2 Copyright(c) 2010-2016 Intel Corporation. All rights reserved.
5 Redistribution and use in source and binary forms, with or without
6 modification, are permitted provided that the following conditions
9 * Redistributions of source code must retain the above copyright
10 notice, this list of conditions and the following disclaimer.
11 * Redistributions in binary form must reproduce the above copyright
12 notice, this list of conditions and the following disclaimer in
13 the documentation and/or other materials provided with the
15 * Neither the name of Intel Corporation nor the names of its
16 contributors may be used to endorse or promote products derived
17 from this software without specific prior written permission.
19 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
20 "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
21 LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
22 A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
23 OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
24 SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
25 LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
26 DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
27 THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
28 (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
29 OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
33 Testpmd Runtime Functions
34 =========================
36 Where the testpmd application is started in interactive mode, (``-i|--interactive``),
37 it displays a prompt that can be used to start and stop forwarding,
38 configure the application, display statistics (including the extended NIC
39 statistics aka xstats) , set the Flow Director and other tasks::
43 The testpmd prompt has some, limited, readline support.
44 Common bash command-line functions such as ``Ctrl+a`` and ``Ctrl+e`` to go to the start and end of the prompt line are supported
45 as well as access to the command history via the up-arrow.
47 There is also support for tab completion.
48 If you type a partial command and hit ``<TAB>`` you get a list of the available completions:
50 .. code-block:: console
52 testpmd> show port <TAB>
54 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
55 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
56 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X
57 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all
63 Some examples in this document are too long to fit on one line are are shown wrapped at `"\\"` for display purposes::
65 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
66 (pause_time) (send_xon) (port_id)
68 In the real ``testpmd>`` prompt these commands should be on a single line.
73 The testpmd has on-line help for the functions that are available at runtime.
74 These are divided into sections and can be accessed using help, help section or help all:
76 .. code-block:: console
80 help control : Start and stop forwarding.
81 help display : Displaying port, stats and config information.
82 help config : Configuration information.
83 help ports : Configuring ports.
84 help registers : Reading and setting port registers.
85 help filters : Filters configuration help.
86 help all : All of the above sections.
89 Command File Functions
90 ----------------------
92 To facilitate loading large number of commands or to avoid cutting and pasting where not
93 practical or possible testpmd supports alternative methods for executing commands.
95 * If started with the ``--cmdline-file=FILENAME`` command line argument testpmd
96 will execute all CLI commands contained within the file immediately before
97 starting packet forwarding or entering interactive mode.
99 .. code-block:: console
101 ./testpmd -n4 -r2 ... -- -i --cmdline-file=/home/ubuntu/flow-create-commands.txt
102 Interactive-mode selected
103 CLI commands to be read from /home/ubuntu/flow-create-commands.txt
104 Configuring Port 0 (socket 0)
105 Port 0: 7C:FE:90:CB:74:CE
106 Configuring Port 1 (socket 0)
107 Port 1: 7C:FE:90:CB:74:CA
108 Checking link statuses...
109 Port 0 Link Up - speed 10000 Mbps - full-duplex
110 Port 1 Link Up - speed 10000 Mbps - full-duplex
116 Flow rule #498 created
117 Flow rule #499 created
118 Read all CLI commands from /home/ubuntu/flow-create-commands.txt
122 * At run-time additional commands can be loaded in bulk by invoking the ``load FILENAME``
125 .. code-block:: console
127 testpmd> load /home/ubuntu/flow-create-commands.txt
132 Flow rule #498 created
133 Flow rule #499 created
134 Read all CLI commands from /home/ubuntu/flow-create-commands.txt
138 In all cases output from any included command will be displayed as standard output.
139 Execution will continue until the end of the file is reached regardless of
140 whether any errors occur. The end user must examine the output to determine if
141 any failures occurred.
150 Start packet forwarding with current configuration::
157 Start packet forwarding with current configuration after sending specified number of bursts of packets::
159 testpmd> start tx_first (""|burst_num)
161 The default burst number is 1 when ``burst_num`` not presented.
166 Stop packet forwarding, and display accumulated statistics::
181 The functions in the following sections are used to display information about the
182 testpmd configuration or the NIC status.
187 Display information for a given port or all ports::
189 testpmd> show port (info|stats|xstats|fdir|stat_qmap|dcb_tc|cap) (port_id|all)
191 The available information categories are:
193 * ``info``: General port information such as MAC address.
195 * ``stats``: RX/TX statistics.
197 * ``xstats``: RX/TX extended NIC statistics.
199 * ``fdir``: Flow Director information and statistics.
201 * ``stat_qmap``: Queue statistics mapping.
203 * ``dcb_tc``: DCB information such as TC mapping.
205 * ``cap``: Supported offload capabilities.
209 .. code-block:: console
211 testpmd> show port info 0
213 ********************* Infos for port 0 *********************
215 MAC address: XX:XX:XX:XX:XX:XX
217 memory allocation on the socket: 0
219 Link speed: 40000 Mbps
220 Link duplex: full-duplex
221 Promiscuous mode: enabled
222 Allmulticast mode: disabled
223 Maximum number of MAC addresses: 64
224 Maximum number of MAC addresses of hash filtering: 0
229 Redirection table size: 512
230 Supported flow types:
250 Display the rss redirection table entry indicated by masks on port X::
252 testpmd> show port (port_id) rss reta (size) (mask0, mask1...)
254 size is used to indicate the hardware supported reta size
259 Display the RSS hash functions and RSS hash key of a port::
261 testpmd> show port (port_id) rss-hash ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2-payload|ipv6-ex|ipv6-tcp-ex|ipv6-udp-ex [key]
266 Clear the port statistics for a given port or for all ports::
268 testpmd> clear port (info|stats|xstats|fdir|stat_qmap) (port_id|all)
272 testpmd> clear port stats all
277 Display information for a given port's RX/TX queue::
279 testpmd> show (rxq|txq) info (port_id) (queue_id)
284 Displays the configuration of the application.
285 The configuration comes from the command-line, the runtime or the application defaults::
287 testpmd> show config (rxtx|cores|fwd|txpkts)
289 The available information categories are:
291 * ``rxtx``: RX/TX configuration items.
293 * ``cores``: List of forwarding cores.
295 * ``fwd``: Packet forwarding configuration.
297 * ``txpkts``: Packets to TX configuration.
301 .. code-block:: console
303 testpmd> show config rxtx
305 io packet forwarding - CRC stripping disabled - packets/burst=16
306 nb forwarding cores=2 - nb forwarding ports=1
307 RX queues=1 - RX desc=128 - RX free threshold=0
308 RX threshold registers: pthresh=8 hthresh=8 wthresh=4
309 TX queues=1 - TX desc=512 - TX free threshold=0
310 TX threshold registers: pthresh=36 hthresh=0 wthresh=0
311 TX RS bit threshold=0 - TXQ flags=0x0
316 Set the packet forwarding mode::
318 testpmd> set fwd (io|mac|macswap|flowgen| \
319 rxonly|txonly|csum|icmpecho) (""|retry)
321 ``retry`` can be specified for forwarding engines except ``rx_only``.
323 The available information categories are:
325 * ``io``: Forwards packets "as-is" in I/O mode.
326 This is the fastest possible forwarding operation as it does not access packets data.
327 This is the default mode.
329 * ``mac``: Changes the source and the destination Ethernet addresses of packets before forwarding them.
330 Default application behaviour is to set source Ethernet address to that of the transmitting interface, and destination
331 address to a dummy value (set during init). The user may specify a target destination Ethernet address via the 'eth-peer' or
332 'eth-peer-configfile' command-line options. It is not currently possible to specify a specific source Ethernet address.
334 * ``macswap``: MAC swap forwarding mode.
335 Swaps the source and the destination Ethernet addresses of packets before forwarding them.
337 * ``flowgen``: Multi-flow generation mode.
338 Originates a number of flows (with varying destination IP addresses), and terminate receive traffic.
340 * ``rxonly``: Receives packets but doesn't transmit them.
342 * ``txonly``: Generates and transmits packets without receiving any.
344 * ``csum``: Changes the checksum field with hardware or software methods depending on the offload flags on the packet.
346 * ``icmpecho``: Receives a burst of packets, lookup for IMCP echo requests and, if any, send back ICMP echo replies.
348 * ``ieee1588``: Demonstrate L2 IEEE1588 V2 PTP timestamping for RX and TX. Requires ``CONFIG_RTE_LIBRTE_IEEE1588=y``.
350 Note: TX timestamping is only available in the "Full Featured" TX path. To force ``testpmd`` into this mode set ``--txqflags=0``.
354 testpmd> set fwd rxonly
356 Set rxonly packet forwarding mode
362 Display an RX descriptor for a port RX queue::
364 testpmd> read rxd (port_id) (queue_id) (rxd_id)
368 testpmd> read rxd 0 0 4
369 0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180
374 Display a TX descriptor for a port TX queue::
376 testpmd> read txd (port_id) (queue_id) (txd_id)
380 testpmd> read txd 0 0 4
381 0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C
386 Get loaded dynamic device personalization (DDP) package info list::
388 testpmd> ddp get list (port_id)
393 Display information about dynamic device personalization (DDP) profile::
395 testpmd> ddp get info (profile_path)
400 Display VF statistics::
402 testpmd> show vf stats (port_id) (vf_id)
407 Reset VF statistics::
409 testpmd> clear vf stats (port_id) (vf_id)
411 show port pctype mapping
412 ~~~~~~~~~~~~~~~~~~~~~~~~
414 List all items from the pctype mapping table::
416 testpmd> show port (port_id) pctype mapping
419 Configuration Functions
420 -----------------------
422 The testpmd application can be configured from the runtime as well as from the command-line.
424 This section details the available configuration functions that are available.
428 Configuration changes only become active when forwarding is started/restarted.
433 Reset forwarding to the default configuration::
440 Set the debug verbosity level::
442 testpmd> set verbose (level)
444 Currently the only available levels are 0 (silent except for error) and 1 (fully verbose).
449 Set the number of ports used by the application:
453 This is equivalent to the ``--nb-ports`` command-line option.
458 Set the number of cores used by the application::
460 testpmd> set nbcore (num)
462 This is equivalent to the ``--nb-cores`` command-line option.
466 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port.
471 Set the forwarding cores hexadecimal mask::
473 testpmd> set coremask (mask)
475 This is equivalent to the ``--coremask`` command-line option.
479 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding.
484 Set the forwarding ports hexadecimal mask::
486 testpmd> set portmask (mask)
488 This is equivalent to the ``--portmask`` command-line option.
493 Set number of packets per burst::
495 testpmd> set burst (num)
497 This is equivalent to the ``--burst command-line`` option.
499 When retry is enabled, the transmit delay time and number of retries can also be set::
501 testpmd> set burst tx delay (microseconds) retry (num)
506 Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode::
508 testpmd> set txpkts (x[,y]*)
510 Where x[,y]* represents a CSV list of values, without white space.
515 Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes::
517 testpmd> set txsplit (off|on|rand)
521 * ``off`` disable packet copy & split for CSUM mode.
523 * ``on`` split outgoing packet into multiple segments. Size of each segment
524 and number of segments per packet is determined by ``set txpkts`` command
527 * ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments.
532 Set the list of forwarding cores::
534 testpmd> set corelist (x[,y]*)
536 For example, to change the forwarding cores:
538 .. code-block:: console
540 testpmd> set corelist 3,1
541 testpmd> show config fwd
543 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled
544 Logical Core 3 (socket 0) forwards packets on 1 streams:
545 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01
546 Logical Core 1 (socket 0) forwards packets on 1 streams:
547 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
551 The cores are used in the same order as specified on the command line.
556 Set the list of forwarding ports::
558 testpmd> set portlist (x[,y]*)
560 For example, to change the port forwarding:
562 .. code-block:: console
564 testpmd> set portlist 0,2,1,3
565 testpmd> show config fwd
567 io packet forwarding - ports=4 - cores=1 - streams=4
568 Logical Core 3 (socket 0) forwards packets on 4 streams:
569 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01
570 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
571 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03
572 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02
577 Enable/disable tx loopback::
579 testpmd> set tx loopback (port_id) (on|off)
584 set drop enable bit for all queues::
586 testpmd> set all queues drop (port_id) (on|off)
588 set split drop enable (for VF)
589 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
591 set split drop enable bit for VF from PF::
593 testpmd> set vf split drop (port_id) (vf_id) (on|off)
595 set mac antispoof (for VF)
596 ~~~~~~~~~~~~~~~~~~~~~~~~~~
598 Set mac antispoof for a VF from the PF::
600 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off)
605 Enable/disable MACsec offload::
607 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off)
608 testpmd> set macsec offload (port_id) off
613 Configure MACsec secure connection (SC)::
615 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi)
619 The pi argument is ignored for tx.
620 Check the NIC Datasheet for hardware limits.
625 Configure MACsec secure association (SA)::
627 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key)
631 The IDX value must be 0 or 1.
632 Check the NIC Datasheet for hardware limits.
634 set broadcast mode (for VF)
635 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
637 Set broadcast mode for a VF from the PF::
639 testpmd> set vf broadcast (port_id) (vf_id) (on|off)
644 Set the VLAN strip on a port::
646 testpmd> vlan set strip (on|off) (port_id)
651 Set the VLAN strip for a queue on a port::
653 testpmd> vlan set stripq (on|off) (port_id,queue_id)
655 vlan set stripq (for VF)
656 ~~~~~~~~~~~~~~~~~~~~~~~~
658 Set VLAN strip for all queues in a pool for a VF from the PF::
660 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off)
662 vlan set insert (for VF)
663 ~~~~~~~~~~~~~~~~~~~~~~~~
665 Set VLAN insert for a VF from the PF::
667 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id)
669 vlan set tag (for VF)
670 ~~~~~~~~~~~~~~~~~~~~~
672 Set VLAN tag for a VF from the PF::
674 testpmd> set vf vlan tag (port_id) (vf_id) (on|off)
676 vlan set antispoof (for VF)
677 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
679 Set VLAN antispoof for a VF from the PF::
681 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off)
686 Set the VLAN filter on a port::
688 testpmd> vlan set filter (on|off) (port_id)
693 Set the VLAN QinQ (extended queue in queue) on for a port::
695 testpmd> vlan set qinq (on|off) (port_id)
700 Set the inner or outer VLAN TPID for packet filtering on a port::
702 testpmd> vlan set (inner|outer) tpid (value) (port_id)
706 TPID value must be a 16-bit number (value <= 65536).
711 Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID::
713 testpmd> rx_vlan add (vlan_id|all) (port_id)
717 VLAN filter must be set on that port. VLAN ID < 4096.
718 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries
719 in VFTA table. This is important if enabling all vlan_ids.
724 Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID::
726 testpmd> rx_vlan rm (vlan_id|all) (port_id)
731 Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID::
733 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask)
738 Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID::
740 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask)
745 Add a tunnel filter on a port::
747 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \
748 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
749 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
751 The available information categories are:
753 * ``vxlan``: Set tunnel type as VXLAN.
755 * ``nvgre``: Set tunnel type as NVGRE.
757 * ``ipingre``: Set tunnel type as IP-in-GRE.
759 * ``imac-ivlan``: Set filter type as Inner MAC and VLAN.
761 * ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID.
763 * ``imac-tenid``: Set filter type as Inner MAC and tenant ID.
765 * ``imac``: Set filter type as Inner MAC.
767 * ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID.
769 * ``oip``: Set filter type as Outer IP.
771 * ``iip``: Set filter type as Inner IP.
775 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \
776 192.168.2.2 0 ipingre oip 1 1
778 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP.
783 Remove a tunnel filter on a port::
785 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \
786 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
787 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
792 Add an UDP port for VXLAN packet filter on a port::
794 testpmd> rx_vxlan_port add (udp_port) (port_id)
799 Remove an UDP port for VXLAN packet filter on a port::
801 testpmd> rx_vxlan_port rm (udp_port) (port_id)
806 Set hardware insertion of VLAN IDs in packets sent on a port::
808 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer]
810 For example, set a single VLAN ID (5) insertion on port 0::
814 Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1::
822 Set port based hardware insertion of VLAN ID in packets sent on a port::
824 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off)
829 Disable hardware insertion of a VLAN header in packets sent on a port::
831 testpmd> tx_vlan reset (port_id)
836 Select hardware or software calculation of the checksum when
837 transmitting a packet using the ``csum`` forwarding engine::
839 testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id)
843 * ``ip|udp|tcp|sctp`` always relate to the inner layer.
845 * ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized
846 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are
847 supported). See also the ``csum parse-tunnel`` command.
851 Check the NIC Datasheet for hardware limits.
856 Define how tunneled packets should be handled by the csum forward
859 testpmd> csum parse-tunnel (on|off) (tx_port_id)
861 If enabled, the csum forward engine will try to recognize supported
862 tunnel headers (vxlan, gre, ipip).
864 If disabled, treat tunnel packets as non-tunneled packets (a inner
865 header is handled as a packet payload).
869 The port argument is the TX port like in the ``csum set`` command.
873 Consider a packet in packet like the following::
875 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in
877 * If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
878 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the
879 ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``).
881 * If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
882 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``.
887 Display tx checksum offload configuration::
889 testpmd> csum show (port_id)
894 Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine::
896 testpmd> tso set (segsize) (port_id)
900 Check the NIC datasheet for hardware limits.
905 Display the status of TCP Segmentation Offload::
907 testpmd> tso show (port_id)
912 Enable or disable GRO in ``csum`` forwarding engine::
914 testpmd> set port <port_id> gro on|off
916 If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4
917 packets received from the given port.
919 If disabled, packets received from the given port won't be performed
920 GRO. By default, GRO is disabled for all ports.
924 When enable GRO for a port, TCP/IPv4 packets received from the port
925 will be performed GRO. After GRO, all merged packets have bad
926 checksums, since the GRO library doesn't re-calculate checksums for
927 the merged packets. Therefore, if users want the merged packets to
928 have correct checksums, please select HW IP checksum calculation and
929 HW TCP checksum calculation for the port which the merged packets are
935 Display GRO configuration for a given port::
937 testpmd> show port <port_id> gro
942 Set the cycle to flush the GROed packets from reassembly tables::
944 testpmd> set gro flush <cycles>
946 When enable GRO, the csum forwarding engine performs GRO on received
947 packets, and the GROed packets are stored in reassembly tables. Users
948 can use this command to determine when the GROed packets are flushed
949 from the reassembly tables.
951 The ``cycles`` is measured in GRO operation times. The csum forwarding
952 engine flushes the GROed packets from the tables every ``cycles`` GRO
955 By default, the value of ``cycles`` is 1, which means flush GROed packets
956 from the reassembly tables as soon as one GRO operation finishes. The value
957 of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``.
959 Please note that the large value of ``cycles`` may cause the poor TCP/IP
960 stack performance. Because the GROed packets are delayed to arrive the
961 stack, thus causing more duplicated ACKs and TCP retransmissions.
966 Add an alternative MAC address to a port::
968 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX)
973 Remove a MAC address from a port::
975 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX)
977 mac_addr add (for VF)
978 ~~~~~~~~~~~~~~~~~~~~~
980 Add an alternative MAC address for a VF to a port::
982 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX)
987 Set the default MAC address for a port::
989 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX)
991 mac_addr set (for VF)
992 ~~~~~~~~~~~~~~~~~~~~~
994 Set the MAC address for a VF from the PF::
996 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX)
1001 Set the unicast hash filter(s) on/off for a port::
1003 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
1008 Set the promiscuous mode on for a port or for all ports.
1009 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1011 testpmd> set promisc (port_id|all) (on|off)
1016 Set the allmulti mode for a port or for all ports::
1018 testpmd> set allmulti (port_id|all) (on|off)
1020 Same as the ifconfig (8) option. Controls how multicast packets are handled.
1022 set promisc (for VF)
1023 ~~~~~~~~~~~~~~~~~~~~
1025 Set the unicast promiscuous mode for a VF from PF.
1026 It's supported by Intel i40e NICs now.
1027 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1029 testpmd> set vf promisc (port_id) (vf_id) (on|off)
1031 set allmulticast (for VF)
1032 ~~~~~~~~~~~~~~~~~~~~~~~~~
1034 Set the multicast promiscuous mode for a VF from PF.
1035 It's supported by Intel i40e NICs now.
1036 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
1038 testpmd> set vf allmulti (port_id) (vf_id) (on|off)
1040 set tx max bandwidth (for VF)
1041 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1043 Set TX max absolute bandwidth (Mbps) for a VF from PF::
1045 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth)
1047 set tc tx min bandwidth (for VF)
1048 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1050 Set all TCs' TX min relative bandwidth (%) for a VF from PF::
1052 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...)
1054 set tc tx max bandwidth (for VF)
1055 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1057 Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF::
1059 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth)
1061 set tc strict link priority mode
1062 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1064 Set some TCs' strict link priority mode on a physical port::
1066 testpmd> set tx strict-link-priority (port_id) (tc_bitmap)
1068 set tc tx min bandwidth
1069 ~~~~~~~~~~~~~~~~~~~~~~~
1071 Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs::
1073 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...)
1078 Set the link flow control parameter on a port::
1080 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1081 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
1082 autoneg (on|off) (port_id)
1086 * ``high_water`` (integer): High threshold value to trigger XOFF.
1088 * ``low_water`` (integer): Low threshold value to trigger XON.
1090 * ``pause_time`` (integer): Pause quota in the Pause frame.
1092 * ``send_xon`` (0/1): Send XON frame.
1094 * ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
1096 * ``autoneg``: Change the auto-negotiation parameter.
1101 Set the priority flow control parameter on a port::
1103 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1104 (pause_time) (priority) (port_id)
1108 * ``high_water`` (integer): High threshold value.
1110 * ``low_water`` (integer): Low threshold value.
1112 * ``pause_time`` (integer): Pause quota in the Pause frame.
1114 * ``priority`` (0-7): VLAN User Priority.
1119 Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
1121 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
1123 For example, to set rx queue 2 on port 0 to mapping 5::
1125 testpmd>set stat_qmap rx 0 2 5
1127 set port - rx/tx (for VF)
1128 ~~~~~~~~~~~~~~~~~~~~~~~~~
1130 Set VF receive/transmit from a port::
1132 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
1134 set port - mac address filter (for VF)
1135 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1137 Add/Remove unicast or multicast MAC addr filter for a VF::
1139 testpmd> set port (port_id) vf (vf_id) (mac_addr) \
1140 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
1142 set port - rx mode(for VF)
1143 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1145 Set the VF receive mode of a port::
1147 testpmd> set port (port_id) vf (vf_id) \
1148 rxmode (AUPE|ROPE|BAM|MPE) (on|off)
1150 The available receive modes are:
1152 * ``AUPE``: Accepts untagged VLAN.
1154 * ``ROPE``: Accepts unicast hash.
1156 * ``BAM``: Accepts broadcast packets.
1158 * ``MPE``: Accepts all multicast packets.
1160 set port - tx_rate (for Queue)
1161 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1163 Set TX rate limitation for a queue on a port::
1165 testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1167 set port - tx_rate (for VF)
1168 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
1170 Set TX rate limitation for queues in VF on a port::
1172 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1174 set port - mirror rule
1175 ~~~~~~~~~~~~~~~~~~~~~~
1177 Set pool or vlan type mirror rule for a port::
1179 testpmd> set port (port_id) mirror-rule (rule_id) \
1180 (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1181 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1183 Set link mirror rule for a port::
1185 testpmd> set port (port_id) mirror-rule (rule_id) \
1186 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1188 For example to enable mirror traffic with vlan 0,1 to pool 0::
1190 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1192 reset port - mirror rule
1193 ~~~~~~~~~~~~~~~~~~~~~~~~
1195 Reset a mirror rule for a port::
1197 testpmd> reset port (port_id) mirror-rule (rule_id)
1202 Set the flush on RX streams before forwarding.
1203 The default is flush ``on``.
1204 Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1206 testpmd> set flush_rx off
1211 Set the bypass mode for the lowest port on bypass enabled NIC::
1213 testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1218 Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1220 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1221 mode (normal|bypass|isolate) (port_id)
1225 * ``timeout``: Enable bypass after watchdog timeout.
1227 * ``os_on``: Enable bypass when OS/board is powered on.
1229 * ``os_off``: Enable bypass when OS/board is powered off.
1231 * ``power_on``: Enable bypass when power supply is turned on.
1233 * ``power_off``: Enable bypass when power supply is turned off.
1239 Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1241 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1246 Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1248 testpmd> show bypass config (port_id)
1253 Set link up for a port::
1255 testpmd> set link-up port (port id)
1260 Set link down for a port::
1262 testpmd> set link-down port (port id)
1267 Enable E-tag insertion for a VF on a port::
1269 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1271 Disable E-tag insertion for a VF on a port::
1273 testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1275 Enable/disable E-tag stripping on a port::
1277 testpmd> E-tag set stripping (on|off) port (port_id)
1279 Enable/disable E-tag based forwarding on a port::
1281 testpmd> E-tag set forwarding (on|off) port (port_id)
1283 Add an E-tag forwarding filter on a port::
1285 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1287 Delete an E-tag forwarding filter on a port::
1288 testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1293 Load a dynamic device personalization (DDP) package::
1295 testpmd> ddp add (port_id) (package_path[,output_path])
1300 Delete a dynamic device personalization package::
1302 testpmd> ddp del (port_id) (package_path)
1307 List all items from the ptype mapping table::
1309 testpmd> ptype mapping get (port_id) (valid_only)
1313 * ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0).
1315 Replace a specific or a group of software defined ptype with a new one::
1317 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type)
1321 * ``target``: A specific software ptype or a mask to represent a group of software ptypes.
1323 * ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1).
1325 * ``pkt_type``: The new software ptype to replace the old ones.
1327 Update hardware defined ptype to software defined packet type mapping table::
1329 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype)
1333 * ``hw_ptype``: hardware ptype as the index of the ptype mapping table.
1335 * ``sw_ptype``: software ptype as the value of the ptype mapping table.
1337 Reset ptype mapping table::
1339 testpmd> ptype mapping reset (port_id)
1344 The following sections show functions for configuring ports.
1348 Port configuration changes only become active when forwarding is started/restarted.
1353 Attach a port specified by pci address or virtual device args::
1355 testpmd> port attach (identifier)
1357 To attach a new pci device, the device should be recognized by kernel first.
1358 Then it should be moved under DPDK management.
1359 Finally the port can be attached to testpmd.
1361 For example, to move a pci device using ixgbe under DPDK management:
1363 .. code-block:: console
1365 # Check the status of the available devices.
1366 ./usertools/dpdk-devbind.py --status
1368 Network devices using DPDK-compatible driver
1369 ============================================
1372 Network devices using kernel driver
1373 ===================================
1374 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1377 # Bind the device to igb_uio.
1378 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1381 # Recheck the status of the devices.
1382 ./usertools/dpdk-devbind.py --status
1383 Network devices using DPDK-compatible driver
1384 ============================================
1385 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1387 To attach a port created by virtual device, above steps are not needed.
1389 For example, to attach a port whose pci address is 0000:0a:00.0.
1391 .. code-block:: console
1393 testpmd> port attach 0000:0a:00.0
1394 Attaching a new port...
1395 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1396 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1397 EAL: PCI memory mapped at 0x7f83bfa00000
1398 EAL: PCI memory mapped at 0x7f83bfa80000
1399 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1400 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1401 Port 0 is attached. Now total ports is 1
1404 For example, to attach a port created by pcap PMD.
1406 .. code-block:: console
1408 testpmd> port attach net_pcap0
1409 Attaching a new port...
1410 PMD: Initializing pmd_pcap for net_pcap0
1411 PMD: Creating pcap-backed ethdev on numa socket 0
1412 Port 0 is attached. Now total ports is 1
1415 In this case, identifier is ``net_pcap0``.
1416 This identifier format is the same as ``--vdev`` format of DPDK applications.
1418 For example, to re-attach a bonded port which has been previously detached,
1419 the mode and slave parameters must be given.
1421 .. code-block:: console
1423 testpmd> port attach net_bond_0,mode=0,slave=1
1424 Attaching a new port...
1425 EAL: Initializing pmd_bond for net_bond_0
1426 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1427 Port 0 is attached. Now total ports is 1
1434 Detach a specific port::
1436 testpmd> port detach (port_id)
1438 Before detaching a port, the port should be stopped and closed.
1440 For example, to detach a pci device port 0.
1442 .. code-block:: console
1444 testpmd> port stop 0
1447 testpmd> port close 0
1451 testpmd> port detach 0
1453 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1454 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1455 EAL: PCI memory unmapped at 0x7f83bfa00000
1456 EAL: PCI memory unmapped at 0x7f83bfa80000
1460 For example, to detach a virtual device port 0.
1462 .. code-block:: console
1464 testpmd> port stop 0
1467 testpmd> port close 0
1471 testpmd> port detach 0
1473 PMD: Closing pcap ethdev on numa socket 0
1474 Port 'net_pcap0' is detached. Now total ports is 0
1477 To remove a pci device completely from the system, first detach the port from testpmd.
1478 Then the device should be moved under kernel management.
1479 Finally the device can be removed using kernel pci hotplug functionality.
1481 For example, to move a pci device under kernel management:
1483 .. code-block:: console
1485 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1487 ./usertools/dpdk-devbind.py --status
1489 Network devices using DPDK-compatible driver
1490 ============================================
1493 Network devices using kernel driver
1494 ===================================
1495 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1497 To remove a port created by a virtual device, above steps are not needed.
1502 Start all ports or a specific port::
1504 testpmd> port start (port_id|all)
1509 Stop all ports or a specific port::
1511 testpmd> port stop (port_id|all)
1516 Close all ports or a specific port::
1518 testpmd> port close (port_id|all)
1520 port start/stop queue
1521 ~~~~~~~~~~~~~~~~~~~~~
1523 Start/stop a rx/tx queue on a specific port::
1525 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1527 Only take effect when port is started.
1532 Set the speed and duplex mode for all ports or a specific port::
1534 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1535 duplex (half|full|auto)
1537 port config - queues/descriptors
1538 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1540 Set number of queues/descriptors for rxq, txq, rxd and txd::
1542 testpmd> port config all (rxq|txq|rxd|txd) (value)
1544 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1546 port config - max-pkt-len
1547 ~~~~~~~~~~~~~~~~~~~~~~~~~
1549 Set the maximum packet length::
1551 testpmd> port config all max-pkt-len (value)
1553 This is equivalent to the ``--max-pkt-len`` command-line option.
1555 port config - CRC Strip
1556 ~~~~~~~~~~~~~~~~~~~~~~~
1558 Set hardware CRC stripping on or off for all ports::
1560 testpmd> port config all crc-strip (on|off)
1562 CRC stripping is on by default.
1564 The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option.
1566 port config - scatter
1567 ~~~~~~~~~~~~~~~~~~~~~~~
1569 Set RX scatter mode on or off for all ports::
1571 testpmd> port config all scatter (on|off)
1573 RX scatter mode is off by default.
1575 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1577 port config - TX queue flags
1578 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1580 Set a hexadecimal bitmap of TX queue flags for all ports::
1582 testpmd> port config all txqflags value
1584 This command is equivalent to the ``--txqflags`` command-line option.
1586 port config - RX Checksum
1587 ~~~~~~~~~~~~~~~~~~~~~~~~~
1589 Set hardware RX checksum offload to on or off for all ports::
1591 testpmd> port config all rx-cksum (on|off)
1593 Checksum offload is off by default.
1595 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1600 Set hardware VLAN on or off for all ports::
1602 testpmd> port config all hw-vlan (on|off)
1604 Hardware VLAN is on by default.
1606 The ``off`` option is equivalent to the ``--disable-hw-vlan`` command-line option.
1608 port config - VLAN filter
1609 ~~~~~~~~~~~~~~~~~~~~~~~~~
1611 Set hardware VLAN filter on or off for all ports::
1613 testpmd> port config all hw-vlan-filter (on|off)
1615 Hardware VLAN filter is on by default.
1617 The ``off`` option is equivalent to the ``--disable-hw-vlan-filter`` command-line option.
1619 port config - VLAN strip
1620 ~~~~~~~~~~~~~~~~~~~~~~~~
1622 Set hardware VLAN strip on or off for all ports::
1624 testpmd> port config all hw-vlan-strip (on|off)
1626 Hardware VLAN strip is on by default.
1628 The ``off`` option is equivalent to the ``--disable-hw-vlan-strip`` command-line option.
1630 port config - VLAN extend
1631 ~~~~~~~~~~~~~~~~~~~~~~~~~
1633 Set hardware VLAN extend on or off for all ports::
1635 testpmd> port config all hw-vlan-extend (on|off)
1637 Hardware VLAN extend is off by default.
1639 The ``off`` option is equivalent to the ``--disable-hw-vlan-extend`` command-line option.
1641 port config - Drop Packets
1642 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1644 Set packet drop for packets with no descriptors on or off for all ports::
1646 testpmd> port config all drop-en (on|off)
1648 Packet dropping for packets with no descriptors is off by default.
1650 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1655 Set the RSS (Receive Side Scaling) mode on or off::
1657 testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1659 RSS is on by default.
1661 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1663 port config - RSS Reta
1664 ~~~~~~~~~~~~~~~~~~~~~~
1666 Set the RSS (Receive Side Scaling) redirection table::
1668 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1673 Set the DCB mode for an individual port::
1675 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1677 The traffic class should be 4 or 8.
1682 Set the number of packets per burst::
1684 testpmd> port config all burst (value)
1686 This is equivalent to the ``--burst`` command-line option.
1688 port config - Threshold
1689 ~~~~~~~~~~~~~~~~~~~~~~~
1691 Set thresholds for TX/RX queues::
1693 testpmd> port config all (threshold) (value)
1695 Where the threshold type can be:
1697 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1699 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1701 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1703 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1705 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1707 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1709 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1711 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1713 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1715 These threshold options are also available from the command-line.
1720 Set the value of ether-type for E-tag::
1722 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1724 Enable/disable the E-tag support::
1726 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1728 port config pctype mapping
1729 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1731 Reset pctype mapping table::
1733 testpmd> port config (port_id) pctype mapping reset
1735 Update hardware defined pctype to software defined flow type mapping table::
1737 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id)
1741 * ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table.
1743 * ``flow_type_id``: software flow type id as the index of the pctype mapping table.
1746 Link Bonding Functions
1747 ----------------------
1749 The Link Bonding functions make it possible to dynamically create and
1750 manage link bonding devices from within testpmd interactive prompt.
1752 create bonded device
1753 ~~~~~~~~~~~~~~~~~~~~
1755 Create a new bonding device::
1757 testpmd> create bonded device (mode) (socket)
1759 For example, to create a bonded device in mode 1 on socket 0::
1761 testpmd> create bonded 1 0
1762 created new bonded device (port X)
1767 Adds Ethernet device to a Link Bonding device::
1769 testpmd> add bonding slave (slave id) (port id)
1771 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
1773 testpmd> add bonding slave 6 10
1776 remove bonding slave
1777 ~~~~~~~~~~~~~~~~~~~~
1779 Removes an Ethernet slave device from a Link Bonding device::
1781 testpmd> remove bonding slave (slave id) (port id)
1783 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
1785 testpmd> remove bonding slave 6 10
1790 Set the Link Bonding mode of a Link Bonding device::
1792 testpmd> set bonding mode (value) (port id)
1794 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
1796 testpmd> set bonding mode 3 10
1801 Set an Ethernet slave device as the primary device on a Link Bonding device::
1803 testpmd> set bonding primary (slave id) (port id)
1805 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
1807 testpmd> set bonding primary 6 10
1812 Set the MAC address of a Link Bonding device::
1814 testpmd> set bonding mac (port id) (mac)
1816 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
1818 testpmd> set bonding mac 10 00:00:00:00:00:01
1820 set bonding xmit_balance_policy
1821 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1823 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
1825 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
1827 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
1829 testpmd> set bonding xmit_balance_policy 10 l34
1832 set bonding mon_period
1833 ~~~~~~~~~~~~~~~~~~~~~~
1835 Set the link status monitoring polling period in milliseconds for a bonding device.
1837 This adds support for PMD slave devices which do not support link status interrupts.
1838 When the mon_period is set to a value greater than 0 then all PMD's which do not support
1839 link status ISR will be queried every polling interval to check if their link status has changed::
1841 testpmd> set bonding mon_period (port_id) (value)
1843 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
1845 testpmd> set bonding mon_period 5 150
1848 set bonding lacp dedicated_queue
1849 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1851 Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic
1852 when in mode 4 (link-aggregration-802.3ad)::
1854 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable)
1857 set bonding agg_mode
1858 ~~~~~~~~~~~~~~~~~~~~
1860 Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad)::
1862 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable)
1868 Show the current configuration of a Link Bonding device::
1870 testpmd> show bonding config (port id)
1873 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
1874 in balance mode with a transmission policy of layer 2+3::
1876 testpmd> show bonding config 9
1878 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
1880 Active Slaves (3): [1 3 4]
1887 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
1888 This is mainly useful for debugging purposes.
1889 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
1890 and fields that can be accessed.
1895 Display the value of a port register::
1897 testpmd> read reg (port_id) (address)
1899 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
1901 testpmd> read reg 0 0xEE00
1902 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
1907 Display a port register bit field::
1909 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
1911 For example, reading the lowest two bits from the register in the example above::
1913 testpmd> read regfield 0 0xEE00 0 1
1914 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
1919 Display a single port register bit::
1921 testpmd> read regbit (port_id) (address) (bit_x)
1923 For example, reading the lowest bit from the register in the example above::
1925 testpmd> read regbit 0 0xEE00 0
1926 port 0 PCI register at offset 0xEE00: bit 0=1
1931 Set the value of a port register::
1933 testpmd> write reg (port_id) (address) (value)
1935 For example, to clear a register::
1937 testpmd> write reg 0 0xEE00 0x0
1938 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
1943 Set bit field of a port register::
1945 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
1947 For example, writing to the register cleared in the example above::
1949 testpmd> write regfield 0 0xEE00 0 1 2
1950 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
1955 Set single bit value of a port register::
1957 testpmd> write regbit (port_id) (address) (bit_x) (value)
1959 For example, to set the high bit in the register from the example above::
1961 testpmd> write regbit 0 0xEE00 31 1
1962 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
1968 This section details the available filter functions that are available.
1970 Note these functions interface the deprecated legacy filtering framework,
1971 superseded by *rte_flow*. See `Flow rules management`_.
1974 ~~~~~~~~~~~~~~~~~~~~
1976 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
1978 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
1979 ethertype (ether_type) (drop|fwd) queue (queue_id)
1981 The available information parameters are:
1983 * ``port_id``: The port which the Ethertype filter assigned on.
1985 * ``mac_addr``: Compare destination mac address.
1987 * ``mac_ignr``: Ignore destination mac address match.
1989 * ``mac_address``: Destination mac address to match.
1991 * ``ether_type``: The EtherType value want to match,
1992 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
1994 * ``queue_id``: The receive queue associated with this EtherType filter.
1995 It is meaningless when deleting or dropping.
1997 Example, to add/remove an ethertype filter rule::
1999 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
2000 ethertype 0x0806 fwd queue 3
2002 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
2003 ethertype 0x0806 fwd queue 3
2008 Add or delete a 2-tuple filter,
2009 which identifies packets by specific protocol and destination TCP/UDP port
2010 and forwards packets into one of the receive queues::
2012 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
2013 protocol (protocol_value) mask (mask_value) \
2014 tcp_flags (tcp_flags_value) priority (prio_value) \
2017 The available information parameters are:
2019 * ``port_id``: The port which the 2-tuple filter assigned on.
2021 * ``dst_port_value``: Destination port in L4.
2023 * ``protocol_value``: IP L4 protocol.
2025 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
2027 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
2029 * ``prio_value``: Priority of this filter.
2031 * ``queue_id``: The receive queue associated with this 2-tuple filter.
2033 Example, to add/remove an 2tuple filter rule::
2035 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
2036 tcp_flags 0x02 priority 3 queue 3
2038 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
2039 tcp_flags 0x02 priority 3 queue 3
2044 Add or delete a 5-tuple filter,
2045 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
2046 and routes packets into one of the receive queues::
2048 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
2049 (src_address) dst_port (dst_port_value) \
2050 src_port (src_port_value) protocol (protocol_value) \
2051 mask (mask_value) tcp_flags (tcp_flags_value) \
2052 priority (prio_value) queue (queue_id)
2054 The available information parameters are:
2056 * ``port_id``: The port which the 5-tuple filter assigned on.
2058 * ``dst_address``: Destination IP address.
2060 * ``src_address``: Source IP address.
2062 * ``dst_port_value``: TCP/UDP destination port.
2064 * ``src_port_value``: TCP/UDP source port.
2066 * ``protocol_value``: L4 protocol.
2068 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
2070 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
2072 * ``prio_value``: The priority of this filter.
2074 * ``queue_id``: The receive queue associated with this 5-tuple filter.
2076 Example, to add/remove an 5tuple filter rule::
2078 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2079 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2080 flags 0x0 priority 3 queue 3
2082 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
2083 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
2084 flags 0x0 priority 3 queue 3
2089 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
2091 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
2093 The available information parameters are:
2095 * ``port_id``: The port which the SYN filter assigned on.
2097 * ``high``: This SYN filter has higher priority than other filters.
2099 * ``low``: This SYN filter has lower priority than other filters.
2101 * ``queue_id``: The receive queue associated with this SYN filter
2105 testpmd> syn_filter 0 add priority high queue 3
2110 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
2111 and routed into one of the receive queues::
2113 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
2114 mask (mask_value) priority (prio_value) queue (queue_id)
2116 The available information parameters are:
2118 * ``port_id``: The port which the Flex filter is assigned on.
2120 * ``len_value``: Filter length in bytes, no greater than 128.
2122 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
2124 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
2126 * ``prio_value``: The priority of this filter.
2128 * ``queue_id``: The receive queue associated with this Flex filter.
2132 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
2133 mask 000C priority 3 queue 3
2135 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
2136 mask 000C priority 3 queue 3
2139 .. _testpmd_flow_director:
2141 flow_director_filter
2142 ~~~~~~~~~~~~~~~~~~~~
2144 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
2146 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
2147 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
2149 * Perfect match filters.
2150 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2151 The masked fields are for IP flow.
2153 * Signature filters.
2154 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
2156 * Perfect-mac-vlan match filters.
2157 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2158 The masked fields are for MAC VLAN flow.
2160 * Perfect-tunnel match filters.
2161 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2162 The masked fields are for tunnel flow.
2164 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
2165 per flow type and the flexible payload.
2167 The Flow Director can also mask out parts of all of these fields so that filters
2168 are only applied to certain fields or parts of the fields.
2170 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
2172 # Commands to add flow director filters of different flow types::
2174 flow_director_filter (port_id) mode IP (add|del|update) \
2175 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
2176 src (src_ip_address) dst (dst_ip_address) \
2177 tos (tos_value) proto (proto_value) ttl (ttl_value) \
2178 vlan (vlan_value) flexbytes (flexbytes_value) \
2179 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
2182 flow_director_filter (port_id) mode IP (add|del|update) \
2183 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
2184 src (src_ip_address) (src_port) \
2185 dst (dst_ip_address) (dst_port) \
2186 tos (tos_value) ttl (ttl_value) \
2187 vlan (vlan_value) flexbytes (flexbytes_value) \
2188 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
2191 flow_director_filter (port_id) mode IP (add|del|update) \
2192 flow (ipv4-sctp|ipv6-sctp) \
2193 src (src_ip_address) (src_port) \
2194 dst (dst_ip_address) (dst_port) \
2195 tos (tos_value) ttl (ttl_value) \
2196 tag (verification_tag) vlan (vlan_value) \
2197 flexbytes (flexbytes_value) (drop|fwd) \
2198 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
2200 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
2201 ether (ethertype) flexbytes (flexbytes_value) \
2202 (drop|fwd) pf|vf(vf_id) queue (queue_id)
2205 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
2206 mac (mac_address) vlan (vlan_value) \
2207 flexbytes (flexbytes_value) (drop|fwd) \
2208 queue (queue_id) fd_id (fd_id_value)
2210 flow_director_filter (port_id) mode Tunnel (add|del|update) \
2211 mac (mac_address) vlan (vlan_value) \
2212 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
2213 flexbytes (flexbytes_value) (drop|fwd) \
2214 queue (queue_id) fd_id (fd_id_value)
2216 For example, to add an ipv4-udp flow type filter::
2218 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
2219 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
2220 fwd pf queue 1 fd_id 1
2222 For example, add an ipv4-other flow type filter::
2224 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
2225 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
2226 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
2231 Flush all flow director filters on a device::
2233 testpmd> flush_flow_director (port_id)
2235 Example, to flush all flow director filter on port 0::
2237 testpmd> flush_flow_director 0
2242 Set flow director's input masks::
2244 flow_director_mask (port_id) mode IP vlan (vlan_value) \
2245 src_mask (ipv4_src) (ipv6_src) (src_port) \
2246 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2248 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2250 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2251 mac (mac_value) tunnel-type (tunnel_type_value) \
2252 tunnel-id (tunnel_id_value)
2254 Example, to set flow director mask on port 0::
2256 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2257 src_mask 255.255.255.255 \
2258 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2259 dst_mask 255.255.255.255 \
2260 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2262 flow_director_flex_mask
2263 ~~~~~~~~~~~~~~~~~~~~~~~
2265 set masks of flow director's flexible payload based on certain flow type::
2267 testpmd> flow_director_flex_mask (port_id) \
2268 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2269 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2270 l2_payload|all) (mask)
2272 Example, to set flow director's flex mask for all flow type on port 0::
2274 testpmd> flow_director_flex_mask 0 flow all \
2275 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2278 flow_director_flex_payload
2279 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2281 Configure flexible payload selection::
2283 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2285 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2287 testpmd> flow_director_flex_payload 0 l4 \
2288 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2290 get_sym_hash_ena_per_port
2291 ~~~~~~~~~~~~~~~~~~~~~~~~~
2293 Get symmetric hash enable configuration per port::
2295 get_sym_hash_ena_per_port (port_id)
2297 For example, to get symmetric hash enable configuration of port 1::
2299 testpmd> get_sym_hash_ena_per_port 1
2301 set_sym_hash_ena_per_port
2302 ~~~~~~~~~~~~~~~~~~~~~~~~~
2304 Set symmetric hash enable configuration per port to enable or disable::
2306 set_sym_hash_ena_per_port (port_id) (enable|disable)
2308 For example, to set symmetric hash enable configuration of port 1 to enable::
2310 testpmd> set_sym_hash_ena_per_port 1 enable
2312 get_hash_global_config
2313 ~~~~~~~~~~~~~~~~~~~~~~
2315 Get the global configurations of hash filters::
2317 get_hash_global_config (port_id)
2319 For example, to get the global configurations of hash filters of port 1::
2321 testpmd> get_hash_global_config 1
2323 set_hash_global_config
2324 ~~~~~~~~~~~~~~~~~~~~~~
2326 Set the global configurations of hash filters::
2328 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2329 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2330 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload) \
2333 For example, to enable simple_xor for flow type of ipv6 on port 2::
2335 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2340 Set the input set for hash::
2342 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2343 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2344 l2_payload) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2345 ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2346 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2347 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2348 fld-8th|none) (select|add)
2350 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2352 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2357 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2358 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2360 Set the input set for flow director::
2362 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2363 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2364 l2_payload) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2365 ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2366 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2367 sctp-dst-port|sctp-veri-tag|none) (select|add)
2369 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2371 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2376 Set different GRE key length for input set::
2378 global_config (port_id) gre-key-len (number in bytes)
2380 For example to set GRE key length for input set to 4 bytes on port 0::
2382 testpmd> global_config 0 gre-key-len 4
2385 .. _testpmd_rte_flow:
2387 Flow rules management
2388 ---------------------
2390 Control of the generic flow API (*rte_flow*) is fully exposed through the
2391 ``flow`` command (validation, creation, destruction, queries and operation
2394 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2395 features simultaneously may cause undefined side-effects and is therefore
2401 Because the ``flow`` command uses dynamic tokens to handle the large number
2402 of possible flow rules combinations, its behavior differs slightly from
2403 other commands, in particular:
2405 - Pressing *?* or the *<tab>* key displays contextual help for the current
2406 token, not that of the entire command.
2408 - Optional and repeated parameters are supported (provided they are listed
2409 in the contextual help).
2411 The first parameter stands for the operation mode. Possible operations and
2412 their general syntax are described below. They are covered in detail in the
2415 - Check whether a flow rule can be created::
2417 flow validate {port_id}
2418 [group {group_id}] [priority {level}] [ingress] [egress]
2419 pattern {item} [/ {item} [...]] / end
2420 actions {action} [/ {action} [...]] / end
2422 - Create a flow rule::
2424 flow create {port_id}
2425 [group {group_id}] [priority {level}] [ingress] [egress]
2426 pattern {item} [/ {item} [...]] / end
2427 actions {action} [/ {action} [...]] / end
2429 - Destroy specific flow rules::
2431 flow destroy {port_id} rule {rule_id} [...]
2433 - Destroy all flow rules::
2435 flow flush {port_id}
2437 - Query an existing flow rule::
2439 flow query {port_id} {rule_id} {action}
2441 - List existing flow rules sorted by priority, filtered by group
2444 flow list {port_id} [group {group_id}] [...]
2446 - Restrict ingress traffic to the defined flow rules::
2448 flow isolate {port_id} {boolean}
2450 Validating flow rules
2451 ~~~~~~~~~~~~~~~~~~~~~
2453 ``flow validate`` reports whether a flow rule would be accepted by the
2454 underlying device in its current state but stops short of creating it. It is
2455 bound to ``rte_flow_validate()``::
2457 flow validate {port_id}
2458 [group {group_id}] [priority {level}] [ingress] [egress]
2459 pattern {item} [/ {item} [...]] / end
2460 actions {action} [/ {action} [...]] / end
2462 If successful, it will show::
2466 Otherwise it will show an error message of the form::
2468 Caught error type [...] ([...]): [...]
2470 This command uses the same parameters as ``flow create``, their format is
2471 described in `Creating flow rules`_.
2473 Check whether redirecting any Ethernet packet received on port 0 to RX queue
2474 index 6 is supported::
2476 testpmd> flow validate 0 ingress pattern eth / end
2477 actions queue index 6 / end
2481 Port 0 does not support TCPv6 rules::
2483 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
2485 Caught error type 9 (specific pattern item): Invalid argument
2491 ``flow create`` validates and creates the specified flow rule. It is bound
2492 to ``rte_flow_create()``::
2494 flow create {port_id}
2495 [group {group_id}] [priority {level}] [ingress] [egress]
2496 pattern {item} [/ {item} [...]] / end
2497 actions {action} [/ {action} [...]] / end
2499 If successful, it will return a flow rule ID usable with other commands::
2501 Flow rule #[...] created
2503 Otherwise it will show an error message of the form::
2505 Caught error type [...] ([...]): [...]
2507 Parameters describe in the following order:
2509 - Attributes (*group*, *priority*, *ingress*, *egress* tokens).
2510 - A matching pattern, starting with the *pattern* token and terminated by an
2512 - Actions, starting with the *actions* token and terminated by an *end*
2515 These translate directly to *rte_flow* objects provided as-is to the
2516 underlying functions.
2518 The shortest valid definition only comprises mandatory tokens::
2520 testpmd> flow create 0 pattern end actions end
2522 Note that PMDs may refuse rules that essentially do nothing such as this
2525 **All unspecified object values are automatically initialized to 0.**
2530 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
2531 specified before the ``pattern`` token.
2533 - ``group {group id}``: priority group.
2534 - ``priority {level}``: priority level within group.
2535 - ``ingress``: rule applies to ingress traffic.
2536 - ``egress``: rule applies to egress traffic.
2538 Each instance of an attribute specified several times overrides the previous
2539 value as shown below (group 4 is used)::
2541 testpmd> flow create 0 group 42 group 24 group 4 [...]
2543 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
2545 While not specifying a direction is an error, some rules may allow both
2548 Most rules affect RX therefore contain the ``ingress`` token::
2550 testpmd> flow create 0 ingress pattern [...]
2555 A matching pattern starts after the ``pattern`` token. It is made of pattern
2556 items and is terminated by a mandatory ``end`` item.
2558 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
2559 rte_flow_item_type``).
2561 The ``/`` token is used as a separator between pattern items as shown
2564 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
2566 Note that protocol items like these must be stacked from lowest to highest
2567 layer to make sense. For instance, the following rule is either invalid or
2568 unlikely to match any packet::
2570 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
2572 More information on these restrictions can be found in the *rte_flow*
2575 Several items support additional specification structures, for example
2576 ``ipv4`` allows specifying source and destination addresses as follows::
2578 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
2579 dst is 10.2.0.0 / end [...]
2581 This rule matches all IPv4 traffic with the specified properties.
2583 In this example, ``src`` and ``dst`` are field names of the underlying
2584 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
2585 in a similar fashion.
2587 The ``is`` token means that the subsequent value must be matched exactly,
2588 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
2589 accordingly. Possible assignment tokens are:
2591 - ``is``: match value perfectly (with full bit-mask).
2592 - ``spec``: match value according to configured bit-mask.
2593 - ``last``: specify upper bound to establish a range.
2594 - ``mask``: specify bit-mask with relevant bits set to one.
2595 - ``prefix``: generate bit-mask from a prefix length.
2597 These yield identical results::
2599 ipv4 src is 10.1.1.1
2603 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
2607 ipv4 src spec 10.1.1.1 src prefix 32
2611 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
2615 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
2617 Inclusive ranges can be defined with ``last``::
2619 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
2621 Note that ``mask`` affects both ``spec`` and ``last``::
2623 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
2624 # matches 10.1.0.0 to 10.2.255.255
2626 Properties can be modified multiple times::
2628 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
2632 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
2637 This section lists supported pattern items and their attributes, if any.
2639 - ``end``: end list of pattern items.
2641 - ``void``: no-op pattern item.
2643 - ``invert``: perform actions when pattern does not match.
2645 - ``any``: match any protocol for the current layer.
2647 - ``num {unsigned}``: number of layers covered.
2649 - ``pf``: match packets addressed to the physical function.
2651 - ``vf``: match packets addressed to a virtual function ID.
2653 - ``id {unsigned}``: destination VF ID.
2655 - ``port``: device-specific physical port index to use.
2657 - ``index {unsigned}``: physical port index.
2659 - ``raw``: match an arbitrary byte string.
2661 - ``relative {boolean}``: look for pattern after the previous item.
2662 - ``search {boolean}``: search pattern from offset (see also limit).
2663 - ``offset {integer}``: absolute or relative offset for pattern.
2664 - ``limit {unsigned}``: search area limit for start of pattern.
2665 - ``pattern {string}``: byte string to look for.
2667 - ``eth``: match Ethernet header.
2669 - ``dst {MAC-48}``: destination MAC.
2670 - ``src {MAC-48}``: source MAC.
2671 - ``type {unsigned}``: EtherType.
2673 - ``vlan``: match 802.1Q/ad VLAN tag.
2675 - ``tpid {unsigned}``: tag protocol identifier.
2676 - ``tci {unsigned}``: tag control information.
2677 - ``pcp {unsigned}``: priority code point.
2678 - ``dei {unsigned}``: drop eligible indicator.
2679 - ``vid {unsigned}``: VLAN identifier.
2681 - ``ipv4``: match IPv4 header.
2683 - ``tos {unsigned}``: type of service.
2684 - ``ttl {unsigned}``: time to live.
2685 - ``proto {unsigned}``: next protocol ID.
2686 - ``src {ipv4 address}``: source address.
2687 - ``dst {ipv4 address}``: destination address.
2689 - ``ipv6``: match IPv6 header.
2691 - ``tc {unsigned}``: traffic class.
2692 - ``flow {unsigned}``: flow label.
2693 - ``proto {unsigned}``: protocol (next header).
2694 - ``hop {unsigned}``: hop limit.
2695 - ``src {ipv6 address}``: source address.
2696 - ``dst {ipv6 address}``: destination address.
2698 - ``icmp``: match ICMP header.
2700 - ``type {unsigned}``: ICMP packet type.
2701 - ``code {unsigned}``: ICMP packet code.
2703 - ``udp``: match UDP header.
2705 - ``src {unsigned}``: UDP source port.
2706 - ``dst {unsigned}``: UDP destination port.
2708 - ``tcp``: match TCP header.
2710 - ``src {unsigned}``: TCP source port.
2711 - ``dst {unsigned}``: TCP destination port.
2713 - ``sctp``: match SCTP header.
2715 - ``src {unsigned}``: SCTP source port.
2716 - ``dst {unsigned}``: SCTP destination port.
2717 - ``tag {unsigned}``: validation tag.
2718 - ``cksum {unsigned}``: checksum.
2720 - ``vxlan``: match VXLAN header.
2722 - ``vni {unsigned}``: VXLAN identifier.
2724 - ``e_tag``: match IEEE 802.1BR E-Tag header.
2726 - ``grp_ecid_b {unsigned}``: GRP and E-CID base.
2728 - ``nvgre``: match NVGRE header.
2730 - ``tni {unsigned}``: virtual subnet ID.
2732 - ``mpls``: match MPLS header.
2734 - ``label {unsigned}``: MPLS label.
2736 - ``gre``: match GRE header.
2738 - ``protocol {unsigned}``: protocol type.
2740 - ``fuzzy``: fuzzy pattern match, expect faster than default.
2742 - ``thresh {unsigned}``: accuracy threshold.
2744 - ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header.
2746 - ``teid {unsigned}``: tunnel endpoint identifier.
2751 A list of actions starts after the ``actions`` token in the same fashion as
2752 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
2753 terminated by a mandatory ``end`` action.
2755 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
2756 rte_flow_action_type``).
2758 Dropping all incoming UDPv4 packets can be expressed as follows::
2760 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2763 Several actions have configurable properties which must be specified when
2764 there is no valid default value. For example, ``queue`` requires a target
2767 This rule redirects incoming UDPv4 traffic to queue index 6::
2769 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2770 actions queue index 6 / end
2772 While this one could be rejected by PMDs (unspecified queue index)::
2774 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2777 As defined by *rte_flow*, the list is not ordered, all actions of a given
2778 rule are performed simultaneously. These are equivalent::
2780 queue index 6 / void / mark id 42 / end
2784 void / mark id 42 / queue index 6 / end
2786 All actions in a list should have different types, otherwise only the last
2787 action of a given type is taken into account::
2789 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
2793 drop / drop / drop / end # drop is performed only once
2797 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
2799 Considering they are performed simultaneously, opposite and overlapping
2800 actions can sometimes be combined when the end result is unambiguous::
2802 drop / queue index 6 / end # drop has no effect
2806 drop / dup index 6 / end # same as above
2810 queue index 6 / rss queues 6 7 8 / end # queue has no effect
2814 drop / passthru / end # drop has no effect
2816 Note that PMDs may still refuse such combinations.
2821 This section lists supported actions and their attributes, if any.
2823 - ``end``: end list of actions.
2825 - ``void``: no-op action.
2827 - ``passthru``: let subsequent rule process matched packets.
2829 - ``mark``: attach 32 bit value to packets.
2831 - ``id {unsigned}``: 32 bit value to return with packets.
2833 - ``flag``: flag packets.
2835 - ``queue``: assign packets to a given queue index.
2837 - ``index {unsigned}``: queue index to use.
2839 - ``drop``: drop packets (note: passthru has priority).
2841 - ``count``: enable counters for this rule.
2843 - ``dup``: duplicate packets to a given queue index.
2845 - ``index {unsigned}``: queue index to duplicate packets to.
2847 - ``rss``: spread packets among several queues.
2849 - ``queues [{unsigned} [...]] end``: queue indices to use.
2851 - ``pf``: redirect packets to physical device function.
2853 - ``vf``: redirect packets to virtual device function.
2855 - ``original {boolean}``: use original VF ID if possible.
2856 - ``id {unsigned}``: VF ID to redirect packets to.
2858 Destroying flow rules
2859 ~~~~~~~~~~~~~~~~~~~~~
2861 ``flow destroy`` destroys one or more rules from their rule ID (as returned
2862 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
2863 times as necessary::
2865 flow destroy {port_id} rule {rule_id} [...]
2867 If successful, it will show::
2869 Flow rule #[...] destroyed
2871 It does not report anything for rule IDs that do not exist. The usual error
2872 message is shown when a rule cannot be destroyed::
2874 Caught error type [...] ([...]): [...]
2876 ``flow flush`` destroys all rules on a device and does not take extra
2877 arguments. It is bound to ``rte_flow_flush()``::
2879 flow flush {port_id}
2881 Any errors are reported as above.
2883 Creating several rules and destroying them::
2885 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2886 actions queue index 2 / end
2887 Flow rule #0 created
2888 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2889 actions queue index 3 / end
2890 Flow rule #1 created
2891 testpmd> flow destroy 0 rule 0 rule 1
2892 Flow rule #1 destroyed
2893 Flow rule #0 destroyed
2896 The same result can be achieved using ``flow flush``::
2898 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2899 actions queue index 2 / end
2900 Flow rule #0 created
2901 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2902 actions queue index 3 / end
2903 Flow rule #1 created
2904 testpmd> flow flush 0
2907 Non-existent rule IDs are ignored::
2909 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2910 actions queue index 2 / end
2911 Flow rule #0 created
2912 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2913 actions queue index 3 / end
2914 Flow rule #1 created
2915 testpmd> flow destroy 0 rule 42 rule 10 rule 2
2917 testpmd> flow destroy 0 rule 0
2918 Flow rule #0 destroyed
2924 ``flow query`` queries a specific action of a flow rule having that
2925 ability. Such actions collect information that can be reported using this
2926 command. It is bound to ``rte_flow_query()``::
2928 flow query {port_id} {rule_id} {action}
2930 If successful, it will display either the retrieved data for known actions
2931 or the following message::
2933 Cannot display result for action type [...] ([...])
2935 Otherwise, it will complain either that the rule does not exist or that some
2938 Flow rule #[...] not found
2942 Caught error type [...] ([...]): [...]
2944 Currently only the ``count`` action is supported. This action reports the
2945 number of packets that hit the flow rule and the total number of bytes. Its
2946 output has the following format::
2949 hits_set: [...] # whether "hits" contains a valid value
2950 bytes_set: [...] # whether "bytes" contains a valid value
2951 hits: [...] # number of packets
2952 bytes: [...] # number of bytes
2954 Querying counters for TCPv6 packets redirected to queue 6::
2956 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
2957 actions queue index 6 / count / end
2958 Flow rule #4 created
2959 testpmd> flow query 0 4 count
2970 ``flow list`` lists existing flow rules sorted by priority and optionally
2971 filtered by group identifiers::
2973 flow list {port_id} [group {group_id}] [...]
2975 This command only fails with the following message if the device does not
2980 Output consists of a header line followed by a short description of each
2981 flow rule, one per line. There is no output at all when no flow rules are
2982 configured on the device::
2984 ID Group Prio Attr Rule
2985 [...] [...] [...] [...] [...]
2987 ``Attr`` column flags:
2989 - ``i`` for ``ingress``.
2990 - ``e`` for ``egress``.
2992 Creating several flow rules and listing them::
2994 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2995 actions queue index 6 / end
2996 Flow rule #0 created
2997 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2998 actions queue index 2 / end
2999 Flow rule #1 created
3000 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
3001 actions rss queues 6 7 8 end / end
3002 Flow rule #2 created
3003 testpmd> flow list 0
3004 ID Group Prio Attr Rule
3005 0 0 0 i- ETH IPV4 => QUEUE
3006 1 0 0 i- ETH IPV6 => QUEUE
3007 2 0 5 i- ETH IPV4 UDP => RSS
3010 Rules are sorted by priority (i.e. group ID first, then priority level)::
3012 testpmd> flow list 1
3013 ID Group Prio Attr Rule
3014 0 0 0 i- ETH => COUNT
3015 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3016 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3017 1 24 0 i- ETH IPV4 UDP => QUEUE
3018 4 24 10 i- ETH IPV4 TCP => DROP
3019 3 24 20 i- ETH IPV4 => DROP
3020 2 24 42 i- ETH IPV4 UDP => QUEUE
3021 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3024 Output can be limited to specific groups::
3026 testpmd> flow list 1 group 0 group 63
3027 ID Group Prio Attr Rule
3028 0 0 0 i- ETH => COUNT
3029 6 0 500 i- ETH IPV6 TCP => DROP COUNT
3030 5 0 1000 i- ETH IPV6 ICMP => QUEUE
3031 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
3034 Toggling isolated mode
3035 ~~~~~~~~~~~~~~~~~~~~~~
3037 ``flow isolate`` can be used to tell the underlying PMD that ingress traffic
3038 must only be injected from the defined flow rules; that no default traffic
3039 is expected outside those rules and the driver is free to assign more
3040 resources to handle them. It is bound to ``rte_flow_isolate()``::
3042 flow isolate {port_id} {boolean}
3044 If successful, enabling or disabling isolated mode shows either::
3046 Ingress traffic on port [...]
3047 is now restricted to the defined flow rules
3051 Ingress traffic on port [...]
3052 is not restricted anymore to the defined flow rules
3054 Otherwise, in case of error::
3056 Caught error type [...] ([...]): [...]
3058 Mainly due to its side effects, PMDs supporting this mode may not have the
3059 ability to toggle it more than once without reinitializing affected ports
3060 first (e.g. by exiting testpmd).
3062 Enabling isolated mode::
3064 testpmd> flow isolate 0 true
3065 Ingress traffic on port 0 is now restricted to the defined flow rules
3068 Disabling isolated mode::
3070 testpmd> flow isolate 0 false
3071 Ingress traffic on port 0 is not restricted anymore to the defined flow rules
3074 Sample QinQ flow rules
3075 ~~~~~~~~~~~~~~~~~~~~~~
3077 Before creating QinQ rule(s) the following commands should be issued to enable QinQ::
3079 testpmd> port stop 0
3080 testpmd> vlan set qinq on 0
3082 The above command sets the inner and outer TPID's to 0x8100.
3084 To change the TPID's the following commands should be used::
3086 testpmd> vlan set outer tpid 0xa100 0
3087 testpmd> vlan set inner tpid 0x9100 0
3088 testpmd> port start 0
3090 Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM.
3094 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 /
3095 vlan tci is 456 / end actions vf id 1 / queue index 0 / end
3096 Flow rule #0 validated
3098 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 /
3099 vlan tci is 456 / end actions vf id 123 / queue index 0 / end
3100 Flow rule #0 created
3102 testpmd> flow list 0
3103 ID Group Prio Attr Rule
3104 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3106 Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host.
3110 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 /
3111 vlan tci is 654 / end actions pf / queue index 0 / end
3112 Flow rule #1 validated
3114 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 /
3115 vlan tci is 654 / end actions pf / queue index 1 / end
3116 Flow rule #1 created
3118 testpmd> flow list 0
3119 ID Group Prio Attr Rule
3120 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3121 1 0 0 i- ETH VLAN VLAN=>PF QUEUE