` command-line option.
1704 port config - VLAN filter
1705 ~~~~~~~~~~~~~~~~~~~~~~~~~
1707 Set hardware VLAN filter on or off for all ports::
1709 testpmd> port config all hw-vlan-filter (on|off)
1711 Hardware VLAN filter is on by default.
1713 The ``off`` option is equivalent to the ``--disable-hw-vlan-filter`` command-line option.
1715 port config - VLAN strip
1716 ~~~~~~~~~~~~~~~~~~~~~~~~
1718 Set hardware VLAN strip on or off for all ports::
1720 testpmd> port config all hw-vlan-strip (on|off)
1722 Hardware VLAN strip is on by default.
1724 The ``off`` option is equivalent to the ``--disable-hw-vlan-strip`` command-line option.
1726 port config - VLAN extend
1727 ~~~~~~~~~~~~~~~~~~~~~~~~~
1729 Set hardware VLAN extend on or off for all ports::
1731 testpmd> port config all hw-vlan-extend (on|off)
1733 Hardware VLAN extend is off by default.
1735 The ``off`` option is equivalent to the ``--disable-hw-vlan-extend`` command-line option.
1737 port config - Drop Packets
1738 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1740 Set packet drop for packets with no descriptors on or off for all ports::
1742 testpmd> port config all drop-en (on|off)
1744 Packet dropping for packets with no descriptors is off by default.
1746 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1751 Set the RSS (Receive Side Scaling) mode on or off::
1753 testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1755 RSS is on by default.
1757 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1759 port config - RSS Reta
1760 ~~~~~~~~~~~~~~~~~~~~~~
1762 Set the RSS (Receive Side Scaling) redirection table::
1764 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1769 Set the DCB mode for an individual port::
1771 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1773 The traffic class should be 4 or 8.
1778 Set the number of packets per burst::
1780 testpmd> port config all burst (value)
1782 This is equivalent to the ``--burst`` command-line option.
1784 port config - Threshold
1785 ~~~~~~~~~~~~~~~~~~~~~~~
1787 Set thresholds for TX/RX queues::
1789 testpmd> port config all (threshold) (value)
1791 Where the threshold type can be:
1793 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1795 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1797 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1799 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1801 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1803 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1805 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1807 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1809 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1811 These threshold options are also available from the command-line.
1816 Set the value of ether-type for E-tag::
1818 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1820 Enable/disable the E-tag support::
1822 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1824 port config pctype mapping
1825 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1827 Reset pctype mapping table::
1829 testpmd> port config (port_id) pctype mapping reset
1831 Update hardware defined pctype to software defined flow type mapping table::
1833 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id)
1837 * ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table.
1839 * ``flow_type_id``: software flow type id as the index of the pctype mapping table.
1842 Link Bonding Functions
1843 ----------------------
1845 The Link Bonding functions make it possible to dynamically create and
1846 manage link bonding devices from within testpmd interactive prompt.
1848 create bonded device
1849 ~~~~~~~~~~~~~~~~~~~~
1851 Create a new bonding device::
1853 testpmd> create bonded device (mode) (socket)
1855 For example, to create a bonded device in mode 1 on socket 0::
1857 testpmd> create bonded 1 0
1858 created new bonded device (port X)
1863 Adds Ethernet device to a Link Bonding device::
1865 testpmd> add bonding slave (slave id) (port id)
1867 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
1869 testpmd> add bonding slave 6 10
1872 remove bonding slave
1873 ~~~~~~~~~~~~~~~~~~~~
1875 Removes an Ethernet slave device from a Link Bonding device::
1877 testpmd> remove bonding slave (slave id) (port id)
1879 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
1881 testpmd> remove bonding slave 6 10
1886 Set the Link Bonding mode of a Link Bonding device::
1888 testpmd> set bonding mode (value) (port id)
1890 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
1892 testpmd> set bonding mode 3 10
1897 Set an Ethernet slave device as the primary device on a Link Bonding device::
1899 testpmd> set bonding primary (slave id) (port id)
1901 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
1903 testpmd> set bonding primary 6 10
1908 Set the MAC address of a Link Bonding device::
1910 testpmd> set bonding mac (port id) (mac)
1912 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
1914 testpmd> set bonding mac 10 00:00:00:00:00:01
1916 set bonding xmit_balance_policy
1917 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1919 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
1921 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
1923 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
1925 testpmd> set bonding xmit_balance_policy 10 l34
1928 set bonding mon_period
1929 ~~~~~~~~~~~~~~~~~~~~~~
1931 Set the link status monitoring polling period in milliseconds for a bonding device.
1933 This adds support for PMD slave devices which do not support link status interrupts.
1934 When the mon_period is set to a value greater than 0 then all PMD's which do not support
1935 link status ISR will be queried every polling interval to check if their link status has changed::
1937 testpmd> set bonding mon_period (port_id) (value)
1939 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
1941 testpmd> set bonding mon_period 5 150
1944 set bonding lacp dedicated_queue
1945 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1947 Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic
1948 when in mode 4 (link-aggregration-802.3ad)::
1950 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable)
1953 set bonding agg_mode
1954 ~~~~~~~~~~~~~~~~~~~~
1956 Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad)::
1958 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable)
1964 Show the current configuration of a Link Bonding device::
1966 testpmd> show bonding config (port id)
1969 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
1970 in balance mode with a transmission policy of layer 2+3::
1972 testpmd> show bonding config 9
1974 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
1976 Active Slaves (3): [1 3 4]
1983 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
1984 This is mainly useful for debugging purposes.
1985 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
1986 and fields that can be accessed.
1991 Display the value of a port register::
1993 testpmd> read reg (port_id) (address)
1995 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
1997 testpmd> read reg 0 0xEE00
1998 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
2003 Display a port register bit field::
2005 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
2007 For example, reading the lowest two bits from the register in the example above::
2009 testpmd> read regfield 0 0xEE00 0 1
2010 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
2015 Display a single port register bit::
2017 testpmd> read regbit (port_id) (address) (bit_x)
2019 For example, reading the lowest bit from the register in the example above::
2021 testpmd> read regbit 0 0xEE00 0
2022 port 0 PCI register at offset 0xEE00: bit 0=1
2027 Set the value of a port register::
2029 testpmd> write reg (port_id) (address) (value)
2031 For example, to clear a register::
2033 testpmd> write reg 0 0xEE00 0x0
2034 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
2039 Set bit field of a port register::
2041 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
2043 For example, writing to the register cleared in the example above::
2045 testpmd> write regfield 0 0xEE00 0 1 2
2046 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
2051 Set single bit value of a port register::
2053 testpmd> write regbit (port_id) (address) (bit_x) (value)
2055 For example, to set the high bit in the register from the example above::
2057 testpmd> write regbit 0 0xEE00 31 1
2058 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
2064 The following section shows functions for configuring traffic management on
2065 on the ethernet device through the use of generic TM API.
2067 show port traffic management capability
2068 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2070 Show traffic management capability of the port::
2072 testpmd> show port tm cap (port_id)
2074 show port traffic management capability (hierarchy level)
2075 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2077 Show traffic management hierarchy level capability of the port::
2079 testpmd> show port tm cap (port_id) (level_id)
2081 show port traffic management capability (hierarchy node level)
2082 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2084 Show the traffic management hierarchy node capability of the port::
2086 testpmd> show port tm cap (port_id) (node_id)
2088 show port traffic management hierarchy node type
2089 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2091 Show the port traffic management hierarchy node type::
2093 testpmd> show port tm node type (port_id) (node_id)
2095 show port traffic management hierarchy node stats
2096 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2098 Show the port traffic management hierarchy node statistics::
2100 testpmd> show port tm node stats (port_id) (node_id) (clear)
2104 * ``clear``: When this parameter has a non-zero value, the statistics counters
2105 are cleared (i.e. set to zero) immediately after they have been read,
2106 otherwise the statistics counters are left untouched.
2108 Add port traffic management private shaper profile
2109 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2111 Add the port traffic management private shaper profile::
2113 testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \
2114 (tb_rate) (tb_size) (packet_length_adjust)
2118 * ``shaper_profile id``: Shaper profile ID for the new profile.
2119 * ``tb_rate``: Token bucket rate (bytes per second).
2120 * ``tb_size``: Token bucket size (bytes).
2121 * ``packet_length_adjust``: The value (bytes) to be added to the length of
2122 each packet for the purpose of shaping. This parameter value can be used to
2123 correct the packet length with the framing overhead bytes that are consumed
2126 Delete port traffic management private shaper profile
2127 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2129 Delete the port traffic management private shaper::
2131 testpmd> del port tm node shaper profile (port_id) (shaper_profile_id)
2135 * ``shaper_profile id``: Shaper profile ID that needs to be deleted.
2137 Add port traffic management shared shaper
2138 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2140 Create the port traffic management shared shaper::
2142 testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \
2147 * ``shared_shaper_id``: Shared shaper ID to be created.
2148 * ``shaper_profile id``: Shaper profile ID for shared shaper.
2150 Set port traffic management shared shaper
2151 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2153 Update the port traffic management shared shaper::
2155 testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \
2160 * ``shared_shaper_id``: Shared shaper ID to be update.
2161 * ``shaper_profile id``: Shaper profile ID for shared shaper.
2163 Delete port traffic management shared shaper
2164 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2166 Delete the port traffic management shared shaper::
2168 testpmd> del port tm node shared shaper (port_id) (shared_shaper_id)
2172 * ``shared_shaper_id``: Shared shaper ID to be deleted.
2174 Set port traffic management hiearchy node private shaper
2175 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2177 set the port traffic management hierarchy node private shaper::
2179 testpmd> set port tm node shaper profile (port_id) (node_id) \
2184 * ``shaper_profile id``: Private shaper profile ID to be enabled on the
2187 Add port traffic management WRED profile
2188 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2190 Create a new WRED profile::
2192 testpmd> add port tm node wred profile (port_id) (wred_profile_id) \
2193 (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \
2194 (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \
2195 (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r)
2199 * ``wred_profile id``: Identifier for the newly create WRED profile
2200 * ``color_g``: Packet color (green)
2201 * ``min_th_g``: Minimum queue threshold for packet with green color
2202 * ``max_th_g``: Minimum queue threshold for packet with green color
2203 * ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp)
2204 * ``wq_log2_g``: Negated log2 of queue weight (wq)
2205 * ``color_y``: Packet color (yellow)
2206 * ``min_th_y``: Minimum queue threshold for packet with yellow color
2207 * ``max_th_y``: Minimum queue threshold for packet with yellow color
2208 * ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp)
2209 * ``wq_log2_y``: Negated log2 of queue weight (wq)
2210 * ``color_r``: Packet color (red)
2211 * ``min_th_r``: Minimum queue threshold for packet with yellow color
2212 * ``max_th_r``: Minimum queue threshold for packet with yellow color
2213 * ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp)
2214 * ``wq_log2_r``: Negated log2 of queue weight (wq)
2216 Delete port traffic management WRED profile
2217 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2219 Delete the WRED profile::
2221 testpmd> del port tm node wred profile (port_id) (wred_profile_id)
2223 Add port traffic management hierarchy nonleaf node
2224 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2226 Add nonleaf node to port traffic management hiearchy::
2228 testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \
2229 (priority) (weight) (level_id) (shaper_profile_id) \
2230 (shared_shaper_id) (n_shared_shapers) (n_sp_priorities) (stats_mask) \
2234 * ``parent_node_id``: Node ID of the parent.
2235 * ``priority``: Node priority (highest node priority is zero). This is used by
2236 the SP algorithm running on the parent node for scheduling this node.
2237 * ``weight``: Node weight (lowest weight is one). The node weight is relative
2238 to the weight sum of all siblings that have the same priority. It is used by
2239 the WFQ algorithm running on the parent node for scheduling this node.
2240 * ``level_id``: Hiearchy level of the node.
2241 * ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2243 * ``shared_shaper_id``: Shared shaper id.
2244 * ``n_shared_shapers``: Number of shared shapers.
2245 * ``n_sp_priorities``: Number of strict priorities.
2246 * ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2248 Add port traffic management hierarchy leaf node
2249 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2251 Add leaf node to port traffic management hiearchy::
2253 testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \
2254 (priority) (weight) (level_id) (shaper_profile_id) \
2255 (shared_shaper_id) (n_shared_shapers) (cman_mode) \
2256 (wred_profile_id) (stats_mask) \
2260 * ``parent_node_id``: Node ID of the parent.
2261 * ``priority``: Node priority (highest node priority is zero). This is used by
2262 the SP algorithm running on the parent node for scheduling this node.
2263 * ``weight``: Node weight (lowest weight is one). The node weight is relative
2264 to the weight sum of all siblings that have the same priority. It is used by
2265 the WFQ algorithm running on the parent node for scheduling this node.
2266 * ``level_id``: Hiearchy level of the node.
2267 * ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2269 * ``shared_shaper_id``: Shared shaper id.
2270 * ``n_shared_shapers``: Number of shared shapers.
2271 * ``cman_mode``: Congestion management mode to be enabled for this node.
2272 * ``wred_profile_id``: WRED profile id to be enabled for this node.
2273 * ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2275 Delete port traffic management hierarchy node
2276 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2278 Delete node from port traffic management hiearchy::
2280 testpmd> del port tm node (port_id) (node_id)
2282 Update port traffic management hierarchy parent node
2283 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2285 Update port traffic management hierarchy parent node::
2287 testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \
2290 This function can only be called after the hierarchy commit invocation. Its
2291 success depends on the port support for this operation, as advertised through
2292 the port capability set. This function is valid for all nodes of the traffic
2293 management hierarchy except root node.
2295 Commit port traffic management hierarchy
2296 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2298 Commit the traffic management hierarchy on the port::
2300 testpmd> port tm hierarchy commit (port_id) (clean_on_fail)
2304 * ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function
2305 call failure. On the other hand, hierarchy is preserved when this parameter
2308 Set port traffic management default hierarchy (tm forwarding mode)
2309 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2311 set the traffic management default hierarchy on the port::
2313 testpmd> set port tm hierarchy default (port_id)
2318 This section details the available filter functions that are available.
2320 Note these functions interface the deprecated legacy filtering framework,
2321 superseded by *rte_flow*. See `Flow rules management`_.
2324 ~~~~~~~~~~~~~~~~~~~~
2326 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
2328 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
2329 ethertype (ether_type) (drop|fwd) queue (queue_id)
2331 The available information parameters are:
2333 * ``port_id``: The port which the Ethertype filter assigned on.
2335 * ``mac_addr``: Compare destination mac address.
2337 * ``mac_ignr``: Ignore destination mac address match.
2339 * ``mac_address``: Destination mac address to match.
2341 * ``ether_type``: The EtherType value want to match,
2342 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
2344 * ``queue_id``: The receive queue associated with this EtherType filter.
2345 It is meaningless when deleting or dropping.
2347 Example, to add/remove an ethertype filter rule::
2349 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
2350 ethertype 0x0806 fwd queue 3
2352 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
2353 ethertype 0x0806 fwd queue 3
2358 Add or delete a 2-tuple filter,
2359 which identifies packets by specific protocol and destination TCP/UDP port
2360 and forwards packets into one of the receive queues::
2362 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
2363 protocol (protocol_value) mask (mask_value) \
2364 tcp_flags (tcp_flags_value) priority (prio_value) \
2367 The available information parameters are:
2369 * ``port_id``: The port which the 2-tuple filter assigned on.
2371 * ``dst_port_value``: Destination port in L4.
2373 * ``protocol_value``: IP L4 protocol.
2375 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
2377 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
2379 * ``prio_value``: Priority of this filter.
2381 * ``queue_id``: The receive queue associated with this 2-tuple filter.
2383 Example, to add/remove an 2tuple filter rule::
2385 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
2386 tcp_flags 0x02 priority 3 queue 3
2388 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
2389 tcp_flags 0x02 priority 3 queue 3
2394 Add or delete a 5-tuple filter,
2395 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
2396 and routes packets into one of the receive queues::
2398 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
2399 (src_address) dst_port (dst_port_value) \
2400 src_port (src_port_value) protocol (protocol_value) \
2401 mask (mask_value) tcp_flags (tcp_flags_value) \
2402 priority (prio_value) queue (queue_id)
2404 The available information parameters are:
2406 * ``port_id``: The port which the 5-tuple filter assigned on.
2408 * ``dst_address``: Destination IP address.
2410 * ``src_address``: Source IP address.
2412 * ``dst_port_value``: TCP/UDP destination port.
2414 * ``src_port_value``: TCP/UDP source port.
2416 * ``protocol_value``: L4 protocol.
2418 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
2420 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
2422 * ``prio_value``: The priority of this filter.
2424 * ``queue_id``: The receive queue associated with this 5-tuple filter.
2426 Example, to add/remove an 5tuple filter rule::
2428 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2429 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2430 flags 0x0 priority 3 queue 3
2432 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2433 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2434 flags 0x0 priority 3 queue 3
2439 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
2441 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
2443 The available information parameters are:
2445 * ``port_id``: The port which the SYN filter assigned on.
2447 * ``high``: This SYN filter has higher priority than other filters.
2449 * ``low``: This SYN filter has lower priority than other filters.
2451 * ``queue_id``: The receive queue associated with this SYN filter
2455 testpmd> syn_filter 0 add priority high queue 3
2460 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
2461 and routed into one of the receive queues::
2463 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
2464 mask (mask_value) priority (prio_value) queue (queue_id)
2466 The available information parameters are:
2468 * ``port_id``: The port which the Flex filter is assigned on.
2470 * ``len_value``: Filter length in bytes, no greater than 128.
2472 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
2474 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
2476 * ``prio_value``: The priority of this filter.
2478 * ``queue_id``: The receive queue associated with this Flex filter.
2482 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
2483 mask 000C priority 3 queue 3
2485 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
2486 mask 000C priority 3 queue 3
2489 .. _testpmd_flow_director:
2491 flow_director_filter
2492 ~~~~~~~~~~~~~~~~~~~~
2494 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
2496 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
2497 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
2499 * Perfect match filters.
2500 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2501 The masked fields are for IP flow.
2503 * Signature filters.
2504 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
2506 * Perfect-mac-vlan match filters.
2507 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2508 The masked fields are for MAC VLAN flow.
2510 * Perfect-tunnel match filters.
2511 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2512 The masked fields are for tunnel flow.
2514 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
2515 per flow type and the flexible payload.
2517 The Flow Director can also mask out parts of all of these fields so that filters
2518 are only applied to certain fields or parts of the fields.
2520 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
2522 # Commands to add flow director filters of different flow types::
2524 flow_director_filter (port_id) mode IP (add|del|update) \
2525 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
2526 src (src_ip_address) dst (dst_ip_address) \
2527 tos (tos_value) proto (proto_value) ttl (ttl_value) \
2528 vlan (vlan_value) flexbytes (flexbytes_value) \
2529 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
2532 flow_director_filter (port_id) mode IP (add|del|update) \
2533 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
2534 src (src_ip_address) (src_port) \
2535 dst (dst_ip_address) (dst_port) \
2536 tos (tos_value) ttl (ttl_value) \
2537 vlan (vlan_value) flexbytes (flexbytes_value) \
2538 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
2541 flow_director_filter (port_id) mode IP (add|del|update) \
2542 flow (ipv4-sctp|ipv6-sctp) \
2543 src (src_ip_address) (src_port) \
2544 dst (dst_ip_address) (dst_port) \
2545 tos (tos_value) ttl (ttl_value) \
2546 tag (verification_tag) vlan (vlan_value) \
2547 flexbytes (flexbytes_value) (drop|fwd) \
2548 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
2550 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
2551 ether (ethertype) flexbytes (flexbytes_value) \
2552 (drop|fwd) pf|vf(vf_id) queue (queue_id)
2555 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
2556 mac (mac_address) vlan (vlan_value) \
2557 flexbytes (flexbytes_value) (drop|fwd) \
2558 queue (queue_id) fd_id (fd_id_value)
2560 flow_director_filter (port_id) mode Tunnel (add|del|update) \
2561 mac (mac_address) vlan (vlan_value) \
2562 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
2563 flexbytes (flexbytes_value) (drop|fwd) \
2564 queue (queue_id) fd_id (fd_id_value)
2566 For example, to add an ipv4-udp flow type filter::
2568 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
2569 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
2570 fwd pf queue 1 fd_id 1
2572 For example, add an ipv4-other flow type filter::
2574 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
2575 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
2576 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
2581 Flush all flow director filters on a device::
2583 testpmd> flush_flow_director (port_id)
2585 Example, to flush all flow director filter on port 0::
2587 testpmd> flush_flow_director 0
2592 Set flow director's input masks::
2594 flow_director_mask (port_id) mode IP vlan (vlan_value) \
2595 src_mask (ipv4_src) (ipv6_src) (src_port) \
2596 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2598 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2600 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2601 mac (mac_value) tunnel-type (tunnel_type_value) \
2602 tunnel-id (tunnel_id_value)
2604 Example, to set flow director mask on port 0::
2606 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2607 src_mask 255.255.255.255 \
2608 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2609 dst_mask 255.255.255.255 \
2610 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2612 flow_director_flex_mask
2613 ~~~~~~~~~~~~~~~~~~~~~~~
2615 set masks of flow director's flexible payload based on certain flow type::
2617 testpmd> flow_director_flex_mask (port_id) \
2618 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2619 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2620 l2_payload|all) (mask)
2622 Example, to set flow director's flex mask for all flow type on port 0::
2624 testpmd> flow_director_flex_mask 0 flow all \
2625 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2628 flow_director_flex_payload
2629 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2631 Configure flexible payload selection::
2633 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2635 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2637 testpmd> flow_director_flex_payload 0 l4 \
2638 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2640 get_sym_hash_ena_per_port
2641 ~~~~~~~~~~~~~~~~~~~~~~~~~
2643 Get symmetric hash enable configuration per port::
2645 get_sym_hash_ena_per_port (port_id)
2647 For example, to get symmetric hash enable configuration of port 1::
2649 testpmd> get_sym_hash_ena_per_port 1
2651 set_sym_hash_ena_per_port
2652 ~~~~~~~~~~~~~~~~~~~~~~~~~
2654 Set symmetric hash enable configuration per port to enable or disable::
2656 set_sym_hash_ena_per_port (port_id) (enable|disable)
2658 For example, to set symmetric hash enable configuration of port 1 to enable::
2660 testpmd> set_sym_hash_ena_per_port 1 enable
2662 get_hash_global_config
2663 ~~~~~~~~~~~~~~~~~~~~~~
2665 Get the global configurations of hash filters::
2667 get_hash_global_config (port_id)
2669 For example, to get the global configurations of hash filters of port 1::
2671 testpmd> get_hash_global_config 1
2673 set_hash_global_config
2674 ~~~~~~~~~~~~~~~~~~~~~~
2676 Set the global configurations of hash filters::
2678 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2679 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2680 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload) \
2683 For example, to enable simple_xor for flow type of ipv6 on port 2::
2685 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2690 Set the input set for hash::
2692 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2693 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2694 l2_payload) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2695 ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2696 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2697 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2698 fld-8th|none) (select|add)
2700 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2702 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2707 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2708 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2710 Set the input set for flow director::
2712 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2713 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2714 l2_payload) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2715 ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2716 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2717 sctp-dst-port|sctp-veri-tag|none) (select|add)
2719 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2721 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2726 Set different GRE key length for input set::
2728 global_config (port_id) gre-key-len (number in bytes)
2730 For example to set GRE key length for input set to 4 bytes on port 0::
2732 testpmd> global_config 0 gre-key-len 4
2735 .. _testpmd_rte_flow:
2737 Flow rules management
2738 ---------------------
2740 Control of the generic flow API (*rte_flow*) is fully exposed through the
2741 ``flow`` command (validation, creation, destruction, queries and operation
2744 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2745 features simultaneously may cause undefined side-effects and is therefore
2751 Because the ``flow`` command uses dynamic tokens to handle the large number
2752 of possible flow rules combinations, its behavior differs slightly from
2753 other commands, in particular:
2755 - Pressing *?* or the *<tab>* key displays contextual help for the current
2756 token, not that of the entire command.
2758 - Optional and repeated parameters are supported (provided they are listed
2759 in the contextual help).
2761 The first parameter stands for the operation mode. Possible operations and
2762 their general syntax are described below. They are covered in detail in the
2765 - Check whether a flow rule can be created::
2767 flow validate {port_id}
2768 [group {group_id}] [priority {level}] [ingress] [egress]
2769 pattern {item} [/ {item} [...]] / end
2770 actions {action} [/ {action} [...]] / end
2772 - Create a flow rule::
2774 flow create {port_id}
2775 [group {group_id}] [priority {level}] [ingress] [egress]
2776 pattern {item} [/ {item} [...]] / end
2777 actions {action} [/ {action} [...]] / end
2779 - Destroy specific flow rules::
2781 flow destroy {port_id} rule {rule_id} [...]
2783 - Destroy all flow rules::
2785 flow flush {port_id}
2787 - Query an existing flow rule::
2789 flow query {port_id} {rule_id} {action}
2791 - List existing flow rules sorted by priority, filtered by group
2794 flow list {port_id} [group {group_id}] [...]
2796 - Restrict ingress traffic to the defined flow rules::
2798 flow isolate {port_id} {boolean}
2800 Validating flow rules
2801 ~~~~~~~~~~~~~~~~~~~~~
2803 ``flow validate`` reports whether a flow rule would be accepted by the
2804 underlying device in its current state but stops short of creating it. It is
2805 bound to ``rte_flow_validate()``::
2807 flow validate {port_id}
2808 [group {group_id}] [priority {level}] [ingress] [egress]
2809 pattern {item} [/ {item} [...]] / end
2810 actions {action} [/ {action} [...]] / end
2812 If successful, it will show::
2816 Otherwise it will show an error message of the form::
2818 Caught error type [...] ([...]): [...]
2820 This command uses the same parameters as ``flow create``, their format is
2821 described in `Creating flow rules`_.
2823 Check whether redirecting any Ethernet packet received on port 0 to RX queue
2824 index 6 is supported::
2826 testpmd> flow validate 0 ingress pattern eth / end
2827 actions queue index 6 / end
2831 Port 0 does not support TCPv6 rules::
2833 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
2835 Caught error type 9 (specific pattern item): Invalid argument
2841 ``flow create`` validates and creates the specified flow rule. It is bound
2842 to ``rte_flow_create()``::
2844 flow create {port_id}
2845 [group {group_id}] [priority {level}] [ingress] [egress]
2846 pattern {item} [/ {item} [...]] / end
2847 actions {action} [/ {action} [...]] / end
2849 If successful, it will return a flow rule ID usable with other commands::
2851 Flow rule #[...] created
2853 Otherwise it will show an error message of the form::
2855 Caught error type [...] ([...]): [...]
2857 Parameters describe in the following order:
2859 - Attributes (*group*, *priority*, *ingress*, *egress* tokens).
2860 - A matching pattern, starting with the *pattern* token and terminated by an
2862 - Actions, starting with the *actions* token and terminated by an *end*
2865 These translate directly to *rte_flow* objects provided as-is to the
2866 underlying functions.
2868 The shortest valid definition only comprises mandatory tokens::
2870 testpmd> flow create 0 pattern end actions end
2872 Note that PMDs may refuse rules that essentially do nothing such as this
2875 **All unspecified object values are automatically initialized to 0.**
2880 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
2881 specified before the ``pattern`` token.
2883 - ``group {group id}``: priority group.
2884 - ``priority {level}``: priority level within group.
2885 - ``ingress``: rule applies to ingress traffic.
2886 - ``egress``: rule applies to egress traffic.
2888 Each instance of an attribute specified several times overrides the previous
2889 value as shown below (group 4 is used)::
2891 testpmd> flow create 0 group 42 group 24 group 4 [...]
2893 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
2895 While not specifying a direction is an error, some rules may allow both
2898 Most rules affect RX therefore contain the ``ingress`` token::
2900 testpmd> flow create 0 ingress pattern [...]
2905 A matching pattern starts after the ``pattern`` token. It is made of pattern
2906 items and is terminated by a mandatory ``end`` item.
2908 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
2909 rte_flow_item_type``).
2911 The ``/`` token is used as a separator between pattern items as shown
2914 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
2916 Note that protocol items like these must be stacked from lowest to highest
2917 layer to make sense. For instance, the following rule is either invalid or
2918 unlikely to match any packet::
2920 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
2922 More information on these restrictions can be found in the *rte_flow*
2925 Several items support additional specification structures, for example
2926 ``ipv4`` allows specifying source and destination addresses as follows::
2928 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
2929 dst is 10.2.0.0 / end [...]
2931 This rule matches all IPv4 traffic with the specified properties.
2933 In this example, ``src`` and ``dst`` are field names of the underlying
2934 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
2935 in a similar fashion.
2937 The ``is`` token means that the subsequent value must be matched exactly,
2938 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
2939 accordingly. Possible assignment tokens are:
2941 - ``is``: match value perfectly (with full bit-mask).
2942 - ``spec``: match value according to configured bit-mask.
2943 - ``last``: specify upper bound to establish a range.
2944 - ``mask``: specify bit-mask with relevant bits set to one.
2945 - ``prefix``: generate bit-mask from a prefix length.
2947 These yield identical results::
2949 ipv4 src is 10.1.1.1
2953 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
2957 ipv4 src spec 10.1.1.1 src prefix 32
2961 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
2965 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
2967 Inclusive ranges can be defined with ``last``::
2969 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
2971 Note that ``mask`` affects both ``spec`` and ``last``::
2973 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
2974 # matches 10.1.0.0 to 10.2.255.255
2976 Properties can be modified multiple times::
2978 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
2982 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
2987 This section lists supported pattern items and their attributes, if any.
2989 - ``end``: end list of pattern items.
2991 - ``void``: no-op pattern item.
2993 - ``invert``: perform actions when pattern does not match.
2995 - ``any``: match any protocol for the current layer.
2997 - ``num {unsigned}``: number of layers covered.
2999 - ``pf``: match packets addressed to the physical function.
3001 - ``vf``: match packets addressed to a virtual function ID.
3003 - ``id {unsigned}``: destination VF ID.
3005 - ``port``: device-specific physical port index to use.
3007 - ``index {unsigned}``: physical port index.
3009 - ``raw``: match an arbitrary byte string.
3011 - ``relative {boolean}``: look for pattern after the previous item.
3012 - ``search {boolean}``: search pattern from offset (see also limit).
3013 - ``offset {integer}``: absolute or relative offset for pattern.
3014 - ``limit {unsigned}``: search area limit for start of pattern.
3015 - ``pattern {string}``: byte string to look for.
3017 - ``eth``: match Ethernet header.
3019 - ``dst {MAC-48}``: destination MAC.
3020 - ``src {MAC-48}``: source MAC.
3021 - ``type {unsigned}``: EtherType.
3023 - ``vlan``: match 802.1Q/ad VLAN tag.
3025 - ``tpid {unsigned}``: tag protocol identifier.
3026 - ``tci {unsigned}``: tag control information.
3027 - ``pcp {unsigned}``: priority code point.
3028 - ``dei {unsigned}``: drop eligible indicator.
3029 - ``vid {unsigned}``: VLAN identifier.
3031 - ``ipv4``: match IPv4 header.
3033 - ``tos {unsigned}``: type of service.
3034 - ``ttl {unsigned}``: time to live.
3035 - ``proto {unsigned}``: next protocol ID.
3036 - ``src {ipv4 address}``: source address.
3037 - ``dst {ipv4 address}``: destination address.
3039 - ``ipv6``: match IPv6 header.
3041 - ``tc {unsigned}``: traffic class.
3042 - ``flow {unsigned}``: flow label.
3043 - ``proto {unsigned}``: protocol (next header).
3044 - ``hop {unsigned}``: hop limit.
3045 - ``src {ipv6 address}``: source address.
3046 - ``dst {ipv6 address}``: destination address.
3048 - ``icmp``: match ICMP header.
3050 - ``type {unsigned}``: ICMP packet type.
3051 - ``code {unsigned}``: ICMP packet code.
3053 - ``udp``: match UDP header.
3055 - ``src {unsigned}``: UDP source port.
3056 - ``dst {unsigned}``: UDP destination port.
3058 - ``tcp``: match TCP header.
3060 - ``src {unsigned}``: TCP source port.
3061 - ``dst {unsigned}``: TCP destination port.
3063 - ``sctp``: match SCTP header.
3065 - ``src {unsigned}``: SCTP source port.
3066 - ``dst {unsigned}``: SCTP destination port.
3067 - ``tag {unsigned}``: validation tag.
3068 - ``cksum {unsigned}``: checksum.
3070 - ``vxlan``: match VXLAN header.
3072 - ``vni {unsigned}``: VXLAN identifier.
3074 - ``e_tag``: match IEEE 802.1BR E-Tag header.
3076 - ``grp_ecid_b {unsigned}``: GRP and E-CID base.
3078 - ``nvgre``: match NVGRE header.
3080 - ``tni {unsigned}``: virtual subnet ID.
3082 - ``mpls``: match MPLS header.
3084 - ``label {unsigned}``: MPLS label.
3086 - ``gre``: match GRE header.
3088 - ``protocol {unsigned}``: protocol type.
3090 - ``fuzzy``: fuzzy pattern match, expect faster than default.
3092 - ``thresh {unsigned}``: accuracy threshold.
3094 - ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header.
3096 - ``teid {unsigned}``: tunnel endpoint identifier.
3101 A list of actions starts after the ``actions`` token in the same fashion as
3102 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
3103 terminated by a mandatory ``end`` action.
3105 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
3106 rte_flow_action_type``).
3108 Dropping all incoming UDPv4 packets can be expressed as follows::
3110 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3113 Several actions have configurable properties which must be specified when
3114 there is no valid default value. For example, ``queue`` requires a target
3117 This rule redirects incoming UDPv4 traffic to queue index 6::
3119 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3120 actions queue index 6 / end
3122 While this one could be rejected by PMDs (unspecified queue index)::
3124 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3127 As defined by *rte_flow*, the list is not ordered, all actions of a given
3128 rule are performed simultaneously. These are equivalent::
3130 queue index 6 / void / mark id 42 / end
3134 void / mark id 42 / queue index 6 / end
3136 All actions in a list should have different types, otherwise only the last
3137 action of a given type is taken into account::
3139 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
3143 drop / drop / drop / end # drop is performed only once
3147 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
3149 Considering they are performed simultaneously, opposite and overlapping
3150 actions can sometimes be combined when the end result is unambiguous::
3152 drop / queue index 6 / end # drop has no effect
3156 drop / dup index 6 / end # same as above
3160 queue index 6 / rss queues 6 7 8 / end # queue has no effect
3164 drop / passthru / end # drop has no effect
3166 Note that PMDs may still refuse such combinations.
3171 This section lists supported actions and their attributes, if any.
3173 - ``end``: end list of actions.
3175 - ``void``: no-op action.
3177 - ``passthru``: let subsequent rule process matched packets.
3179 - ``mark``: attach 32 bit value to packets.
3181 - ``id {unsigned}``: 32 bit value to return with packets.
3183 - ``flag``: flag packets.
3185 - ``queue``: assign packets to a given queue index.
3187 - ``index {unsigned}``: queue index to use.
3189 - ``drop``: drop packets (note: passthru has priority).
3191 - ``count``: enable counters for this rule.
3193 - ``dup``: duplicate packets to a given queue index.
3195 - ``index {unsigned}``: queue index to duplicate packets to.
3197 - ``rss``: spread packets among several queues.
3199 - ``queues [{unsigned} [...]] end``: queue indices to use.
3201 - ``pf``: redirect packets to physical device function.
3203 - ``vf``: redirect packets to virtual device function.
3205 - ``original {boolean}``: use original VF ID if possible.
3206 - ``id {unsigned}``: VF ID to redirect packets to.
3208 Destroying flow rules
3209 ~~~~~~~~~~~~~~~~~~~~~
3211 ``flow destroy`` destroys one or more rules from their rule ID (as returned
3212 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
3213 times as necessary::
3215 flow destroy {port_id} rule {rule_id} [...]
3217 If successful, it will show::
3219 Flow rule #[...] destroyed
3221 It does not report anything for rule IDs that do not exist. The usual error
3222 message is shown when a rule cannot be destroyed::
3224 Caught error type [...] ([...]): [...]
3226 ``flow flush`` destroys all rules on a device and does not take extra
3227 arguments. It is bound to ``rte_flow_flush()``::
3229 flow flush {port_id}
3231 Any errors are reported as above.
3233 Creating several rules and destroying them::
3235 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3236 actions queue index 2 / end
3237 Flow rule #0 created
3238 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3239 actions queue index 3 / end
3240 Flow rule #1 created
3241 testpmd> flow destroy 0 rule 0 rule 1
3242 Flow rule #1 destroyed
3243 Flow rule #0 destroyed
3246 The same result can be achieved using ``flow flush``::
3248 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3249 actions queue index 2 / end
3250 Flow rule #0 created
3251 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3252 actions queue index 3 / end
3253 Flow rule #1 created
3254 testpmd> flow flush 0
3257 Non-existent rule IDs are ignored::
3259 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3260 actions queue index 2 / end
3261 Flow rule #0 created
3262 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3263 actions queue index 3 / end
3264 Flow rule #1 created
3265 testpmd> flow destroy 0 rule 42 rule 10 rule 2
3267 testpmd> flow destroy 0 rule 0
3268 Flow rule #0 destroyed
3274 ``flow query`` queries a specific action of a flow rule having that
3275 ability. Such actions collect information that can be reported using this
3276 command. It is bound to ``rte_flow_query()``::
3278 flow query {port_id} {rule_id} {action}
3280 If successful, it will display either the retrieved data for known actions
3281 or the following message::
3283 Cannot display result for action type [...] ([...])
3285 Otherwise, it will complain either that the rule does not exist or that some
3288 Flow rule #[...] not found
3292 Caught error type [...] ([...]): [...]
3294 Currently only the ``count`` action is supported. This action reports the
3295 number of packets that hit the flow rule and the total number of bytes. Its
3296 output has the following format::
3299 hits_set: [...] # whether "hits" contains a valid value
3300 bytes_set: [...] # whether "bytes" contains a valid value
3301 hits: [...] # number of packets
3302 bytes: [...] # number of bytes
3304 Querying counters for TCPv6 packets redirected to queue 6::
3306 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
3307 actions queue index 6 / count / end
3308 Flow rule #4 created
3309 testpmd> flow query 0 4 count
3320 ``flow list`` lists existing flow rules sorted by priority and optionally
3321 filtered by group identifiers::
3323 flow list {port_id} [group {group_id}] [...]
3325 This command only fails with the following message if the device does not
3330 Output consists of a header line followed by a short description of each
3331 flow rule, one per line. There is no output at all when no flow rules are
3332 configured on the device::
3334 ID Group Prio Attr Rule
3335 [...] [...] [...] [...] [...]
3337 ``Attr`` column flags:
3339 - ``i`` for ``ingress``.
3340 - ``e`` for ``egress``.
3342 Creating several flow rules and listing them::
3344 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3345 actions queue index 6 / end
3346 Flow rule #0 created
3347 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3348 actions queue index 2 / end
3349 Flow rule #1 created
3350 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
3351 actions rss queues 6 7 8 end / end
3352 Flow rule #2 created
3353 testpmd> flow list 0
3354 ID Group Prio Attr Rule
3355 0 0 0 i- ETH IPV4 => QUEUE
3356 1 0 0 i- ETH IPV6 => QUEUE
3357 2 0 5 i- ETH IPV4 UDP => RSS
3360 Rules are sorted by priority (i.e. group ID first, then priority level)::
3362 testpmd> flow list 1
3363 ID Group Prio Attr Rule
3364 0 0 0 i- ETH => COUNT
3365 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3366 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3367 1 24 0 i- ETH IPV4 UDP => QUEUE
3368 4 24 10 i- ETH IPV4 TCP => DROP
3369 3 24 20 i- ETH IPV4 => DROP
3370 2 24 42 i- ETH IPV4 UDP => QUEUE
3371 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3374 Output can be limited to specific groups::
3376 testpmd> flow list 1 group 0 group 63
3377 ID Group Prio Attr Rule
3378 0 0 0 i- ETH => COUNT
3379 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3380 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3381 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3384 Toggling isolated mode
3385 ~~~~~~~~~~~~~~~~~~~~~~
3387 ``flow isolate`` can be used to tell the underlying PMD that ingress traffic
3388 must only be injected from the defined flow rules; that no default traffic
3389 is expected outside those rules and the driver is free to assign more
3390 resources to handle them. It is bound to ``rte_flow_isolate()``::
3392 flow isolate {port_id} {boolean}
3394 If successful, enabling or disabling isolated mode shows either::
3396 Ingress traffic on port [...]
3397 is now restricted to the defined flow rules
3401 Ingress traffic on port [...]
3402 is not restricted anymore to the defined flow rules
3404 Otherwise, in case of error::
3406 Caught error type [...] ([...]): [...]
3408 Mainly due to its side effects, PMDs supporting this mode may not have the
3409 ability to toggle it more than once without reinitializing affected ports
3410 first (e.g. by exiting testpmd).
3412 Enabling isolated mode::
3414 testpmd> flow isolate 0 true
3415 Ingress traffic on port 0 is now restricted to the defined flow rules
3418 Disabling isolated mode::
3420 testpmd> flow isolate 0 false
3421 Ingress traffic on port 0 is not restricted anymore to the defined flow rules
3424 Sample QinQ flow rules
3425 ~~~~~~~~~~~~~~~~~~~~~~
3427 Before creating QinQ rule(s) the following commands should be issued to enable QinQ::
3429 testpmd> port stop 0
3430 testpmd> vlan set qinq on 0
3432 The above command sets the inner and outer TPID's to 0x8100.
3434 To change the TPID's the following commands should be used::
3436 testpmd> vlan set outer tpid 0xa100 0
3437 testpmd> vlan set inner tpid 0x9100 0
3438 testpmd> port start 0
3440 Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM.
3444 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 /
3445 vlan tci is 456 / end actions vf id 1 / queue index 0 / end
3446 Flow rule #0 validated
3448 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 /
3449 vlan tci is 456 / end actions vf id 123 / queue index 0 / end
3450 Flow rule #0 created
3452 testpmd> flow list 0
3453 ID Group Prio Attr Rule
3454 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3456 Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host.
3460 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 /
3461 vlan tci is 654 / end actions pf / queue index 0 / end
3462 Flow rule #1 validated
3464 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 /
3465 vlan tci is 654 / end actions pf / queue index 1 / end
3466 Flow rule #1 created
3468 testpmd> flow list 0
3469 ID Group Prio Attr Rule
3470 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3471 1 0 0 i- ETH VLAN VLAN=>PF QUEUE