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 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 ./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 202 filter on 203 qinq(extend) off 204 Redirection table size: 512 205 Supported flow types: 206 ipv4-frag 207 ipv4-tcp 208 ipv4-udp 209 ipv4-sctp 210 ipv4-other 211 ipv6-frag 212 ipv6-tcp 213 ipv6-udp 214 ipv6-sctp 215 ipv6-other 216 l2_payload 217 port 218 vxlan 219 geneve 220 nvgre 221 222show port rss reta 223~~~~~~~~~~~~~~~~~~ 224 225Display the rss redirection table entry indicated by masks on port X:: 226 227 testpmd> show port (port_id) rss reta (size) (mask0, mask1...) 228 229size is used to indicate the hardware supported reta size 230 231show port rss-hash 232~~~~~~~~~~~~~~~~~~ 233 234Display the RSS hash functions and RSS hash key of a port:: 235 236 testpmd> show port (port_id) rss-hash [key] 237 238clear port 239~~~~~~~~~~ 240 241Clear the port statistics for a given port or for all ports:: 242 243 testpmd> clear port (info|stats|xstats|fdir|stat_qmap) (port_id|all) 244 245For example:: 246 247 testpmd> clear port stats all 248 249show (rxq|txq) 250~~~~~~~~~~~~~~ 251 252Display information for a given port's RX/TX queue:: 253 254 testpmd> show (rxq|txq) info (port_id) (queue_id) 255 256show config 257~~~~~~~~~~~ 258 259Displays the configuration of the application. 260The configuration comes from the command-line, the runtime or the application defaults:: 261 262 testpmd> show config (rxtx|cores|fwd|txpkts) 263 264The available information categories are: 265 266* ``rxtx``: RX/TX configuration items. 267 268* ``cores``: List of forwarding cores. 269 270* ``fwd``: Packet forwarding configuration. 271 272* ``txpkts``: Packets to TX configuration. 273 274For example: 275 276.. code-block:: console 277 278 testpmd> show config rxtx 279 280 io packet forwarding - CRC stripping disabled - packets/burst=16 281 nb forwarding cores=2 - nb forwarding ports=1 282 RX queues=1 - RX desc=128 - RX free threshold=0 283 RX threshold registers: pthresh=8 hthresh=8 wthresh=4 284 TX queues=1 - TX desc=512 - TX free threshold=0 285 TX threshold registers: pthresh=36 hthresh=0 wthresh=0 286 TX RS bit threshold=0 - TXQ flags=0x0 287 288set fwd 289~~~~~~~ 290 291Set the packet forwarding mode:: 292 293 testpmd> set fwd (io|mac|macswap|flowgen| \ 294 rxonly|txonly|csum|icmpecho) (""|retry) 295 296``retry`` can be specified for forwarding engines except ``rx_only``. 297 298The available information categories are: 299 300* ``io``: Forwards packets "as-is" in I/O mode. 301 This is the fastest possible forwarding operation as it does not access packets data. 302 This is the default mode. 303 304* ``mac``: Changes the source and the destination Ethernet addresses of packets before forwarding them. 305 Default application behaviour is to set source Ethernet address to that of the transmitting interface, and destination 306 address to a dummy value (set during init). The user may specify a target destination Ethernet address via the 'eth-peer' or 307 'eth-peer-configfile' command-line options. It is not currently possible to specify a specific source Ethernet address. 308 309* ``macswap``: MAC swap forwarding mode. 310 Swaps the source and the destination Ethernet addresses of packets before forwarding them. 311 312* ``flowgen``: Multi-flow generation mode. 313 Originates a number of flows (with varying destination IP addresses), and terminate receive traffic. 314 315* ``rxonly``: Receives packets but doesn't transmit them. 316 317* ``txonly``: Generates and transmits packets without receiving any. 318 319* ``csum``: Changes the checksum field with hardware or software methods depending on the offload flags on the packet. 320 321* ``icmpecho``: Receives a burst of packets, lookup for IMCP echo requests and, if any, send back ICMP echo replies. 322 323* ``ieee1588``: Demonstrate L2 IEEE1588 V2 PTP timestamping for RX and TX. Requires ``CONFIG_RTE_LIBRTE_IEEE1588=y``. 324 325* ``softnic``: Demonstrates the softnic forwarding operation. In this mode, packet forwarding is 326 similar to I/O mode except for the fact that packets are loopback to the softnic ports only. Therefore, portmask parameter should be set to softnic port only. The various software based custom NIC pipelines specified through the softnic firmware (DPDK packet framework script) can be tested in this mode. Furthermore, it allows to build 5-level hierarchical QoS scheduler as a default option that can be enabled through CLI once testpmd application is initialised. The user can modify the default scheduler hierarchy or can specify the new QoS Scheduler hierarchy through CLI. Requires ``CONFIG_RTE_LIBRTE_PMD_SOFTNIC=y``. 327 328Example:: 329 330 testpmd> set fwd rxonly 331 332 Set rxonly packet forwarding mode 333 334 335read rxd 336~~~~~~~~ 337 338Display an RX descriptor for a port RX queue:: 339 340 testpmd> read rxd (port_id) (queue_id) (rxd_id) 341 342For example:: 343 344 testpmd> read rxd 0 0 4 345 0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180 346 347read txd 348~~~~~~~~ 349 350Display a TX descriptor for a port TX queue:: 351 352 testpmd> read txd (port_id) (queue_id) (txd_id) 353 354For example:: 355 356 testpmd> read txd 0 0 4 357 0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C 358 359ddp get list 360~~~~~~~~~~~~ 361 362Get loaded dynamic device personalization (DDP) package info list:: 363 364 testpmd> ddp get list (port_id) 365 366ddp get info 367~~~~~~~~~~~~ 368 369Display information about dynamic device personalization (DDP) profile:: 370 371 testpmd> ddp get info (profile_path) 372 373show vf stats 374~~~~~~~~~~~~~ 375 376Display VF statistics:: 377 378 testpmd> show vf stats (port_id) (vf_id) 379 380clear vf stats 381~~~~~~~~~~~~~~ 382 383Reset VF statistics:: 384 385 testpmd> clear vf stats (port_id) (vf_id) 386 387show port pctype mapping 388~~~~~~~~~~~~~~~~~~~~~~~~ 389 390List all items from the pctype mapping table:: 391 392 testpmd> show port (port_id) pctype mapping 393 394show rx offloading capabilities 395~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 396 397List all per queue and per port Rx offloading capabilities of a port:: 398 399 testpmd> show port (port_id) rx_offload capabilities 400 401show rx offloading configuration 402~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 403 404List port level and all queue level Rx offloading configuration:: 405 406 testpmd> show port (port_id) rx_offload configuration 407 408show tx offloading capabilities 409~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 410 411List all per queue and per port Tx offloading capabilities of a port:: 412 413 testpmd> show port (port_id) tx_offload capabilities 414 415show tx offloading configuration 416~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 417 418List port level and all queue level Tx offloading configuration:: 419 420 testpmd> show port (port_id) tx_offload configuration 421 422 423Configuration Functions 424----------------------- 425 426The testpmd application can be configured from the runtime as well as from the command-line. 427 428This section details the available configuration functions that are available. 429 430.. note:: 431 432 Configuration changes only become active when forwarding is started/restarted. 433 434set default 435~~~~~~~~~~~ 436 437Reset forwarding to the default configuration:: 438 439 testpmd> set default 440 441set verbose 442~~~~~~~~~~~ 443 444Set the debug verbosity level:: 445 446 testpmd> set verbose (level) 447 448Currently the only available levels are 0 (silent except for error) and 1 (fully verbose). 449 450set log 451~~~~~~~ 452 453Set the log level for a log type:: 454 455 testpmd> set log global|(type) (level) 456 457Where: 458 459* ``type`` is the log name. 460 461* ``level`` is the log level. 462 463For example, to change the global log level:: 464 testpmd> set log global (level) 465 466Regexes can also be used for type. To change log level of user1, user2 and user3:: 467 testpmd> set log user[1-3] (level) 468 469set nbport 470~~~~~~~~~~ 471 472Set the number of ports used by the application: 473 474set nbport (num) 475 476This is equivalent to the ``--nb-ports`` command-line option. 477 478set nbcore 479~~~~~~~~~~ 480 481Set the number of cores used by the application:: 482 483 testpmd> set nbcore (num) 484 485This is equivalent to the ``--nb-cores`` command-line option. 486 487.. note:: 488 489 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port. 490 491set coremask 492~~~~~~~~~~~~ 493 494Set the forwarding cores hexadecimal mask:: 495 496 testpmd> set coremask (mask) 497 498This is equivalent to the ``--coremask`` command-line option. 499 500.. note:: 501 502 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding. 503 504set portmask 505~~~~~~~~~~~~ 506 507Set the forwarding ports hexadecimal mask:: 508 509 testpmd> set portmask (mask) 510 511This is equivalent to the ``--portmask`` command-line option. 512 513set burst 514~~~~~~~~~ 515 516Set number of packets per burst:: 517 518 testpmd> set burst (num) 519 520This is equivalent to the ``--burst command-line`` option. 521 522When retry is enabled, the transmit delay time and number of retries can also be set:: 523 524 testpmd> set burst tx delay (microseconds) retry (num) 525 526set txpkts 527~~~~~~~~~~ 528 529Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode:: 530 531 testpmd> set txpkts (x[,y]*) 532 533Where x[,y]* represents a CSV list of values, without white space. 534 535set txsplit 536~~~~~~~~~~~ 537 538Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes:: 539 540 testpmd> set txsplit (off|on|rand) 541 542Where: 543 544* ``off`` disable packet copy & split for CSUM mode. 545 546* ``on`` split outgoing packet into multiple segments. Size of each segment 547 and number of segments per packet is determined by ``set txpkts`` command 548 (see above). 549 550* ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments. 551 552set corelist 553~~~~~~~~~~~~ 554 555Set the list of forwarding cores:: 556 557 testpmd> set corelist (x[,y]*) 558 559For example, to change the forwarding cores: 560 561.. code-block:: console 562 563 testpmd> set corelist 3,1 564 testpmd> show config fwd 565 566 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled 567 Logical Core 3 (socket 0) forwards packets on 1 streams: 568 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01 569 Logical Core 1 (socket 0) forwards packets on 1 streams: 570 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00 571 572.. note:: 573 574 The cores are used in the same order as specified on the command line. 575 576set portlist 577~~~~~~~~~~~~ 578 579Set the list of forwarding ports:: 580 581 testpmd> set portlist (x[,y]*) 582 583For example, to change the port forwarding: 584 585.. code-block:: console 586 587 testpmd> set portlist 0,2,1,3 588 testpmd> show config fwd 589 590 io packet forwarding - ports=4 - cores=1 - streams=4 591 Logical Core 3 (socket 0) forwards packets on 4 streams: 592 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01 593 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00 594 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03 595 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02 596 597set tx loopback 598~~~~~~~~~~~~~~~ 599 600Enable/disable tx loopback:: 601 602 testpmd> set tx loopback (port_id) (on|off) 603 604set drop enable 605~~~~~~~~~~~~~~~ 606 607set drop enable bit for all queues:: 608 609 testpmd> set all queues drop (port_id) (on|off) 610 611set split drop enable (for VF) 612~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 613 614set split drop enable bit for VF from PF:: 615 616 testpmd> set vf split drop (port_id) (vf_id) (on|off) 617 618set mac antispoof (for VF) 619~~~~~~~~~~~~~~~~~~~~~~~~~~ 620 621Set mac antispoof for a VF from the PF:: 622 623 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off) 624 625set macsec offload 626~~~~~~~~~~~~~~~~~~ 627 628Enable/disable MACsec offload:: 629 630 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off) 631 testpmd> set macsec offload (port_id) off 632 633set macsec sc 634~~~~~~~~~~~~~ 635 636Configure MACsec secure connection (SC):: 637 638 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi) 639 640.. note:: 641 642 The pi argument is ignored for tx. 643 Check the NIC Datasheet for hardware limits. 644 645set macsec sa 646~~~~~~~~~~~~~ 647 648Configure MACsec secure association (SA):: 649 650 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key) 651 652.. note:: 653 654 The IDX value must be 0 or 1. 655 Check the NIC Datasheet for hardware limits. 656 657set broadcast mode (for VF) 658~~~~~~~~~~~~~~~~~~~~~~~~~~~ 659 660Set broadcast mode for a VF from the PF:: 661 662 testpmd> set vf broadcast (port_id) (vf_id) (on|off) 663 664vlan set strip 665~~~~~~~~~~~~~~ 666 667Set the VLAN strip on a port:: 668 669 testpmd> vlan set strip (on|off) (port_id) 670 671vlan set stripq 672~~~~~~~~~~~~~~~ 673 674Set the VLAN strip for a queue on a port:: 675 676 testpmd> vlan set stripq (on|off) (port_id,queue_id) 677 678vlan set stripq (for VF) 679~~~~~~~~~~~~~~~~~~~~~~~~ 680 681Set VLAN strip for all queues in a pool for a VF from the PF:: 682 683 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off) 684 685vlan set insert (for VF) 686~~~~~~~~~~~~~~~~~~~~~~~~ 687 688Set VLAN insert for a VF from the PF:: 689 690 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id) 691 692vlan set tag (for VF) 693~~~~~~~~~~~~~~~~~~~~~ 694 695Set VLAN tag for a VF from the PF:: 696 697 testpmd> set vf vlan tag (port_id) (vf_id) (on|off) 698 699vlan set antispoof (for VF) 700~~~~~~~~~~~~~~~~~~~~~~~~~~~ 701 702Set VLAN antispoof for a VF from the PF:: 703 704 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off) 705 706vlan set filter 707~~~~~~~~~~~~~~~ 708 709Set the VLAN filter on a port:: 710 711 testpmd> vlan set filter (on|off) (port_id) 712 713vlan set qinq 714~~~~~~~~~~~~~ 715 716Set the VLAN QinQ (extended queue in queue) on for a port:: 717 718 testpmd> vlan set qinq (on|off) (port_id) 719 720vlan set tpid 721~~~~~~~~~~~~~ 722 723Set the inner or outer VLAN TPID for packet filtering on a port:: 724 725 testpmd> vlan set (inner|outer) tpid (value) (port_id) 726 727.. note:: 728 729 TPID value must be a 16-bit number (value <= 65536). 730 731rx_vlan add 732~~~~~~~~~~~ 733 734Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID:: 735 736 testpmd> rx_vlan add (vlan_id|all) (port_id) 737 738.. note:: 739 740 VLAN filter must be set on that port. VLAN ID < 4096. 741 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries 742 in VFTA table. This is important if enabling all vlan_ids. 743 744rx_vlan rm 745~~~~~~~~~~ 746 747Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID:: 748 749 testpmd> rx_vlan rm (vlan_id|all) (port_id) 750 751rx_vlan add (for VF) 752~~~~~~~~~~~~~~~~~~~~ 753 754Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID:: 755 756 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask) 757 758rx_vlan rm (for VF) 759~~~~~~~~~~~~~~~~~~~ 760 761Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID:: 762 763 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask) 764 765tunnel_filter add 766~~~~~~~~~~~~~~~~~ 767 768Add a tunnel filter on a port:: 769 770 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \ 771 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\ 772 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id) 773 774The available information categories are: 775 776* ``vxlan``: Set tunnel type as VXLAN. 777 778* ``nvgre``: Set tunnel type as NVGRE. 779 780* ``ipingre``: Set tunnel type as IP-in-GRE. 781 782* ``imac-ivlan``: Set filter type as Inner MAC and VLAN. 783 784* ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID. 785 786* ``imac-tenid``: Set filter type as Inner MAC and tenant ID. 787 788* ``imac``: Set filter type as Inner MAC. 789 790* ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID. 791 792* ``oip``: Set filter type as Outer IP. 793 794* ``iip``: Set filter type as Inner IP. 795 796Example:: 797 798 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \ 799 192.168.2.2 0 ipingre oip 1 1 800 801 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP. 802 803tunnel_filter remove 804~~~~~~~~~~~~~~~~~~~~ 805 806Remove a tunnel filter on a port:: 807 808 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \ 809 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\ 810 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id) 811 812rx_vxlan_port add 813~~~~~~~~~~~~~~~~~ 814 815Add an UDP port for VXLAN packet filter on a port:: 816 817 testpmd> rx_vxlan_port add (udp_port) (port_id) 818 819rx_vxlan_port remove 820~~~~~~~~~~~~~~~~~~~~ 821 822Remove an UDP port for VXLAN packet filter on a port:: 823 824 testpmd> rx_vxlan_port rm (udp_port) (port_id) 825 826tx_vlan set 827~~~~~~~~~~~ 828 829Set hardware insertion of VLAN IDs in packets sent on a port:: 830 831 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer] 832 833For example, set a single VLAN ID (5) insertion on port 0:: 834 835 tx_vlan set 0 5 836 837Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1:: 838 839 tx_vlan set 1 2 3 840 841 842tx_vlan set pvid 843~~~~~~~~~~~~~~~~ 844 845Set port based hardware insertion of VLAN ID in packets sent on a port:: 846 847 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off) 848 849tx_vlan reset 850~~~~~~~~~~~~~ 851 852Disable hardware insertion of a VLAN header in packets sent on a port:: 853 854 testpmd> tx_vlan reset (port_id) 855 856csum set 857~~~~~~~~ 858 859Select hardware or software calculation of the checksum when 860transmitting a packet using the ``csum`` forwarding engine:: 861 862 testpmd> csum set (ip|udp|tcp|sctp|outer-ip|outer-udp) (hw|sw) (port_id) 863 864Where: 865 866* ``ip|udp|tcp|sctp`` always relate to the inner layer. 867 868* ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized 869 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are 870 supported). See also the ``csum parse-tunnel`` command. 871 872* ``outer-udp`` relates to the outer UDP layer in the case where the packet is recognized 873 as a tunnel packet by the forwarding engine (vxlan, vxlan-gpe are 874 supported). See also the ``csum parse-tunnel`` command. 875 876.. note:: 877 878 Check the NIC Datasheet for hardware limits. 879 880RSS queue region 881~~~~~~~~~~~~~~~~ 882 883Set RSS queue region span on a port:: 884 885 testpmd> set port (port_id) queue-region region_id (value) \ 886 queue_start_index (value) queue_num (value) 887 888Set flowtype mapping on a RSS queue region on a port:: 889 890 testpmd> set port (port_id) queue-region region_id (value) flowtype (value) 891 892where: 893 894* For the flowtype(pctype) of packet,the specific index for each type has 895 been defined in file i40e_type.h as enum i40e_filter_pctype. 896 897Set user priority mapping on a RSS queue region on a port:: 898 899 testpmd> set port (port_id) queue-region UP (value) region_id (value) 900 901Flush all queue region related configuration on a port:: 902 903 testpmd> set port (port_id) queue-region flush (on|off) 904 905where: 906 907* "on"is just an enable function which server for other configuration, 908 it is for all configuration about queue region from up layer, 909 at first will only keep in DPDK softwarestored in driver, 910 only after "flush on", it commit all configuration to HW. 911 "off" is just clean all configuration about queue region just now, 912 and restore all to DPDK i40e driver default config when start up. 913 914Show all queue region related configuration info on a port:: 915 916 testpmd> show port (port_id) queue-region 917 918.. note:: 919 920 Queue region only support on PF by now, so these command is 921 only for configuration of queue region on PF port. 922 923csum parse-tunnel 924~~~~~~~~~~~~~~~~~ 925 926Define how tunneled packets should be handled by the csum forward 927engine:: 928 929 testpmd> csum parse-tunnel (on|off) (tx_port_id) 930 931If enabled, the csum forward engine will try to recognize supported 932tunnel headers (vxlan, gre, ipip). 933 934If disabled, treat tunnel packets as non-tunneled packets (a inner 935header is handled as a packet payload). 936 937.. note:: 938 939 The port argument is the TX port like in the ``csum set`` command. 940 941Example: 942 943Consider a packet in packet like the following:: 944 945 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in 946 947* If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set`` 948 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the 949 ``outer-ip|outer-udp`` parameter relates to the outer headers (here ``ipv4_out`` and ``udp_out``). 950 951* If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set`` 952 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``. 953 954csum show 955~~~~~~~~~ 956 957Display tx checksum offload configuration:: 958 959 testpmd> csum show (port_id) 960 961tso set 962~~~~~~~ 963 964Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine:: 965 966 testpmd> tso set (segsize) (port_id) 967 968.. note:: 969 970 Check the NIC datasheet for hardware limits. 971 972tso show 973~~~~~~~~ 974 975Display the status of TCP Segmentation Offload:: 976 977 testpmd> tso show (port_id) 978 979set port - gro 980~~~~~~~~~~~~~~ 981 982Enable or disable GRO in ``csum`` forwarding engine:: 983 984 testpmd> set port <port_id> gro on|off 985 986If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4 987packets received from the given port. 988 989If disabled, packets received from the given port won't be performed 990GRO. By default, GRO is disabled for all ports. 991 992.. note:: 993 994 When enable GRO for a port, TCP/IPv4 packets received from the port 995 will be performed GRO. After GRO, all merged packets have bad 996 checksums, since the GRO library doesn't re-calculate checksums for 997 the merged packets. Therefore, if users want the merged packets to 998 have correct checksums, please select HW IP checksum calculation and 999 HW TCP checksum calculation for the port which the merged packets are 1000 transmitted to. 1001 1002show port - gro 1003~~~~~~~~~~~~~~~ 1004 1005Display GRO configuration for a given port:: 1006 1007 testpmd> show port <port_id> gro 1008 1009set gro flush 1010~~~~~~~~~~~~~ 1011 1012Set the cycle to flush the GROed packets from reassembly tables:: 1013 1014 testpmd> set gro flush <cycles> 1015 1016When enable GRO, the csum forwarding engine performs GRO on received 1017packets, and the GROed packets are stored in reassembly tables. Users 1018can use this command to determine when the GROed packets are flushed 1019from the reassembly tables. 1020 1021The ``cycles`` is measured in GRO operation times. The csum forwarding 1022engine flushes the GROed packets from the tables every ``cycles`` GRO 1023operations. 1024 1025By default, the value of ``cycles`` is 1, which means flush GROed packets 1026from the reassembly tables as soon as one GRO operation finishes. The value 1027of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``. 1028 1029Please note that the large value of ``cycles`` may cause the poor TCP/IP 1030stack performance. Because the GROed packets are delayed to arrive the 1031stack, thus causing more duplicated ACKs and TCP retransmissions. 1032 1033set port - gso 1034~~~~~~~~~~~~~~ 1035 1036Toggle per-port GSO support in ``csum`` forwarding engine:: 1037 1038 testpmd> set port <port_id> gso on|off 1039 1040If enabled, the csum forwarding engine will perform GSO on supported IPv4 1041packets, transmitted on the given port. 1042 1043If disabled, packets transmitted on the given port will not undergo GSO. 1044By default, GSO is disabled for all ports. 1045 1046.. note:: 1047 1048 When GSO is enabled on a port, supported IPv4 packets transmitted on that 1049 port undergo GSO. Afterwards, the segmented packets are represented by 1050 multi-segment mbufs; however, the csum forwarding engine doesn't calculation 1051 of checksums for GSO'd segments in SW. As a result, if users want correct 1052 checksums in GSO segments, they should enable HW checksum calculation for 1053 GSO-enabled ports. 1054 1055 For example, HW checksum calculation for VxLAN GSO'd packets may be enabled 1056 by setting the following options in the csum forwarding engine: 1057 1058 testpmd> csum set outer_ip hw <port_id> 1059 1060 testpmd> csum set ip hw <port_id> 1061 1062 testpmd> csum set tcp hw <port_id> 1063 1064 UDP GSO is the same as IP fragmentation, which treats the UDP header 1065 as the payload and does not modify it during segmentation. That is, 1066 after UDP GSO, only the first output fragment has the original UDP 1067 header. Therefore, users need to enable HW IP checksum calculation 1068 and SW UDP checksum calculation for GSO-enabled ports, if they want 1069 correct checksums for UDP/IPv4 packets. 1070 1071set gso segsz 1072~~~~~~~~~~~~~ 1073 1074Set the maximum GSO segment size (measured in bytes), which includes the 1075packet header and the packet payload for GSO-enabled ports (global):: 1076 1077 testpmd> set gso segsz <length> 1078 1079show port - gso 1080~~~~~~~~~~~~~~~ 1081 1082Display the status of Generic Segmentation Offload for a given port:: 1083 1084 testpmd> show port <port_id> gso 1085 1086mac_addr add 1087~~~~~~~~~~~~ 1088 1089Add an alternative MAC address to a port:: 1090 1091 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX) 1092 1093mac_addr remove 1094~~~~~~~~~~~~~~~ 1095 1096Remove a MAC address from a port:: 1097 1098 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX) 1099 1100mac_addr add (for VF) 1101~~~~~~~~~~~~~~~~~~~~~ 1102 1103Add an alternative MAC address for a VF to a port:: 1104 1105 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX) 1106 1107mac_addr set 1108~~~~~~~~~~~~ 1109 1110Set the default MAC address for a port:: 1111 1112 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX) 1113 1114mac_addr set (for VF) 1115~~~~~~~~~~~~~~~~~~~~~ 1116 1117Set the MAC address for a VF from the PF:: 1118 1119 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX) 1120 1121set eth-peer 1122~~~~~~~~~~~~ 1123 1124Set the forwarding peer address for certain port:: 1125 1126 testpmd> set eth-peer (port_id) (perr_addr) 1127 1128This is equivalent to the ``--eth-peer`` command-line option. 1129 1130set port-uta 1131~~~~~~~~~~~~ 1132 1133Set the unicast hash filter(s) on/off for a port:: 1134 1135 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off) 1136 1137set promisc 1138~~~~~~~~~~~ 1139 1140Set the promiscuous mode on for a port or for all ports. 1141In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1142 1143 testpmd> set promisc (port_id|all) (on|off) 1144 1145set allmulti 1146~~~~~~~~~~~~ 1147 1148Set the allmulti mode for a port or for all ports:: 1149 1150 testpmd> set allmulti (port_id|all) (on|off) 1151 1152Same as the ifconfig (8) option. Controls how multicast packets are handled. 1153 1154set promisc (for VF) 1155~~~~~~~~~~~~~~~~~~~~ 1156 1157Set the unicast promiscuous mode for a VF from PF. 1158It's supported by Intel i40e NICs now. 1159In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1160 1161 testpmd> set vf promisc (port_id) (vf_id) (on|off) 1162 1163set allmulticast (for VF) 1164~~~~~~~~~~~~~~~~~~~~~~~~~ 1165 1166Set the multicast promiscuous mode for a VF from PF. 1167It's supported by Intel i40e NICs now. 1168In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1169 1170 testpmd> set vf allmulti (port_id) (vf_id) (on|off) 1171 1172set tx max bandwidth (for VF) 1173~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1174 1175Set TX max absolute bandwidth (Mbps) for a VF from PF:: 1176 1177 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth) 1178 1179set tc tx min bandwidth (for VF) 1180~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1181 1182Set all TCs' TX min relative bandwidth (%) for a VF from PF:: 1183 1184 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...) 1185 1186set tc tx max bandwidth (for VF) 1187~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1188 1189Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF:: 1190 1191 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth) 1192 1193set tc strict link priority mode 1194~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1195 1196Set some TCs' strict link priority mode on a physical port:: 1197 1198 testpmd> set tx strict-link-priority (port_id) (tc_bitmap) 1199 1200set tc tx min bandwidth 1201~~~~~~~~~~~~~~~~~~~~~~~ 1202 1203Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs:: 1204 1205 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...) 1206 1207set flow_ctrl rx 1208~~~~~~~~~~~~~~~~ 1209 1210Set the link flow control parameter on a port:: 1211 1212 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 1213 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \ 1214 autoneg (on|off) (port_id) 1215 1216Where: 1217 1218* ``high_water`` (integer): High threshold value to trigger XOFF. 1219 1220* ``low_water`` (integer): Low threshold value to trigger XON. 1221 1222* ``pause_time`` (integer): Pause quota in the Pause frame. 1223 1224* ``send_xon`` (0/1): Send XON frame. 1225 1226* ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames. 1227 1228* ``autoneg``: Change the auto-negotiation parameter. 1229 1230set pfc_ctrl rx 1231~~~~~~~~~~~~~~~ 1232 1233Set the priority flow control parameter on a port:: 1234 1235 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 1236 (pause_time) (priority) (port_id) 1237 1238Where: 1239 1240* ``high_water`` (integer): High threshold value. 1241 1242* ``low_water`` (integer): Low threshold value. 1243 1244* ``pause_time`` (integer): Pause quota in the Pause frame. 1245 1246* ``priority`` (0-7): VLAN User Priority. 1247 1248set stat_qmap 1249~~~~~~~~~~~~~ 1250 1251Set statistics mapping (qmapping 0..15) for RX/TX queue on port:: 1252 1253 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping) 1254 1255For example, to set rx queue 2 on port 0 to mapping 5:: 1256 1257 testpmd>set stat_qmap rx 0 2 5 1258 1259set xstats-hide-zero 1260~~~~~~~~~~~~~~~~~~~~ 1261 1262Set the option to hide zero values for xstats display:: 1263 1264 testpmd> set xstats-hide-zero on|off 1265 1266.. note:: 1267 1268 By default, the zero values are displayed for xstats. 1269 1270set port - rx/tx (for VF) 1271~~~~~~~~~~~~~~~~~~~~~~~~~ 1272 1273Set VF receive/transmit from a port:: 1274 1275 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off) 1276 1277set port - mac address filter (for VF) 1278~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1279 1280Add/Remove unicast or multicast MAC addr filter for a VF:: 1281 1282 testpmd> set port (port_id) vf (vf_id) (mac_addr) \ 1283 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off) 1284 1285set port - rx mode(for VF) 1286~~~~~~~~~~~~~~~~~~~~~~~~~~ 1287 1288Set the VF receive mode of a port:: 1289 1290 testpmd> set port (port_id) vf (vf_id) \ 1291 rxmode (AUPE|ROPE|BAM|MPE) (on|off) 1292 1293The available receive modes are: 1294 1295* ``AUPE``: Accepts untagged VLAN. 1296 1297* ``ROPE``: Accepts unicast hash. 1298 1299* ``BAM``: Accepts broadcast packets. 1300 1301* ``MPE``: Accepts all multicast packets. 1302 1303set port - tx_rate (for Queue) 1304~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1305 1306Set TX rate limitation for a queue on a port:: 1307 1308 testpmd> set port (port_id) queue (queue_id) rate (rate_value) 1309 1310set port - tx_rate (for VF) 1311~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1312 1313Set TX rate limitation for queues in VF on a port:: 1314 1315 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask) 1316 1317set port - mirror rule 1318~~~~~~~~~~~~~~~~~~~~~~ 1319 1320Set pool or vlan type mirror rule for a port:: 1321 1322 testpmd> set port (port_id) mirror-rule (rule_id) \ 1323 (pool-mirror-up|pool-mirror-down|vlan-mirror) \ 1324 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off) 1325 1326Set link mirror rule for a port:: 1327 1328 testpmd> set port (port_id) mirror-rule (rule_id) \ 1329 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off) 1330 1331For example to enable mirror traffic with vlan 0,1 to pool 0:: 1332 1333 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on 1334 1335reset port - mirror rule 1336~~~~~~~~~~~~~~~~~~~~~~~~ 1337 1338Reset a mirror rule for a port:: 1339 1340 testpmd> reset port (port_id) mirror-rule (rule_id) 1341 1342set flush_rx 1343~~~~~~~~~~~~ 1344 1345Set the flush on RX streams before forwarding. 1346The default is flush ``on``. 1347Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams:: 1348 1349 testpmd> set flush_rx off 1350 1351set bypass mode 1352~~~~~~~~~~~~~~~ 1353 1354Set the bypass mode for the lowest port on bypass enabled NIC:: 1355 1356 testpmd> set bypass mode (normal|bypass|isolate) (port_id) 1357 1358set bypass event 1359~~~~~~~~~~~~~~~~ 1360 1361Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled:: 1362 1363 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \ 1364 mode (normal|bypass|isolate) (port_id) 1365 1366Where: 1367 1368* ``timeout``: Enable bypass after watchdog timeout. 1369 1370* ``os_on``: Enable bypass when OS/board is powered on. 1371 1372* ``os_off``: Enable bypass when OS/board is powered off. 1373 1374* ``power_on``: Enable bypass when power supply is turned on. 1375 1376* ``power_off``: Enable bypass when power supply is turned off. 1377 1378 1379set bypass timeout 1380~~~~~~~~~~~~~~~~~~ 1381 1382Set the bypass watchdog timeout to ``n`` seconds where 0 = instant:: 1383 1384 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32) 1385 1386show bypass config 1387~~~~~~~~~~~~~~~~~~ 1388 1389Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC:: 1390 1391 testpmd> show bypass config (port_id) 1392 1393set link up 1394~~~~~~~~~~~ 1395 1396Set link up for a port:: 1397 1398 testpmd> set link-up port (port id) 1399 1400set link down 1401~~~~~~~~~~~~~ 1402 1403Set link down for a port:: 1404 1405 testpmd> set link-down port (port id) 1406 1407E-tag set 1408~~~~~~~~~ 1409 1410Enable E-tag insertion for a VF on a port:: 1411 1412 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id) 1413 1414Disable E-tag insertion for a VF on a port:: 1415 1416 testpmd> E-tag set insertion off port (port_id) vf (vf_id) 1417 1418Enable/disable E-tag stripping on a port:: 1419 1420 testpmd> E-tag set stripping (on|off) port (port_id) 1421 1422Enable/disable E-tag based forwarding on a port:: 1423 1424 testpmd> E-tag set forwarding (on|off) port (port_id) 1425 1426Add an E-tag forwarding filter on a port:: 1427 1428 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id) 1429 1430Delete an E-tag forwarding filter on a port:: 1431 testpmd> E-tag set filter del e-tag-id (value) port (port_id) 1432 1433ddp add 1434~~~~~~~ 1435 1436Load a dynamic device personalization (DDP) profile and store backup profile:: 1437 1438 testpmd> ddp add (port_id) (profile_path[,backup_profile_path]) 1439 1440ddp del 1441~~~~~~~ 1442 1443Delete a dynamic device personalization profile and restore backup profile:: 1444 1445 testpmd> ddp del (port_id) (backup_profile_path) 1446 1447ptype mapping 1448~~~~~~~~~~~~~ 1449 1450List all items from the ptype mapping table:: 1451 1452 testpmd> ptype mapping get (port_id) (valid_only) 1453 1454Where: 1455 1456* ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0). 1457 1458Replace a specific or a group of software defined ptype with a new one:: 1459 1460 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type) 1461 1462where: 1463 1464* ``target``: A specific software ptype or a mask to represent a group of software ptypes. 1465 1466* ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1). 1467 1468* ``pkt_type``: The new software ptype to replace the old ones. 1469 1470Update hardware defined ptype to software defined packet type mapping table:: 1471 1472 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype) 1473 1474where: 1475 1476* ``hw_ptype``: hardware ptype as the index of the ptype mapping table. 1477 1478* ``sw_ptype``: software ptype as the value of the ptype mapping table. 1479 1480Reset ptype mapping table:: 1481 1482 testpmd> ptype mapping reset (port_id) 1483 1484config per port Rx offloading 1485~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1486 1487Enable or disable a per port Rx offloading on all Rx queues of a port:: 1488 1489 testpmd> port config (port_id) rx_offload (offloading) on|off 1490 1491* ``offloading``: can be any of these offloading capability: 1492 vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro, 1493 qinq_strip, outer_ipv4_cksum, macsec_strip, 1494 header_split, vlan_filter, vlan_extend, jumbo_frame, 1495 crc_strip, scatter, timestamp, security, keep_crc 1496 1497This command should be run when the port is stopped, or else it will fail. 1498 1499config per queue Rx offloading 1500~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1501 1502Enable or disable a per queue Rx offloading only on a specific Rx queue:: 1503 1504 testpmd> port (port_id) rxq (queue_id) rx_offload (offloading) on|off 1505 1506* ``offloading``: can be any of these offloading capability: 1507 vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro, 1508 qinq_strip, outer_ipv4_cksum, macsec_strip, 1509 header_split, vlan_filter, vlan_extend, jumbo_frame, 1510 crc_strip, scatter, timestamp, security, keep_crc 1511 1512This command should be run when the port is stopped, or else it will fail. 1513 1514config per port Tx offloading 1515~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1516 1517Enable or disable a per port Tx offloading on all Tx queues of a port:: 1518 1519 testpmd> port config (port_id) tx_offload (offloading) on|off 1520 1521* ``offloading``: can be any of these offloading capability: 1522 vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum, 1523 sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum, 1524 qinq_insert, vxlan_tnl_tso, gre_tnl_tso, 1525 ipip_tnl_tso, geneve_tnl_tso, macsec_insert, 1526 mt_lockfree, multi_segs, mbuf_fast_free, security 1527 1528This command should be run when the port is stopped, or else it will fail. 1529 1530config per queue Tx offloading 1531~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1532 1533Enable or disable a per queue Tx offloading only on a specific Tx queue:: 1534 1535 testpmd> port (port_id) txq (queue_id) tx_offload (offloading) on|off 1536 1537* ``offloading``: can be any of these offloading capability: 1538 vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum, 1539 sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum, 1540 qinq_insert, vxlan_tnl_tso, gre_tnl_tso, 1541 ipip_tnl_tso, geneve_tnl_tso, macsec_insert, 1542 mt_lockfree, multi_segs, mbuf_fast_free, security 1543 1544This command should be run when the port is stopped, or else it will fail. 1545 1546Config VXLAN Encap outer layers 1547~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1548 1549Configure the outer layer to encapsulate a packet inside a VXLAN tunnel:: 1550 1551 set vxlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \ 1552 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) eth-src (eth-src) \ 1553 eth-dst (eth-dst) 1554 1555 set vxlan-with-vlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \ 1556 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) vlan-tci (vlan-tci) \ 1557 eth-src (eth-src) eth-dst (eth-dst) 1558 1559Those command will set an internal configuration inside testpmd, any following 1560flow rule using the action vxlan_encap will use the last configuration set. 1561To have a different encapsulation header, one of those commands must be called 1562before the flow rule creation. 1563 1564Config NVGRE Encap outer layers 1565~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1566 1567Configure the outer layer to encapsulate a packet inside a NVGRE tunnel:: 1568 1569 set nvgre ip-version (ipv4|ipv6) tni (tni) ip-src (ip-src) ip-dst (ip-dst) \ 1570 eth-src (eth-src) eth-dst (eth-dst) 1571 set nvgre-with-vlan ip-version (ipv4|ipv6) tni (tni) ip-src (ip-src) \ 1572 ip-dst (ip-dst) vlan-tci (vlan-tci) eth-src (eth-src) eth-dst (eth-dst) 1573 1574Those command will set an internal configuration inside testpmd, any following 1575flow rule using the action nvgre_encap will use the last configuration set. 1576To have a different encapsulation header, one of those commands must be called 1577before the flow rule creation. 1578 1579Port Functions 1580-------------- 1581 1582The following sections show functions for configuring ports. 1583 1584.. note:: 1585 1586 Port configuration changes only become active when forwarding is started/restarted. 1587 1588port attach 1589~~~~~~~~~~~ 1590 1591Attach a port specified by pci address or virtual device args:: 1592 1593 testpmd> port attach (identifier) 1594 1595To attach a new pci device, the device should be recognized by kernel first. 1596Then it should be moved under DPDK management. 1597Finally the port can be attached to testpmd. 1598 1599For example, to move a pci device using ixgbe under DPDK management: 1600 1601.. code-block:: console 1602 1603 # Check the status of the available devices. 1604 ./usertools/dpdk-devbind.py --status 1605 1606 Network devices using DPDK-compatible driver 1607 ============================================ 1608 <none> 1609 1610 Network devices using kernel driver 1611 =================================== 1612 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused= 1613 1614 1615 # Bind the device to igb_uio. 1616 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0 1617 1618 1619 # Recheck the status of the devices. 1620 ./usertools/dpdk-devbind.py --status 1621 Network devices using DPDK-compatible driver 1622 ============================================ 1623 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused= 1624 1625To attach a port created by virtual device, above steps are not needed. 1626 1627For example, to attach a port whose pci address is 0000:0a:00.0. 1628 1629.. code-block:: console 1630 1631 testpmd> port attach 0000:0a:00.0 1632 Attaching a new port... 1633 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1634 EAL: probe driver: 8086:10fb rte_ixgbe_pmd 1635 EAL: PCI memory mapped at 0x7f83bfa00000 1636 EAL: PCI memory mapped at 0x7f83bfa80000 1637 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5 1638 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb 1639 Port 0 is attached. Now total ports is 1 1640 Done 1641 1642For example, to attach a port created by pcap PMD. 1643 1644.. code-block:: console 1645 1646 testpmd> port attach net_pcap0 1647 Attaching a new port... 1648 PMD: Initializing pmd_pcap for net_pcap0 1649 PMD: Creating pcap-backed ethdev on numa socket 0 1650 Port 0 is attached. Now total ports is 1 1651 Done 1652 1653In this case, identifier is ``net_pcap0``. 1654This identifier format is the same as ``--vdev`` format of DPDK applications. 1655 1656For example, to re-attach a bonded port which has been previously detached, 1657the mode and slave parameters must be given. 1658 1659.. code-block:: console 1660 1661 testpmd> port attach net_bond_0,mode=0,slave=1 1662 Attaching a new port... 1663 EAL: Initializing pmd_bond for net_bond_0 1664 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0. 1665 Port 0 is attached. Now total ports is 1 1666 Done 1667 1668 1669port detach 1670~~~~~~~~~~~ 1671 1672Detach a specific port:: 1673 1674 testpmd> port detach (port_id) 1675 1676Before detaching a port, the port should be stopped and closed. 1677 1678For example, to detach a pci device port 0. 1679 1680.. code-block:: console 1681 1682 testpmd> port stop 0 1683 Stopping ports... 1684 Done 1685 testpmd> port close 0 1686 Closing ports... 1687 Done 1688 1689 testpmd> port detach 0 1690 Detaching a port... 1691 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1692 EAL: remove driver: 8086:10fb rte_ixgbe_pmd 1693 EAL: PCI memory unmapped at 0x7f83bfa00000 1694 EAL: PCI memory unmapped at 0x7f83bfa80000 1695 Done 1696 1697 1698For example, to detach a virtual device port 0. 1699 1700.. code-block:: console 1701 1702 testpmd> port stop 0 1703 Stopping ports... 1704 Done 1705 testpmd> port close 0 1706 Closing ports... 1707 Done 1708 1709 testpmd> port detach 0 1710 Detaching a port... 1711 PMD: Closing pcap ethdev on numa socket 0 1712 Port 'net_pcap0' is detached. Now total ports is 0 1713 Done 1714 1715To remove a pci device completely from the system, first detach the port from testpmd. 1716Then the device should be moved under kernel management. 1717Finally the device can be removed using kernel pci hotplug functionality. 1718 1719For example, to move a pci device under kernel management: 1720 1721.. code-block:: console 1722 1723 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0 1724 1725 ./usertools/dpdk-devbind.py --status 1726 1727 Network devices using DPDK-compatible driver 1728 ============================================ 1729 <none> 1730 1731 Network devices using kernel driver 1732 =================================== 1733 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio 1734 1735To remove a port created by a virtual device, above steps are not needed. 1736 1737port start 1738~~~~~~~~~~ 1739 1740Start all ports or a specific port:: 1741 1742 testpmd> port start (port_id|all) 1743 1744port stop 1745~~~~~~~~~ 1746 1747Stop all ports or a specific port:: 1748 1749 testpmd> port stop (port_id|all) 1750 1751port close 1752~~~~~~~~~~ 1753 1754Close all ports or a specific port:: 1755 1756 testpmd> port close (port_id|all) 1757 1758port config - queue ring size 1759~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1760 1761Configure a rx/tx queue ring size:: 1762 1763 testpmd> port (port_id) (rxq|txq) (queue_id) ring_size (value) 1764 1765Only take effect after command that (re-)start the port or command that setup specific queue. 1766 1767port start/stop queue 1768~~~~~~~~~~~~~~~~~~~~~ 1769 1770Start/stop a rx/tx queue on a specific port:: 1771 1772 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop) 1773 1774port config - queue deferred start 1775~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1776 1777Switch on/off deferred start of a specific port queue:: 1778 1779 testpmd> port (port_id) (rxq|txq) (queue_id) deferred_start (on|off) 1780 1781port setup queue 1782~~~~~~~~~~~~~~~~~~~~~ 1783 1784Setup a rx/tx queue on a specific port:: 1785 1786 testpmd> port (port_id) (rxq|txq) (queue_id) setup 1787 1788Only take effect when port is started. 1789 1790port config - speed 1791~~~~~~~~~~~~~~~~~~~ 1792 1793Set the speed and duplex mode for all ports or a specific port:: 1794 1795 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \ 1796 duplex (half|full|auto) 1797 1798port config - queues/descriptors 1799~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1800 1801Set number of queues/descriptors for rxq, txq, rxd and txd:: 1802 1803 testpmd> port config all (rxq|txq|rxd|txd) (value) 1804 1805This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options. 1806 1807port config - max-pkt-len 1808~~~~~~~~~~~~~~~~~~~~~~~~~ 1809 1810Set the maximum packet length:: 1811 1812 testpmd> port config all max-pkt-len (value) 1813 1814This is equivalent to the ``--max-pkt-len`` command-line option. 1815 1816port config - CRC Strip 1817~~~~~~~~~~~~~~~~~~~~~~~ 1818 1819Set hardware CRC stripping on or off for all ports:: 1820 1821 testpmd> port config all crc-strip (on|off) 1822 1823CRC stripping is on by default. 1824 1825The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option. 1826 1827port config - scatter 1828~~~~~~~~~~~~~~~~~~~~~~~ 1829 1830Set RX scatter mode on or off for all ports:: 1831 1832 testpmd> port config all scatter (on|off) 1833 1834RX scatter mode is off by default. 1835 1836The ``on`` option is equivalent to the ``--enable-scatter`` command-line option. 1837 1838port config - RX Checksum 1839~~~~~~~~~~~~~~~~~~~~~~~~~ 1840 1841Set hardware RX checksum offload to on or off for all ports:: 1842 1843 testpmd> port config all rx-cksum (on|off) 1844 1845Checksum offload is off by default. 1846 1847The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option. 1848 1849port config - VLAN 1850~~~~~~~~~~~~~~~~~~ 1851 1852Set hardware VLAN on or off for all ports:: 1853 1854 testpmd> port config all hw-vlan (on|off) 1855 1856Hardware VLAN is off by default. 1857 1858The ``on`` option is equivalent to the ``--enable-hw-vlan`` command-line option. 1859 1860port config - VLAN filter 1861~~~~~~~~~~~~~~~~~~~~~~~~~ 1862 1863Set hardware VLAN filter on or off for all ports:: 1864 1865 testpmd> port config all hw-vlan-filter (on|off) 1866 1867Hardware VLAN filter is off by default. 1868 1869The ``on`` option is equivalent to the ``--enable-hw-vlan-filter`` command-line option. 1870 1871port config - VLAN strip 1872~~~~~~~~~~~~~~~~~~~~~~~~ 1873 1874Set hardware VLAN strip on or off for all ports:: 1875 1876 testpmd> port config all hw-vlan-strip (on|off) 1877 1878Hardware VLAN strip is off by default. 1879 1880The ``on`` option is equivalent to the ``--enable-hw-vlan-strip`` command-line option. 1881 1882port config - VLAN extend 1883~~~~~~~~~~~~~~~~~~~~~~~~~ 1884 1885Set hardware VLAN extend on or off for all ports:: 1886 1887 testpmd> port config all hw-vlan-extend (on|off) 1888 1889Hardware VLAN extend is off by default. 1890 1891The ``on`` option is equivalent to the ``--enable-hw-vlan-extend`` command-line option. 1892 1893port config - Drop Packets 1894~~~~~~~~~~~~~~~~~~~~~~~~~~ 1895 1896Set packet drop for packets with no descriptors on or off for all ports:: 1897 1898 testpmd> port config all drop-en (on|off) 1899 1900Packet dropping for packets with no descriptors is off by default. 1901 1902The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option. 1903 1904port config - RSS 1905~~~~~~~~~~~~~~~~~ 1906 1907Set the RSS (Receive Side Scaling) mode on or off:: 1908 1909 testpmd> port config all rss (all|default|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none) 1910 1911RSS is on by default. 1912 1913The ``all`` option is equivalent to ip|tcp|udp|sctp|ether. 1914The ``default`` option enables all supported RSS types reported by device info. 1915The ``none`` option is equivalent to the ``--disable-rss`` command-line option. 1916 1917port config - RSS Reta 1918~~~~~~~~~~~~~~~~~~~~~~ 1919 1920Set the RSS (Receive Side Scaling) redirection table:: 1921 1922 testpmd> port config all rss reta (hash,queue)[,(hash,queue)] 1923 1924port config - DCB 1925~~~~~~~~~~~~~~~~~ 1926 1927Set the DCB mode for an individual port:: 1928 1929 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off) 1930 1931The traffic class should be 4 or 8. 1932 1933port config - Burst 1934~~~~~~~~~~~~~~~~~~~ 1935 1936Set the number of packets per burst:: 1937 1938 testpmd> port config all burst (value) 1939 1940This is equivalent to the ``--burst`` command-line option. 1941 1942port config - Threshold 1943~~~~~~~~~~~~~~~~~~~~~~~ 1944 1945Set thresholds for TX/RX queues:: 1946 1947 testpmd> port config all (threshold) (value) 1948 1949Where the threshold type can be: 1950 1951* ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255. 1952 1953* ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255. 1954 1955* ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255. 1956 1957* ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255. 1958 1959* ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255. 1960 1961* ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255. 1962 1963* ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd. 1964 1965* ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd. 1966 1967* ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd. 1968 1969These threshold options are also available from the command-line. 1970 1971port config - E-tag 1972~~~~~~~~~~~~~~~~~~~ 1973 1974Set the value of ether-type for E-tag:: 1975 1976 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value) 1977 1978Enable/disable the E-tag support:: 1979 1980 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable) 1981 1982port config pctype mapping 1983~~~~~~~~~~~~~~~~~~~~~~~~~~ 1984 1985Reset pctype mapping table:: 1986 1987 testpmd> port config (port_id) pctype mapping reset 1988 1989Update hardware defined pctype to software defined flow type mapping table:: 1990 1991 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id) 1992 1993where: 1994 1995* ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table. 1996 1997* ``flow_type_id``: software flow type id as the index of the pctype mapping table. 1998 1999port config input set 2000~~~~~~~~~~~~~~~~~~~~~ 2001 2002Config RSS/FDIR/FDIR flexible payload input set for some pctype:: 2003 testpmd> port config (port_id) pctype (pctype_id) \ 2004 (hash_inset|fdir_inset|fdir_flx_inset) \ 2005 (get|set|clear) field (field_idx) 2006 2007Clear RSS/FDIR/FDIR flexible payload input set for some pctype:: 2008 testpmd> port config (port_id) pctype (pctype_id) \ 2009 (hash_inset|fdir_inset|fdir_flx_inset) clear all 2010 2011where: 2012 2013* ``pctype_id``: hardware packet classification types. 2014* ``field_idx``: hardware field index. 2015 2016port config udp_tunnel_port 2017~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2018 2019Add/remove UDP tunnel port for VXLAN/GENEVE tunneling protocols:: 2020 testpmd> port config (port_id) udp_tunnel_port add|rm vxlan|geneve (udp_port) 2021 2022Link Bonding Functions 2023---------------------- 2024 2025The Link Bonding functions make it possible to dynamically create and 2026manage link bonding devices from within testpmd interactive prompt. 2027 2028create bonded device 2029~~~~~~~~~~~~~~~~~~~~ 2030 2031Create a new bonding device:: 2032 2033 testpmd> create bonded device (mode) (socket) 2034 2035For example, to create a bonded device in mode 1 on socket 0:: 2036 2037 testpmd> create bonded device 1 0 2038 created new bonded device (port X) 2039 2040add bonding slave 2041~~~~~~~~~~~~~~~~~ 2042 2043Adds Ethernet device to a Link Bonding device:: 2044 2045 testpmd> add bonding slave (slave id) (port id) 2046 2047For example, to add Ethernet device (port 6) to a Link Bonding device (port 10):: 2048 2049 testpmd> add bonding slave 6 10 2050 2051 2052remove bonding slave 2053~~~~~~~~~~~~~~~~~~~~ 2054 2055Removes an Ethernet slave device from a Link Bonding device:: 2056 2057 testpmd> remove bonding slave (slave id) (port id) 2058 2059For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10):: 2060 2061 testpmd> remove bonding slave 6 10 2062 2063set bonding mode 2064~~~~~~~~~~~~~~~~ 2065 2066Set the Link Bonding mode of a Link Bonding device:: 2067 2068 testpmd> set bonding mode (value) (port id) 2069 2070For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3):: 2071 2072 testpmd> set bonding mode 3 10 2073 2074set bonding primary 2075~~~~~~~~~~~~~~~~~~~ 2076 2077Set an Ethernet slave device as the primary device on a Link Bonding device:: 2078 2079 testpmd> set bonding primary (slave id) (port id) 2080 2081For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10):: 2082 2083 testpmd> set bonding primary 6 10 2084 2085set bonding mac 2086~~~~~~~~~~~~~~~ 2087 2088Set the MAC address of a Link Bonding device:: 2089 2090 testpmd> set bonding mac (port id) (mac) 2091 2092For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01:: 2093 2094 testpmd> set bonding mac 10 00:00:00:00:00:01 2095 2096set bonding xmit_balance_policy 2097~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2098 2099Set the transmission policy for a Link Bonding device when it is in Balance XOR mode:: 2100 2101 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34) 2102 2103For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports):: 2104 2105 testpmd> set bonding xmit_balance_policy 10 l34 2106 2107 2108set bonding mon_period 2109~~~~~~~~~~~~~~~~~~~~~~ 2110 2111Set the link status monitoring polling period in milliseconds for a bonding device. 2112 2113This adds support for PMD slave devices which do not support link status interrupts. 2114When the mon_period is set to a value greater than 0 then all PMD's which do not support 2115link status ISR will be queried every polling interval to check if their link status has changed:: 2116 2117 testpmd> set bonding mon_period (port_id) (value) 2118 2119For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms:: 2120 2121 testpmd> set bonding mon_period 5 150 2122 2123 2124set bonding lacp dedicated_queue 2125~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2126 2127Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic 2128when in mode 4 (link-aggregration-802.3ad):: 2129 2130 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable) 2131 2132 2133set bonding agg_mode 2134~~~~~~~~~~~~~~~~~~~~ 2135 2136Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad):: 2137 2138 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable) 2139 2140 2141show bonding config 2142~~~~~~~~~~~~~~~~~~~ 2143 2144Show the current configuration of a Link Bonding device:: 2145 2146 testpmd> show bonding config (port id) 2147 2148For example, 2149to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4) 2150in balance mode with a transmission policy of layer 2+3:: 2151 2152 testpmd> show bonding config 9 2153 Bonding mode: 2 2154 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23 2155 Slaves (3): [1 3 4] 2156 Active Slaves (3): [1 3 4] 2157 Primary: [3] 2158 2159 2160Register Functions 2161------------------ 2162 2163The Register Functions can be used to read from and write to registers on the network card referenced by a port number. 2164This is mainly useful for debugging purposes. 2165Reference should be made to the appropriate datasheet for the network card for details on the register addresses 2166and fields that can be accessed. 2167 2168read reg 2169~~~~~~~~ 2170 2171Display the value of a port register:: 2172 2173 testpmd> read reg (port_id) (address) 2174 2175For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller:: 2176 2177 testpmd> read reg 0 0xEE00 2178 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241) 2179 2180read regfield 2181~~~~~~~~~~~~~ 2182 2183Display a port register bit field:: 2184 2185 testpmd> read regfield (port_id) (address) (bit_x) (bit_y) 2186 2187For example, reading the lowest two bits from the register in the example above:: 2188 2189 testpmd> read regfield 0 0xEE00 0 1 2190 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1) 2191 2192read regbit 2193~~~~~~~~~~~ 2194 2195Display a single port register bit:: 2196 2197 testpmd> read regbit (port_id) (address) (bit_x) 2198 2199For example, reading the lowest bit from the register in the example above:: 2200 2201 testpmd> read regbit 0 0xEE00 0 2202 port 0 PCI register at offset 0xEE00: bit 0=1 2203 2204write reg 2205~~~~~~~~~ 2206 2207Set the value of a port register:: 2208 2209 testpmd> write reg (port_id) (address) (value) 2210 2211For example, to clear a register:: 2212 2213 testpmd> write reg 0 0xEE00 0x0 2214 port 0 PCI register at offset 0xEE00: 0x00000000 (0) 2215 2216write regfield 2217~~~~~~~~~~~~~~ 2218 2219Set bit field of a port register:: 2220 2221 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value) 2222 2223For example, writing to the register cleared in the example above:: 2224 2225 testpmd> write regfield 0 0xEE00 0 1 2 2226 port 0 PCI register at offset 0xEE00: 0x00000002 (2) 2227 2228write regbit 2229~~~~~~~~~~~~ 2230 2231Set single bit value of a port register:: 2232 2233 testpmd> write regbit (port_id) (address) (bit_x) (value) 2234 2235For example, to set the high bit in the register from the example above:: 2236 2237 testpmd> write regbit 0 0xEE00 31 1 2238 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658) 2239 2240Traffic Metering and Policing 2241----------------------------- 2242 2243The following section shows functions for configuring traffic metering and 2244policing on the ethernet device through the use of generic ethdev API. 2245 2246show port traffic management capability 2247~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2248 2249Show traffic metering and policing capability of the port:: 2250 2251 testpmd> show port meter cap (port_id) 2252 2253add port meter profile (srTCM rfc2967) 2254~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2255 2256Add meter profile (srTCM rfc2697) to the ethernet device:: 2257 2258 testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \ 2259 (cir) (cbs) (ebs) 2260 2261where: 2262 2263* ``profile_id``: ID for the meter profile. 2264* ``cir``: Committed Information Rate (CIR) (bytes/second). 2265* ``cbs``: Committed Burst Size (CBS) (bytes). 2266* ``ebs``: Excess Burst Size (EBS) (bytes). 2267 2268add port meter profile (trTCM rfc2968) 2269~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2270 2271Add meter profile (srTCM rfc2698) to the ethernet device:: 2272 2273 testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \ 2274 (cir) (pir) (cbs) (pbs) 2275 2276where: 2277 2278* ``profile_id``: ID for the meter profile. 2279* ``cir``: Committed information rate (bytes/second). 2280* ``pir``: Peak information rate (bytes/second). 2281* ``cbs``: Committed burst size (bytes). 2282* ``pbs``: Peak burst size (bytes). 2283 2284add port meter profile (trTCM rfc4115) 2285~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2286 2287Add meter profile (trTCM rfc4115) to the ethernet device:: 2288 2289 testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \ 2290 (cir) (eir) (cbs) (ebs) 2291 2292where: 2293 2294* ``profile_id``: ID for the meter profile. 2295* ``cir``: Committed information rate (bytes/second). 2296* ``eir``: Excess information rate (bytes/second). 2297* ``cbs``: Committed burst size (bytes). 2298* ``ebs``: Excess burst size (bytes). 2299 2300delete port meter profile 2301~~~~~~~~~~~~~~~~~~~~~~~~~ 2302 2303Delete meter profile from the ethernet device:: 2304 2305 testpmd> del port meter profile (port_id) (profile_id) 2306 2307create port meter 2308~~~~~~~~~~~~~~~~~ 2309 2310Create new meter object for the ethernet device:: 2311 2312 testpmd> create port meter (port_id) (mtr_id) (profile_id) \ 2313 (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \ 2314 (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\ 2315 (dscp_tbl_entry63)] 2316 2317where: 2318 2319* ``mtr_id``: meter object ID. 2320* ``profile_id``: ID for the meter profile. 2321* ``meter_enable``: When this parameter has a non-zero value, the meter object 2322 gets enabled at the time of creation, otherwise remains disabled. 2323* ``g_action``: Policer action for the packet with green color. 2324* ``y_action``: Policer action for the packet with yellow color. 2325* ``r_action``: Policer action for the packet with red color. 2326* ``stats_mask``: Mask of statistics counter types to be enabled for the 2327 meter object. 2328* ``shared``: When this parameter has a non-zero value, the meter object is 2329 shared by multiple flows. Otherwise, meter object is used by single flow. 2330* ``use_pre_meter_color``: When this parameter has a non-zero value, the 2331 input color for the current meter object is determined by the latest meter 2332 object in the same flow. Otherwise, the current meter object uses the 2333 *dscp_table* to determine the input color. 2334* ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input 2335 color, 0 <= x <= 63. 2336 2337enable port meter 2338~~~~~~~~~~~~~~~~~ 2339 2340Enable meter for the ethernet device:: 2341 2342 testpmd> enable port meter (port_id) (mtr_id) 2343 2344disable port meter 2345~~~~~~~~~~~~~~~~~~ 2346 2347Disable meter for the ethernet device:: 2348 2349 testpmd> disable port meter (port_id) (mtr_id) 2350 2351delete port meter 2352~~~~~~~~~~~~~~~~~ 2353 2354Delete meter for the ethernet device:: 2355 2356 testpmd> del port meter (port_id) (mtr_id) 2357 2358Set port meter profile 2359~~~~~~~~~~~~~~~~~~~~~~ 2360 2361Set meter profile for the ethernet device:: 2362 2363 testpmd> set port meter profile (port_id) (mtr_id) (profile_id) 2364 2365set port meter dscp table 2366~~~~~~~~~~~~~~~~~~~~~~~~~ 2367 2368Set meter dscp table for the ethernet device:: 2369 2370 testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \ 2371 (dscp_tbl_entry1)...(dscp_tbl_entry63)] 2372 2373set port meter policer action 2374~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2375 2376Set meter policer action for the ethernet device:: 2377 2378 testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \ 2379 (action0) [(action1) (action1)] 2380 2381where: 2382 2383* ``action_mask``: Bit mask indicating which policer actions need to be 2384 updated. One or more policer actions can be updated in a single function 2385 invocation. To update the policer action associated with color C, bit 2386 (1 << C) needs to be set in *action_mask* and element at position C 2387 in the *actions* array needs to be valid. 2388* ``actionx``: Policer action for the color x, 2389 RTE_MTR_GREEN <= x < RTE_MTR_COLORS 2390 2391set port meter stats mask 2392~~~~~~~~~~~~~~~~~~~~~~~~~ 2393 2394Set meter stats mask for the ethernet device:: 2395 2396 testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask) 2397 2398where: 2399 2400* ``stats_mask``: Bit mask indicating statistics counter types to be enabled. 2401 2402show port meter stats 2403~~~~~~~~~~~~~~~~~~~~~ 2404 2405Show meter stats of the ethernet device:: 2406 2407 testpmd> show port meter stats (port_id) (mtr_id) (clear) 2408 2409where: 2410 2411* ``clear``: Flag that indicates whether the statistics counters should 2412 be cleared (i.e. set to zero) immediately after they have been read or not. 2413 2414Traffic Management 2415------------------ 2416 2417The following section shows functions for configuring traffic management on 2418on the ethernet device through the use of generic TM API. 2419 2420show port traffic management capability 2421~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2422 2423Show traffic management capability of the port:: 2424 2425 testpmd> show port tm cap (port_id) 2426 2427show port traffic management capability (hierarchy level) 2428~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2429 2430Show traffic management hierarchy level capability of the port:: 2431 2432 testpmd> show port tm level cap (port_id) (level_id) 2433 2434show port traffic management capability (hierarchy node level) 2435~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2436 2437Show the traffic management hierarchy node capability of the port:: 2438 2439 testpmd> show port tm node cap (port_id) (node_id) 2440 2441show port traffic management hierarchy node type 2442~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2443 2444Show the port traffic management hierarchy node type:: 2445 2446 testpmd> show port tm node type (port_id) (node_id) 2447 2448show port traffic management hierarchy node stats 2449~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2450 2451Show the port traffic management hierarchy node statistics:: 2452 2453 testpmd> show port tm node stats (port_id) (node_id) (clear) 2454 2455where: 2456 2457* ``clear``: When this parameter has a non-zero value, the statistics counters 2458 are cleared (i.e. set to zero) immediately after they have been read, 2459 otherwise the statistics counters are left untouched. 2460 2461Add port traffic management private shaper profile 2462~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2463 2464Add the port traffic management private shaper profile:: 2465 2466 testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \ 2467 (tb_rate) (tb_size) (packet_length_adjust) 2468 2469where: 2470 2471* ``shaper_profile id``: Shaper profile ID for the new profile. 2472* ``tb_rate``: Token bucket rate (bytes per second). 2473* ``tb_size``: Token bucket size (bytes). 2474* ``packet_length_adjust``: The value (bytes) to be added to the length of 2475 each packet for the purpose of shaping. This parameter value can be used to 2476 correct the packet length with the framing overhead bytes that are consumed 2477 on the wire. 2478 2479Delete port traffic management private shaper profile 2480~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2481 2482Delete the port traffic management private shaper:: 2483 2484 testpmd> del port tm node shaper profile (port_id) (shaper_profile_id) 2485 2486where: 2487 2488* ``shaper_profile id``: Shaper profile ID that needs to be deleted. 2489 2490Add port traffic management shared shaper 2491~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2492 2493Create the port traffic management shared shaper:: 2494 2495 testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \ 2496 (shaper_profile_id) 2497 2498where: 2499 2500* ``shared_shaper_id``: Shared shaper ID to be created. 2501* ``shaper_profile id``: Shaper profile ID for shared shaper. 2502 2503Set port traffic management shared shaper 2504~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2505 2506Update the port traffic management shared shaper:: 2507 2508 testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \ 2509 (shaper_profile_id) 2510 2511where: 2512 2513* ``shared_shaper_id``: Shared shaper ID to be update. 2514* ``shaper_profile id``: Shaper profile ID for shared shaper. 2515 2516Delete port traffic management shared shaper 2517~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2518 2519Delete the port traffic management shared shaper:: 2520 2521 testpmd> del port tm node shared shaper (port_id) (shared_shaper_id) 2522 2523where: 2524 2525* ``shared_shaper_id``: Shared shaper ID to be deleted. 2526 2527Set port traffic management hiearchy node private shaper 2528~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2529 2530set the port traffic management hierarchy node private shaper:: 2531 2532 testpmd> set port tm node shaper profile (port_id) (node_id) \ 2533 (shaper_profile_id) 2534 2535where: 2536 2537* ``shaper_profile id``: Private shaper profile ID to be enabled on the 2538 hierarchy node. 2539 2540Add port traffic management WRED profile 2541~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2542 2543Create a new WRED profile:: 2544 2545 testpmd> add port tm node wred profile (port_id) (wred_profile_id) \ 2546 (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \ 2547 (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \ 2548 (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r) 2549 2550where: 2551 2552* ``wred_profile id``: Identifier for the newly create WRED profile 2553* ``color_g``: Packet color (green) 2554* ``min_th_g``: Minimum queue threshold for packet with green color 2555* ``max_th_g``: Minimum queue threshold for packet with green color 2556* ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp) 2557* ``wq_log2_g``: Negated log2 of queue weight (wq) 2558* ``color_y``: Packet color (yellow) 2559* ``min_th_y``: Minimum queue threshold for packet with yellow color 2560* ``max_th_y``: Minimum queue threshold for packet with yellow color 2561* ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp) 2562* ``wq_log2_y``: Negated log2 of queue weight (wq) 2563* ``color_r``: Packet color (red) 2564* ``min_th_r``: Minimum queue threshold for packet with yellow color 2565* ``max_th_r``: Minimum queue threshold for packet with yellow color 2566* ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp) 2567* ``wq_log2_r``: Negated log2 of queue weight (wq) 2568 2569Delete port traffic management WRED profile 2570~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2571 2572Delete the WRED profile:: 2573 2574 testpmd> del port tm node wred profile (port_id) (wred_profile_id) 2575 2576Add port traffic management hierarchy nonleaf node 2577~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2578 2579Add nonleaf node to port traffic management hiearchy:: 2580 2581 testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \ 2582 (priority) (weight) (level_id) (shaper_profile_id) \ 2583 (n_sp_priorities) (stats_mask) (n_shared_shapers) \ 2584 [(shared_shaper_0) (shared_shaper_1) ...] \ 2585 2586where: 2587 2588* ``parent_node_id``: Node ID of the parent. 2589* ``priority``: Node priority (highest node priority is zero). This is used by 2590 the SP algorithm running on the parent node for scheduling this node. 2591* ``weight``: Node weight (lowest weight is one). The node weight is relative 2592 to the weight sum of all siblings that have the same priority. It is used by 2593 the WFQ algorithm running on the parent node for scheduling this node. 2594* ``level_id``: Hiearchy level of the node. 2595* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by 2596 the node. 2597* ``n_sp_priorities``: Number of strict priorities. 2598* ``stats_mask``: Mask of statistics counter types to be enabled for this node. 2599* ``n_shared_shapers``: Number of shared shapers. 2600* ``shared_shaper_id``: Shared shaper id. 2601 2602Add port traffic management hierarchy leaf node 2603~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2604 2605Add leaf node to port traffic management hiearchy:: 2606 2607 testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \ 2608 (priority) (weight) (level_id) (shaper_profile_id) \ 2609 (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \ 2610 [(shared_shaper_id) (shared_shaper_id) ...] \ 2611 2612where: 2613 2614* ``parent_node_id``: Node ID of the parent. 2615* ``priority``: Node priority (highest node priority is zero). This is used by 2616 the SP algorithm running on the parent node for scheduling this node. 2617* ``weight``: Node weight (lowest weight is one). The node weight is relative 2618 to the weight sum of all siblings that have the same priority. It is used by 2619 the WFQ algorithm running on the parent node for scheduling this node. 2620* ``level_id``: Hiearchy level of the node. 2621* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by 2622 the node. 2623* ``cman_mode``: Congestion management mode to be enabled for this node. 2624* ``wred_profile_id``: WRED profile id to be enabled for this node. 2625* ``stats_mask``: Mask of statistics counter types to be enabled for this node. 2626* ``n_shared_shapers``: Number of shared shapers. 2627* ``shared_shaper_id``: Shared shaper id. 2628 2629Delete port traffic management hierarchy node 2630~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2631 2632Delete node from port traffic management hiearchy:: 2633 2634 testpmd> del port tm node (port_id) (node_id) 2635 2636Update port traffic management hierarchy parent node 2637~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2638 2639Update port traffic management hierarchy parent node:: 2640 2641 testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \ 2642 (priority) (weight) 2643 2644This function can only be called after the hierarchy commit invocation. Its 2645success depends on the port support for this operation, as advertised through 2646the port capability set. This function is valid for all nodes of the traffic 2647management hierarchy except root node. 2648 2649Suspend port traffic management hierarchy node 2650~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2651 2652 testpmd> suspend port tm node (port_id) (node_id) 2653 2654Resume port traffic management hierarchy node 2655~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2656 2657 testpmd> resume port tm node (port_id) (node_id) 2658 2659Commit port traffic management hierarchy 2660~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2661 2662Commit the traffic management hierarchy on the port:: 2663 2664 testpmd> port tm hierarchy commit (port_id) (clean_on_fail) 2665 2666where: 2667 2668* ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function 2669 call failure. On the other hand, hierarchy is preserved when this parameter 2670 is equal to zero. 2671 2672Set port traffic management default hierarchy (softnic forwarding mode) 2673~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2674 2675set the traffic management default hierarchy on the port:: 2676 2677 testpmd> set port tm hierarchy default (port_id) 2678 2679Filter Functions 2680---------------- 2681 2682This section details the available filter functions that are available. 2683 2684Note these functions interface the deprecated legacy filtering framework, 2685superseded by *rte_flow*. See `Flow rules management`_. 2686 2687ethertype_filter 2688~~~~~~~~~~~~~~~~~~~~ 2689 2690Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue:: 2691 2692 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \ 2693 ethertype (ether_type) (drop|fwd) queue (queue_id) 2694 2695The available information parameters are: 2696 2697* ``port_id``: The port which the Ethertype filter assigned on. 2698 2699* ``mac_addr``: Compare destination mac address. 2700 2701* ``mac_ignr``: Ignore destination mac address match. 2702 2703* ``mac_address``: Destination mac address to match. 2704 2705* ``ether_type``: The EtherType value want to match, 2706 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid. 2707 2708* ``queue_id``: The receive queue associated with this EtherType filter. 2709 It is meaningless when deleting or dropping. 2710 2711Example, to add/remove an ethertype filter rule:: 2712 2713 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \ 2714 ethertype 0x0806 fwd queue 3 2715 2716 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \ 2717 ethertype 0x0806 fwd queue 3 2718 27192tuple_filter 2720~~~~~~~~~~~~~~~~~ 2721 2722Add or delete a 2-tuple filter, 2723which identifies packets by specific protocol and destination TCP/UDP port 2724and forwards packets into one of the receive queues:: 2725 2726 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \ 2727 protocol (protocol_value) mask (mask_value) \ 2728 tcp_flags (tcp_flags_value) priority (prio_value) \ 2729 queue (queue_id) 2730 2731The available information parameters are: 2732 2733* ``port_id``: The port which the 2-tuple filter assigned on. 2734 2735* ``dst_port_value``: Destination port in L4. 2736 2737* ``protocol_value``: IP L4 protocol. 2738 2739* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate. 2740 2741* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP). 2742 2743* ``prio_value``: Priority of this filter. 2744 2745* ``queue_id``: The receive queue associated with this 2-tuple filter. 2746 2747Example, to add/remove an 2tuple filter rule:: 2748 2749 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \ 2750 tcp_flags 0x02 priority 3 queue 3 2751 2752 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \ 2753 tcp_flags 0x02 priority 3 queue 3 2754 27555tuple_filter 2756~~~~~~~~~~~~~~~~~ 2757 2758Add or delete a 5-tuple filter, 2759which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port) 2760and routes packets into one of the receive queues:: 2761 2762 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \ 2763 (src_address) dst_port (dst_port_value) \ 2764 src_port (src_port_value) protocol (protocol_value) \ 2765 mask (mask_value) tcp_flags (tcp_flags_value) \ 2766 priority (prio_value) queue (queue_id) 2767 2768The available information parameters are: 2769 2770* ``port_id``: The port which the 5-tuple filter assigned on. 2771 2772* ``dst_address``: Destination IP address. 2773 2774* ``src_address``: Source IP address. 2775 2776* ``dst_port_value``: TCP/UDP destination port. 2777 2778* ``src_port_value``: TCP/UDP source port. 2779 2780* ``protocol_value``: L4 protocol. 2781 2782* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate 2783 2784* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP). 2785 2786* ``prio_value``: The priority of this filter. 2787 2788* ``queue_id``: The receive queue associated with this 5-tuple filter. 2789 2790Example, to add/remove an 5tuple filter rule:: 2791 2792 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 2793 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 2794 flags 0x0 priority 3 queue 3 2795 2796 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 2797 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 2798 flags 0x0 priority 3 queue 3 2799 2800syn_filter 2801~~~~~~~~~~ 2802 2803Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue:: 2804 2805 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id) 2806 2807The available information parameters are: 2808 2809* ``port_id``: The port which the SYN filter assigned on. 2810 2811* ``high``: This SYN filter has higher priority than other filters. 2812 2813* ``low``: This SYN filter has lower priority than other filters. 2814 2815* ``queue_id``: The receive queue associated with this SYN filter 2816 2817Example:: 2818 2819 testpmd> syn_filter 0 add priority high queue 3 2820 2821flex_filter 2822~~~~~~~~~~~ 2823 2824With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet 2825and routed into one of the receive queues:: 2826 2827 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \ 2828 mask (mask_value) priority (prio_value) queue (queue_id) 2829 2830The available information parameters are: 2831 2832* ``port_id``: The port which the Flex filter is assigned on. 2833 2834* ``len_value``: Filter length in bytes, no greater than 128. 2835 2836* ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match. 2837 2838* ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match. 2839 2840* ``prio_value``: The priority of this filter. 2841 2842* ``queue_id``: The receive queue associated with this Flex filter. 2843 2844Example:: 2845 2846 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \ 2847 mask 000C priority 3 queue 3 2848 2849 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \ 2850 mask 000C priority 3 queue 3 2851 2852 2853.. _testpmd_flow_director: 2854 2855flow_director_filter 2856~~~~~~~~~~~~~~~~~~~~ 2857 2858The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues. 2859 2860Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and 2861Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter: 2862 2863* Perfect match filters. 2864 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2865 The masked fields are for IP flow. 2866 2867* Signature filters. 2868 The hardware checks a match between a hash-based signature of the masked fields of the received packet. 2869 2870* Perfect-mac-vlan match filters. 2871 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2872 The masked fields are for MAC VLAN flow. 2873 2874* Perfect-tunnel match filters. 2875 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2876 The masked fields are for tunnel flow. 2877 2878* Perfect-raw-flow-type match filters. 2879 The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet. 2880 The masked fields are specified by input sets. 2881 2882The Flow Director filters can match the different fields for different type of packet: flow type, specific input set 2883per flow type and the flexible payload. 2884 2885The Flow Director can also mask out parts of all of these fields so that filters 2886are only applied to certain fields or parts of the fields. 2887 2888Note that for raw flow type mode the source and destination fields in the 2889raw packet buffer need to be presented in a reversed order with respect 2890to the expected received packets. 2891For example: IP source and destination addresses or TCP/UDP/SCTP 2892source and destination ports 2893 2894Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information. 2895 2896# Commands to add flow director filters of different flow types:: 2897 2898 flow_director_filter (port_id) mode IP (add|del|update) \ 2899 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \ 2900 src (src_ip_address) dst (dst_ip_address) \ 2901 tos (tos_value) proto (proto_value) ttl (ttl_value) \ 2902 vlan (vlan_value) flexbytes (flexbytes_value) \ 2903 (drop|fwd) pf|vf(vf_id) queue (queue_id) \ 2904 fd_id (fd_id_value) 2905 2906 flow_director_filter (port_id) mode IP (add|del|update) \ 2907 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \ 2908 src (src_ip_address) (src_port) \ 2909 dst (dst_ip_address) (dst_port) \ 2910 tos (tos_value) ttl (ttl_value) \ 2911 vlan (vlan_value) flexbytes (flexbytes_value) \ 2912 (drop|fwd) queue pf|vf(vf_id) (queue_id) \ 2913 fd_id (fd_id_value) 2914 2915 flow_director_filter (port_id) mode IP (add|del|update) \ 2916 flow (ipv4-sctp|ipv6-sctp) \ 2917 src (src_ip_address) (src_port) \ 2918 dst (dst_ip_address) (dst_port) \ 2919 tos (tos_value) ttl (ttl_value) \ 2920 tag (verification_tag) vlan (vlan_value) \ 2921 flexbytes (flexbytes_value) (drop|fwd) \ 2922 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value) 2923 2924 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \ 2925 ether (ethertype) flexbytes (flexbytes_value) \ 2926 (drop|fwd) pf|vf(vf_id) queue (queue_id) 2927 fd_id (fd_id_value) 2928 2929 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \ 2930 mac (mac_address) vlan (vlan_value) \ 2931 flexbytes (flexbytes_value) (drop|fwd) \ 2932 queue (queue_id) fd_id (fd_id_value) 2933 2934 flow_director_filter (port_id) mode Tunnel (add|del|update) \ 2935 mac (mac_address) vlan (vlan_value) \ 2936 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \ 2937 flexbytes (flexbytes_value) (drop|fwd) \ 2938 queue (queue_id) fd_id (fd_id_value) 2939 2940 flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \ 2941 (drop|fwd) queue (queue_id) fd_id (fd_id_value) \ 2942 packet (packet file name) 2943 2944For example, to add an ipv4-udp flow type filter:: 2945 2946 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \ 2947 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \ 2948 fwd pf queue 1 fd_id 1 2949 2950For example, add an ipv4-other flow type filter:: 2951 2952 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \ 2953 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \ 2954 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1 2955 2956flush_flow_director 2957~~~~~~~~~~~~~~~~~~~ 2958 2959Flush all flow director filters on a device:: 2960 2961 testpmd> flush_flow_director (port_id) 2962 2963Example, to flush all flow director filter on port 0:: 2964 2965 testpmd> flush_flow_director 0 2966 2967flow_director_mask 2968~~~~~~~~~~~~~~~~~~ 2969 2970Set flow director's input masks:: 2971 2972 flow_director_mask (port_id) mode IP vlan (vlan_value) \ 2973 src_mask (ipv4_src) (ipv6_src) (src_port) \ 2974 dst_mask (ipv4_dst) (ipv6_dst) (dst_port) 2975 2976 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value) 2977 2978 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \ 2979 mac (mac_value) tunnel-type (tunnel_type_value) \ 2980 tunnel-id (tunnel_id_value) 2981 2982Example, to set flow director mask on port 0:: 2983 2984 testpmd> flow_director_mask 0 mode IP vlan 0xefff \ 2985 src_mask 255.255.255.255 \ 2986 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \ 2987 dst_mask 255.255.255.255 \ 2988 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF 2989 2990flow_director_flex_mask 2991~~~~~~~~~~~~~~~~~~~~~~~ 2992 2993set masks of flow director's flexible payload based on certain flow type:: 2994 2995 testpmd> flow_director_flex_mask (port_id) \ 2996 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2997 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \ 2998 l2_payload|all) (mask) 2999 3000Example, to set flow director's flex mask for all flow type on port 0:: 3001 3002 testpmd> flow_director_flex_mask 0 flow all \ 3003 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0) 3004 3005 3006flow_director_flex_payload 3007~~~~~~~~~~~~~~~~~~~~~~~~~~ 3008 3009Configure flexible payload selection:: 3010 3011 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config) 3012 3013For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload:: 3014 3015 testpmd> flow_director_flex_payload 0 l4 \ 3016 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19) 3017 3018get_sym_hash_ena_per_port 3019~~~~~~~~~~~~~~~~~~~~~~~~~ 3020 3021Get symmetric hash enable configuration per port:: 3022 3023 get_sym_hash_ena_per_port (port_id) 3024 3025For example, to get symmetric hash enable configuration of port 1:: 3026 3027 testpmd> get_sym_hash_ena_per_port 1 3028 3029set_sym_hash_ena_per_port 3030~~~~~~~~~~~~~~~~~~~~~~~~~ 3031 3032Set symmetric hash enable configuration per port to enable or disable:: 3033 3034 set_sym_hash_ena_per_port (port_id) (enable|disable) 3035 3036For example, to set symmetric hash enable configuration of port 1 to enable:: 3037 3038 testpmd> set_sym_hash_ena_per_port 1 enable 3039 3040get_hash_global_config 3041~~~~~~~~~~~~~~~~~~~~~~ 3042 3043Get the global configurations of hash filters:: 3044 3045 get_hash_global_config (port_id) 3046 3047For example, to get the global configurations of hash filters of port 1:: 3048 3049 testpmd> get_hash_global_config 1 3050 3051set_hash_global_config 3052~~~~~~~~~~~~~~~~~~~~~~ 3053 3054Set the global configurations of hash filters:: 3055 3056 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \ 3057 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \ 3058 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \ 3059 (enable|disable) 3060 3061For example, to enable simple_xor for flow type of ipv6 on port 2:: 3062 3063 testpmd> set_hash_global_config 2 simple_xor ipv6 enable 3064 3065set_hash_input_set 3066~~~~~~~~~~~~~~~~~~ 3067 3068Set the input set for hash:: 3069 3070 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 3071 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 3072 l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \ 3073 ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \ 3074 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \ 3075 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \ 3076 fld-8th|none) (select|add) 3077 3078For example, to add source IP to hash input set for flow type of ipv4-udp on port 0:: 3079 3080 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add 3081 3082set_fdir_input_set 3083~~~~~~~~~~~~~~~~~~ 3084 3085The Flow Director filters can match the different fields for different type of packet, i.e. specific input set 3086on per flow type and the flexible payload. This command can be used to change input set for each flow type. 3087 3088Set the input set for flow director:: 3089 3090 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 3091 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 3092 l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \ 3093 ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \ 3094 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \ 3095 sctp-dst-port|sctp-veri-tag|none) (select|add) 3096 3097For example to add source IP to FD input set for flow type of ipv4-udp on port 0:: 3098 3099 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add 3100 3101global_config 3102~~~~~~~~~~~~~ 3103 3104Set different GRE key length for input set:: 3105 3106 global_config (port_id) gre-key-len (number in bytes) 3107 3108For example to set GRE key length for input set to 4 bytes on port 0:: 3109 3110 testpmd> global_config 0 gre-key-len 4 3111 3112 3113.. _testpmd_rte_flow: 3114 3115Flow rules management 3116--------------------- 3117 3118Control of the generic flow API (*rte_flow*) is fully exposed through the 3119``flow`` command (validation, creation, destruction, queries and operation 3120modes). 3121 3122Considering *rte_flow* overlaps with all `Filter Functions`_, using both 3123features simultaneously may cause undefined side-effects and is therefore 3124not recommended. 3125 3126``flow`` syntax 3127~~~~~~~~~~~~~~~ 3128 3129Because the ``flow`` command uses dynamic tokens to handle the large number 3130of possible flow rules combinations, its behavior differs slightly from 3131other commands, in particular: 3132 3133- Pressing *?* or the *<tab>* key displays contextual help for the current 3134 token, not that of the entire command. 3135 3136- Optional and repeated parameters are supported (provided they are listed 3137 in the contextual help). 3138 3139The first parameter stands for the operation mode. Possible operations and 3140their general syntax are described below. They are covered in detail in the 3141following sections. 3142 3143- Check whether a flow rule can be created:: 3144 3145 flow validate {port_id} 3146 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3147 pattern {item} [/ {item} [...]] / end 3148 actions {action} [/ {action} [...]] / end 3149 3150- Create a flow rule:: 3151 3152 flow create {port_id} 3153 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3154 pattern {item} [/ {item} [...]] / end 3155 actions {action} [/ {action} [...]] / end 3156 3157- Destroy specific flow rules:: 3158 3159 flow destroy {port_id} rule {rule_id} [...] 3160 3161- Destroy all flow rules:: 3162 3163 flow flush {port_id} 3164 3165- Query an existing flow rule:: 3166 3167 flow query {port_id} {rule_id} {action} 3168 3169- List existing flow rules sorted by priority, filtered by group 3170 identifiers:: 3171 3172 flow list {port_id} [group {group_id}] [...] 3173 3174- Restrict ingress traffic to the defined flow rules:: 3175 3176 flow isolate {port_id} {boolean} 3177 3178Validating flow rules 3179~~~~~~~~~~~~~~~~~~~~~ 3180 3181``flow validate`` reports whether a flow rule would be accepted by the 3182underlying device in its current state but stops short of creating it. It is 3183bound to ``rte_flow_validate()``:: 3184 3185 flow validate {port_id} 3186 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3187 pattern {item} [/ {item} [...]] / end 3188 actions {action} [/ {action} [...]] / end 3189 3190If successful, it will show:: 3191 3192 Flow rule validated 3193 3194Otherwise it will show an error message of the form:: 3195 3196 Caught error type [...] ([...]): [...] 3197 3198This command uses the same parameters as ``flow create``, their format is 3199described in `Creating flow rules`_. 3200 3201Check whether redirecting any Ethernet packet received on port 0 to RX queue 3202index 6 is supported:: 3203 3204 testpmd> flow validate 0 ingress pattern eth / end 3205 actions queue index 6 / end 3206 Flow rule validated 3207 testpmd> 3208 3209Port 0 does not support TCPv6 rules:: 3210 3211 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end 3212 actions drop / end 3213 Caught error type 9 (specific pattern item): Invalid argument 3214 testpmd> 3215 3216Creating flow rules 3217~~~~~~~~~~~~~~~~~~~ 3218 3219``flow create`` validates and creates the specified flow rule. It is bound 3220to ``rte_flow_create()``:: 3221 3222 flow create {port_id} 3223 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3224 pattern {item} [/ {item} [...]] / end 3225 actions {action} [/ {action} [...]] / end 3226 3227If successful, it will return a flow rule ID usable with other commands:: 3228 3229 Flow rule #[...] created 3230 3231Otherwise it will show an error message of the form:: 3232 3233 Caught error type [...] ([...]): [...] 3234 3235Parameters describe in the following order: 3236 3237- Attributes (*group*, *priority*, *ingress*, *egress*, *transfer* tokens). 3238- A matching pattern, starting with the *pattern* token and terminated by an 3239 *end* pattern item. 3240- Actions, starting with the *actions* token and terminated by an *end* 3241 action. 3242 3243These translate directly to *rte_flow* objects provided as-is to the 3244underlying functions. 3245 3246The shortest valid definition only comprises mandatory tokens:: 3247 3248 testpmd> flow create 0 pattern end actions end 3249 3250Note that PMDs may refuse rules that essentially do nothing such as this 3251one. 3252 3253**All unspecified object values are automatically initialized to 0.** 3254 3255Attributes 3256^^^^^^^^^^ 3257 3258These tokens affect flow rule attributes (``struct rte_flow_attr``) and are 3259specified before the ``pattern`` token. 3260 3261- ``group {group id}``: priority group. 3262- ``priority {level}``: priority level within group. 3263- ``ingress``: rule applies to ingress traffic. 3264- ``egress``: rule applies to egress traffic. 3265- ``transfer``: apply rule directly to endpoints found in pattern. 3266 3267Each instance of an attribute specified several times overrides the previous 3268value as shown below (group 4 is used):: 3269 3270 testpmd> flow create 0 group 42 group 24 group 4 [...] 3271 3272Note that once enabled, ``ingress`` and ``egress`` cannot be disabled. 3273 3274While not specifying a direction is an error, some rules may allow both 3275simultaneously. 3276 3277Most rules affect RX therefore contain the ``ingress`` token:: 3278 3279 testpmd> flow create 0 ingress pattern [...] 3280 3281Matching pattern 3282^^^^^^^^^^^^^^^^ 3283 3284A matching pattern starts after the ``pattern`` token. It is made of pattern 3285items and is terminated by a mandatory ``end`` item. 3286 3287Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum 3288rte_flow_item_type``). 3289 3290The ``/`` token is used as a separator between pattern items as shown 3291below:: 3292 3293 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...] 3294 3295Note that protocol items like these must be stacked from lowest to highest 3296layer to make sense. For instance, the following rule is either invalid or 3297unlikely to match any packet:: 3298 3299 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...] 3300 3301More information on these restrictions can be found in the *rte_flow* 3302documentation. 3303 3304Several items support additional specification structures, for example 3305``ipv4`` allows specifying source and destination addresses as follows:: 3306 3307 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1 3308 dst is 10.2.0.0 / end [...] 3309 3310This rule matches all IPv4 traffic with the specified properties. 3311 3312In this example, ``src`` and ``dst`` are field names of the underlying 3313``struct rte_flow_item_ipv4`` object. All item properties can be specified 3314in a similar fashion. 3315 3316The ``is`` token means that the subsequent value must be matched exactly, 3317and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item`` 3318accordingly. Possible assignment tokens are: 3319 3320- ``is``: match value perfectly (with full bit-mask). 3321- ``spec``: match value according to configured bit-mask. 3322- ``last``: specify upper bound to establish a range. 3323- ``mask``: specify bit-mask with relevant bits set to one. 3324- ``prefix``: generate bit-mask from a prefix length. 3325 3326These yield identical results:: 3327 3328 ipv4 src is 10.1.1.1 3329 3330:: 3331 3332 ipv4 src spec 10.1.1.1 src mask 255.255.255.255 3333 3334:: 3335 3336 ipv4 src spec 10.1.1.1 src prefix 32 3337 3338:: 3339 3340 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value 3341 3342:: 3343 3344 ipv4 src is 10.1.1.1 src last 0 # 0 disables range 3345 3346Inclusive ranges can be defined with ``last``:: 3347 3348 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4 3349 3350Note that ``mask`` affects both ``spec`` and ``last``:: 3351 3352 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0 3353 # matches 10.1.0.0 to 10.2.255.255 3354 3355Properties can be modified multiple times:: 3356 3357 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4 3358 3359:: 3360 3361 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16 3362 3363Pattern items 3364^^^^^^^^^^^^^ 3365 3366This section lists supported pattern items and their attributes, if any. 3367 3368- ``end``: end list of pattern items. 3369 3370- ``void``: no-op pattern item. 3371 3372- ``invert``: perform actions when pattern does not match. 3373 3374- ``any``: match any protocol for the current layer. 3375 3376 - ``num {unsigned}``: number of layers covered. 3377 3378- ``pf``: match traffic from/to the physical function. 3379 3380- ``vf``: match traffic from/to a virtual function ID. 3381 3382 - ``id {unsigned}``: VF ID. 3383 3384- ``phy_port``: match traffic from/to a specific physical port. 3385 3386 - ``index {unsigned}``: physical port index. 3387 3388- ``port_id``: match traffic from/to a given DPDK port ID. 3389 3390 - ``id {unsigned}``: DPDK port ID. 3391 3392- ``mark``: match value set in previously matched flow rule using the mark action. 3393 3394 - ``id {unsigned}``: arbitrary integer value. 3395 3396- ``raw``: match an arbitrary byte string. 3397 3398 - ``relative {boolean}``: look for pattern after the previous item. 3399 - ``search {boolean}``: search pattern from offset (see also limit). 3400 - ``offset {integer}``: absolute or relative offset for pattern. 3401 - ``limit {unsigned}``: search area limit for start of pattern. 3402 - ``pattern {string}``: byte string to look for. 3403 3404- ``eth``: match Ethernet header. 3405 3406 - ``dst {MAC-48}``: destination MAC. 3407 - ``src {MAC-48}``: source MAC. 3408 - ``type {unsigned}``: EtherType or TPID. 3409 3410- ``vlan``: match 802.1Q/ad VLAN tag. 3411 3412 - ``tci {unsigned}``: tag control information. 3413 - ``pcp {unsigned}``: priority code point. 3414 - ``dei {unsigned}``: drop eligible indicator. 3415 - ``vid {unsigned}``: VLAN identifier. 3416 - ``inner_type {unsigned}``: inner EtherType or TPID. 3417 3418- ``ipv4``: match IPv4 header. 3419 3420 - ``tos {unsigned}``: type of service. 3421 - ``ttl {unsigned}``: time to live. 3422 - ``proto {unsigned}``: next protocol ID. 3423 - ``src {ipv4 address}``: source address. 3424 - ``dst {ipv4 address}``: destination address. 3425 3426- ``ipv6``: match IPv6 header. 3427 3428 - ``tc {unsigned}``: traffic class. 3429 - ``flow {unsigned}``: flow label. 3430 - ``proto {unsigned}``: protocol (next header). 3431 - ``hop {unsigned}``: hop limit. 3432 - ``src {ipv6 address}``: source address. 3433 - ``dst {ipv6 address}``: destination address. 3434 3435- ``icmp``: match ICMP header. 3436 3437 - ``type {unsigned}``: ICMP packet type. 3438 - ``code {unsigned}``: ICMP packet code. 3439 3440- ``udp``: match UDP header. 3441 3442 - ``src {unsigned}``: UDP source port. 3443 - ``dst {unsigned}``: UDP destination port. 3444 3445- ``tcp``: match TCP header. 3446 3447 - ``src {unsigned}``: TCP source port. 3448 - ``dst {unsigned}``: TCP destination port. 3449 3450- ``sctp``: match SCTP header. 3451 3452 - ``src {unsigned}``: SCTP source port. 3453 - ``dst {unsigned}``: SCTP destination port. 3454 - ``tag {unsigned}``: validation tag. 3455 - ``cksum {unsigned}``: checksum. 3456 3457- ``vxlan``: match VXLAN header. 3458 3459 - ``vni {unsigned}``: VXLAN identifier. 3460 3461- ``e_tag``: match IEEE 802.1BR E-Tag header. 3462 3463 - ``grp_ecid_b {unsigned}``: GRP and E-CID base. 3464 3465- ``nvgre``: match NVGRE header. 3466 3467 - ``tni {unsigned}``: virtual subnet ID. 3468 3469- ``mpls``: match MPLS header. 3470 3471 - ``label {unsigned}``: MPLS label. 3472 3473- ``gre``: match GRE header. 3474 3475 - ``protocol {unsigned}``: protocol type. 3476 3477- ``fuzzy``: fuzzy pattern match, expect faster than default. 3478 3479 - ``thresh {unsigned}``: accuracy threshold. 3480 3481- ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header. 3482 3483 - ``teid {unsigned}``: tunnel endpoint identifier. 3484 3485- ``geneve``: match GENEVE header. 3486 3487 - ``vni {unsigned}``: virtual network identifier. 3488 - ``protocol {unsigned}``: protocol type. 3489 3490- ``vxlan-gpe``: match VXLAN-GPE header. 3491 3492 - ``vni {unsigned}``: VXLAN-GPE identifier. 3493 3494- ``arp_eth_ipv4``: match ARP header for Ethernet/IPv4. 3495 3496 - ``sha {MAC-48}``: sender hardware address. 3497 - ``spa {ipv4 address}``: sender IPv4 address. 3498 - ``tha {MAC-48}``: target hardware address. 3499 - ``tpa {ipv4 address}``: target IPv4 address. 3500 3501- ``ipv6_ext``: match presence of any IPv6 extension header. 3502 3503 - ``next_hdr {unsigned}``: next header. 3504 3505- ``icmp6``: match any ICMPv6 header. 3506 3507 - ``type {unsigned}``: ICMPv6 type. 3508 - ``code {unsigned}``: ICMPv6 code. 3509 3510- ``icmp6_nd_ns``: match ICMPv6 neighbor discovery solicitation. 3511 3512 - ``target_addr {ipv6 address}``: target address. 3513 3514- ``icmp6_nd_na``: match ICMPv6 neighbor discovery advertisement. 3515 3516 - ``target_addr {ipv6 address}``: target address. 3517 3518- ``icmp6_nd_opt``: match presence of any ICMPv6 neighbor discovery option. 3519 3520 - ``type {unsigned}``: ND option type. 3521 3522- ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery source Ethernet 3523 link-layer address option. 3524 3525 - ``sla {MAC-48}``: source Ethernet LLA. 3526 3527- ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery target Ethernet 3528 link-layer address option. 3529 3530 - ``tla {MAC-48}``: target Ethernet LLA. 3531 3532Actions list 3533^^^^^^^^^^^^ 3534 3535A list of actions starts after the ``actions`` token in the same fashion as 3536`Matching pattern`_; actions are separated by ``/`` tokens and the list is 3537terminated by a mandatory ``end`` action. 3538 3539Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum 3540rte_flow_action_type``). 3541 3542Dropping all incoming UDPv4 packets can be expressed as follows:: 3543 3544 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3545 actions drop / end 3546 3547Several actions have configurable properties which must be specified when 3548there is no valid default value. For example, ``queue`` requires a target 3549queue index. 3550 3551This rule redirects incoming UDPv4 traffic to queue index 6:: 3552 3553 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3554 actions queue index 6 / end 3555 3556While this one could be rejected by PMDs (unspecified queue index):: 3557 3558 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3559 actions queue / end 3560 3561As defined by *rte_flow*, the list is not ordered, all actions of a given 3562rule are performed simultaneously. These are equivalent:: 3563 3564 queue index 6 / void / mark id 42 / end 3565 3566:: 3567 3568 void / mark id 42 / queue index 6 / end 3569 3570All actions in a list should have different types, otherwise only the last 3571action of a given type is taken into account:: 3572 3573 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6 3574 3575:: 3576 3577 drop / drop / drop / end # drop is performed only once 3578 3579:: 3580 3581 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24 3582 3583Considering they are performed simultaneously, opposite and overlapping 3584actions can sometimes be combined when the end result is unambiguous:: 3585 3586 drop / queue index 6 / end # drop has no effect 3587 3588:: 3589 3590 queue index 6 / rss queues 6 7 8 / end # queue has no effect 3591 3592:: 3593 3594 drop / passthru / end # drop has no effect 3595 3596Note that PMDs may still refuse such combinations. 3597 3598Actions 3599^^^^^^^ 3600 3601This section lists supported actions and their attributes, if any. 3602 3603- ``end``: end list of actions. 3604 3605- ``void``: no-op action. 3606 3607- ``passthru``: let subsequent rule process matched packets. 3608 3609- ``jump``: redirect traffic to group on device. 3610 3611 - ``group {unsigned}``: group to redirect to. 3612 3613- ``mark``: attach 32 bit value to packets. 3614 3615 - ``id {unsigned}``: 32 bit value to return with packets. 3616 3617- ``flag``: flag packets. 3618 3619- ``queue``: assign packets to a given queue index. 3620 3621 - ``index {unsigned}``: queue index to use. 3622 3623- ``drop``: drop packets (note: passthru has priority). 3624 3625- ``count``: enable counters for this rule. 3626 3627- ``rss``: spread packets among several queues. 3628 3629 - ``func {hash function}``: RSS hash function to apply, allowed tokens are 3630 the same as `set_hash_global_config`_. 3631 3632 - ``level {unsigned}``: encapsulation level for ``types``. 3633 3634 - ``types [{RSS hash type} [...]] end``: specific RSS hash types, allowed 3635 tokens are the same as `set_hash_input_set`_, except that an empty list 3636 does not disable RSS but instead requests unspecified "best-effort" 3637 settings. 3638 3639 - ``key {string}``: RSS hash key, overrides ``key_len``. 3640 3641 - ``key_len {unsigned}``: RSS hash key length in bytes, can be used in 3642 conjunction with ``key`` to pad or truncate it. 3643 3644 - ``queues [{unsigned} [...]] end``: queue indices to use. 3645 3646- ``pf``: direct traffic to physical function. 3647 3648- ``vf``: direct traffic to a virtual function ID. 3649 3650 - ``original {boolean}``: use original VF ID if possible. 3651 - ``id {unsigned}``: VF ID. 3652 3653- ``phy_port``: direct packets to physical port index. 3654 3655 - ``original {boolean}``: use original port index if possible. 3656 - ``index {unsigned}``: physical port index. 3657 3658- ``port_id``: direct matching traffic to a given DPDK port ID. 3659 3660 - ``original {boolean}``: use original DPDK port ID if possible. 3661 - ``id {unsigned}``: DPDK port ID. 3662 3663- ``of_set_mpls_ttl``: OpenFlow's ``OFPAT_SET_MPLS_TTL``. 3664 3665 - ``mpls_ttl``: MPLS TTL. 3666 3667- ``of_dec_mpls_ttl``: OpenFlow's ``OFPAT_DEC_MPLS_TTL``. 3668 3669- ``of_set_nw_ttl``: OpenFlow's ``OFPAT_SET_NW_TTL``. 3670 3671 - ``nw_ttl``: IP TTL. 3672 3673- ``of_dec_nw_ttl``: OpenFlow's ``OFPAT_DEC_NW_TTL``. 3674 3675- ``of_copy_ttl_out``: OpenFlow's ``OFPAT_COPY_TTL_OUT``. 3676 3677- ``of_copy_ttl_in``: OpenFlow's ``OFPAT_COPY_TTL_IN``. 3678 3679- ``of_pop_vlan``: OpenFlow's ``OFPAT_POP_VLAN``. 3680 3681- ``of_push_vlan``: OpenFlow's ``OFPAT_PUSH_VLAN``. 3682 3683 - ``ethertype``: Ethertype. 3684 3685- ``of_set_vlan_vid``: OpenFlow's ``OFPAT_SET_VLAN_VID``. 3686 3687 - ``vlan_vid``: VLAN id. 3688 3689- ``of_set_vlan_pcp``: OpenFlow's ``OFPAT_SET_VLAN_PCP``. 3690 3691 - ``vlan_pcp``: VLAN priority. 3692 3693- ``of_pop_mpls``: OpenFlow's ``OFPAT_POP_MPLS``. 3694 3695 - ``ethertype``: Ethertype. 3696 3697- ``of_push_mpls``: OpenFlow's ``OFPAT_PUSH_MPLS``. 3698 3699 - ``ethertype``: Ethertype. 3700 3701- ``vxlan_encap``: Performs a VXLAN encapsulation, outer layer configuration 3702 is done through `Config VXLAN Encap outer layers`_. 3703 3704- ``vxlan_decap``: Performs a decapsulation action by stripping all headers of 3705 the VXLAN tunnel network overlay from the matched flow. 3706 3707- ``nvgre_encap``: Performs a NVGRE encapsulation, outer layer configuration 3708 is done through `Config NVGRE Encap outer layers`_. 3709 3710- ``nvgre_decap``: Performs a decapsulation action by stripping all headers of 3711 the NVGRE tunnel network overlay from the matched flow. 3712 3713- ``set_ipv4_src``: Set a new IPv4 source address in the outermost IPv4 header. 3714 3715 - ``ipv4_addr``: New IPv4 source address. 3716 3717- ``set_ipv4_dst``: Set a new IPv4 destination address in the outermost IPv4 3718 header. 3719 3720 - ``ipv4_addr``: New IPv4 destination address. 3721 3722- ``set_ipv6_src``: Set a new IPv6 source address in the outermost IPv6 header. 3723 3724 - ``ipv6_addr``: New IPv6 source address. 3725 3726- ``set_ipv6_dst``: Set a new IPv6 destination address in the outermost IPv6 3727 header. 3728 3729 - ``ipv6_addr``: New IPv6 destination address. 3730 3731- ``of_set_tp_src``: Set a new source port number in the outermost TCP/UDP 3732 header. 3733 3734 - ``port``: New TCP/UDP source port number. 3735 3736- ``of_set_tp_dst``: Set a new destination port number in the outermost TCP/UDP 3737 header. 3738 3739 - ``port``: New TCP/UDP destination port number. 3740 3741- ``mac_swap``: Swap the source and destination MAC addresses in the outermost 3742 Ethernet header. 3743 3744Destroying flow rules 3745~~~~~~~~~~~~~~~~~~~~~ 3746 3747``flow destroy`` destroys one or more rules from their rule ID (as returned 3748by ``flow create``), this command calls ``rte_flow_destroy()`` as many 3749times as necessary:: 3750 3751 flow destroy {port_id} rule {rule_id} [...] 3752 3753If successful, it will show:: 3754 3755 Flow rule #[...] destroyed 3756 3757It does not report anything for rule IDs that do not exist. The usual error 3758message is shown when a rule cannot be destroyed:: 3759 3760 Caught error type [...] ([...]): [...] 3761 3762``flow flush`` destroys all rules on a device and does not take extra 3763arguments. It is bound to ``rte_flow_flush()``:: 3764 3765 flow flush {port_id} 3766 3767Any errors are reported as above. 3768 3769Creating several rules and destroying them:: 3770 3771 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3772 actions queue index 2 / end 3773 Flow rule #0 created 3774 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3775 actions queue index 3 / end 3776 Flow rule #1 created 3777 testpmd> flow destroy 0 rule 0 rule 1 3778 Flow rule #1 destroyed 3779 Flow rule #0 destroyed 3780 testpmd> 3781 3782The same result can be achieved using ``flow flush``:: 3783 3784 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3785 actions queue index 2 / end 3786 Flow rule #0 created 3787 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3788 actions queue index 3 / end 3789 Flow rule #1 created 3790 testpmd> flow flush 0 3791 testpmd> 3792 3793Non-existent rule IDs are ignored:: 3794 3795 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3796 actions queue index 2 / end 3797 Flow rule #0 created 3798 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3799 actions queue index 3 / end 3800 Flow rule #1 created 3801 testpmd> flow destroy 0 rule 42 rule 10 rule 2 3802 testpmd> 3803 testpmd> flow destroy 0 rule 0 3804 Flow rule #0 destroyed 3805 testpmd> 3806 3807Querying flow rules 3808~~~~~~~~~~~~~~~~~~~ 3809 3810``flow query`` queries a specific action of a flow rule having that 3811ability. Such actions collect information that can be reported using this 3812command. It is bound to ``rte_flow_query()``:: 3813 3814 flow query {port_id} {rule_id} {action} 3815 3816If successful, it will display either the retrieved data for known actions 3817or the following message:: 3818 3819 Cannot display result for action type [...] ([...]) 3820 3821Otherwise, it will complain either that the rule does not exist or that some 3822error occurred:: 3823 3824 Flow rule #[...] not found 3825 3826:: 3827 3828 Caught error type [...] ([...]): [...] 3829 3830Currently only the ``count`` action is supported. This action reports the 3831number of packets that hit the flow rule and the total number of bytes. Its 3832output has the following format:: 3833 3834 count: 3835 hits_set: [...] # whether "hits" contains a valid value 3836 bytes_set: [...] # whether "bytes" contains a valid value 3837 hits: [...] # number of packets 3838 bytes: [...] # number of bytes 3839 3840Querying counters for TCPv6 packets redirected to queue 6:: 3841 3842 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end 3843 actions queue index 6 / count / end 3844 Flow rule #4 created 3845 testpmd> flow query 0 4 count 3846 count: 3847 hits_set: 1 3848 bytes_set: 0 3849 hits: 386446 3850 bytes: 0 3851 testpmd> 3852 3853Listing flow rules 3854~~~~~~~~~~~~~~~~~~ 3855 3856``flow list`` lists existing flow rules sorted by priority and optionally 3857filtered by group identifiers:: 3858 3859 flow list {port_id} [group {group_id}] [...] 3860 3861This command only fails with the following message if the device does not 3862exist:: 3863 3864 Invalid port [...] 3865 3866Output consists of a header line followed by a short description of each 3867flow rule, one per line. There is no output at all when no flow rules are 3868configured on the device:: 3869 3870 ID Group Prio Attr Rule 3871 [...] [...] [...] [...] [...] 3872 3873``Attr`` column flags: 3874 3875- ``i`` for ``ingress``. 3876- ``e`` for ``egress``. 3877 3878Creating several flow rules and listing them:: 3879 3880 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3881 actions queue index 6 / end 3882 Flow rule #0 created 3883 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3884 actions queue index 2 / end 3885 Flow rule #1 created 3886 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end 3887 actions rss queues 6 7 8 end / end 3888 Flow rule #2 created 3889 testpmd> flow list 0 3890 ID Group Prio Attr Rule 3891 0 0 0 i- ETH IPV4 => QUEUE 3892 1 0 0 i- ETH IPV6 => QUEUE 3893 2 0 5 i- ETH IPV4 UDP => RSS 3894 testpmd> 3895 3896Rules are sorted by priority (i.e. group ID first, then priority level):: 3897 3898 testpmd> flow list 1 3899 ID Group Prio Attr Rule 3900 0 0 0 i- ETH => COUNT 3901 6 0 500 i- ETH IPV6 TCP => DROP COUNT 3902 5 0 1000 i- ETH IPV6 ICMP => QUEUE 3903 1 24 0 i- ETH IPV4 UDP => QUEUE 3904 4 24 10 i- ETH IPV4 TCP => DROP 3905 3 24 20 i- ETH IPV4 => DROP 3906 2 24 42 i- ETH IPV4 UDP => QUEUE 3907 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 3908 testpmd> 3909 3910Output can be limited to specific groups:: 3911 3912 testpmd> flow list 1 group 0 group 63 3913 ID Group Prio Attr Rule 3914 0 0 0 i- ETH => COUNT 3915 6 0 500 i- ETH IPV6 TCP => DROP COUNT 3916 5 0 1000 i- ETH IPV6 ICMP => QUEUE 3917 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 3918 testpmd> 3919 3920Toggling isolated mode 3921~~~~~~~~~~~~~~~~~~~~~~ 3922 3923``flow isolate`` can be used to tell the underlying PMD that ingress traffic 3924must only be injected from the defined flow rules; that no default traffic 3925is expected outside those rules and the driver is free to assign more 3926resources to handle them. It is bound to ``rte_flow_isolate()``:: 3927 3928 flow isolate {port_id} {boolean} 3929 3930If successful, enabling or disabling isolated mode shows either:: 3931 3932 Ingress traffic on port [...] 3933 is now restricted to the defined flow rules 3934 3935Or:: 3936 3937 Ingress traffic on port [...] 3938 is not restricted anymore to the defined flow rules 3939 3940Otherwise, in case of error:: 3941 3942 Caught error type [...] ([...]): [...] 3943 3944Mainly due to its side effects, PMDs supporting this mode may not have the 3945ability to toggle it more than once without reinitializing affected ports 3946first (e.g. by exiting testpmd). 3947 3948Enabling isolated mode:: 3949 3950 testpmd> flow isolate 0 true 3951 Ingress traffic on port 0 is now restricted to the defined flow rules 3952 testpmd> 3953 3954Disabling isolated mode:: 3955 3956 testpmd> flow isolate 0 false 3957 Ingress traffic on port 0 is not restricted anymore to the defined flow rules 3958 testpmd> 3959 3960Sample QinQ flow rules 3961~~~~~~~~~~~~~~~~~~~~~~ 3962 3963Before creating QinQ rule(s) the following commands should be issued to enable QinQ:: 3964 3965 testpmd> port stop 0 3966 testpmd> vlan set qinq on 0 3967 3968The above command sets the inner and outer TPID's to 0x8100. 3969 3970To change the TPID's the following commands should be used:: 3971 3972 testpmd> vlan set outer tpid 0xa100 0 3973 testpmd> vlan set inner tpid 0x9100 0 3974 testpmd> port start 0 3975 3976Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM. 3977 3978:: 3979 3980 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 / 3981 vlan tci is 456 / end actions vf id 1 / queue index 0 / end 3982 Flow rule #0 validated 3983 3984 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 / 3985 vlan tci is 456 / end actions vf id 123 / queue index 0 / end 3986 Flow rule #0 created 3987 3988 testpmd> flow list 0 3989 ID Group Prio Attr Rule 3990 0 0 0 i- ETH VLAN VLAN=>VF QUEUE 3991 3992Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host. 3993 3994:: 3995 3996 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 / 3997 vlan tci is 654 / end actions pf / queue index 0 / end 3998 Flow rule #1 validated 3999 4000 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 / 4001 vlan tci is 654 / end actions pf / queue index 1 / end 4002 Flow rule #1 created 4003 4004 testpmd> flow list 0 4005 ID Group Prio Attr Rule 4006 0 0 0 i- ETH VLAN VLAN=>VF QUEUE 4007 1 0 0 i- ETH VLAN VLAN=>PF QUEUE 4008 4009Sample VXLAN encapsulation rule 4010~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4011 4012VXLAN encapsulation outer layer has default value pre-configured in testpmd 4013source code, those can be changed by using the following commands 4014 4015IPv4 VXLAN outer header:: 4016 4017 testpmd> set vxlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src 127.0.0.1 4018 ip-dst 128.0.0.1 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4019 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 4020 queue index 0 / end 4021 4022 testpmd> set vxlan-with-vlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src 4023 127.0.0.1 ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11 4024 eth-dst 22:22:22:22:22:22 4025 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 4026 queue index 0 / end 4027 4028IPv6 VXLAN outer header:: 4029 4030 testpmd> set vxlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4 ip-src ::1 4031 ip-dst ::2222 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4032 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 4033 queue index 0 / end 4034 4035 testpmd> set vxlan-with-vlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4 4036 ip-src ::1 ip-dst ::2222 vlan-tci 34 eth-src 11:11:11:11:11:11 4037 eth-dst 22:22:22:22:22:22 4038 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 4039 queue index 0 / end 4040 4041Sample NVGRE encapsulation rule 4042~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4043 4044NVGRE encapsulation outer layer has default value pre-configured in testpmd 4045source code, those can be changed by using the following commands 4046 4047IPv4 NVGRE outer header:: 4048 4049 testpmd> set nvgre ip-version ipv4 tni 4 ip-src 127.0.0.1 ip-dst 128.0.0.1 4050 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4051 testpmd> flow create 0 ingress pattern end actions nvgre_encap / 4052 queue index 0 / end 4053 4054 testpmd> set nvgre-with-vlan ip-version ipv4 tni 4 ip-src 127.0.0.1 4055 ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11 4056 eth-dst 22:22:22:22:22:22 4057 testpmd> flow create 0 ingress pattern end actions nvgre_encap / 4058 queue index 0 / end 4059 4060IPv6 NVGRE outer header:: 4061 4062 testpmd> set nvgre ip-version ipv6 tni 4 ip-src ::1 ip-dst ::2222 4063 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4064 testpmd> flow create 0 ingress pattern end actions nvgre_encap / 4065 queue index 0 / end 4066 4067 testpmd> set nvgre-with-vlan ip-version ipv6 tni 4 ip-src ::1 ip-dst ::2222 4068 vlan-tci 34 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4069 testpmd> flow create 0 ingress pattern end actions nvgre_encap / 4070 queue index 0 / end 4071 4072BPF Functions 4073-------------- 4074 4075The following sections show functions to load/unload eBPF based filters. 4076 4077bpf-load 4078~~~~~~~~ 4079 4080Load an eBPF program as a callback for partciular RX/TX queue:: 4081 4082 testpmd> bpf-load rx|tx (portid) (queueid) (load-flags) (bpf-prog-filename) 4083 4084The available load-flags are: 4085 4086* ``J``: use JIT generated native code, otherwise BPF interpreter will be used. 4087 4088* ``M``: assume input parameter is a pointer to rte_mbuf, otherwise assume it is a pointer to first segment's data. 4089 4090* ``-``: none. 4091 4092.. note:: 4093 4094 You'll need clang v3.7 or above to build bpf program you'd like to load 4095 4096For example: 4097 4098.. code-block:: console 4099 4100 cd test/bpf 4101 clang -O2 -target bpf -c t1.c 4102 4103Then to load (and JIT compile) t1.o at RX queue 0, port 1:: 4104 4105.. code-block:: console 4106 4107 testpmd> bpf-load rx 1 0 J ./dpdk.org/test/bpf/t1.o 4108 4109To load (not JITed) t1.o at TX queue 0, port 0:: 4110 4111.. code-block:: console 4112 4113 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o 4114 4115bpf-unload 4116~~~~~~~~~~ 4117 4118Unload previously loaded eBPF program for partciular RX/TX queue:: 4119 4120 testpmd> bpf-unload rx|tx (portid) (queueid) 4121 4122For example to unload BPF filter from TX queue 0, port 0: 4123 4124.. code-block:: console 4125 4126 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o 4127