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 VF statistics::
395 testpmd> show vf stats (port_id) (vf_id)
400 Reset VF statistics::
402 testpmd> clear vf stats (port_id) (vf_id)
404 Configuration Functions
405 -----------------------
407 The testpmd application can be configured from the runtime as well as from the command-line.
409 This section details the available configuration functions that are available.
413 Configuration changes only become active when forwarding is started/restarted.
418 Reset forwarding to the default configuration::
425 Set the debug verbosity level::
427 testpmd> set verbose (level)
429 Currently the only available levels are 0 (silent except for error) and 1 (fully verbose).
434 Set the number of ports used by the application:
438 This is equivalent to the ``--nb-ports`` command-line option.
443 Set the number of cores used by the application::
445 testpmd> set nbcore (num)
447 This is equivalent to the ``--nb-cores`` command-line option.
451 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port.
456 Set the forwarding cores hexadecimal mask::
458 testpmd> set coremask (mask)
460 This is equivalent to the ``--coremask`` command-line option.
464 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding.
469 Set the forwarding ports hexadecimal mask::
471 testpmd> set portmask (mask)
473 This is equivalent to the ``--portmask`` command-line option.
478 Set number of packets per burst::
480 testpmd> set burst (num)
482 This is equivalent to the ``--burst command-line`` option.
484 When retry is enabled, the transmit delay time and number of retries can also be set::
486 testpmd> set burst tx delay (microseconds) retry (num)
491 Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode::
493 testpmd> set txpkts (x[,y]*)
495 Where x[,y]* represents a CSV list of values, without white space.
500 Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes::
502 testpmd> set txsplit (off|on|rand)
506 * ``off`` disable packet copy & split for CSUM mode.
508 * ``on`` split outgoing packet into multiple segments. Size of each segment
509 and number of segments per packet is determined by ``set txpkts`` command
512 * ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments.
517 Set the list of forwarding cores::
519 testpmd> set corelist (x[,y]*)
521 For example, to change the forwarding cores:
523 .. code-block:: console
525 testpmd> set corelist 3,1
526 testpmd> show config fwd
528 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled
529 Logical Core 3 (socket 0) forwards packets on 1 streams:
530 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01
531 Logical Core 1 (socket 0) forwards packets on 1 streams:
532 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
536 The cores are used in the same order as specified on the command line.
541 Set the list of forwarding ports::
543 testpmd> set portlist (x[,y]*)
545 For example, to change the port forwarding:
547 .. code-block:: console
549 testpmd> set portlist 0,2,1,3
550 testpmd> show config fwd
552 io packet forwarding - ports=4 - cores=1 - streams=4
553 Logical Core 3 (socket 0) forwards packets on 4 streams:
554 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01
555 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
556 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03
557 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02
562 Enable/disable tx loopback::
564 testpmd> set tx loopback (port_id) (on|off)
569 set drop enable bit for all queues::
571 testpmd> set all queues drop (port_id) (on|off)
573 set split drop enable (for VF)
574 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
576 set split drop enable bit for VF from PF::
578 testpmd> set vf split drop (port_id) (vf_id) (on|off)
580 set mac antispoof (for VF)
581 ~~~~~~~~~~~~~~~~~~~~~~~~~~
583 Set mac antispoof for a VF from the PF::
585 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off)
590 Enable/disable MACsec offload::
592 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off)
593 testpmd> set macsec offload (port_id) off
598 Configure MACsec secure connection (SC)::
600 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi)
604 The pi argument is ignored for tx.
605 Check the NIC Datasheet for hardware limits.
610 Configure MACsec secure association (SA)::
612 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key)
616 The IDX value must be 0 or 1.
617 Check the NIC Datasheet for hardware limits.
619 set broadcast mode (for VF)
620 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
622 Set broadcast mode for a VF from the PF::
624 testpmd> set vf broadcast (port_id) (vf_id) (on|off)
629 Set the VLAN strip on a port::
631 testpmd> vlan set strip (on|off) (port_id)
636 Set the VLAN strip for a queue on a port::
638 testpmd> vlan set stripq (on|off) (port_id,queue_id)
640 vlan set stripq (for VF)
641 ~~~~~~~~~~~~~~~~~~~~~~~~
643 Set VLAN strip for all queues in a pool for a VF from the PF::
645 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off)
647 vlan set insert (for VF)
648 ~~~~~~~~~~~~~~~~~~~~~~~~
650 Set VLAN insert for a VF from the PF::
652 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id)
654 vlan set tag (for VF)
655 ~~~~~~~~~~~~~~~~~~~~~
657 Set VLAN tag for a VF from the PF::
659 testpmd> set vf vlan tag (port_id) (vf_id) (on|off)
661 vlan set antispoof (for VF)
662 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
664 Set VLAN antispoof for a VF from the PF::
666 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off)
671 Set the VLAN filter on a port::
673 testpmd> vlan set filter (on|off) (port_id)
678 Set the VLAN QinQ (extended queue in queue) on for a port::
680 testpmd> vlan set qinq (on|off) (port_id)
685 Set the inner or outer VLAN TPID for packet filtering on a port::
687 testpmd> vlan set (inner|outer) tpid (value) (port_id)
691 TPID value must be a 16-bit number (value <= 65536).
696 Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID::
698 testpmd> rx_vlan add (vlan_id|all) (port_id)
702 VLAN filter must be set on that port. VLAN ID < 4096.
703 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries
704 in VFTA table. This is important if enabling all vlan_ids.
709 Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID::
711 testpmd> rx_vlan rm (vlan_id|all) (port_id)
716 Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID::
718 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask)
723 Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID::
725 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask)
730 Add a tunnel filter on a port::
732 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \
733 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
734 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
736 The available information categories are:
738 * ``vxlan``: Set tunnel type as VXLAN.
740 * ``nvgre``: Set tunnel type as NVGRE.
742 * ``ipingre``: Set tunnel type as IP-in-GRE.
744 * ``imac-ivlan``: Set filter type as Inner MAC and VLAN.
746 * ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID.
748 * ``imac-tenid``: Set filter type as Inner MAC and tenant ID.
750 * ``imac``: Set filter type as Inner MAC.
752 * ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID.
754 * ``oip``: Set filter type as Outer IP.
756 * ``iip``: Set filter type as Inner IP.
760 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \
761 192.168.2.2 0 ipingre oip 1 1
763 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP.
768 Remove a tunnel filter on a port::
770 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \
771 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
772 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
777 Add an UDP port for VXLAN packet filter on a port::
779 testpmd> rx_vxlan_port add (udp_port) (port_id)
784 Remove an UDP port for VXLAN packet filter on a port::
786 testpmd> rx_vxlan_port rm (udp_port) (port_id)
791 Set hardware insertion of VLAN IDs in packets sent on a port::
793 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer]
795 For example, set a single VLAN ID (5) insertion on port 0::
799 Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1::
807 Set port based hardware insertion of VLAN ID in packets sent on a port::
809 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off)
814 Disable hardware insertion of a VLAN header in packets sent on a port::
816 testpmd> tx_vlan reset (port_id)
821 Select hardware or software calculation of the checksum when
822 transmitting a packet using the ``csum`` forwarding engine::
824 testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id)
828 * ``ip|udp|tcp|sctp`` always relate to the inner layer.
830 * ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized
831 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are
832 supported). See also the ``csum parse-tunnel`` command.
836 Check the NIC Datasheet for hardware limits.
841 Define how tunneled packets should be handled by the csum forward
844 testpmd> csum parse-tunnel (on|off) (tx_port_id)
846 If enabled, the csum forward engine will try to recognize supported
847 tunnel headers (vxlan, gre, ipip).
849 If disabled, treat tunnel packets as non-tunneled packets (a inner
850 header is handled as a packet payload).
854 The port argument is the TX port like in the ``csum set`` command.
858 Consider a packet in packet like the following::
860 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in
862 * If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
863 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the
864 ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``).
866 * If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
867 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``.
872 Display tx checksum offload configuration::
874 testpmd> csum show (port_id)
879 Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine::
881 testpmd> tso set (segsize) (port_id)
885 Check the NIC datasheet for hardware limits.
890 Display the status of TCP Segmentation Offload::
892 testpmd> tso show (port_id)
897 Add an alternative MAC address to a port::
899 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX)
904 Remove a MAC address from a port::
906 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX)
908 mac_addr add (for VF)
909 ~~~~~~~~~~~~~~~~~~~~~
911 Add an alternative MAC address for a VF to a port::
913 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX)
918 Set the default MAC address for a port::
920 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX)
922 mac_addr set (for VF)
923 ~~~~~~~~~~~~~~~~~~~~~
925 Set the MAC address for a VF from the PF::
927 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX)
932 Set the unicast hash filter(s) on/off for a port::
934 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
939 Set the promiscuous mode on for a port or for all ports.
940 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
942 testpmd> set promisc (port_id|all) (on|off)
947 Set the allmulti mode for a port or for all ports::
949 testpmd> set allmulti (port_id|all) (on|off)
951 Same as the ifconfig (8) option. Controls how multicast packets are handled.
956 Set the unicast promiscuous mode for a VF from PF.
957 It's supported by Intel i40e NICs now.
958 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
960 testpmd> set vf promisc (port_id) (vf_id) (on|off)
962 set allmulticast (for VF)
963 ~~~~~~~~~~~~~~~~~~~~~~~~~
965 Set the multicast promiscuous mode for a VF from PF.
966 It's supported by Intel i40e NICs now.
967 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
969 testpmd> set vf allmulti (port_id) (vf_id) (on|off)
971 set tx max bandwidth (for VF)
972 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
974 Set TX max absolute bandwidth (Mbps) for a VF from PF::
976 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth)
978 set tc tx min bandwidth (for VF)
979 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
981 Set all TCs' TX min relative bandwidth (%) for a VF from PF::
983 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...)
985 set tc tx max bandwidth (for VF)
986 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
988 Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF::
990 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth)
992 set tc strict link priority mode
993 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
995 Set some TCs' strict link priority mode on a physical port::
997 testpmd> set tx strict-link-priority (port_id) (tc_bitmap)
999 set tc tx min bandwidth
1000 ~~~~~~~~~~~~~~~~~~~~~~~
1002 Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs::
1004 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...)
1009 Set the link flow control parameter on a port::
1011 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1012 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
1013 autoneg (on|off) (port_id)
1017 * ``high_water`` (integer): High threshold value to trigger XOFF.
1019 * ``low_water`` (integer): Low threshold value to trigger XON.
1021 * ``pause_time`` (integer): Pause quota in the Pause frame.
1023 * ``send_xon`` (0/1): Send XON frame.
1025 * ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
1027 * ``autoneg``: Change the auto-negotiation parameter.
1032 Set the priority flow control parameter on a port::
1034 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1035 (pause_time) (priority) (port_id)
1039 * ``high_water`` (integer): High threshold value.
1041 * ``low_water`` (integer): Low threshold value.
1043 * ``pause_time`` (integer): Pause quota in the Pause frame.
1045 * ``priority`` (0-7): VLAN User Priority.
1050 Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
1052 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
1054 For example, to set rx queue 2 on port 0 to mapping 5::
1056 testpmd>set stat_qmap rx 0 2 5
1058 set port - rx/tx (for VF)
1059 ~~~~~~~~~~~~~~~~~~~~~~~~~
1061 Set VF receive/transmit from a port::
1063 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
1065 set port - mac address filter (for VF)
1066 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1068 Add/Remove unicast or multicast MAC addr filter for a VF::
1070 testpmd> set port (port_id) vf (vf_id) (mac_addr) \
1071 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
1073 set port - rx mode(for VF)
1074 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1076 Set the VF receive mode of a port::
1078 testpmd> set port (port_id) vf (vf_id) \
1079 rxmode (AUPE|ROPE|BAM|MPE) (on|off)
1081 The available receive modes are:
1083 * ``AUPE``: Accepts untagged VLAN.
1085 * ``ROPE``: Accepts unicast hash.
1087 * ``BAM``: Accepts broadcast packets.
1089 * ``MPE``: Accepts all multicast packets.
1091 set port - tx_rate (for Queue)
1092 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1094 Set TX rate limitation for a queue on a port::
1096 testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1098 set port - tx_rate (for VF)
1099 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
1101 Set TX rate limitation for queues in VF on a port::
1103 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1105 set port - mirror rule
1106 ~~~~~~~~~~~~~~~~~~~~~~
1108 Set pool or vlan type mirror rule for a port::
1110 testpmd> set port (port_id) mirror-rule (rule_id) \
1111 (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1112 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1114 Set link mirror rule for a port::
1116 testpmd> set port (port_id) mirror-rule (rule_id) \
1117 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1119 For example to enable mirror traffic with vlan 0,1 to pool 0::
1121 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1123 reset port - mirror rule
1124 ~~~~~~~~~~~~~~~~~~~~~~~~
1126 Reset a mirror rule for a port::
1128 testpmd> reset port (port_id) mirror-rule (rule_id)
1133 Set the flush on RX streams before forwarding.
1134 The default is flush ``on``.
1135 Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1137 testpmd> set flush_rx off
1142 Set the bypass mode for the lowest port on bypass enabled NIC::
1144 testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1149 Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1151 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1152 mode (normal|bypass|isolate) (port_id)
1156 * ``timeout``: Enable bypass after watchdog timeout.
1158 * ``os_on``: Enable bypass when OS/board is powered on.
1160 * ``os_off``: Enable bypass when OS/board is powered off.
1162 * ``power_on``: Enable bypass when power supply is turned on.
1164 * ``power_off``: Enable bypass when power supply is turned off.
1170 Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1172 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1177 Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1179 testpmd> show bypass config (port_id)
1184 Set link up for a port::
1186 testpmd> set link-up port (port id)
1191 Set link down for a port::
1193 testpmd> set link-down port (port id)
1198 Enable E-tag insertion for a VF on a port::
1200 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1202 Disable E-tag insertion for a VF on a port::
1204 testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1206 Enable/disable E-tag stripping on a port::
1208 testpmd> E-tag set stripping (on|off) port (port_id)
1210 Enable/disable E-tag based forwarding on a port::
1212 testpmd> E-tag set forwarding (on|off) port (port_id)
1214 Add an E-tag forwarding filter on a port::
1216 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1218 Delete an E-tag forwarding filter on a port::
1219 testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1224 Load a dynamic device personalization (DDP) package::
1226 testpmd> ddp add (port_id) (package_path)
1231 List all items from the ptype mapping table::
1233 testpmd> ptype mapping get (port_id) (valid_only)
1237 * ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0).
1239 Replace a specific or a group of software defined ptype with a new one::
1241 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type)
1245 * ``target``: A specific software ptype or a mask to represent a group of software ptypes.
1247 * ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1).
1249 * ``pkt_type``: The new software ptype to replace the old ones.
1251 Update hardware defined ptype to software defined packet type mapping table::
1253 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype)
1257 * ``hw_ptype``: hardware ptype as the index of the ptype mapping table.
1259 * ``sw_ptype``: software ptype as the value of the ptype mapping table.
1261 Reset ptype mapping table::
1263 testpmd> ptype mapping reset (port_id)
1268 The following sections show functions for configuring ports.
1272 Port configuration changes only become active when forwarding is started/restarted.
1277 Attach a port specified by pci address or virtual device args::
1279 testpmd> port attach (identifier)
1281 To attach a new pci device, the device should be recognized by kernel first.
1282 Then it should be moved under DPDK management.
1283 Finally the port can be attached to testpmd.
1285 For example, to move a pci device using ixgbe under DPDK management:
1287 .. code-block:: console
1289 # Check the status of the available devices.
1290 ./usertools/dpdk-devbind.py --status
1292 Network devices using DPDK-compatible driver
1293 ============================================
1296 Network devices using kernel driver
1297 ===================================
1298 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1301 # Bind the device to igb_uio.
1302 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1305 # Recheck the status of the devices.
1306 ./usertools/dpdk-devbind.py --status
1307 Network devices using DPDK-compatible driver
1308 ============================================
1309 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1311 To attach a port created by virtual device, above steps are not needed.
1313 For example, to attach a port whose pci address is 0000:0a:00.0.
1315 .. code-block:: console
1317 testpmd> port attach 0000:0a:00.0
1318 Attaching a new port...
1319 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1320 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1321 EAL: PCI memory mapped at 0x7f83bfa00000
1322 EAL: PCI memory mapped at 0x7f83bfa80000
1323 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1324 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1325 Port 0 is attached. Now total ports is 1
1328 For example, to attach a port created by pcap PMD.
1330 .. code-block:: console
1332 testpmd> port attach net_pcap0
1333 Attaching a new port...
1334 PMD: Initializing pmd_pcap for net_pcap0
1335 PMD: Creating pcap-backed ethdev on numa socket 0
1336 Port 0 is attached. Now total ports is 1
1339 In this case, identifier is ``net_pcap0``.
1340 This identifier format is the same as ``--vdev`` format of DPDK applications.
1342 For example, to re-attach a bonded port which has been previously detached,
1343 the mode and slave parameters must be given.
1345 .. code-block:: console
1347 testpmd> port attach net_bond_0,mode=0,slave=1
1348 Attaching a new port...
1349 EAL: Initializing pmd_bond for net_bond_0
1350 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1351 Port 0 is attached. Now total ports is 1
1358 Detach a specific port::
1360 testpmd> port detach (port_id)
1362 Before detaching a port, the port should be stopped and closed.
1364 For example, to detach a pci device port 0.
1366 .. code-block:: console
1368 testpmd> port stop 0
1371 testpmd> port close 0
1375 testpmd> port detach 0
1377 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1378 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1379 EAL: PCI memory unmapped at 0x7f83bfa00000
1380 EAL: PCI memory unmapped at 0x7f83bfa80000
1384 For example, to detach a virtual device port 0.
1386 .. code-block:: console
1388 testpmd> port stop 0
1391 testpmd> port close 0
1395 testpmd> port detach 0
1397 PMD: Closing pcap ethdev on numa socket 0
1398 Port 'net_pcap0' is detached. Now total ports is 0
1401 To remove a pci device completely from the system, first detach the port from testpmd.
1402 Then the device should be moved under kernel management.
1403 Finally the device can be removed using kernel pci hotplug functionality.
1405 For example, to move a pci device under kernel management:
1407 .. code-block:: console
1409 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1411 ./usertools/dpdk-devbind.py --status
1413 Network devices using DPDK-compatible driver
1414 ============================================
1417 Network devices using kernel driver
1418 ===================================
1419 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1421 To remove a port created by a virtual device, above steps are not needed.
1426 Start all ports or a specific port::
1428 testpmd> port start (port_id|all)
1433 Stop all ports or a specific port::
1435 testpmd> port stop (port_id|all)
1440 Close all ports or a specific port::
1442 testpmd> port close (port_id|all)
1444 port start/stop queue
1445 ~~~~~~~~~~~~~~~~~~~~~
1447 Start/stop a rx/tx queue on a specific port::
1449 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1451 Only take effect when port is started.
1456 Set the speed and duplex mode for all ports or a specific port::
1458 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1459 duplex (half|full|auto)
1461 port config - queues/descriptors
1462 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1464 Set number of queues/descriptors for rxq, txq, rxd and txd::
1466 testpmd> port config all (rxq|txq|rxd|txd) (value)
1468 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1470 port config - max-pkt-len
1471 ~~~~~~~~~~~~~~~~~~~~~~~~~
1473 Set the maximum packet length::
1475 testpmd> port config all max-pkt-len (value)
1477 This is equivalent to the ``--max-pkt-len`` command-line option.
1479 port config - CRC Strip
1480 ~~~~~~~~~~~~~~~~~~~~~~~
1482 Set hardware CRC stripping on or off for all ports::
1484 testpmd> port config all crc-strip (on|off)
1486 CRC stripping is on by default.
1488 The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option.
1490 port config - scatter
1491 ~~~~~~~~~~~~~~~~~~~~~~~
1493 Set RX scatter mode on or off for all ports::
1495 testpmd> port config all scatter (on|off)
1497 RX scatter mode is off by default.
1499 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1501 port config - TX queue flags
1502 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1504 Set a hexadecimal bitmap of TX queue flags for all ports::
1506 testpmd> port config all txqflags value
1508 This command is equivalent to the ``--txqflags`` command-line option.
1510 port config - RX Checksum
1511 ~~~~~~~~~~~~~~~~~~~~~~~~~
1513 Set hardware RX checksum offload to on or off for all ports::
1515 testpmd> port config all rx-cksum (on|off)
1517 Checksum offload is off by default.
1519 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1524 Set hardware VLAN on or off for all ports::
1526 testpmd> port config all hw-vlan (on|off)
1528 Hardware VLAN is on by default.
1530 The ``off`` option is equivalent to the ``--disable-hw-vlan`` command-line option.
1532 port config - VLAN filter
1533 ~~~~~~~~~~~~~~~~~~~~~~~~~
1535 Set hardware VLAN filter on or off for all ports::
1537 testpmd> port config all hw-vlan-filter (on|off)
1539 Hardware VLAN filter is on by default.
1541 The ``off`` option is equivalent to the ``--disable-hw-vlan-filter`` command-line option.
1543 port config - VLAN strip
1544 ~~~~~~~~~~~~~~~~~~~~~~~~
1546 Set hardware VLAN strip on or off for all ports::
1548 testpmd> port config all hw-vlan-strip (on|off)
1550 Hardware VLAN strip is on by default.
1552 The ``off`` option is equivalent to the ``--disable-hw-vlan-strip`` command-line option.
1554 port config - VLAN extend
1555 ~~~~~~~~~~~~~~~~~~~~~~~~~
1557 Set hardware VLAN extend on or off for all ports::
1559 testpmd> port config all hw-vlan-extend (on|off)
1561 Hardware VLAN extend is off by default.
1563 The ``off`` option is equivalent to the ``--disable-hw-vlan-extend`` command-line option.
1565 port config - Drop Packets
1566 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1568 Set packet drop for packets with no descriptors on or off for all ports::
1570 testpmd> port config all drop-en (on|off)
1572 Packet dropping for packets with no descriptors is off by default.
1574 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1579 Set the RSS (Receive Side Scaling) mode on or off::
1581 testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1583 RSS is on by default.
1585 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1587 port config - RSS Reta
1588 ~~~~~~~~~~~~~~~~~~~~~~
1590 Set the RSS (Receive Side Scaling) redirection table::
1592 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1597 Set the DCB mode for an individual port::
1599 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1601 The traffic class should be 4 or 8.
1606 Set the number of packets per burst::
1608 testpmd> port config all burst (value)
1610 This is equivalent to the ``--burst`` command-line option.
1612 port config - Threshold
1613 ~~~~~~~~~~~~~~~~~~~~~~~
1615 Set thresholds for TX/RX queues::
1617 testpmd> port config all (threshold) (value)
1619 Where the threshold type can be:
1621 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1623 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1625 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1627 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1629 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1631 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1633 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1635 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1637 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1639 These threshold options are also available from the command-line.
1644 Set the value of ether-type for E-tag::
1646 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1648 Enable/disable the E-tag support::
1650 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1653 Link Bonding Functions
1654 ----------------------
1656 The Link Bonding functions make it possible to dynamically create and
1657 manage link bonding devices from within testpmd interactive prompt.
1659 create bonded device
1660 ~~~~~~~~~~~~~~~~~~~~
1662 Create a new bonding device::
1664 testpmd> create bonded device (mode) (socket)
1666 For example, to create a bonded device in mode 1 on socket 0::
1668 testpmd> create bonded 1 0
1669 created new bonded device (port X)
1674 Adds Ethernet device to a Link Bonding device::
1676 testpmd> add bonding slave (slave id) (port id)
1678 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
1680 testpmd> add bonding slave 6 10
1683 remove bonding slave
1684 ~~~~~~~~~~~~~~~~~~~~
1686 Removes an Ethernet slave device from a Link Bonding device::
1688 testpmd> remove bonding slave (slave id) (port id)
1690 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
1692 testpmd> remove bonding slave 6 10
1697 Set the Link Bonding mode of a Link Bonding device::
1699 testpmd> set bonding mode (value) (port id)
1701 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
1703 testpmd> set bonding mode 3 10
1708 Set an Ethernet slave device as the primary device on a Link Bonding device::
1710 testpmd> set bonding primary (slave id) (port id)
1712 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
1714 testpmd> set bonding primary 6 10
1719 Set the MAC address of a Link Bonding device::
1721 testpmd> set bonding mac (port id) (mac)
1723 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
1725 testpmd> set bonding mac 10 00:00:00:00:00:01
1727 set bonding xmit_balance_policy
1728 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1730 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
1732 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
1734 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
1736 testpmd> set bonding xmit_balance_policy 10 l34
1739 set bonding mon_period
1740 ~~~~~~~~~~~~~~~~~~~~~~
1742 Set the link status monitoring polling period in milliseconds for a bonding device.
1744 This adds support for PMD slave devices which do not support link status interrupts.
1745 When the mon_period is set to a value greater than 0 then all PMD's which do not support
1746 link status ISR will be queried every polling interval to check if their link status has changed::
1748 testpmd> set bonding mon_period (port_id) (value)
1750 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
1752 testpmd> set bonding mon_period 5 150
1758 Show the current configuration of a Link Bonding device::
1760 testpmd> show bonding config (port id)
1763 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
1764 in balance mode with a transmission policy of layer 2+3::
1766 testpmd> show bonding config 9
1768 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
1770 Active Slaves (3): [1 3 4]
1777 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
1778 This is mainly useful for debugging purposes.
1779 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
1780 and fields that can be accessed.
1785 Display the value of a port register::
1787 testpmd> read reg (port_id) (address)
1789 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
1791 testpmd> read reg 0 0xEE00
1792 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
1797 Display a port register bit field::
1799 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
1801 For example, reading the lowest two bits from the register in the example above::
1803 testpmd> read regfield 0 0xEE00 0 1
1804 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
1809 Display a single port register bit::
1811 testpmd> read regbit (port_id) (address) (bit_x)
1813 For example, reading the lowest bit from the register in the example above::
1815 testpmd> read regbit 0 0xEE00 0
1816 port 0 PCI register at offset 0xEE00: bit 0=1
1821 Set the value of a port register::
1823 testpmd> write reg (port_id) (address) (value)
1825 For example, to clear a register::
1827 testpmd> write reg 0 0xEE00 0x0
1828 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
1833 Set bit field of a port register::
1835 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
1837 For example, writing to the register cleared in the example above::
1839 testpmd> write regfield 0 0xEE00 0 1 2
1840 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
1845 Set single bit value of a port register::
1847 testpmd> write regbit (port_id) (address) (bit_x) (value)
1849 For example, to set the high bit in the register from the example above::
1851 testpmd> write regbit 0 0xEE00 31 1
1852 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
1858 This section details the available filter functions that are available.
1860 Note these functions interface the deprecated legacy filtering framework,
1861 superseded by *rte_flow*. See `Flow rules management`_.
1864 ~~~~~~~~~~~~~~~~~~~~
1866 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
1868 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
1869 ethertype (ether_type) (drop|fwd) queue (queue_id)
1871 The available information parameters are:
1873 * ``port_id``: The port which the Ethertype filter assigned on.
1875 * ``mac_addr``: Compare destination mac address.
1877 * ``mac_ignr``: Ignore destination mac address match.
1879 * ``mac_address``: Destination mac address to match.
1881 * ``ether_type``: The EtherType value want to match,
1882 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
1884 * ``queue_id``: The receive queue associated with this EtherType filter.
1885 It is meaningless when deleting or dropping.
1887 Example, to add/remove an ethertype filter rule::
1889 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
1890 ethertype 0x0806 fwd queue 3
1892 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
1893 ethertype 0x0806 fwd queue 3
1898 Add or delete a 2-tuple filter,
1899 which identifies packets by specific protocol and destination TCP/UDP port
1900 and forwards packets into one of the receive queues::
1902 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
1903 protocol (protocol_value) mask (mask_value) \
1904 tcp_flags (tcp_flags_value) priority (prio_value) \
1907 The available information parameters are:
1909 * ``port_id``: The port which the 2-tuple filter assigned on.
1911 * ``dst_port_value``: Destination port in L4.
1913 * ``protocol_value``: IP L4 protocol.
1915 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
1917 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
1919 * ``prio_value``: Priority of this filter.
1921 * ``queue_id``: The receive queue associated with this 2-tuple filter.
1923 Example, to add/remove an 2tuple filter rule::
1925 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
1926 tcp_flags 0x02 priority 3 queue 3
1928 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
1929 tcp_flags 0x02 priority 3 queue 3
1934 Add or delete a 5-tuple filter,
1935 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
1936 and routes packets into one of the receive queues::
1938 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
1939 (src_address) dst_port (dst_port_value) \
1940 src_port (src_port_value) protocol (protocol_value) \
1941 mask (mask_value) tcp_flags (tcp_flags_value) \
1942 priority (prio_value) queue (queue_id)
1944 The available information parameters are:
1946 * ``port_id``: The port which the 5-tuple filter assigned on.
1948 * ``dst_address``: Destination IP address.
1950 * ``src_address``: Source IP address.
1952 * ``dst_port_value``: TCP/UDP destination port.
1954 * ``src_port_value``: TCP/UDP source port.
1956 * ``protocol_value``: L4 protocol.
1958 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
1960 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
1962 * ``prio_value``: The priority of this filter.
1964 * ``queue_id``: The receive queue associated with this 5-tuple filter.
1966 Example, to add/remove an 5tuple filter rule::
1968 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1969 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1970 flags 0x0 priority 3 queue 3
1972 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1973 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1974 flags 0x0 priority 3 queue 3
1979 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
1981 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
1983 The available information parameters are:
1985 * ``port_id``: The port which the SYN filter assigned on.
1987 * ``high``: This SYN filter has higher priority than other filters.
1989 * ``low``: This SYN filter has lower priority than other filters.
1991 * ``queue_id``: The receive queue associated with this SYN filter
1995 testpmd> syn_filter 0 add priority high queue 3
2000 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
2001 and routed into one of the receive queues::
2003 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
2004 mask (mask_value) priority (prio_value) queue (queue_id)
2006 The available information parameters are:
2008 * ``port_id``: The port which the Flex filter is assigned on.
2010 * ``len_value``: Filter length in bytes, no greater than 128.
2012 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
2014 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
2016 * ``prio_value``: The priority of this filter.
2018 * ``queue_id``: The receive queue associated with this Flex filter.
2022 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
2023 mask 000C priority 3 queue 3
2025 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
2026 mask 000C priority 3 queue 3
2029 .. _testpmd_flow_director:
2031 flow_director_filter
2032 ~~~~~~~~~~~~~~~~~~~~
2034 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
2036 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
2037 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
2039 * Perfect match filters.
2040 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2041 The masked fields are for IP flow.
2043 * Signature filters.
2044 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
2046 * Perfect-mac-vlan match filters.
2047 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2048 The masked fields are for MAC VLAN flow.
2050 * Perfect-tunnel match filters.
2051 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2052 The masked fields are for tunnel flow.
2054 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
2055 per flow type and the flexible payload.
2057 The Flow Director can also mask out parts of all of these fields so that filters
2058 are only applied to certain fields or parts of the fields.
2060 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
2062 # Commands to add flow director filters of different flow types::
2064 flow_director_filter (port_id) mode IP (add|del|update) \
2065 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
2066 src (src_ip_address) dst (dst_ip_address) \
2067 tos (tos_value) proto (proto_value) ttl (ttl_value) \
2068 vlan (vlan_value) flexbytes (flexbytes_value) \
2069 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
2072 flow_director_filter (port_id) mode IP (add|del|update) \
2073 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
2074 src (src_ip_address) (src_port) \
2075 dst (dst_ip_address) (dst_port) \
2076 tos (tos_value) ttl (ttl_value) \
2077 vlan (vlan_value) flexbytes (flexbytes_value) \
2078 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
2081 flow_director_filter (port_id) mode IP (add|del|update) \
2082 flow (ipv4-sctp|ipv6-sctp) \
2083 src (src_ip_address) (src_port) \
2084 dst (dst_ip_address) (dst_port) \
2085 tos (tos_value) ttl (ttl_value) \
2086 tag (verification_tag) vlan (vlan_value) \
2087 flexbytes (flexbytes_value) (drop|fwd) \
2088 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
2090 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
2091 ether (ethertype) flexbytes (flexbytes_value) \
2092 (drop|fwd) pf|vf(vf_id) queue (queue_id)
2095 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
2096 mac (mac_address) vlan (vlan_value) \
2097 flexbytes (flexbytes_value) (drop|fwd) \
2098 queue (queue_id) fd_id (fd_id_value)
2100 flow_director_filter (port_id) mode Tunnel (add|del|update) \
2101 mac (mac_address) vlan (vlan_value) \
2102 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
2103 flexbytes (flexbytes_value) (drop|fwd) \
2104 queue (queue_id) fd_id (fd_id_value)
2106 For example, to add an ipv4-udp flow type filter::
2108 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
2109 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
2110 fwd pf queue 1 fd_id 1
2112 For example, add an ipv4-other flow type filter::
2114 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
2115 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
2116 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
2121 Flush all flow director filters on a device::
2123 testpmd> flush_flow_director (port_id)
2125 Example, to flush all flow director filter on port 0::
2127 testpmd> flush_flow_director 0
2132 Set flow director's input masks::
2134 flow_director_mask (port_id) mode IP vlan (vlan_value) \
2135 src_mask (ipv4_src) (ipv6_src) (src_port) \
2136 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2138 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2140 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2141 mac (mac_value) tunnel-type (tunnel_type_value) \
2142 tunnel-id (tunnel_id_value)
2144 Example, to set flow director mask on port 0::
2146 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2147 src_mask 255.255.255.255 \
2148 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2149 dst_mask 255.255.255.255 \
2150 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2152 flow_director_flex_mask
2153 ~~~~~~~~~~~~~~~~~~~~~~~
2155 set masks of flow director's flexible payload based on certain flow type::
2157 testpmd> flow_director_flex_mask (port_id) \
2158 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2159 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2160 l2_payload|all) (mask)
2162 Example, to set flow director's flex mask for all flow type on port 0::
2164 testpmd> flow_director_flex_mask 0 flow all \
2165 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2168 flow_director_flex_payload
2169 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2171 Configure flexible payload selection::
2173 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2175 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2177 testpmd> flow_director_flex_payload 0 l4 \
2178 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2180 get_sym_hash_ena_per_port
2181 ~~~~~~~~~~~~~~~~~~~~~~~~~
2183 Get symmetric hash enable configuration per port::
2185 get_sym_hash_ena_per_port (port_id)
2187 For example, to get symmetric hash enable configuration of port 1::
2189 testpmd> get_sym_hash_ena_per_port 1
2191 set_sym_hash_ena_per_port
2192 ~~~~~~~~~~~~~~~~~~~~~~~~~
2194 Set symmetric hash enable configuration per port to enable or disable::
2196 set_sym_hash_ena_per_port (port_id) (enable|disable)
2198 For example, to set symmetric hash enable configuration of port 1 to enable::
2200 testpmd> set_sym_hash_ena_per_port 1 enable
2202 get_hash_global_config
2203 ~~~~~~~~~~~~~~~~~~~~~~
2205 Get the global configurations of hash filters::
2207 get_hash_global_config (port_id)
2209 For example, to get the global configurations of hash filters of port 1::
2211 testpmd> get_hash_global_config 1
2213 set_hash_global_config
2214 ~~~~~~~~~~~~~~~~~~~~~~
2216 Set the global configurations of hash filters::
2218 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2219 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2220 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload) \
2223 For example, to enable simple_xor for flow type of ipv6 on port 2::
2225 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2230 Set the input set for hash::
2232 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2233 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2234 l2_payload) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2235 ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2236 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2237 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2238 fld-8th|none) (select|add)
2240 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2242 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2247 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2248 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2250 Set the input set for flow director::
2252 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2253 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2254 l2_payload) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2255 ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2256 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2257 sctp-dst-port|sctp-veri-tag|none) (select|add)
2259 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2261 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2266 Set different GRE key length for input set::
2268 global_config (port_id) gre-key-len (number in bytes)
2270 For example to set GRE key length for input set to 4 bytes on port 0::
2272 testpmd> global_config 0 gre-key-len 4
2275 .. _testpmd_rte_flow:
2277 Flow rules management
2278 ---------------------
2280 Control of the generic flow API (*rte_flow*) is fully exposed through the
2281 ``flow`` command (validation, creation, destruction, queries and operation
2284 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2285 features simultaneously may cause undefined side-effects and is therefore
2291 Because the ``flow`` command uses dynamic tokens to handle the large number
2292 of possible flow rules combinations, its behavior differs slightly from
2293 other commands, in particular:
2295 - Pressing *?* or the *<tab>* key displays contextual help for the current
2296 token, not that of the entire command.
2298 - Optional and repeated parameters are supported (provided they are listed
2299 in the contextual help).
2301 The first parameter stands for the operation mode. Possible operations and
2302 their general syntax are described below. They are covered in detail in the
2305 - Check whether a flow rule can be created::
2307 flow validate {port_id}
2308 [group {group_id}] [priority {level}] [ingress] [egress]
2309 pattern {item} [/ {item} [...]] / end
2310 actions {action} [/ {action} [...]] / end
2312 - Create a flow rule::
2314 flow create {port_id}
2315 [group {group_id}] [priority {level}] [ingress] [egress]
2316 pattern {item} [/ {item} [...]] / end
2317 actions {action} [/ {action} [...]] / end
2319 - Destroy specific flow rules::
2321 flow destroy {port_id} rule {rule_id} [...]
2323 - Destroy all flow rules::
2325 flow flush {port_id}
2327 - Query an existing flow rule::
2329 flow query {port_id} {rule_id} {action}
2331 - List existing flow rules sorted by priority, filtered by group
2334 flow list {port_id} [group {group_id}] [...]
2336 - Restrict ingress traffic to the defined flow rules::
2338 flow isolate {port_id} {boolean}
2340 Validating flow rules
2341 ~~~~~~~~~~~~~~~~~~~~~
2343 ``flow validate`` reports whether a flow rule would be accepted by the
2344 underlying device in its current state but stops short of creating it. It is
2345 bound to ``rte_flow_validate()``::
2347 flow validate {port_id}
2348 [group {group_id}] [priority {level}] [ingress] [egress]
2349 pattern {item} [/ {item} [...]] / end
2350 actions {action} [/ {action} [...]] / end
2352 If successful, it will show::
2356 Otherwise it will show an error message of the form::
2358 Caught error type [...] ([...]): [...]
2360 This command uses the same parameters as ``flow create``, their format is
2361 described in `Creating flow rules`_.
2363 Check whether redirecting any Ethernet packet received on port 0 to RX queue
2364 index 6 is supported::
2366 testpmd> flow validate 0 ingress pattern eth / end
2367 actions queue index 6 / end
2371 Port 0 does not support TCPv6 rules::
2373 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
2375 Caught error type 9 (specific pattern item): Invalid argument
2381 ``flow create`` validates and creates the specified flow rule. It is bound
2382 to ``rte_flow_create()``::
2384 flow create {port_id}
2385 [group {group_id}] [priority {level}] [ingress] [egress]
2386 pattern {item} [/ {item} [...]] / end
2387 actions {action} [/ {action} [...]] / end
2389 If successful, it will return a flow rule ID usable with other commands::
2391 Flow rule #[...] created
2393 Otherwise it will show an error message of the form::
2395 Caught error type [...] ([...]): [...]
2397 Parameters describe in the following order:
2399 - Attributes (*group*, *priority*, *ingress*, *egress* tokens).
2400 - A matching pattern, starting with the *pattern* token and terminated by an
2402 - Actions, starting with the *actions* token and terminated by an *end*
2405 These translate directly to *rte_flow* objects provided as-is to the
2406 underlying functions.
2408 The shortest valid definition only comprises mandatory tokens::
2410 testpmd> flow create 0 pattern end actions end
2412 Note that PMDs may refuse rules that essentially do nothing such as this
2415 **All unspecified object values are automatically initialized to 0.**
2420 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
2421 specified before the ``pattern`` token.
2423 - ``group {group id}``: priority group.
2424 - ``priority {level}``: priority level within group.
2425 - ``ingress``: rule applies to ingress traffic.
2426 - ``egress``: rule applies to egress traffic.
2428 Each instance of an attribute specified several times overrides the previous
2429 value as shown below (group 4 is used)::
2431 testpmd> flow create 0 group 42 group 24 group 4 [...]
2433 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
2435 While not specifying a direction is an error, some rules may allow both
2438 Most rules affect RX therefore contain the ``ingress`` token::
2440 testpmd> flow create 0 ingress pattern [...]
2445 A matching pattern starts after the ``pattern`` token. It is made of pattern
2446 items and is terminated by a mandatory ``end`` item.
2448 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
2449 rte_flow_item_type``).
2451 The ``/`` token is used as a separator between pattern items as shown
2454 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
2456 Note that protocol items like these must be stacked from lowest to highest
2457 layer to make sense. For instance, the following rule is either invalid or
2458 unlikely to match any packet::
2460 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
2462 More information on these restrictions can be found in the *rte_flow*
2465 Several items support additional specification structures, for example
2466 ``ipv4`` allows specifying source and destination addresses as follows::
2468 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
2469 dst is 10.2.0.0 / end [...]
2471 This rule matches all IPv4 traffic with the specified properties.
2473 In this example, ``src`` and ``dst`` are field names of the underlying
2474 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
2475 in a similar fashion.
2477 The ``is`` token means that the subsequent value must be matched exactly,
2478 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
2479 accordingly. Possible assignment tokens are:
2481 - ``is``: match value perfectly (with full bit-mask).
2482 - ``spec``: match value according to configured bit-mask.
2483 - ``last``: specify upper bound to establish a range.
2484 - ``mask``: specify bit-mask with relevant bits set to one.
2485 - ``prefix``: generate bit-mask from a prefix length.
2487 These yield identical results::
2489 ipv4 src is 10.1.1.1
2493 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
2497 ipv4 src spec 10.1.1.1 src prefix 32
2501 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
2505 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
2507 Inclusive ranges can be defined with ``last``::
2509 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
2511 Note that ``mask`` affects both ``spec`` and ``last``::
2513 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
2514 # matches 10.1.0.0 to 10.2.255.255
2516 Properties can be modified multiple times::
2518 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
2522 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
2527 This section lists supported pattern items and their attributes, if any.
2529 - ``end``: end list of pattern items.
2531 - ``void``: no-op pattern item.
2533 - ``invert``: perform actions when pattern does not match.
2535 - ``any``: match any protocol for the current layer.
2537 - ``num {unsigned}``: number of layers covered.
2539 - ``pf``: match packets addressed to the physical function.
2541 - ``vf``: match packets addressed to a virtual function ID.
2543 - ``id {unsigned}``: destination VF ID.
2545 - ``port``: device-specific physical port index to use.
2547 - ``index {unsigned}``: physical port index.
2549 - ``raw``: match an arbitrary byte string.
2551 - ``relative {boolean}``: look for pattern after the previous item.
2552 - ``search {boolean}``: search pattern from offset (see also limit).
2553 - ``offset {integer}``: absolute or relative offset for pattern.
2554 - ``limit {unsigned}``: search area limit for start of pattern.
2555 - ``pattern {string}``: byte string to look for.
2557 - ``eth``: match Ethernet header.
2559 - ``dst {MAC-48}``: destination MAC.
2560 - ``src {MAC-48}``: source MAC.
2561 - ``type {unsigned}``: EtherType.
2563 - ``vlan``: match 802.1Q/ad VLAN tag.
2565 - ``tpid {unsigned}``: tag protocol identifier.
2566 - ``tci {unsigned}``: tag control information.
2567 - ``pcp {unsigned}``: priority code point.
2568 - ``dei {unsigned}``: drop eligible indicator.
2569 - ``vid {unsigned}``: VLAN identifier.
2571 - ``ipv4``: match IPv4 header.
2573 - ``tos {unsigned}``: type of service.
2574 - ``ttl {unsigned}``: time to live.
2575 - ``proto {unsigned}``: next protocol ID.
2576 - ``src {ipv4 address}``: source address.
2577 - ``dst {ipv4 address}``: destination address.
2579 - ``ipv6``: match IPv6 header.
2581 - ``tc {unsigned}``: traffic class.
2582 - ``flow {unsigned}``: flow label.
2583 - ``proto {unsigned}``: protocol (next header).
2584 - ``hop {unsigned}``: hop limit.
2585 - ``src {ipv6 address}``: source address.
2586 - ``dst {ipv6 address}``: destination address.
2588 - ``icmp``: match ICMP header.
2590 - ``type {unsigned}``: ICMP packet type.
2591 - ``code {unsigned}``: ICMP packet code.
2593 - ``udp``: match UDP header.
2595 - ``src {unsigned}``: UDP source port.
2596 - ``dst {unsigned}``: UDP destination port.
2598 - ``tcp``: match TCP header.
2600 - ``src {unsigned}``: TCP source port.
2601 - ``dst {unsigned}``: TCP destination port.
2603 - ``sctp``: match SCTP header.
2605 - ``src {unsigned}``: SCTP source port.
2606 - ``dst {unsigned}``: SCTP destination port.
2607 - ``tag {unsigned}``: validation tag.
2608 - ``cksum {unsigned}``: checksum.
2610 - ``vxlan``: match VXLAN header.
2612 - ``vni {unsigned}``: VXLAN identifier.
2614 - ``e_tag``: match IEEE 802.1BR E-Tag header.
2616 - ``grp_ecid_b {unsigned}``: GRP and E-CID base.
2618 - ``nvgre``: match NVGRE header.
2620 - ``tni {unsigned}``: virtual subnet ID.
2622 - ``mpls``: match MPLS header.
2624 - ``label {unsigned}``: MPLS label.
2626 - ``gre``: match GRE header.
2628 - ``protocol {unsigned}``: protocol type.
2630 - ``fuzzy``: fuzzy pattern match, expect faster than default.
2632 - ``thresh {unsigned}``: accuracy threshold.
2637 A list of actions starts after the ``actions`` token in the same fashion as
2638 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
2639 terminated by a mandatory ``end`` action.
2641 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
2642 rte_flow_action_type``).
2644 Dropping all incoming UDPv4 packets can be expressed as follows::
2646 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2649 Several actions have configurable properties which must be specified when
2650 there is no valid default value. For example, ``queue`` requires a target
2653 This rule redirects incoming UDPv4 traffic to queue index 6::
2655 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2656 actions queue index 6 / end
2658 While this one could be rejected by PMDs (unspecified queue index)::
2660 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2663 As defined by *rte_flow*, the list is not ordered, all actions of a given
2664 rule are performed simultaneously. These are equivalent::
2666 queue index 6 / void / mark id 42 / end
2670 void / mark id 42 / queue index 6 / end
2672 All actions in a list should have different types, otherwise only the last
2673 action of a given type is taken into account::
2675 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
2679 drop / drop / drop / end # drop is performed only once
2683 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
2685 Considering they are performed simultaneously, opposite and overlapping
2686 actions can sometimes be combined when the end result is unambiguous::
2688 drop / queue index 6 / end # drop has no effect
2692 drop / dup index 6 / end # same as above
2696 queue index 6 / rss queues 6 7 8 / end # queue has no effect
2700 drop / passthru / end # drop has no effect
2702 Note that PMDs may still refuse such combinations.
2707 This section lists supported actions and their attributes, if any.
2709 - ``end``: end list of actions.
2711 - ``void``: no-op action.
2713 - ``passthru``: let subsequent rule process matched packets.
2715 - ``mark``: attach 32 bit value to packets.
2717 - ``id {unsigned}``: 32 bit value to return with packets.
2719 - ``flag``: flag packets.
2721 - ``queue``: assign packets to a given queue index.
2723 - ``index {unsigned}``: queue index to use.
2725 - ``drop``: drop packets (note: passthru has priority).
2727 - ``count``: enable counters for this rule.
2729 - ``dup``: duplicate packets to a given queue index.
2731 - ``index {unsigned}``: queue index to duplicate packets to.
2733 - ``rss``: spread packets among several queues.
2735 - ``queues [{unsigned} [...]] end``: queue indices to use.
2737 - ``pf``: redirect packets to physical device function.
2739 - ``vf``: redirect packets to virtual device function.
2741 - ``original {boolean}``: use original VF ID if possible.
2742 - ``id {unsigned}``: VF ID to redirect packets to.
2744 Destroying flow rules
2745 ~~~~~~~~~~~~~~~~~~~~~
2747 ``flow destroy`` destroys one or more rules from their rule ID (as returned
2748 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
2749 times as necessary::
2751 flow destroy {port_id} rule {rule_id} [...]
2753 If successful, it will show::
2755 Flow rule #[...] destroyed
2757 It does not report anything for rule IDs that do not exist. The usual error
2758 message is shown when a rule cannot be destroyed::
2760 Caught error type [...] ([...]): [...]
2762 ``flow flush`` destroys all rules on a device and does not take extra
2763 arguments. It is bound to ``rte_flow_flush()``::
2765 flow flush {port_id}
2767 Any errors are reported as above.
2769 Creating several rules and destroying them::
2771 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2772 actions queue index 2 / end
2773 Flow rule #0 created
2774 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2775 actions queue index 3 / end
2776 Flow rule #1 created
2777 testpmd> flow destroy 0 rule 0 rule 1
2778 Flow rule #1 destroyed
2779 Flow rule #0 destroyed
2782 The same result can be achieved using ``flow flush``::
2784 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2785 actions queue index 2 / end
2786 Flow rule #0 created
2787 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2788 actions queue index 3 / end
2789 Flow rule #1 created
2790 testpmd> flow flush 0
2793 Non-existent rule IDs are ignored::
2795 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2796 actions queue index 2 / end
2797 Flow rule #0 created
2798 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2799 actions queue index 3 / end
2800 Flow rule #1 created
2801 testpmd> flow destroy 0 rule 42 rule 10 rule 2
2803 testpmd> flow destroy 0 rule 0
2804 Flow rule #0 destroyed
2810 ``flow query`` queries a specific action of a flow rule having that
2811 ability. Such actions collect information that can be reported using this
2812 command. It is bound to ``rte_flow_query()``::
2814 flow query {port_id} {rule_id} {action}
2816 If successful, it will display either the retrieved data for known actions
2817 or the following message::
2819 Cannot display result for action type [...] ([...])
2821 Otherwise, it will complain either that the rule does not exist or that some
2824 Flow rule #[...] not found
2828 Caught error type [...] ([...]): [...]
2830 Currently only the ``count`` action is supported. This action reports the
2831 number of packets that hit the flow rule and the total number of bytes. Its
2832 output has the following format::
2835 hits_set: [...] # whether "hits" contains a valid value
2836 bytes_set: [...] # whether "bytes" contains a valid value
2837 hits: [...] # number of packets
2838 bytes: [...] # number of bytes
2840 Querying counters for TCPv6 packets redirected to queue 6::
2842 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
2843 actions queue index 6 / count / end
2844 Flow rule #4 created
2845 testpmd> flow query 0 4 count
2856 ``flow list`` lists existing flow rules sorted by priority and optionally
2857 filtered by group identifiers::
2859 flow list {port_id} [group {group_id}] [...]
2861 This command only fails with the following message if the device does not
2866 Output consists of a header line followed by a short description of each
2867 flow rule, one per line. There is no output at all when no flow rules are
2868 configured on the device::
2870 ID Group Prio Attr Rule
2871 [...] [...] [...] [...] [...]
2873 ``Attr`` column flags:
2875 - ``i`` for ``ingress``.
2876 - ``e`` for ``egress``.
2878 Creating several flow rules and listing them::
2880 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2881 actions queue index 6 / end
2882 Flow rule #0 created
2883 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2884 actions queue index 2 / end
2885 Flow rule #1 created
2886 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
2887 actions rss queues 6 7 8 end / end
2888 Flow rule #2 created
2889 testpmd> flow list 0
2890 ID Group Prio Attr Rule
2891 0 0 0 i- ETH IPV4 => QUEUE
2892 1 0 0 i- ETH IPV6 => QUEUE
2893 2 0 5 i- ETH IPV4 UDP => RSS
2896 Rules are sorted by priority (i.e. group ID first, then priority level)::
2898 testpmd> flow list 1
2899 ID Group Prio Attr Rule
2900 0 0 0 i- ETH => COUNT
2901 6 0 500 i- ETH IPV6 TCP => DROP COUNT
2902 5 0 1000 i- ETH IPV6 ICMP => QUEUE
2903 1 24 0 i- ETH IPV4 UDP => QUEUE
2904 4 24 10 i- ETH IPV4 TCP => DROP
2905 3 24 20 i- ETH IPV4 => DROP
2906 2 24 42 i- ETH IPV4 UDP => QUEUE
2907 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
2910 Output can be limited to specific groups::
2912 testpmd> flow list 1 group 0 group 63
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 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
2920 Toggling isolated mode
2921 ~~~~~~~~~~~~~~~~~~~~~~
2923 ``flow isolate`` can be used to tell the underlying PMD that ingress traffic
2924 must only be injected from the defined flow rules; that no default traffic
2925 is expected outside those rules and the driver is free to assign more
2926 resources to handle them. It is bound to ``rte_flow_isolate()``::
2928 flow isolate {port_id} {boolean}
2930 If successful, enabling or disabling isolated mode shows either::
2932 Ingress traffic on port [...]
2933 is now restricted to the defined flow rules
2937 Ingress traffic on port [...]
2938 is not restricted anymore to the defined flow rules
2940 Otherwise, in case of error::
2942 Caught error type [...] ([...]): [...]
2944 Mainly due to its side effects, PMDs supporting this mode may not have the
2945 ability to toggle it more than once without reinitializing affected ports
2946 first (e.g. by exiting testpmd).
2948 Enabling isolated mode::
2950 testpmd> flow isolate 0 true
2951 Ingress traffic on port 0 is now restricted to the defined flow rules
2954 Disabling isolated mode::
2956 testpmd> flow isolate 0 false
2957 Ingress traffic on port 0 is not restricted anymore to the defined flow rules
2960 Sample QinQ flow rules
2961 ~~~~~~~~~~~~~~~~~~~~~~
2963 Before creating QinQ rule(s) the following commands should be issued to enable QinQ::
2965 testpmd> port stop 0
2966 testpmd> vlan set qinq on 0
2968 The above command sets the inner and outer TPID's to 0x8100.
2970 To change the TPID's the following commands should be used::
2972 testpmd> vlan set outer tpid 0xa100 0
2973 testpmd> vlan set inner tpid 0x9100 0
2974 testpmd> port start 0
2976 Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM.
2980 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 /
2981 vlan tci is 456 / end actions vf id 1 / queue index 0 / end
2982 Flow rule #0 validated
2984 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 /
2985 vlan tci is 456 / end actions vf id 123 / queue index 0 / end
2986 Flow rule #0 created
2988 testpmd> flow list 0
2989 ID Group Prio Attr Rule
2990 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
2992 Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host.
2996 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 /
2997 vlan tci is 654 / end actions pf / queue index 0 / end
2998 Flow rule #1 validated
3000 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 /
3001 vlan tci is 654 / end actions pf / queue index 1 / end
3002 Flow rule #1 created
3004 testpmd> flow list 0
3005 ID Group Prio Attr Rule
3006 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
3007 1 0 0 i- ETH VLAN VLAN=>PF QUEUE