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
1772 Show the current configuration of a Link Bonding device::
1774 testpmd> show bonding config (port id)
1777 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
1778 in balance mode with a transmission policy of layer 2+3::
1780 testpmd> show bonding config 9
1782 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
1784 Active Slaves (3): [1 3 4]
1791 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
1792 This is mainly useful for debugging purposes.
1793 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
1794 and fields that can be accessed.
1799 Display the value of a port register::
1801 testpmd> read reg (port_id) (address)
1803 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
1805 testpmd> read reg 0 0xEE00
1806 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
1811 Display a port register bit field::
1813 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
1815 For example, reading the lowest two bits from the register in the example above::
1817 testpmd> read regfield 0 0xEE00 0 1
1818 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
1823 Display a single port register bit::
1825 testpmd> read regbit (port_id) (address) (bit_x)
1827 For example, reading the lowest bit from the register in the example above::
1829 testpmd> read regbit 0 0xEE00 0
1830 port 0 PCI register at offset 0xEE00: bit 0=1
1835 Set the value of a port register::
1837 testpmd> write reg (port_id) (address) (value)
1839 For example, to clear a register::
1841 testpmd> write reg 0 0xEE00 0x0
1842 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
1847 Set bit field of a port register::
1849 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
1851 For example, writing to the register cleared in the example above::
1853 testpmd> write regfield 0 0xEE00 0 1 2
1854 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
1859 Set single bit value of a port register::
1861 testpmd> write regbit (port_id) (address) (bit_x) (value)
1863 For example, to set the high bit in the register from the example above::
1865 testpmd> write regbit 0 0xEE00 31 1
1866 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
1872 This section details the available filter functions that are available.
1874 Note these functions interface the deprecated legacy filtering framework,
1875 superseded by *rte_flow*. See `Flow rules management`_.
1878 ~~~~~~~~~~~~~~~~~~~~
1880 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
1882 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
1883 ethertype (ether_type) (drop|fwd) queue (queue_id)
1885 The available information parameters are:
1887 * ``port_id``: The port which the Ethertype filter assigned on.
1889 * ``mac_addr``: Compare destination mac address.
1891 * ``mac_ignr``: Ignore destination mac address match.
1893 * ``mac_address``: Destination mac address to match.
1895 * ``ether_type``: The EtherType value want to match,
1896 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
1898 * ``queue_id``: The receive queue associated with this EtherType filter.
1899 It is meaningless when deleting or dropping.
1901 Example, to add/remove an ethertype filter rule::
1903 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
1904 ethertype 0x0806 fwd queue 3
1906 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
1907 ethertype 0x0806 fwd queue 3
1912 Add or delete a 2-tuple filter,
1913 which identifies packets by specific protocol and destination TCP/UDP port
1914 and forwards packets into one of the receive queues::
1916 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
1917 protocol (protocol_value) mask (mask_value) \
1918 tcp_flags (tcp_flags_value) priority (prio_value) \
1921 The available information parameters are:
1923 * ``port_id``: The port which the 2-tuple filter assigned on.
1925 * ``dst_port_value``: Destination port in L4.
1927 * ``protocol_value``: IP L4 protocol.
1929 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
1931 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
1933 * ``prio_value``: Priority of this filter.
1935 * ``queue_id``: The receive queue associated with this 2-tuple filter.
1937 Example, to add/remove an 2tuple filter rule::
1939 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
1940 tcp_flags 0x02 priority 3 queue 3
1942 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
1943 tcp_flags 0x02 priority 3 queue 3
1948 Add or delete a 5-tuple filter,
1949 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
1950 and routes packets into one of the receive queues::
1952 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
1953 (src_address) dst_port (dst_port_value) \
1954 src_port (src_port_value) protocol (protocol_value) \
1955 mask (mask_value) tcp_flags (tcp_flags_value) \
1956 priority (prio_value) queue (queue_id)
1958 The available information parameters are:
1960 * ``port_id``: The port which the 5-tuple filter assigned on.
1962 * ``dst_address``: Destination IP address.
1964 * ``src_address``: Source IP address.
1966 * ``dst_port_value``: TCP/UDP destination port.
1968 * ``src_port_value``: TCP/UDP source port.
1970 * ``protocol_value``: L4 protocol.
1972 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
1974 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
1976 * ``prio_value``: The priority of this filter.
1978 * ``queue_id``: The receive queue associated with this 5-tuple filter.
1980 Example, to add/remove an 5tuple filter rule::
1982 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1983 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1984 flags 0x0 priority 3 queue 3
1986 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1987 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1988 flags 0x0 priority 3 queue 3
1993 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
1995 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
1997 The available information parameters are:
1999 * ``port_id``: The port which the SYN filter assigned on.
2001 * ``high``: This SYN filter has higher priority than other filters.
2003 * ``low``: This SYN filter has lower priority than other filters.
2005 * ``queue_id``: The receive queue associated with this SYN filter
2009 testpmd> syn_filter 0 add priority high queue 3
2014 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
2015 and routed into one of the receive queues::
2017 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
2018 mask (mask_value) priority (prio_value) queue (queue_id)
2020 The available information parameters are:
2022 * ``port_id``: The port which the Flex filter is assigned on.
2024 * ``len_value``: Filter length in bytes, no greater than 128.
2026 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
2028 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
2030 * ``prio_value``: The priority of this filter.
2032 * ``queue_id``: The receive queue associated with this Flex filter.
2036 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
2037 mask 000C priority 3 queue 3
2039 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
2040 mask 000C priority 3 queue 3
2043 .. _testpmd_flow_director:
2045 flow_director_filter
2046 ~~~~~~~~~~~~~~~~~~~~
2048 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
2050 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
2051 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
2053 * Perfect match filters.
2054 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2055 The masked fields are for IP flow.
2057 * Signature filters.
2058 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
2060 * Perfect-mac-vlan match filters.
2061 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2062 The masked fields are for MAC VLAN flow.
2064 * Perfect-tunnel match filters.
2065 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2066 The masked fields are for tunnel flow.
2068 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
2069 per flow type and the flexible payload.
2071 The Flow Director can also mask out parts of all of these fields so that filters
2072 are only applied to certain fields or parts of the fields.
2074 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
2076 # Commands to add flow director filters of different flow types::
2078 flow_director_filter (port_id) mode IP (add|del|update) \
2079 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
2080 src (src_ip_address) dst (dst_ip_address) \
2081 tos (tos_value) proto (proto_value) ttl (ttl_value) \
2082 vlan (vlan_value) flexbytes (flexbytes_value) \
2083 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
2086 flow_director_filter (port_id) mode IP (add|del|update) \
2087 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
2088 src (src_ip_address) (src_port) \
2089 dst (dst_ip_address) (dst_port) \
2090 tos (tos_value) ttl (ttl_value) \
2091 vlan (vlan_value) flexbytes (flexbytes_value) \
2092 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
2095 flow_director_filter (port_id) mode IP (add|del|update) \
2096 flow (ipv4-sctp|ipv6-sctp) \
2097 src (src_ip_address) (src_port) \
2098 dst (dst_ip_address) (dst_port) \
2099 tos (tos_value) ttl (ttl_value) \
2100 tag (verification_tag) vlan (vlan_value) \
2101 flexbytes (flexbytes_value) (drop|fwd) \
2102 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
2104 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
2105 ether (ethertype) flexbytes (flexbytes_value) \
2106 (drop|fwd) pf|vf(vf_id) queue (queue_id)
2109 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
2110 mac (mac_address) vlan (vlan_value) \
2111 flexbytes (flexbytes_value) (drop|fwd) \
2112 queue (queue_id) fd_id (fd_id_value)
2114 flow_director_filter (port_id) mode Tunnel (add|del|update) \
2115 mac (mac_address) vlan (vlan_value) \
2116 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
2117 flexbytes (flexbytes_value) (drop|fwd) \
2118 queue (queue_id) fd_id (fd_id_value)
2120 For example, to add an ipv4-udp flow type filter::
2122 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
2123 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
2124 fwd pf queue 1 fd_id 1
2126 For example, add an ipv4-other flow type filter::
2128 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
2129 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
2130 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
2135 Flush all flow director filters on a device::
2137 testpmd> flush_flow_director (port_id)
2139 Example, to flush all flow director filter on port 0::
2141 testpmd> flush_flow_director 0
2146 Set flow director's input masks::
2148 flow_director_mask (port_id) mode IP vlan (vlan_value) \
2149 src_mask (ipv4_src) (ipv6_src) (src_port) \
2150 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2152 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2154 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2155 mac (mac_value) tunnel-type (tunnel_type_value) \
2156 tunnel-id (tunnel_id_value)
2158 Example, to set flow director mask on port 0::
2160 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2161 src_mask 255.255.255.255 \
2162 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2163 dst_mask 255.255.255.255 \
2164 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2166 flow_director_flex_mask
2167 ~~~~~~~~~~~~~~~~~~~~~~~
2169 set masks of flow director's flexible payload based on certain flow type::
2171 testpmd> flow_director_flex_mask (port_id) \
2172 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2173 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2174 l2_payload|all) (mask)
2176 Example, to set flow director's flex mask for all flow type on port 0::
2178 testpmd> flow_director_flex_mask 0 flow all \
2179 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2182 flow_director_flex_payload
2183 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2185 Configure flexible payload selection::
2187 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2189 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2191 testpmd> flow_director_flex_payload 0 l4 \
2192 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2194 get_sym_hash_ena_per_port
2195 ~~~~~~~~~~~~~~~~~~~~~~~~~
2197 Get symmetric hash enable configuration per port::
2199 get_sym_hash_ena_per_port (port_id)
2201 For example, to get symmetric hash enable configuration of port 1::
2203 testpmd> get_sym_hash_ena_per_port 1
2205 set_sym_hash_ena_per_port
2206 ~~~~~~~~~~~~~~~~~~~~~~~~~
2208 Set symmetric hash enable configuration per port to enable or disable::
2210 set_sym_hash_ena_per_port (port_id) (enable|disable)
2212 For example, to set symmetric hash enable configuration of port 1 to enable::
2214 testpmd> set_sym_hash_ena_per_port 1 enable
2216 get_hash_global_config
2217 ~~~~~~~~~~~~~~~~~~~~~~
2219 Get the global configurations of hash filters::
2221 get_hash_global_config (port_id)
2223 For example, to get the global configurations of hash filters of port 1::
2225 testpmd> get_hash_global_config 1
2227 set_hash_global_config
2228 ~~~~~~~~~~~~~~~~~~~~~~
2230 Set the global configurations of hash filters::
2232 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2233 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2234 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload) \
2237 For example, to enable simple_xor for flow type of ipv6 on port 2::
2239 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2244 Set the input set for hash::
2246 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2247 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2248 l2_payload) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2249 ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2250 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2251 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2252 fld-8th|none) (select|add)
2254 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2256 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2261 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2262 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2264 Set the input set for flow director::
2266 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2267 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2268 l2_payload) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2269 ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2270 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2271 sctp-dst-port|sctp-veri-tag|none) (select|add)
2273 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2275 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2280 Set different GRE key length for input set::
2282 global_config (port_id) gre-key-len (number in bytes)
2284 For example to set GRE key length for input set to 4 bytes on port 0::
2286 testpmd> global_config 0 gre-key-len 4
2289 .. _testpmd_rte_flow:
2291 Flow rules management
2292 ---------------------
2294 Control of the generic flow API (*rte_flow*) is fully exposed through the
2295 ``flow`` command (validation, creation, destruction, queries and operation
2298 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2299 features simultaneously may cause undefined side-effects and is therefore
2305 Because the ``flow`` command uses dynamic tokens to handle the large number
2306 of possible flow rules combinations, its behavior differs slightly from
2307 other commands, in particular:
2309 - Pressing *?* or the *<tab>* key displays contextual help for the current
2310 token, not that of the entire command.
2312 - Optional and repeated parameters are supported (provided they are listed
2313 in the contextual help).
2315 The first parameter stands for the operation mode. Possible operations and
2316 their general syntax are described below. They are covered in detail in the
2319 - Check whether a flow rule can be created::
2321 flow validate {port_id}
2322 [group {group_id}] [priority {level}] [ingress] [egress]
2323 pattern {item} [/ {item} [...]] / end
2324 actions {action} [/ {action} [...]] / end
2326 - Create a flow rule::
2328 flow create {port_id}
2329 [group {group_id}] [priority {level}] [ingress] [egress]
2330 pattern {item} [/ {item} [...]] / end
2331 actions {action} [/ {action} [...]] / end
2333 - Destroy specific flow rules::
2335 flow destroy {port_id} rule {rule_id} [...]
2337 - Destroy all flow rules::
2339 flow flush {port_id}
2341 - Query an existing flow rule::
2343 flow query {port_id} {rule_id} {action}
2345 - List existing flow rules sorted by priority, filtered by group
2348 flow list {port_id} [group {group_id}] [...]
2350 - Restrict ingress traffic to the defined flow rules::
2352 flow isolate {port_id} {boolean}
2354 Validating flow rules
2355 ~~~~~~~~~~~~~~~~~~~~~
2357 ``flow validate`` reports whether a flow rule would be accepted by the
2358 underlying device in its current state but stops short of creating it. It is
2359 bound to ``rte_flow_validate()``::
2361 flow validate {port_id}
2362 [group {group_id}] [priority {level}] [ingress] [egress]
2363 pattern {item} [/ {item} [...]] / end
2364 actions {action} [/ {action} [...]] / end
2366 If successful, it will show::
2370 Otherwise it will show an error message of the form::
2372 Caught error type [...] ([...]): [...]
2374 This command uses the same parameters as ``flow create``, their format is
2375 described in `Creating flow rules`_.
2377 Check whether redirecting any Ethernet packet received on port 0 to RX queue
2378 index 6 is supported::
2380 testpmd> flow validate 0 ingress pattern eth / end
2381 actions queue index 6 / end
2385 Port 0 does not support TCPv6 rules::
2387 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
2389 Caught error type 9 (specific pattern item): Invalid argument
2395 ``flow create`` validates and creates the specified flow rule. It is bound
2396 to ``rte_flow_create()``::
2398 flow create {port_id}
2399 [group {group_id}] [priority {level}] [ingress] [egress]
2400 pattern {item} [/ {item} [...]] / end
2401 actions {action} [/ {action} [...]] / end
2403 If successful, it will return a flow rule ID usable with other commands::
2405 Flow rule #[...] created
2407 Otherwise it will show an error message of the form::
2409 Caught error type [...] ([...]): [...]
2411 Parameters describe in the following order:
2413 - Attributes (*group*, *priority*, *ingress*, *egress* tokens).
2414 - A matching pattern, starting with the *pattern* token and terminated by an
2416 - Actions, starting with the *actions* token and terminated by an *end*
2419 These translate directly to *rte_flow* objects provided as-is to the
2420 underlying functions.
2422 The shortest valid definition only comprises mandatory tokens::
2424 testpmd> flow create 0 pattern end actions end
2426 Note that PMDs may refuse rules that essentially do nothing such as this
2429 **All unspecified object values are automatically initialized to 0.**
2434 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
2435 specified before the ``pattern`` token.
2437 - ``group {group id}``: priority group.
2438 - ``priority {level}``: priority level within group.
2439 - ``ingress``: rule applies to ingress traffic.
2440 - ``egress``: rule applies to egress traffic.
2442 Each instance of an attribute specified several times overrides the previous
2443 value as shown below (group 4 is used)::
2445 testpmd> flow create 0 group 42 group 24 group 4 [...]
2447 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
2449 While not specifying a direction is an error, some rules may allow both
2452 Most rules affect RX therefore contain the ``ingress`` token::
2454 testpmd> flow create 0 ingress pattern [...]
2459 A matching pattern starts after the ``pattern`` token. It is made of pattern
2460 items and is terminated by a mandatory ``end`` item.
2462 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
2463 rte_flow_item_type``).
2465 The ``/`` token is used as a separator between pattern items as shown
2468 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
2470 Note that protocol items like these must be stacked from lowest to highest
2471 layer to make sense. For instance, the following rule is either invalid or
2472 unlikely to match any packet::
2474 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
2476 More information on these restrictions can be found in the *rte_flow*
2479 Several items support additional specification structures, for example
2480 ``ipv4`` allows specifying source and destination addresses as follows::
2482 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
2483 dst is 10.2.0.0 / end [...]
2485 This rule matches all IPv4 traffic with the specified properties.
2487 In this example, ``src`` and ``dst`` are field names of the underlying
2488 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
2489 in a similar fashion.
2491 The ``is`` token means that the subsequent value must be matched exactly,
2492 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
2493 accordingly. Possible assignment tokens are:
2495 - ``is``: match value perfectly (with full bit-mask).
2496 - ``spec``: match value according to configured bit-mask.
2497 - ``last``: specify upper bound to establish a range.
2498 - ``mask``: specify bit-mask with relevant bits set to one.
2499 - ``prefix``: generate bit-mask from a prefix length.
2501 These yield identical results::
2503 ipv4 src is 10.1.1.1
2507 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
2511 ipv4 src spec 10.1.1.1 src prefix 32
2515 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
2519 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
2521 Inclusive ranges can be defined with ``last``::
2523 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
2525 Note that ``mask`` affects both ``spec`` and ``last``::
2527 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
2528 # matches 10.1.0.0 to 10.2.255.255
2530 Properties can be modified multiple times::
2532 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
2536 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
2541 This section lists supported pattern items and their attributes, if any.
2543 - ``end``: end list of pattern items.
2545 - ``void``: no-op pattern item.
2547 - ``invert``: perform actions when pattern does not match.
2549 - ``any``: match any protocol for the current layer.
2551 - ``num {unsigned}``: number of layers covered.
2553 - ``pf``: match packets addressed to the physical function.
2555 - ``vf``: match packets addressed to a virtual function ID.
2557 - ``id {unsigned}``: destination VF ID.
2559 - ``port``: device-specific physical port index to use.
2561 - ``index {unsigned}``: physical port index.
2563 - ``raw``: match an arbitrary byte string.
2565 - ``relative {boolean}``: look for pattern after the previous item.
2566 - ``search {boolean}``: search pattern from offset (see also limit).
2567 - ``offset {integer}``: absolute or relative offset for pattern.
2568 - ``limit {unsigned}``: search area limit for start of pattern.
2569 - ``pattern {string}``: byte string to look for.
2571 - ``eth``: match Ethernet header.
2573 - ``dst {MAC-48}``: destination MAC.
2574 - ``src {MAC-48}``: source MAC.
2575 - ``type {unsigned}``: EtherType.
2577 - ``vlan``: match 802.1Q/ad VLAN tag.
2579 - ``tpid {unsigned}``: tag protocol identifier.
2580 - ``tci {unsigned}``: tag control information.
2581 - ``pcp {unsigned}``: priority code point.
2582 - ``dei {unsigned}``: drop eligible indicator.
2583 - ``vid {unsigned}``: VLAN identifier.
2585 - ``ipv4``: match IPv4 header.
2587 - ``tos {unsigned}``: type of service.
2588 - ``ttl {unsigned}``: time to live.
2589 - ``proto {unsigned}``: next protocol ID.
2590 - ``src {ipv4 address}``: source address.
2591 - ``dst {ipv4 address}``: destination address.
2593 - ``ipv6``: match IPv6 header.
2595 - ``tc {unsigned}``: traffic class.
2596 - ``flow {unsigned}``: flow label.
2597 - ``proto {unsigned}``: protocol (next header).
2598 - ``hop {unsigned}``: hop limit.
2599 - ``src {ipv6 address}``: source address.
2600 - ``dst {ipv6 address}``: destination address.
2602 - ``icmp``: match ICMP header.
2604 - ``type {unsigned}``: ICMP packet type.
2605 - ``code {unsigned}``: ICMP packet code.
2607 - ``udp``: match UDP header.
2609 - ``src {unsigned}``: UDP source port.
2610 - ``dst {unsigned}``: UDP destination port.
2612 - ``tcp``: match TCP header.
2614 - ``src {unsigned}``: TCP source port.
2615 - ``dst {unsigned}``: TCP destination port.
2617 - ``sctp``: match SCTP header.
2619 - ``src {unsigned}``: SCTP source port.
2620 - ``dst {unsigned}``: SCTP destination port.
2621 - ``tag {unsigned}``: validation tag.
2622 - ``cksum {unsigned}``: checksum.
2624 - ``vxlan``: match VXLAN header.
2626 - ``vni {unsigned}``: VXLAN identifier.
2628 - ``e_tag``: match IEEE 802.1BR E-Tag header.
2630 - ``grp_ecid_b {unsigned}``: GRP and E-CID base.
2632 - ``nvgre``: match NVGRE header.
2634 - ``tni {unsigned}``: virtual subnet ID.
2636 - ``mpls``: match MPLS header.
2638 - ``label {unsigned}``: MPLS label.
2640 - ``gre``: match GRE header.
2642 - ``protocol {unsigned}``: protocol type.
2644 - ``fuzzy``: fuzzy pattern match, expect faster than default.
2646 - ``thresh {unsigned}``: accuracy threshold.
2651 A list of actions starts after the ``actions`` token in the same fashion as
2652 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
2653 terminated by a mandatory ``end`` action.
2655 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
2656 rte_flow_action_type``).
2658 Dropping all incoming UDPv4 packets can be expressed as follows::
2660 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2663 Several actions have configurable properties which must be specified when
2664 there is no valid default value. For example, ``queue`` requires a target
2667 This rule redirects incoming UDPv4 traffic to queue index 6::
2669 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2670 actions queue index 6 / end
2672 While this one could be rejected by PMDs (unspecified queue index)::
2674 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2677 As defined by *rte_flow*, the list is not ordered, all actions of a given
2678 rule are performed simultaneously. These are equivalent::
2680 queue index 6 / void / mark id 42 / end
2684 void / mark id 42 / queue index 6 / end
2686 All actions in a list should have different types, otherwise only the last
2687 action of a given type is taken into account::
2689 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
2693 drop / drop / drop / end # drop is performed only once
2697 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
2699 Considering they are performed simultaneously, opposite and overlapping
2700 actions can sometimes be combined when the end result is unambiguous::
2702 drop / queue index 6 / end # drop has no effect
2706 drop / dup index 6 / end # same as above
2710 queue index 6 / rss queues 6 7 8 / end # queue has no effect
2714 drop / passthru / end # drop has no effect
2716 Note that PMDs may still refuse such combinations.
2721 This section lists supported actions and their attributes, if any.
2723 - ``end``: end list of actions.
2725 - ``void``: no-op action.
2727 - ``passthru``: let subsequent rule process matched packets.
2729 - ``mark``: attach 32 bit value to packets.
2731 - ``id {unsigned}``: 32 bit value to return with packets.
2733 - ``flag``: flag packets.
2735 - ``queue``: assign packets to a given queue index.
2737 - ``index {unsigned}``: queue index to use.
2739 - ``drop``: drop packets (note: passthru has priority).
2741 - ``count``: enable counters for this rule.
2743 - ``dup``: duplicate packets to a given queue index.
2745 - ``index {unsigned}``: queue index to duplicate packets to.
2747 - ``rss``: spread packets among several queues.
2749 - ``queues [{unsigned} [...]] end``: queue indices to use.
2751 - ``pf``: redirect packets to physical device function.
2753 - ``vf``: redirect packets to virtual device function.
2755 - ``original {boolean}``: use original VF ID if possible.
2756 - ``id {unsigned}``: VF ID to redirect packets to.
2758 Destroying flow rules
2759 ~~~~~~~~~~~~~~~~~~~~~
2761 ``flow destroy`` destroys one or more rules from their rule ID (as returned
2762 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
2763 times as necessary::
2765 flow destroy {port_id} rule {rule_id} [...]
2767 If successful, it will show::
2769 Flow rule #[...] destroyed
2771 It does not report anything for rule IDs that do not exist. The usual error
2772 message is shown when a rule cannot be destroyed::
2774 Caught error type [...] ([...]): [...]
2776 ``flow flush`` destroys all rules on a device and does not take extra
2777 arguments. It is bound to ``rte_flow_flush()``::
2779 flow flush {port_id}
2781 Any errors are reported as above.
2783 Creating several rules and destroying them::
2785 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2786 actions queue index 2 / end
2787 Flow rule #0 created
2788 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2789 actions queue index 3 / end
2790 Flow rule #1 created
2791 testpmd> flow destroy 0 rule 0 rule 1
2792 Flow rule #1 destroyed
2793 Flow rule #0 destroyed
2796 The same result can be achieved using ``flow flush``::
2798 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2799 actions queue index 2 / end
2800 Flow rule #0 created
2801 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2802 actions queue index 3 / end
2803 Flow rule #1 created
2804 testpmd> flow flush 0
2807 Non-existent rule IDs are ignored::
2809 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2810 actions queue index 2 / end
2811 Flow rule #0 created
2812 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2813 actions queue index 3 / end
2814 Flow rule #1 created
2815 testpmd> flow destroy 0 rule 42 rule 10 rule 2
2817 testpmd> flow destroy 0 rule 0
2818 Flow rule #0 destroyed
2824 ``flow query`` queries a specific action of a flow rule having that
2825 ability. Such actions collect information that can be reported using this
2826 command. It is bound to ``rte_flow_query()``::
2828 flow query {port_id} {rule_id} {action}
2830 If successful, it will display either the retrieved data for known actions
2831 or the following message::
2833 Cannot display result for action type [...] ([...])
2835 Otherwise, it will complain either that the rule does not exist or that some
2838 Flow rule #[...] not found
2842 Caught error type [...] ([...]): [...]
2844 Currently only the ``count`` action is supported. This action reports the
2845 number of packets that hit the flow rule and the total number of bytes. Its
2846 output has the following format::
2849 hits_set: [...] # whether "hits" contains a valid value
2850 bytes_set: [...] # whether "bytes" contains a valid value
2851 hits: [...] # number of packets
2852 bytes: [...] # number of bytes
2854 Querying counters for TCPv6 packets redirected to queue 6::
2856 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
2857 actions queue index 6 / count / end
2858 Flow rule #4 created
2859 testpmd> flow query 0 4 count
2870 ``flow list`` lists existing flow rules sorted by priority and optionally
2871 filtered by group identifiers::
2873 flow list {port_id} [group {group_id}] [...]
2875 This command only fails with the following message if the device does not
2880 Output consists of a header line followed by a short description of each
2881 flow rule, one per line. There is no output at all when no flow rules are
2882 configured on the device::
2884 ID Group Prio Attr Rule
2885 [...] [...] [...] [...] [...]
2887 ``Attr`` column flags:
2889 - ``i`` for ``ingress``.
2890 - ``e`` for ``egress``.
2892 Creating several flow rules and listing them::
2894 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2895 actions queue index 6 / end
2896 Flow rule #0 created
2897 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2898 actions queue index 2 / end
2899 Flow rule #1 created
2900 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
2901 actions rss queues 6 7 8 end / end
2902 Flow rule #2 created
2903 testpmd> flow list 0
2904 ID Group Prio Attr Rule
2905 0 0 0 i- ETH IPV4 => QUEUE
2906 1 0 0 i- ETH IPV6 => QUEUE
2907 2 0 5 i- ETH IPV4 UDP => RSS
2910 Rules are sorted by priority (i.e. group ID first, then priority level)::
2912 testpmd> flow list 1
2913 ID Group Prio Attr Rule
2914 0 0 0 i- ETH => COUNT
2915 6 0 500 i- ETH IPV6 TCP => DROP COUNT
2916 5 0 1000 i- ETH IPV6 ICMP => QUEUE
2917 1 24 0 i- ETH IPV4 UDP => QUEUE
2918 4 24 10 i- ETH IPV4 TCP => DROP
2919 3 24 20 i- ETH IPV4 => DROP
2920 2 24 42 i- ETH IPV4 UDP => QUEUE
2921 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
2924 Output can be limited to specific groups::
2926 testpmd> flow list 1 group 0 group 63
2927 ID Group Prio Attr Rule
2928 0 0 0 i- ETH => COUNT
2929 6 0 500 i- ETH IPV6 TCP => DROP COUNT
2930 5 0 1000 i- ETH IPV6 ICMP => QUEUE
2931 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
2934 Toggling isolated mode
2935 ~~~~~~~~~~~~~~~~~~~~~~
2937 ``flow isolate`` can be used to tell the underlying PMD that ingress traffic
2938 must only be injected from the defined flow rules; that no default traffic
2939 is expected outside those rules and the driver is free to assign more
2940 resources to handle them. It is bound to ``rte_flow_isolate()``::
2942 flow isolate {port_id} {boolean}
2944 If successful, enabling or disabling isolated mode shows either::
2946 Ingress traffic on port [...]
2947 is now restricted to the defined flow rules
2951 Ingress traffic on port [...]
2952 is not restricted anymore to the defined flow rules
2954 Otherwise, in case of error::
2956 Caught error type [...] ([...]): [...]
2958 Mainly due to its side effects, PMDs supporting this mode may not have the
2959 ability to toggle it more than once without reinitializing affected ports
2960 first (e.g. by exiting testpmd).
2962 Enabling isolated mode::
2964 testpmd> flow isolate 0 true
2965 Ingress traffic on port 0 is now restricted to the defined flow rules
2968 Disabling isolated mode::
2970 testpmd> flow isolate 0 false
2971 Ingress traffic on port 0 is not restricted anymore to the defined flow rules
2974 Sample QinQ flow rules
2975 ~~~~~~~~~~~~~~~~~~~~~~
2977 Before creating QinQ rule(s) the following commands should be issued to enable QinQ::
2979 testpmd> port stop 0
2980 testpmd> vlan set qinq on 0
2982 The above command sets the inner and outer TPID's to 0x8100.
2984 To change the TPID's the following commands should be used::
2986 testpmd> vlan set outer tpid 0xa100 0
2987 testpmd> vlan set inner tpid 0x9100 0
2988 testpmd> port start 0
2990 Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM.
2994 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 /
2995 vlan tci is 456 / end actions vf id 1 / queue index 0 / end
2996 Flow rule #0 validated
2998 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 /
2999 vlan tci is 456 / end actions vf id 123 / queue index 0 / end
3000 Flow rule #0 created
3002 testpmd> flow list 0
3003 ID Group Prio Attr Rule
3004 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3006 Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host.
3010 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 /
3011 vlan tci is 654 / end actions pf / queue index 0 / end
3012 Flow rule #1 validated
3014 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 /
3015 vlan tci is 654 / end actions pf / queue index 1 / end
3016 Flow rule #1 created
3018 testpmd> flow list 0
3019 ID Group Prio Attr Rule
3020 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3021 1 0 0 i- ETH VLAN VLAN=>PF QUEUE