2 Copyright(c) 2010-2016 Intel Corporation. All rights reserved.
5 Redistribution and use in source and binary forms, with or without
6 modification, are permitted provided that the following conditions
9 * Redistributions of source code must retain the above copyright
10 notice, this list of conditions and the following disclaimer.
11 * Redistributions in binary form must reproduce the above copyright
12 notice, this list of conditions and the following disclaimer in
13 the documentation and/or other materials provided with the
15 * Neither the name of Intel Corporation nor the names of its
16 contributors may be used to endorse or promote products derived
17 from this software without specific prior written permission.
19 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
20 "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
21 LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
22 A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
23 OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
24 SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
25 LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
26 DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
27 THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
28 (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
29 OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
33 Testpmd Runtime Functions
34 =========================
36 Where the testpmd application is started in interactive mode, (``-i|--interactive``),
37 it displays a prompt that can be used to start and stop forwarding,
38 configure the application, display statistics (including the extended NIC
39 statistics aka xstats) , set the Flow Director and other tasks::
43 The testpmd prompt has some, limited, readline support.
44 Common bash command-line functions such as ``Ctrl+a`` and ``Ctrl+e`` to go to the start and end of the prompt line are supported
45 as well as access to the command history via the up-arrow.
47 There is also support for tab completion.
48 If you type a partial command and hit ``<TAB>`` you get a list of the available completions:
50 .. code-block:: console
52 testpmd> show port <TAB>
54 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
55 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
56 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
57 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
63 Some examples in this document are too long to fit on one line are are shown wrapped at `"\\"` for display purposes::
65 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
66 (pause_time) (send_xon) (port_id)
68 In the real ``testpmd>`` prompt these commands should be on a single line.
73 The testpmd has on-line help for the functions that are available at runtime.
74 These are divided into sections and can be accessed using help, help section or help all:
76 .. code-block:: console
80 help control : Start and stop forwarding.
81 help display : Displaying port, stats and config information.
82 help config : Configuration information.
83 help ports : Configuring ports.
84 help registers : Reading and setting port registers.
85 help filters : Filters configuration help.
86 help all : All of the above sections.
89 Command File Functions
90 ----------------------
92 To facilitate loading large number of commands or to avoid cutting and pasting where not
93 practical or possible testpmd supports alternative methods for executing commands.
95 * If started with the ``--cmdline-file=FILENAME`` command line argument testpmd
96 will execute all CLI commands contained within the file immediately before
97 starting packet forwarding or entering interactive mode.
99 .. code-block:: console
101 ./testpmd -n4 -r2 ... -- -i --cmdline-file=/home/ubuntu/flow-create-commands.txt
102 Interactive-mode selected
103 CLI commands to be read from /home/ubuntu/flow-create-commands.txt
104 Configuring Port 0 (socket 0)
105 Port 0: 7C:FE:90:CB:74:CE
106 Configuring Port 1 (socket 0)
107 Port 1: 7C:FE:90:CB:74:CA
108 Checking link statuses...
109 Port 0 Link Up - speed 10000 Mbps - full-duplex
110 Port 1 Link Up - speed 10000 Mbps - full-duplex
116 Flow rule #498 created
117 Flow rule #499 created
118 Read all CLI commands from /home/ubuntu/flow-create-commands.txt
122 * At run-time additional commands can be loaded in bulk by invoking the ``load FILENAME``
125 .. code-block:: console
127 testpmd> load /home/ubuntu/flow-create-commands.txt
132 Flow rule #498 created
133 Flow rule #499 created
134 Read all CLI commands from /home/ubuntu/flow-create-commands.txt
138 In all cases output from any included command will be displayed as standard output.
139 Execution will continue until the end of the file is reached regardless of
140 whether any errors occur. The end user must examine the output to determine if
141 any failures occurred.
150 Start packet forwarding with current configuration::
157 Start packet forwarding with current configuration after sending specified number of bursts of packets::
159 testpmd> start tx_first (""|burst_num)
161 The default burst number is 1 when ``burst_num`` not presented.
166 Stop packet forwarding, and display accumulated statistics::
181 The functions in the following sections are used to display information about the
182 testpmd configuration or the NIC status.
187 Display information for a given port or all ports::
189 testpmd> show port (info|stats|xstats|fdir|stat_qmap|dcb_tc|cap) (port_id|all)
191 The available information categories are:
193 * ``info``: General port information such as MAC address.
195 * ``stats``: RX/TX statistics.
197 * ``xstats``: RX/TX extended NIC statistics.
199 * ``fdir``: Flow Director information and statistics.
201 * ``stat_qmap``: Queue statistics mapping.
203 * ``dcb_tc``: DCB information such as TC mapping.
205 * ``cap``: Supported offload capabilities.
209 .. code-block:: console
211 testpmd> show port info 0
213 ********************* Infos for port 0 *********************
215 MAC address: XX:XX:XX:XX:XX:XX
217 memory allocation on the socket: 0
219 Link speed: 40000 Mbps
220 Link duplex: full-duplex
221 Promiscuous mode: enabled
222 Allmulticast mode: disabled
223 Maximum number of MAC addresses: 64
224 Maximum number of MAC addresses of hash filtering: 0
229 Redirection table size: 512
230 Supported flow types:
250 Display the rss redirection table entry indicated by masks on port X::
252 testpmd> show port (port_id) rss reta (size) (mask0, mask1...)
254 size is used to indicate the hardware supported reta size
259 Display the RSS hash functions and RSS hash key of a port::
261 testpmd> show port (port_id) rss-hash ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2-payload|ipv6-ex|ipv6-tcp-ex|ipv6-udp-ex [key]
266 Clear the port statistics for a given port or for all ports::
268 testpmd> clear port (info|stats|xstats|fdir|stat_qmap) (port_id|all)
272 testpmd> clear port stats all
277 Display information for a given port's RX/TX queue::
279 testpmd> show (rxq|txq) info (port_id) (queue_id)
284 Displays the configuration of the application.
285 The configuration comes from the command-line, the runtime or the application defaults::
287 testpmd> show config (rxtx|cores|fwd|txpkts)
289 The available information categories are:
291 * ``rxtx``: RX/TX configuration items.
293 * ``cores``: List of forwarding cores.
295 * ``fwd``: Packet forwarding configuration.
297 * ``txpkts``: Packets to TX configuration.
301 .. code-block:: console
303 testpmd> show config rxtx
305 io packet forwarding - CRC stripping disabled - packets/burst=16
306 nb forwarding cores=2 - nb forwarding ports=1
307 RX queues=1 - RX desc=128 - RX free threshold=0
308 RX threshold registers: pthresh=8 hthresh=8 wthresh=4
309 TX queues=1 - TX desc=512 - TX free threshold=0
310 TX threshold registers: pthresh=36 hthresh=0 wthresh=0
311 TX RS bit threshold=0 - TXQ flags=0x0
316 Set the packet forwarding mode::
318 testpmd> set fwd (io|mac|macswap|flowgen| \
319 rxonly|txonly|csum|icmpecho) (""|retry)
321 ``retry`` can be specified for forwarding engines except ``rx_only``.
323 The available information categories are:
325 * ``io``: Forwards packets "as-is" in I/O mode.
326 This is the fastest possible forwarding operation as it does not access packets data.
327 This is the default mode.
329 * ``mac``: Changes the source and the destination Ethernet addresses of packets before forwarding them.
330 Default application behaviour is to set source Ethernet address to that of the transmitting interface, and destination
331 address to a dummy value (set during init). The user may specify a target destination Ethernet address via the 'eth-peer' or
332 'eth-peer-configfile' command-line options. It is not currently possible to specify a specific source Ethernet address.
334 * ``macswap``: MAC swap forwarding mode.
335 Swaps the source and the destination Ethernet addresses of packets before forwarding them.
337 * ``flowgen``: Multi-flow generation mode.
338 Originates a number of flows (with varying destination IP addresses), and terminate receive traffic.
340 * ``rxonly``: Receives packets but doesn't transmit them.
342 * ``txonly``: Generates and transmits packets without receiving any.
344 * ``csum``: Changes the checksum field with hardware or software methods depending on the offload flags on the packet.
346 * ``icmpecho``: Receives a burst of packets, lookup for IMCP echo requests and, if any, send back ICMP echo replies.
348 * ``ieee1588``: Demonstrate L2 IEEE1588 V2 PTP timestamping for RX and TX. Requires ``CONFIG_RTE_LIBRTE_IEEE1588=y``.
350 * ``tm``: Traffic Management forwarding mode
351 Demonstrates the use of ethdev traffic management APIs and softnic PMD for
352 QoS traffic management. In this mode, 5-level hierarchical QoS scheduler is
353 available as an default option that can be enabled through CLI. The user can
354 also modify the default hierarchy or specify the new hierarchy through CLI for
355 implementing QoS scheduler. Requires ``CONFIG_RTE_LIBRTE_PMD_SOFTNIC=y`` ``CONFIG_RTE_LIBRTE_SCHED=y``.
359 testpmd> set fwd rxonly
361 Set rxonly packet forwarding mode
367 Display an RX descriptor for a port RX queue::
369 testpmd> read rxd (port_id) (queue_id) (rxd_id)
373 testpmd> read rxd 0 0 4
374 0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180
379 Display a TX descriptor for a port TX queue::
381 testpmd> read txd (port_id) (queue_id) (txd_id)
385 testpmd> read txd 0 0 4
386 0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C
391 Get loaded dynamic device personalization (DDP) package info list::
393 testpmd> ddp get list (port_id)
398 Display information about dynamic device personalization (DDP) profile::
400 testpmd> ddp get info (profile_path)
405 Display VF statistics::
407 testpmd> show vf stats (port_id) (vf_id)
412 Reset VF statistics::
414 testpmd> clear vf stats (port_id) (vf_id)
416 show port pctype mapping
417 ~~~~~~~~~~~~~~~~~~~~~~~~
419 List all items from the pctype mapping table::
421 testpmd> show port (port_id) pctype mapping
424 Configuration Functions
425 -----------------------
427 The testpmd application can be configured from the runtime as well as from the command-line.
429 This section details the available configuration functions that are available.
433 Configuration changes only become active when forwarding is started/restarted.
438 Reset forwarding to the default configuration::
445 Set the debug verbosity level::
447 testpmd> set verbose (level)
449 Currently the only available levels are 0 (silent except for error) and 1 (fully verbose).
454 Set the number of ports used by the application:
458 This is equivalent to the ``--nb-ports`` command-line option.
463 Set the number of cores used by the application::
465 testpmd> set nbcore (num)
467 This is equivalent to the ``--nb-cores`` command-line option.
471 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port.
476 Set the forwarding cores hexadecimal mask::
478 testpmd> set coremask (mask)
480 This is equivalent to the ``--coremask`` command-line option.
484 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding.
489 Set the forwarding ports hexadecimal mask::
491 testpmd> set portmask (mask)
493 This is equivalent to the ``--portmask`` command-line option.
498 Set number of packets per burst::
500 testpmd> set burst (num)
502 This is equivalent to the ``--burst command-line`` option.
504 When retry is enabled, the transmit delay time and number of retries can also be set::
506 testpmd> set burst tx delay (microseconds) retry (num)
511 Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode::
513 testpmd> set txpkts (x[,y]*)
515 Where x[,y]* represents a CSV list of values, without white space.
520 Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes::
522 testpmd> set txsplit (off|on|rand)
526 * ``off`` disable packet copy & split for CSUM mode.
528 * ``on`` split outgoing packet into multiple segments. Size of each segment
529 and number of segments per packet is determined by ``set txpkts`` command
532 * ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments.
537 Set the list of forwarding cores::
539 testpmd> set corelist (x[,y]*)
541 For example, to change the forwarding cores:
543 .. code-block:: console
545 testpmd> set corelist 3,1
546 testpmd> show config fwd
548 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled
549 Logical Core 3 (socket 0) forwards packets on 1 streams:
550 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01
551 Logical Core 1 (socket 0) forwards packets on 1 streams:
552 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
556 The cores are used in the same order as specified on the command line.
561 Set the list of forwarding ports::
563 testpmd> set portlist (x[,y]*)
565 For example, to change the port forwarding:
567 .. code-block:: console
569 testpmd> set portlist 0,2,1,3
570 testpmd> show config fwd
572 io packet forwarding - ports=4 - cores=1 - streams=4
573 Logical Core 3 (socket 0) forwards packets on 4 streams:
574 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01
575 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
576 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03
577 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02
582 Enable/disable tx loopback::
584 testpmd> set tx loopback (port_id) (on|off)
589 set drop enable bit for all queues::
591 testpmd> set all queues drop (port_id) (on|off)
593 set split drop enable (for VF)
594 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
596 set split drop enable bit for VF from PF::
598 testpmd> set vf split drop (port_id) (vf_id) (on|off)
600 set mac antispoof (for VF)
601 ~~~~~~~~~~~~~~~~~~~~~~~~~~
603 Set mac antispoof for a VF from the PF::
605 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off)
610 Enable/disable MACsec offload::
612 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off)
613 testpmd> set macsec offload (port_id) off
618 Configure MACsec secure connection (SC)::
620 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi)
624 The pi argument is ignored for tx.
625 Check the NIC Datasheet for hardware limits.
630 Configure MACsec secure association (SA)::
632 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key)
636 The IDX value must be 0 or 1.
637 Check the NIC Datasheet for hardware limits.
639 set broadcast mode (for VF)
640 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
642 Set broadcast mode for a VF from the PF::
644 testpmd> set vf broadcast (port_id) (vf_id) (on|off)
649 Set the VLAN strip on a port::
651 testpmd> vlan set strip (on|off) (port_id)
656 Set the VLAN strip for a queue on a port::
658 testpmd> vlan set stripq (on|off) (port_id,queue_id)
660 vlan set stripq (for VF)
661 ~~~~~~~~~~~~~~~~~~~~~~~~
663 Set VLAN strip for all queues in a pool for a VF from the PF::
665 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off)
667 vlan set insert (for VF)
668 ~~~~~~~~~~~~~~~~~~~~~~~~
670 Set VLAN insert for a VF from the PF::
672 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id)
674 vlan set tag (for VF)
675 ~~~~~~~~~~~~~~~~~~~~~
677 Set VLAN tag for a VF from the PF::
679 testpmd> set vf vlan tag (port_id) (vf_id) (on|off)
681 vlan set antispoof (for VF)
682 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
684 Set VLAN antispoof for a VF from the PF::
686 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off)
691 Set the VLAN filter on a port::
693 testpmd> vlan set filter (on|off) (port_id)
698 Set the VLAN QinQ (extended queue in queue) on for a port::
700 testpmd> vlan set qinq (on|off) (port_id)
705 Set the inner or outer VLAN TPID for packet filtering on a port::
707 testpmd> vlan set (inner|outer) tpid (value) (port_id)
711 TPID value must be a 16-bit number (value <= 65536).
716 Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID::
718 testpmd> rx_vlan add (vlan_id|all) (port_id)
722 VLAN filter must be set on that port. VLAN ID < 4096.
723 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries
724 in VFTA table. This is important if enabling all vlan_ids.
729 Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID::
731 testpmd> rx_vlan rm (vlan_id|all) (port_id)
736 Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID::
738 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask)
743 Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID::
745 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask)
750 Add a tunnel filter on a port::
752 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \
753 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
754 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
756 The available information categories are:
758 * ``vxlan``: Set tunnel type as VXLAN.
760 * ``nvgre``: Set tunnel type as NVGRE.
762 * ``ipingre``: Set tunnel type as IP-in-GRE.
764 * ``imac-ivlan``: Set filter type as Inner MAC and VLAN.
766 * ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID.
768 * ``imac-tenid``: Set filter type as Inner MAC and tenant ID.
770 * ``imac``: Set filter type as Inner MAC.
772 * ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID.
774 * ``oip``: Set filter type as Outer IP.
776 * ``iip``: Set filter type as Inner IP.
780 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \
781 192.168.2.2 0 ipingre oip 1 1
783 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP.
788 Remove a tunnel filter on a port::
790 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \
791 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
792 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
797 Add an UDP port for VXLAN packet filter on a port::
799 testpmd> rx_vxlan_port add (udp_port) (port_id)
804 Remove an UDP port for VXLAN packet filter on a port::
806 testpmd> rx_vxlan_port rm (udp_port) (port_id)
811 Set hardware insertion of VLAN IDs in packets sent on a port::
813 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer]
815 For example, set a single VLAN ID (5) insertion on port 0::
819 Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1::
827 Set port based hardware insertion of VLAN ID in packets sent on a port::
829 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off)
834 Disable hardware insertion of a VLAN header in packets sent on a port::
836 testpmd> tx_vlan reset (port_id)
841 Select hardware or software calculation of the checksum when
842 transmitting a packet using the ``csum`` forwarding engine::
844 testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id)
848 * ``ip|udp|tcp|sctp`` always relate to the inner layer.
850 * ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized
851 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are
852 supported). See also the ``csum parse-tunnel`` command.
856 Check the NIC Datasheet for hardware limits.
861 Set RSS queue region span on a port::
863 testpmd> set port (port_id) queue-region region_id (value) \
864 queue_start_index (value) queue_num (value)
866 Set flowtype mapping on a RSS queue region on a port::
868 testpmd> set port (port_id) queue-region region_id (value) flowtype (value)
872 * For the flowtype(pctype) of packet,the specific index for each type has
873 been defined in file i40e_type.h as enum i40e_filter_pctype.
875 Set user priority mapping on a RSS queue region on a port::
877 testpmd> set port (port_id) queue-region UP (value) region_id (value)
879 Flush all queue region related configuration on a port::
881 testpmd> set port (port_id) queue-region flush (on|off)
885 * "on"is just an enable function which server for other configuration,
886 it is for all configuration about queue region from up layer,
887 at first will only keep in DPDK softwarestored in driver,
888 only after "flush on", it commit all configuration to HW.
889 "off" is just clean all configuration about queue region just now,
890 and restore all to DPDK i40e driver default config when start up.
892 Show all queue region related configuration info on a port::
894 testpmd> show port (port_id) queue-region
898 Queue region only support on PF by now, so these command is
899 only for configuration of queue region on PF port.
904 Define how tunneled packets should be handled by the csum forward
907 testpmd> csum parse-tunnel (on|off) (tx_port_id)
909 If enabled, the csum forward engine will try to recognize supported
910 tunnel headers (vxlan, gre, ipip).
912 If disabled, treat tunnel packets as non-tunneled packets (a inner
913 header is handled as a packet payload).
917 The port argument is the TX port like in the ``csum set`` command.
921 Consider a packet in packet like the following::
923 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in
925 * If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
926 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the
927 ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``).
929 * If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
930 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``.
935 Display tx checksum offload configuration::
937 testpmd> csum show (port_id)
942 Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine::
944 testpmd> tso set (segsize) (port_id)
948 Check the NIC datasheet for hardware limits.
953 Display the status of TCP Segmentation Offload::
955 testpmd> tso show (port_id)
960 Enable or disable GRO in ``csum`` forwarding engine::
962 testpmd> set port <port_id> gro on|off
964 If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4
965 packets received from the given port.
967 If disabled, packets received from the given port won't be performed
968 GRO. By default, GRO is disabled for all ports.
972 When enable GRO for a port, TCP/IPv4 packets received from the port
973 will be performed GRO. After GRO, all merged packets have bad
974 checksums, since the GRO library doesn't re-calculate checksums for
975 the merged packets. Therefore, if users want the merged packets to
976 have correct checksums, please select HW IP checksum calculation and
977 HW TCP checksum calculation for the port which the merged packets are
983 Display GRO configuration for a given port::
985 testpmd> show port <port_id> gro
990 Set the cycle to flush the GROed packets from reassembly tables::
992 testpmd> set gro flush <cycles>
994 When enable GRO, the csum forwarding engine performs GRO on received
995 packets, and the GROed packets are stored in reassembly tables. Users
996 can use this command to determine when the GROed packets are flushed
997 from the reassembly tables.
999 The ``cycles`` is measured in GRO operation times. The csum forwarding
1000 engine flushes the GROed packets from the tables every ``cycles`` GRO
1003 By default, the value of ``cycles`` is 1, which means flush GROed packets
1004 from the reassembly tables as soon as one GRO operation finishes. The value
1005 of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``.
1007 Please note that the large value of ``cycles`` may cause the poor TCP/IP
1008 stack performance. Because the GROed packets are delayed to arrive the
1009 stack, thus causing more duplicated ACKs and TCP retransmissions.
1014 Toggle per-port GSO support in ``csum`` forwarding engine::
1016 testpmd> set port <port_id> gso on|off
1018 If enabled, the csum forwarding engine will perform GSO on supported IPv4
1019 packets, transmitted on the given port.
1021 If disabled, packets transmitted on the given port will not undergo GSO.
1022 By default, GSO is disabled for all ports.
1026 When GSO is enabled on a port, supported IPv4 packets transmitted on that
1027 port undergo GSO. Afterwards, the segmented packets are represented by
1028 multi-segment mbufs; however, the csum forwarding engine doesn't calculation
1029 of checksums for GSO'd segments in SW. As a result, if users want correct
1030 checksums in GSO segments, they should enable HW checksum calculation for
1033 For example, HW checksum calculation for VxLAN GSO'd packets may be enabled
1034 by setting the following options in the csum forwarding engine:
1036 testpmd> csum set outer_ip hw <port_id>
1038 testpmd> csum set ip hw <port_id>
1040 testpmd> csum set tcp hw <port_id>
1045 Set the maximum GSO segment size (measured in bytes), which includes the
1046 packet header and the packet payload for GSO-enabled ports (global)::
1048 testpmd> set gso segsz <length>
1053 Display the status of Generic Segmentation Offload for a given port::
1055 testpmd> show port <port_id> gso
1060 Add an alternative MAC address to a port::
1062 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX)
1067 Remove a MAC address from a port::
1069 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX)
1071 mac_addr add (for VF)
1072 ~~~~~~~~~~~~~~~~~~~~~
1074 Add an alternative MAC address for a VF to a port::
1076 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX)
1081 Set the default MAC address for a port::
1083 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX)
1085 mac_addr set (for VF)
1086 ~~~~~~~~~~~~~~~~~~~~~
1088 Set the MAC address for a VF from the PF::
1090 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX)
1095 Set the forwarding peer address for certain port::
1097 testpmd> set eth-peer (port_id) (perr_addr)
1099 This is equivalent to the ``--eth-peer`` command-line option.
1104 Set the unicast hash filter(s) on/off for a port::
1106 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
1111 Set the promiscuous mode on for a port or for all ports.
1112 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1114 testpmd> set promisc (port_id|all) (on|off)
1119 Set the allmulti mode for a port or for all ports::
1121 testpmd> set allmulti (port_id|all) (on|off)
1123 Same as the ifconfig (8) option. Controls how multicast packets are handled.
1125 set promisc (for VF)
1126 ~~~~~~~~~~~~~~~~~~~~
1128 Set the unicast promiscuous mode for a VF from PF.
1129 It's supported by Intel i40e NICs now.
1130 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1132 testpmd> set vf promisc (port_id) (vf_id) (on|off)
1134 set allmulticast (for VF)
1135 ~~~~~~~~~~~~~~~~~~~~~~~~~
1137 Set the multicast promiscuous mode for a VF from PF.
1138 It's supported by Intel i40e NICs now.
1139 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1141 testpmd> set vf allmulti (port_id) (vf_id) (on|off)
1143 set tx max bandwidth (for VF)
1144 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1146 Set TX max absolute bandwidth (Mbps) for a VF from PF::
1148 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth)
1150 set tc tx min bandwidth (for VF)
1151 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1153 Set all TCs' TX min relative bandwidth (%) for a VF from PF::
1155 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...)
1157 set tc tx max bandwidth (for VF)
1158 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1160 Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF::
1162 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth)
1164 set tc strict link priority mode
1165 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1167 Set some TCs' strict link priority mode on a physical port::
1169 testpmd> set tx strict-link-priority (port_id) (tc_bitmap)
1171 set tc tx min bandwidth
1172 ~~~~~~~~~~~~~~~~~~~~~~~
1174 Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs::
1176 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...)
1181 Set the link flow control parameter on a port::
1183 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1184 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
1185 autoneg (on|off) (port_id)
1189 * ``high_water`` (integer): High threshold value to trigger XOFF.
1191 * ``low_water`` (integer): Low threshold value to trigger XON.
1193 * ``pause_time`` (integer): Pause quota in the Pause frame.
1195 * ``send_xon`` (0/1): Send XON frame.
1197 * ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
1199 * ``autoneg``: Change the auto-negotiation parameter.
1204 Set the priority flow control parameter on a port::
1206 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1207 (pause_time) (priority) (port_id)
1211 * ``high_water`` (integer): High threshold value.
1213 * ``low_water`` (integer): Low threshold value.
1215 * ``pause_time`` (integer): Pause quota in the Pause frame.
1217 * ``priority`` (0-7): VLAN User Priority.
1222 Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
1224 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
1226 For example, to set rx queue 2 on port 0 to mapping 5::
1228 testpmd>set stat_qmap rx 0 2 5
1230 set xstats-hide-zero
1231 ~~~~~~~~~~~~~~~~~~~~
1233 Set the option to hide zero values for xstats display::
1235 testpmd> set xstats-hide-zero on|off
1239 By default, the zero values are displayed for xstats.
1241 set port - rx/tx (for VF)
1242 ~~~~~~~~~~~~~~~~~~~~~~~~~
1244 Set VF receive/transmit from a port::
1246 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
1248 set port - mac address filter (for VF)
1249 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1251 Add/Remove unicast or multicast MAC addr filter for a VF::
1253 testpmd> set port (port_id) vf (vf_id) (mac_addr) \
1254 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
1256 set port - rx mode(for VF)
1257 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1259 Set the VF receive mode of a port::
1261 testpmd> set port (port_id) vf (vf_id) \
1262 rxmode (AUPE|ROPE|BAM|MPE) (on|off)
1264 The available receive modes are:
1266 * ``AUPE``: Accepts untagged VLAN.
1268 * ``ROPE``: Accepts unicast hash.
1270 * ``BAM``: Accepts broadcast packets.
1272 * ``MPE``: Accepts all multicast packets.
1274 set port - tx_rate (for Queue)
1275 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1277 Set TX rate limitation for a queue on a port::
1279 testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1281 set port - tx_rate (for VF)
1282 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
1284 Set TX rate limitation for queues in VF on a port::
1286 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1288 set port - mirror rule
1289 ~~~~~~~~~~~~~~~~~~~~~~
1291 Set pool or vlan type mirror rule for a port::
1293 testpmd> set port (port_id) mirror-rule (rule_id) \
1294 (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1295 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1297 Set link mirror rule for a port::
1299 testpmd> set port (port_id) mirror-rule (rule_id) \
1300 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1302 For example to enable mirror traffic with vlan 0,1 to pool 0::
1304 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1306 reset port - mirror rule
1307 ~~~~~~~~~~~~~~~~~~~~~~~~
1309 Reset a mirror rule for a port::
1311 testpmd> reset port (port_id) mirror-rule (rule_id)
1316 Set the flush on RX streams before forwarding.
1317 The default is flush ``on``.
1318 Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1320 testpmd> set flush_rx off
1325 Set the bypass mode for the lowest port on bypass enabled NIC::
1327 testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1332 Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1334 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1335 mode (normal|bypass|isolate) (port_id)
1339 * ``timeout``: Enable bypass after watchdog timeout.
1341 * ``os_on``: Enable bypass when OS/board is powered on.
1343 * ``os_off``: Enable bypass when OS/board is powered off.
1345 * ``power_on``: Enable bypass when power supply is turned on.
1347 * ``power_off``: Enable bypass when power supply is turned off.
1353 Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1355 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1360 Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1362 testpmd> show bypass config (port_id)
1367 Set link up for a port::
1369 testpmd> set link-up port (port id)
1374 Set link down for a port::
1376 testpmd> set link-down port (port id)
1381 Enable E-tag insertion for a VF on a port::
1383 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1385 Disable E-tag insertion for a VF on a port::
1387 testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1389 Enable/disable E-tag stripping on a port::
1391 testpmd> E-tag set stripping (on|off) port (port_id)
1393 Enable/disable E-tag based forwarding on a port::
1395 testpmd> E-tag set forwarding (on|off) port (port_id)
1397 Add an E-tag forwarding filter on a port::
1399 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1401 Delete an E-tag forwarding filter on a port::
1402 testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1407 Load a dynamic device personalization (DDP) package::
1409 testpmd> ddp add (port_id) (package_path[,output_path])
1414 Delete a dynamic device personalization package::
1416 testpmd> ddp del (port_id) (package_path)
1421 List all items from the ptype mapping table::
1423 testpmd> ptype mapping get (port_id) (valid_only)
1427 * ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0).
1429 Replace a specific or a group of software defined ptype with a new one::
1431 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type)
1435 * ``target``: A specific software ptype or a mask to represent a group of software ptypes.
1437 * ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1).
1439 * ``pkt_type``: The new software ptype to replace the old ones.
1441 Update hardware defined ptype to software defined packet type mapping table::
1443 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype)
1447 * ``hw_ptype``: hardware ptype as the index of the ptype mapping table.
1449 * ``sw_ptype``: software ptype as the value of the ptype mapping table.
1451 Reset ptype mapping table::
1453 testpmd> ptype mapping reset (port_id)
1458 The following sections show functions for configuring ports.
1462 Port configuration changes only become active when forwarding is started/restarted.
1467 Attach a port specified by pci address or virtual device args::
1469 testpmd> port attach (identifier)
1471 To attach a new pci device, the device should be recognized by kernel first.
1472 Then it should be moved under DPDK management.
1473 Finally the port can be attached to testpmd.
1475 For example, to move a pci device using ixgbe under DPDK management:
1477 .. code-block:: console
1479 # Check the status of the available devices.
1480 ./usertools/dpdk-devbind.py --status
1482 Network devices using DPDK-compatible driver
1483 ============================================
1486 Network devices using kernel driver
1487 ===================================
1488 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1491 # Bind the device to igb_uio.
1492 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1495 # Recheck the status of the devices.
1496 ./usertools/dpdk-devbind.py --status
1497 Network devices using DPDK-compatible driver
1498 ============================================
1499 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1501 To attach a port created by virtual device, above steps are not needed.
1503 For example, to attach a port whose pci address is 0000:0a:00.0.
1505 .. code-block:: console
1507 testpmd> port attach 0000:0a:00.0
1508 Attaching a new port...
1509 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1510 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1511 EAL: PCI memory mapped at 0x7f83bfa00000
1512 EAL: PCI memory mapped at 0x7f83bfa80000
1513 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1514 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1515 Port 0 is attached. Now total ports is 1
1518 For example, to attach a port created by pcap PMD.
1520 .. code-block:: console
1522 testpmd> port attach net_pcap0
1523 Attaching a new port...
1524 PMD: Initializing pmd_pcap for net_pcap0
1525 PMD: Creating pcap-backed ethdev on numa socket 0
1526 Port 0 is attached. Now total ports is 1
1529 In this case, identifier is ``net_pcap0``.
1530 This identifier format is the same as ``--vdev`` format of DPDK applications.
1532 For example, to re-attach a bonded port which has been previously detached,
1533 the mode and slave parameters must be given.
1535 .. code-block:: console
1537 testpmd> port attach net_bond_0,mode=0,slave=1
1538 Attaching a new port...
1539 EAL: Initializing pmd_bond for net_bond_0
1540 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1541 Port 0 is attached. Now total ports is 1
1548 Detach a specific port::
1550 testpmd> port detach (port_id)
1552 Before detaching a port, the port should be stopped and closed.
1554 For example, to detach a pci device port 0.
1556 .. code-block:: console
1558 testpmd> port stop 0
1561 testpmd> port close 0
1565 testpmd> port detach 0
1567 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1568 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1569 EAL: PCI memory unmapped at 0x7f83bfa00000
1570 EAL: PCI memory unmapped at 0x7f83bfa80000
1574 For example, to detach a virtual device port 0.
1576 .. code-block:: console
1578 testpmd> port stop 0
1581 testpmd> port close 0
1585 testpmd> port detach 0
1587 PMD: Closing pcap ethdev on numa socket 0
1588 Port 'net_pcap0' is detached. Now total ports is 0
1591 To remove a pci device completely from the system, first detach the port from testpmd.
1592 Then the device should be moved under kernel management.
1593 Finally the device can be removed using kernel pci hotplug functionality.
1595 For example, to move a pci device under kernel management:
1597 .. code-block:: console
1599 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1601 ./usertools/dpdk-devbind.py --status
1603 Network devices using DPDK-compatible driver
1604 ============================================
1607 Network devices using kernel driver
1608 ===================================
1609 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1611 To remove a port created by a virtual device, above steps are not needed.
1616 Start all ports or a specific port::
1618 testpmd> port start (port_id|all)
1623 Stop all ports or a specific port::
1625 testpmd> port stop (port_id|all)
1630 Close all ports or a specific port::
1632 testpmd> port close (port_id|all)
1634 port start/stop queue
1635 ~~~~~~~~~~~~~~~~~~~~~
1637 Start/stop a rx/tx queue on a specific port::
1639 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1641 Only take effect when port is started.
1646 Set the speed and duplex mode for all ports or a specific port::
1648 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1649 duplex (half|full|auto)
1651 port config - queues/descriptors
1652 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1654 Set number of queues/descriptors for rxq, txq, rxd and txd::
1656 testpmd> port config all (rxq|txq|rxd|txd) (value)
1658 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1660 port config - max-pkt-len
1661 ~~~~~~~~~~~~~~~~~~~~~~~~~
1663 Set the maximum packet length::
1665 testpmd> port config all max-pkt-len (value)
1667 This is equivalent to the ``--max-pkt-len`` command-line option.
1669 port config - CRC Strip
1670 ~~~~~~~~~~~~~~~~~~~~~~~
1672 Set hardware CRC stripping on or off for all ports::
1674 testpmd> port config all crc-strip (on|off)
1676 CRC stripping is on by default.
1678 The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option.
1680 port config - scatter
1681 ~~~~~~~~~~~~~~~~~~~~~~~
1683 Set RX scatter mode on or off for all ports::
1685 testpmd> port config all scatter (on|off)
1687 RX scatter mode is off by default.
1689 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1691 port config - RX Checksum
1692 ~~~~~~~~~~~~~~~~~~~~~~~~~
1694 Set hardware RX checksum offload to on or off for all ports::
1696 testpmd> port config all rx-cksum (on|off)
1698 Checksum offload is off by default.
1700 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1705 Set hardware VLAN on or off for all ports::
1707 testpmd> port config all hw-vlan (on|off)
1709 Hardware VLAN is on by default.
1711 The ``off`` option is equivalent to the ``--disable-hw-vlan`` command-line option.
1713 port config - VLAN filter
1714 ~~~~~~~~~~~~~~~~~~~~~~~~~
1716 Set hardware VLAN filter on or off for all ports::
1718 testpmd> port config all hw-vlan-filter (on|off)
1720 Hardware VLAN filter is on by default.
1722 The ``off`` option is equivalent to the ``--disable-hw-vlan-filter`` command-line option.
1724 port config - VLAN strip
1725 ~~~~~~~~~~~~~~~~~~~~~~~~
1727 Set hardware VLAN strip on or off for all ports::
1729 testpmd> port config all hw-vlan-strip (on|off)
1731 Hardware VLAN strip is on by default.
1733 The ``off`` option is equivalent to the ``--disable-hw-vlan-strip`` command-line option.
1735 port config - VLAN extend
1736 ~~~~~~~~~~~~~~~~~~~~~~~~~
1738 Set hardware VLAN extend on or off for all ports::
1740 testpmd> port config all hw-vlan-extend (on|off)
1742 Hardware VLAN extend is off by default.
1744 The ``off`` option is equivalent to the ``--disable-hw-vlan-extend`` command-line option.
1746 port config - Drop Packets
1747 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1749 Set packet drop for packets with no descriptors on or off for all ports::
1751 testpmd> port config all drop-en (on|off)
1753 Packet dropping for packets with no descriptors is off by default.
1755 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1760 Set the RSS (Receive Side Scaling) mode on or off::
1762 testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1764 RSS is on by default.
1766 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1768 port config - RSS Reta
1769 ~~~~~~~~~~~~~~~~~~~~~~
1771 Set the RSS (Receive Side Scaling) redirection table::
1773 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1778 Set the DCB mode for an individual port::
1780 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1782 The traffic class should be 4 or 8.
1787 Set the number of packets per burst::
1789 testpmd> port config all burst (value)
1791 This is equivalent to the ``--burst`` command-line option.
1793 port config - Threshold
1794 ~~~~~~~~~~~~~~~~~~~~~~~
1796 Set thresholds for TX/RX queues::
1798 testpmd> port config all (threshold) (value)
1800 Where the threshold type can be:
1802 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1804 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1806 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1808 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1810 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1812 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1814 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1816 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1818 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1820 These threshold options are also available from the command-line.
1825 Set the value of ether-type for E-tag::
1827 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1829 Enable/disable the E-tag support::
1831 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1833 port config pctype mapping
1834 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1836 Reset pctype mapping table::
1838 testpmd> port config (port_id) pctype mapping reset
1840 Update hardware defined pctype to software defined flow type mapping table::
1842 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id)
1846 * ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table.
1848 * ``flow_type_id``: software flow type id as the index of the pctype mapping table.
1850 port config input set
1851 ~~~~~~~~~~~~~~~~~~~~~
1853 Config RSS/FDIR/FDIR flexible payload input set for some pctype::
1854 testpmd> port config (port_id) pctype (pctype_id) \
1855 (hash_inset|fdir_inset|fdir_flx_inset) \
1856 (get|set|clear) field (field_idx)
1858 Clear RSS/FDIR/FDIR flexible payload input set for some pctype::
1859 testpmd> port config (port_id) pctype (pctype_id) \
1860 (hash_inset|fdir_inset|fdir_flx_inset) clear all
1864 * ``pctype_id``: hardware packet classification types.
1865 * ``field_idx``: hardware field index.
1867 Link Bonding Functions
1868 ----------------------
1870 The Link Bonding functions make it possible to dynamically create and
1871 manage link bonding devices from within testpmd interactive prompt.
1873 create bonded device
1874 ~~~~~~~~~~~~~~~~~~~~
1876 Create a new bonding device::
1878 testpmd> create bonded device (mode) (socket)
1880 For example, to create a bonded device in mode 1 on socket 0::
1882 testpmd> create bonded 1 0
1883 created new bonded device (port X)
1888 Adds Ethernet device to a Link Bonding device::
1890 testpmd> add bonding slave (slave id) (port id)
1892 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
1894 testpmd> add bonding slave 6 10
1897 remove bonding slave
1898 ~~~~~~~~~~~~~~~~~~~~
1900 Removes an Ethernet slave device from a Link Bonding device::
1902 testpmd> remove bonding slave (slave id) (port id)
1904 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
1906 testpmd> remove bonding slave 6 10
1911 Set the Link Bonding mode of a Link Bonding device::
1913 testpmd> set bonding mode (value) (port id)
1915 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
1917 testpmd> set bonding mode 3 10
1922 Set an Ethernet slave device as the primary device on a Link Bonding device::
1924 testpmd> set bonding primary (slave id) (port id)
1926 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
1928 testpmd> set bonding primary 6 10
1933 Set the MAC address of a Link Bonding device::
1935 testpmd> set bonding mac (port id) (mac)
1937 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
1939 testpmd> set bonding mac 10 00:00:00:00:00:01
1941 set bonding xmit_balance_policy
1942 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1944 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
1946 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
1948 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
1950 testpmd> set bonding xmit_balance_policy 10 l34
1953 set bonding mon_period
1954 ~~~~~~~~~~~~~~~~~~~~~~
1956 Set the link status monitoring polling period in milliseconds for a bonding device.
1958 This adds support for PMD slave devices which do not support link status interrupts.
1959 When the mon_period is set to a value greater than 0 then all PMD's which do not support
1960 link status ISR will be queried every polling interval to check if their link status has changed::
1962 testpmd> set bonding mon_period (port_id) (value)
1964 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
1966 testpmd> set bonding mon_period 5 150
1969 set bonding lacp dedicated_queue
1970 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1972 Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic
1973 when in mode 4 (link-aggregration-802.3ad)::
1975 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable)
1978 set bonding agg_mode
1979 ~~~~~~~~~~~~~~~~~~~~
1981 Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad)::
1983 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable)
1989 Show the current configuration of a Link Bonding device::
1991 testpmd> show bonding config (port id)
1994 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
1995 in balance mode with a transmission policy of layer 2+3::
1997 testpmd> show bonding config 9
1999 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
2001 Active Slaves (3): [1 3 4]
2008 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
2009 This is mainly useful for debugging purposes.
2010 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
2011 and fields that can be accessed.
2016 Display the value of a port register::
2018 testpmd> read reg (port_id) (address)
2020 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
2022 testpmd> read reg 0 0xEE00
2023 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
2028 Display a port register bit field::
2030 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
2032 For example, reading the lowest two bits from the register in the example above::
2034 testpmd> read regfield 0 0xEE00 0 1
2035 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
2040 Display a single port register bit::
2042 testpmd> read regbit (port_id) (address) (bit_x)
2044 For example, reading the lowest bit from the register in the example above::
2046 testpmd> read regbit 0 0xEE00 0
2047 port 0 PCI register at offset 0xEE00: bit 0=1
2052 Set the value of a port register::
2054 testpmd> write reg (port_id) (address) (value)
2056 For example, to clear a register::
2058 testpmd> write reg 0 0xEE00 0x0
2059 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
2064 Set bit field of a port register::
2066 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
2068 For example, writing to the register cleared in the example above::
2070 testpmd> write regfield 0 0xEE00 0 1 2
2071 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
2076 Set single bit value of a port register::
2078 testpmd> write regbit (port_id) (address) (bit_x) (value)
2080 For example, to set the high bit in the register from the example above::
2082 testpmd> write regbit 0 0xEE00 31 1
2083 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
2085 Traffic Metering and Policing
2086 -----------------------------
2088 The following section shows functions for configuring traffic metering and
2089 policing on the ethernet device through the use of generic ethdev API.
2091 show port traffic management capability
2092 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2094 Show traffic metering and policing capability of the port::
2096 testpmd> show port meter cap (port_id)
2098 add port meter profile (srTCM rfc2967)
2099 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2101 Add meter profile (srTCM rfc2697) to the ethernet device::
2103 testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \
2108 * ``profile_id``: ID for the meter profile.
2109 * ``cir``: Committed Information Rate (CIR) (bytes/second).
2110 * ``cbs``: Committed Burst Size (CBS) (bytes).
2111 * ``ebs``: Excess Burst Size (EBS) (bytes).
2113 add port meter profile (trTCM rfc2968)
2114 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2116 Add meter profile (srTCM rfc2698) to the ethernet device::
2118 testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \
2119 (cir) (pir) (cbs) (pbs)
2123 * ``profile_id``: ID for the meter profile.
2124 * ``cir``: Committed information rate (bytes/second).
2125 * ``pir``: Peak information rate (bytes/second).
2126 * ``cbs``: Committed burst size (bytes).
2127 * ``pbs``: Peak burst size (bytes).
2129 add port meter profile (trTCM rfc4115)
2130 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2132 Add meter profile (trTCM rfc4115) to the ethernet device::
2134 testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \
2135 (cir) (eir) (cbs) (ebs)
2139 * ``profile_id``: ID for the meter profile.
2140 * ``cir``: Committed information rate (bytes/second).
2141 * ``eir``: Excess information rate (bytes/second).
2142 * ``cbs``: Committed burst size (bytes).
2143 * ``ebs``: Excess burst size (bytes).
2145 delete port meter profile
2146 ~~~~~~~~~~~~~~~~~~~~~~~~~
2148 Delete meter profile from the ethernet device::
2150 testpmd> del port meter profile (port_id) (profile_id)
2155 Create new meter object for the ethernet device::
2157 testpmd> create port meter (port_id) (mtr_id) (profile_id) \
2158 (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \
2159 (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\
2164 * ``mtr_id``: meter object ID.
2165 * ``profile_id``: ID for the meter profile.
2166 * ``meter_enable``: When this parameter has a non-zero value, the meter object
2167 gets enabled at the time of creation, otherwise remains disabled.
2168 * ``g_action``: Policer action for the packet with green color.
2169 * ``y_action``: Policer action for the packet with yellow color.
2170 * ``r_action``: Policer action for the packet with red color.
2171 * ``stats_mask``: Mask of statistics counter types to be enabled for the
2173 * ``shared``: When this parameter has a non-zero value, the meter object is
2174 shared by multiple flows. Otherwise, meter object is used by single flow.
2175 * ``use_pre_meter_color``: When this parameter has a non-zero value, the
2176 input color for the current meter object is determined by the latest meter
2177 object in the same flow. Otherwise, the current meter object uses the
2178 *dscp_table* to determine the input color.
2179 * ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input
2180 color, 0 <= x <= 63.
2185 Enable meter for the ethernet device::
2187 testpmd> enable port meter (port_id) (mtr_id)
2192 Disable meter for the ethernet device::
2194 testpmd> disable port meter (port_id) (mtr_id)
2199 Delete meter for the ethernet device::
2201 testpmd> del port meter (port_id) (mtr_id)
2203 Set port meter profile
2204 ~~~~~~~~~~~~~~~~~~~~~~
2206 Set meter profile for the ethernet device::
2208 testpmd> set port meter profile (port_id) (mtr_id) (profile_id)
2210 set port meter dscp table
2211 ~~~~~~~~~~~~~~~~~~~~~~~~~
2213 Set meter dscp table for the ethernet device::
2215 testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \
2216 (dscp_tbl_entry1)...(dscp_tbl_entry63)]
2218 set port meter policer action
2219 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2221 Set meter policer action for the ethernet device::
2223 testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \
2224 (action0) [(action1) (action1)]
2228 * ``action_mask``: Bit mask indicating which policer actions need to be
2229 updated. One or more policer actions can be updated in a single function
2230 invocation. To update the policer action associated with color C, bit
2231 (1 << C) needs to be set in *action_mask* and element at position C
2232 in the *actions* array needs to be valid.
2233 * ``actionx``: Policer action for the color x,
2234 RTE_MTR_GREEN <= x < RTE_MTR_COLORS
2236 set port meter stats mask
2237 ~~~~~~~~~~~~~~~~~~~~~~~~~
2239 Set meter stats mask for the ethernet device::
2241 testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask)
2245 * ``stats_mask``: Bit mask indicating statistics counter types to be enabled.
2247 show port meter stats
2248 ~~~~~~~~~~~~~~~~~~~~~
2250 Show meter stats of the ethernet device::
2252 testpmd> show port meter stats (port_id) (mtr_id) (clear)
2256 * ``clear``: Flag that indicates whether the statistics counters should
2257 be cleared (i.e. set to zero) immediately after they have been read or not.
2262 The following section shows functions for configuring traffic management on
2263 on the ethernet device through the use of generic TM API.
2265 show port traffic management capability
2266 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2268 Show traffic management capability of the port::
2270 testpmd> show port tm cap (port_id)
2272 show port traffic management capability (hierarchy level)
2273 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2275 Show traffic management hierarchy level capability of the port::
2277 testpmd> show port tm cap (port_id) (level_id)
2279 show port traffic management capability (hierarchy node level)
2280 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2282 Show the traffic management hierarchy node capability of the port::
2284 testpmd> show port tm cap (port_id) (node_id)
2286 show port traffic management hierarchy node type
2287 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2289 Show the port traffic management hierarchy node type::
2291 testpmd> show port tm node type (port_id) (node_id)
2293 show port traffic management hierarchy node stats
2294 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2296 Show the port traffic management hierarchy node statistics::
2298 testpmd> show port tm node stats (port_id) (node_id) (clear)
2302 * ``clear``: When this parameter has a non-zero value, the statistics counters
2303 are cleared (i.e. set to zero) immediately after they have been read,
2304 otherwise the statistics counters are left untouched.
2306 Add port traffic management private shaper profile
2307 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2309 Add the port traffic management private shaper profile::
2311 testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \
2312 (tb_rate) (tb_size) (packet_length_adjust)
2316 * ``shaper_profile id``: Shaper profile ID for the new profile.
2317 * ``tb_rate``: Token bucket rate (bytes per second).
2318 * ``tb_size``: Token bucket size (bytes).
2319 * ``packet_length_adjust``: The value (bytes) to be added to the length of
2320 each packet for the purpose of shaping. This parameter value can be used to
2321 correct the packet length with the framing overhead bytes that are consumed
2324 Delete port traffic management private shaper profile
2325 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2327 Delete the port traffic management private shaper::
2329 testpmd> del port tm node shaper profile (port_id) (shaper_profile_id)
2333 * ``shaper_profile id``: Shaper profile ID that needs to be deleted.
2335 Add port traffic management shared shaper
2336 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2338 Create the port traffic management shared shaper::
2340 testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \
2345 * ``shared_shaper_id``: Shared shaper ID to be created.
2346 * ``shaper_profile id``: Shaper profile ID for shared shaper.
2348 Set port traffic management shared shaper
2349 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2351 Update the port traffic management shared shaper::
2353 testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \
2358 * ``shared_shaper_id``: Shared shaper ID to be update.
2359 * ``shaper_profile id``: Shaper profile ID for shared shaper.
2361 Delete port traffic management shared shaper
2362 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2364 Delete the port traffic management shared shaper::
2366 testpmd> del port tm node shared shaper (port_id) (shared_shaper_id)
2370 * ``shared_shaper_id``: Shared shaper ID to be deleted.
2372 Set port traffic management hiearchy node private shaper
2373 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2375 set the port traffic management hierarchy node private shaper::
2377 testpmd> set port tm node shaper profile (port_id) (node_id) \
2382 * ``shaper_profile id``: Private shaper profile ID to be enabled on the
2385 Add port traffic management WRED profile
2386 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2388 Create a new WRED profile::
2390 testpmd> add port tm node wred profile (port_id) (wred_profile_id) \
2391 (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \
2392 (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \
2393 (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r)
2397 * ``wred_profile id``: Identifier for the newly create WRED profile
2398 * ``color_g``: Packet color (green)
2399 * ``min_th_g``: Minimum queue threshold for packet with green color
2400 * ``max_th_g``: Minimum queue threshold for packet with green color
2401 * ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp)
2402 * ``wq_log2_g``: Negated log2 of queue weight (wq)
2403 * ``color_y``: Packet color (yellow)
2404 * ``min_th_y``: Minimum queue threshold for packet with yellow color
2405 * ``max_th_y``: Minimum queue threshold for packet with yellow color
2406 * ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp)
2407 * ``wq_log2_y``: Negated log2 of queue weight (wq)
2408 * ``color_r``: Packet color (red)
2409 * ``min_th_r``: Minimum queue threshold for packet with yellow color
2410 * ``max_th_r``: Minimum queue threshold for packet with yellow color
2411 * ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp)
2412 * ``wq_log2_r``: Negated log2 of queue weight (wq)
2414 Delete port traffic management WRED profile
2415 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2417 Delete the WRED profile::
2419 testpmd> del port tm node wred profile (port_id) (wred_profile_id)
2421 Add port traffic management hierarchy nonleaf node
2422 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2424 Add nonleaf node to port traffic management hiearchy::
2426 testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \
2427 (priority) (weight) (level_id) (shaper_profile_id) \
2428 (n_sp_priorities) (stats_mask) (n_shared_shapers) \
2429 [(shared_shaper_0) (shared_shaper_1) ...] \
2433 * ``parent_node_id``: Node ID of the parent.
2434 * ``priority``: Node priority (highest node priority is zero). This is used by
2435 the SP algorithm running on the parent node for scheduling this node.
2436 * ``weight``: Node weight (lowest weight is one). The node weight is relative
2437 to the weight sum of all siblings that have the same priority. It is used by
2438 the WFQ algorithm running on the parent node for scheduling this node.
2439 * ``level_id``: Hiearchy level of the node.
2440 * ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2442 * ``n_sp_priorities``: Number of strict priorities.
2443 * ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2444 * ``n_shared_shapers``: Number of shared shapers.
2445 * ``shared_shaper_id``: Shared shaper id.
2447 Add port traffic management hierarchy leaf node
2448 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2450 Add leaf node to port traffic management hiearchy::
2452 testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \
2453 (priority) (weight) (level_id) (shaper_profile_id) \
2454 (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \
2455 [(shared_shaper_id) (shared_shaper_id) ...] \
2459 * ``parent_node_id``: Node ID of the parent.
2460 * ``priority``: Node priority (highest node priority is zero). This is used by
2461 the SP algorithm running on the parent node for scheduling this node.
2462 * ``weight``: Node weight (lowest weight is one). The node weight is relative
2463 to the weight sum of all siblings that have the same priority. It is used by
2464 the WFQ algorithm running on the parent node for scheduling this node.
2465 * ``level_id``: Hiearchy level of the node.
2466 * ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by
2468 * ``cman_mode``: Congestion management mode to be enabled for this node.
2469 * ``wred_profile_id``: WRED profile id to be enabled for this node.
2470 * ``stats_mask``: Mask of statistics counter types to be enabled for this node.
2471 * ``n_shared_shapers``: Number of shared shapers.
2472 * ``shared_shaper_id``: Shared shaper id.
2474 Delete port traffic management hierarchy node
2475 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2477 Delete node from port traffic management hiearchy::
2479 testpmd> del port tm node (port_id) (node_id)
2481 Update port traffic management hierarchy parent node
2482 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2484 Update port traffic management hierarchy parent node::
2486 testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \
2489 This function can only be called after the hierarchy commit invocation. Its
2490 success depends on the port support for this operation, as advertised through
2491 the port capability set. This function is valid for all nodes of the traffic
2492 management hierarchy except root node.
2494 Commit port traffic management hierarchy
2495 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2497 Commit the traffic management hierarchy on the port::
2499 testpmd> port tm hierarchy commit (port_id) (clean_on_fail)
2503 * ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function
2504 call failure. On the other hand, hierarchy is preserved when this parameter
2507 Set port traffic management default hierarchy (tm forwarding mode)
2508 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2510 set the traffic management default hierarchy on the port::
2512 testpmd> set port tm hierarchy default (port_id)
2517 This section details the available filter functions that are available.
2519 Note these functions interface the deprecated legacy filtering framework,
2520 superseded by *rte_flow*. See `Flow rules management`_.
2523 ~~~~~~~~~~~~~~~~~~~~
2525 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
2527 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
2528 ethertype (ether_type) (drop|fwd) queue (queue_id)
2530 The available information parameters are:
2532 * ``port_id``: The port which the Ethertype filter assigned on.
2534 * ``mac_addr``: Compare destination mac address.
2536 * ``mac_ignr``: Ignore destination mac address match.
2538 * ``mac_address``: Destination mac address to match.
2540 * ``ether_type``: The EtherType value want to match,
2541 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
2543 * ``queue_id``: The receive queue associated with this EtherType filter.
2544 It is meaningless when deleting or dropping.
2546 Example, to add/remove an ethertype filter rule::
2548 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
2549 ethertype 0x0806 fwd queue 3
2551 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
2552 ethertype 0x0806 fwd queue 3
2557 Add or delete a 2-tuple filter,
2558 which identifies packets by specific protocol and destination TCP/UDP port
2559 and forwards packets into one of the receive queues::
2561 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
2562 protocol (protocol_value) mask (mask_value) \
2563 tcp_flags (tcp_flags_value) priority (prio_value) \
2566 The available information parameters are:
2568 * ``port_id``: The port which the 2-tuple filter assigned on.
2570 * ``dst_port_value``: Destination port in L4.
2572 * ``protocol_value``: IP L4 protocol.
2574 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
2576 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
2578 * ``prio_value``: Priority of this filter.
2580 * ``queue_id``: The receive queue associated with this 2-tuple filter.
2582 Example, to add/remove an 2tuple filter rule::
2584 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
2585 tcp_flags 0x02 priority 3 queue 3
2587 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
2588 tcp_flags 0x02 priority 3 queue 3
2593 Add or delete a 5-tuple filter,
2594 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
2595 and routes packets into one of the receive queues::
2597 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
2598 (src_address) dst_port (dst_port_value) \
2599 src_port (src_port_value) protocol (protocol_value) \
2600 mask (mask_value) tcp_flags (tcp_flags_value) \
2601 priority (prio_value) queue (queue_id)
2603 The available information parameters are:
2605 * ``port_id``: The port which the 5-tuple filter assigned on.
2607 * ``dst_address``: Destination IP address.
2609 * ``src_address``: Source IP address.
2611 * ``dst_port_value``: TCP/UDP destination port.
2613 * ``src_port_value``: TCP/UDP source port.
2615 * ``protocol_value``: L4 protocol.
2617 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
2619 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
2621 * ``prio_value``: The priority of this filter.
2623 * ``queue_id``: The receive queue associated with this 5-tuple filter.
2625 Example, to add/remove an 5tuple filter rule::
2627 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2628 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2629 flags 0x0 priority 3 queue 3
2631 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2632 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2633 flags 0x0 priority 3 queue 3
2638 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
2640 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
2642 The available information parameters are:
2644 * ``port_id``: The port which the SYN filter assigned on.
2646 * ``high``: This SYN filter has higher priority than other filters.
2648 * ``low``: This SYN filter has lower priority than other filters.
2650 * ``queue_id``: The receive queue associated with this SYN filter
2654 testpmd> syn_filter 0 add priority high queue 3
2659 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
2660 and routed into one of the receive queues::
2662 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
2663 mask (mask_value) priority (prio_value) queue (queue_id)
2665 The available information parameters are:
2667 * ``port_id``: The port which the Flex filter is assigned on.
2669 * ``len_value``: Filter length in bytes, no greater than 128.
2671 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
2673 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
2675 * ``prio_value``: The priority of this filter.
2677 * ``queue_id``: The receive queue associated with this Flex filter.
2681 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
2682 mask 000C priority 3 queue 3
2684 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
2685 mask 000C priority 3 queue 3
2688 .. _testpmd_flow_director:
2690 flow_director_filter
2691 ~~~~~~~~~~~~~~~~~~~~
2693 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
2695 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
2696 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
2698 * Perfect match filters.
2699 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2700 The masked fields are for IP flow.
2702 * Signature filters.
2703 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
2705 * Perfect-mac-vlan match filters.
2706 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2707 The masked fields are for MAC VLAN flow.
2709 * Perfect-tunnel match filters.
2710 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2711 The masked fields are for tunnel flow.
2713 * Perfect-raw-flow-type match filters.
2714 The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet.
2715 The masked fields are specified by input sets.
2717 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
2718 per flow type and the flexible payload.
2720 The Flow Director can also mask out parts of all of these fields so that filters
2721 are only applied to certain fields or parts of the fields.
2723 Note that for raw flow type mode the source and destination fields in the
2724 raw packet buffer need to be presented in a reversed order with respect
2725 to the expected received packets.
2726 For example: IP source and destination addresses or TCP/UDP/SCTP
2727 source and destination ports
2729 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
2731 # Commands to add flow director filters of different flow types::
2733 flow_director_filter (port_id) mode IP (add|del|update) \
2734 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
2735 src (src_ip_address) dst (dst_ip_address) \
2736 tos (tos_value) proto (proto_value) ttl (ttl_value) \
2737 vlan (vlan_value) flexbytes (flexbytes_value) \
2738 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
2741 flow_director_filter (port_id) mode IP (add|del|update) \
2742 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
2743 src (src_ip_address) (src_port) \
2744 dst (dst_ip_address) (dst_port) \
2745 tos (tos_value) ttl (ttl_value) \
2746 vlan (vlan_value) flexbytes (flexbytes_value) \
2747 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
2750 flow_director_filter (port_id) mode IP (add|del|update) \
2751 flow (ipv4-sctp|ipv6-sctp) \
2752 src (src_ip_address) (src_port) \
2753 dst (dst_ip_address) (dst_port) \
2754 tos (tos_value) ttl (ttl_value) \
2755 tag (verification_tag) vlan (vlan_value) \
2756 flexbytes (flexbytes_value) (drop|fwd) \
2757 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
2759 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
2760 ether (ethertype) flexbytes (flexbytes_value) \
2761 (drop|fwd) pf|vf(vf_id) queue (queue_id)
2764 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
2765 mac (mac_address) vlan (vlan_value) \
2766 flexbytes (flexbytes_value) (drop|fwd) \
2767 queue (queue_id) fd_id (fd_id_value)
2769 flow_director_filter (port_id) mode Tunnel (add|del|update) \
2770 mac (mac_address) vlan (vlan_value) \
2771 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
2772 flexbytes (flexbytes_value) (drop|fwd) \
2773 queue (queue_id) fd_id (fd_id_value)
2775 flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \
2776 (drop|fwd) queue (queue_id) fd_id (fd_id_value) \
2777 packet (packet file name)
2779 For example, to add an ipv4-udp flow type filter::
2781 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
2782 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
2783 fwd pf queue 1 fd_id 1
2785 For example, add an ipv4-other flow type filter::
2787 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
2788 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
2789 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
2794 Flush all flow director filters on a device::
2796 testpmd> flush_flow_director (port_id)
2798 Example, to flush all flow director filter on port 0::
2800 testpmd> flush_flow_director 0
2805 Set flow director's input masks::
2807 flow_director_mask (port_id) mode IP vlan (vlan_value) \
2808 src_mask (ipv4_src) (ipv6_src) (src_port) \
2809 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2811 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2813 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2814 mac (mac_value) tunnel-type (tunnel_type_value) \
2815 tunnel-id (tunnel_id_value)
2817 Example, to set flow director mask on port 0::
2819 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2820 src_mask 255.255.255.255 \
2821 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2822 dst_mask 255.255.255.255 \
2823 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2825 flow_director_flex_mask
2826 ~~~~~~~~~~~~~~~~~~~~~~~
2828 set masks of flow director's flexible payload based on certain flow type::
2830 testpmd> flow_director_flex_mask (port_id) \
2831 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2832 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2833 l2_payload|all) (mask)
2835 Example, to set flow director's flex mask for all flow type on port 0::
2837 testpmd> flow_director_flex_mask 0 flow all \
2838 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2841 flow_director_flex_payload
2842 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2844 Configure flexible payload selection::
2846 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2848 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2850 testpmd> flow_director_flex_payload 0 l4 \
2851 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2853 get_sym_hash_ena_per_port
2854 ~~~~~~~~~~~~~~~~~~~~~~~~~
2856 Get symmetric hash enable configuration per port::
2858 get_sym_hash_ena_per_port (port_id)
2860 For example, to get symmetric hash enable configuration of port 1::
2862 testpmd> get_sym_hash_ena_per_port 1
2864 set_sym_hash_ena_per_port
2865 ~~~~~~~~~~~~~~~~~~~~~~~~~
2867 Set symmetric hash enable configuration per port to enable or disable::
2869 set_sym_hash_ena_per_port (port_id) (enable|disable)
2871 For example, to set symmetric hash enable configuration of port 1 to enable::
2873 testpmd> set_sym_hash_ena_per_port 1 enable
2875 get_hash_global_config
2876 ~~~~~~~~~~~~~~~~~~~~~~
2878 Get the global configurations of hash filters::
2880 get_hash_global_config (port_id)
2882 For example, to get the global configurations of hash filters of port 1::
2884 testpmd> get_hash_global_config 1
2886 set_hash_global_config
2887 ~~~~~~~~~~~~~~~~~~~~~~
2889 Set the global configurations of hash filters::
2891 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2892 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2893 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \
2896 For example, to enable simple_xor for flow type of ipv6 on port 2::
2898 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2903 Set the input set for hash::
2905 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2906 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2907 l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
2908 ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2909 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2910 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2911 fld-8th|none) (select|add)
2913 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2915 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2920 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2921 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2923 Set the input set for flow director::
2925 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2926 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2927 l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \
2928 ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2929 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2930 sctp-dst-port|sctp-veri-tag|none) (select|add)
2932 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2934 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2939 Set different GRE key length for input set::
2941 global_config (port_id) gre-key-len (number in bytes)
2943 For example to set GRE key length for input set to 4 bytes on port 0::
2945 testpmd> global_config 0 gre-key-len 4
2948 .. _testpmd_rte_flow:
2950 Flow rules management
2951 ---------------------
2953 Control of the generic flow API (*rte_flow*) is fully exposed through the
2954 ``flow`` command (validation, creation, destruction, queries and operation
2957 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2958 features simultaneously may cause undefined side-effects and is therefore
2964 Because the ``flow`` command uses dynamic tokens to handle the large number
2965 of possible flow rules combinations, its behavior differs slightly from
2966 other commands, in particular:
2968 - Pressing *?* or the *<tab>* key displays contextual help for the current
2969 token, not that of the entire command.
2971 - Optional and repeated parameters are supported (provided they are listed
2972 in the contextual help).
2974 The first parameter stands for the operation mode. Possible operations and
2975 their general syntax are described below. They are covered in detail in the
2978 - Check whether a flow rule can be created::
2980 flow validate {port_id}
2981 [group {group_id}] [priority {level}] [ingress] [egress]
2982 pattern {item} [/ {item} [...]] / end
2983 actions {action} [/ {action} [...]] / end
2985 - Create a flow rule::
2987 flow create {port_id}
2988 [group {group_id}] [priority {level}] [ingress] [egress]
2989 pattern {item} [/ {item} [...]] / end
2990 actions {action} [/ {action} [...]] / end
2992 - Destroy specific flow rules::
2994 flow destroy {port_id} rule {rule_id} [...]
2996 - Destroy all flow rules::
2998 flow flush {port_id}
3000 - Query an existing flow rule::
3002 flow query {port_id} {rule_id} {action}
3004 - List existing flow rules sorted by priority, filtered by group
3007 flow list {port_id} [group {group_id}] [...]
3009 - Restrict ingress traffic to the defined flow rules::
3011 flow isolate {port_id} {boolean}
3013 Validating flow rules
3014 ~~~~~~~~~~~~~~~~~~~~~
3016 ``flow validate`` reports whether a flow rule would be accepted by the
3017 underlying device in its current state but stops short of creating it. It is
3018 bound to ``rte_flow_validate()``::
3020 flow validate {port_id}
3021 [group {group_id}] [priority {level}] [ingress] [egress]
3022 pattern {item} [/ {item} [...]] / end
3023 actions {action} [/ {action} [...]] / end
3025 If successful, it will show::
3029 Otherwise it will show an error message of the form::
3031 Caught error type [...] ([...]): [...]
3033 This command uses the same parameters as ``flow create``, their format is
3034 described in `Creating flow rules`_.
3036 Check whether redirecting any Ethernet packet received on port 0 to RX queue
3037 index 6 is supported::
3039 testpmd> flow validate 0 ingress pattern eth / end
3040 actions queue index 6 / end
3044 Port 0 does not support TCPv6 rules::
3046 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
3048 Caught error type 9 (specific pattern item): Invalid argument
3054 ``flow create`` validates and creates the specified flow rule. It is bound
3055 to ``rte_flow_create()``::
3057 flow create {port_id}
3058 [group {group_id}] [priority {level}] [ingress] [egress]
3059 pattern {item} [/ {item} [...]] / end
3060 actions {action} [/ {action} [...]] / end
3062 If successful, it will return a flow rule ID usable with other commands::
3064 Flow rule #[...] created
3066 Otherwise it will show an error message of the form::
3068 Caught error type [...] ([...]): [...]
3070 Parameters describe in the following order:
3072 - Attributes (*group*, *priority*, *ingress*, *egress* tokens).
3073 - A matching pattern, starting with the *pattern* token and terminated by an
3075 - Actions, starting with the *actions* token and terminated by an *end*
3078 These translate directly to *rte_flow* objects provided as-is to the
3079 underlying functions.
3081 The shortest valid definition only comprises mandatory tokens::
3083 testpmd> flow create 0 pattern end actions end
3085 Note that PMDs may refuse rules that essentially do nothing such as this
3088 **All unspecified object values are automatically initialized to 0.**
3093 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
3094 specified before the ``pattern`` token.
3096 - ``group {group id}``: priority group.
3097 - ``priority {level}``: priority level within group.
3098 - ``ingress``: rule applies to ingress traffic.
3099 - ``egress``: rule applies to egress traffic.
3101 Each instance of an attribute specified several times overrides the previous
3102 value as shown below (group 4 is used)::
3104 testpmd> flow create 0 group 42 group 24 group 4 [...]
3106 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
3108 While not specifying a direction is an error, some rules may allow both
3111 Most rules affect RX therefore contain the ``ingress`` token::
3113 testpmd> flow create 0 ingress pattern [...]
3118 A matching pattern starts after the ``pattern`` token. It is made of pattern
3119 items and is terminated by a mandatory ``end`` item.
3121 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
3122 rte_flow_item_type``).
3124 The ``/`` token is used as a separator between pattern items as shown
3127 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
3129 Note that protocol items like these must be stacked from lowest to highest
3130 layer to make sense. For instance, the following rule is either invalid or
3131 unlikely to match any packet::
3133 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
3135 More information on these restrictions can be found in the *rte_flow*
3138 Several items support additional specification structures, for example
3139 ``ipv4`` allows specifying source and destination addresses as follows::
3141 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
3142 dst is 10.2.0.0 / end [...]
3144 This rule matches all IPv4 traffic with the specified properties.
3146 In this example, ``src`` and ``dst`` are field names of the underlying
3147 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
3148 in a similar fashion.
3150 The ``is`` token means that the subsequent value must be matched exactly,
3151 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
3152 accordingly. Possible assignment tokens are:
3154 - ``is``: match value perfectly (with full bit-mask).
3155 - ``spec``: match value according to configured bit-mask.
3156 - ``last``: specify upper bound to establish a range.
3157 - ``mask``: specify bit-mask with relevant bits set to one.
3158 - ``prefix``: generate bit-mask from a prefix length.
3160 These yield identical results::
3162 ipv4 src is 10.1.1.1
3166 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
3170 ipv4 src spec 10.1.1.1 src prefix 32
3174 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
3178 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
3180 Inclusive ranges can be defined with ``last``::
3182 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
3184 Note that ``mask`` affects both ``spec`` and ``last``::
3186 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
3187 # matches 10.1.0.0 to 10.2.255.255
3189 Properties can be modified multiple times::
3191 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
3195 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
3200 This section lists supported pattern items and their attributes, if any.
3202 - ``end``: end list of pattern items.
3204 - ``void``: no-op pattern item.
3206 - ``invert``: perform actions when pattern does not match.
3208 - ``any``: match any protocol for the current layer.
3210 - ``num {unsigned}``: number of layers covered.
3212 - ``pf``: match packets addressed to the physical function.
3214 - ``vf``: match packets addressed to a virtual function ID.
3216 - ``id {unsigned}``: destination VF ID.
3218 - ``port``: device-specific physical port index to use.
3220 - ``index {unsigned}``: physical port index.
3222 - ``raw``: match an arbitrary byte string.
3224 - ``relative {boolean}``: look for pattern after the previous item.
3225 - ``search {boolean}``: search pattern from offset (see also limit).
3226 - ``offset {integer}``: absolute or relative offset for pattern.
3227 - ``limit {unsigned}``: search area limit for start of pattern.
3228 - ``pattern {string}``: byte string to look for.
3230 - ``eth``: match Ethernet header.
3232 - ``dst {MAC-48}``: destination MAC.
3233 - ``src {MAC-48}``: source MAC.
3234 - ``type {unsigned}``: EtherType.
3236 - ``vlan``: match 802.1Q/ad VLAN tag.
3238 - ``tpid {unsigned}``: tag protocol identifier.
3239 - ``tci {unsigned}``: tag control information.
3240 - ``pcp {unsigned}``: priority code point.
3241 - ``dei {unsigned}``: drop eligible indicator.
3242 - ``vid {unsigned}``: VLAN identifier.
3244 - ``ipv4``: match IPv4 header.
3246 - ``tos {unsigned}``: type of service.
3247 - ``ttl {unsigned}``: time to live.
3248 - ``proto {unsigned}``: next protocol ID.
3249 - ``src {ipv4 address}``: source address.
3250 - ``dst {ipv4 address}``: destination address.
3252 - ``ipv6``: match IPv6 header.
3254 - ``tc {unsigned}``: traffic class.
3255 - ``flow {unsigned}``: flow label.
3256 - ``proto {unsigned}``: protocol (next header).
3257 - ``hop {unsigned}``: hop limit.
3258 - ``src {ipv6 address}``: source address.
3259 - ``dst {ipv6 address}``: destination address.
3261 - ``icmp``: match ICMP header.
3263 - ``type {unsigned}``: ICMP packet type.
3264 - ``code {unsigned}``: ICMP packet code.
3266 - ``udp``: match UDP header.
3268 - ``src {unsigned}``: UDP source port.
3269 - ``dst {unsigned}``: UDP destination port.
3271 - ``tcp``: match TCP header.
3273 - ``src {unsigned}``: TCP source port.
3274 - ``dst {unsigned}``: TCP destination port.
3276 - ``sctp``: match SCTP header.
3278 - ``src {unsigned}``: SCTP source port.
3279 - ``dst {unsigned}``: SCTP destination port.
3280 - ``tag {unsigned}``: validation tag.
3281 - ``cksum {unsigned}``: checksum.
3283 - ``vxlan``: match VXLAN header.
3285 - ``vni {unsigned}``: VXLAN identifier.
3287 - ``e_tag``: match IEEE 802.1BR E-Tag header.
3289 - ``grp_ecid_b {unsigned}``: GRP and E-CID base.
3291 - ``nvgre``: match NVGRE header.
3293 - ``tni {unsigned}``: virtual subnet ID.
3295 - ``mpls``: match MPLS header.
3297 - ``label {unsigned}``: MPLS label.
3299 - ``gre``: match GRE header.
3301 - ``protocol {unsigned}``: protocol type.
3303 - ``fuzzy``: fuzzy pattern match, expect faster than default.
3305 - ``thresh {unsigned}``: accuracy threshold.
3307 - ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header.
3309 - ``teid {unsigned}``: tunnel endpoint identifier.
3311 - ``geneve``: match GENEVE header.
3313 - ``vni {unsigned}``: virtual network identifier.
3314 - ``protocol {unsigned}``: protocol type.
3319 A list of actions starts after the ``actions`` token in the same fashion as
3320 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
3321 terminated by a mandatory ``end`` action.
3323 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
3324 rte_flow_action_type``).
3326 Dropping all incoming UDPv4 packets can be expressed as follows::
3328 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3331 Several actions have configurable properties which must be specified when
3332 there is no valid default value. For example, ``queue`` requires a target
3335 This rule redirects incoming UDPv4 traffic to queue index 6::
3337 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3338 actions queue index 6 / end
3340 While this one could be rejected by PMDs (unspecified queue index)::
3342 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
3345 As defined by *rte_flow*, the list is not ordered, all actions of a given
3346 rule are performed simultaneously. These are equivalent::
3348 queue index 6 / void / mark id 42 / end
3352 void / mark id 42 / queue index 6 / end
3354 All actions in a list should have different types, otherwise only the last
3355 action of a given type is taken into account::
3357 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
3361 drop / drop / drop / end # drop is performed only once
3365 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
3367 Considering they are performed simultaneously, opposite and overlapping
3368 actions can sometimes be combined when the end result is unambiguous::
3370 drop / queue index 6 / end # drop has no effect
3374 drop / dup index 6 / end # same as above
3378 queue index 6 / rss queues 6 7 8 / end # queue has no effect
3382 drop / passthru / end # drop has no effect
3384 Note that PMDs may still refuse such combinations.
3389 This section lists supported actions and their attributes, if any.
3391 - ``end``: end list of actions.
3393 - ``void``: no-op action.
3395 - ``passthru``: let subsequent rule process matched packets.
3397 - ``mark``: attach 32 bit value to packets.
3399 - ``id {unsigned}``: 32 bit value to return with packets.
3401 - ``flag``: flag packets.
3403 - ``queue``: assign packets to a given queue index.
3405 - ``index {unsigned}``: queue index to use.
3407 - ``drop``: drop packets (note: passthru has priority).
3409 - ``count``: enable counters for this rule.
3411 - ``dup``: duplicate packets to a given queue index.
3413 - ``index {unsigned}``: queue index to duplicate packets to.
3415 - ``rss``: spread packets among several queues.
3417 - ``queues [{unsigned} [...]] end``: queue indices to use.
3419 - ``pf``: redirect packets to physical device function.
3421 - ``vf``: redirect packets to virtual device function.
3423 - ``original {boolean}``: use original VF ID if possible.
3424 - ``id {unsigned}``: VF ID to redirect packets to.
3426 Destroying flow rules
3427 ~~~~~~~~~~~~~~~~~~~~~
3429 ``flow destroy`` destroys one or more rules from their rule ID (as returned
3430 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
3431 times as necessary::
3433 flow destroy {port_id} rule {rule_id} [...]
3435 If successful, it will show::
3437 Flow rule #[...] destroyed
3439 It does not report anything for rule IDs that do not exist. The usual error
3440 message is shown when a rule cannot be destroyed::
3442 Caught error type [...] ([...]): [...]
3444 ``flow flush`` destroys all rules on a device and does not take extra
3445 arguments. It is bound to ``rte_flow_flush()``::
3447 flow flush {port_id}
3449 Any errors are reported as above.
3451 Creating several rules and destroying them::
3453 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3454 actions queue index 2 / end
3455 Flow rule #0 created
3456 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3457 actions queue index 3 / end
3458 Flow rule #1 created
3459 testpmd> flow destroy 0 rule 0 rule 1
3460 Flow rule #1 destroyed
3461 Flow rule #0 destroyed
3464 The same result can be achieved using ``flow flush``::
3466 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3467 actions queue index 2 / end
3468 Flow rule #0 created
3469 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3470 actions queue index 3 / end
3471 Flow rule #1 created
3472 testpmd> flow flush 0
3475 Non-existent rule IDs are ignored::
3477 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3478 actions queue index 2 / end
3479 Flow rule #0 created
3480 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3481 actions queue index 3 / end
3482 Flow rule #1 created
3483 testpmd> flow destroy 0 rule 42 rule 10 rule 2
3485 testpmd> flow destroy 0 rule 0
3486 Flow rule #0 destroyed
3492 ``flow query`` queries a specific action of a flow rule having that
3493 ability. Such actions collect information that can be reported using this
3494 command. It is bound to ``rte_flow_query()``::
3496 flow query {port_id} {rule_id} {action}
3498 If successful, it will display either the retrieved data for known actions
3499 or the following message::
3501 Cannot display result for action type [...] ([...])
3503 Otherwise, it will complain either that the rule does not exist or that some
3506 Flow rule #[...] not found
3510 Caught error type [...] ([...]): [...]
3512 Currently only the ``count`` action is supported. This action reports the
3513 number of packets that hit the flow rule and the total number of bytes. Its
3514 output has the following format::
3517 hits_set: [...] # whether "hits" contains a valid value
3518 bytes_set: [...] # whether "bytes" contains a valid value
3519 hits: [...] # number of packets
3520 bytes: [...] # number of bytes
3522 Querying counters for TCPv6 packets redirected to queue 6::
3524 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
3525 actions queue index 6 / count / end
3526 Flow rule #4 created
3527 testpmd> flow query 0 4 count
3538 ``flow list`` lists existing flow rules sorted by priority and optionally
3539 filtered by group identifiers::
3541 flow list {port_id} [group {group_id}] [...]
3543 This command only fails with the following message if the device does not
3548 Output consists of a header line followed by a short description of each
3549 flow rule, one per line. There is no output at all when no flow rules are
3550 configured on the device::
3552 ID Group Prio Attr Rule
3553 [...] [...] [...] [...] [...]
3555 ``Attr`` column flags:
3557 - ``i`` for ``ingress``.
3558 - ``e`` for ``egress``.
3560 Creating several flow rules and listing them::
3562 testpmd> flow create 0 ingress pattern eth / ipv4 / end
3563 actions queue index 6 / end
3564 Flow rule #0 created
3565 testpmd> flow create 0 ingress pattern eth / ipv6 / end
3566 actions queue index 2 / end
3567 Flow rule #1 created
3568 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
3569 actions rss queues 6 7 8 end / end
3570 Flow rule #2 created
3571 testpmd> flow list 0
3572 ID Group Prio Attr Rule
3573 0 0 0 i- ETH IPV4 => QUEUE
3574 1 0 0 i- ETH IPV6 => QUEUE
3575 2 0 5 i- ETH IPV4 UDP => RSS
3578 Rules are sorted by priority (i.e. group ID first, then priority level)::
3580 testpmd> flow list 1
3581 ID Group Prio Attr Rule
3582 0 0 0 i- ETH => COUNT
3583 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3584 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3585 1 24 0 i- ETH IPV4 UDP => QUEUE
3586 4 24 10 i- ETH IPV4 TCP => DROP
3587 3 24 20 i- ETH IPV4 => DROP
3588 2 24 42 i- ETH IPV4 UDP => QUEUE
3589 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3592 Output can be limited to specific groups::
3594 testpmd> flow list 1 group 0 group 63
3595 ID Group Prio Attr Rule
3596 0 0 0 i- ETH => COUNT
3597 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3598 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3599 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3602 Toggling isolated mode
3603 ~~~~~~~~~~~~~~~~~~~~~~
3605 ``flow isolate`` can be used to tell the underlying PMD that ingress traffic
3606 must only be injected from the defined flow rules; that no default traffic
3607 is expected outside those rules and the driver is free to assign more
3608 resources to handle them. It is bound to ``rte_flow_isolate()``::
3610 flow isolate {port_id} {boolean}
3612 If successful, enabling or disabling isolated mode shows either::
3614 Ingress traffic on port [...]
3615 is now restricted to the defined flow rules
3619 Ingress traffic on port [...]
3620 is not restricted anymore to the defined flow rules
3622 Otherwise, in case of error::
3624 Caught error type [...] ([...]): [...]
3626 Mainly due to its side effects, PMDs supporting this mode may not have the
3627 ability to toggle it more than once without reinitializing affected ports
3628 first (e.g. by exiting testpmd).
3630 Enabling isolated mode::
3632 testpmd> flow isolate 0 true
3633 Ingress traffic on port 0 is now restricted to the defined flow rules
3636 Disabling isolated mode::
3638 testpmd> flow isolate 0 false
3639 Ingress traffic on port 0 is not restricted anymore to the defined flow rules
3642 Sample QinQ flow rules
3643 ~~~~~~~~~~~~~~~~~~~~~~
3645 Before creating QinQ rule(s) the following commands should be issued to enable QinQ::
3647 testpmd> port stop 0
3648 testpmd> vlan set qinq on 0
3650 The above command sets the inner and outer TPID's to 0x8100.
3652 To change the TPID's the following commands should be used::
3654 testpmd> vlan set outer tpid 0xa100 0
3655 testpmd> vlan set inner tpid 0x9100 0
3656 testpmd> port start 0
3658 Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM.
3662 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 /
3663 vlan tci is 456 / end actions vf id 1 / queue index 0 / end
3664 Flow rule #0 validated
3666 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 /
3667 vlan tci is 456 / end actions vf id 123 / queue index 0 / end
3668 Flow rule #0 created
3670 testpmd> flow list 0
3671 ID Group Prio Attr Rule
3672 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3674 Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host.
3678 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 /
3679 vlan tci is 654 / end actions pf / queue index 0 / end
3680 Flow rule #1 validated
3682 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 /
3683 vlan tci is 654 / end actions pf / queue index 1 / end
3684 Flow rule #1 created
3686 testpmd> flow list 0
3687 ID Group Prio Attr Rule
3688 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3689 1 0 0 i- ETH VLAN VLAN=>PF QUEUE