1.. SPDX-License-Identifier: BSD-3-Clause 2 Copyright(c) 2010-2016 Intel Corporation. 3 4.. _testpmd_runtime: 5 6Testpmd Runtime Functions 7========================= 8 9Where the testpmd application is started in interactive mode, (``-i|--interactive``), 10it displays a prompt that can be used to start and stop forwarding, 11configure the application, display statistics (including the extended NIC 12statistics aka xstats) , set the Flow Director and other tasks:: 13 14 testpmd> 15 16The testpmd prompt has some, limited, readline support. 17Common bash command-line functions such as ``Ctrl+a`` and ``Ctrl+e`` to go to the start and end of the prompt line are supported 18as well as access to the command history via the up-arrow. 19 20There is also support for tab completion. 21If you type a partial command and hit ``<TAB>`` you get a list of the available completions: 22 23.. code-block:: console 24 25 testpmd> show port <TAB> 26 27 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X 28 info [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all 29 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap X 30 stats [Mul-choice STRING]: show|clear port info|stats|xstats|fdir|stat_qmap|dcb_tc|cap all 31 ... 32 33 34.. note:: 35 36 Some examples in this document are too long to fit on one line are shown wrapped at `"\\"` for display purposes:: 37 38 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 39 (pause_time) (send_xon) (port_id) 40 41In the real ``testpmd>`` prompt these commands should be on a single line. 42 43Help Functions 44-------------- 45 46The testpmd has on-line help for the functions that are available at runtime. 47These are divided into sections and can be accessed using help, help section or help all: 48 49.. code-block:: console 50 51 testpmd> help 52 53 help control : Start and stop forwarding. 54 help display : Displaying port, stats and config information. 55 help config : Configuration information. 56 help ports : Configuring ports. 57 help registers : Reading and setting port registers. 58 help filters : Filters configuration help. 59 help all : All of the above sections. 60 61 62Command File Functions 63---------------------- 64 65To facilitate loading large number of commands or to avoid cutting and pasting where not 66practical or possible testpmd supports alternative methods for executing commands. 67 68* If started with the ``--cmdline-file=FILENAME`` command line argument testpmd 69 will execute all CLI commands contained within the file immediately before 70 starting packet forwarding or entering interactive mode. 71 72.. code-block:: console 73 74 ./dpdk-testpmd -n4 -r2 ... -- -i --cmdline-file=/home/ubuntu/flow-create-commands.txt 75 Interactive-mode selected 76 CLI commands to be read from /home/ubuntu/flow-create-commands.txt 77 Configuring Port 0 (socket 0) 78 Port 0: 7C:FE:90:CB:74:CE 79 Configuring Port 1 (socket 0) 80 Port 1: 7C:FE:90:CB:74:CA 81 Checking link statuses... 82 Port 0 Link Up - speed 10000 Mbps - full-duplex 83 Port 1 Link Up - speed 10000 Mbps - full-duplex 84 Done 85 Flow rule #0 created 86 Flow rule #1 created 87 ... 88 ... 89 Flow rule #498 created 90 Flow rule #499 created 91 Read all CLI commands from /home/ubuntu/flow-create-commands.txt 92 testpmd> 93 94 95* At run-time additional commands can be loaded in bulk by invoking the ``load FILENAME`` 96 command. 97 98.. code-block:: console 99 100 testpmd> load /home/ubuntu/flow-create-commands.txt 101 Flow rule #0 created 102 Flow rule #1 created 103 ... 104 ... 105 Flow rule #498 created 106 Flow rule #499 created 107 Read all CLI commands from /home/ubuntu/flow-create-commands.txt 108 testpmd> 109 110 111In all cases output from any included command will be displayed as standard output. 112Execution will continue until the end of the file is reached regardless of 113whether any errors occur. The end user must examine the output to determine if 114any failures occurred. 115 116 117Control Functions 118----------------- 119 120start 121~~~~~ 122 123Start packet forwarding with current configuration:: 124 125 testpmd> start 126 127start tx_first 128~~~~~~~~~~~~~~ 129 130Start packet forwarding with current configuration after sending specified number of bursts of packets:: 131 132 testpmd> start tx_first (""|burst_num) 133 134The default burst number is 1 when ``burst_num`` not presented. 135 136stop 137~~~~ 138 139Stop packet forwarding, and display accumulated statistics:: 140 141 testpmd> stop 142 143quit 144~~~~ 145 146Quit to prompt:: 147 148 testpmd> quit 149 150 151Display Functions 152----------------- 153 154The functions in the following sections are used to display information about the 155testpmd configuration or the NIC status. 156 157show port 158~~~~~~~~~ 159 160Display information for a given port or all ports:: 161 162 testpmd> show port (info|summary|stats|xstats|fdir|stat_qmap|dcb_tc|cap) (port_id|all) 163 164The available information categories are: 165 166* ``info``: General port information such as MAC address. 167 168* ``summary``: Brief port summary such as Device Name, Driver Name etc. 169 170* ``stats``: RX/TX statistics. 171 172* ``xstats``: RX/TX extended NIC statistics. 173 174* ``fdir``: Flow Director information and statistics. 175 176* ``stat_qmap``: Queue statistics mapping. 177 178* ``dcb_tc``: DCB information such as TC mapping. 179 180* ``cap``: Supported offload capabilities. 181 182For example: 183 184.. code-block:: console 185 186 testpmd> show port info 0 187 188 ********************* Infos for port 0 ********************* 189 190 MAC address: XX:XX:XX:XX:XX:XX 191 Connect to socket: 0 192 memory allocation on the socket: 0 193 Link status: up 194 Link speed: 40000 Mbps 195 Link duplex: full-duplex 196 Promiscuous mode: enabled 197 Allmulticast mode: disabled 198 Maximum number of MAC addresses: 64 199 Maximum number of MAC addresses of hash filtering: 0 200 VLAN offload: 201 strip on, filter on, extend off, qinq strip off 202 Redirection table size: 512 203 Supported flow types: 204 ipv4-frag 205 ipv4-tcp 206 ipv4-udp 207 ipv4-sctp 208 ipv4-other 209 ipv6-frag 210 ipv6-tcp 211 ipv6-udp 212 ipv6-sctp 213 ipv6-other 214 l2_payload 215 port 216 vxlan 217 geneve 218 nvgre 219 vxlan-gpe 220 221show port (module_eeprom|eeprom) 222~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 223 224Display the EEPROM information of a port:: 225 226 testpmd> show port (port_id) (module_eeprom|eeprom) 227 228show port rss reta 229~~~~~~~~~~~~~~~~~~ 230 231Display the rss redirection table entry indicated by masks on port X:: 232 233 testpmd> show port (port_id) rss reta (size) (mask0, mask1...) 234 235size is used to indicate the hardware supported reta size 236 237show port rss-hash 238~~~~~~~~~~~~~~~~~~ 239 240Display the RSS hash functions and RSS hash key of a port:: 241 242 testpmd> show port (port_id) rss-hash [key] 243 244clear port 245~~~~~~~~~~ 246 247Clear the port statistics and forward engine statistics for a given port or for all ports:: 248 249 testpmd> clear port (info|stats|xstats|fdir|stat_qmap) (port_id|all) 250 251For example:: 252 253 testpmd> clear port stats all 254 255show (rxq|txq) 256~~~~~~~~~~~~~~ 257 258Display information for a given port's RX/TX queue:: 259 260 testpmd> show (rxq|txq) info (port_id) (queue_id) 261 262show desc status(rxq|txq) 263~~~~~~~~~~~~~~~~~~~~~~~~~ 264 265Display information for a given port's RX/TX descriptor status:: 266 267 testpmd> show port (port_id) (rxq|txq) (queue_id) desc (desc_id) status 268 269 270show config 271~~~~~~~~~~~ 272 273Displays the configuration of the application. 274The configuration comes from the command-line, the runtime or the application defaults:: 275 276 testpmd> show config (rxtx|cores|fwd|rxoffs|rxpkts|txpkts|txtimes) 277 278The available information categories are: 279 280* ``rxtx``: RX/TX configuration items. 281 282* ``cores``: List of forwarding cores. 283 284* ``fwd``: Packet forwarding configuration. 285 286* ``rxoffs``: Packet offsets for RX split. 287 288* ``rxpkts``: Packets to RX split configuration. 289 290* ``txpkts``: Packets to TX configuration. 291 292* ``txtimes``: Burst time pattern for Tx only mode. 293 294For example: 295 296.. code-block:: console 297 298 testpmd> show config rxtx 299 300 io packet forwarding - CRC stripping disabled - packets/burst=16 301 nb forwarding cores=2 - nb forwarding ports=1 302 RX queues=1 - RX desc=128 - RX free threshold=0 303 RX threshold registers: pthresh=8 hthresh=8 wthresh=4 304 TX queues=1 - TX desc=512 - TX free threshold=0 305 TX threshold registers: pthresh=36 hthresh=0 wthresh=0 306 TX RS bit threshold=0 - TXQ flags=0x0 307 308set fwd 309~~~~~~~ 310 311Set the packet forwarding mode:: 312 313 testpmd> set fwd (io|mac|macswap|flowgen| \ 314 rxonly|txonly|csum|icmpecho|noisy|5tswap) (""|retry) 315 316``retry`` can be specified for forwarding engines except ``rx_only``. 317 318The available information categories are: 319 320* ``io``: Forwards packets "as-is" in I/O mode. 321 This is the fastest possible forwarding operation as it does not access packets data. 322 This is the default mode. 323 324* ``mac``: Changes the source and the destination Ethernet addresses of packets before forwarding them. 325 Default application behavior is to set source Ethernet address to that of the transmitting interface, and destination 326 address to a dummy value (set during init). The user may specify a target destination Ethernet address via the 'eth-peer' or 327 'eth-peers-configfile' command-line options. It is not currently possible to specify a specific source Ethernet address. 328 329* ``macswap``: MAC swap forwarding mode. 330 Swaps the source and the destination Ethernet addresses of packets before forwarding them. 331 332* ``flowgen``: Multi-flow generation mode. 333 Originates a number of flows (with varying destination IP addresses), and terminate receive traffic. 334 335* ``rxonly``: Receives packets but doesn't transmit them. 336 337* ``txonly``: Generates and transmits packets without receiving any. 338 339* ``csum``: Changes the checksum field with hardware or software methods depending on the offload flags on the packet. 340 341* ``icmpecho``: Receives a burst of packets, lookup for ICMP echo requests and, if any, send back ICMP echo replies. 342 343* ``ieee1588``: Demonstrate L2 IEEE1588 V2 PTP timestamping for RX and TX. 344 345* ``noisy``: Noisy neighbor simulation. 346 Simulate more realistic behavior of a guest machine engaged in receiving 347 and sending packets performing Virtual Network Function (VNF). 348 349* ``5tswap``: Swap the source and destination of L2,L3,L4 if they exist. 350 351 L2 swaps the source address and destination address of Ethernet, as same as ``macswap``. 352 353 L3 swaps the source address and destination address of IP (v4 and v6). 354 355 L4 swaps the source port and destination port of transport layer (TCP and UDP). 356 357Example:: 358 359 testpmd> set fwd rxonly 360 361 Set rxonly packet forwarding mode 362 363 364show fwd 365~~~~~~~~ 366 367When running, forwarding engines maintain statistics from the time they have been started. 368Example for the io forwarding engine, with some packet drops on the tx side:: 369 370 testpmd> show fwd stats all 371 372 ------- Forward Stats for RX Port= 0/Queue= 0 -> TX Port= 1/Queue= 0 ------- 373 RX-packets: 274293770 TX-packets: 274293642 TX-dropped: 128 374 375 ------- Forward Stats for RX Port= 1/Queue= 0 -> TX Port= 0/Queue= 0 ------- 376 RX-packets: 274301850 TX-packets: 274301850 TX-dropped: 0 377 378 ---------------------- Forward statistics for port 0 ---------------------- 379 RX-packets: 274293802 RX-dropped: 0 RX-total: 274293802 380 TX-packets: 274301862 TX-dropped: 0 TX-total: 274301862 381 ---------------------------------------------------------------------------- 382 383 ---------------------- Forward statistics for port 1 ---------------------- 384 RX-packets: 274301894 RX-dropped: 0 RX-total: 274301894 385 TX-packets: 274293706 TX-dropped: 128 TX-total: 274293834 386 ---------------------------------------------------------------------------- 387 388 +++++++++++++++ Accumulated forward statistics for all ports+++++++++++++++ 389 RX-packets: 548595696 RX-dropped: 0 RX-total: 548595696 390 TX-packets: 548595568 TX-dropped: 128 TX-total: 548595696 391 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 392 393 394clear fwd 395~~~~~~~~~ 396 397Clear the forwarding engines statistics:: 398 399 testpmd> clear fwd stats all 400 401read rxd 402~~~~~~~~ 403 404Display an RX descriptor for a port RX queue:: 405 406 testpmd> read rxd (port_id) (queue_id) (rxd_id) 407 408For example:: 409 410 testpmd> read rxd 0 0 4 411 0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180 412 413read txd 414~~~~~~~~ 415 416Display a TX descriptor for a port TX queue:: 417 418 testpmd> read txd (port_id) (queue_id) (txd_id) 419 420For example:: 421 422 testpmd> read txd 0 0 4 423 0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C 424 425ddp get list 426~~~~~~~~~~~~ 427 428Get loaded dynamic device personalization (DDP) package info list:: 429 430 testpmd> ddp get list (port_id) 431 432ddp get info 433~~~~~~~~~~~~ 434 435Display information about dynamic device personalization (DDP) profile:: 436 437 testpmd> ddp get info (profile_path) 438 439show vf stats 440~~~~~~~~~~~~~ 441 442Display VF statistics:: 443 444 testpmd> show vf stats (port_id) (vf_id) 445 446clear vf stats 447~~~~~~~~~~~~~~ 448 449Reset VF statistics:: 450 451 testpmd> clear vf stats (port_id) (vf_id) 452 453show port pctype mapping 454~~~~~~~~~~~~~~~~~~~~~~~~ 455 456List all items from the pctype mapping table:: 457 458 testpmd> show port (port_id) pctype mapping 459 460show rx offloading capabilities 461~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 462 463List all per queue and per port Rx offloading capabilities of a port:: 464 465 testpmd> show port (port_id) rx_offload capabilities 466 467show rx offloading configuration 468~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 469 470List port level and all queue level Rx offloading configuration:: 471 472 testpmd> show port (port_id) rx_offload configuration 473 474show tx offloading capabilities 475~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 476 477List all per queue and per port Tx offloading capabilities of a port:: 478 479 testpmd> show port (port_id) tx_offload capabilities 480 481show tx offloading configuration 482~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 483 484List port level and all queue level Tx offloading configuration:: 485 486 testpmd> show port (port_id) tx_offload configuration 487 488show tx metadata setting 489~~~~~~~~~~~~~~~~~~~~~~~~ 490 491Show Tx metadata value set for a specific port:: 492 493 testpmd> show port (port_id) tx_metadata 494 495show port supported ptypes 496~~~~~~~~~~~~~~~~~~~~~~~~~~ 497 498Show ptypes supported for a specific port:: 499 500 testpmd> show port (port_id) ptypes 501 502set port supported ptypes 503~~~~~~~~~~~~~~~~~~~~~~~~~ 504 505set packet types classification for a specific port:: 506 507 testpmd> set port (port_id) ptypes_mask (mask) 508 509show port mac addresses info 510~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 511 512Show mac addresses added for a specific port:: 513 514 testpmd> show port (port_id) macs 515 516 517show port multicast mac addresses info 518~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 519 520Show multicast mac addresses added for a specific port:: 521 522 testpmd> show port (port_id) mcast_macs 523 524show device info 525~~~~~~~~~~~~~~~~ 526 527Show general information about devices probed:: 528 529 testpmd> show device info (<identifier>|all) 530 531For example: 532 533.. code-block:: console 534 535 testpmd> show device info net_pcap0 536 537 ********************* Infos for device net_pcap0 ********************* 538 Bus name: vdev 539 Driver name: net_pcap 540 Devargs: iface=enP2p6s0,phy_mac=1 541 Connect to socket: -1 542 543 Port id: 2 544 MAC address: 1E:37:93:28:04:B8 545 Device name: net_pcap0 546 547dump physmem 548~~~~~~~~~~~~ 549 550Dumps all physical memory segment layouts:: 551 552 testpmd> dump_physmem 553 554dump memzone 555~~~~~~~~~~~~ 556 557Dumps the layout of all memory zones:: 558 559 testpmd> dump_memzone 560 561dump socket memory 562~~~~~~~~~~~~~~~~~~ 563 564Dumps the memory usage of all sockets:: 565 566 testpmd> dump_socket_mem 567 568dump struct size 569~~~~~~~~~~~~~~~~ 570 571Dumps the size of all memory structures:: 572 573 testpmd> dump_struct_sizes 574 575dump ring 576~~~~~~~~~ 577 578Dumps the status of all or specific element in DPDK rings:: 579 580 testpmd> dump_ring [ring_name] 581 582dump mempool 583~~~~~~~~~~~~ 584 585Dumps the statistics of all or specific memory pool:: 586 587 testpmd> dump_mempool [mempool_name] 588 589dump devargs 590~~~~~~~~~~~~ 591 592Dumps the user device list:: 593 594 testpmd> dump_devargs 595 596dump log types 597~~~~~~~~~~~~~~ 598 599Dumps the log level for all the dpdk modules:: 600 601 testpmd> dump_log_types 602 603show (raw_encap|raw_decap) 604~~~~~~~~~~~~~~~~~~~~~~~~~~ 605 606Display content of raw_encap/raw_decap buffers in hex:: 607 608 testpmd> show <raw_encap|raw_decap> <index> 609 testpmd> show <raw_encap|raw_decap> all 610 611For example:: 612 613 testpmd> show raw_encap 6 614 615 index: 6 at [0x1c565b0], len=50 616 00000000: 00 00 00 00 00 00 16 26 36 46 56 66 08 00 45 00 | .......&6FVf..E. 617 00000010: 00 00 00 00 00 00 00 11 00 00 C0 A8 01 06 C0 A8 | ................ 618 00000020: 03 06 00 00 00 FA 00 00 00 00 08 00 00 00 00 00 | ................ 619 00000030: 06 00 | .. 620 621show fec capabilities 622~~~~~~~~~~~~~~~~~~~~~ 623 624Show fec capabilities of a port:: 625 626 testpmd> show port (port_id) fec capabilities 627 628show fec mode 629~~~~~~~~~~~~~ 630 631Show fec mode of a port:: 632 633 testpmd> show port (port_id) fec_mode 634 635 636Configuration Functions 637----------------------- 638 639The testpmd application can be configured from the runtime as well as from the command-line. 640 641This section details the available configuration functions that are available. 642 643.. note:: 644 645 Configuration changes only become active when forwarding is started/restarted. 646 647set default 648~~~~~~~~~~~ 649 650Reset forwarding to the default configuration:: 651 652 testpmd> set default 653 654set verbose 655~~~~~~~~~~~ 656 657Set the debug verbosity level:: 658 659 testpmd> set verbose (level) 660 661Available levels are as following: 662 663* ``0`` silent except for error. 664* ``1`` fully verbose except for Tx packets. 665* ``2`` fully verbose except for Rx packets. 666* ``> 2`` fully verbose. 667 668set log 669~~~~~~~ 670 671Set the log level for a log type:: 672 673 testpmd> set log global|(type) (level) 674 675Where: 676 677* ``type`` is the log name. 678 679* ``level`` is the log level. 680 681For example, to change the global log level:: 682 683 testpmd> set log global (level) 684 685Regexes can also be used for type. To change log level of user1, user2 and user3:: 686 687 testpmd> set log user[1-3] (level) 688 689set nbport 690~~~~~~~~~~ 691 692Set the number of ports used by the application: 693 694set nbport (num) 695 696This is equivalent to the ``--nb-ports`` command-line option. 697 698set nbcore 699~~~~~~~~~~ 700 701Set the number of cores used by the application:: 702 703 testpmd> set nbcore (num) 704 705This is equivalent to the ``--nb-cores`` command-line option. 706 707.. note:: 708 709 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port. 710 711set coremask 712~~~~~~~~~~~~ 713 714Set the forwarding cores hexadecimal mask:: 715 716 testpmd> set coremask (mask) 717 718This is equivalent to the ``--coremask`` command-line option. 719 720.. note:: 721 722 The main lcore is reserved for command line parsing only and cannot be masked on for packet forwarding. 723 724set portmask 725~~~~~~~~~~~~ 726 727Set the forwarding ports hexadecimal mask:: 728 729 testpmd> set portmask (mask) 730 731This is equivalent to the ``--portmask`` command-line option. 732 733set record-core-cycles 734~~~~~~~~~~~~~~~~~~~~~~ 735 736Set the recording of CPU cycles:: 737 738 testpmd> set record-core-cycles (on|off) 739 740Where: 741 742* ``on`` enables measurement of CPU cycles per packet. 743 744* ``off`` disables measurement of CPU cycles per packet. 745 746This is equivalent to the ``--record-core-cycles command-line`` option. 747 748set record-burst-stats 749~~~~~~~~~~~~~~~~~~~~~~ 750 751Set the displaying of RX and TX bursts:: 752 753 testpmd> set record-burst-stats (on|off) 754 755Where: 756 757* ``on`` enables display of RX and TX bursts. 758 759* ``off`` disables display of RX and TX bursts. 760 761This is equivalent to the ``--record-burst-stats command-line`` option. 762 763set burst 764~~~~~~~~~ 765 766Set number of packets per burst:: 767 768 testpmd> set burst (num) 769 770This is equivalent to the ``--burst command-line`` option. 771 772When retry is enabled, the transmit delay time and number of retries can also be set:: 773 774 testpmd> set burst tx delay (microseconds) retry (num) 775 776set rxoffs 777~~~~~~~~~~ 778 779Set the offsets of segments relating to the data buffer beginning on receiving 780if split feature is engaged. Affects only the queues configured with split 781offloads (currently BUFFER_SPLIT is supported only). 782 783 testpmd> set rxoffs (x[,y]*) 784 785Where x[,y]* represents a CSV list of values, without white space. If the list 786of offsets is shorter than the list of segments the zero offsets will be used 787for the remaining segments. 788 789set rxpkts 790~~~~~~~~~~ 791 792Set the length of segments to scatter packets on receiving if split 793feature is engaged. Affects only the queues configured with split offloads 794(currently BUFFER_SPLIT is supported only). Optionally the multiple memory 795pools can be specified with --mbuf-size command line parameter and the mbufs 796to receive will be allocated sequentially from these extra memory pools (the 797mbuf for the first segment is allocated from the first pool, the second one 798from the second pool, and so on, if segment number is greater then pool's the 799mbuf for remaining segments will be allocated from the last valid pool). 800 801 testpmd> set rxpkts (x[,y]*) 802 803Where x[,y]* represents a CSV list of values, without white space. Zero value 804means to use the corresponding memory pool data buffer size. 805 806set txpkts 807~~~~~~~~~~ 808 809Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode:: 810 811 testpmd> set txpkts (x[,y]*) 812 813Where x[,y]* represents a CSV list of values, without white space. 814 815set txtimes 816~~~~~~~~~~~ 817 818Configure the timing burst pattern for Tx only mode. This command enables 819the packet send scheduling on dynamic timestamp mbuf field and configures 820timing pattern in Tx only mode. In this mode, if scheduling is enabled 821application provides timestamps in the packets being sent. It is possible 822to configure delay (in unspecified device clock units) between bursts 823and between the packets within the burst:: 824 825 testpmd> set txtimes (inter),(intra) 826 827where: 828 829* ``inter`` is the delay between the bursts in the device clock units. 830 If ``intra`` is zero, this is the time between the beginnings of the 831 first packets in the neighbour bursts, if ``intra`` is not zero, 832 ``inter`` specifies the time between the beginning of the first packet 833 of the current burst and the beginning of the last packet of the 834 previous burst. If ``inter`` parameter is zero the send scheduling 835 on timestamps is disabled (default). 836 837* ``intra`` is the delay between the packets within the burst specified 838 in the device clock units. The number of packets in the burst is defined 839 by regular burst setting. If ``intra`` parameter is zero no timestamps 840 provided in the packets excepting the first one in the burst. 841 842As the result the bursts of packet will be transmitted with specific 843delays between the packets within the burst and specific delay between 844the bursts. The rte_eth_read_clock() must be supported by the device(s) 845and is supposed to be engaged to get the current device clock value 846and provide the reference for the timestamps. If there is no supported 847rte_eth_read_clock() there will be no send scheduling provided on the port. 848 849set txsplit 850~~~~~~~~~~~ 851 852Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes:: 853 854 testpmd> set txsplit (off|on|rand) 855 856Where: 857 858* ``off`` disable packet copy & split for CSUM mode. 859 860* ``on`` split outgoing packet into multiple segments. Size of each segment 861 and number of segments per packet is determined by ``set txpkts`` command 862 (see above). 863 864* ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments. 865 866set corelist 867~~~~~~~~~~~~ 868 869Set the list of forwarding cores:: 870 871 testpmd> set corelist (x[,y]*) 872 873For example, to change the forwarding cores: 874 875.. code-block:: console 876 877 testpmd> set corelist 3,1 878 testpmd> show config fwd 879 880 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled 881 Logical Core 3 (socket 0) forwards packets on 1 streams: 882 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01 883 Logical Core 1 (socket 0) forwards packets on 1 streams: 884 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00 885 886.. note:: 887 888 The cores are used in the same order as specified on the command line. 889 890set portlist 891~~~~~~~~~~~~ 892 893Set the list of forwarding ports:: 894 895 testpmd> set portlist (x[,y]*) 896 897For example, to change the port forwarding: 898 899.. code-block:: console 900 901 testpmd> set portlist 0,2,1,3 902 testpmd> show config fwd 903 904 io packet forwarding - ports=4 - cores=1 - streams=4 905 Logical Core 3 (socket 0) forwards packets on 4 streams: 906 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01 907 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00 908 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03 909 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02 910 911set port setup on 912~~~~~~~~~~~~~~~~~ 913 914Select how to retrieve new ports created after "port attach" command:: 915 916 testpmd> set port setup on (iterator|event) 917 918For each new port, a setup is done. 919It will find the probed ports via RTE_ETH_FOREACH_MATCHING_DEV loop 920in iterator mode, or via RTE_ETH_EVENT_NEW in event mode. 921 922set tx loopback 923~~~~~~~~~~~~~~~ 924 925Enable/disable tx loopback:: 926 927 testpmd> set tx loopback (port_id) (on|off) 928 929set drop enable 930~~~~~~~~~~~~~~~ 931 932set drop enable bit for all queues:: 933 934 testpmd> set all queues drop (port_id) (on|off) 935 936set split drop enable (for VF) 937~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 938 939set split drop enable bit for VF from PF:: 940 941 testpmd> set vf split drop (port_id) (vf_id) (on|off) 942 943set mac antispoof (for VF) 944~~~~~~~~~~~~~~~~~~~~~~~~~~ 945 946Set mac antispoof for a VF from the PF:: 947 948 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off) 949 950set macsec offload 951~~~~~~~~~~~~~~~~~~ 952 953Enable/disable MACsec offload:: 954 955 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off) 956 testpmd> set macsec offload (port_id) off 957 958set macsec sc 959~~~~~~~~~~~~~ 960 961Configure MACsec secure connection (SC):: 962 963 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi) 964 965.. note:: 966 967 The pi argument is ignored for tx. 968 Check the NIC Datasheet for hardware limits. 969 970set macsec sa 971~~~~~~~~~~~~~ 972 973Configure MACsec secure association (SA):: 974 975 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key) 976 977.. note:: 978 979 The IDX value must be 0 or 1. 980 Check the NIC Datasheet for hardware limits. 981 982set broadcast mode (for VF) 983~~~~~~~~~~~~~~~~~~~~~~~~~~~ 984 985Set broadcast mode for a VF from the PF:: 986 987 testpmd> set vf broadcast (port_id) (vf_id) (on|off) 988 989vlan set stripq 990~~~~~~~~~~~~~~~ 991 992Set the VLAN strip for a queue on a port:: 993 994 testpmd> vlan set stripq (on|off) (port_id,queue_id) 995 996vlan set stripq (for VF) 997~~~~~~~~~~~~~~~~~~~~~~~~ 998 999Set VLAN strip for all queues in a pool for a VF from the PF:: 1000 1001 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off) 1002 1003vlan set insert (for VF) 1004~~~~~~~~~~~~~~~~~~~~~~~~ 1005 1006Set VLAN insert for a VF from the PF:: 1007 1008 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id) 1009 1010vlan set tag (for VF) 1011~~~~~~~~~~~~~~~~~~~~~ 1012 1013Set VLAN tag for a VF from the PF:: 1014 1015 testpmd> set vf vlan tag (port_id) (vf_id) (on|off) 1016 1017vlan set antispoof (for VF) 1018~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1019 1020Set VLAN antispoof for a VF from the PF:: 1021 1022 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off) 1023 1024vlan set (strip|filter|qinq_strip|extend) 1025~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1026Set the VLAN strip/filter/QinQ strip/extend on for a port:: 1027 1028 testpmd> vlan set (strip|filter|qinq_strip|extend) (on|off) (port_id) 1029 1030vlan set tpid 1031~~~~~~~~~~~~~ 1032 1033Set the inner or outer VLAN TPID for packet filtering on a port:: 1034 1035 testpmd> vlan set (inner|outer) tpid (value) (port_id) 1036 1037.. note:: 1038 1039 TPID value must be a 16-bit number (value <= 65536). 1040 1041rx_vlan add 1042~~~~~~~~~~~ 1043 1044Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID:: 1045 1046 testpmd> rx_vlan add (vlan_id|all) (port_id) 1047 1048.. note:: 1049 1050 VLAN filter must be set on that port. VLAN ID < 4096. 1051 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries 1052 in VFTA table. This is important if enabling all vlan_ids. 1053 1054rx_vlan rm 1055~~~~~~~~~~ 1056 1057Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID:: 1058 1059 testpmd> rx_vlan rm (vlan_id|all) (port_id) 1060 1061rx_vlan add (for VF) 1062~~~~~~~~~~~~~~~~~~~~ 1063 1064Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID:: 1065 1066 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask) 1067 1068rx_vlan rm (for VF) 1069~~~~~~~~~~~~~~~~~~~ 1070 1071Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID:: 1072 1073 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask) 1074 1075rx_vxlan_port add 1076~~~~~~~~~~~~~~~~~ 1077 1078Add an UDP port for VXLAN packet filter on a port:: 1079 1080 testpmd> rx_vxlan_port add (udp_port) (port_id) 1081 1082rx_vxlan_port remove 1083~~~~~~~~~~~~~~~~~~~~ 1084 1085Remove an UDP port for VXLAN packet filter on a port:: 1086 1087 testpmd> rx_vxlan_port rm (udp_port) (port_id) 1088 1089tx_vlan set 1090~~~~~~~~~~~ 1091 1092Set hardware insertion of VLAN IDs in packets sent on a port:: 1093 1094 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer] 1095 1096For example, set a single VLAN ID (5) insertion on port 0:: 1097 1098 tx_vlan set 0 5 1099 1100Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1:: 1101 1102 tx_vlan set 1 2 3 1103 1104 1105tx_vlan set pvid 1106~~~~~~~~~~~~~~~~ 1107 1108Set port based hardware insertion of VLAN ID in packets sent on a port:: 1109 1110 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off) 1111 1112tx_vlan reset 1113~~~~~~~~~~~~~ 1114 1115Disable hardware insertion of a VLAN header in packets sent on a port:: 1116 1117 testpmd> tx_vlan reset (port_id) 1118 1119csum set 1120~~~~~~~~ 1121 1122Select hardware or software calculation of the checksum when 1123transmitting a packet using the ``csum`` forwarding engine:: 1124 1125 testpmd> csum set (ip|udp|tcp|sctp|outer-ip|outer-udp) (hw|sw) (port_id) 1126 1127Where: 1128 1129* ``ip|udp|tcp|sctp`` always relate to the inner layer. 1130 1131* ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized 1132 as a tunnel packet by the forwarding engine (geneve, gre, gtp, ipip, vxlan and vxlan-gpe are 1133 supported). See also the ``csum parse-tunnel`` command. 1134 1135* ``outer-udp`` relates to the outer UDP layer in the case where the packet is recognized 1136 as a tunnel packet by the forwarding engine (geneve, gtp, vxlan and vxlan-gpe are 1137 supported). See also the ``csum parse-tunnel`` command. 1138 1139.. note:: 1140 1141 Check the NIC Datasheet for hardware limits. 1142 1143RSS queue region 1144~~~~~~~~~~~~~~~~ 1145 1146Set RSS queue region span on a port:: 1147 1148 testpmd> set port (port_id) queue-region region_id (value) \ 1149 queue_start_index (value) queue_num (value) 1150 1151Set flowtype mapping on a RSS queue region on a port:: 1152 1153 testpmd> set port (port_id) queue-region region_id (value) flowtype (value) 1154 1155where: 1156 1157* For the flowtype(pctype) of packet,the specific index for each type has 1158 been defined in file i40e_type.h as enum i40e_filter_pctype. 1159 1160Set user priority mapping on a RSS queue region on a port:: 1161 1162 testpmd> set port (port_id) queue-region UP (value) region_id (value) 1163 1164Flush all queue region related configuration on a port:: 1165 1166 testpmd> set port (port_id) queue-region flush (on|off) 1167 1168where: 1169 1170* ``on``: is just an enable function which server for other configuration, 1171 it is for all configuration about queue region from up layer, 1172 at first will only keep in DPDK software stored in driver, 1173 only after "flush on", it commit all configuration to HW. 1174 1175* ``"off``: is just clean all configuration about queue region just now, 1176 and restore all to DPDK i40e driver default config when start up. 1177 1178Show all queue region related configuration info on a port:: 1179 1180 testpmd> show port (port_id) queue-region 1181 1182.. note:: 1183 1184 Queue region only support on PF by now, so these command is 1185 only for configuration of queue region on PF port. 1186 1187csum parse-tunnel 1188~~~~~~~~~~~~~~~~~ 1189 1190Define how tunneled packets should be handled by the csum forward 1191engine:: 1192 1193 testpmd> csum parse-tunnel (on|off) (tx_port_id) 1194 1195If enabled, the csum forward engine will try to recognize supported 1196tunnel headers (geneve, gtp, gre, ipip, vxlan, vxlan-gpe). 1197 1198If disabled, treat tunnel packets as non-tunneled packets (a inner 1199header is handled as a packet payload). 1200 1201.. note:: 1202 1203 The port argument is the TX port like in the ``csum set`` command. 1204 1205Example: 1206 1207Consider a packet in packet like the following:: 1208 1209 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in 1210 1211* If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set`` 1212 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the 1213 ``outer-ip|outer-udp`` parameter relates to the outer headers (here ``ipv4_out`` and ``udp_out``). 1214 1215* If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set`` 1216 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``. 1217 1218csum show 1219~~~~~~~~~ 1220 1221Display tx checksum offload configuration:: 1222 1223 testpmd> csum show (port_id) 1224 1225tso set 1226~~~~~~~ 1227 1228Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine:: 1229 1230 testpmd> tso set (segsize) (port_id) 1231 1232.. note:: 1233 1234 Check the NIC datasheet for hardware limits. 1235 1236tso show 1237~~~~~~~~ 1238 1239Display the status of TCP Segmentation Offload:: 1240 1241 testpmd> tso show (port_id) 1242 1243tunnel tso set 1244~~~~~~~~~~~~~~ 1245 1246Set tso segment size of tunneled packets for a port in csum engine:: 1247 1248 testpmd> tunnel_tso set (tso_segsz) (port_id) 1249 1250tunnel tso show 1251~~~~~~~~~~~~~~~ 1252 1253Display the status of tunneled TCP Segmentation Offload for a port:: 1254 1255 testpmd> tunnel_tso show (port_id) 1256 1257set port - gro 1258~~~~~~~~~~~~~~ 1259 1260Enable or disable GRO in ``csum`` forwarding engine:: 1261 1262 testpmd> set port <port_id> gro on|off 1263 1264If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4 1265packets received from the given port. 1266 1267If disabled, packets received from the given port won't be performed 1268GRO. By default, GRO is disabled for all ports. 1269 1270.. note:: 1271 1272 When enable GRO for a port, TCP/IPv4 packets received from the port 1273 will be performed GRO. After GRO, all merged packets have bad 1274 checksums, since the GRO library doesn't re-calculate checksums for 1275 the merged packets. Therefore, if users want the merged packets to 1276 have correct checksums, please select HW IP checksum calculation and 1277 HW TCP checksum calculation for the port which the merged packets are 1278 transmitted to. 1279 1280show port - gro 1281~~~~~~~~~~~~~~~ 1282 1283Display GRO configuration for a given port:: 1284 1285 testpmd> show port <port_id> gro 1286 1287set gro flush 1288~~~~~~~~~~~~~ 1289 1290Set the cycle to flush the GROed packets from reassembly tables:: 1291 1292 testpmd> set gro flush <cycles> 1293 1294When enable GRO, the csum forwarding engine performs GRO on received 1295packets, and the GROed packets are stored in reassembly tables. Users 1296can use this command to determine when the GROed packets are flushed 1297from the reassembly tables. 1298 1299The ``cycles`` is measured in GRO operation times. The csum forwarding 1300engine flushes the GROed packets from the tables every ``cycles`` GRO 1301operations. 1302 1303By default, the value of ``cycles`` is 1, which means flush GROed packets 1304from the reassembly tables as soon as one GRO operation finishes. The value 1305of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``. 1306 1307Please note that the large value of ``cycles`` may cause the poor TCP/IP 1308stack performance. Because the GROed packets are delayed to arrive the 1309stack, thus causing more duplicated ACKs and TCP retransmissions. 1310 1311set port - gso 1312~~~~~~~~~~~~~~ 1313 1314Toggle per-port GSO support in ``csum`` forwarding engine:: 1315 1316 testpmd> set port <port_id> gso on|off 1317 1318If enabled, the csum forwarding engine will perform GSO on supported IPv4 1319packets, transmitted on the given port. 1320 1321If disabled, packets transmitted on the given port will not undergo GSO. 1322By default, GSO is disabled for all ports. 1323 1324.. note:: 1325 1326 When GSO is enabled on a port, supported IPv4 packets transmitted on that 1327 port undergo GSO. Afterwards, the segmented packets are represented by 1328 multi-segment mbufs; however, the csum forwarding engine doesn't calculation 1329 of checksums for GSO'd segments in SW. As a result, if users want correct 1330 checksums in GSO segments, they should enable HW checksum calculation for 1331 GSO-enabled ports. 1332 1333 For example, HW checksum calculation for VxLAN GSO'd packets may be enabled 1334 by setting the following options in the csum forwarding engine: 1335 1336 testpmd> csum set outer_ip hw <port_id> 1337 1338 testpmd> csum set ip hw <port_id> 1339 1340 testpmd> csum set tcp hw <port_id> 1341 1342 UDP GSO is the same as IP fragmentation, which treats the UDP header 1343 as the payload and does not modify it during segmentation. That is, 1344 after UDP GSO, only the first output fragment has the original UDP 1345 header. Therefore, users need to enable HW IP checksum calculation 1346 and SW UDP checksum calculation for GSO-enabled ports, if they want 1347 correct checksums for UDP/IPv4 packets. 1348 1349set gso segsz 1350~~~~~~~~~~~~~ 1351 1352Set the maximum GSO segment size (measured in bytes), which includes the 1353packet header and the packet payload for GSO-enabled ports (global):: 1354 1355 testpmd> set gso segsz <length> 1356 1357show port - gso 1358~~~~~~~~~~~~~~~ 1359 1360Display the status of Generic Segmentation Offload for a given port:: 1361 1362 testpmd> show port <port_id> gso 1363 1364mac_addr add 1365~~~~~~~~~~~~ 1366 1367Add an alternative MAC address to a port:: 1368 1369 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX) 1370 1371mac_addr remove 1372~~~~~~~~~~~~~~~ 1373 1374Remove a MAC address from a port:: 1375 1376 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX) 1377 1378mcast_addr add 1379~~~~~~~~~~~~~~ 1380 1381To add the multicast MAC address to/from the set of multicast addresses 1382filtered by port:: 1383 1384 testpmd> mcast_addr add (port_id) (mcast_addr) 1385 1386mcast_addr remove 1387~~~~~~~~~~~~~~~~~ 1388 1389To remove the multicast MAC address to/from the set of multicast addresses 1390filtered by port:: 1391 1392 testpmd> mcast_addr remove (port_id) (mcast_addr) 1393 1394mac_addr add (for VF) 1395~~~~~~~~~~~~~~~~~~~~~ 1396 1397Add an alternative MAC address for a VF to a port:: 1398 1399 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX) 1400 1401mac_addr set 1402~~~~~~~~~~~~ 1403 1404Set the default MAC address for a port:: 1405 1406 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX) 1407 1408mac_addr set (for VF) 1409~~~~~~~~~~~~~~~~~~~~~ 1410 1411Set the MAC address for a VF from the PF:: 1412 1413 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX) 1414 1415set eth-peer 1416~~~~~~~~~~~~ 1417 1418Set the forwarding peer address for certain port:: 1419 1420 testpmd> set eth-peer (port_id) (peer_addr) 1421 1422This is equivalent to the ``--eth-peer`` command-line option. 1423 1424set port-uta 1425~~~~~~~~~~~~ 1426 1427Set the unicast hash filter(s) on/off for a port:: 1428 1429 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off) 1430 1431set promisc 1432~~~~~~~~~~~ 1433 1434Set the promiscuous mode on for a port or for all ports. 1435In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1436 1437 testpmd> set promisc (port_id|all) (on|off) 1438 1439set allmulti 1440~~~~~~~~~~~~ 1441 1442Set the allmulti mode for a port or for all ports:: 1443 1444 testpmd> set allmulti (port_id|all) (on|off) 1445 1446Same as the ifconfig (8) option. Controls how multicast packets are handled. 1447 1448set promisc (for VF) 1449~~~~~~~~~~~~~~~~~~~~ 1450 1451Set the unicast promiscuous mode for a VF from PF. 1452It's supported by Intel i40e NICs now. 1453In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1454 1455 testpmd> set vf promisc (port_id) (vf_id) (on|off) 1456 1457set allmulticast (for VF) 1458~~~~~~~~~~~~~~~~~~~~~~~~~ 1459 1460Set the multicast promiscuous mode for a VF from PF. 1461It's supported by Intel i40e NICs now. 1462In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1463 1464 testpmd> set vf allmulti (port_id) (vf_id) (on|off) 1465 1466set tx max bandwidth (for VF) 1467~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1468 1469Set TX max absolute bandwidth (Mbps) for a VF from PF:: 1470 1471 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth) 1472 1473set tc tx min bandwidth (for VF) 1474~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1475 1476Set all TCs' TX min relative bandwidth (%) for a VF from PF:: 1477 1478 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...) 1479 1480set tc tx max bandwidth (for VF) 1481~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1482 1483Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF:: 1484 1485 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth) 1486 1487set tc strict link priority mode 1488~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1489 1490Set some TCs' strict link priority mode on a physical port:: 1491 1492 testpmd> set tx strict-link-priority (port_id) (tc_bitmap) 1493 1494set tc tx min bandwidth 1495~~~~~~~~~~~~~~~~~~~~~~~ 1496 1497Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs:: 1498 1499 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...) 1500 1501set flow_ctrl rx 1502~~~~~~~~~~~~~~~~ 1503 1504Set the link flow control parameter on a port:: 1505 1506 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 1507 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \ 1508 autoneg (on|off) (port_id) 1509 1510Where: 1511 1512* ``high_water`` (integer): High threshold value to trigger XOFF. 1513 1514* ``low_water`` (integer): Low threshold value to trigger XON. 1515 1516* ``pause_time`` (integer): Pause quota in the Pause frame. 1517 1518* ``send_xon`` (0/1): Send XON frame. 1519 1520* ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames. 1521 1522* ``autoneg``: Change the auto-negotiation parameter. 1523 1524set pfc_ctrl rx 1525~~~~~~~~~~~~~~~ 1526 1527Set the priority flow control parameter on a port:: 1528 1529 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 1530 (pause_time) (priority) (port_id) 1531 1532Where: 1533 1534* ``high_water`` (integer): High threshold value. 1535 1536* ``low_water`` (integer): Low threshold value. 1537 1538* ``pause_time`` (integer): Pause quota in the Pause frame. 1539 1540* ``priority`` (0-7): VLAN User Priority. 1541 1542set stat_qmap 1543~~~~~~~~~~~~~ 1544 1545Set statistics mapping (qmapping 0..15) for RX/TX queue on port:: 1546 1547 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping) 1548 1549For example, to set rx queue 2 on port 0 to mapping 5:: 1550 1551 testpmd>set stat_qmap rx 0 2 5 1552 1553set xstats-hide-zero 1554~~~~~~~~~~~~~~~~~~~~ 1555 1556Set the option to hide zero values for xstats display:: 1557 1558 testpmd> set xstats-hide-zero on|off 1559 1560.. note:: 1561 1562 By default, the zero values are displayed for xstats. 1563 1564set port - rx/tx (for VF) 1565~~~~~~~~~~~~~~~~~~~~~~~~~ 1566 1567Set VF receive/transmit from a port:: 1568 1569 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off) 1570 1571set port - rx mode(for VF) 1572~~~~~~~~~~~~~~~~~~~~~~~~~~ 1573 1574Set the VF receive mode of a port:: 1575 1576 testpmd> set port (port_id) vf (vf_id) \ 1577 rxmode (AUPE|ROPE|BAM|MPE) (on|off) 1578 1579The available receive modes are: 1580 1581* ``AUPE``: Accepts untagged VLAN. 1582 1583* ``ROPE``: Accepts unicast hash. 1584 1585* ``BAM``: Accepts broadcast packets. 1586 1587* ``MPE``: Accepts all multicast packets. 1588 1589set port - tx_rate (for Queue) 1590~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1591 1592Set TX rate limitation for a queue on a port:: 1593 1594 testpmd> set port (port_id) queue (queue_id) rate (rate_value) 1595 1596set port - tx_rate (for VF) 1597~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1598 1599Set TX rate limitation for queues in VF on a port:: 1600 1601 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask) 1602 1603set port - mirror rule 1604~~~~~~~~~~~~~~~~~~~~~~ 1605 1606Set pool or vlan type mirror rule for a port:: 1607 1608 testpmd> set port (port_id) mirror-rule (rule_id) \ 1609 (pool-mirror-up|pool-mirror-down|vlan-mirror) \ 1610 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off) 1611 1612Set link mirror rule for a port:: 1613 1614 testpmd> set port (port_id) mirror-rule (rule_id) \ 1615 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off) 1616 1617For example to enable mirror traffic with vlan 0,1 to pool 0:: 1618 1619 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on 1620 1621reset port - mirror rule 1622~~~~~~~~~~~~~~~~~~~~~~~~ 1623 1624Reset a mirror rule for a port:: 1625 1626 testpmd> reset port (port_id) mirror-rule (rule_id) 1627 1628set flush_rx 1629~~~~~~~~~~~~ 1630 1631Set the flush on RX streams before forwarding. 1632The default is flush ``on``. 1633Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams:: 1634 1635 testpmd> set flush_rx off 1636 1637set bypass mode 1638~~~~~~~~~~~~~~~ 1639 1640Set the bypass mode for the lowest port on bypass enabled NIC:: 1641 1642 testpmd> set bypass mode (normal|bypass|isolate) (port_id) 1643 1644set bypass event 1645~~~~~~~~~~~~~~~~ 1646 1647Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled:: 1648 1649 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \ 1650 mode (normal|bypass|isolate) (port_id) 1651 1652Where: 1653 1654* ``timeout``: Enable bypass after watchdog timeout. 1655 1656* ``os_on``: Enable bypass when OS/board is powered on. 1657 1658* ``os_off``: Enable bypass when OS/board is powered off. 1659 1660* ``power_on``: Enable bypass when power supply is turned on. 1661 1662* ``power_off``: Enable bypass when power supply is turned off. 1663 1664 1665set bypass timeout 1666~~~~~~~~~~~~~~~~~~ 1667 1668Set the bypass watchdog timeout to ``n`` seconds where 0 = instant:: 1669 1670 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32) 1671 1672show bypass config 1673~~~~~~~~~~~~~~~~~~ 1674 1675Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC:: 1676 1677 testpmd> show bypass config (port_id) 1678 1679set link up 1680~~~~~~~~~~~ 1681 1682Set link up for a port:: 1683 1684 testpmd> set link-up port (port id) 1685 1686set link down 1687~~~~~~~~~~~~~ 1688 1689Set link down for a port:: 1690 1691 testpmd> set link-down port (port id) 1692 1693E-tag set 1694~~~~~~~~~ 1695 1696Enable E-tag insertion for a VF on a port:: 1697 1698 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id) 1699 1700Disable E-tag insertion for a VF on a port:: 1701 1702 testpmd> E-tag set insertion off port (port_id) vf (vf_id) 1703 1704Enable/disable E-tag stripping on a port:: 1705 1706 testpmd> E-tag set stripping (on|off) port (port_id) 1707 1708Enable/disable E-tag based forwarding on a port:: 1709 1710 testpmd> E-tag set forwarding (on|off) port (port_id) 1711 1712ddp add 1713~~~~~~~ 1714 1715Load a dynamic device personalization (DDP) profile and store backup profile:: 1716 1717 testpmd> ddp add (port_id) (profile_path[,backup_profile_path]) 1718 1719ddp del 1720~~~~~~~ 1721 1722Delete a dynamic device personalization profile and restore backup profile:: 1723 1724 testpmd> ddp del (port_id) (backup_profile_path) 1725 1726ptype mapping 1727~~~~~~~~~~~~~ 1728 1729List all items from the ptype mapping table:: 1730 1731 testpmd> ptype mapping get (port_id) (valid_only) 1732 1733Where: 1734 1735* ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0). 1736 1737Replace a specific or a group of software defined ptype with a new one:: 1738 1739 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type) 1740 1741where: 1742 1743* ``target``: A specific software ptype or a mask to represent a group of software ptypes. 1744 1745* ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1). 1746 1747* ``pkt_type``: The new software ptype to replace the old ones. 1748 1749Update hardware defined ptype to software defined packet type mapping table:: 1750 1751 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype) 1752 1753where: 1754 1755* ``hw_ptype``: hardware ptype as the index of the ptype mapping table. 1756 1757* ``sw_ptype``: software ptype as the value of the ptype mapping table. 1758 1759Reset ptype mapping table:: 1760 1761 testpmd> ptype mapping reset (port_id) 1762 1763config per port Rx offloading 1764~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1765 1766Enable or disable a per port Rx offloading on all Rx queues of a port:: 1767 1768 testpmd> port config (port_id) rx_offload (offloading) on|off 1769 1770* ``offloading``: can be any of these offloading capability: 1771 vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro, 1772 qinq_strip, outer_ipv4_cksum, macsec_strip, 1773 header_split, vlan_filter, vlan_extend, jumbo_frame, 1774 scatter, timestamp, security, keep_crc, rss_hash 1775 1776This command should be run when the port is stopped, or else it will fail. 1777 1778config per queue Rx offloading 1779~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1780 1781Enable or disable a per queue Rx offloading only on a specific Rx queue:: 1782 1783 testpmd> port (port_id) rxq (queue_id) rx_offload (offloading) on|off 1784 1785* ``offloading``: can be any of these offloading capability: 1786 vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro, 1787 qinq_strip, outer_ipv4_cksum, macsec_strip, 1788 header_split, vlan_filter, vlan_extend, jumbo_frame, 1789 scatter, timestamp, security, keep_crc 1790 1791This command should be run when the port is stopped, or else it will fail. 1792 1793config per port Tx offloading 1794~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1795 1796Enable or disable a per port Tx offloading on all Tx queues of a port:: 1797 1798 testpmd> port config (port_id) tx_offload (offloading) on|off 1799 1800* ``offloading``: can be any of these offloading capability: 1801 vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum, 1802 sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum, 1803 qinq_insert, vxlan_tnl_tso, gre_tnl_tso, 1804 ipip_tnl_tso, geneve_tnl_tso, macsec_insert, 1805 mt_lockfree, multi_segs, mbuf_fast_free, security 1806 1807This command should be run when the port is stopped, or else it will fail. 1808 1809config per queue Tx offloading 1810~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1811 1812Enable or disable a per queue Tx offloading only on a specific Tx queue:: 1813 1814 testpmd> port (port_id) txq (queue_id) tx_offload (offloading) on|off 1815 1816* ``offloading``: can be any of these offloading capability: 1817 vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum, 1818 sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum, 1819 qinq_insert, vxlan_tnl_tso, gre_tnl_tso, 1820 ipip_tnl_tso, geneve_tnl_tso, macsec_insert, 1821 mt_lockfree, multi_segs, mbuf_fast_free, security 1822 1823This command should be run when the port is stopped, or else it will fail. 1824 1825Config VXLAN Encap outer layers 1826~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1827 1828Configure the outer layer to encapsulate a packet inside a VXLAN tunnel:: 1829 1830 set vxlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \ 1831 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) eth-src (eth-src) \ 1832 eth-dst (eth-dst) 1833 1834 set vxlan-with-vlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \ 1835 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) vlan-tci (vlan-tci) \ 1836 eth-src (eth-src) eth-dst (eth-dst) 1837 1838 set vxlan-tos-ttl ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \ 1839 udp-dst (udp-dst) ip-tos (ip-tos) ip-ttl (ip-ttl) ip-src (ip-src) \ 1840 ip-dst (ip-dst) eth-src (eth-src) eth-dst (eth-dst) 1841 1842These commands will set an internal configuration inside testpmd, any following 1843flow rule using the action vxlan_encap will use the last configuration set. 1844To have a different encapsulation header, one of those commands must be called 1845before the flow rule creation. 1846 1847Config NVGRE Encap outer layers 1848~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1849 1850Configure the outer layer to encapsulate a packet inside a NVGRE tunnel:: 1851 1852 set nvgre ip-version (ipv4|ipv6) tni (tni) ip-src (ip-src) ip-dst (ip-dst) \ 1853 eth-src (eth-src) eth-dst (eth-dst) 1854 set nvgre-with-vlan ip-version (ipv4|ipv6) tni (tni) ip-src (ip-src) \ 1855 ip-dst (ip-dst) vlan-tci (vlan-tci) eth-src (eth-src) eth-dst (eth-dst) 1856 1857These commands will set an internal configuration inside testpmd, any following 1858flow rule using the action nvgre_encap will use the last configuration set. 1859To have a different encapsulation header, one of those commands must be called 1860before the flow rule creation. 1861 1862Config L2 Encap 1863~~~~~~~~~~~~~~~ 1864 1865Configure the l2 to be used when encapsulating a packet with L2:: 1866 1867 set l2_encap ip-version (ipv4|ipv6) eth-src (eth-src) eth-dst (eth-dst) 1868 set l2_encap-with-vlan ip-version (ipv4|ipv6) vlan-tci (vlan-tci) \ 1869 eth-src (eth-src) eth-dst (eth-dst) 1870 1871Those commands will set an internal configuration inside testpmd, any following 1872flow rule using the action l2_encap will use the last configuration set. 1873To have a different encapsulation header, one of those commands must be called 1874before the flow rule creation. 1875 1876Config L2 Decap 1877~~~~~~~~~~~~~~~ 1878 1879Configure the l2 to be removed when decapsulating a packet with L2:: 1880 1881 set l2_decap ip-version (ipv4|ipv6) 1882 set l2_decap-with-vlan ip-version (ipv4|ipv6) 1883 1884Those commands will set an internal configuration inside testpmd, any following 1885flow rule using the action l2_decap will use the last configuration set. 1886To have a different encapsulation header, one of those commands must be called 1887before the flow rule creation. 1888 1889Config MPLSoGRE Encap outer layers 1890~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1891 1892Configure the outer layer to encapsulate a packet inside a MPLSoGRE tunnel:: 1893 1894 set mplsogre_encap ip-version (ipv4|ipv6) label (label) \ 1895 ip-src (ip-src) ip-dst (ip-dst) eth-src (eth-src) eth-dst (eth-dst) 1896 set mplsogre_encap-with-vlan ip-version (ipv4|ipv6) label (label) \ 1897 ip-src (ip-src) ip-dst (ip-dst) vlan-tci (vlan-tci) \ 1898 eth-src (eth-src) eth-dst (eth-dst) 1899 1900These commands will set an internal configuration inside testpmd, any following 1901flow rule using the action mplsogre_encap will use the last configuration set. 1902To have a different encapsulation header, one of those commands must be called 1903before the flow rule creation. 1904 1905Config MPLSoGRE Decap outer layers 1906~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1907 1908Configure the outer layer to decapsulate MPLSoGRE packet:: 1909 1910 set mplsogre_decap ip-version (ipv4|ipv6) 1911 set mplsogre_decap-with-vlan ip-version (ipv4|ipv6) 1912 1913These commands will set an internal configuration inside testpmd, any following 1914flow rule using the action mplsogre_decap will use the last configuration set. 1915To have a different decapsulation header, one of those commands must be called 1916before the flow rule creation. 1917 1918Config MPLSoUDP Encap outer layers 1919~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1920 1921Configure the outer layer to encapsulate a packet inside a MPLSoUDP tunnel:: 1922 1923 set mplsoudp_encap ip-version (ipv4|ipv6) label (label) udp-src (udp-src) \ 1924 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) \ 1925 eth-src (eth-src) eth-dst (eth-dst) 1926 set mplsoudp_encap-with-vlan ip-version (ipv4|ipv6) label (label) \ 1927 udp-src (udp-src) udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) \ 1928 vlan-tci (vlan-tci) eth-src (eth-src) eth-dst (eth-dst) 1929 1930These commands will set an internal configuration inside testpmd, any following 1931flow rule using the action mplsoudp_encap will use the last configuration set. 1932To have a different encapsulation header, one of those commands must be called 1933before the flow rule creation. 1934 1935Config MPLSoUDP Decap outer layers 1936~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1937 1938Configure the outer layer to decapsulate MPLSoUDP packet:: 1939 1940 set mplsoudp_decap ip-version (ipv4|ipv6) 1941 set mplsoudp_decap-with-vlan ip-version (ipv4|ipv6) 1942 1943These commands will set an internal configuration inside testpmd, any following 1944flow rule using the action mplsoudp_decap will use the last configuration set. 1945To have a different decapsulation header, one of those commands must be called 1946before the flow rule creation. 1947 1948Config Raw Encapsulation 1949~~~~~~~~~~~~~~~~~~~~~~~~~ 1950 1951Configure the raw data to be used when encapsulating a packet by 1952rte_flow_action_raw_encap:: 1953 1954 set raw_encap {index} {item} [/ {item} [...]] / end_set 1955 1956There are multiple global buffers for ``raw_encap``, this command will set one 1957internal buffer index by ``{index}``. 1958If there is no ``{index}`` specified:: 1959 1960 set raw_encap {item} [/ {item} [...]] / end_set 1961 1962the default index ``0`` is used. 1963In order to use different encapsulating header, ``index`` must be specified 1964during the flow rule creation:: 1965 1966 testpmd> flow create 0 egress pattern eth / ipv4 / end actions 1967 raw_encap index 2 / end 1968 1969Otherwise the default index ``0`` is used. 1970 1971Config Raw Decapsulation 1972~~~~~~~~~~~~~~~~~~~~~~~~ 1973 1974Configure the raw data to be used when decapsulating a packet by 1975rte_flow_action_raw_decap:: 1976 1977 set raw_decap {index} {item} [/ {item} [...]] / end_set 1978 1979There are multiple global buffers for ``raw_decap``, this command will set 1980one internal buffer index by ``{index}``. 1981If there is no ``{index}`` specified:: 1982 1983 set raw_decap {item} [/ {item} [...]] / end_set 1984 1985the default index ``0`` is used. 1986In order to use different decapsulating header, ``index`` must be specified 1987during the flow rule creation:: 1988 1989 testpmd> flow create 0 egress pattern eth / ipv4 / end actions 1990 raw_encap index 3 / end 1991 1992Otherwise the default index ``0`` is used. 1993 1994Set fec mode 1995~~~~~~~~~~~~ 1996 1997Set fec mode for a specific port:: 1998 1999 testpmd> set port (port_id) fec_mode auto|off|rs|baser 2000 2001 2002Port Functions 2003-------------- 2004 2005The following sections show functions for configuring ports. 2006 2007.. note:: 2008 2009 Port configuration changes only become active when forwarding is started/restarted. 2010 2011port attach 2012~~~~~~~~~~~ 2013 2014Attach a port specified by pci address or virtual device args:: 2015 2016 testpmd> port attach (identifier) 2017 2018To attach a new pci device, the device should be recognized by kernel first. 2019Then it should be moved under DPDK management. 2020Finally the port can be attached to testpmd. 2021 2022For example, to move a pci device using ixgbe under DPDK management: 2023 2024.. code-block:: console 2025 2026 # Check the status of the available devices. 2027 ./usertools/dpdk-devbind.py --status 2028 2029 Network devices using DPDK-compatible driver 2030 ============================================ 2031 <none> 2032 2033 Network devices using kernel driver 2034 =================================== 2035 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused= 2036 2037 2038 # Bind the device to igb_uio. 2039 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0 2040 2041 2042 # Recheck the status of the devices. 2043 ./usertools/dpdk-devbind.py --status 2044 Network devices using DPDK-compatible driver 2045 ============================================ 2046 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused= 2047 2048To attach a port created by virtual device, above steps are not needed. 2049 2050For example, to attach a port whose pci address is 0000:0a:00.0. 2051 2052.. code-block:: console 2053 2054 testpmd> port attach 0000:0a:00.0 2055 Attaching a new port... 2056 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 2057 EAL: probe driver: 8086:10fb rte_ixgbe_pmd 2058 EAL: PCI memory mapped at 0x7f83bfa00000 2059 EAL: PCI memory mapped at 0x7f83bfa80000 2060 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5 2061 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb 2062 Port 0 is attached. Now total ports is 1 2063 Done 2064 2065For example, to attach a port created by pcap PMD. 2066 2067.. code-block:: console 2068 2069 testpmd> port attach net_pcap0 2070 Attaching a new port... 2071 PMD: Initializing pmd_pcap for net_pcap0 2072 PMD: Creating pcap-backed ethdev on numa socket 0 2073 Port 0 is attached. Now total ports is 1 2074 Done 2075 2076In this case, identifier is ``net_pcap0``. 2077This identifier format is the same as ``--vdev`` format of DPDK applications. 2078 2079For example, to re-attach a bonded port which has been previously detached, 2080the mode and slave parameters must be given. 2081 2082.. code-block:: console 2083 2084 testpmd> port attach net_bond_0,mode=0,slave=1 2085 Attaching a new port... 2086 EAL: Initializing pmd_bond for net_bond_0 2087 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0. 2088 Port 0 is attached. Now total ports is 1 2089 Done 2090 2091 2092port detach 2093~~~~~~~~~~~ 2094 2095Detach a specific port:: 2096 2097 testpmd> port detach (port_id) 2098 2099Before detaching a port, the port should be stopped and closed. 2100 2101For example, to detach a pci device port 0. 2102 2103.. code-block:: console 2104 2105 testpmd> port stop 0 2106 Stopping ports... 2107 Done 2108 testpmd> port close 0 2109 Closing ports... 2110 Done 2111 2112 testpmd> port detach 0 2113 Detaching a port... 2114 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 2115 EAL: remove driver: 8086:10fb rte_ixgbe_pmd 2116 EAL: PCI memory unmapped at 0x7f83bfa00000 2117 EAL: PCI memory unmapped at 0x7f83bfa80000 2118 Done 2119 2120 2121For example, to detach a virtual device port 0. 2122 2123.. code-block:: console 2124 2125 testpmd> port stop 0 2126 Stopping ports... 2127 Done 2128 testpmd> port close 0 2129 Closing ports... 2130 Done 2131 2132 testpmd> port detach 0 2133 Detaching a port... 2134 PMD: Closing pcap ethdev on numa socket 0 2135 Port 'net_pcap0' is detached. Now total ports is 0 2136 Done 2137 2138To remove a pci device completely from the system, first detach the port from testpmd. 2139Then the device should be moved under kernel management. 2140Finally the device can be removed using kernel pci hotplug functionality. 2141 2142For example, to move a pci device under kernel management: 2143 2144.. code-block:: console 2145 2146 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0 2147 2148 ./usertools/dpdk-devbind.py --status 2149 2150 Network devices using DPDK-compatible driver 2151 ============================================ 2152 <none> 2153 2154 Network devices using kernel driver 2155 =================================== 2156 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio 2157 2158To remove a port created by a virtual device, above steps are not needed. 2159 2160port start 2161~~~~~~~~~~ 2162 2163Start all ports or a specific port:: 2164 2165 testpmd> port start (port_id|all) 2166 2167port stop 2168~~~~~~~~~ 2169 2170Stop all ports or a specific port:: 2171 2172 testpmd> port stop (port_id|all) 2173 2174port close 2175~~~~~~~~~~ 2176 2177Close all ports or a specific port:: 2178 2179 testpmd> port close (port_id|all) 2180 2181port reset 2182~~~~~~~~~~ 2183 2184Reset all ports or a specific port:: 2185 2186 testpmd> port reset (port_id|all) 2187 2188User should stop port(s) before resetting and (re-)start after reset. 2189 2190port config - queue ring size 2191~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2192 2193Configure a rx/tx queue ring size:: 2194 2195 testpmd> port (port_id) (rxq|txq) (queue_id) ring_size (value) 2196 2197Only take effect after command that (re-)start the port or command that setup specific queue. 2198 2199port start/stop queue 2200~~~~~~~~~~~~~~~~~~~~~ 2201 2202Start/stop a rx/tx queue on a specific port:: 2203 2204 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop) 2205 2206port config - queue deferred start 2207~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2208 2209Switch on/off deferred start of a specific port queue:: 2210 2211 testpmd> port (port_id) (rxq|txq) (queue_id) deferred_start (on|off) 2212 2213port setup queue 2214~~~~~~~~~~~~~~~~~~~~~ 2215 2216Setup a rx/tx queue on a specific port:: 2217 2218 testpmd> port (port_id) (rxq|txq) (queue_id) setup 2219 2220Only take effect when port is started. 2221 2222port config - speed 2223~~~~~~~~~~~~~~~~~~~ 2224 2225Set the speed and duplex mode for all ports or a specific port:: 2226 2227 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|200000|auto) \ 2228 duplex (half|full|auto) 2229 2230port config - queues/descriptors 2231~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2232 2233Set number of queues/descriptors for rxq, txq, rxd and txd:: 2234 2235 testpmd> port config all (rxq|txq|rxd|txd) (value) 2236 2237This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options. 2238 2239port config - max-pkt-len 2240~~~~~~~~~~~~~~~~~~~~~~~~~ 2241 2242Set the maximum packet length:: 2243 2244 testpmd> port config all max-pkt-len (value) 2245 2246This is equivalent to the ``--max-pkt-len`` command-line option. 2247 2248port config - max-lro-pkt-size 2249~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2250 2251Set the maximum LRO aggregated packet size:: 2252 2253 testpmd> port config all max-lro-pkt-size (value) 2254 2255This is equivalent to the ``--max-lro-pkt-size`` command-line option. 2256 2257port config - Drop Packets 2258~~~~~~~~~~~~~~~~~~~~~~~~~~ 2259 2260Enable or disable packet drop on all RX queues of all ports when no receive buffers available:: 2261 2262 testpmd> port config all drop-en (on|off) 2263 2264Packet dropping when no receive buffers available is off by default. 2265 2266The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option. 2267 2268port config - RSS 2269~~~~~~~~~~~~~~~~~ 2270 2271Set the RSS (Receive Side Scaling) mode on or off:: 2272 2273 testpmd> port config all rss (all|default|eth|vlan|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|vxlan-gpe|l2tpv3|esp|ah|pfcp|none) 2274 2275RSS is on by default. 2276 2277The ``all`` option is equivalent to eth|vlan|ip|tcp|udp|sctp|ether|l2tpv3|esp|ah|pfcp. 2278 2279The ``default`` option enables all supported RSS types reported by device info. 2280 2281The ``none`` option is equivalent to the ``--disable-rss`` command-line option. 2282 2283port config - RSS Reta 2284~~~~~~~~~~~~~~~~~~~~~~ 2285 2286Set the RSS (Receive Side Scaling) redirection table:: 2287 2288 testpmd> port config all rss reta (hash,queue)[,(hash,queue)] 2289 2290port config - DCB 2291~~~~~~~~~~~~~~~~~ 2292 2293Set the DCB mode for an individual port:: 2294 2295 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off) 2296 2297The traffic class should be 4 or 8. 2298 2299port config - Burst 2300~~~~~~~~~~~~~~~~~~~ 2301 2302Set the number of packets per burst:: 2303 2304 testpmd> port config all burst (value) 2305 2306This is equivalent to the ``--burst`` command-line option. 2307 2308port config - Threshold 2309~~~~~~~~~~~~~~~~~~~~~~~ 2310 2311Set thresholds for TX/RX queues:: 2312 2313 testpmd> port config all (threshold) (value) 2314 2315Where the threshold type can be: 2316 2317* ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255. 2318 2319* ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255. 2320 2321* ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255. 2322 2323* ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255. 2324 2325* ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255. 2326 2327* ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255. 2328 2329* ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd. 2330 2331* ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd. 2332 2333* ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd. 2334 2335These threshold options are also available from the command-line. 2336 2337port config - E-tag 2338~~~~~~~~~~~~~~~~~~~ 2339 2340Set the value of ether-type for E-tag:: 2341 2342 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value) 2343 2344Enable/disable the E-tag support:: 2345 2346 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable) 2347 2348port config pctype mapping 2349~~~~~~~~~~~~~~~~~~~~~~~~~~ 2350 2351Reset pctype mapping table:: 2352 2353 testpmd> port config (port_id) pctype mapping reset 2354 2355Update hardware defined pctype to software defined flow type mapping table:: 2356 2357 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id) 2358 2359where: 2360 2361* ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table. 2362 2363* ``flow_type_id``: software flow type id as the index of the pctype mapping table. 2364 2365port config input set 2366~~~~~~~~~~~~~~~~~~~~~ 2367 2368Config RSS/FDIR/FDIR flexible payload input set for some pctype:: 2369 2370 testpmd> port config (port_id) pctype (pctype_id) \ 2371 (hash_inset|fdir_inset|fdir_flx_inset) \ 2372 (get|set|clear) field (field_idx) 2373 2374Clear RSS/FDIR/FDIR flexible payload input set for some pctype:: 2375 2376 testpmd> port config (port_id) pctype (pctype_id) \ 2377 (hash_inset|fdir_inset|fdir_flx_inset) clear all 2378 2379where: 2380 2381* ``pctype_id``: hardware packet classification types. 2382* ``field_idx``: hardware field index. 2383 2384port config udp_tunnel_port 2385~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2386 2387Add/remove UDP tunnel port for VXLAN/GENEVE tunneling protocols:: 2388 2389 testpmd> port config (port_id) udp_tunnel_port add|rm vxlan|geneve|vxlan-gpe (udp_port) 2390 2391port config tx_metadata 2392~~~~~~~~~~~~~~~~~~~~~~~ 2393 2394Set Tx metadata value per port. 2395testpmd will add this value to any Tx packet sent from this port:: 2396 2397 testpmd> port config (port_id) tx_metadata (value) 2398 2399port config dynf 2400~~~~~~~~~~~~~~~~ 2401 2402Set/clear dynamic flag per port. 2403testpmd will register this flag in the mbuf (same registration 2404for both Tx and Rx). Then set/clear this flag for each Tx 2405packet sent from this port. The set bit only works for Tx packet:: 2406 2407 testpmd> port config (port_id) dynf (name) (set|clear) 2408 2409port config mtu 2410~~~~~~~~~~~~~~~ 2411 2412To configure MTU(Maximum Transmission Unit) on devices using testpmd:: 2413 2414 testpmd> port config mtu (port_id) (value) 2415 2416port config rss hash key 2417~~~~~~~~~~~~~~~~~~~~~~~~ 2418 2419To configure the RSS hash key used to compute the RSS 2420hash of input [IP] packets received on port:: 2421 2422 testpmd> port config <port_id> rss-hash-key (ipv4|ipv4-frag|\ 2423 ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|\ 2424 ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|\ 2425 ipv6-other|l2-payload|ipv6-ex|ipv6-tcp-ex|\ 2426 ipv6-udp-ex <string of hex digits \ 2427 (variable length, NIC dependent)>) 2428 2429Device Functions 2430---------------- 2431 2432The following sections show functions for device operations. 2433 2434device detach 2435~~~~~~~~~~~~~ 2436 2437Detach a device specified by pci address or virtual device args:: 2438 2439 testpmd> device detach (identifier) 2440 2441Before detaching a device associated with ports, the ports should be stopped and closed. 2442 2443For example, to detach a pci device whose address is 0002:03:00.0. 2444 2445.. code-block:: console 2446 2447 testpmd> device detach 0002:03:00.0 2448 Removing a device... 2449 Port 1 is now closed 2450 EAL: Releasing pci mapped resource for 0002:03:00.0 2451 EAL: Calling pci_unmap_resource for 0002:03:00.0 at 0x218a050000 2452 EAL: Calling pci_unmap_resource for 0002:03:00.0 at 0x218c050000 2453 Device 0002:03:00.0 is detached 2454 Now total ports is 1 2455 2456For example, to detach a port created by pcap PMD. 2457 2458.. code-block:: console 2459 2460 testpmd> device detach net_pcap0 2461 Removing a device... 2462 Port 0 is now closed 2463 Device net_pcap0 is detached 2464 Now total ports is 0 2465 Done 2466 2467In this case, identifier is ``net_pcap0``. 2468This identifier format is the same as ``--vdev`` format of DPDK applications. 2469 2470Link Bonding Functions 2471---------------------- 2472 2473The Link Bonding functions make it possible to dynamically create and 2474manage link bonding devices from within testpmd interactive prompt. 2475 2476create bonded device 2477~~~~~~~~~~~~~~~~~~~~ 2478 2479Create a new bonding device:: 2480 2481 testpmd> create bonded device (mode) (socket) 2482 2483For example, to create a bonded device in mode 1 on socket 0:: 2484 2485 testpmd> create bonded device 1 0 2486 created new bonded device (port X) 2487 2488add bonding slave 2489~~~~~~~~~~~~~~~~~ 2490 2491Adds Ethernet device to a Link Bonding device:: 2492 2493 testpmd> add bonding slave (slave id) (port id) 2494 2495For example, to add Ethernet device (port 6) to a Link Bonding device (port 10):: 2496 2497 testpmd> add bonding slave 6 10 2498 2499 2500remove bonding slave 2501~~~~~~~~~~~~~~~~~~~~ 2502 2503Removes an Ethernet slave device from a Link Bonding device:: 2504 2505 testpmd> remove bonding slave (slave id) (port id) 2506 2507For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10):: 2508 2509 testpmd> remove bonding slave 6 10 2510 2511set bonding mode 2512~~~~~~~~~~~~~~~~ 2513 2514Set the Link Bonding mode of a Link Bonding device:: 2515 2516 testpmd> set bonding mode (value) (port id) 2517 2518For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3):: 2519 2520 testpmd> set bonding mode 3 10 2521 2522set bonding primary 2523~~~~~~~~~~~~~~~~~~~ 2524 2525Set an Ethernet slave device as the primary device on a Link Bonding device:: 2526 2527 testpmd> set bonding primary (slave id) (port id) 2528 2529For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10):: 2530 2531 testpmd> set bonding primary 6 10 2532 2533set bonding mac 2534~~~~~~~~~~~~~~~ 2535 2536Set the MAC address of a Link Bonding device:: 2537 2538 testpmd> set bonding mac (port id) (mac) 2539 2540For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01:: 2541 2542 testpmd> set bonding mac 10 00:00:00:00:00:01 2543 2544set bonding balance_xmit_policy 2545~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2546 2547Set the transmission policy for a Link Bonding device when it is in Balance XOR mode:: 2548 2549 testpmd> set bonding balance_xmit_policy (port_id) (l2|l23|l34) 2550 2551For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports):: 2552 2553 testpmd> set bonding balance_xmit_policy 10 l34 2554 2555 2556set bonding mon_period 2557~~~~~~~~~~~~~~~~~~~~~~ 2558 2559Set the link status monitoring polling period in milliseconds for a bonding device. 2560 2561This adds support for PMD slave devices which do not support link status interrupts. 2562When the mon_period is set to a value greater than 0 then all PMD's which do not support 2563link status ISR will be queried every polling interval to check if their link status has changed:: 2564 2565 testpmd> set bonding mon_period (port_id) (value) 2566 2567For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms:: 2568 2569 testpmd> set bonding mon_period 5 150 2570 2571 2572set bonding lacp dedicated_queue 2573~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2574 2575Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic 2576when in mode 4 (link-aggregation-802.3ad):: 2577 2578 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable) 2579 2580 2581set bonding agg_mode 2582~~~~~~~~~~~~~~~~~~~~ 2583 2584Enable one of the specific aggregators mode when in mode 4 (link-aggregation-802.3ad):: 2585 2586 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable) 2587 2588 2589show bonding config 2590~~~~~~~~~~~~~~~~~~~ 2591 2592Show the current configuration of a Link Bonding device:: 2593 2594 testpmd> show bonding config (port id) 2595 2596For example, 2597to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4) 2598in balance mode with a transmission policy of layer 2+3:: 2599 2600 testpmd> show bonding config 9 2601 Bonding mode: 2 2602 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23 2603 Slaves (3): [1 3 4] 2604 Active Slaves (3): [1 3 4] 2605 Primary: [3] 2606 2607 2608Register Functions 2609------------------ 2610 2611The Register Functions can be used to read from and write to registers on the network card referenced by a port number. 2612This is mainly useful for debugging purposes. 2613Reference should be made to the appropriate datasheet for the network card for details on the register addresses 2614and fields that can be accessed. 2615 2616read reg 2617~~~~~~~~ 2618 2619Display the value of a port register:: 2620 2621 testpmd> read reg (port_id) (address) 2622 2623For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller:: 2624 2625 testpmd> read reg 0 0xEE00 2626 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241) 2627 2628read regfield 2629~~~~~~~~~~~~~ 2630 2631Display a port register bit field:: 2632 2633 testpmd> read regfield (port_id) (address) (bit_x) (bit_y) 2634 2635For example, reading the lowest two bits from the register in the example above:: 2636 2637 testpmd> read regfield 0 0xEE00 0 1 2638 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1) 2639 2640read regbit 2641~~~~~~~~~~~ 2642 2643Display a single port register bit:: 2644 2645 testpmd> read regbit (port_id) (address) (bit_x) 2646 2647For example, reading the lowest bit from the register in the example above:: 2648 2649 testpmd> read regbit 0 0xEE00 0 2650 port 0 PCI register at offset 0xEE00: bit 0=1 2651 2652write reg 2653~~~~~~~~~ 2654 2655Set the value of a port register:: 2656 2657 testpmd> write reg (port_id) (address) (value) 2658 2659For example, to clear a register:: 2660 2661 testpmd> write reg 0 0xEE00 0x0 2662 port 0 PCI register at offset 0xEE00: 0x00000000 (0) 2663 2664write regfield 2665~~~~~~~~~~~~~~ 2666 2667Set bit field of a port register:: 2668 2669 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value) 2670 2671For example, writing to the register cleared in the example above:: 2672 2673 testpmd> write regfield 0 0xEE00 0 1 2 2674 port 0 PCI register at offset 0xEE00: 0x00000002 (2) 2675 2676write regbit 2677~~~~~~~~~~~~ 2678 2679Set single bit value of a port register:: 2680 2681 testpmd> write regbit (port_id) (address) (bit_x) (value) 2682 2683For example, to set the high bit in the register from the example above:: 2684 2685 testpmd> write regbit 0 0xEE00 31 1 2686 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658) 2687 2688Traffic Metering and Policing 2689----------------------------- 2690 2691The following section shows functions for configuring traffic metering and 2692policing on the ethernet device through the use of generic ethdev API. 2693 2694show port traffic management capability 2695~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2696 2697Show traffic metering and policing capability of the port:: 2698 2699 testpmd> show port meter cap (port_id) 2700 2701add port meter profile (srTCM rfc2967) 2702~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2703 2704Add meter profile (srTCM rfc2697) to the ethernet device:: 2705 2706 testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \ 2707 (cir) (cbs) (ebs) 2708 2709where: 2710 2711* ``profile_id``: ID for the meter profile. 2712* ``cir``: Committed Information Rate (CIR) (bytes/second). 2713* ``cbs``: Committed Burst Size (CBS) (bytes). 2714* ``ebs``: Excess Burst Size (EBS) (bytes). 2715 2716add port meter profile (trTCM rfc2968) 2717~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2718 2719Add meter profile (srTCM rfc2698) to the ethernet device:: 2720 2721 testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \ 2722 (cir) (pir) (cbs) (pbs) 2723 2724where: 2725 2726* ``profile_id``: ID for the meter profile. 2727* ``cir``: Committed information rate (bytes/second). 2728* ``pir``: Peak information rate (bytes/second). 2729* ``cbs``: Committed burst size (bytes). 2730* ``pbs``: Peak burst size (bytes). 2731 2732add port meter profile (trTCM rfc4115) 2733~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2734 2735Add meter profile (trTCM rfc4115) to the ethernet device:: 2736 2737 testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \ 2738 (cir) (eir) (cbs) (ebs) 2739 2740where: 2741 2742* ``profile_id``: ID for the meter profile. 2743* ``cir``: Committed information rate (bytes/second). 2744* ``eir``: Excess information rate (bytes/second). 2745* ``cbs``: Committed burst size (bytes). 2746* ``ebs``: Excess burst size (bytes). 2747 2748delete port meter profile 2749~~~~~~~~~~~~~~~~~~~~~~~~~ 2750 2751Delete meter profile from the ethernet device:: 2752 2753 testpmd> del port meter profile (port_id) (profile_id) 2754 2755create port meter 2756~~~~~~~~~~~~~~~~~ 2757 2758Create new meter object for the ethernet device:: 2759 2760 testpmd> create port meter (port_id) (mtr_id) (profile_id) \ 2761 (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \ 2762 (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\ 2763 (dscp_tbl_entry63)] 2764 2765where: 2766 2767* ``mtr_id``: meter object ID. 2768* ``profile_id``: ID for the meter profile. 2769* ``meter_enable``: When this parameter has a non-zero value, the meter object 2770 gets enabled at the time of creation, otherwise remains disabled. 2771* ``g_action``: Policer action for the packet with green color. 2772* ``y_action``: Policer action for the packet with yellow color. 2773* ``r_action``: Policer action for the packet with red color. 2774* ``stats_mask``: Mask of statistics counter types to be enabled for the 2775 meter object. 2776* ``shared``: When this parameter has a non-zero value, the meter object is 2777 shared by multiple flows. Otherwise, meter object is used by single flow. 2778* ``use_pre_meter_color``: When this parameter has a non-zero value, the 2779 input color for the current meter object is determined by the latest meter 2780 object in the same flow. Otherwise, the current meter object uses the 2781 *dscp_table* to determine the input color. 2782* ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input 2783 color, 0 <= x <= 63. 2784 2785enable port meter 2786~~~~~~~~~~~~~~~~~ 2787 2788Enable meter for the ethernet device:: 2789 2790 testpmd> enable port meter (port_id) (mtr_id) 2791 2792disable port meter 2793~~~~~~~~~~~~~~~~~~ 2794 2795Disable meter for the ethernet device:: 2796 2797 testpmd> disable port meter (port_id) (mtr_id) 2798 2799delete port meter 2800~~~~~~~~~~~~~~~~~ 2801 2802Delete meter for the ethernet device:: 2803 2804 testpmd> del port meter (port_id) (mtr_id) 2805 2806Set port meter profile 2807~~~~~~~~~~~~~~~~~~~~~~ 2808 2809Set meter profile for the ethernet device:: 2810 2811 testpmd> set port meter profile (port_id) (mtr_id) (profile_id) 2812 2813set port meter dscp table 2814~~~~~~~~~~~~~~~~~~~~~~~~~ 2815 2816Set meter dscp table for the ethernet device:: 2817 2818 testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \ 2819 (dscp_tbl_entry1)...(dscp_tbl_entry63)] 2820 2821set port meter policer action 2822~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2823 2824Set meter policer action for the ethernet device:: 2825 2826 testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \ 2827 (action0) [(action1) (action1)] 2828 2829where: 2830 2831* ``action_mask``: Bit mask indicating which policer actions need to be 2832 updated. One or more policer actions can be updated in a single function 2833 invocation. To update the policer action associated with color C, bit 2834 (1 << C) needs to be set in *action_mask* and element at position C 2835 in the *actions* array needs to be valid. 2836* ``actionx``: Policer action for the color x, 2837 RTE_MTR_GREEN <= x < RTE_MTR_COLORS 2838 2839set port meter stats mask 2840~~~~~~~~~~~~~~~~~~~~~~~~~ 2841 2842Set meter stats mask for the ethernet device:: 2843 2844 testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask) 2845 2846where: 2847 2848* ``stats_mask``: Bit mask indicating statistics counter types to be enabled. 2849 2850show port meter stats 2851~~~~~~~~~~~~~~~~~~~~~ 2852 2853Show meter stats of the ethernet device:: 2854 2855 testpmd> show port meter stats (port_id) (mtr_id) (clear) 2856 2857where: 2858 2859* ``clear``: Flag that indicates whether the statistics counters should 2860 be cleared (i.e. set to zero) immediately after they have been read or not. 2861 2862Traffic Management 2863------------------ 2864 2865The following section shows functions for configuring traffic management on 2866the ethernet device through the use of generic TM API. 2867 2868show port traffic management capability 2869~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2870 2871Show traffic management capability of the port:: 2872 2873 testpmd> show port tm cap (port_id) 2874 2875show port traffic management capability (hierarchy level) 2876~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2877 2878Show traffic management hierarchy level capability of the port:: 2879 2880 testpmd> show port tm level cap (port_id) (level_id) 2881 2882show port traffic management capability (hierarchy node level) 2883~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2884 2885Show the traffic management hierarchy node capability of the port:: 2886 2887 testpmd> show port tm node cap (port_id) (node_id) 2888 2889show port traffic management hierarchy node type 2890~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2891 2892Show the port traffic management hierarchy node type:: 2893 2894 testpmd> show port tm node type (port_id) (node_id) 2895 2896show port traffic management hierarchy node stats 2897~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2898 2899Show the port traffic management hierarchy node statistics:: 2900 2901 testpmd> show port tm node stats (port_id) (node_id) (clear) 2902 2903where: 2904 2905* ``clear``: When this parameter has a non-zero value, the statistics counters 2906 are cleared (i.e. set to zero) immediately after they have been read, 2907 otherwise the statistics counters are left untouched. 2908 2909Add port traffic management private shaper profile 2910~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2911 2912Add the port traffic management private shaper profile:: 2913 2914 testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \ 2915 (cmit_tb_rate) (cmit_tb_size) (peak_tb_rate) (peak_tb_size) \ 2916 (packet_length_adjust) (packet_mode) 2917 2918where: 2919 2920* ``shaper_profile id``: Shaper profile ID for the new profile. 2921* ``cmit_tb_rate``: Committed token bucket rate (bytes per second or packets per second). 2922* ``cmit_tb_size``: Committed token bucket size (bytes or packets). 2923* ``peak_tb_rate``: Peak token bucket rate (bytes per second or packets per second). 2924* ``peak_tb_size``: Peak token bucket size (bytes or packets). 2925* ``packet_length_adjust``: The value (bytes) to be added to the length of 2926 each packet for the purpose of shaping. This parameter value can be used to 2927 correct the packet length with the framing overhead bytes that are consumed 2928 on the wire. 2929* ``packet_mode``: Shaper configured in packet mode. This parameter value if 2930 zero, configures shaper in byte mode and if non-zero configures it in packet 2931 mode. 2932 2933Delete port traffic management private shaper profile 2934~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2935 2936Delete the port traffic management private shaper:: 2937 2938 testpmd> del port tm node shaper profile (port_id) (shaper_profile_id) 2939 2940where: 2941 2942* ``shaper_profile id``: Shaper profile ID that needs to be deleted. 2943 2944Add port traffic management shared shaper 2945~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2946 2947Create the port traffic management shared shaper:: 2948 2949 testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \ 2950 (shaper_profile_id) 2951 2952where: 2953 2954* ``shared_shaper_id``: Shared shaper ID to be created. 2955* ``shaper_profile id``: Shaper profile ID for shared shaper. 2956 2957Set port traffic management shared shaper 2958~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2959 2960Update the port traffic management shared shaper:: 2961 2962 testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \ 2963 (shaper_profile_id) 2964 2965where: 2966 2967* ``shared_shaper_id``: Shared shaper ID to be update. 2968* ``shaper_profile id``: Shaper profile ID for shared shaper. 2969 2970Delete port traffic management shared shaper 2971~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2972 2973Delete the port traffic management shared shaper:: 2974 2975 testpmd> del port tm node shared shaper (port_id) (shared_shaper_id) 2976 2977where: 2978 2979* ``shared_shaper_id``: Shared shaper ID to be deleted. 2980 2981Set port traffic management hierarchy node private shaper 2982~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2983 2984set the port traffic management hierarchy node private shaper:: 2985 2986 testpmd> set port tm node shaper profile (port_id) (node_id) \ 2987 (shaper_profile_id) 2988 2989where: 2990 2991* ``shaper_profile id``: Private shaper profile ID to be enabled on the 2992 hierarchy node. 2993 2994Add port traffic management WRED profile 2995~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2996 2997Create a new WRED profile:: 2998 2999 testpmd> add port tm node wred profile (port_id) (wred_profile_id) \ 3000 (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \ 3001 (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \ 3002 (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r) 3003 3004where: 3005 3006* ``wred_profile id``: Identifier for the newly create WRED profile 3007* ``color_g``: Packet color (green) 3008* ``min_th_g``: Minimum queue threshold for packet with green color 3009* ``max_th_g``: Minimum queue threshold for packet with green color 3010* ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp) 3011* ``wq_log2_g``: Negated log2 of queue weight (wq) 3012* ``color_y``: Packet color (yellow) 3013* ``min_th_y``: Minimum queue threshold for packet with yellow color 3014* ``max_th_y``: Minimum queue threshold for packet with yellow color 3015* ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp) 3016* ``wq_log2_y``: Negated log2 of queue weight (wq) 3017* ``color_r``: Packet color (red) 3018* ``min_th_r``: Minimum queue threshold for packet with yellow color 3019* ``max_th_r``: Minimum queue threshold for packet with yellow color 3020* ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp) 3021* ``wq_log2_r``: Negated log2 of queue weight (wq) 3022 3023Delete port traffic management WRED profile 3024~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 3025 3026Delete the WRED profile:: 3027 3028 testpmd> del port tm node wred profile (port_id) (wred_profile_id) 3029 3030Add port traffic management hierarchy nonleaf node 3031~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 3032 3033Add nonleaf node to port traffic management hierarchy:: 3034 3035 testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \ 3036 (priority) (weight) (level_id) (shaper_profile_id) \ 3037 (n_sp_priorities) (stats_mask) (n_shared_shapers) \ 3038 [(shared_shaper_0) (shared_shaper_1) ...] \ 3039 3040where: 3041 3042* ``parent_node_id``: Node ID of the parent. 3043* ``priority``: Node priority (highest node priority is zero). This is used by 3044 the SP algorithm running on the parent node for scheduling this node. 3045* ``weight``: Node weight (lowest weight is one). The node weight is relative 3046 to the weight sum of all siblings that have the same priority. It is used by 3047 the WFQ algorithm running on the parent node for scheduling this node. 3048* ``level_id``: Hierarchy level of the node. 3049* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by 3050 the node. 3051* ``n_sp_priorities``: Number of strict priorities. 3052* ``stats_mask``: Mask of statistics counter types to be enabled for this node. 3053* ``n_shared_shapers``: Number of shared shapers. 3054* ``shared_shaper_id``: Shared shaper id. 3055 3056Add port traffic management hierarchy nonleaf node with packet mode 3057~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 3058 3059Add nonleaf node with packet mode to port traffic management hierarchy:: 3060 3061 testpmd> add port tm nonleaf node pktmode (port_id) (node_id) (parent_node_id) \ 3062 (priority) (weight) (level_id) (shaper_profile_id) \ 3063 (n_sp_priorities) (stats_mask) (n_shared_shapers) \ 3064 [(shared_shaper_0) (shared_shaper_1) ...] \ 3065 3066where: 3067 3068* ``parent_node_id``: Node ID of the parent. 3069* ``priority``: Node priority (highest node priority is zero). This is used by 3070 the SP algorithm running on the parent node for scheduling this node. 3071* ``weight``: Node weight (lowest weight is one). The node weight is relative 3072 to the weight sum of all siblings that have the same priority. It is used by 3073 the WFQ algorithm running on the parent node for scheduling this node. 3074* ``level_id``: Hierarchy level of the node. 3075* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by 3076 the node. 3077* ``n_sp_priorities``: Number of strict priorities. Packet mode is enabled on 3078 all of them. 3079* ``stats_mask``: Mask of statistics counter types to be enabled for this node. 3080* ``n_shared_shapers``: Number of shared shapers. 3081* ``shared_shaper_id``: Shared shaper id. 3082 3083Add port traffic management hierarchy leaf node 3084~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 3085 3086Add leaf node to port traffic management hierarchy:: 3087 3088 testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \ 3089 (priority) (weight) (level_id) (shaper_profile_id) \ 3090 (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \ 3091 [(shared_shaper_id) (shared_shaper_id) ...] \ 3092 3093where: 3094 3095* ``parent_node_id``: Node ID of the parent. 3096* ``priority``: Node priority (highest node priority is zero). This is used by 3097 the SP algorithm running on the parent node for scheduling this node. 3098* ``weight``: Node weight (lowest weight is one). The node weight is relative 3099 to the weight sum of all siblings that have the same priority. It is used by 3100 the WFQ algorithm running on the parent node for scheduling this node. 3101* ``level_id``: Hierarchy level of the node. 3102* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by 3103 the node. 3104* ``cman_mode``: Congestion management mode to be enabled for this node. 3105* ``wred_profile_id``: WRED profile id to be enabled for this node. 3106* ``stats_mask``: Mask of statistics counter types to be enabled for this node. 3107* ``n_shared_shapers``: Number of shared shapers. 3108* ``shared_shaper_id``: Shared shaper id. 3109 3110Delete port traffic management hierarchy node 3111~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 3112 3113Delete node from port traffic management hierarchy:: 3114 3115 testpmd> del port tm node (port_id) (node_id) 3116 3117Update port traffic management hierarchy parent node 3118~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 3119 3120Update port traffic management hierarchy parent node:: 3121 3122 testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \ 3123 (priority) (weight) 3124 3125This function can only be called after the hierarchy commit invocation. Its 3126success depends on the port support for this operation, as advertised through 3127the port capability set. This function is valid for all nodes of the traffic 3128management hierarchy except root node. 3129 3130Suspend port traffic management hierarchy node 3131~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 3132 3133 testpmd> suspend port tm node (port_id) (node_id) 3134 3135Resume port traffic management hierarchy node 3136~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 3137 3138 testpmd> resume port tm node (port_id) (node_id) 3139 3140Commit port traffic management hierarchy 3141~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 3142 3143Commit the traffic management hierarchy on the port:: 3144 3145 testpmd> port tm hierarchy commit (port_id) (clean_on_fail) 3146 3147where: 3148 3149* ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function 3150 call failure. On the other hand, hierarchy is preserved when this parameter 3151 is equal to zero. 3152 3153Set port traffic management mark VLAN dei 3154~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 3155 3156Enables/Disables the traffic management marking on the port for VLAN packets:: 3157 3158 testpmd> set port tm mark vlan_dei <port_id> <green> <yellow> <red> 3159 3160where: 3161 3162* ``port_id``: The port which on which VLAN packets marked as ``green`` or 3163 ``yellow`` or ``red`` will have dei bit enabled 3164 3165* ``green`` enable 1, disable 0 marking for dei bit of VLAN packets marked as green 3166 3167* ``yellow`` enable 1, disable 0 marking for dei bit of VLAN packets marked as yellow 3168 3169* ``red`` enable 1, disable 0 marking for dei bit of VLAN packets marked as red 3170 3171Set port traffic management mark IP dscp 3172~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 3173 3174Enables/Disables the traffic management marking on the port for IP dscp packets:: 3175 3176 testpmd> set port tm mark ip_dscp <port_id> <green> <yellow> <red> 3177 3178where: 3179 3180* ``port_id``: The port which on which IP packets marked as ``green`` or 3181 ``yellow`` or ``red`` will have IP dscp bits updated 3182 3183* ``green`` enable 1, disable 0 marking IP dscp to low drop precedence for green packets 3184 3185* ``yellow`` enable 1, disable 0 marking IP dscp to medium drop precedence for yellow packets 3186 3187* ``red`` enable 1, disable 0 marking IP dscp to high drop precedence for red packets 3188 3189Set port traffic management mark IP ecn 3190~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 3191 3192Enables/Disables the traffic management marking on the port for IP ecn packets:: 3193 3194 testpmd> set port tm mark ip_ecn <port_id> <green> <yellow> <red> 3195 3196where: 3197 3198* ``port_id``: The port which on which IP packets marked as ``green`` or 3199 ``yellow`` or ``red`` will have IP ecn bits updated 3200 3201* ``green`` enable 1, disable 0 marking IP ecn for green marked packets with ecn of 2'b01 or 2'b10 3202 to ecn of 2'b11 when IP is caring TCP or SCTP 3203 3204* ``yellow`` enable 1, disable 0 marking IP ecn for yellow marked packets with ecn of 2'b01 or 2'b10 3205 to ecn of 2'b11 when IP is caring TCP or SCTP 3206 3207* ``red`` enable 1, disable 0 marking IP ecn for yellow marked packets with ecn of 2'b01 or 2'b10 3208 to ecn of 2'b11 when IP is caring TCP or SCTP 3209 3210Filter Functions 3211---------------- 3212 3213This section details the available filter functions that are available. 3214 3215Note these functions interface the deprecated legacy filtering framework, 3216superseded by *rte_flow*. See `Flow rules management`_. 3217 3218.. _testpmd_flow_director: 3219 3220flow_director_filter 3221~~~~~~~~~~~~~~~~~~~~ 3222 3223The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues. 3224 3225Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and 3226Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter: 3227 3228* Perfect match filters. 3229 The hardware checks a match between the masked fields of the received packets and the programmed filters. 3230 The masked fields are for IP flow. 3231 3232* Signature filters. 3233 The hardware checks a match between a hash-based signature of the masked fields of the received packet. 3234 3235* Perfect-mac-vlan match filters. 3236 The hardware checks a match between the masked fields of the received packets and the programmed filters. 3237 The masked fields are for MAC VLAN flow. 3238 3239* Perfect-tunnel match filters. 3240 The hardware checks a match between the masked fields of the received packets and the programmed filters. 3241 The masked fields are for tunnel flow. 3242 3243* Perfect-raw-flow-type match filters. 3244 The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet. 3245 The masked fields are specified by input sets. 3246 3247The Flow Director filters can match the different fields for different type of packet: flow type, specific input set 3248per flow type and the flexible payload. 3249 3250The Flow Director can also mask out parts of all of these fields so that filters 3251are only applied to certain fields or parts of the fields. 3252 3253Note that for raw flow type mode the source and destination fields in the 3254raw packet buffer need to be presented in a reversed order with respect 3255to the expected received packets. 3256For example: IP source and destination addresses or TCP/UDP/SCTP 3257source and destination ports 3258 3259Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information. 3260 3261# Commands to add flow director filters of different flow types:: 3262 3263 flow_director_filter (port_id) mode IP (add|del|update) \ 3264 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \ 3265 src (src_ip_address) dst (dst_ip_address) \ 3266 tos (tos_value) proto (proto_value) ttl (ttl_value) \ 3267 vlan (vlan_value) flexbytes (flexbytes_value) \ 3268 (drop|fwd) pf|vf(vf_id) queue (queue_id) \ 3269 fd_id (fd_id_value) 3270 3271 flow_director_filter (port_id) mode IP (add|del|update) \ 3272 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \ 3273 src (src_ip_address) (src_port) \ 3274 dst (dst_ip_address) (dst_port) \ 3275 tos (tos_value) ttl (ttl_value) \ 3276 vlan (vlan_value) flexbytes (flexbytes_value) \ 3277 (drop|fwd) queue pf|vf(vf_id) (queue_id) \ 3278 fd_id (fd_id_value) 3279 3280 flow_director_filter (port_id) mode IP (add|del|update) \ 3281 flow (ipv4-sctp|ipv6-sctp) \ 3282 src (src_ip_address) (src_port) \ 3283 dst (dst_ip_address) (dst_port) \ 3284 tos (tos_value) ttl (ttl_value) \ 3285 tag (verification_tag) vlan (vlan_value) \ 3286 flexbytes (flexbytes_value) (drop|fwd) \ 3287 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value) 3288 3289 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \ 3290 ether (ethertype) flexbytes (flexbytes_value) \ 3291 (drop|fwd) pf|vf(vf_id) queue (queue_id) 3292 fd_id (fd_id_value) 3293 3294 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \ 3295 mac (mac_address) vlan (vlan_value) \ 3296 flexbytes (flexbytes_value) (drop|fwd) \ 3297 queue (queue_id) fd_id (fd_id_value) 3298 3299 flow_director_filter (port_id) mode Tunnel (add|del|update) \ 3300 mac (mac_address) vlan (vlan_value) \ 3301 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \ 3302 flexbytes (flexbytes_value) (drop|fwd) \ 3303 queue (queue_id) fd_id (fd_id_value) 3304 3305 flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \ 3306 (drop|fwd) queue (queue_id) fd_id (fd_id_value) \ 3307 packet (packet file name) 3308 3309For example, to add an ipv4-udp flow type filter:: 3310 3311 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \ 3312 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \ 3313 fwd pf queue 1 fd_id 1 3314 3315For example, add an ipv4-other flow type filter:: 3316 3317 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \ 3318 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \ 3319 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1 3320 3321flush_flow_director 3322~~~~~~~~~~~~~~~~~~~ 3323 3324Flush all flow director filters on a device:: 3325 3326 testpmd> flush_flow_director (port_id) 3327 3328Example, to flush all flow director filter on port 0:: 3329 3330 testpmd> flush_flow_director 0 3331 3332flow_director_mask 3333~~~~~~~~~~~~~~~~~~ 3334 3335Set flow director's input masks:: 3336 3337 flow_director_mask (port_id) mode IP vlan (vlan_value) \ 3338 src_mask (ipv4_src) (ipv6_src) (src_port) \ 3339 dst_mask (ipv4_dst) (ipv6_dst) (dst_port) 3340 3341 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value) 3342 3343 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \ 3344 mac (mac_value) tunnel-type (tunnel_type_value) \ 3345 tunnel-id (tunnel_id_value) 3346 3347Example, to set flow director mask on port 0:: 3348 3349 testpmd> flow_director_mask 0 mode IP vlan 0xefff \ 3350 src_mask 255.255.255.255 \ 3351 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \ 3352 dst_mask 255.255.255.255 \ 3353 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF 3354 3355flow_director_flex_mask 3356~~~~~~~~~~~~~~~~~~~~~~~ 3357 3358set masks of flow director's flexible payload based on certain flow type:: 3359 3360 testpmd> flow_director_flex_mask (port_id) \ 3361 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 3362 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \ 3363 l2_payload|all) (mask) 3364 3365Example, to set flow director's flex mask for all flow type on port 0:: 3366 3367 testpmd> flow_director_flex_mask 0 flow all \ 3368 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0) 3369 3370 3371flow_director_flex_payload 3372~~~~~~~~~~~~~~~~~~~~~~~~~~ 3373 3374Configure flexible payload selection:: 3375 3376 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config) 3377 3378For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload:: 3379 3380 testpmd> flow_director_flex_payload 0 l4 \ 3381 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19) 3382 3383set_fdir_input_set 3384~~~~~~~~~~~~~~~~~~ 3385 3386The Flow Director filters can match the different fields for different type of packet, i.e. specific input set 3387on per flow type and the flexible payload. This command can be used to change input set for each flow type. 3388 3389Set the input set for flow director:: 3390 3391 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 3392 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 3393 l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \ 3394 ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \ 3395 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \ 3396 sctp-dst-port|sctp-veri-tag|none) (select|add) 3397 3398For example to add source IP to FD input set for flow type of ipv4-udp on port 0:: 3399 3400 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add 3401 3402 3403.. _testpmd_rte_flow: 3404 3405Flow rules management 3406--------------------- 3407 3408Control of the generic flow API (*rte_flow*) is fully exposed through the 3409``flow`` command (validation, creation, destruction, queries and operation 3410modes). 3411 3412Considering *rte_flow* overlaps with all `Filter Functions`_, using both 3413features simultaneously may cause undefined side-effects and is therefore 3414not recommended. 3415 3416``flow`` syntax 3417~~~~~~~~~~~~~~~ 3418 3419Because the ``flow`` command uses dynamic tokens to handle the large number 3420of possible flow rules combinations, its behavior differs slightly from 3421other commands, in particular: 3422 3423- Pressing *?* or the *<tab>* key displays contextual help for the current 3424 token, not that of the entire command. 3425 3426- Optional and repeated parameters are supported (provided they are listed 3427 in the contextual help). 3428 3429The first parameter stands for the operation mode. Possible operations and 3430their general syntax are described below. They are covered in detail in the 3431following sections. 3432 3433- Check whether a flow rule can be created:: 3434 3435 flow validate {port_id} 3436 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3437 pattern {item} [/ {item} [...]] / end 3438 actions {action} [/ {action} [...]] / end 3439 3440- Create a flow rule:: 3441 3442 flow create {port_id} 3443 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3444 pattern {item} [/ {item} [...]] / end 3445 actions {action} [/ {action} [...]] / end 3446 3447- Destroy specific flow rules:: 3448 3449 flow destroy {port_id} rule {rule_id} [...] 3450 3451- Destroy all flow rules:: 3452 3453 flow flush {port_id} 3454 3455- Query an existing flow rule:: 3456 3457 flow query {port_id} {rule_id} {action} 3458 3459- List existing flow rules sorted by priority, filtered by group 3460 identifiers:: 3461 3462 flow list {port_id} [group {group_id}] [...] 3463 3464- Restrict ingress traffic to the defined flow rules:: 3465 3466 flow isolate {port_id} {boolean} 3467 3468- Dump internal representation information of all flows in hardware:: 3469 3470 flow dump {port_id} {output_file} 3471 3472- List and destroy aged flow rules:: 3473 3474 flow aged {port_id} [destroy] 3475 3476- Tunnel offload - create a tunnel stub:: 3477 3478 flow tunnel create {port_id} type {tunnel_type} 3479 3480- Tunnel offload - destroy a tunnel stub:: 3481 3482 flow tunnel destroy {port_id} id {tunnel_id} 3483 3484- Tunnel offload - list port tunnel stubs:: 3485 3486 flow tunnel list {port_id} 3487 3488Creating a tunnel stub for offload 3489~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 3490 3491``flow tunnel create`` setup a tunnel stub for tunnel offload flow rules:: 3492 3493 flow tunnel create {port_id} type {tunnel_type} 3494 3495If successful, it will return a tunnel stub ID usable with other commands:: 3496 3497 port [...]: flow tunnel #[...] type [...] 3498 3499Tunnel stub ID is relative to a port. 3500 3501Destroying tunnel offload stub 3502~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 3503 3504``flow tunnel destroy`` destroy port tunnel stub:: 3505 3506 flow tunnel destroy {port_id} id {tunnel_id} 3507 3508Listing tunnel offload stubs 3509~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 3510 3511``flow tunnel list`` list port tunnel offload stubs:: 3512 3513 flow tunnel list {port_id} 3514 3515Validating flow rules 3516~~~~~~~~~~~~~~~~~~~~~ 3517 3518``flow validate`` reports whether a flow rule would be accepted by the 3519underlying device in its current state but stops short of creating it. It is 3520bound to ``rte_flow_validate()``:: 3521 3522 flow validate {port_id} 3523 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3524 pattern {item} [/ {item} [...]] / end 3525 actions {action} [/ {action} [...]] / end 3526 3527If successful, it will show:: 3528 3529 Flow rule validated 3530 3531Otherwise it will show an error message of the form:: 3532 3533 Caught error type [...] ([...]): [...] 3534 3535This command uses the same parameters as ``flow create``, their format is 3536described in `Creating flow rules`_. 3537 3538Check whether redirecting any Ethernet packet received on port 0 to RX queue 3539index 6 is supported:: 3540 3541 testpmd> flow validate 0 ingress pattern eth / end 3542 actions queue index 6 / end 3543 Flow rule validated 3544 testpmd> 3545 3546Port 0 does not support TCPv6 rules:: 3547 3548 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end 3549 actions drop / end 3550 Caught error type 9 (specific pattern item): Invalid argument 3551 testpmd> 3552 3553Creating flow rules 3554~~~~~~~~~~~~~~~~~~~ 3555 3556``flow create`` validates and creates the specified flow rule. It is bound 3557to ``rte_flow_create()``:: 3558 3559 flow create {port_id} 3560 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3561 [tunnel_set {tunnel_id}] [tunnel_match {tunnel_id}] 3562 pattern {item} [/ {item} [...]] / end 3563 actions {action} [/ {action} [...]] / end 3564 3565If successful, it will return a flow rule ID usable with other commands:: 3566 3567 Flow rule #[...] created 3568 3569Otherwise it will show an error message of the form:: 3570 3571 Caught error type [...] ([...]): [...] 3572 3573Parameters describe in the following order: 3574 3575- Attributes (*group*, *priority*, *ingress*, *egress*, *transfer* tokens). 3576- Tunnel offload specification (tunnel_set, tunnel_match) 3577- A matching pattern, starting with the *pattern* token and terminated by an 3578 *end* pattern item. 3579- Actions, starting with the *actions* token and terminated by an *end* 3580 action. 3581 3582These translate directly to *rte_flow* objects provided as-is to the 3583underlying functions. 3584 3585The shortest valid definition only comprises mandatory tokens:: 3586 3587 testpmd> flow create 0 pattern end actions end 3588 3589Note that PMDs may refuse rules that essentially do nothing such as this 3590one. 3591 3592**All unspecified object values are automatically initialized to 0.** 3593 3594Attributes 3595^^^^^^^^^^ 3596 3597These tokens affect flow rule attributes (``struct rte_flow_attr``) and are 3598specified before the ``pattern`` token. 3599 3600- ``group {group id}``: priority group. 3601- ``priority {level}``: priority level within group. 3602- ``ingress``: rule applies to ingress traffic. 3603- ``egress``: rule applies to egress traffic. 3604- ``transfer``: apply rule directly to endpoints found in pattern. 3605 3606Each instance of an attribute specified several times overrides the previous 3607value as shown below (group 4 is used):: 3608 3609 testpmd> flow create 0 group 42 group 24 group 4 [...] 3610 3611Note that once enabled, ``ingress`` and ``egress`` cannot be disabled. 3612 3613While not specifying a direction is an error, some rules may allow both 3614simultaneously. 3615 3616Most rules affect RX therefore contain the ``ingress`` token:: 3617 3618 testpmd> flow create 0 ingress pattern [...] 3619 3620Tunnel offload 3621^^^^^^^^^^^^^^ 3622 3623Indicate tunnel offload rule type 3624 3625- ``tunnel_set {tunnel_id}``: mark rule as tunnel offload decap_set type. 3626- ``tunnel_match {tunnel_id}``: mark rule as tunel offload match type. 3627 3628Matching pattern 3629^^^^^^^^^^^^^^^^ 3630 3631A matching pattern starts after the ``pattern`` token. It is made of pattern 3632items and is terminated by a mandatory ``end`` item. 3633 3634Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum 3635rte_flow_item_type``). 3636 3637The ``/`` token is used as a separator between pattern items as shown 3638below:: 3639 3640 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...] 3641 3642Note that protocol items like these must be stacked from lowest to highest 3643layer to make sense. For instance, the following rule is either invalid or 3644unlikely to match any packet:: 3645 3646 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...] 3647 3648More information on these restrictions can be found in the *rte_flow* 3649documentation. 3650 3651Several items support additional specification structures, for example 3652``ipv4`` allows specifying source and destination addresses as follows:: 3653 3654 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1 3655 dst is 10.2.0.0 / end [...] 3656 3657This rule matches all IPv4 traffic with the specified properties. 3658 3659In this example, ``src`` and ``dst`` are field names of the underlying 3660``struct rte_flow_item_ipv4`` object. All item properties can be specified 3661in a similar fashion. 3662 3663The ``is`` token means that the subsequent value must be matched exactly, 3664and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item`` 3665accordingly. Possible assignment tokens are: 3666 3667- ``is``: match value perfectly (with full bit-mask). 3668- ``spec``: match value according to configured bit-mask. 3669- ``last``: specify upper bound to establish a range. 3670- ``mask``: specify bit-mask with relevant bits set to one. 3671- ``prefix``: generate bit-mask with <prefix-length> most-significant bits set to one. 3672 3673These yield identical results:: 3674 3675 ipv4 src is 10.1.1.1 3676 3677:: 3678 3679 ipv4 src spec 10.1.1.1 src mask 255.255.255.255 3680 3681:: 3682 3683 ipv4 src spec 10.1.1.1 src prefix 32 3684 3685:: 3686 3687 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value 3688 3689:: 3690 3691 ipv4 src is 10.1.1.1 src last 0 # 0 disables range 3692 3693Inclusive ranges can be defined with ``last``:: 3694 3695 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4 3696 3697Note that ``mask`` affects both ``spec`` and ``last``:: 3698 3699 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0 3700 # matches 10.1.0.0 to 10.2.255.255 3701 3702Properties can be modified multiple times:: 3703 3704 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4 3705 3706:: 3707 3708 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16 3709 3710Pattern items 3711^^^^^^^^^^^^^ 3712 3713This section lists supported pattern items and their attributes, if any. 3714 3715- ``end``: end list of pattern items. 3716 3717- ``void``: no-op pattern item. 3718 3719- ``invert``: perform actions when pattern does not match. 3720 3721- ``any``: match any protocol for the current layer. 3722 3723 - ``num {unsigned}``: number of layers covered. 3724 3725- ``pf``: match traffic from/to the physical function. 3726 3727- ``vf``: match traffic from/to a virtual function ID. 3728 3729 - ``id {unsigned}``: VF ID. 3730 3731- ``phy_port``: match traffic from/to a specific physical port. 3732 3733 - ``index {unsigned}``: physical port index. 3734 3735- ``port_id``: match traffic from/to a given DPDK port ID. 3736 3737 - ``id {unsigned}``: DPDK port ID. 3738 3739- ``mark``: match value set in previously matched flow rule using the mark action. 3740 3741 - ``id {unsigned}``: arbitrary integer value. 3742 3743- ``raw``: match an arbitrary byte string. 3744 3745 - ``relative {boolean}``: look for pattern after the previous item. 3746 - ``search {boolean}``: search pattern from offset (see also limit). 3747 - ``offset {integer}``: absolute or relative offset for pattern. 3748 - ``limit {unsigned}``: search area limit for start of pattern. 3749 - ``pattern {string}``: byte string to look for. 3750 3751- ``eth``: match Ethernet header. 3752 3753 - ``dst {MAC-48}``: destination MAC. 3754 - ``src {MAC-48}``: source MAC. 3755 - ``type {unsigned}``: EtherType or TPID. 3756 3757- ``vlan``: match 802.1Q/ad VLAN tag. 3758 3759 - ``tci {unsigned}``: tag control information. 3760 - ``pcp {unsigned}``: priority code point. 3761 - ``dei {unsigned}``: drop eligible indicator. 3762 - ``vid {unsigned}``: VLAN identifier. 3763 - ``inner_type {unsigned}``: inner EtherType or TPID. 3764 3765- ``ipv4``: match IPv4 header. 3766 3767 - ``tos {unsigned}``: type of service. 3768 - ``ttl {unsigned}``: time to live. 3769 - ``proto {unsigned}``: next protocol ID. 3770 - ``src {ipv4 address}``: source address. 3771 - ``dst {ipv4 address}``: destination address. 3772 3773- ``ipv6``: match IPv6 header. 3774 3775 - ``tc {unsigned}``: traffic class. 3776 - ``flow {unsigned}``: flow label. 3777 - ``proto {unsigned}``: protocol (next header). 3778 - ``hop {unsigned}``: hop limit. 3779 - ``src {ipv6 address}``: source address. 3780 - ``dst {ipv6 address}``: destination address. 3781 3782- ``icmp``: match ICMP header. 3783 3784 - ``type {unsigned}``: ICMP packet type. 3785 - ``code {unsigned}``: ICMP packet code. 3786 3787- ``udp``: match UDP header. 3788 3789 - ``src {unsigned}``: UDP source port. 3790 - ``dst {unsigned}``: UDP destination port. 3791 3792- ``tcp``: match TCP header. 3793 3794 - ``src {unsigned}``: TCP source port. 3795 - ``dst {unsigned}``: TCP destination port. 3796 3797- ``sctp``: match SCTP header. 3798 3799 - ``src {unsigned}``: SCTP source port. 3800 - ``dst {unsigned}``: SCTP destination port. 3801 - ``tag {unsigned}``: validation tag. 3802 - ``cksum {unsigned}``: checksum. 3803 3804- ``vxlan``: match VXLAN header. 3805 3806 - ``vni {unsigned}``: VXLAN identifier. 3807 3808- ``e_tag``: match IEEE 802.1BR E-Tag header. 3809 3810 - ``grp_ecid_b {unsigned}``: GRP and E-CID base. 3811 3812- ``nvgre``: match NVGRE header. 3813 3814 - ``tni {unsigned}``: virtual subnet ID. 3815 3816- ``mpls``: match MPLS header. 3817 3818 - ``label {unsigned}``: MPLS label. 3819 3820- ``gre``: match GRE header. 3821 3822 - ``protocol {unsigned}``: protocol type. 3823 3824- ``gre_key``: match GRE optional key field. 3825 3826 - ``value {unsigned}``: key value. 3827 3828- ``fuzzy``: fuzzy pattern match, expect faster than default. 3829 3830 - ``thresh {unsigned}``: accuracy threshold. 3831 3832- ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header. 3833 3834 - ``teid {unsigned}``: tunnel endpoint identifier. 3835 3836- ``geneve``: match GENEVE header. 3837 3838 - ``vni {unsigned}``: virtual network identifier. 3839 - ``protocol {unsigned}``: protocol type. 3840 3841- ``vxlan-gpe``: match VXLAN-GPE header. 3842 3843 - ``vni {unsigned}``: VXLAN-GPE identifier. 3844 3845- ``arp_eth_ipv4``: match ARP header for Ethernet/IPv4. 3846 3847 - ``sha {MAC-48}``: sender hardware address. 3848 - ``spa {ipv4 address}``: sender IPv4 address. 3849 - ``tha {MAC-48}``: target hardware address. 3850 - ``tpa {ipv4 address}``: target IPv4 address. 3851 3852- ``ipv6_ext``: match presence of any IPv6 extension header. 3853 3854 - ``next_hdr {unsigned}``: next header. 3855 3856- ``icmp6``: match any ICMPv6 header. 3857 3858 - ``type {unsigned}``: ICMPv6 type. 3859 - ``code {unsigned}``: ICMPv6 code. 3860 3861- ``icmp6_nd_ns``: match ICMPv6 neighbor discovery solicitation. 3862 3863 - ``target_addr {ipv6 address}``: target address. 3864 3865- ``icmp6_nd_na``: match ICMPv6 neighbor discovery advertisement. 3866 3867 - ``target_addr {ipv6 address}``: target address. 3868 3869- ``icmp6_nd_opt``: match presence of any ICMPv6 neighbor discovery option. 3870 3871 - ``type {unsigned}``: ND option type. 3872 3873- ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery source Ethernet 3874 link-layer address option. 3875 3876 - ``sla {MAC-48}``: source Ethernet LLA. 3877 3878- ``icmp6_nd_opt_tla_eth``: match ICMPv6 neighbor discovery target Ethernet 3879 link-layer address option. 3880 3881 - ``tla {MAC-48}``: target Ethernet LLA. 3882 3883- ``meta``: match application specific metadata. 3884 3885 - ``data {unsigned}``: metadata value. 3886 3887- ``gtp_psc``: match GTP PDU extension header with type 0x85. 3888 3889 - ``pdu_type {unsigned}``: PDU type. 3890 - ``qfi {unsigned}``: QoS flow identifier. 3891 3892- ``pppoes``, ``pppoed``: match PPPoE header. 3893 3894 - ``session_id {unsigned}``: session identifier. 3895 3896- ``pppoe_proto_id``: match PPPoE session protocol identifier. 3897 3898 - ``proto_id {unsigned}``: PPP protocol identifier. 3899 3900- ``l2tpv3oip``: match L2TPv3 over IP header. 3901 3902 - ``session_id {unsigned}``: L2TPv3 over IP session identifier. 3903 3904- ``ah``: match AH header. 3905 3906 - ``spi {unsigned}``: security parameters index. 3907 3908- ``pfcp``: match PFCP header. 3909 3910 - ``s_field {unsigned}``: S field. 3911 - ``seid {unsigned}``: session endpoint identifier. 3912 3913Actions list 3914^^^^^^^^^^^^ 3915 3916A list of actions starts after the ``actions`` token in the same fashion as 3917`Matching pattern`_; actions are separated by ``/`` tokens and the list is 3918terminated by a mandatory ``end`` action. 3919 3920Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum 3921rte_flow_action_type``). 3922 3923Dropping all incoming UDPv4 packets can be expressed as follows:: 3924 3925 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3926 actions drop / end 3927 3928Several actions have configurable properties which must be specified when 3929there is no valid default value. For example, ``queue`` requires a target 3930queue index. 3931 3932This rule redirects incoming UDPv4 traffic to queue index 6:: 3933 3934 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3935 actions queue index 6 / end 3936 3937While this one could be rejected by PMDs (unspecified queue index):: 3938 3939 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3940 actions queue / end 3941 3942As defined by *rte_flow*, the list is not ordered, all actions of a given 3943rule are performed simultaneously. These are equivalent:: 3944 3945 queue index 6 / void / mark id 42 / end 3946 3947:: 3948 3949 void / mark id 42 / queue index 6 / end 3950 3951All actions in a list should have different types, otherwise only the last 3952action of a given type is taken into account:: 3953 3954 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6 3955 3956:: 3957 3958 drop / drop / drop / end # drop is performed only once 3959 3960:: 3961 3962 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24 3963 3964Considering they are performed simultaneously, opposite and overlapping 3965actions can sometimes be combined when the end result is unambiguous:: 3966 3967 drop / queue index 6 / end # drop has no effect 3968 3969:: 3970 3971 queue index 6 / rss queues 6 7 8 / end # queue has no effect 3972 3973:: 3974 3975 drop / passthru / end # drop has no effect 3976 3977Note that PMDs may still refuse such combinations. 3978 3979Actions 3980^^^^^^^ 3981 3982This section lists supported actions and their attributes, if any. 3983 3984- ``end``: end list of actions. 3985 3986- ``void``: no-op action. 3987 3988- ``passthru``: let subsequent rule process matched packets. 3989 3990- ``jump``: redirect traffic to group on device. 3991 3992 - ``group {unsigned}``: group to redirect to. 3993 3994- ``mark``: attach 32 bit value to packets. 3995 3996 - ``id {unsigned}``: 32 bit value to return with packets. 3997 3998- ``flag``: flag packets. 3999 4000- ``queue``: assign packets to a given queue index. 4001 4002 - ``index {unsigned}``: queue index to use. 4003 4004- ``drop``: drop packets (note: passthru has priority). 4005 4006- ``count``: enable counters for this rule. 4007 4008- ``rss``: spread packets among several queues. 4009 4010 - ``func {hash function}``: RSS hash function to apply, allowed tokens are 4011 ``toeplitz``, ``simple_xor``, ``symmetric_toeplitz`` and ``default``. 4012 4013 - ``level {unsigned}``: encapsulation level for ``types``. 4014 4015 - ``types [{RSS hash type} [...]] end``: specific RSS hash types. 4016 Note that an empty list does not disable RSS but instead requests 4017 unspecified "best-effort" settings. 4018 4019 - ``key {string}``: RSS hash key, overrides ``key_len``. 4020 4021 - ``key_len {unsigned}``: RSS hash key length in bytes, can be used in 4022 conjunction with ``key`` to pad or truncate it. 4023 4024 - ``queues [{unsigned} [...]] end``: queue indices to use. 4025 4026- ``pf``: direct traffic to physical function. 4027 4028- ``vf``: direct traffic to a virtual function ID. 4029 4030 - ``original {boolean}``: use original VF ID if possible. 4031 - ``id {unsigned}``: VF ID. 4032 4033- ``phy_port``: direct packets to physical port index. 4034 4035 - ``original {boolean}``: use original port index if possible. 4036 - ``index {unsigned}``: physical port index. 4037 4038- ``port_id``: direct matching traffic to a given DPDK port ID. 4039 4040 - ``original {boolean}``: use original DPDK port ID if possible. 4041 - ``id {unsigned}``: DPDK port ID. 4042 4043- ``of_set_mpls_ttl``: OpenFlow's ``OFPAT_SET_MPLS_TTL``. 4044 4045 - ``mpls_ttl``: MPLS TTL. 4046 4047- ``of_dec_mpls_ttl``: OpenFlow's ``OFPAT_DEC_MPLS_TTL``. 4048 4049- ``of_set_nw_ttl``: OpenFlow's ``OFPAT_SET_NW_TTL``. 4050 4051 - ``nw_ttl``: IP TTL. 4052 4053- ``of_dec_nw_ttl``: OpenFlow's ``OFPAT_DEC_NW_TTL``. 4054 4055- ``of_copy_ttl_out``: OpenFlow's ``OFPAT_COPY_TTL_OUT``. 4056 4057- ``of_copy_ttl_in``: OpenFlow's ``OFPAT_COPY_TTL_IN``. 4058 4059- ``of_pop_vlan``: OpenFlow's ``OFPAT_POP_VLAN``. 4060 4061- ``of_push_vlan``: OpenFlow's ``OFPAT_PUSH_VLAN``. 4062 4063 - ``ethertype``: Ethertype. 4064 4065- ``of_set_vlan_vid``: OpenFlow's ``OFPAT_SET_VLAN_VID``. 4066 4067 - ``vlan_vid``: VLAN id. 4068 4069- ``of_set_vlan_pcp``: OpenFlow's ``OFPAT_SET_VLAN_PCP``. 4070 4071 - ``vlan_pcp``: VLAN priority. 4072 4073- ``of_pop_mpls``: OpenFlow's ``OFPAT_POP_MPLS``. 4074 4075 - ``ethertype``: Ethertype. 4076 4077- ``of_push_mpls``: OpenFlow's ``OFPAT_PUSH_MPLS``. 4078 4079 - ``ethertype``: Ethertype. 4080 4081- ``vxlan_encap``: Performs a VXLAN encapsulation, outer layer configuration 4082 is done through `Config VXLAN Encap outer layers`_. 4083 4084- ``vxlan_decap``: Performs a decapsulation action by stripping all headers of 4085 the VXLAN tunnel network overlay from the matched flow. 4086 4087- ``nvgre_encap``: Performs a NVGRE encapsulation, outer layer configuration 4088 is done through `Config NVGRE Encap outer layers`_. 4089 4090- ``nvgre_decap``: Performs a decapsulation action by stripping all headers of 4091 the NVGRE tunnel network overlay from the matched flow. 4092 4093- ``l2_encap``: Performs a L2 encapsulation, L2 configuration 4094 is done through `Config L2 Encap`_. 4095 4096- ``l2_decap``: Performs a L2 decapsulation, L2 configuration 4097 is done through `Config L2 Decap`_. 4098 4099- ``mplsogre_encap``: Performs a MPLSoGRE encapsulation, outer layer 4100 configuration is done through `Config MPLSoGRE Encap outer layers`_. 4101 4102- ``mplsogre_decap``: Performs a MPLSoGRE decapsulation, outer layer 4103 configuration is done through `Config MPLSoGRE Decap outer layers`_. 4104 4105- ``mplsoudp_encap``: Performs a MPLSoUDP encapsulation, outer layer 4106 configuration is done through `Config MPLSoUDP Encap outer layers`_. 4107 4108- ``mplsoudp_decap``: Performs a MPLSoUDP decapsulation, outer layer 4109 configuration is done through `Config MPLSoUDP Decap outer layers`_. 4110 4111- ``set_ipv4_src``: Set a new IPv4 source address in the outermost IPv4 header. 4112 4113 - ``ipv4_addr``: New IPv4 source address. 4114 4115- ``set_ipv4_dst``: Set a new IPv4 destination address in the outermost IPv4 4116 header. 4117 4118 - ``ipv4_addr``: New IPv4 destination address. 4119 4120- ``set_ipv6_src``: Set a new IPv6 source address in the outermost IPv6 header. 4121 4122 - ``ipv6_addr``: New IPv6 source address. 4123 4124- ``set_ipv6_dst``: Set a new IPv6 destination address in the outermost IPv6 4125 header. 4126 4127 - ``ipv6_addr``: New IPv6 destination address. 4128 4129- ``set_tp_src``: Set a new source port number in the outermost TCP/UDP 4130 header. 4131 4132 - ``port``: New TCP/UDP source port number. 4133 4134- ``set_tp_dst``: Set a new destination port number in the outermost TCP/UDP 4135 header. 4136 4137 - ``port``: New TCP/UDP destination port number. 4138 4139- ``mac_swap``: Swap the source and destination MAC addresses in the outermost 4140 Ethernet header. 4141 4142- ``dec_ttl``: Performs a decrease TTL value action 4143 4144- ``set_ttl``: Set TTL value with specified value 4145 - ``ttl_value {unsigned}``: The new TTL value to be set 4146 4147- ``set_mac_src``: set source MAC address 4148 4149 - ``mac_addr {MAC-48}``: new source MAC address 4150 4151- ``set_mac_dst``: set destination MAC address 4152 4153 - ``mac_addr {MAC-48}``: new destination MAC address 4154 4155- ``inc_tcp_seq``: Increase sequence number in the outermost TCP header. 4156 4157 - ``value {unsigned}``: Value to increase TCP sequence number by. 4158 4159- ``dec_tcp_seq``: Decrease sequence number in the outermost TCP header. 4160 4161 - ``value {unsigned}``: Value to decrease TCP sequence number by. 4162 4163- ``inc_tcp_ack``: Increase acknowledgment number in the outermost TCP header. 4164 4165 - ``value {unsigned}``: Value to increase TCP acknowledgment number by. 4166 4167- ``dec_tcp_ack``: Decrease acknowledgment number in the outermost TCP header. 4168 4169 - ``value {unsigned}``: Value to decrease TCP acknowledgment number by. 4170 4171- ``set_ipv4_dscp``: Set IPv4 DSCP value with specified value 4172 4173 - ``dscp_value {unsigned}``: The new DSCP value to be set 4174 4175- ``set_ipv6_dscp``: Set IPv6 DSCP value with specified value 4176 4177 - ``dscp_value {unsigned}``: The new DSCP value to be set 4178 4179- ``shared``: Use shared action created via 4180 ``flow shared_action {port_id} create`` 4181 4182 - ``shared_action_id {unsigned}``: Shared action ID to use 4183 4184Destroying flow rules 4185~~~~~~~~~~~~~~~~~~~~~ 4186 4187``flow destroy`` destroys one or more rules from their rule ID (as returned 4188by ``flow create``), this command calls ``rte_flow_destroy()`` as many 4189times as necessary:: 4190 4191 flow destroy {port_id} rule {rule_id} [...] 4192 4193If successful, it will show:: 4194 4195 Flow rule #[...] destroyed 4196 4197It does not report anything for rule IDs that do not exist. The usual error 4198message is shown when a rule cannot be destroyed:: 4199 4200 Caught error type [...] ([...]): [...] 4201 4202``flow flush`` destroys all rules on a device and does not take extra 4203arguments. It is bound to ``rte_flow_flush()``:: 4204 4205 flow flush {port_id} 4206 4207Any errors are reported as above. 4208 4209Creating several rules and destroying them:: 4210 4211 testpmd> flow create 0 ingress pattern eth / ipv6 / end 4212 actions queue index 2 / end 4213 Flow rule #0 created 4214 testpmd> flow create 0 ingress pattern eth / ipv4 / end 4215 actions queue index 3 / end 4216 Flow rule #1 created 4217 testpmd> flow destroy 0 rule 0 rule 1 4218 Flow rule #1 destroyed 4219 Flow rule #0 destroyed 4220 testpmd> 4221 4222The same result can be achieved using ``flow flush``:: 4223 4224 testpmd> flow create 0 ingress pattern eth / ipv6 / end 4225 actions queue index 2 / end 4226 Flow rule #0 created 4227 testpmd> flow create 0 ingress pattern eth / ipv4 / end 4228 actions queue index 3 / end 4229 Flow rule #1 created 4230 testpmd> flow flush 0 4231 testpmd> 4232 4233Non-existent rule IDs are ignored:: 4234 4235 testpmd> flow create 0 ingress pattern eth / ipv6 / end 4236 actions queue index 2 / end 4237 Flow rule #0 created 4238 testpmd> flow create 0 ingress pattern eth / ipv4 / end 4239 actions queue index 3 / end 4240 Flow rule #1 created 4241 testpmd> flow destroy 0 rule 42 rule 10 rule 2 4242 testpmd> 4243 testpmd> flow destroy 0 rule 0 4244 Flow rule #0 destroyed 4245 testpmd> 4246 4247Querying flow rules 4248~~~~~~~~~~~~~~~~~~~ 4249 4250``flow query`` queries a specific action of a flow rule having that 4251ability. Such actions collect information that can be reported using this 4252command. It is bound to ``rte_flow_query()``:: 4253 4254 flow query {port_id} {rule_id} {action} 4255 4256If successful, it will display either the retrieved data for known actions 4257or the following message:: 4258 4259 Cannot display result for action type [...] ([...]) 4260 4261Otherwise, it will complain either that the rule does not exist or that some 4262error occurred:: 4263 4264 Flow rule #[...] not found 4265 4266:: 4267 4268 Caught error type [...] ([...]): [...] 4269 4270Currently only the ``count`` action is supported. This action reports the 4271number of packets that hit the flow rule and the total number of bytes. Its 4272output has the following format:: 4273 4274 count: 4275 hits_set: [...] # whether "hits" contains a valid value 4276 bytes_set: [...] # whether "bytes" contains a valid value 4277 hits: [...] # number of packets 4278 bytes: [...] # number of bytes 4279 4280Querying counters for TCPv6 packets redirected to queue 6:: 4281 4282 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end 4283 actions queue index 6 / count / end 4284 Flow rule #4 created 4285 testpmd> flow query 0 4 count 4286 count: 4287 hits_set: 1 4288 bytes_set: 0 4289 hits: 386446 4290 bytes: 0 4291 testpmd> 4292 4293Listing flow rules 4294~~~~~~~~~~~~~~~~~~ 4295 4296``flow list`` lists existing flow rules sorted by priority and optionally 4297filtered by group identifiers:: 4298 4299 flow list {port_id} [group {group_id}] [...] 4300 4301This command only fails with the following message if the device does not 4302exist:: 4303 4304 Invalid port [...] 4305 4306Output consists of a header line followed by a short description of each 4307flow rule, one per line. There is no output at all when no flow rules are 4308configured on the device:: 4309 4310 ID Group Prio Attr Rule 4311 [...] [...] [...] [...] [...] 4312 4313``Attr`` column flags: 4314 4315- ``i`` for ``ingress``. 4316- ``e`` for ``egress``. 4317 4318Creating several flow rules and listing them:: 4319 4320 testpmd> flow create 0 ingress pattern eth / ipv4 / end 4321 actions queue index 6 / end 4322 Flow rule #0 created 4323 testpmd> flow create 0 ingress pattern eth / ipv6 / end 4324 actions queue index 2 / end 4325 Flow rule #1 created 4326 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end 4327 actions rss queues 6 7 8 end / end 4328 Flow rule #2 created 4329 testpmd> flow list 0 4330 ID Group Prio Attr Rule 4331 0 0 0 i- ETH IPV4 => QUEUE 4332 1 0 0 i- ETH IPV6 => QUEUE 4333 2 0 5 i- ETH IPV4 UDP => RSS 4334 testpmd> 4335 4336Rules are sorted by priority (i.e. group ID first, then priority level):: 4337 4338 testpmd> flow list 1 4339 ID Group Prio Attr Rule 4340 0 0 0 i- ETH => COUNT 4341 6 0 500 i- ETH IPV6 TCP => DROP COUNT 4342 5 0 1000 i- ETH IPV6 ICMP => QUEUE 4343 1 24 0 i- ETH IPV4 UDP => QUEUE 4344 4 24 10 i- ETH IPV4 TCP => DROP 4345 3 24 20 i- ETH IPV4 => DROP 4346 2 24 42 i- ETH IPV4 UDP => QUEUE 4347 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 4348 testpmd> 4349 4350Output can be limited to specific groups:: 4351 4352 testpmd> flow list 1 group 0 group 63 4353 ID Group Prio Attr Rule 4354 0 0 0 i- ETH => COUNT 4355 6 0 500 i- ETH IPV6 TCP => DROP COUNT 4356 5 0 1000 i- ETH IPV6 ICMP => QUEUE 4357 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 4358 testpmd> 4359 4360Toggling isolated mode 4361~~~~~~~~~~~~~~~~~~~~~~ 4362 4363``flow isolate`` can be used to tell the underlying PMD that ingress traffic 4364must only be injected from the defined flow rules; that no default traffic 4365is expected outside those rules and the driver is free to assign more 4366resources to handle them. It is bound to ``rte_flow_isolate()``:: 4367 4368 flow isolate {port_id} {boolean} 4369 4370If successful, enabling or disabling isolated mode shows either:: 4371 4372 Ingress traffic on port [...] 4373 is now restricted to the defined flow rules 4374 4375Or:: 4376 4377 Ingress traffic on port [...] 4378 is not restricted anymore to the defined flow rules 4379 4380Otherwise, in case of error:: 4381 4382 Caught error type [...] ([...]): [...] 4383 4384Mainly due to its side effects, PMDs supporting this mode may not have the 4385ability to toggle it more than once without reinitializing affected ports 4386first (e.g. by exiting testpmd). 4387 4388Enabling isolated mode:: 4389 4390 testpmd> flow isolate 0 true 4391 Ingress traffic on port 0 is now restricted to the defined flow rules 4392 testpmd> 4393 4394Disabling isolated mode:: 4395 4396 testpmd> flow isolate 0 false 4397 Ingress traffic on port 0 is not restricted anymore to the defined flow rules 4398 testpmd> 4399 4400Dumping HW internal information 4401~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4402 4403``flow dump`` dumps the hardware's internal representation information of 4404all flows. It is bound to ``rte_flow_dev_dump()``:: 4405 4406 flow dump {port_id} {output_file} 4407 4408If successful, it will show:: 4409 4410 Flow dump finished 4411 4412Otherwise, it will complain error occurred:: 4413 4414 Caught error type [...] ([...]): [...] 4415 4416Listing and destroying aged flow rules 4417~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4418 4419``flow aged`` simply lists aged flow rules be get from api ``rte_flow_get_aged_flows``, 4420and ``destroy`` parameter can be used to destroy those flow rules in PMD. 4421 4422 flow aged {port_id} [destroy] 4423 4424Listing current aged flow rules:: 4425 4426 testpmd> flow aged 0 4427 Port 0 total aged flows: 0 4428 testpmd> flow create 0 ingress pattern eth / ipv4 src is 2.2.2.14 / end 4429 actions age timeout 5 / queue index 0 / end 4430 Flow rule #0 created 4431 testpmd> flow create 0 ingress pattern eth / ipv4 src is 2.2.2.15 / end 4432 actions age timeout 4 / queue index 0 / end 4433 Flow rule #1 created 4434 testpmd> flow create 0 ingress pattern eth / ipv4 src is 2.2.2.16 / end 4435 actions age timeout 2 / queue index 0 / end 4436 Flow rule #2 created 4437 testpmd> flow create 0 ingress pattern eth / ipv4 src is 2.2.2.17 / end 4438 actions age timeout 3 / queue index 0 / end 4439 Flow rule #3 created 4440 4441 4442Aged Rules are simply list as command ``flow list {port_id}``, but strip the detail rule 4443information, all the aged flows are sorted by the longest timeout time. For example, if 4444those rules be configured in the same time, ID 2 will be the first aged out rule, the next 4445will be ID 3, ID 1, ID 0:: 4446 4447 testpmd> flow aged 0 4448 Port 0 total aged flows: 4 4449 ID Group Prio Attr 4450 2 0 0 i-- 4451 3 0 0 i-- 4452 1 0 0 i-- 4453 0 0 0 i-- 4454 4455If attach ``destroy`` parameter, the command will destroy all the list aged flow rules. 4456 4457 testpmd> flow aged 0 destroy 4458 Port 0 total aged flows: 4 4459 ID Group Prio Attr 4460 2 0 0 i-- 4461 3 0 0 i-- 4462 1 0 0 i-- 4463 0 0 0 i-- 4464 4465 Flow rule #2 destroyed 4466 Flow rule #3 destroyed 4467 Flow rule #1 destroyed 4468 Flow rule #0 destroyed 4469 4 flows be destroyed 4470 testpmd> flow aged 0 4471 Port 0 total aged flows: 0 4472 4473Creating shared actions 4474~~~~~~~~~~~~~~~~~~~~~~~ 4475``flow shared_action {port_id} create`` creates shared action with optional 4476shared action ID. It is bound to ``rte_flow_shared_action_create()``:: 4477 4478 flow shared_action {port_id} create [action_id {shared_action_id}] 4479 [ingress] [egress] action {action} / end 4480 4481If successful, it will show:: 4482 4483 Shared action #[...] created 4484 4485Otherwise, it will complain either that shared action already exists or that 4486some error occurred:: 4487 4488 Shared action #[...] is already assigned, delete it first 4489 4490:: 4491 4492 Caught error type [...] ([...]): [...] 4493 4494Create shared rss action with id 100 to queues 1 and 2 on port 0:: 4495 4496 testpmd> flow shared_action 0 create action_id 100 \ 4497 ingress action rss queues 1 2 end / end 4498 4499Create shared rss action with id assigned by testpmd to queues 1 and 2 on 4500port 0:: 4501 4502 testpmd> flow shared_action 0 create action_id \ 4503 ingress action rss queues 0 1 end / end 4504 4505Updating shared actions 4506~~~~~~~~~~~~~~~~~~~~~~~ 4507``flow shared_action {port_id} update`` updates configuration of the shared 4508action from its shared action ID (as returned by 4509``flow shared_action {port_id} create``). It is bound to 4510``rte_flow_shared_action_update()``:: 4511 4512 flow shared_action {port_id} update {shared_action_id} 4513 action {action} / end 4514 4515If successful, it will show:: 4516 4517 Shared action #[...] updated 4518 4519Otherwise, it will complain either that shared action not found or that some 4520error occurred:: 4521 4522 Failed to find shared action #[...] on port [...] 4523 4524:: 4525 4526 Caught error type [...] ([...]): [...] 4527 4528Update shared rss action having id 100 on port 0 with rss to queues 0 and 3 4529(in create example above rss queues were 1 and 2):: 4530 4531 testpmd> flow shared_action 0 update 100 action rss queues 0 3 end / end 4532 4533Destroying shared actions 4534~~~~~~~~~~~~~~~~~~~~~~~~~ 4535``flow shared_action {port_id} update`` destroys one or more shared actions 4536from their shared action IDs (as returned by 4537``flow shared_action {port_id} create``). It is bound to 4538``rte_flow_shared_action_destroy()``:: 4539 4540 flow shared_action {port_id} destroy action_id {shared_action_id} [...] 4541 4542If successful, it will show:: 4543 4544 Shared action #[...] destroyed 4545 4546It does not report anything for shared action IDs that do not exist. 4547The usual error message is shown when a shared action cannot be destroyed:: 4548 4549 Caught error type [...] ([...]): [...] 4550 4551Destroy shared actions having id 100 & 101:: 4552 4553 testpmd> flow shared_action 0 destroy action_id 100 action_id 101 4554 4555Query shared actions 4556~~~~~~~~~~~~~~~~~~~~ 4557``flow shared_action {port_id} query`` queries the shared action from its 4558shared action ID (as returned by ``flow shared_action {port_id} create``). 4559It is bound to ``rte_flow_shared_action_query()``:: 4560 4561 flow shared_action {port_id} query {shared_action_id} 4562 4563Currently only rss shared action supported. If successful, it will show:: 4564 4565 Shared RSS action: 4566 refs:[...] 4567 4568Otherwise, it will complain either that shared action not found or that some 4569error occurred:: 4570 4571 Failed to find shared action #[...] on port [...] 4572 4573:: 4574 4575 Caught error type [...] ([...]): [...] 4576 4577Query shared action having id 100:: 4578 4579 testpmd> flow shared_action 0 query 100 4580 4581Sample QinQ flow rules 4582~~~~~~~~~~~~~~~~~~~~~~ 4583 4584Before creating QinQ rule(s) the following commands should be issued to enable QinQ:: 4585 4586 testpmd> port stop 0 4587 testpmd> vlan set qinq_strip on 0 4588 4589The above command sets the inner and outer TPID's to 0x8100. 4590 4591To change the TPID's the following commands should be used:: 4592 4593 testpmd> vlan set outer tpid 0xa100 0 4594 testpmd> vlan set inner tpid 0x9100 0 4595 testpmd> port start 0 4596 4597Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM. 4598 4599:: 4600 4601 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 / 4602 vlan tci is 456 / end actions vf id 1 / queue index 0 / end 4603 Flow rule #0 validated 4604 4605 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 / 4606 vlan tci is 456 / end actions vf id 123 / queue index 0 / end 4607 Flow rule #0 created 4608 4609 testpmd> flow list 0 4610 ID Group Prio Attr Rule 4611 0 0 0 i- ETH VLAN VLAN=>VF QUEUE 4612 4613Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host. 4614 4615:: 4616 4617 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 / 4618 vlan tci is 654 / end actions pf / queue index 0 / end 4619 Flow rule #1 validated 4620 4621 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 / 4622 vlan tci is 654 / end actions pf / queue index 1 / end 4623 Flow rule #1 created 4624 4625 testpmd> flow list 0 4626 ID Group Prio Attr Rule 4627 0 0 0 i- ETH VLAN VLAN=>VF QUEUE 4628 1 0 0 i- ETH VLAN VLAN=>PF QUEUE 4629 4630Sample VXLAN flow rules 4631~~~~~~~~~~~~~~~~~~~~~~~ 4632 4633Before creating VXLAN rule(s), the UDP port should be added for VXLAN packet 4634filter on a port:: 4635 4636 testpmd> rx_vxlan_port add 4789 0 4637 4638Create VXLAN rules on port 0 to steer traffic to PF queues. 4639 4640:: 4641 4642 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / vxlan / 4643 eth dst is 00:11:22:33:44:55 / end actions pf / queue index 1 / end 4644 Flow rule #0 created 4645 4646 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / vxlan vni is 3 / 4647 eth dst is 00:11:22:33:44:55 / end actions pf / queue index 2 / end 4648 Flow rule #1 created 4649 4650 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / vxlan / 4651 eth dst is 00:11:22:33:44:55 / vlan tci is 10 / end actions pf / 4652 queue index 3 / end 4653 Flow rule #2 created 4654 4655 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / vxlan vni is 5 / 4656 eth dst is 00:11:22:33:44:55 / vlan tci is 20 / end actions pf / 4657 queue index 4 / end 4658 Flow rule #3 created 4659 4660 testpmd> flow create 0 ingress pattern eth dst is 00:00:00:00:01:00 / ipv4 / 4661 udp / vxlan vni is 6 / eth dst is 00:11:22:33:44:55 / end actions pf / 4662 queue index 5 / end 4663 Flow rule #4 created 4664 4665 testpmd> flow list 0 4666 ID Group Prio Attr Rule 4667 0 0 0 i- ETH IPV4 UDP VXLAN ETH => QUEUE 4668 1 0 0 i- ETH IPV4 UDP VXLAN ETH => QUEUE 4669 2 0 0 i- ETH IPV4 UDP VXLAN ETH VLAN => QUEUE 4670 3 0 0 i- ETH IPV4 UDP VXLAN ETH VLAN => QUEUE 4671 4 0 0 i- ETH IPV4 UDP VXLAN ETH => QUEUE 4672 4673Sample VXLAN encapsulation rule 4674~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4675 4676VXLAN encapsulation outer layer has default value pre-configured in testpmd 4677source code, those can be changed by using the following commands 4678 4679IPv4 VXLAN outer header:: 4680 4681 testpmd> set vxlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src 127.0.0.1 4682 ip-dst 128.0.0.1 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4683 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 4684 queue index 0 / end 4685 4686 testpmd> set vxlan-with-vlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src 4687 127.0.0.1 ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11 4688 eth-dst 22:22:22:22:22:22 4689 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 4690 queue index 0 / end 4691 4692 testpmd> set vxlan-tos-ttl ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-tos 0 4693 ip-ttl 255 ip-src 127.0.0.1 ip-dst 128.0.0.1 eth-src 11:11:11:11:11:11 4694 eth-dst 22:22:22:22:22:22 4695 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 4696 queue index 0 / end 4697 4698IPv6 VXLAN outer header:: 4699 4700 testpmd> set vxlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4 ip-src ::1 4701 ip-dst ::2222 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4702 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 4703 queue index 0 / end 4704 4705 testpmd> set vxlan-with-vlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4 4706 ip-src ::1 ip-dst ::2222 vlan-tci 34 eth-src 11:11:11:11:11:11 4707 eth-dst 22:22:22:22:22:22 4708 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 4709 queue index 0 / end 4710 4711 testpmd> set vxlan-tos-ttl ip-version ipv6 vni 4 udp-src 4 udp-dst 4 4712 ip-tos 0 ip-ttl 255 ::1 ip-dst ::2222 eth-src 11:11:11:11:11:11 4713 eth-dst 22:22:22:22:22:22 4714 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 4715 queue index 0 / end 4716 4717Sample NVGRE encapsulation rule 4718~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4719 4720NVGRE encapsulation outer layer has default value pre-configured in testpmd 4721source code, those can be changed by using the following commands 4722 4723IPv4 NVGRE outer header:: 4724 4725 testpmd> set nvgre ip-version ipv4 tni 4 ip-src 127.0.0.1 ip-dst 128.0.0.1 4726 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4727 testpmd> flow create 0 ingress pattern end actions nvgre_encap / 4728 queue index 0 / end 4729 4730 testpmd> set nvgre-with-vlan ip-version ipv4 tni 4 ip-src 127.0.0.1 4731 ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11 4732 eth-dst 22:22:22:22:22:22 4733 testpmd> flow create 0 ingress pattern end actions nvgre_encap / 4734 queue index 0 / end 4735 4736IPv6 NVGRE outer header:: 4737 4738 testpmd> set nvgre ip-version ipv6 tni 4 ip-src ::1 ip-dst ::2222 4739 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4740 testpmd> flow create 0 ingress pattern end actions nvgre_encap / 4741 queue index 0 / end 4742 4743 testpmd> set nvgre-with-vlan ip-version ipv6 tni 4 ip-src ::1 ip-dst ::2222 4744 vlan-tci 34 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4745 testpmd> flow create 0 ingress pattern end actions nvgre_encap / 4746 queue index 0 / end 4747 4748Sample L2 encapsulation rule 4749~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4750 4751L2 encapsulation has default value pre-configured in testpmd 4752source code, those can be changed by using the following commands 4753 4754L2 header:: 4755 4756 testpmd> set l2_encap ip-version ipv4 4757 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4758 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / mpls / end actions 4759 mplsoudp_decap / l2_encap / end 4760 4761L2 with VXLAN header:: 4762 4763 testpmd> set l2_encap-with-vlan ip-version ipv4 vlan-tci 34 4764 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4765 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / mpls / end actions 4766 mplsoudp_decap / l2_encap / end 4767 4768Sample L2 decapsulation rule 4769~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4770 4771L2 decapsulation has default value pre-configured in testpmd 4772source code, those can be changed by using the following commands 4773 4774L2 header:: 4775 4776 testpmd> set l2_decap 4777 testpmd> flow create 0 egress pattern eth / end actions l2_decap / mplsoudp_encap / 4778 queue index 0 / end 4779 4780L2 with VXLAN header:: 4781 4782 testpmd> set l2_encap-with-vlan 4783 testpmd> flow create 0 egress pattern eth / end actions l2_encap / mplsoudp_encap / 4784 queue index 0 / end 4785 4786Sample MPLSoGRE encapsulation rule 4787~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4788 4789MPLSoGRE encapsulation outer layer has default value pre-configured in testpmd 4790source code, those can be changed by using the following commands 4791 4792IPv4 MPLSoGRE outer header:: 4793 4794 testpmd> set mplsogre_encap ip-version ipv4 label 4 4795 ip-src 127.0.0.1 ip-dst 128.0.0.1 eth-src 11:11:11:11:11:11 4796 eth-dst 22:22:22:22:22:22 4797 testpmd> flow create 0 egress pattern eth / end actions l2_decap / 4798 mplsogre_encap / end 4799 4800IPv4 MPLSoGRE with VLAN outer header:: 4801 4802 testpmd> set mplsogre_encap-with-vlan ip-version ipv4 label 4 4803 ip-src 127.0.0.1 ip-dst 128.0.0.1 vlan-tci 34 4804 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4805 testpmd> flow create 0 egress pattern eth / end actions l2_decap / 4806 mplsogre_encap / end 4807 4808IPv6 MPLSoGRE outer header:: 4809 4810 testpmd> set mplsogre_encap ip-version ipv6 mask 4 4811 ip-src ::1 ip-dst ::2222 eth-src 11:11:11:11:11:11 4812 eth-dst 22:22:22:22:22:22 4813 testpmd> flow create 0 egress pattern eth / end actions l2_decap / 4814 mplsogre_encap / end 4815 4816IPv6 MPLSoGRE with VLAN outer header:: 4817 4818 testpmd> set mplsogre_encap-with-vlan ip-version ipv6 mask 4 4819 ip-src ::1 ip-dst ::2222 vlan-tci 34 4820 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4821 testpmd> flow create 0 egress pattern eth / end actions l2_decap / 4822 mplsogre_encap / end 4823 4824Sample MPLSoGRE decapsulation rule 4825~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4826 4827MPLSoGRE decapsulation outer layer has default value pre-configured in testpmd 4828source code, those can be changed by using the following commands 4829 4830IPv4 MPLSoGRE outer header:: 4831 4832 testpmd> set mplsogre_decap ip-version ipv4 4833 testpmd> flow create 0 ingress pattern eth / ipv4 / gre / mpls / end actions 4834 mplsogre_decap / l2_encap / end 4835 4836IPv4 MPLSoGRE with VLAN outer header:: 4837 4838 testpmd> set mplsogre_decap-with-vlan ip-version ipv4 4839 testpmd> flow create 0 ingress pattern eth / vlan / ipv4 / gre / mpls / end 4840 actions mplsogre_decap / l2_encap / end 4841 4842IPv6 MPLSoGRE outer header:: 4843 4844 testpmd> set mplsogre_decap ip-version ipv6 4845 testpmd> flow create 0 ingress pattern eth / ipv6 / gre / mpls / end 4846 actions mplsogre_decap / l2_encap / end 4847 4848IPv6 MPLSoGRE with VLAN outer header:: 4849 4850 testpmd> set mplsogre_decap-with-vlan ip-version ipv6 4851 testpmd> flow create 0 ingress pattern eth / vlan / ipv6 / gre / mpls / end 4852 actions mplsogre_decap / l2_encap / end 4853 4854Sample MPLSoUDP encapsulation rule 4855~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4856 4857MPLSoUDP encapsulation outer layer has default value pre-configured in testpmd 4858source code, those can be changed by using the following commands 4859 4860IPv4 MPLSoUDP outer header:: 4861 4862 testpmd> set mplsoudp_encap ip-version ipv4 label 4 udp-src 5 udp-dst 10 4863 ip-src 127.0.0.1 ip-dst 128.0.0.1 eth-src 11:11:11:11:11:11 4864 eth-dst 22:22:22:22:22:22 4865 testpmd> flow create 0 egress pattern eth / end actions l2_decap / 4866 mplsoudp_encap / end 4867 4868IPv4 MPLSoUDP with VLAN outer header:: 4869 4870 testpmd> set mplsoudp_encap-with-vlan ip-version ipv4 label 4 udp-src 5 4871 udp-dst 10 ip-src 127.0.0.1 ip-dst 128.0.0.1 vlan-tci 34 4872 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4873 testpmd> flow create 0 egress pattern eth / end actions l2_decap / 4874 mplsoudp_encap / end 4875 4876IPv6 MPLSoUDP outer header:: 4877 4878 testpmd> set mplsoudp_encap ip-version ipv6 mask 4 udp-src 5 udp-dst 10 4879 ip-src ::1 ip-dst ::2222 eth-src 11:11:11:11:11:11 4880 eth-dst 22:22:22:22:22:22 4881 testpmd> flow create 0 egress pattern eth / end actions l2_decap / 4882 mplsoudp_encap / end 4883 4884IPv6 MPLSoUDP with VLAN outer header:: 4885 4886 testpmd> set mplsoudp_encap-with-vlan ip-version ipv6 mask 4 udp-src 5 4887 udp-dst 10 ip-src ::1 ip-dst ::2222 vlan-tci 34 4888 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4889 testpmd> flow create 0 egress pattern eth / end actions l2_decap / 4890 mplsoudp_encap / end 4891 4892Sample MPLSoUDP decapsulation rule 4893~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4894 4895MPLSoUDP decapsulation outer layer has default value pre-configured in testpmd 4896source code, those can be changed by using the following commands 4897 4898IPv4 MPLSoUDP outer header:: 4899 4900 testpmd> set mplsoudp_decap ip-version ipv4 4901 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / mpls / end actions 4902 mplsoudp_decap / l2_encap / end 4903 4904IPv4 MPLSoUDP with VLAN outer header:: 4905 4906 testpmd> set mplsoudp_decap-with-vlan ip-version ipv4 4907 testpmd> flow create 0 ingress pattern eth / vlan / ipv4 / udp / mpls / end 4908 actions mplsoudp_decap / l2_encap / end 4909 4910IPv6 MPLSoUDP outer header:: 4911 4912 testpmd> set mplsoudp_decap ip-version ipv6 4913 testpmd> flow create 0 ingress pattern eth / ipv6 / udp / mpls / end 4914 actions mplsoudp_decap / l2_encap / end 4915 4916IPv6 MPLSoUDP with VLAN outer header:: 4917 4918 testpmd> set mplsoudp_decap-with-vlan ip-version ipv6 4919 testpmd> flow create 0 ingress pattern eth / vlan / ipv6 / udp / mpls / end 4920 actions mplsoudp_decap / l2_encap / end 4921 4922Sample Raw encapsulation rule 4923~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4924 4925Raw encapsulation configuration can be set by the following commands 4926 4927Eecapsulating VxLAN:: 4928 4929 testpmd> set raw_encap 4 eth src is 10:11:22:33:44:55 / vlan tci is 1 4930 inner_type is 0x0800 / ipv4 / udp dst is 4789 / vxlan vni 4931 is 2 / end_set 4932 testpmd> flow create 0 egress pattern eth / ipv4 / end actions 4933 raw_encap index 4 / end 4934 4935Sample Raw decapsulation rule 4936~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4937 4938Raw decapsulation configuration can be set by the following commands 4939 4940Decapsulating VxLAN:: 4941 4942 testpmd> set raw_decap eth / ipv4 / udp / vxlan / end_set 4943 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / vxlan / eth / ipv4 / 4944 end actions raw_decap / queue index 0 / end 4945 4946Sample ESP rules 4947~~~~~~~~~~~~~~~~ 4948 4949ESP rules can be created by the following commands:: 4950 4951 testpmd> flow create 0 ingress pattern eth / ipv4 / esp spi is 1 / end actions 4952 queue index 3 / end 4953 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / esp spi is 1 / end 4954 actions queue index 3 / end 4955 testpmd> flow create 0 ingress pattern eth / ipv6 / esp spi is 1 / end actions 4956 queue index 3 / end 4957 testpmd> flow create 0 ingress pattern eth / ipv6 / udp / esp spi is 1 / end 4958 actions queue index 3 / end 4959 4960Sample AH rules 4961~~~~~~~~~~~~~~~~ 4962 4963AH rules can be created by the following commands:: 4964 4965 testpmd> flow create 0 ingress pattern eth / ipv4 / ah spi is 1 / end actions 4966 queue index 3 / end 4967 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / ah spi is 1 / end 4968 actions queue index 3 / end 4969 testpmd> flow create 0 ingress pattern eth / ipv6 / ah spi is 1 / end actions 4970 queue index 3 / end 4971 testpmd> flow create 0 ingress pattern eth / ipv6 / udp / ah spi is 1 / end 4972 actions queue index 3 / end 4973 4974Sample PFCP rules 4975~~~~~~~~~~~~~~~~~ 4976 4977PFCP rules can be created by the following commands(s_field need to be 1 4978if seid is set):: 4979 4980 testpmd> flow create 0 ingress pattern eth / ipv4 / pfcp s_field is 0 / end 4981 actions queue index 3 / end 4982 testpmd> flow create 0 ingress pattern eth / ipv4 / pfcp s_field is 1 4983 seid is 1 / end actions queue index 3 / end 4984 testpmd> flow create 0 ingress pattern eth / ipv6 / pfcp s_field is 0 / end 4985 actions queue index 3 / end 4986 testpmd> flow create 0 ingress pattern eth / ipv6 / pfcp s_field is 1 4987 seid is 1 / end actions queue index 3 / end 4988 4989BPF Functions 4990-------------- 4991 4992The following sections show functions to load/unload eBPF based filters. 4993 4994bpf-load 4995~~~~~~~~ 4996 4997Load an eBPF program as a callback for particular RX/TX queue:: 4998 4999 testpmd> bpf-load rx|tx (portid) (queueid) (load-flags) (bpf-prog-filename) 5000 5001The available load-flags are: 5002 5003* ``J``: use JIT generated native code, otherwise BPF interpreter will be used. 5004 5005* ``M``: assume input parameter is a pointer to rte_mbuf, otherwise assume it is a pointer to first segment's data. 5006 5007* ``-``: none. 5008 5009.. note:: 5010 5011 You'll need clang v3.7 or above to build bpf program you'd like to load 5012 5013For example: 5014 5015.. code-block:: console 5016 5017 cd examples/bpf 5018 clang -O2 -target bpf -c t1.c 5019 5020Then to load (and JIT compile) t1.o at RX queue 0, port 1: 5021 5022.. code-block:: console 5023 5024 testpmd> bpf-load rx 1 0 J ./dpdk.org/examples/bpf/t1.o 5025 5026To load (not JITed) t1.o at TX queue 0, port 0: 5027 5028.. code-block:: console 5029 5030 testpmd> bpf-load tx 0 0 - ./dpdk.org/examples/bpf/t1.o 5031 5032bpf-unload 5033~~~~~~~~~~ 5034 5035Unload previously loaded eBPF program for particular RX/TX queue:: 5036 5037 testpmd> bpf-unload rx|tx (portid) (queueid) 5038 5039For example to unload BPF filter from TX queue 0, port 0: 5040 5041.. code-block:: console 5042 5043 testpmd> bpf-unload tx 0 0 5044