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_patch)
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 Configuration Functions
412 -----------------------
414 The testpmd application can be configured from the runtime as well as from the command-line.
416 This section details the available configuration functions that are available.
420 Configuration changes only become active when forwarding is started/restarted.
425 Reset forwarding to the default configuration::
432 Set the debug verbosity level::
434 testpmd> set verbose (level)
436 Currently the only available levels are 0 (silent except for error) and 1 (fully verbose).
441 Set the number of ports used by the application:
445 This is equivalent to the ``--nb-ports`` command-line option.
450 Set the number of cores used by the application::
452 testpmd> set nbcore (num)
454 This is equivalent to the ``--nb-cores`` command-line option.
458 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port.
463 Set the forwarding cores hexadecimal mask::
465 testpmd> set coremask (mask)
467 This is equivalent to the ``--coremask`` command-line option.
471 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding.
476 Set the forwarding ports hexadecimal mask::
478 testpmd> set portmask (mask)
480 This is equivalent to the ``--portmask`` command-line option.
485 Set number of packets per burst::
487 testpmd> set burst (num)
489 This is equivalent to the ``--burst command-line`` option.
491 When retry is enabled, the transmit delay time and number of retries can also be set::
493 testpmd> set burst tx delay (microseconds) retry (num)
498 Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode::
500 testpmd> set txpkts (x[,y]*)
502 Where x[,y]* represents a CSV list of values, without white space.
507 Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes::
509 testpmd> set txsplit (off|on|rand)
513 * ``off`` disable packet copy & split for CSUM mode.
515 * ``on`` split outgoing packet into multiple segments. Size of each segment
516 and number of segments per packet is determined by ``set txpkts`` command
519 * ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments.
524 Set the list of forwarding cores::
526 testpmd> set corelist (x[,y]*)
528 For example, to change the forwarding cores:
530 .. code-block:: console
532 testpmd> set corelist 3,1
533 testpmd> show config fwd
535 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled
536 Logical Core 3 (socket 0) forwards packets on 1 streams:
537 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01
538 Logical Core 1 (socket 0) forwards packets on 1 streams:
539 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
543 The cores are used in the same order as specified on the command line.
548 Set the list of forwarding ports::
550 testpmd> set portlist (x[,y]*)
552 For example, to change the port forwarding:
554 .. code-block:: console
556 testpmd> set portlist 0,2,1,3
557 testpmd> show config fwd
559 io packet forwarding - ports=4 - cores=1 - streams=4
560 Logical Core 3 (socket 0) forwards packets on 4 streams:
561 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01
562 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
563 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03
564 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02
569 Enable/disable tx loopback::
571 testpmd> set tx loopback (port_id) (on|off)
576 set drop enable bit for all queues::
578 testpmd> set all queues drop (port_id) (on|off)
580 set split drop enable (for VF)
581 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
583 set split drop enable bit for VF from PF::
585 testpmd> set vf split drop (port_id) (vf_id) (on|off)
587 set mac antispoof (for VF)
588 ~~~~~~~~~~~~~~~~~~~~~~~~~~
590 Set mac antispoof for a VF from the PF::
592 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off)
597 Enable/disable MACsec offload::
599 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off)
600 testpmd> set macsec offload (port_id) off
605 Configure MACsec secure connection (SC)::
607 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi)
611 The pi argument is ignored for tx.
612 Check the NIC Datasheet for hardware limits.
617 Configure MACsec secure association (SA)::
619 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key)
623 The IDX value must be 0 or 1.
624 Check the NIC Datasheet for hardware limits.
626 set broadcast mode (for VF)
627 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
629 Set broadcast mode for a VF from the PF::
631 testpmd> set vf broadcast (port_id) (vf_id) (on|off)
636 Set the VLAN strip on a port::
638 testpmd> vlan set strip (on|off) (port_id)
643 Set the VLAN strip for a queue on a port::
645 testpmd> vlan set stripq (on|off) (port_id,queue_id)
647 vlan set stripq (for VF)
648 ~~~~~~~~~~~~~~~~~~~~~~~~
650 Set VLAN strip for all queues in a pool for a VF from the PF::
652 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off)
654 vlan set insert (for VF)
655 ~~~~~~~~~~~~~~~~~~~~~~~~
657 Set VLAN insert for a VF from the PF::
659 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id)
661 vlan set tag (for VF)
662 ~~~~~~~~~~~~~~~~~~~~~
664 Set VLAN tag for a VF from the PF::
666 testpmd> set vf vlan tag (port_id) (vf_id) (on|off)
668 vlan set antispoof (for VF)
669 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
671 Set VLAN antispoof for a VF from the PF::
673 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off)
678 Set the VLAN filter on a port::
680 testpmd> vlan set filter (on|off) (port_id)
685 Set the VLAN QinQ (extended queue in queue) on for a port::
687 testpmd> vlan set qinq (on|off) (port_id)
692 Set the inner or outer VLAN TPID for packet filtering on a port::
694 testpmd> vlan set (inner|outer) tpid (value) (port_id)
698 TPID value must be a 16-bit number (value <= 65536).
703 Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID::
705 testpmd> rx_vlan add (vlan_id|all) (port_id)
709 VLAN filter must be set on that port. VLAN ID < 4096.
710 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries
711 in VFTA table. This is important if enabling all vlan_ids.
716 Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID::
718 testpmd> rx_vlan rm (vlan_id|all) (port_id)
723 Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID::
725 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask)
730 Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID::
732 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask)
737 Add a tunnel filter on a port::
739 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \
740 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
741 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
743 The available information categories are:
745 * ``vxlan``: Set tunnel type as VXLAN.
747 * ``nvgre``: Set tunnel type as NVGRE.
749 * ``ipingre``: Set tunnel type as IP-in-GRE.
751 * ``imac-ivlan``: Set filter type as Inner MAC and VLAN.
753 * ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID.
755 * ``imac-tenid``: Set filter type as Inner MAC and tenant ID.
757 * ``imac``: Set filter type as Inner MAC.
759 * ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID.
761 * ``oip``: Set filter type as Outer IP.
763 * ``iip``: Set filter type as Inner IP.
767 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \
768 192.168.2.2 0 ipingre oip 1 1
770 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP.
775 Remove a tunnel filter on a port::
777 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \
778 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
779 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
784 Add an UDP port for VXLAN packet filter on a port::
786 testpmd> rx_vxlan_port add (udp_port) (port_id)
791 Remove an UDP port for VXLAN packet filter on a port::
793 testpmd> rx_vxlan_port rm (udp_port) (port_id)
798 Set hardware insertion of VLAN IDs in packets sent on a port::
800 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer]
802 For example, set a single VLAN ID (5) insertion on port 0::
806 Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1::
814 Set port based hardware insertion of VLAN ID in packets sent on a port::
816 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off)
821 Disable hardware insertion of a VLAN header in packets sent on a port::
823 testpmd> tx_vlan reset (port_id)
828 Select hardware or software calculation of the checksum when
829 transmitting a packet using the ``csum`` forwarding engine::
831 testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id)
835 * ``ip|udp|tcp|sctp`` always relate to the inner layer.
837 * ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized
838 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are
839 supported). See also the ``csum parse-tunnel`` command.
843 Check the NIC Datasheet for hardware limits.
848 Define how tunneled packets should be handled by the csum forward
851 testpmd> csum parse-tunnel (on|off) (tx_port_id)
853 If enabled, the csum forward engine will try to recognize supported
854 tunnel headers (vxlan, gre, ipip).
856 If disabled, treat tunnel packets as non-tunneled packets (a inner
857 header is handled as a packet payload).
861 The port argument is the TX port like in the ``csum set`` command.
865 Consider a packet in packet like the following::
867 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in
869 * If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
870 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the
871 ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``).
873 * If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
874 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``.
879 Display tx checksum offload configuration::
881 testpmd> csum show (port_id)
886 Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine::
888 testpmd> tso set (segsize) (port_id)
892 Check the NIC datasheet for hardware limits.
897 Display the status of TCP Segmentation Offload::
899 testpmd> tso show (port_id)
904 Add an alternative MAC address to a port::
906 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX)
911 Remove a MAC address from a port::
913 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX)
915 mac_addr add (for VF)
916 ~~~~~~~~~~~~~~~~~~~~~
918 Add an alternative MAC address for a VF to a port::
920 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX)
925 Set the default MAC address for a port::
927 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX)
929 mac_addr set (for VF)
930 ~~~~~~~~~~~~~~~~~~~~~
932 Set the MAC address for a VF from the PF::
934 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX)
939 Set the unicast hash filter(s) on/off for a port::
941 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
946 Set the promiscuous mode on for a port or for all ports.
947 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
949 testpmd> set promisc (port_id|all) (on|off)
954 Set the allmulti mode for a port or for all ports::
956 testpmd> set allmulti (port_id|all) (on|off)
958 Same as the ifconfig (8) option. Controls how multicast packets are handled.
963 Set the unicast promiscuous mode for a VF from PF.
964 It's supported by Intel i40e NICs now.
965 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
967 testpmd> set vf promisc (port_id) (vf_id) (on|off)
969 set allmulticast (for VF)
970 ~~~~~~~~~~~~~~~~~~~~~~~~~
972 Set the multicast promiscuous mode for a VF from PF.
973 It's supported by Intel i40e NICs now.
974 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
976 testpmd> set vf allmulti (port_id) (vf_id) (on|off)
978 set tx max bandwidth (for VF)
979 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
981 Set TX max absolute bandwidth (Mbps) for a VF from PF::
983 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth)
985 set tc tx min bandwidth (for VF)
986 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
988 Set all TCs' TX min relative bandwidth (%) for a VF from PF::
990 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...)
992 set tc tx max bandwidth (for VF)
993 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
995 Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF::
997 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth)
999 set tc strict link priority mode
1000 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1002 Set some TCs' strict link priority mode on a physical port::
1004 testpmd> set tx strict-link-priority (port_id) (tc_bitmap)
1006 set tc tx min bandwidth
1007 ~~~~~~~~~~~~~~~~~~~~~~~
1009 Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs::
1011 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...)
1016 Set the link flow control parameter on a port::
1018 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1019 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
1020 autoneg (on|off) (port_id)
1024 * ``high_water`` (integer): High threshold value to trigger XOFF.
1026 * ``low_water`` (integer): Low threshold value to trigger XON.
1028 * ``pause_time`` (integer): Pause quota in the Pause frame.
1030 * ``send_xon`` (0/1): Send XON frame.
1032 * ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
1034 * ``autoneg``: Change the auto-negotiation parameter.
1039 Set the priority flow control parameter on a port::
1041 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1042 (pause_time) (priority) (port_id)
1046 * ``high_water`` (integer): High threshold value.
1048 * ``low_water`` (integer): Low threshold value.
1050 * ``pause_time`` (integer): Pause quota in the Pause frame.
1052 * ``priority`` (0-7): VLAN User Priority.
1057 Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
1059 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
1061 For example, to set rx queue 2 on port 0 to mapping 5::
1063 testpmd>set stat_qmap rx 0 2 5
1065 set port - rx/tx (for VF)
1066 ~~~~~~~~~~~~~~~~~~~~~~~~~
1068 Set VF receive/transmit from a port::
1070 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
1072 set port - mac address filter (for VF)
1073 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1075 Add/Remove unicast or multicast MAC addr filter for a VF::
1077 testpmd> set port (port_id) vf (vf_id) (mac_addr) \
1078 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
1080 set port - rx mode(for VF)
1081 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1083 Set the VF receive mode of a port::
1085 testpmd> set port (port_id) vf (vf_id) \
1086 rxmode (AUPE|ROPE|BAM|MPE) (on|off)
1088 The available receive modes are:
1090 * ``AUPE``: Accepts untagged VLAN.
1092 * ``ROPE``: Accepts unicast hash.
1094 * ``BAM``: Accepts broadcast packets.
1096 * ``MPE``: Accepts all multicast packets.
1098 set port - tx_rate (for Queue)
1099 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1101 Set TX rate limitation for a queue on a port::
1103 testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1105 set port - tx_rate (for VF)
1106 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
1108 Set TX rate limitation for queues in VF on a port::
1110 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1112 set port - mirror rule
1113 ~~~~~~~~~~~~~~~~~~~~~~
1115 Set pool or vlan type mirror rule for a port::
1117 testpmd> set port (port_id) mirror-rule (rule_id) \
1118 (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1119 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1121 Set link mirror rule for a port::
1123 testpmd> set port (port_id) mirror-rule (rule_id) \
1124 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1126 For example to enable mirror traffic with vlan 0,1 to pool 0::
1128 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1130 reset port - mirror rule
1131 ~~~~~~~~~~~~~~~~~~~~~~~~
1133 Reset a mirror rule for a port::
1135 testpmd> reset port (port_id) mirror-rule (rule_id)
1140 Set the flush on RX streams before forwarding.
1141 The default is flush ``on``.
1142 Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1144 testpmd> set flush_rx off
1149 Set the bypass mode for the lowest port on bypass enabled NIC::
1151 testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1156 Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1158 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1159 mode (normal|bypass|isolate) (port_id)
1163 * ``timeout``: Enable bypass after watchdog timeout.
1165 * ``os_on``: Enable bypass when OS/board is powered on.
1167 * ``os_off``: Enable bypass when OS/board is powered off.
1169 * ``power_on``: Enable bypass when power supply is turned on.
1171 * ``power_off``: Enable bypass when power supply is turned off.
1177 Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1179 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1184 Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1186 testpmd> show bypass config (port_id)
1191 Set link up for a port::
1193 testpmd> set link-up port (port id)
1198 Set link down for a port::
1200 testpmd> set link-down port (port id)
1205 Enable E-tag insertion for a VF on a port::
1207 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1209 Disable E-tag insertion for a VF on a port::
1211 testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1213 Enable/disable E-tag stripping on a port::
1215 testpmd> E-tag set stripping (on|off) port (port_id)
1217 Enable/disable E-tag based forwarding on a port::
1219 testpmd> E-tag set forwarding (on|off) port (port_id)
1221 Add an E-tag forwarding filter on a port::
1223 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1225 Delete an E-tag forwarding filter on a port::
1226 testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1231 Load a dynamic device personalization (DDP) package::
1233 testpmd> ddp add (port_id) (package_path[,output_path])
1238 Delete a dynamic device personalization package::
1240 testpmd> ddp del (port_id) (package_path)
1245 List all items from the ptype mapping table::
1247 testpmd> ptype mapping get (port_id) (valid_only)
1251 * ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0).
1253 Replace a specific or a group of software defined ptype with a new one::
1255 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type)
1259 * ``target``: A specific software ptype or a mask to represent a group of software ptypes.
1261 * ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1).
1263 * ``pkt_type``: The new software ptype to replace the old ones.
1265 Update hardware defined ptype to software defined packet type mapping table::
1267 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype)
1271 * ``hw_ptype``: hardware ptype as the index of the ptype mapping table.
1273 * ``sw_ptype``: software ptype as the value of the ptype mapping table.
1275 Reset ptype mapping table::
1277 testpmd> ptype mapping reset (port_id)
1282 The following sections show functions for configuring ports.
1286 Port configuration changes only become active when forwarding is started/restarted.
1291 Attach a port specified by pci address or virtual device args::
1293 testpmd> port attach (identifier)
1295 To attach a new pci device, the device should be recognized by kernel first.
1296 Then it should be moved under DPDK management.
1297 Finally the port can be attached to testpmd.
1299 For example, to move a pci device using ixgbe under DPDK management:
1301 .. code-block:: console
1303 # Check the status of the available devices.
1304 ./usertools/dpdk-devbind.py --status
1306 Network devices using DPDK-compatible driver
1307 ============================================
1310 Network devices using kernel driver
1311 ===================================
1312 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1315 # Bind the device to igb_uio.
1316 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1319 # Recheck the status of the devices.
1320 ./usertools/dpdk-devbind.py --status
1321 Network devices using DPDK-compatible driver
1322 ============================================
1323 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1325 To attach a port created by virtual device, above steps are not needed.
1327 For example, to attach a port whose pci address is 0000:0a:00.0.
1329 .. code-block:: console
1331 testpmd> port attach 0000:0a:00.0
1332 Attaching a new port...
1333 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1334 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1335 EAL: PCI memory mapped at 0x7f83bfa00000
1336 EAL: PCI memory mapped at 0x7f83bfa80000
1337 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1338 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1339 Port 0 is attached. Now total ports is 1
1342 For example, to attach a port created by pcap PMD.
1344 .. code-block:: console
1346 testpmd> port attach net_pcap0
1347 Attaching a new port...
1348 PMD: Initializing pmd_pcap for net_pcap0
1349 PMD: Creating pcap-backed ethdev on numa socket 0
1350 Port 0 is attached. Now total ports is 1
1353 In this case, identifier is ``net_pcap0``.
1354 This identifier format is the same as ``--vdev`` format of DPDK applications.
1356 For example, to re-attach a bonded port which has been previously detached,
1357 the mode and slave parameters must be given.
1359 .. code-block:: console
1361 testpmd> port attach net_bond_0,mode=0,slave=1
1362 Attaching a new port...
1363 EAL: Initializing pmd_bond for net_bond_0
1364 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1365 Port 0 is attached. Now total ports is 1
1372 Detach a specific port::
1374 testpmd> port detach (port_id)
1376 Before detaching a port, the port should be stopped and closed.
1378 For example, to detach a pci device port 0.
1380 .. code-block:: console
1382 testpmd> port stop 0
1385 testpmd> port close 0
1389 testpmd> port detach 0
1391 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1392 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1393 EAL: PCI memory unmapped at 0x7f83bfa00000
1394 EAL: PCI memory unmapped at 0x7f83bfa80000
1398 For example, to detach a virtual device port 0.
1400 .. code-block:: console
1402 testpmd> port stop 0
1405 testpmd> port close 0
1409 testpmd> port detach 0
1411 PMD: Closing pcap ethdev on numa socket 0
1412 Port 'net_pcap0' is detached. Now total ports is 0
1415 To remove a pci device completely from the system, first detach the port from testpmd.
1416 Then the device should be moved under kernel management.
1417 Finally the device can be removed using kernel pci hotplug functionality.
1419 For example, to move a pci device under kernel management:
1421 .. code-block:: console
1423 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1425 ./usertools/dpdk-devbind.py --status
1427 Network devices using DPDK-compatible driver
1428 ============================================
1431 Network devices using kernel driver
1432 ===================================
1433 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1435 To remove a port created by a virtual device, above steps are not needed.
1440 Start all ports or a specific port::
1442 testpmd> port start (port_id|all)
1447 Stop all ports or a specific port::
1449 testpmd> port stop (port_id|all)
1454 Close all ports or a specific port::
1456 testpmd> port close (port_id|all)
1458 port start/stop queue
1459 ~~~~~~~~~~~~~~~~~~~~~
1461 Start/stop a rx/tx queue on a specific port::
1463 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1465 Only take effect when port is started.
1470 Set the speed and duplex mode for all ports or a specific port::
1472 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1473 duplex (half|full|auto)
1475 port config - queues/descriptors
1476 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1478 Set number of queues/descriptors for rxq, txq, rxd and txd::
1480 testpmd> port config all (rxq|txq|rxd|txd) (value)
1482 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1484 port config - max-pkt-len
1485 ~~~~~~~~~~~~~~~~~~~~~~~~~
1487 Set the maximum packet length::
1489 testpmd> port config all max-pkt-len (value)
1491 This is equivalent to the ``--max-pkt-len`` command-line option.
1493 port config - CRC Strip
1494 ~~~~~~~~~~~~~~~~~~~~~~~
1496 Set hardware CRC stripping on or off for all ports::
1498 testpmd> port config all crc-strip (on|off)
1500 CRC stripping is on by default.
1502 The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option.
1504 port config - scatter
1505 ~~~~~~~~~~~~~~~~~~~~~~~
1507 Set RX scatter mode on or off for all ports::
1509 testpmd> port config all scatter (on|off)
1511 RX scatter mode is off by default.
1513 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1515 port config - TX queue flags
1516 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1518 Set a hexadecimal bitmap of TX queue flags for all ports::
1520 testpmd> port config all txqflags value
1522 This command is equivalent to the ``--txqflags`` command-line option.
1524 port config - RX Checksum
1525 ~~~~~~~~~~~~~~~~~~~~~~~~~
1527 Set hardware RX checksum offload to on or off for all ports::
1529 testpmd> port config all rx-cksum (on|off)
1531 Checksum offload is off by default.
1533 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1538 Set hardware VLAN on or off for all ports::
1540 testpmd> port config all hw-vlan (on|off)
1542 Hardware VLAN is on by default.
1544 The ``off`` option is equivalent to the ``--disable-hw-vlan`` command-line option.
1546 port config - VLAN filter
1547 ~~~~~~~~~~~~~~~~~~~~~~~~~
1549 Set hardware VLAN filter on or off for all ports::
1551 testpmd> port config all hw-vlan-filter (on|off)
1553 Hardware VLAN filter is on by default.
1555 The ``off`` option is equivalent to the ``--disable-hw-vlan-filter`` command-line option.
1557 port config - VLAN strip
1558 ~~~~~~~~~~~~~~~~~~~~~~~~
1560 Set hardware VLAN strip on or off for all ports::
1562 testpmd> port config all hw-vlan-strip (on|off)
1564 Hardware VLAN strip is on by default.
1566 The ``off`` option is equivalent to the ``--disable-hw-vlan-strip`` command-line option.
1568 port config - VLAN extend
1569 ~~~~~~~~~~~~~~~~~~~~~~~~~
1571 Set hardware VLAN extend on or off for all ports::
1573 testpmd> port config all hw-vlan-extend (on|off)
1575 Hardware VLAN extend is off by default.
1577 The ``off`` option is equivalent to the ``--disable-hw-vlan-extend`` command-line option.
1579 port config - Drop Packets
1580 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1582 Set packet drop for packets with no descriptors on or off for all ports::
1584 testpmd> port config all drop-en (on|off)
1586 Packet dropping for packets with no descriptors is off by default.
1588 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1593 Set the RSS (Receive Side Scaling) mode on or off::
1595 testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1597 RSS is on by default.
1599 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1601 port config - RSS Reta
1602 ~~~~~~~~~~~~~~~~~~~~~~
1604 Set the RSS (Receive Side Scaling) redirection table::
1606 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1611 Set the DCB mode for an individual port::
1613 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1615 The traffic class should be 4 or 8.
1620 Set the number of packets per burst::
1622 testpmd> port config all burst (value)
1624 This is equivalent to the ``--burst`` command-line option.
1626 port config - Threshold
1627 ~~~~~~~~~~~~~~~~~~~~~~~
1629 Set thresholds for TX/RX queues::
1631 testpmd> port config all (threshold) (value)
1633 Where the threshold type can be:
1635 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1637 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1639 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1641 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1643 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1645 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1647 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1649 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1651 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1653 These threshold options are also available from the command-line.
1658 Set the value of ether-type for E-tag::
1660 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1662 Enable/disable the E-tag support::
1664 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1667 Link Bonding Functions
1668 ----------------------
1670 The Link Bonding functions make it possible to dynamically create and
1671 manage link bonding devices from within testpmd interactive prompt.
1673 create bonded device
1674 ~~~~~~~~~~~~~~~~~~~~
1676 Create a new bonding device::
1678 testpmd> create bonded device (mode) (socket)
1680 For example, to create a bonded device in mode 1 on socket 0::
1682 testpmd> create bonded 1 0
1683 created new bonded device (port X)
1688 Adds Ethernet device to a Link Bonding device::
1690 testpmd> add bonding slave (slave id) (port id)
1692 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
1694 testpmd> add bonding slave 6 10
1697 remove bonding slave
1698 ~~~~~~~~~~~~~~~~~~~~
1700 Removes an Ethernet slave device from a Link Bonding device::
1702 testpmd> remove bonding slave (slave id) (port id)
1704 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
1706 testpmd> remove bonding slave 6 10
1711 Set the Link Bonding mode of a Link Bonding device::
1713 testpmd> set bonding mode (value) (port id)
1715 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
1717 testpmd> set bonding mode 3 10
1722 Set an Ethernet slave device as the primary device on a Link Bonding device::
1724 testpmd> set bonding primary (slave id) (port id)
1726 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
1728 testpmd> set bonding primary 6 10
1733 Set the MAC address of a Link Bonding device::
1735 testpmd> set bonding mac (port id) (mac)
1737 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
1739 testpmd> set bonding mac 10 00:00:00:00:00:01
1741 set bonding xmit_balance_policy
1742 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1744 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
1746 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
1748 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
1750 testpmd> set bonding xmit_balance_policy 10 l34
1753 set bonding mon_period
1754 ~~~~~~~~~~~~~~~~~~~~~~
1756 Set the link status monitoring polling period in milliseconds for a bonding device.
1758 This adds support for PMD slave devices which do not support link status interrupts.
1759 When the mon_period is set to a value greater than 0 then all PMD's which do not support
1760 link status ISR will be queried every polling interval to check if their link status has changed::
1762 testpmd> set bonding mon_period (port_id) (value)
1764 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
1766 testpmd> set bonding mon_period 5 150
1769 set bonding lacp dedicated_queue
1770 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1772 Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic
1773 when in mode 4 (link-aggregration-802.3ad)
1775 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable)
1781 Show the current configuration of a Link Bonding device::
1783 testpmd> show bonding config (port id)
1786 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
1787 in balance mode with a transmission policy of layer 2+3::
1789 testpmd> show bonding config 9
1791 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
1793 Active Slaves (3): [1 3 4]
1800 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
1801 This is mainly useful for debugging purposes.
1802 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
1803 and fields that can be accessed.
1808 Display the value of a port register::
1810 testpmd> read reg (port_id) (address)
1812 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
1814 testpmd> read reg 0 0xEE00
1815 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
1820 Display a port register bit field::
1822 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
1824 For example, reading the lowest two bits from the register in the example above::
1826 testpmd> read regfield 0 0xEE00 0 1
1827 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
1832 Display a single port register bit::
1834 testpmd> read regbit (port_id) (address) (bit_x)
1836 For example, reading the lowest bit from the register in the example above::
1838 testpmd> read regbit 0 0xEE00 0
1839 port 0 PCI register at offset 0xEE00: bit 0=1
1844 Set the value of a port register::
1846 testpmd> write reg (port_id) (address) (value)
1848 For example, to clear a register::
1850 testpmd> write reg 0 0xEE00 0x0
1851 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
1856 Set bit field of a port register::
1858 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
1860 For example, writing to the register cleared in the example above::
1862 testpmd> write regfield 0 0xEE00 0 1 2
1863 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
1868 Set single bit value of a port register::
1870 testpmd> write regbit (port_id) (address) (bit_x) (value)
1872 For example, to set the high bit in the register from the example above::
1874 testpmd> write regbit 0 0xEE00 31 1
1875 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
1881 This section details the available filter functions that are available.
1883 Note these functions interface the deprecated legacy filtering framework,
1884 superseded by *rte_flow*. See `Flow rules management`_.
1887 ~~~~~~~~~~~~~~~~~~~~
1889 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
1891 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
1892 ethertype (ether_type) (drop|fwd) queue (queue_id)
1894 The available information parameters are:
1896 * ``port_id``: The port which the Ethertype filter assigned on.
1898 * ``mac_addr``: Compare destination mac address.
1900 * ``mac_ignr``: Ignore destination mac address match.
1902 * ``mac_address``: Destination mac address to match.
1904 * ``ether_type``: The EtherType value want to match,
1905 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
1907 * ``queue_id``: The receive queue associated with this EtherType filter.
1908 It is meaningless when deleting or dropping.
1910 Example, to add/remove an ethertype filter rule::
1912 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
1913 ethertype 0x0806 fwd queue 3
1915 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
1916 ethertype 0x0806 fwd queue 3
1921 Add or delete a 2-tuple filter,
1922 which identifies packets by specific protocol and destination TCP/UDP port
1923 and forwards packets into one of the receive queues::
1925 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
1926 protocol (protocol_value) mask (mask_value) \
1927 tcp_flags (tcp_flags_value) priority (prio_value) \
1930 The available information parameters are:
1932 * ``port_id``: The port which the 2-tuple filter assigned on.
1934 * ``dst_port_value``: Destination port in L4.
1936 * ``protocol_value``: IP L4 protocol.
1938 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
1940 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
1942 * ``prio_value``: Priority of this filter.
1944 * ``queue_id``: The receive queue associated with this 2-tuple filter.
1946 Example, to add/remove an 2tuple filter rule::
1948 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
1949 tcp_flags 0x02 priority 3 queue 3
1951 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
1952 tcp_flags 0x02 priority 3 queue 3
1957 Add or delete a 5-tuple filter,
1958 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
1959 and routes packets into one of the receive queues::
1961 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
1962 (src_address) dst_port (dst_port_value) \
1963 src_port (src_port_value) protocol (protocol_value) \
1964 mask (mask_value) tcp_flags (tcp_flags_value) \
1965 priority (prio_value) queue (queue_id)
1967 The available information parameters are:
1969 * ``port_id``: The port which the 5-tuple filter assigned on.
1971 * ``dst_address``: Destination IP address.
1973 * ``src_address``: Source IP address.
1975 * ``dst_port_value``: TCP/UDP destination port.
1977 * ``src_port_value``: TCP/UDP source port.
1979 * ``protocol_value``: L4 protocol.
1981 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
1983 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
1985 * ``prio_value``: The priority of this filter.
1987 * ``queue_id``: The receive queue associated with this 5-tuple filter.
1989 Example, to add/remove an 5tuple filter rule::
1991 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1992 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1993 flags 0x0 priority 3 queue 3
1995 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1996 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1997 flags 0x0 priority 3 queue 3
2002 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
2004 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
2006 The available information parameters are:
2008 * ``port_id``: The port which the SYN filter assigned on.
2010 * ``high``: This SYN filter has higher priority than other filters.
2012 * ``low``: This SYN filter has lower priority than other filters.
2014 * ``queue_id``: The receive queue associated with this SYN filter
2018 testpmd> syn_filter 0 add priority high queue 3
2023 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
2024 and routed into one of the receive queues::
2026 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
2027 mask (mask_value) priority (prio_value) queue (queue_id)
2029 The available information parameters are:
2031 * ``port_id``: The port which the Flex filter is assigned on.
2033 * ``len_value``: Filter length in bytes, no greater than 128.
2035 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
2037 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
2039 * ``prio_value``: The priority of this filter.
2041 * ``queue_id``: The receive queue associated with this Flex filter.
2045 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
2046 mask 000C priority 3 queue 3
2048 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
2049 mask 000C priority 3 queue 3
2052 .. _testpmd_flow_director:
2054 flow_director_filter
2055 ~~~~~~~~~~~~~~~~~~~~
2057 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
2059 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
2060 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
2062 * Perfect match filters.
2063 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2064 The masked fields are for IP flow.
2066 * Signature filters.
2067 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
2069 * Perfect-mac-vlan match filters.
2070 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2071 The masked fields are for MAC VLAN flow.
2073 * Perfect-tunnel match filters.
2074 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2075 The masked fields are for tunnel flow.
2077 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
2078 per flow type and the flexible payload.
2080 The Flow Director can also mask out parts of all of these fields so that filters
2081 are only applied to certain fields or parts of the fields.
2083 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
2085 # Commands to add flow director filters of different flow types::
2087 flow_director_filter (port_id) mode IP (add|del|update) \
2088 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
2089 src (src_ip_address) dst (dst_ip_address) \
2090 tos (tos_value) proto (proto_value) ttl (ttl_value) \
2091 vlan (vlan_value) flexbytes (flexbytes_value) \
2092 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
2095 flow_director_filter (port_id) mode IP (add|del|update) \
2096 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
2097 src (src_ip_address) (src_port) \
2098 dst (dst_ip_address) (dst_port) \
2099 tos (tos_value) ttl (ttl_value) \
2100 vlan (vlan_value) flexbytes (flexbytes_value) \
2101 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
2104 flow_director_filter (port_id) mode IP (add|del|update) \
2105 flow (ipv4-sctp|ipv6-sctp) \
2106 src (src_ip_address) (src_port) \
2107 dst (dst_ip_address) (dst_port) \
2108 tos (tos_value) ttl (ttl_value) \
2109 tag (verification_tag) vlan (vlan_value) \
2110 flexbytes (flexbytes_value) (drop|fwd) \
2111 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
2113 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
2114 ether (ethertype) flexbytes (flexbytes_value) \
2115 (drop|fwd) pf|vf(vf_id) queue (queue_id)
2118 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
2119 mac (mac_address) vlan (vlan_value) \
2120 flexbytes (flexbytes_value) (drop|fwd) \
2121 queue (queue_id) fd_id (fd_id_value)
2123 flow_director_filter (port_id) mode Tunnel (add|del|update) \
2124 mac (mac_address) vlan (vlan_value) \
2125 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
2126 flexbytes (flexbytes_value) (drop|fwd) \
2127 queue (queue_id) fd_id (fd_id_value)
2129 For example, to add an ipv4-udp flow type filter::
2131 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
2132 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
2133 fwd pf queue 1 fd_id 1
2135 For example, add an ipv4-other flow type filter::
2137 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
2138 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
2139 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
2144 Flush all flow director filters on a device::
2146 testpmd> flush_flow_director (port_id)
2148 Example, to flush all flow director filter on port 0::
2150 testpmd> flush_flow_director 0
2155 Set flow director's input masks::
2157 flow_director_mask (port_id) mode IP vlan (vlan_value) \
2158 src_mask (ipv4_src) (ipv6_src) (src_port) \
2159 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2161 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2163 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2164 mac (mac_value) tunnel-type (tunnel_type_value) \
2165 tunnel-id (tunnel_id_value)
2167 Example, to set flow director mask on port 0::
2169 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2170 src_mask 255.255.255.255 \
2171 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2172 dst_mask 255.255.255.255 \
2173 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2175 flow_director_flex_mask
2176 ~~~~~~~~~~~~~~~~~~~~~~~
2178 set masks of flow director's flexible payload based on certain flow type::
2180 testpmd> flow_director_flex_mask (port_id) \
2181 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2182 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2183 l2_payload|all) (mask)
2185 Example, to set flow director's flex mask for all flow type on port 0::
2187 testpmd> flow_director_flex_mask 0 flow all \
2188 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2191 flow_director_flex_payload
2192 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2194 Configure flexible payload selection::
2196 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2198 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2200 testpmd> flow_director_flex_payload 0 l4 \
2201 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2203 get_sym_hash_ena_per_port
2204 ~~~~~~~~~~~~~~~~~~~~~~~~~
2206 Get symmetric hash enable configuration per port::
2208 get_sym_hash_ena_per_port (port_id)
2210 For example, to get symmetric hash enable configuration of port 1::
2212 testpmd> get_sym_hash_ena_per_port 1
2214 set_sym_hash_ena_per_port
2215 ~~~~~~~~~~~~~~~~~~~~~~~~~
2217 Set symmetric hash enable configuration per port to enable or disable::
2219 set_sym_hash_ena_per_port (port_id) (enable|disable)
2221 For example, to set symmetric hash enable configuration of port 1 to enable::
2223 testpmd> set_sym_hash_ena_per_port 1 enable
2225 get_hash_global_config
2226 ~~~~~~~~~~~~~~~~~~~~~~
2228 Get the global configurations of hash filters::
2230 get_hash_global_config (port_id)
2232 For example, to get the global configurations of hash filters of port 1::
2234 testpmd> get_hash_global_config 1
2236 set_hash_global_config
2237 ~~~~~~~~~~~~~~~~~~~~~~
2239 Set the global configurations of hash filters::
2241 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2242 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2243 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload) \
2246 For example, to enable simple_xor for flow type of ipv6 on port 2::
2248 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2253 Set the input set for hash::
2255 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2256 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2257 l2_payload) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2258 ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2259 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2260 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2261 fld-8th|none) (select|add)
2263 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2265 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2270 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2271 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2273 Set the input set for flow director::
2275 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2276 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2277 l2_payload) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2278 ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2279 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2280 sctp-dst-port|sctp-veri-tag|none) (select|add)
2282 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2284 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2289 Set different GRE key length for input set::
2291 global_config (port_id) gre-key-len (number in bytes)
2293 For example to set GRE key length for input set to 4 bytes on port 0::
2295 testpmd> global_config 0 gre-key-len 4
2298 .. _testpmd_rte_flow:
2300 Flow rules management
2301 ---------------------
2303 Control of the generic flow API (*rte_flow*) is fully exposed through the
2304 ``flow`` command (validation, creation, destruction, queries and operation
2307 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2308 features simultaneously may cause undefined side-effects and is therefore
2314 Because the ``flow`` command uses dynamic tokens to handle the large number
2315 of possible flow rules combinations, its behavior differs slightly from
2316 other commands, in particular:
2318 - Pressing *?* or the *<tab>* key displays contextual help for the current
2319 token, not that of the entire command.
2321 - Optional and repeated parameters are supported (provided they are listed
2322 in the contextual help).
2324 The first parameter stands for the operation mode. Possible operations and
2325 their general syntax are described below. They are covered in detail in the
2328 - Check whether a flow rule can be created::
2330 flow validate {port_id}
2331 [group {group_id}] [priority {level}] [ingress] [egress]
2332 pattern {item} [/ {item} [...]] / end
2333 actions {action} [/ {action} [...]] / end
2335 - Create a flow rule::
2337 flow create {port_id}
2338 [group {group_id}] [priority {level}] [ingress] [egress]
2339 pattern {item} [/ {item} [...]] / end
2340 actions {action} [/ {action} [...]] / end
2342 - Destroy specific flow rules::
2344 flow destroy {port_id} rule {rule_id} [...]
2346 - Destroy all flow rules::
2348 flow flush {port_id}
2350 - Query an existing flow rule::
2352 flow query {port_id} {rule_id} {action}
2354 - List existing flow rules sorted by priority, filtered by group
2357 flow list {port_id} [group {group_id}] [...]
2359 - Restrict ingress traffic to the defined flow rules::
2361 flow isolate {port_id} {boolean}
2363 Validating flow rules
2364 ~~~~~~~~~~~~~~~~~~~~~
2366 ``flow validate`` reports whether a flow rule would be accepted by the
2367 underlying device in its current state but stops short of creating it. It is
2368 bound to ``rte_flow_validate()``::
2370 flow validate {port_id}
2371 [group {group_id}] [priority {level}] [ingress] [egress]
2372 pattern {item} [/ {item} [...]] / end
2373 actions {action} [/ {action} [...]] / end
2375 If successful, it will show::
2379 Otherwise it will show an error message of the form::
2381 Caught error type [...] ([...]): [...]
2383 This command uses the same parameters as ``flow create``, their format is
2384 described in `Creating flow rules`_.
2386 Check whether redirecting any Ethernet packet received on port 0 to RX queue
2387 index 6 is supported::
2389 testpmd> flow validate 0 ingress pattern eth / end
2390 actions queue index 6 / end
2394 Port 0 does not support TCPv6 rules::
2396 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
2398 Caught error type 9 (specific pattern item): Invalid argument
2404 ``flow create`` validates and creates the specified flow rule. It is bound
2405 to ``rte_flow_create()``::
2407 flow create {port_id}
2408 [group {group_id}] [priority {level}] [ingress] [egress]
2409 pattern {item} [/ {item} [...]] / end
2410 actions {action} [/ {action} [...]] / end
2412 If successful, it will return a flow rule ID usable with other commands::
2414 Flow rule #[...] created
2416 Otherwise it will show an error message of the form::
2418 Caught error type [...] ([...]): [...]
2420 Parameters describe in the following order:
2422 - Attributes (*group*, *priority*, *ingress*, *egress* tokens).
2423 - A matching pattern, starting with the *pattern* token and terminated by an
2425 - Actions, starting with the *actions* token and terminated by an *end*
2428 These translate directly to *rte_flow* objects provided as-is to the
2429 underlying functions.
2431 The shortest valid definition only comprises mandatory tokens::
2433 testpmd> flow create 0 pattern end actions end
2435 Note that PMDs may refuse rules that essentially do nothing such as this
2438 **All unspecified object values are automatically initialized to 0.**
2443 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
2444 specified before the ``pattern`` token.
2446 - ``group {group id}``: priority group.
2447 - ``priority {level}``: priority level within group.
2448 - ``ingress``: rule applies to ingress traffic.
2449 - ``egress``: rule applies to egress traffic.
2451 Each instance of an attribute specified several times overrides the previous
2452 value as shown below (group 4 is used)::
2454 testpmd> flow create 0 group 42 group 24 group 4 [...]
2456 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
2458 While not specifying a direction is an error, some rules may allow both
2461 Most rules affect RX therefore contain the ``ingress`` token::
2463 testpmd> flow create 0 ingress pattern [...]
2468 A matching pattern starts after the ``pattern`` token. It is made of pattern
2469 items and is terminated by a mandatory ``end`` item.
2471 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
2472 rte_flow_item_type``).
2474 The ``/`` token is used as a separator between pattern items as shown
2477 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
2479 Note that protocol items like these must be stacked from lowest to highest
2480 layer to make sense. For instance, the following rule is either invalid or
2481 unlikely to match any packet::
2483 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
2485 More information on these restrictions can be found in the *rte_flow*
2488 Several items support additional specification structures, for example
2489 ``ipv4`` allows specifying source and destination addresses as follows::
2491 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
2492 dst is 10.2.0.0 / end [...]
2494 This rule matches all IPv4 traffic with the specified properties.
2496 In this example, ``src`` and ``dst`` are field names of the underlying
2497 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
2498 in a similar fashion.
2500 The ``is`` token means that the subsequent value must be matched exactly,
2501 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
2502 accordingly. Possible assignment tokens are:
2504 - ``is``: match value perfectly (with full bit-mask).
2505 - ``spec``: match value according to configured bit-mask.
2506 - ``last``: specify upper bound to establish a range.
2507 - ``mask``: specify bit-mask with relevant bits set to one.
2508 - ``prefix``: generate bit-mask from a prefix length.
2510 These yield identical results::
2512 ipv4 src is 10.1.1.1
2516 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
2520 ipv4 src spec 10.1.1.1 src prefix 32
2524 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
2528 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
2530 Inclusive ranges can be defined with ``last``::
2532 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
2534 Note that ``mask`` affects both ``spec`` and ``last``::
2536 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
2537 # matches 10.1.0.0 to 10.2.255.255
2539 Properties can be modified multiple times::
2541 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
2545 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
2550 This section lists supported pattern items and their attributes, if any.
2552 - ``end``: end list of pattern items.
2554 - ``void``: no-op pattern item.
2556 - ``invert``: perform actions when pattern does not match.
2558 - ``any``: match any protocol for the current layer.
2560 - ``num {unsigned}``: number of layers covered.
2562 - ``pf``: match packets addressed to the physical function.
2564 - ``vf``: match packets addressed to a virtual function ID.
2566 - ``id {unsigned}``: destination VF ID.
2568 - ``port``: device-specific physical port index to use.
2570 - ``index {unsigned}``: physical port index.
2572 - ``raw``: match an arbitrary byte string.
2574 - ``relative {boolean}``: look for pattern after the previous item.
2575 - ``search {boolean}``: search pattern from offset (see also limit).
2576 - ``offset {integer}``: absolute or relative offset for pattern.
2577 - ``limit {unsigned}``: search area limit for start of pattern.
2578 - ``pattern {string}``: byte string to look for.
2580 - ``eth``: match Ethernet header.
2582 - ``dst {MAC-48}``: destination MAC.
2583 - ``src {MAC-48}``: source MAC.
2584 - ``type {unsigned}``: EtherType.
2586 - ``vlan``: match 802.1Q/ad VLAN tag.
2588 - ``tpid {unsigned}``: tag protocol identifier.
2589 - ``tci {unsigned}``: tag control information.
2590 - ``pcp {unsigned}``: priority code point.
2591 - ``dei {unsigned}``: drop eligible indicator.
2592 - ``vid {unsigned}``: VLAN identifier.
2594 - ``ipv4``: match IPv4 header.
2596 - ``tos {unsigned}``: type of service.
2597 - ``ttl {unsigned}``: time to live.
2598 - ``proto {unsigned}``: next protocol ID.
2599 - ``src {ipv4 address}``: source address.
2600 - ``dst {ipv4 address}``: destination address.
2602 - ``ipv6``: match IPv6 header.
2604 - ``tc {unsigned}``: traffic class.
2605 - ``flow {unsigned}``: flow label.
2606 - ``proto {unsigned}``: protocol (next header).
2607 - ``hop {unsigned}``: hop limit.
2608 - ``src {ipv6 address}``: source address.
2609 - ``dst {ipv6 address}``: destination address.
2611 - ``icmp``: match ICMP header.
2613 - ``type {unsigned}``: ICMP packet type.
2614 - ``code {unsigned}``: ICMP packet code.
2616 - ``udp``: match UDP header.
2618 - ``src {unsigned}``: UDP source port.
2619 - ``dst {unsigned}``: UDP destination port.
2621 - ``tcp``: match TCP header.
2623 - ``src {unsigned}``: TCP source port.
2624 - ``dst {unsigned}``: TCP destination port.
2626 - ``sctp``: match SCTP header.
2628 - ``src {unsigned}``: SCTP source port.
2629 - ``dst {unsigned}``: SCTP destination port.
2630 - ``tag {unsigned}``: validation tag.
2631 - ``cksum {unsigned}``: checksum.
2633 - ``vxlan``: match VXLAN header.
2635 - ``vni {unsigned}``: VXLAN identifier.
2637 - ``e_tag``: match IEEE 802.1BR E-Tag header.
2639 - ``grp_ecid_b {unsigned}``: GRP and E-CID base.
2641 - ``nvgre``: match NVGRE header.
2643 - ``tni {unsigned}``: virtual subnet ID.
2645 - ``mpls``: match MPLS header.
2647 - ``label {unsigned}``: MPLS label.
2649 - ``gre``: match GRE header.
2651 - ``protocol {unsigned}``: protocol type.
2653 - ``fuzzy``: fuzzy pattern match, expect faster than default.
2655 - ``thresh {unsigned}``: accuracy threshold.
2660 A list of actions starts after the ``actions`` token in the same fashion as
2661 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
2662 terminated by a mandatory ``end`` action.
2664 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
2665 rte_flow_action_type``).
2667 Dropping all incoming UDPv4 packets can be expressed as follows::
2669 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2672 Several actions have configurable properties which must be specified when
2673 there is no valid default value. For example, ``queue`` requires a target
2676 This rule redirects incoming UDPv4 traffic to queue index 6::
2678 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2679 actions queue index 6 / end
2681 While this one could be rejected by PMDs (unspecified queue index)::
2683 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2686 As defined by *rte_flow*, the list is not ordered, all actions of a given
2687 rule are performed simultaneously. These are equivalent::
2689 queue index 6 / void / mark id 42 / end
2693 void / mark id 42 / queue index 6 / end
2695 All actions in a list should have different types, otherwise only the last
2696 action of a given type is taken into account::
2698 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
2702 drop / drop / drop / end # drop is performed only once
2706 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
2708 Considering they are performed simultaneously, opposite and overlapping
2709 actions can sometimes be combined when the end result is unambiguous::
2711 drop / queue index 6 / end # drop has no effect
2715 drop / dup index 6 / end # same as above
2719 queue index 6 / rss queues 6 7 8 / end # queue has no effect
2723 drop / passthru / end # drop has no effect
2725 Note that PMDs may still refuse such combinations.
2730 This section lists supported actions and their attributes, if any.
2732 - ``end``: end list of actions.
2734 - ``void``: no-op action.
2736 - ``passthru``: let subsequent rule process matched packets.
2738 - ``mark``: attach 32 bit value to packets.
2740 - ``id {unsigned}``: 32 bit value to return with packets.
2742 - ``flag``: flag packets.
2744 - ``queue``: assign packets to a given queue index.
2746 - ``index {unsigned}``: queue index to use.
2748 - ``drop``: drop packets (note: passthru has priority).
2750 - ``count``: enable counters for this rule.
2752 - ``dup``: duplicate packets to a given queue index.
2754 - ``index {unsigned}``: queue index to duplicate packets to.
2756 - ``rss``: spread packets among several queues.
2758 - ``queues [{unsigned} [...]] end``: queue indices to use.
2760 - ``pf``: redirect packets to physical device function.
2762 - ``vf``: redirect packets to virtual device function.
2764 - ``original {boolean}``: use original VF ID if possible.
2765 - ``id {unsigned}``: VF ID to redirect packets to.
2767 Destroying flow rules
2768 ~~~~~~~~~~~~~~~~~~~~~
2770 ``flow destroy`` destroys one or more rules from their rule ID (as returned
2771 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
2772 times as necessary::
2774 flow destroy {port_id} rule {rule_id} [...]
2776 If successful, it will show::
2778 Flow rule #[...] destroyed
2780 It does not report anything for rule IDs that do not exist. The usual error
2781 message is shown when a rule cannot be destroyed::
2783 Caught error type [...] ([...]): [...]
2785 ``flow flush`` destroys all rules on a device and does not take extra
2786 arguments. It is bound to ``rte_flow_flush()``::
2788 flow flush {port_id}
2790 Any errors are reported as above.
2792 Creating several rules and destroying them::
2794 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2795 actions queue index 2 / end
2796 Flow rule #0 created
2797 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2798 actions queue index 3 / end
2799 Flow rule #1 created
2800 testpmd> flow destroy 0 rule 0 rule 1
2801 Flow rule #1 destroyed
2802 Flow rule #0 destroyed
2805 The same result can be achieved using ``flow flush``::
2807 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2808 actions queue index 2 / end
2809 Flow rule #0 created
2810 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2811 actions queue index 3 / end
2812 Flow rule #1 created
2813 testpmd> flow flush 0
2816 Non-existent rule IDs are ignored::
2818 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2819 actions queue index 2 / end
2820 Flow rule #0 created
2821 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2822 actions queue index 3 / end
2823 Flow rule #1 created
2824 testpmd> flow destroy 0 rule 42 rule 10 rule 2
2826 testpmd> flow destroy 0 rule 0
2827 Flow rule #0 destroyed
2833 ``flow query`` queries a specific action of a flow rule having that
2834 ability. Such actions collect information that can be reported using this
2835 command. It is bound to ``rte_flow_query()``::
2837 flow query {port_id} {rule_id} {action}
2839 If successful, it will display either the retrieved data for known actions
2840 or the following message::
2842 Cannot display result for action type [...] ([...])
2844 Otherwise, it will complain either that the rule does not exist or that some
2847 Flow rule #[...] not found
2851 Caught error type [...] ([...]): [...]
2853 Currently only the ``count`` action is supported. This action reports the
2854 number of packets that hit the flow rule and the total number of bytes. Its
2855 output has the following format::
2858 hits_set: [...] # whether "hits" contains a valid value
2859 bytes_set: [...] # whether "bytes" contains a valid value
2860 hits: [...] # number of packets
2861 bytes: [...] # number of bytes
2863 Querying counters for TCPv6 packets redirected to queue 6::
2865 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
2866 actions queue index 6 / count / end
2867 Flow rule #4 created
2868 testpmd> flow query 0 4 count
2879 ``flow list`` lists existing flow rules sorted by priority and optionally
2880 filtered by group identifiers::
2882 flow list {port_id} [group {group_id}] [...]
2884 This command only fails with the following message if the device does not
2889 Output consists of a header line followed by a short description of each
2890 flow rule, one per line. There is no output at all when no flow rules are
2891 configured on the device::
2893 ID Group Prio Attr Rule
2894 [...] [...] [...] [...] [...]
2896 ``Attr`` column flags:
2898 - ``i`` for ``ingress``.
2899 - ``e`` for ``egress``.
2901 Creating several flow rules and listing them::
2903 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2904 actions queue index 6 / end
2905 Flow rule #0 created
2906 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2907 actions queue index 2 / end
2908 Flow rule #1 created
2909 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
2910 actions rss queues 6 7 8 end / end
2911 Flow rule #2 created
2912 testpmd> flow list 0
2913 ID Group Prio Attr Rule
2914 0 0 0 i- ETH IPV4 => QUEUE
2915 1 0 0 i- ETH IPV6 => QUEUE
2916 2 0 5 i- ETH IPV4 UDP => RSS
2919 Rules are sorted by priority (i.e. group ID first, then priority level)::
2921 testpmd> flow list 1
2922 ID Group Prio Attr Rule
2923 0 0 0 i- ETH => COUNT
2924 6 0 500 i- ETH IPV6 TCP => DROP COUNT
2925 5 0 1000 i- ETH IPV6 ICMP => QUEUE
2926 1 24 0 i- ETH IPV4 UDP => QUEUE
2927 4 24 10 i- ETH IPV4 TCP => DROP
2928 3 24 20 i- ETH IPV4 => DROP
2929 2 24 42 i- ETH IPV4 UDP => QUEUE
2930 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
2933 Output can be limited to specific groups::
2935 testpmd> flow list 1 group 0 group 63
2936 ID Group Prio Attr Rule
2937 0 0 0 i- ETH => COUNT
2938 6 0 500 i- ETH IPV6 TCP => DROP COUNT
2939 5 0 1000 i- ETH IPV6 ICMP => QUEUE
2940 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
2943 Toggling isolated mode
2944 ~~~~~~~~~~~~~~~~~~~~~~
2946 ``flow isolate`` can be used to tell the underlying PMD that ingress traffic
2947 must only be injected from the defined flow rules; that no default traffic
2948 is expected outside those rules and the driver is free to assign more
2949 resources to handle them. It is bound to ``rte_flow_isolate()``::
2951 flow isolate {port_id} {boolean}
2953 If successful, enabling or disabling isolated mode shows either::
2955 Ingress traffic on port [...]
2956 is now restricted to the defined flow rules
2960 Ingress traffic on port [...]
2961 is not restricted anymore to the defined flow rules
2963 Otherwise, in case of error::
2965 Caught error type [...] ([...]): [...]
2967 Mainly due to its side effects, PMDs supporting this mode may not have the
2968 ability to toggle it more than once without reinitializing affected ports
2969 first (e.g. by exiting testpmd).
2971 Enabling isolated mode::
2973 testpmd> flow isolate 0 true
2974 Ingress traffic on port 0 is now restricted to the defined flow rules
2977 Disabling isolated mode::
2979 testpmd> flow isolate 0 false
2980 Ingress traffic on port 0 is not restricted anymore to the defined flow rules
2983 Sample QinQ flow rules
2984 ~~~~~~~~~~~~~~~~~~~~~~
2986 Before creating QinQ rule(s) the following commands should be issued to enable QinQ::
2988 testpmd> port stop 0
2989 testpmd> vlan set qinq on 0
2991 The above command sets the inner and outer TPID's to 0x8100.
2993 To change the TPID's the following commands should be used::
2995 testpmd> vlan set outer tpid 0xa100 0
2996 testpmd> vlan set inner tpid 0x9100 0
2997 testpmd> port start 0
2999 Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM.
3003 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 /
3004 vlan tci is 456 / end actions vf id 1 / queue index 0 / end
3005 Flow rule #0 validated
3007 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 /
3008 vlan tci is 456 / end actions vf id 123 / queue index 0 / end
3009 Flow rule #0 created
3011 testpmd> flow list 0
3012 ID Group Prio Attr Rule
3013 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3015 Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host.
3019 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 /
3020 vlan tci is 654 / end actions pf / queue index 0 / end
3021 Flow rule #1 validated
3023 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 /
3024 vlan tci is 654 / end actions pf / queue index 1 / end
3025 Flow rule #1 created
3027 testpmd> flow list 0
3028 ID Group Prio Attr Rule
3029 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3030 1 0 0 i- ETH VLAN VLAN=>PF QUEUE