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 In all cases output from any included command will be displayed as standard output.
123 Execution will continue until the end of the file is reached regardless of
124 whether any errors occur. The end user must examine the output to determine if
125 any failures occurred.
134 Start packet forwarding with current configuration::
141 Start packet forwarding with current configuration after sending specified number of bursts of packets::
143 testpmd> start tx_first (""|burst_num)
145 The default burst number is 1 when ``burst_num`` not presented.
150 Stop packet forwarding, and display accumulated statistics::
165 The functions in the following sections are used to display information about the
166 testpmd configuration or the NIC status.
171 Display information for a given port or all ports::
173 testpmd> show port (info|stats|xstats|fdir|stat_qmap|dcb_tc|cap) (port_id|all)
175 The available information categories are:
177 * ``info``: General port information such as MAC address.
179 * ``stats``: RX/TX statistics.
181 * ``xstats``: RX/TX extended NIC statistics.
183 * ``fdir``: Flow Director information and statistics.
185 * ``stat_qmap``: Queue statistics mapping.
187 * ``dcb_tc``: DCB information such as TC mapping.
189 * ``cap``: Supported offload capabilities.
193 .. code-block:: console
195 testpmd> show port info 0
197 ********************* Infos for port 0 *********************
199 MAC address: XX:XX:XX:XX:XX:XX
201 memory allocation on the socket: 0
203 Link speed: 40000 Mbps
204 Link duplex: full-duplex
205 Promiscuous mode: enabled
206 Allmulticast mode: disabled
207 Maximum number of MAC addresses: 64
208 Maximum number of MAC addresses of hash filtering: 0
213 Redirection table size: 512
214 Supported flow types:
234 Display the rss redirection table entry indicated by masks on port X::
236 testpmd> show port (port_id) rss reta (size) (mask0, mask1...)
238 size is used to indicate the hardware supported reta size
243 Display the RSS hash functions and RSS hash key of a port::
245 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]
250 Clear the port statistics for a given port or for all ports::
252 testpmd> clear port (info|stats|xstats|fdir|stat_qmap) (port_id|all)
256 testpmd> clear port stats all
261 Display information for a given port's RX/TX queue::
263 testpmd> show (rxq|txq) info (port_id) (queue_id)
268 Displays the configuration of the application.
269 The configuration comes from the command-line, the runtime or the application defaults::
271 testpmd> show config (rxtx|cores|fwd|txpkts)
273 The available information categories are:
275 * ``rxtx``: RX/TX configuration items.
277 * ``cores``: List of forwarding cores.
279 * ``fwd``: Packet forwarding configuration.
281 * ``txpkts``: Packets to TX configuration.
285 .. code-block:: console
287 testpmd> show config rxtx
289 io packet forwarding - CRC stripping disabled - packets/burst=16
290 nb forwarding cores=2 - nb forwarding ports=1
291 RX queues=1 - RX desc=128 - RX free threshold=0
292 RX threshold registers: pthresh=8 hthresh=8 wthresh=4
293 TX queues=1 - TX desc=512 - TX free threshold=0
294 TX threshold registers: pthresh=36 hthresh=0 wthresh=0
295 TX RS bit threshold=0 - TXQ flags=0x0
300 Set the packet forwarding mode::
302 testpmd> set fwd (io|mac|macswap|flowgen| \
303 rxonly|txonly|csum|icmpecho) (""|retry)
305 ``retry`` can be specified for forwarding engines except ``rx_only``.
307 The available information categories are:
309 * ``io``: Forwards packets "as-is" in I/O mode.
310 This is the fastest possible forwarding operation as it does not access packets data.
311 This is the default mode.
313 * ``mac``: Changes the source and the destination Ethernet addresses of packets before forwarding them.
314 Default application behaviour is to set source Ethernet address to that of the transmitting interface, and destination
315 address to a dummy value (set during init). The user may specify a target destination Ethernet address via the 'eth-peer' or
316 'eth-peer-configfile' command-line options. It is not currently possible to specify a specific source Ethernet address.
318 * ``macswap``: MAC swap forwarding mode.
319 Swaps the source and the destination Ethernet addresses of packets before forwarding them.
321 * ``flowgen``: Multi-flow generation mode.
322 Originates a number of flows (with varying destination IP addresses), and terminate receive traffic.
324 * ``rxonly``: Receives packets but doesn't transmit them.
326 * ``txonly``: Generates and transmits packets without receiving any.
328 * ``csum``: Changes the checksum field with hardware or software methods depending on the offload flags on the packet.
330 * ``icmpecho``: Receives a burst of packets, lookup for IMCP echo requests and, if any, send back ICMP echo replies.
332 * ``ieee1588``: Demonstrate L2 IEEE1588 V2 PTP timestamping for RX and TX. Requires ``CONFIG_RTE_LIBRTE_IEEE1588=y``.
334 Note: TX timestamping is only available in the "Full Featured" TX path. To force ``testpmd`` into this mode set ``--txqflags=0``.
338 testpmd> set fwd rxonly
340 Set rxonly packet forwarding mode
346 Display an RX descriptor for a port RX queue::
348 testpmd> read rxd (port_id) (queue_id) (rxd_id)
352 testpmd> read rxd 0 0 4
353 0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180
358 Display a TX descriptor for a port TX queue::
360 testpmd> read txd (port_id) (queue_id) (txd_id)
364 testpmd> read txd 0 0 4
365 0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C
370 Display VF statistics::
372 testpmd> show vf stats (port_id) (vf_id)
377 Reset VF statistics::
379 testpmd> clear vf stats (port_id) (vf_id)
381 Configuration Functions
382 -----------------------
384 The testpmd application can be configured from the runtime as well as from the command-line.
386 This section details the available configuration functions that are available.
390 Configuration changes only become active when forwarding is started/restarted.
395 Reset forwarding to the default configuration::
402 Set the debug verbosity level::
404 testpmd> set verbose (level)
406 Currently the only available levels are 0 (silent except for error) and 1 (fully verbose).
411 Set the number of ports used by the application:
415 This is equivalent to the ``--nb-ports`` command-line option.
420 Set the number of cores used by the application::
422 testpmd> set nbcore (num)
424 This is equivalent to the ``--nb-cores`` command-line option.
428 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port.
433 Set the forwarding cores hexadecimal mask::
435 testpmd> set coremask (mask)
437 This is equivalent to the ``--coremask`` command-line option.
441 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding.
446 Set the forwarding ports hexadecimal mask::
448 testpmd> set portmask (mask)
450 This is equivalent to the ``--portmask`` command-line option.
455 Set number of packets per burst::
457 testpmd> set burst (num)
459 This is equivalent to the ``--burst command-line`` option.
461 When retry is enabled, the transmit delay time and number of retries can also be set::
463 testpmd> set burst tx delay (microseconds) retry (num)
468 Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode::
470 testpmd> set txpkts (x[,y]*)
472 Where x[,y]* represents a CSV list of values, without white space.
477 Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes::
479 testpmd> set txsplit (off|on|rand)
483 * ``off`` disable packet copy & split for CSUM mode.
485 * ``on`` split outgoing packet into multiple segments. Size of each segment
486 and number of segments per packet is determined by ``set txpkts`` command
489 * ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments.
494 Set the list of forwarding cores::
496 testpmd> set corelist (x[,y]*)
498 For example, to change the forwarding cores:
500 .. code-block:: console
502 testpmd> set corelist 3,1
503 testpmd> show config fwd
505 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled
506 Logical Core 3 (socket 0) forwards packets on 1 streams:
507 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01
508 Logical Core 1 (socket 0) forwards packets on 1 streams:
509 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
513 The cores are used in the same order as specified on the command line.
518 Set the list of forwarding ports::
520 testpmd> set portlist (x[,y]*)
522 For example, to change the port forwarding:
524 .. code-block:: console
526 testpmd> set portlist 0,2,1,3
527 testpmd> show config fwd
529 io packet forwarding - ports=4 - cores=1 - streams=4
530 Logical Core 3 (socket 0) forwards packets on 4 streams:
531 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01
532 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00
533 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03
534 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02
539 Enable/disable tx loopback::
541 testpmd> set tx loopback (port_id) (on|off)
546 set drop enable bit for all queues::
548 testpmd> set all queues drop (port_id) (on|off)
550 set split drop enable (for VF)
551 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
553 set split drop enable bit for VF from PF::
555 testpmd> set vf split drop (port_id) (vf_id) (on|off)
557 set mac antispoof (for VF)
558 ~~~~~~~~~~~~~~~~~~~~~~~~~~
560 Set mac antispoof for a VF from the PF::
562 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off)
567 Enable/disable MACsec offload::
569 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off)
570 testpmd> set macsec offload (port_id) off
575 Configure MACsec secure connection (SC)::
577 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi)
581 The pi argument is ignored for tx.
582 Check the NIC Datasheet for hardware limits.
587 Configure MACsec secure association (SA)::
589 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key)
593 The IDX value must be 0 or 1.
594 Check the NIC Datasheet for hardware limits.
596 set broadcast mode (for VF)
597 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
599 Set broadcast mode for a VF from the PF::
601 testpmd> set vf broadcast (port_id) (vf_id) (on|off)
606 Set the VLAN strip on a port::
608 testpmd> vlan set strip (on|off) (port_id)
613 Set the VLAN strip for a queue on a port::
615 testpmd> vlan set stripq (on|off) (port_id,queue_id)
617 vlan set stripq (for VF)
618 ~~~~~~~~~~~~~~~~~~~~~~~~
620 Set VLAN strip for all queues in a pool for a VF from the PF::
622 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off)
624 vlan set insert (for VF)
625 ~~~~~~~~~~~~~~~~~~~~~~~~
627 Set VLAN insert for a VF from the PF::
629 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id)
631 vlan set tag (for VF)
632 ~~~~~~~~~~~~~~~~~~~~~
634 Set VLAN tag for a VF from the PF::
636 testpmd> set vf vlan tag (port_id) (vf_id) (on|off)
638 vlan set antispoof (for VF)
639 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
641 Set VLAN antispoof for a VF from the PF::
643 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off)
648 Set the VLAN filter on a port::
650 testpmd> vlan set filter (on|off) (port_id)
655 Set the VLAN QinQ (extended queue in queue) on for a port::
657 testpmd> vlan set qinq (on|off) (port_id)
662 Set the inner or outer VLAN TPID for packet filtering on a port::
664 testpmd> vlan set (inner|outer) tpid (value) (port_id)
668 TPID value must be a 16-bit number (value <= 65536).
673 Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID::
675 testpmd> rx_vlan add (vlan_id|all) (port_id)
679 VLAN filter must be set on that port. VLAN ID < 4096.
680 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries
681 in VFTA table. This is important if enabling all vlan_ids.
686 Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID::
688 testpmd> rx_vlan rm (vlan_id|all) (port_id)
693 Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID::
695 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask)
700 Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID::
702 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask)
707 Add a tunnel filter on a port::
709 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \
710 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
711 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
713 The available information categories are:
715 * ``vxlan``: Set tunnel type as VXLAN.
717 * ``nvgre``: Set tunnel type as NVGRE.
719 * ``ipingre``: Set tunnel type as IP-in-GRE.
721 * ``imac-ivlan``: Set filter type as Inner MAC and VLAN.
723 * ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID.
725 * ``imac-tenid``: Set filter type as Inner MAC and tenant ID.
727 * ``imac``: Set filter type as Inner MAC.
729 * ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID.
731 * ``oip``: Set filter type as Outer IP.
733 * ``iip``: Set filter type as Inner IP.
737 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \
738 192.168.2.2 0 ipingre oip 1 1
740 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP.
745 Remove a tunnel filter on a port::
747 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \
748 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\
749 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id)
754 Add an UDP port for VXLAN packet filter on a port::
756 testpmd> rx_vxlan_port add (udp_port) (port_id)
761 Remove an UDP port for VXLAN packet filter on a port::
763 testpmd> rx_vxlan_port rm (udp_port) (port_id)
768 Set hardware insertion of VLAN IDs in packets sent on a port::
770 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer]
772 For example, set a single VLAN ID (5) insertion on port 0::
776 Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1::
784 Set port based hardware insertion of VLAN ID in packets sent on a port::
786 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off)
791 Disable hardware insertion of a VLAN header in packets sent on a port::
793 testpmd> tx_vlan reset (port_id)
798 Select hardware or software calculation of the checksum when
799 transmitting a packet using the ``csum`` forwarding engine::
801 testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id)
805 * ``ip|udp|tcp|sctp`` always relate to the inner layer.
807 * ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized
808 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are
809 supported). See also the ``csum parse-tunnel`` command.
813 Check the NIC Datasheet for hardware limits.
818 Define how tunneled packets should be handled by the csum forward
821 testpmd> csum parse-tunnel (on|off) (tx_port_id)
823 If enabled, the csum forward engine will try to recognize supported
824 tunnel headers (vxlan, gre, ipip).
826 If disabled, treat tunnel packets as non-tunneled packets (a inner
827 header is handled as a packet payload).
831 The port argument is the TX port like in the ``csum set`` command.
835 Consider a packet in packet like the following::
837 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in
839 * If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
840 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the
841 ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``).
843 * If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set``
844 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``.
849 Display tx checksum offload configuration::
851 testpmd> csum show (port_id)
856 Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine::
858 testpmd> tso set (segsize) (port_id)
862 Check the NIC datasheet for hardware limits.
867 Display the status of TCP Segmentation Offload::
869 testpmd> tso show (port_id)
874 Add an alternative MAC address to a port::
876 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX)
881 Remove a MAC address from a port::
883 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX)
885 mac_addr add (for VF)
886 ~~~~~~~~~~~~~~~~~~~~~
888 Add an alternative MAC address for a VF to a port::
890 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX)
895 Set the default MAC address for a port::
897 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX)
899 mac_addr set (for VF)
900 ~~~~~~~~~~~~~~~~~~~~~
902 Set the MAC address for a VF from the PF::
904 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX)
909 Set the unicast hash filter(s) on/off for a port::
911 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off)
916 Set the promiscuous mode on for a port or for all ports.
917 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
919 testpmd> set promisc (port_id|all) (on|off)
924 Set the allmulti mode for a port or for all ports::
926 testpmd> set allmulti (port_id|all) (on|off)
928 Same as the ifconfig (8) option. Controls how multicast packets are handled.
933 Set the unicast promiscuous mode for a VF from PF.
934 It's supported by Intel i40e NICs now.
935 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
937 testpmd> set vf promisc (port_id) (vf_id) (on|off)
939 set allmulticast (for VF)
940 ~~~~~~~~~~~~~~~~~~~~~~~~~
942 Set the multicast promiscuous mode for a VF from PF.
943 It's supported by Intel i40e NICs now.
944 In promiscuous mode packets are not dropped if they aren't for the specified MAC address::
946 testpmd> set vf allmulti (port_id) (vf_id) (on|off)
948 set tx max bandwidth (for VF)
949 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
951 Set TX max absolute bandwidth (Mbps) for a VF from PF::
953 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth)
955 set tc tx min bandwidth (for VF)
956 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
958 Set all TCs' TX min relative bandwidth (%) for a VF from PF::
960 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...)
962 set tc tx max bandwidth (for VF)
963 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
965 Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF::
967 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth)
969 set tc strict link priority mode
970 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
972 Set some TCs' strict link priority mode on a physical port::
974 testpmd> set tx strict-link-priority (port_id) (tc_bitmap)
976 set tc tx min bandwidth
977 ~~~~~~~~~~~~~~~~~~~~~~~
979 Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs::
981 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...)
986 Set the link flow control parameter on a port::
988 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
989 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \
990 autoneg (on|off) (port_id)
994 * ``high_water`` (integer): High threshold value to trigger XOFF.
996 * ``low_water`` (integer): Low threshold value to trigger XON.
998 * ``pause_time`` (integer): Pause quota in the Pause frame.
1000 * ``send_xon`` (0/1): Send XON frame.
1002 * ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames.
1004 * ``autoneg``: Change the auto-negotiation parameter.
1009 Set the priority flow control parameter on a port::
1011 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \
1012 (pause_time) (priority) (port_id)
1016 * ``high_water`` (integer): High threshold value.
1018 * ``low_water`` (integer): Low threshold value.
1020 * ``pause_time`` (integer): Pause quota in the Pause frame.
1022 * ``priority`` (0-7): VLAN User Priority.
1027 Set statistics mapping (qmapping 0..15) for RX/TX queue on port::
1029 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping)
1031 For example, to set rx queue 2 on port 0 to mapping 5::
1033 testpmd>set stat_qmap rx 0 2 5
1035 set port - rx/tx (for VF)
1036 ~~~~~~~~~~~~~~~~~~~~~~~~~
1038 Set VF receive/transmit from a port::
1040 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off)
1042 set port - mac address filter (for VF)
1043 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1045 Add/Remove unicast or multicast MAC addr filter for a VF::
1047 testpmd> set port (port_id) vf (vf_id) (mac_addr) \
1048 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off)
1050 set port - rx mode(for VF)
1051 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1053 Set the VF receive mode of a port::
1055 testpmd> set port (port_id) vf (vf_id) \
1056 rxmode (AUPE|ROPE|BAM|MPE) (on|off)
1058 The available receive modes are:
1060 * ``AUPE``: Accepts untagged VLAN.
1062 * ``ROPE``: Accepts unicast hash.
1064 * ``BAM``: Accepts broadcast packets.
1066 * ``MPE``: Accepts all multicast packets.
1068 set port - tx_rate (for Queue)
1069 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1071 Set TX rate limitation for a queue on a port::
1073 testpmd> set port (port_id) queue (queue_id) rate (rate_value)
1075 set port - tx_rate (for VF)
1076 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
1078 Set TX rate limitation for queues in VF on a port::
1080 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask)
1082 set port - mirror rule
1083 ~~~~~~~~~~~~~~~~~~~~~~
1085 Set pool or vlan type mirror rule for a port::
1087 testpmd> set port (port_id) mirror-rule (rule_id) \
1088 (pool-mirror-up|pool-mirror-down|vlan-mirror) \
1089 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off)
1091 Set link mirror rule for a port::
1093 testpmd> set port (port_id) mirror-rule (rule_id) \
1094 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off)
1096 For example to enable mirror traffic with vlan 0,1 to pool 0::
1098 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on
1100 reset port - mirror rule
1101 ~~~~~~~~~~~~~~~~~~~~~~~~
1103 Reset a mirror rule for a port::
1105 testpmd> reset port (port_id) mirror-rule (rule_id)
1110 Set the flush on RX streams before forwarding.
1111 The default is flush ``on``.
1112 Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams::
1114 testpmd> set flush_rx off
1119 Set the bypass mode for the lowest port on bypass enabled NIC::
1121 testpmd> set bypass mode (normal|bypass|isolate) (port_id)
1126 Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled::
1128 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \
1129 mode (normal|bypass|isolate) (port_id)
1133 * ``timeout``: Enable bypass after watchdog timeout.
1135 * ``os_on``: Enable bypass when OS/board is powered on.
1137 * ``os_off``: Enable bypass when OS/board is powered off.
1139 * ``power_on``: Enable bypass when power supply is turned on.
1141 * ``power_off``: Enable bypass when power supply is turned off.
1147 Set the bypass watchdog timeout to ``n`` seconds where 0 = instant::
1149 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32)
1154 Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC::
1156 testpmd> show bypass config (port_id)
1161 Set link up for a port::
1163 testpmd> set link-up port (port id)
1168 Set link down for a port::
1170 testpmd> set link-down port (port id)
1175 Enable E-tag insertion for a VF on a port::
1177 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id)
1179 Disable E-tag insertion for a VF on a port::
1181 testpmd> E-tag set insertion off port (port_id) vf (vf_id)
1183 Enable/disable E-tag stripping on a port::
1185 testpmd> E-tag set stripping (on|off) port (port_id)
1187 Enable/disable E-tag based forwarding on a port::
1189 testpmd> E-tag set forwarding (on|off) port (port_id)
1191 Add an E-tag forwarding filter on a port::
1193 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id)
1195 Delete an E-tag forwarding filter on a port::
1196 testpmd> E-tag set filter del e-tag-id (value) port (port_id)
1201 List all items from the ptype mapping table::
1203 testpmd> ptype mapping get (port_id) (valid_only)
1207 * ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0).
1209 Replace a specific or a group of software defined ptype with a new one::
1211 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type)
1215 * ``target``: A specific software ptype or a mask to represent a group of software ptypes.
1217 * ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1).
1219 * ``pkt_type``: The new software ptype to replace the old ones.
1221 Update hardware defined ptype to software defined packet type mapping table::
1223 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype)
1227 * ``hw_ptype``: hardware ptype as the index of the ptype mapping table.
1229 * ``sw_ptype``: software ptype as the value of the ptype mapping table.
1231 Reset ptype mapping table::
1233 testpmd> ptype mapping reset (port_id)
1238 The following sections show functions for configuring ports.
1242 Port configuration changes only become active when forwarding is started/restarted.
1247 Attach a port specified by pci address or virtual device args::
1249 testpmd> port attach (identifier)
1251 To attach a new pci device, the device should be recognized by kernel first.
1252 Then it should be moved under DPDK management.
1253 Finally the port can be attached to testpmd.
1255 For example, to move a pci device using ixgbe under DPDK management:
1257 .. code-block:: console
1259 # Check the status of the available devices.
1260 ./usertools/dpdk-devbind.py --status
1262 Network devices using DPDK-compatible driver
1263 ============================================
1266 Network devices using kernel driver
1267 ===================================
1268 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=
1271 # Bind the device to igb_uio.
1272 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0
1275 # Recheck the status of the devices.
1276 ./usertools/dpdk-devbind.py --status
1277 Network devices using DPDK-compatible driver
1278 ============================================
1279 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused=
1281 To attach a port created by virtual device, above steps are not needed.
1283 For example, to attach a port whose pci address is 0000:0a:00.0.
1285 .. code-block:: console
1287 testpmd> port attach 0000:0a:00.0
1288 Attaching a new port...
1289 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1290 EAL: probe driver: 8086:10fb rte_ixgbe_pmd
1291 EAL: PCI memory mapped at 0x7f83bfa00000
1292 EAL: PCI memory mapped at 0x7f83bfa80000
1293 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5
1294 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb
1295 Port 0 is attached. Now total ports is 1
1298 For example, to attach a port created by pcap PMD.
1300 .. code-block:: console
1302 testpmd> port attach net_pcap0
1303 Attaching a new port...
1304 PMD: Initializing pmd_pcap for net_pcap0
1305 PMD: Creating pcap-backed ethdev on numa socket 0
1306 Port 0 is attached. Now total ports is 1
1309 In this case, identifier is ``net_pcap0``.
1310 This identifier format is the same as ``--vdev`` format of DPDK applications.
1312 For example, to re-attach a bonded port which has been previously detached,
1313 the mode and slave parameters must be given.
1315 .. code-block:: console
1317 testpmd> port attach net_bond_0,mode=0,slave=1
1318 Attaching a new port...
1319 EAL: Initializing pmd_bond for net_bond_0
1320 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0.
1321 Port 0 is attached. Now total ports is 1
1328 Detach a specific port::
1330 testpmd> port detach (port_id)
1332 Before detaching a port, the port should be stopped and closed.
1334 For example, to detach a pci device port 0.
1336 .. code-block:: console
1338 testpmd> port stop 0
1341 testpmd> port close 0
1345 testpmd> port detach 0
1347 EAL: PCI device 0000:0a:00.0 on NUMA socket -1
1348 EAL: remove driver: 8086:10fb rte_ixgbe_pmd
1349 EAL: PCI memory unmapped at 0x7f83bfa00000
1350 EAL: PCI memory unmapped at 0x7f83bfa80000
1354 For example, to detach a virtual device port 0.
1356 .. code-block:: console
1358 testpmd> port stop 0
1361 testpmd> port close 0
1365 testpmd> port detach 0
1367 PMD: Closing pcap ethdev on numa socket 0
1368 Port 'net_pcap0' is detached. Now total ports is 0
1371 To remove a pci device completely from the system, first detach the port from testpmd.
1372 Then the device should be moved under kernel management.
1373 Finally the device can be removed using kernel pci hotplug functionality.
1375 For example, to move a pci device under kernel management:
1377 .. code-block:: console
1379 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0
1381 ./usertools/dpdk-devbind.py --status
1383 Network devices using DPDK-compatible driver
1384 ============================================
1387 Network devices using kernel driver
1388 ===================================
1389 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio
1391 To remove a port created by a virtual device, above steps are not needed.
1396 Start all ports or a specific port::
1398 testpmd> port start (port_id|all)
1403 Stop all ports or a specific port::
1405 testpmd> port stop (port_id|all)
1410 Close all ports or a specific port::
1412 testpmd> port close (port_id|all)
1414 port start/stop queue
1415 ~~~~~~~~~~~~~~~~~~~~~
1417 Start/stop a rx/tx queue on a specific port::
1419 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop)
1421 Only take effect when port is started.
1426 Set the speed and duplex mode for all ports or a specific port::
1428 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \
1429 duplex (half|full|auto)
1431 port config - queues/descriptors
1432 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1434 Set number of queues/descriptors for rxq, txq, rxd and txd::
1436 testpmd> port config all (rxq|txq|rxd|txd) (value)
1438 This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options.
1440 port config - max-pkt-len
1441 ~~~~~~~~~~~~~~~~~~~~~~~~~
1443 Set the maximum packet length::
1445 testpmd> port config all max-pkt-len (value)
1447 This is equivalent to the ``--max-pkt-len`` command-line option.
1449 port config - CRC Strip
1450 ~~~~~~~~~~~~~~~~~~~~~~~
1452 Set hardware CRC stripping on or off for all ports::
1454 testpmd> port config all crc-strip (on|off)
1456 CRC stripping is on by default.
1458 The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option.
1460 port config - scatter
1461 ~~~~~~~~~~~~~~~~~~~~~~~
1463 Set RX scatter mode on or off for all ports::
1465 testpmd> port config all scatter (on|off)
1467 RX scatter mode is off by default.
1469 The ``on`` option is equivalent to the ``--enable-scatter`` command-line option.
1471 port config - TX queue flags
1472 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1474 Set a hexadecimal bitmap of TX queue flags for all ports::
1476 testpmd> port config all txqflags value
1478 This command is equivalent to the ``--txqflags`` command-line option.
1480 port config - RX Checksum
1481 ~~~~~~~~~~~~~~~~~~~~~~~~~
1483 Set hardware RX checksum offload to on or off for all ports::
1485 testpmd> port config all rx-cksum (on|off)
1487 Checksum offload is off by default.
1489 The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option.
1494 Set hardware VLAN on or off for all ports::
1496 testpmd> port config all hw-vlan (on|off)
1498 Hardware VLAN is on by default.
1500 The ``off`` option is equivalent to the ``--disable-hw-vlan`` command-line option.
1502 port config - VLAN filter
1503 ~~~~~~~~~~~~~~~~~~~~~~~~~
1505 Set hardware VLAN filter on or off for all ports::
1507 testpmd> port config all hw-vlan-filter (on|off)
1509 Hardware VLAN filter is on by default.
1511 The ``off`` option is equivalent to the ``--disable-hw-vlan-filter`` command-line option.
1513 port config - VLAN strip
1514 ~~~~~~~~~~~~~~~~~~~~~~~~
1516 Set hardware VLAN strip on or off for all ports::
1518 testpmd> port config all hw-vlan-strip (on|off)
1520 Hardware VLAN strip is on by default.
1522 The ``off`` option is equivalent to the ``--disable-hw-vlan-strip`` command-line option.
1524 port config - VLAN extend
1525 ~~~~~~~~~~~~~~~~~~~~~~~~~
1527 Set hardware VLAN extend on or off for all ports::
1529 testpmd> port config all hw-vlan-extend (on|off)
1531 Hardware VLAN extend is off by default.
1533 The ``off`` option is equivalent to the ``--disable-hw-vlan-extend`` command-line option.
1535 port config - Drop Packets
1536 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1538 Set packet drop for packets with no descriptors on or off for all ports::
1540 testpmd> port config all drop-en (on|off)
1542 Packet dropping for packets with no descriptors is off by default.
1544 The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option.
1549 Set the RSS (Receive Side Scaling) mode on or off::
1551 testpmd> port config all rss (all|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none)
1553 RSS is on by default.
1555 The ``none`` option is equivalent to the ``--disable-rss`` command-line option.
1557 port config - RSS Reta
1558 ~~~~~~~~~~~~~~~~~~~~~~
1560 Set the RSS (Receive Side Scaling) redirection table::
1562 testpmd> port config all rss reta (hash,queue)[,(hash,queue)]
1567 Set the DCB mode for an individual port::
1569 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off)
1571 The traffic class should be 4 or 8.
1576 Set the number of packets per burst::
1578 testpmd> port config all burst (value)
1580 This is equivalent to the ``--burst`` command-line option.
1582 port config - Threshold
1583 ~~~~~~~~~~~~~~~~~~~~~~~
1585 Set thresholds for TX/RX queues::
1587 testpmd> port config all (threshold) (value)
1589 Where the threshold type can be:
1591 * ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255.
1593 * ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255.
1595 * ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255.
1597 * ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255.
1599 * ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255.
1601 * ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255.
1603 * ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd.
1605 * ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd.
1607 * ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd.
1609 These threshold options are also available from the command-line.
1614 Set the value of ether-type for E-tag::
1616 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value)
1618 Enable/disable the E-tag support::
1620 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable)
1623 Link Bonding Functions
1624 ----------------------
1626 The Link Bonding functions make it possible to dynamically create and
1627 manage link bonding devices from within testpmd interactive prompt.
1629 create bonded device
1630 ~~~~~~~~~~~~~~~~~~~~
1632 Create a new bonding device::
1634 testpmd> create bonded device (mode) (socket)
1636 For example, to create a bonded device in mode 1 on socket 0::
1638 testpmd> create bonded 1 0
1639 created new bonded device (port X)
1644 Adds Ethernet device to a Link Bonding device::
1646 testpmd> add bonding slave (slave id) (port id)
1648 For example, to add Ethernet device (port 6) to a Link Bonding device (port 10)::
1650 testpmd> add bonding slave 6 10
1653 remove bonding slave
1654 ~~~~~~~~~~~~~~~~~~~~
1656 Removes an Ethernet slave device from a Link Bonding device::
1658 testpmd> remove bonding slave (slave id) (port id)
1660 For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10)::
1662 testpmd> remove bonding slave 6 10
1667 Set the Link Bonding mode of a Link Bonding device::
1669 testpmd> set bonding mode (value) (port id)
1671 For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3)::
1673 testpmd> set bonding mode 3 10
1678 Set an Ethernet slave device as the primary device on a Link Bonding device::
1680 testpmd> set bonding primary (slave id) (port id)
1682 For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10)::
1684 testpmd> set bonding primary 6 10
1689 Set the MAC address of a Link Bonding device::
1691 testpmd> set bonding mac (port id) (mac)
1693 For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01::
1695 testpmd> set bonding mac 10 00:00:00:00:00:01
1697 set bonding xmit_balance_policy
1698 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1700 Set the transmission policy for a Link Bonding device when it is in Balance XOR mode::
1702 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34)
1704 For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports)::
1706 testpmd> set bonding xmit_balance_policy 10 l34
1709 set bonding mon_period
1710 ~~~~~~~~~~~~~~~~~~~~~~
1712 Set the link status monitoring polling period in milliseconds for a bonding device.
1714 This adds support for PMD slave devices which do not support link status interrupts.
1715 When the mon_period is set to a value greater than 0 then all PMD's which do not support
1716 link status ISR will be queried every polling interval to check if their link status has changed::
1718 testpmd> set bonding mon_period (port_id) (value)
1720 For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms::
1722 testpmd> set bonding mon_period 5 150
1728 Show the current configuration of a Link Bonding device::
1730 testpmd> show bonding config (port id)
1733 to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4)
1734 in balance mode with a transmission policy of layer 2+3::
1736 testpmd> show bonding config 9
1738 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23
1740 Active Slaves (3): [1 3 4]
1747 The Register Functions can be used to read from and write to registers on the network card referenced by a port number.
1748 This is mainly useful for debugging purposes.
1749 Reference should be made to the appropriate datasheet for the network card for details on the register addresses
1750 and fields that can be accessed.
1755 Display the value of a port register::
1757 testpmd> read reg (port_id) (address)
1759 For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller::
1761 testpmd> read reg 0 0xEE00
1762 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241)
1767 Display a port register bit field::
1769 testpmd> read regfield (port_id) (address) (bit_x) (bit_y)
1771 For example, reading the lowest two bits from the register in the example above::
1773 testpmd> read regfield 0 0xEE00 0 1
1774 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1)
1779 Display a single port register bit::
1781 testpmd> read regbit (port_id) (address) (bit_x)
1783 For example, reading the lowest bit from the register in the example above::
1785 testpmd> read regbit 0 0xEE00 0
1786 port 0 PCI register at offset 0xEE00: bit 0=1
1791 Set the value of a port register::
1793 testpmd> write reg (port_id) (address) (value)
1795 For example, to clear a register::
1797 testpmd> write reg 0 0xEE00 0x0
1798 port 0 PCI register at offset 0xEE00: 0x00000000 (0)
1803 Set bit field of a port register::
1805 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value)
1807 For example, writing to the register cleared in the example above::
1809 testpmd> write regfield 0 0xEE00 0 1 2
1810 port 0 PCI register at offset 0xEE00: 0x00000002 (2)
1815 Set single bit value of a port register::
1817 testpmd> write regbit (port_id) (address) (bit_x) (value)
1819 For example, to set the high bit in the register from the example above::
1821 testpmd> write regbit 0 0xEE00 31 1
1822 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658)
1828 This section details the available filter functions that are available.
1830 Note these functions interface the deprecated legacy filtering framework,
1831 superseded by *rte_flow*. See `Flow rules management`_.
1834 ~~~~~~~~~~~~~~~~~~~~
1836 Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue::
1838 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \
1839 ethertype (ether_type) (drop|fwd) queue (queue_id)
1841 The available information parameters are:
1843 * ``port_id``: The port which the Ethertype filter assigned on.
1845 * ``mac_addr``: Compare destination mac address.
1847 * ``mac_ignr``: Ignore destination mac address match.
1849 * ``mac_address``: Destination mac address to match.
1851 * ``ether_type``: The EtherType value want to match,
1852 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid.
1854 * ``queue_id``: The receive queue associated with this EtherType filter.
1855 It is meaningless when deleting or dropping.
1857 Example, to add/remove an ethertype filter rule::
1859 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \
1860 ethertype 0x0806 fwd queue 3
1862 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \
1863 ethertype 0x0806 fwd queue 3
1868 Add or delete a 2-tuple filter,
1869 which identifies packets by specific protocol and destination TCP/UDP port
1870 and forwards packets into one of the receive queues::
1872 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \
1873 protocol (protocol_value) mask (mask_value) \
1874 tcp_flags (tcp_flags_value) priority (prio_value) \
1877 The available information parameters are:
1879 * ``port_id``: The port which the 2-tuple filter assigned on.
1881 * ``dst_port_value``: Destination port in L4.
1883 * ``protocol_value``: IP L4 protocol.
1885 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate.
1887 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP).
1889 * ``prio_value``: Priority of this filter.
1891 * ``queue_id``: The receive queue associated with this 2-tuple filter.
1893 Example, to add/remove an 2tuple filter rule::
1895 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \
1896 tcp_flags 0x02 priority 3 queue 3
1898 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \
1899 tcp_flags 0x02 priority 3 queue 3
1904 Add or delete a 5-tuple filter,
1905 which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port)
1906 and routes packets into one of the receive queues::
1908 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \
1909 (src_address) dst_port (dst_port_value) \
1910 src_port (src_port_value) protocol (protocol_value) \
1911 mask (mask_value) tcp_flags (tcp_flags_value) \
1912 priority (prio_value) queue (queue_id)
1914 The available information parameters are:
1916 * ``port_id``: The port which the 5-tuple filter assigned on.
1918 * ``dst_address``: Destination IP address.
1920 * ``src_address``: Source IP address.
1922 * ``dst_port_value``: TCP/UDP destination port.
1924 * ``src_port_value``: TCP/UDP source port.
1926 * ``protocol_value``: L4 protocol.
1928 * ``mask_value``: Participates in the match or not by bit for field above, 1b means participate
1930 * ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP).
1932 * ``prio_value``: The priority of this filter.
1934 * ``queue_id``: The receive queue associated with this 5-tuple filter.
1936 Example, to add/remove an 5tuple filter rule::
1938 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1939 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1940 flags 0x0 priority 3 queue 3
1942 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \
1943 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \
1944 flags 0x0 priority 3 queue 3
1949 Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue::
1951 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id)
1953 The available information parameters are:
1955 * ``port_id``: The port which the SYN filter assigned on.
1957 * ``high``: This SYN filter has higher priority than other filters.
1959 * ``low``: This SYN filter has lower priority than other filters.
1961 * ``queue_id``: The receive queue associated with this SYN filter
1965 testpmd> syn_filter 0 add priority high queue 3
1970 With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet
1971 and routed into one of the receive queues::
1973 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \
1974 mask (mask_value) priority (prio_value) queue (queue_id)
1976 The available information parameters are:
1978 * ``port_id``: The port which the Flex filter is assigned on.
1980 * ``len_value``: Filter length in bytes, no greater than 128.
1982 * ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match.
1984 * ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match.
1986 * ``prio_value``: The priority of this filter.
1988 * ``queue_id``: The receive queue associated with this Flex filter.
1992 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \
1993 mask 000C priority 3 queue 3
1995 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \
1996 mask 000C priority 3 queue 3
1999 .. _testpmd_flow_director:
2001 flow_director_filter
2002 ~~~~~~~~~~~~~~~~~~~~
2004 The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues.
2006 Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and
2007 Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter:
2009 * Perfect match filters.
2010 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2011 The masked fields are for IP flow.
2013 * Signature filters.
2014 The hardware checks a match between a hash-based signature of the masked fields of the received packet.
2016 * Perfect-mac-vlan match filters.
2017 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2018 The masked fields are for MAC VLAN flow.
2020 * Perfect-tunnel match filters.
2021 The hardware checks a match between the masked fields of the received packets and the programmed filters.
2022 The masked fields are for tunnel flow.
2024 The Flow Director filters can match the different fields for different type of packet: flow type, specific input set
2025 per flow type and the flexible payload.
2027 The Flow Director can also mask out parts of all of these fields so that filters
2028 are only applied to certain fields or parts of the fields.
2030 Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information.
2032 # Commands to add flow director filters of different flow types::
2034 flow_director_filter (port_id) mode IP (add|del|update) \
2035 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \
2036 src (src_ip_address) dst (dst_ip_address) \
2037 tos (tos_value) proto (proto_value) ttl (ttl_value) \
2038 vlan (vlan_value) flexbytes (flexbytes_value) \
2039 (drop|fwd) pf|vf(vf_id) queue (queue_id) \
2042 flow_director_filter (port_id) mode IP (add|del|update) \
2043 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \
2044 src (src_ip_address) (src_port) \
2045 dst (dst_ip_address) (dst_port) \
2046 tos (tos_value) ttl (ttl_value) \
2047 vlan (vlan_value) flexbytes (flexbytes_value) \
2048 (drop|fwd) queue pf|vf(vf_id) (queue_id) \
2051 flow_director_filter (port_id) mode IP (add|del|update) \
2052 flow (ipv4-sctp|ipv6-sctp) \
2053 src (src_ip_address) (src_port) \
2054 dst (dst_ip_address) (dst_port) \
2055 tos (tos_value) ttl (ttl_value) \
2056 tag (verification_tag) vlan (vlan_value) \
2057 flexbytes (flexbytes_value) (drop|fwd) \
2058 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value)
2060 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \
2061 ether (ethertype) flexbytes (flexbytes_value) \
2062 (drop|fwd) pf|vf(vf_id) queue (queue_id)
2065 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \
2066 mac (mac_address) vlan (vlan_value) \
2067 flexbytes (flexbytes_value) (drop|fwd) \
2068 queue (queue_id) fd_id (fd_id_value)
2070 flow_director_filter (port_id) mode Tunnel (add|del|update) \
2071 mac (mac_address) vlan (vlan_value) \
2072 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \
2073 flexbytes (flexbytes_value) (drop|fwd) \
2074 queue (queue_id) fd_id (fd_id_value)
2076 For example, to add an ipv4-udp flow type filter::
2078 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \
2079 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \
2080 fwd pf queue 1 fd_id 1
2082 For example, add an ipv4-other flow type filter::
2084 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \
2085 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \
2086 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1
2091 Flush all flow director filters on a device::
2093 testpmd> flush_flow_director (port_id)
2095 Example, to flush all flow director filter on port 0::
2097 testpmd> flush_flow_director 0
2102 Set flow director's input masks::
2104 flow_director_mask (port_id) mode IP vlan (vlan_value) \
2105 src_mask (ipv4_src) (ipv6_src) (src_port) \
2106 dst_mask (ipv4_dst) (ipv6_dst) (dst_port)
2108 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value)
2110 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \
2111 mac (mac_value) tunnel-type (tunnel_type_value) \
2112 tunnel-id (tunnel_id_value)
2114 Example, to set flow director mask on port 0::
2116 testpmd> flow_director_mask 0 mode IP vlan 0xefff \
2117 src_mask 255.255.255.255 \
2118 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \
2119 dst_mask 255.255.255.255 \
2120 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF
2122 flow_director_flex_mask
2123 ~~~~~~~~~~~~~~~~~~~~~~~
2125 set masks of flow director's flexible payload based on certain flow type::
2127 testpmd> flow_director_flex_mask (port_id) \
2128 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2129 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \
2130 l2_payload|all) (mask)
2132 Example, to set flow director's flex mask for all flow type on port 0::
2134 testpmd> flow_director_flex_mask 0 flow all \
2135 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0)
2138 flow_director_flex_payload
2139 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2141 Configure flexible payload selection::
2143 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config)
2145 For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload::
2147 testpmd> flow_director_flex_payload 0 l4 \
2148 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19)
2150 get_sym_hash_ena_per_port
2151 ~~~~~~~~~~~~~~~~~~~~~~~~~
2153 Get symmetric hash enable configuration per port::
2155 get_sym_hash_ena_per_port (port_id)
2157 For example, to get symmetric hash enable configuration of port 1::
2159 testpmd> get_sym_hash_ena_per_port 1
2161 set_sym_hash_ena_per_port
2162 ~~~~~~~~~~~~~~~~~~~~~~~~~
2164 Set symmetric hash enable configuration per port to enable or disable::
2166 set_sym_hash_ena_per_port (port_id) (enable|disable)
2168 For example, to set symmetric hash enable configuration of port 1 to enable::
2170 testpmd> set_sym_hash_ena_per_port 1 enable
2172 get_hash_global_config
2173 ~~~~~~~~~~~~~~~~~~~~~~
2175 Get the global configurations of hash filters::
2177 get_hash_global_config (port_id)
2179 For example, to get the global configurations of hash filters of port 1::
2181 testpmd> get_hash_global_config 1
2183 set_hash_global_config
2184 ~~~~~~~~~~~~~~~~~~~~~~
2186 Set the global configurations of hash filters::
2188 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \
2189 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \
2190 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload) \
2193 For example, to enable simple_xor for flow type of ipv6 on port 2::
2195 testpmd> set_hash_global_config 2 simple_xor ipv6 enable
2200 Set the input set for hash::
2202 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2203 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2204 l2_payload) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2205 ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \
2206 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \
2207 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \
2208 fld-8th|none) (select|add)
2210 For example, to add source IP to hash input set for flow type of ipv4-udp on port 0::
2212 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add
2217 The Flow Director filters can match the different fields for different type of packet, i.e. specific input set
2218 on per flow type and the flexible payload. This command can be used to change input set for each flow type.
2220 Set the input set for flow director::
2222 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \
2223 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \
2224 l2_payload) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6|ipv4-tos| \
2225 ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \
2226 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \
2227 sctp-dst-port|sctp-veri-tag|none) (select|add)
2229 For example to add source IP to FD input set for flow type of ipv4-udp on port 0::
2231 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add
2236 Set different GRE key length for input set::
2238 global_config (port_id) gre-key-len (number in bytes)
2240 For example to set GRE key length for input set to 4 bytes on port 0::
2242 testpmd> global_config 0 gre-key-len 4
2245 .. _testpmd_rte_flow:
2247 Flow rules management
2248 ---------------------
2250 Control of the generic flow API (*rte_flow*) is fully exposed through the
2251 ``flow`` command (validation, creation, destruction and queries).
2253 Considering *rte_flow* overlaps with all `Filter Functions`_, using both
2254 features simultaneously may cause undefined side-effects and is therefore
2260 Because the ``flow`` command uses dynamic tokens to handle the large number
2261 of possible flow rules combinations, its behavior differs slightly from
2262 other commands, in particular:
2264 - Pressing *?* or the *<tab>* key displays contextual help for the current
2265 token, not that of the entire command.
2267 - Optional and repeated parameters are supported (provided they are listed
2268 in the contextual help).
2270 The first parameter stands for the operation mode. Possible operations and
2271 their general syntax are described below. They are covered in detail in the
2274 - Check whether a flow rule can be created::
2276 flow validate {port_id}
2277 [group {group_id}] [priority {level}] [ingress] [egress]
2278 pattern {item} [/ {item} [...]] / end
2279 actions {action} [/ {action} [...]] / end
2281 - Create a flow rule::
2283 flow create {port_id}
2284 [group {group_id}] [priority {level}] [ingress] [egress]
2285 pattern {item} [/ {item} [...]] / end
2286 actions {action} [/ {action} [...]] / end
2288 - Destroy specific flow rules::
2290 flow destroy {port_id} rule {rule_id} [...]
2292 - Destroy all flow rules::
2294 flow flush {port_id}
2296 - Query an existing flow rule::
2298 flow query {port_id} {rule_id} {action}
2300 - List existing flow rules sorted by priority, filtered by group
2303 flow list {port_id} [group {group_id}] [...]
2305 Validating flow rules
2306 ~~~~~~~~~~~~~~~~~~~~~
2308 ``flow validate`` reports whether a flow rule would be accepted by the
2309 underlying device in its current state but stops short of creating it. It is
2310 bound to ``rte_flow_validate()``::
2312 flow validate {port_id}
2313 [group {group_id}] [priority {level}] [ingress] [egress]
2314 pattern {item} [/ {item} [...]] / end
2315 actions {action} [/ {action} [...]] / end
2317 If successful, it will show::
2321 Otherwise it will show an error message of the form::
2323 Caught error type [...] ([...]): [...]
2325 This command uses the same parameters as ``flow create``, their format is
2326 described in `Creating flow rules`_.
2328 Check whether redirecting any Ethernet packet received on port 0 to RX queue
2329 index 6 is supported::
2331 testpmd> flow validate 0 ingress pattern eth / end
2332 actions queue index 6 / end
2336 Port 0 does not support TCPv6 rules::
2338 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end
2340 Caught error type 9 (specific pattern item): Invalid argument
2346 ``flow create`` validates and creates the specified flow rule. It is bound
2347 to ``rte_flow_create()``::
2349 flow create {port_id}
2350 [group {group_id}] [priority {level}] [ingress] [egress]
2351 pattern {item} [/ {item} [...]] / end
2352 actions {action} [/ {action} [...]] / end
2354 If successful, it will return a flow rule ID usable with other commands::
2356 Flow rule #[...] created
2358 Otherwise it will show an error message of the form::
2360 Caught error type [...] ([...]): [...]
2362 Parameters describe in the following order:
2364 - Attributes (*group*, *priority*, *ingress*, *egress* tokens).
2365 - A matching pattern, starting with the *pattern* token and terminated by an
2367 - Actions, starting with the *actions* token and terminated by an *end*
2370 These translate directly to *rte_flow* objects provided as-is to the
2371 underlying functions.
2373 The shortest valid definition only comprises mandatory tokens::
2375 testpmd> flow create 0 pattern end actions end
2377 Note that PMDs may refuse rules that essentially do nothing such as this
2380 **All unspecified object values are automatically initialized to 0.**
2385 These tokens affect flow rule attributes (``struct rte_flow_attr``) and are
2386 specified before the ``pattern`` token.
2388 - ``group {group id}``: priority group.
2389 - ``priority {level}``: priority level within group.
2390 - ``ingress``: rule applies to ingress traffic.
2391 - ``egress``: rule applies to egress traffic.
2393 Each instance of an attribute specified several times overrides the previous
2394 value as shown below (group 4 is used)::
2396 testpmd> flow create 0 group 42 group 24 group 4 [...]
2398 Note that once enabled, ``ingress`` and ``egress`` cannot be disabled.
2400 While not specifying a direction is an error, some rules may allow both
2403 Most rules affect RX therefore contain the ``ingress`` token::
2405 testpmd> flow create 0 ingress pattern [...]
2410 A matching pattern starts after the ``pattern`` token. It is made of pattern
2411 items and is terminated by a mandatory ``end`` item.
2413 Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum
2414 rte_flow_item_type``).
2416 The ``/`` token is used as a separator between pattern items as shown
2419 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...]
2421 Note that protocol items like these must be stacked from lowest to highest
2422 layer to make sense. For instance, the following rule is either invalid or
2423 unlikely to match any packet::
2425 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...]
2427 More information on these restrictions can be found in the *rte_flow*
2430 Several items support additional specification structures, for example
2431 ``ipv4`` allows specifying source and destination addresses as follows::
2433 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1
2434 dst is 10.2.0.0 / end [...]
2436 This rule matches all IPv4 traffic with the specified properties.
2438 In this example, ``src`` and ``dst`` are field names of the underlying
2439 ``struct rte_flow_item_ipv4`` object. All item properties can be specified
2440 in a similar fashion.
2442 The ``is`` token means that the subsequent value must be matched exactly,
2443 and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item``
2444 accordingly. Possible assignment tokens are:
2446 - ``is``: match value perfectly (with full bit-mask).
2447 - ``spec``: match value according to configured bit-mask.
2448 - ``last``: specify upper bound to establish a range.
2449 - ``mask``: specify bit-mask with relevant bits set to one.
2450 - ``prefix``: generate bit-mask from a prefix length.
2452 These yield identical results::
2454 ipv4 src is 10.1.1.1
2458 ipv4 src spec 10.1.1.1 src mask 255.255.255.255
2462 ipv4 src spec 10.1.1.1 src prefix 32
2466 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value
2470 ipv4 src is 10.1.1.1 src last 0 # 0 disables range
2472 Inclusive ranges can be defined with ``last``::
2474 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4
2476 Note that ``mask`` affects both ``spec`` and ``last``::
2478 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0
2479 # matches 10.1.0.0 to 10.2.255.255
2481 Properties can be modified multiple times::
2483 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4
2487 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16
2492 This section lists supported pattern items and their attributes, if any.
2494 - ``end``: end list of pattern items.
2496 - ``void``: no-op pattern item.
2498 - ``invert``: perform actions when pattern does not match.
2500 - ``any``: match any protocol for the current layer.
2502 - ``num {unsigned}``: number of layers covered.
2504 - ``pf``: match packets addressed to the physical function.
2506 - ``vf``: match packets addressed to a virtual function ID.
2508 - ``id {unsigned}``: destination VF ID.
2510 - ``port``: device-specific physical port index to use.
2512 - ``index {unsigned}``: physical port index.
2514 - ``raw``: match an arbitrary byte string.
2516 - ``relative {boolean}``: look for pattern after the previous item.
2517 - ``search {boolean}``: search pattern from offset (see also limit).
2518 - ``offset {integer}``: absolute or relative offset for pattern.
2519 - ``limit {unsigned}``: search area limit for start of pattern.
2520 - ``pattern {string}``: byte string to look for.
2522 - ``eth``: match Ethernet header.
2524 - ``dst {MAC-48}``: destination MAC.
2525 - ``src {MAC-48}``: source MAC.
2526 - ``type {unsigned}``: EtherType.
2528 - ``vlan``: match 802.1Q/ad VLAN tag.
2530 - ``tpid {unsigned}``: tag protocol identifier.
2531 - ``tci {unsigned}``: tag control information.
2532 - ``pcp {unsigned}``: priority code point.
2533 - ``dei {unsigned}``: drop eligible indicator.
2534 - ``vid {unsigned}``: VLAN identifier.
2536 - ``ipv4``: match IPv4 header.
2538 - ``tos {unsigned}``: type of service.
2539 - ``ttl {unsigned}``: time to live.
2540 - ``proto {unsigned}``: next protocol ID.
2541 - ``src {ipv4 address}``: source address.
2542 - ``dst {ipv4 address}``: destination address.
2544 - ``ipv6``: match IPv6 header.
2546 - ``tc {unsigned}``: traffic class.
2547 - ``flow {unsigned}``: flow label.
2548 - ``proto {unsigned}``: protocol (next header).
2549 - ``hop {unsigned}``: hop limit.
2550 - ``src {ipv6 address}``: source address.
2551 - ``dst {ipv6 address}``: destination address.
2553 - ``icmp``: match ICMP header.
2555 - ``type {unsigned}``: ICMP packet type.
2556 - ``code {unsigned}``: ICMP packet code.
2558 - ``udp``: match UDP header.
2560 - ``src {unsigned}``: UDP source port.
2561 - ``dst {unsigned}``: UDP destination port.
2563 - ``tcp``: match TCP header.
2565 - ``src {unsigned}``: TCP source port.
2566 - ``dst {unsigned}``: TCP destination port.
2568 - ``sctp``: match SCTP header.
2570 - ``src {unsigned}``: SCTP source port.
2571 - ``dst {unsigned}``: SCTP destination port.
2572 - ``tag {unsigned}``: validation tag.
2573 - ``cksum {unsigned}``: checksum.
2575 - ``vxlan``: match VXLAN header.
2577 - ``vni {unsigned}``: VXLAN identifier.
2579 - ``e_tag``: match IEEE 802.1BR E-Tag header.
2581 - ``grp_ecid_b {unsigned}``: GRP and E-CID base.
2583 - ``nvgre``: match NVGRE header.
2585 - ``tni {unsigned}``: virtual subnet ID.
2587 - ``mpls``: match MPLS header.
2589 - ``label {unsigned}``: MPLS label.
2591 - ``gre``: match GRE header.
2593 - ``protocol {unsigned}``: protocol type.
2598 A list of actions starts after the ``actions`` token in the same fashion as
2599 `Matching pattern`_; actions are separated by ``/`` tokens and the list is
2600 terminated by a mandatory ``end`` action.
2602 Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum
2603 rte_flow_action_type``).
2605 Dropping all incoming UDPv4 packets can be expressed as follows::
2607 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2610 Several actions have configurable properties which must be specified when
2611 there is no valid default value. For example, ``queue`` requires a target
2614 This rule redirects incoming UDPv4 traffic to queue index 6::
2616 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2617 actions queue index 6 / end
2619 While this one could be rejected by PMDs (unspecified queue index)::
2621 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end
2624 As defined by *rte_flow*, the list is not ordered, all actions of a given
2625 rule are performed simultaneously. These are equivalent::
2627 queue index 6 / void / mark id 42 / end
2631 void / mark id 42 / queue index 6 / end
2633 All actions in a list should have different types, otherwise only the last
2634 action of a given type is taken into account::
2636 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6
2640 drop / drop / drop / end # drop is performed only once
2644 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24
2646 Considering they are performed simultaneously, opposite and overlapping
2647 actions can sometimes be combined when the end result is unambiguous::
2649 drop / queue index 6 / end # drop has no effect
2653 drop / dup index 6 / end # same as above
2657 queue index 6 / rss queues 6 7 8 / end # queue has no effect
2661 drop / passthru / end # drop has no effect
2663 Note that PMDs may still refuse such combinations.
2668 This section lists supported actions and their attributes, if any.
2670 - ``end``: end list of actions.
2672 - ``void``: no-op action.
2674 - ``passthru``: let subsequent rule process matched packets.
2676 - ``mark``: attach 32 bit value to packets.
2678 - ``id {unsigned}``: 32 bit value to return with packets.
2680 - ``flag``: flag packets.
2682 - ``queue``: assign packets to a given queue index.
2684 - ``index {unsigned}``: queue index to use.
2686 - ``drop``: drop packets (note: passthru has priority).
2688 - ``count``: enable counters for this rule.
2690 - ``dup``: duplicate packets to a given queue index.
2692 - ``index {unsigned}``: queue index to duplicate packets to.
2694 - ``rss``: spread packets among several queues.
2696 - ``queues [{unsigned} [...]] end``: queue indices to use.
2698 - ``pf``: redirect packets to physical device function.
2700 - ``vf``: redirect packets to virtual device function.
2702 - ``original {boolean}``: use original VF ID if possible.
2703 - ``id {unsigned}``: VF ID to redirect packets to.
2705 Destroying flow rules
2706 ~~~~~~~~~~~~~~~~~~~~~
2708 ``flow destroy`` destroys one or more rules from their rule ID (as returned
2709 by ``flow create``), this command calls ``rte_flow_destroy()`` as many
2710 times as necessary::
2712 flow destroy {port_id} rule {rule_id} [...]
2714 If successful, it will show::
2716 Flow rule #[...] destroyed
2718 It does not report anything for rule IDs that do not exist. The usual error
2719 message is shown when a rule cannot be destroyed::
2721 Caught error type [...] ([...]): [...]
2723 ``flow flush`` destroys all rules on a device and does not take extra
2724 arguments. It is bound to ``rte_flow_flush()``::
2726 flow flush {port_id}
2728 Any errors are reported as above.
2730 Creating several rules and destroying them::
2732 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2733 actions queue index 2 / end
2734 Flow rule #0 created
2735 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2736 actions queue index 3 / end
2737 Flow rule #1 created
2738 testpmd> flow destroy 0 rule 0 rule 1
2739 Flow rule #1 destroyed
2740 Flow rule #0 destroyed
2743 The same result can be achieved using ``flow flush``::
2745 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2746 actions queue index 2 / end
2747 Flow rule #0 created
2748 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2749 actions queue index 3 / end
2750 Flow rule #1 created
2751 testpmd> flow flush 0
2754 Non-existent rule IDs are ignored::
2756 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2757 actions queue index 2 / end
2758 Flow rule #0 created
2759 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2760 actions queue index 3 / end
2761 Flow rule #1 created
2762 testpmd> flow destroy 0 rule 42 rule 10 rule 2
2764 testpmd> flow destroy 0 rule 0
2765 Flow rule #0 destroyed
2771 ``flow query`` queries a specific action of a flow rule having that
2772 ability. Such actions collect information that can be reported using this
2773 command. It is bound to ``rte_flow_query()``::
2775 flow query {port_id} {rule_id} {action}
2777 If successful, it will display either the retrieved data for known actions
2778 or the following message::
2780 Cannot display result for action type [...] ([...])
2782 Otherwise, it will complain either that the rule does not exist or that some
2785 Flow rule #[...] not found
2789 Caught error type [...] ([...]): [...]
2791 Currently only the ``count`` action is supported. This action reports the
2792 number of packets that hit the flow rule and the total number of bytes. Its
2793 output has the following format::
2796 hits_set: [...] # whether "hits" contains a valid value
2797 bytes_set: [...] # whether "bytes" contains a valid value
2798 hits: [...] # number of packets
2799 bytes: [...] # number of bytes
2801 Querying counters for TCPv6 packets redirected to queue 6::
2803 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end
2804 actions queue index 6 / count / end
2805 Flow rule #4 created
2806 testpmd> flow query 0 4 count
2817 ``flow list`` lists existing flow rules sorted by priority and optionally
2818 filtered by group identifiers::
2820 flow list {port_id} [group {group_id}] [...]
2822 This command only fails with the following message if the device does not
2827 Output consists of a header line followed by a short description of each
2828 flow rule, one per line. There is no output at all when no flow rules are
2829 configured on the device::
2831 ID Group Prio Attr Rule
2832 [...] [...] [...] [...] [...]
2834 ``Attr`` column flags:
2836 - ``i`` for ``ingress``.
2837 - ``e`` for ``egress``.
2839 Creating several flow rules and listing them::
2841 testpmd> flow create 0 ingress pattern eth / ipv4 / end
2842 actions queue index 6 / end
2843 Flow rule #0 created
2844 testpmd> flow create 0 ingress pattern eth / ipv6 / end
2845 actions queue index 2 / end
2846 Flow rule #1 created
2847 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end
2848 actions rss queues 6 7 8 end / end
2849 Flow rule #2 created
2850 testpmd> flow list 0
2851 ID Group Prio Attr Rule
2852 0 0 0 i- ETH IPV4 => QUEUE
2853 1 0 0 i- ETH IPV6 => QUEUE
2854 2 0 5 i- ETH IPV4 UDP => RSS
2857 Rules are sorted by priority (i.e. group ID first, then priority level)::
2859 testpmd> flow list 1
2860 ID Group Prio Attr Rule
2861 0 0 0 i- ETH => COUNT
2862 6 0 500 i- ETH IPV6 TCP => DROP COUNT
2863 5 0 1000 i- ETH IPV6 ICMP => QUEUE
2864 1 24 0 i- ETH IPV4 UDP => QUEUE
2865 4 24 10 i- ETH IPV4 TCP => DROP
2866 3 24 20 i- ETH IPV4 => DROP
2867 2 24 42 i- ETH IPV4 UDP => QUEUE
2868 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
2871 Output can be limited to specific groups::
2873 testpmd> flow list 1 group 0 group 63
2874 ID Group Prio Attr Rule
2875 0 0 0 i- ETH => COUNT
2876 6 0 500 i- ETH IPV6 TCP => DROP COUNT
2877 5 0 1000 i- ETH IPV6 ICMP => QUEUE
2878 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE
2881 Sample QinQ flow rules
2882 ~~~~~~~~~~~~~~~~~~~~~~
2884 Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM.
2888 testpmd> flow validate 0 ingress pattern eth / vlan tpid is 0x8100 tci is 4 /
2889 vlan tpid is 0x8100 tci is 5 / end actions vf id 1 / queue index 0 / end
2890 Flow rule #0 validated
2892 testpmd> flow create 0 ingress pattern eth / vlan tpid is 0x8100 tci is 4 /
2893 vlan tpid is 0x8100 tci is 5 / end actions vf id 1 / queue index 0 / end
2894 Flow rule #0 created
2896 testpmd> flow list 0
2897 ID Group Prio Attr Rule
2898 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
2900 Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host.
2904 testpmd> flow validate 0 ingress pattern eth / vlan tpid is 0x8100 tci is 6 /
2905 vlan tpid is 0x8100 tci is 7 / end actions pf / queue index 0 / end
2906 Flow rule #1 validated
2908 testpmd> flow create 0 ingress pattern eth / vlan tpid is 0x8100 tci is 6 /
2909 vlan tpid is 0x8100 tci is 7 / end actions pf / queue index 1 / end
2910 Flow rule #1 created
2912 testpmd> flow list 0
2913 ID Group Prio Attr Rule
2914 0 0 0 i- ETH VLAN VLAN=>VF QUEUE
2915 1 0 0 i- ETH VLAN VLAN=>PF QUEUE
2917 After creating QinQ rule(s) the following command should be issued to enable QinQ::
2919 testpmd> vlan set qinq on 0