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|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* ``stats``: RX/TX statistics. 169 170* ``xstats``: RX/TX extended NIC statistics. 171 172* ``fdir``: Flow Director information and statistics. 173 174* ``stat_qmap``: Queue statistics mapping. 175 176* ``dcb_tc``: DCB information such as TC mapping. 177 178* ``cap``: Supported offload capabilities. 179 180For example: 181 182.. code-block:: console 183 184 testpmd> show port info 0 185 186 ********************* Infos for port 0 ********************* 187 188 MAC address: XX:XX:XX:XX:XX:XX 189 Connect to socket: 0 190 memory allocation on the socket: 0 191 Link status: up 192 Link speed: 40000 Mbps 193 Link duplex: full-duplex 194 Promiscuous mode: enabled 195 Allmulticast mode: disabled 196 Maximum number of MAC addresses: 64 197 Maximum number of MAC addresses of hash filtering: 0 198 VLAN offload: 199 strip on 200 filter on 201 qinq(extend) off 202 Redirection table size: 512 203 Supported flow types: 204 ipv4-frag 205 ipv4-tcp 206 ipv4-udp 207 ipv4-sctp 208 ipv4-other 209 ipv6-frag 210 ipv6-tcp 211 ipv6-udp 212 ipv6-sctp 213 ipv6-other 214 l2_payload 215 port 216 vxlan 217 geneve 218 nvgre 219 220show port rss reta 221~~~~~~~~~~~~~~~~~~ 222 223Display the rss redirection table entry indicated by masks on port X:: 224 225 testpmd> show port (port_id) rss reta (size) (mask0, mask1...) 226 227size is used to indicate the hardware supported reta size 228 229show port rss-hash 230~~~~~~~~~~~~~~~~~~ 231 232Display the RSS hash functions and RSS hash key of a port:: 233 234 testpmd> show port (port_id) rss-hash ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2-payload|ipv6-ex|ipv6-tcp-ex|ipv6-udp-ex [key] 235 236clear port 237~~~~~~~~~~ 238 239Clear the port statistics for a given port or for all ports:: 240 241 testpmd> clear port (info|stats|xstats|fdir|stat_qmap) (port_id|all) 242 243For example:: 244 245 testpmd> clear port stats all 246 247show (rxq|txq) 248~~~~~~~~~~~~~~ 249 250Display information for a given port's RX/TX queue:: 251 252 testpmd> show (rxq|txq) info (port_id) (queue_id) 253 254show config 255~~~~~~~~~~~ 256 257Displays the configuration of the application. 258The configuration comes from the command-line, the runtime or the application defaults:: 259 260 testpmd> show config (rxtx|cores|fwd|txpkts) 261 262The available information categories are: 263 264* ``rxtx``: RX/TX configuration items. 265 266* ``cores``: List of forwarding cores. 267 268* ``fwd``: Packet forwarding configuration. 269 270* ``txpkts``: Packets to TX configuration. 271 272For example: 273 274.. code-block:: console 275 276 testpmd> show config rxtx 277 278 io packet forwarding - CRC stripping disabled - packets/burst=16 279 nb forwarding cores=2 - nb forwarding ports=1 280 RX queues=1 - RX desc=128 - RX free threshold=0 281 RX threshold registers: pthresh=8 hthresh=8 wthresh=4 282 TX queues=1 - TX desc=512 - TX free threshold=0 283 TX threshold registers: pthresh=36 hthresh=0 wthresh=0 284 TX RS bit threshold=0 - TXQ flags=0x0 285 286set fwd 287~~~~~~~ 288 289Set the packet forwarding mode:: 290 291 testpmd> set fwd (io|mac|macswap|flowgen| \ 292 rxonly|txonly|csum|icmpecho) (""|retry) 293 294``retry`` can be specified for forwarding engines except ``rx_only``. 295 296The available information categories are: 297 298* ``io``: Forwards packets "as-is" in I/O mode. 299 This is the fastest possible forwarding operation as it does not access packets data. 300 This is the default mode. 301 302* ``mac``: Changes the source and the destination Ethernet addresses of packets before forwarding them. 303 Default application behaviour is to set source Ethernet address to that of the transmitting interface, and destination 304 address to a dummy value (set during init). The user may specify a target destination Ethernet address via the 'eth-peer' or 305 'eth-peer-configfile' command-line options. It is not currently possible to specify a specific source Ethernet address. 306 307* ``macswap``: MAC swap forwarding mode. 308 Swaps the source and the destination Ethernet addresses of packets before forwarding them. 309 310* ``flowgen``: Multi-flow generation mode. 311 Originates a number of flows (with varying destination IP addresses), and terminate receive traffic. 312 313* ``rxonly``: Receives packets but doesn't transmit them. 314 315* ``txonly``: Generates and transmits packets without receiving any. 316 317* ``csum``: Changes the checksum field with hardware or software methods depending on the offload flags on the packet. 318 319* ``icmpecho``: Receives a burst of packets, lookup for IMCP echo requests and, if any, send back ICMP echo replies. 320 321* ``ieee1588``: Demonstrate L2 IEEE1588 V2 PTP timestamping for RX and TX. Requires ``CONFIG_RTE_LIBRTE_IEEE1588=y``. 322 323* ``tm``: Traffic Management forwarding mode 324 Demonstrates the use of ethdev traffic management APIs and softnic PMD for 325 QoS traffic management. In this mode, 5-level hierarchical QoS scheduler is 326 available as an default option that can be enabled through CLI. The user can 327 also modify the default hierarchy or specify the new hierarchy through CLI for 328 implementing QoS scheduler. Requires ``CONFIG_RTE_LIBRTE_PMD_SOFTNIC=y`` ``CONFIG_RTE_LIBRTE_SCHED=y``. 329 330Example:: 331 332 testpmd> set fwd rxonly 333 334 Set rxonly packet forwarding mode 335 336 337read rxd 338~~~~~~~~ 339 340Display an RX descriptor for a port RX queue:: 341 342 testpmd> read rxd (port_id) (queue_id) (rxd_id) 343 344For example:: 345 346 testpmd> read rxd 0 0 4 347 0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180 348 349read txd 350~~~~~~~~ 351 352Display a TX descriptor for a port TX queue:: 353 354 testpmd> read txd (port_id) (queue_id) (txd_id) 355 356For example:: 357 358 testpmd> read txd 0 0 4 359 0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C 360 361ddp get list 362~~~~~~~~~~~~ 363 364Get loaded dynamic device personalization (DDP) package info list:: 365 366 testpmd> ddp get list (port_id) 367 368ddp get info 369~~~~~~~~~~~~ 370 371Display information about dynamic device personalization (DDP) profile:: 372 373 testpmd> ddp get info (profile_path) 374 375show vf stats 376~~~~~~~~~~~~~ 377 378Display VF statistics:: 379 380 testpmd> show vf stats (port_id) (vf_id) 381 382clear vf stats 383~~~~~~~~~~~~~~ 384 385Reset VF statistics:: 386 387 testpmd> clear vf stats (port_id) (vf_id) 388 389show port pctype mapping 390~~~~~~~~~~~~~~~~~~~~~~~~ 391 392List all items from the pctype mapping table:: 393 394 testpmd> show port (port_id) pctype mapping 395 396show rx offloading capabilities 397~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 398 399List all per queue and per port Rx offloading capabilities of a port:: 400 401 testpmd> show port (port_id) rx_offload capabilities 402 403show rx offloading configuration 404~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 405 406List port level and all queue level Rx offloading configuration:: 407 408 testpmd> show port (port_id) rx_offload configuration 409 410show tx offloading capabilities 411~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 412 413List all per queue and per port Tx offloading capabilities of a port:: 414 415 testpmd> show port (port_id) tx_offload capabilities 416 417show tx offloading configuration 418~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 419 420List port level and all queue level Tx offloading configuration:: 421 422 testpmd> show port (port_id) tx_offload configuration 423 424 425Configuration Functions 426----------------------- 427 428The testpmd application can be configured from the runtime as well as from the command-line. 429 430This section details the available configuration functions that are available. 431 432.. note:: 433 434 Configuration changes only become active when forwarding is started/restarted. 435 436set default 437~~~~~~~~~~~ 438 439Reset forwarding to the default configuration:: 440 441 testpmd> set default 442 443set verbose 444~~~~~~~~~~~ 445 446Set the debug verbosity level:: 447 448 testpmd> set verbose (level) 449 450Currently the only available levels are 0 (silent except for error) and 1 (fully verbose). 451 452set log 453~~~~~~~ 454 455Set the log level for a log type:: 456 457 testpmd> set log global|(type) (level) 458 459Where: 460 461* ``type`` is the log name. 462 463* ``level`` is the log level. 464 465For example, to change the global log level:: 466 testpmd> set log global (level) 467 468Regexes can also be used for type. To change log level of user1, user2 and user3:: 469 testpmd> set log user[1-3] (level) 470 471set nbport 472~~~~~~~~~~ 473 474Set the number of ports used by the application: 475 476set nbport (num) 477 478This is equivalent to the ``--nb-ports`` command-line option. 479 480set nbcore 481~~~~~~~~~~ 482 483Set the number of cores used by the application:: 484 485 testpmd> set nbcore (num) 486 487This is equivalent to the ``--nb-cores`` command-line option. 488 489.. note:: 490 491 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port. 492 493set coremask 494~~~~~~~~~~~~ 495 496Set the forwarding cores hexadecimal mask:: 497 498 testpmd> set coremask (mask) 499 500This is equivalent to the ``--coremask`` command-line option. 501 502.. note:: 503 504 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding. 505 506set portmask 507~~~~~~~~~~~~ 508 509Set the forwarding ports hexadecimal mask:: 510 511 testpmd> set portmask (mask) 512 513This is equivalent to the ``--portmask`` command-line option. 514 515set burst 516~~~~~~~~~ 517 518Set number of packets per burst:: 519 520 testpmd> set burst (num) 521 522This is equivalent to the ``--burst command-line`` option. 523 524When retry is enabled, the transmit delay time and number of retries can also be set:: 525 526 testpmd> set burst tx delay (microseconds) retry (num) 527 528set txpkts 529~~~~~~~~~~ 530 531Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode:: 532 533 testpmd> set txpkts (x[,y]*) 534 535Where x[,y]* represents a CSV list of values, without white space. 536 537set txsplit 538~~~~~~~~~~~ 539 540Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes:: 541 542 testpmd> set txsplit (off|on|rand) 543 544Where: 545 546* ``off`` disable packet copy & split for CSUM mode. 547 548* ``on`` split outgoing packet into multiple segments. Size of each segment 549 and number of segments per packet is determined by ``set txpkts`` command 550 (see above). 551 552* ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments. 553 554set corelist 555~~~~~~~~~~~~ 556 557Set the list of forwarding cores:: 558 559 testpmd> set corelist (x[,y]*) 560 561For example, to change the forwarding cores: 562 563.. code-block:: console 564 565 testpmd> set corelist 3,1 566 testpmd> show config fwd 567 568 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled 569 Logical Core 3 (socket 0) forwards packets on 1 streams: 570 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01 571 Logical Core 1 (socket 0) forwards packets on 1 streams: 572 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00 573 574.. note:: 575 576 The cores are used in the same order as specified on the command line. 577 578set portlist 579~~~~~~~~~~~~ 580 581Set the list of forwarding ports:: 582 583 testpmd> set portlist (x[,y]*) 584 585For example, to change the port forwarding: 586 587.. code-block:: console 588 589 testpmd> set portlist 0,2,1,3 590 testpmd> show config fwd 591 592 io packet forwarding - ports=4 - cores=1 - streams=4 593 Logical Core 3 (socket 0) forwards packets on 4 streams: 594 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01 595 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00 596 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03 597 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02 598 599set tx loopback 600~~~~~~~~~~~~~~~ 601 602Enable/disable tx loopback:: 603 604 testpmd> set tx loopback (port_id) (on|off) 605 606set drop enable 607~~~~~~~~~~~~~~~ 608 609set drop enable bit for all queues:: 610 611 testpmd> set all queues drop (port_id) (on|off) 612 613set split drop enable (for VF) 614~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 615 616set split drop enable bit for VF from PF:: 617 618 testpmd> set vf split drop (port_id) (vf_id) (on|off) 619 620set mac antispoof (for VF) 621~~~~~~~~~~~~~~~~~~~~~~~~~~ 622 623Set mac antispoof for a VF from the PF:: 624 625 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off) 626 627set macsec offload 628~~~~~~~~~~~~~~~~~~ 629 630Enable/disable MACsec offload:: 631 632 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off) 633 testpmd> set macsec offload (port_id) off 634 635set macsec sc 636~~~~~~~~~~~~~ 637 638Configure MACsec secure connection (SC):: 639 640 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi) 641 642.. note:: 643 644 The pi argument is ignored for tx. 645 Check the NIC Datasheet for hardware limits. 646 647set macsec sa 648~~~~~~~~~~~~~ 649 650Configure MACsec secure association (SA):: 651 652 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key) 653 654.. note:: 655 656 The IDX value must be 0 or 1. 657 Check the NIC Datasheet for hardware limits. 658 659set broadcast mode (for VF) 660~~~~~~~~~~~~~~~~~~~~~~~~~~~ 661 662Set broadcast mode for a VF from the PF:: 663 664 testpmd> set vf broadcast (port_id) (vf_id) (on|off) 665 666vlan set strip 667~~~~~~~~~~~~~~ 668 669Set the VLAN strip on a port:: 670 671 testpmd> vlan set strip (on|off) (port_id) 672 673vlan set stripq 674~~~~~~~~~~~~~~~ 675 676Set the VLAN strip for a queue on a port:: 677 678 testpmd> vlan set stripq (on|off) (port_id,queue_id) 679 680vlan set stripq (for VF) 681~~~~~~~~~~~~~~~~~~~~~~~~ 682 683Set VLAN strip for all queues in a pool for a VF from the PF:: 684 685 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off) 686 687vlan set insert (for VF) 688~~~~~~~~~~~~~~~~~~~~~~~~ 689 690Set VLAN insert for a VF from the PF:: 691 692 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id) 693 694vlan set tag (for VF) 695~~~~~~~~~~~~~~~~~~~~~ 696 697Set VLAN tag for a VF from the PF:: 698 699 testpmd> set vf vlan tag (port_id) (vf_id) (on|off) 700 701vlan set antispoof (for VF) 702~~~~~~~~~~~~~~~~~~~~~~~~~~~ 703 704Set VLAN antispoof for a VF from the PF:: 705 706 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off) 707 708vlan set filter 709~~~~~~~~~~~~~~~ 710 711Set the VLAN filter on a port:: 712 713 testpmd> vlan set filter (on|off) (port_id) 714 715vlan set qinq 716~~~~~~~~~~~~~ 717 718Set the VLAN QinQ (extended queue in queue) on for a port:: 719 720 testpmd> vlan set qinq (on|off) (port_id) 721 722vlan set tpid 723~~~~~~~~~~~~~ 724 725Set the inner or outer VLAN TPID for packet filtering on a port:: 726 727 testpmd> vlan set (inner|outer) tpid (value) (port_id) 728 729.. note:: 730 731 TPID value must be a 16-bit number (value <= 65536). 732 733rx_vlan add 734~~~~~~~~~~~ 735 736Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID:: 737 738 testpmd> rx_vlan add (vlan_id|all) (port_id) 739 740.. note:: 741 742 VLAN filter must be set on that port. VLAN ID < 4096. 743 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries 744 in VFTA table. This is important if enabling all vlan_ids. 745 746rx_vlan rm 747~~~~~~~~~~ 748 749Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID:: 750 751 testpmd> rx_vlan rm (vlan_id|all) (port_id) 752 753rx_vlan add (for VF) 754~~~~~~~~~~~~~~~~~~~~ 755 756Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID:: 757 758 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask) 759 760rx_vlan rm (for VF) 761~~~~~~~~~~~~~~~~~~~ 762 763Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID:: 764 765 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask) 766 767tunnel_filter add 768~~~~~~~~~~~~~~~~~ 769 770Add a tunnel filter on a port:: 771 772 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \ 773 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\ 774 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id) 775 776The available information categories are: 777 778* ``vxlan``: Set tunnel type as VXLAN. 779 780* ``nvgre``: Set tunnel type as NVGRE. 781 782* ``ipingre``: Set tunnel type as IP-in-GRE. 783 784* ``imac-ivlan``: Set filter type as Inner MAC and VLAN. 785 786* ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID. 787 788* ``imac-tenid``: Set filter type as Inner MAC and tenant ID. 789 790* ``imac``: Set filter type as Inner MAC. 791 792* ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID. 793 794* ``oip``: Set filter type as Outer IP. 795 796* ``iip``: Set filter type as Inner IP. 797 798Example:: 799 800 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \ 801 192.168.2.2 0 ipingre oip 1 1 802 803 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP. 804 805tunnel_filter remove 806~~~~~~~~~~~~~~~~~~~~ 807 808Remove a tunnel filter on a port:: 809 810 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \ 811 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\ 812 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id) 813 814rx_vxlan_port add 815~~~~~~~~~~~~~~~~~ 816 817Add an UDP port for VXLAN packet filter on a port:: 818 819 testpmd> rx_vxlan_port add (udp_port) (port_id) 820 821rx_vxlan_port remove 822~~~~~~~~~~~~~~~~~~~~ 823 824Remove an UDP port for VXLAN packet filter on a port:: 825 826 testpmd> rx_vxlan_port rm (udp_port) (port_id) 827 828tx_vlan set 829~~~~~~~~~~~ 830 831Set hardware insertion of VLAN IDs in packets sent on a port:: 832 833 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer] 834 835For example, set a single VLAN ID (5) insertion on port 0:: 836 837 tx_vlan set 0 5 838 839Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1:: 840 841 tx_vlan set 1 2 3 842 843 844tx_vlan set pvid 845~~~~~~~~~~~~~~~~ 846 847Set port based hardware insertion of VLAN ID in packets sent on a port:: 848 849 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off) 850 851tx_vlan reset 852~~~~~~~~~~~~~ 853 854Disable hardware insertion of a VLAN header in packets sent on a port:: 855 856 testpmd> tx_vlan reset (port_id) 857 858csum set 859~~~~~~~~ 860 861Select hardware or software calculation of the checksum when 862transmitting a packet using the ``csum`` forwarding engine:: 863 864 testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id) 865 866Where: 867 868* ``ip|udp|tcp|sctp`` always relate to the inner layer. 869 870* ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized 871 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are 872 supported). See also the ``csum parse-tunnel`` command. 873 874.. note:: 875 876 Check the NIC Datasheet for hardware limits. 877 878RSS queue region 879~~~~~~~~~~~~~~~~ 880 881Set RSS queue region span on a port:: 882 883 testpmd> set port (port_id) queue-region region_id (value) \ 884 queue_start_index (value) queue_num (value) 885 886Set flowtype mapping on a RSS queue region on a port:: 887 888 testpmd> set port (port_id) queue-region region_id (value) flowtype (value) 889 890where: 891 892* For the flowtype(pctype) of packet,the specific index for each type has 893 been defined in file i40e_type.h as enum i40e_filter_pctype. 894 895Set user priority mapping on a RSS queue region on a port:: 896 897 testpmd> set port (port_id) queue-region UP (value) region_id (value) 898 899Flush all queue region related configuration on a port:: 900 901 testpmd> set port (port_id) queue-region flush (on|off) 902 903where: 904 905* "on"is just an enable function which server for other configuration, 906 it is for all configuration about queue region from up layer, 907 at first will only keep in DPDK softwarestored in driver, 908 only after "flush on", it commit all configuration to HW. 909 "off" is just clean all configuration about queue region just now, 910 and restore all to DPDK i40e driver default config when start up. 911 912Show all queue region related configuration info on a port:: 913 914 testpmd> show port (port_id) queue-region 915 916.. note:: 917 918 Queue region only support on PF by now, so these command is 919 only for configuration of queue region on PF port. 920 921csum parse-tunnel 922~~~~~~~~~~~~~~~~~ 923 924Define how tunneled packets should be handled by the csum forward 925engine:: 926 927 testpmd> csum parse-tunnel (on|off) (tx_port_id) 928 929If enabled, the csum forward engine will try to recognize supported 930tunnel headers (vxlan, gre, ipip). 931 932If disabled, treat tunnel packets as non-tunneled packets (a inner 933header is handled as a packet payload). 934 935.. note:: 936 937 The port argument is the TX port like in the ``csum set`` command. 938 939Example: 940 941Consider a packet in packet like the following:: 942 943 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in 944 945* If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set`` 946 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the 947 ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``). 948 949* If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set`` 950 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``. 951 952csum show 953~~~~~~~~~ 954 955Display tx checksum offload configuration:: 956 957 testpmd> csum show (port_id) 958 959tso set 960~~~~~~~ 961 962Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine:: 963 964 testpmd> tso set (segsize) (port_id) 965 966.. note:: 967 968 Check the NIC datasheet for hardware limits. 969 970tso show 971~~~~~~~~ 972 973Display the status of TCP Segmentation Offload:: 974 975 testpmd> tso show (port_id) 976 977set port - gro 978~~~~~~~~~~~~~~ 979 980Enable or disable GRO in ``csum`` forwarding engine:: 981 982 testpmd> set port <port_id> gro on|off 983 984If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4 985packets received from the given port. 986 987If disabled, packets received from the given port won't be performed 988GRO. By default, GRO is disabled for all ports. 989 990.. note:: 991 992 When enable GRO for a port, TCP/IPv4 packets received from the port 993 will be performed GRO. After GRO, all merged packets have bad 994 checksums, since the GRO library doesn't re-calculate checksums for 995 the merged packets. Therefore, if users want the merged packets to 996 have correct checksums, please select HW IP checksum calculation and 997 HW TCP checksum calculation for the port which the merged packets are 998 transmitted to. 999 1000show port - gro 1001~~~~~~~~~~~~~~~ 1002 1003Display GRO configuration for a given port:: 1004 1005 testpmd> show port <port_id> gro 1006 1007set gro flush 1008~~~~~~~~~~~~~ 1009 1010Set the cycle to flush the GROed packets from reassembly tables:: 1011 1012 testpmd> set gro flush <cycles> 1013 1014When enable GRO, the csum forwarding engine performs GRO on received 1015packets, and the GROed packets are stored in reassembly tables. Users 1016can use this command to determine when the GROed packets are flushed 1017from the reassembly tables. 1018 1019The ``cycles`` is measured in GRO operation times. The csum forwarding 1020engine flushes the GROed packets from the tables every ``cycles`` GRO 1021operations. 1022 1023By default, the value of ``cycles`` is 1, which means flush GROed packets 1024from the reassembly tables as soon as one GRO operation finishes. The value 1025of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``. 1026 1027Please note that the large value of ``cycles`` may cause the poor TCP/IP 1028stack performance. Because the GROed packets are delayed to arrive the 1029stack, thus causing more duplicated ACKs and TCP retransmissions. 1030 1031set port - gso 1032~~~~~~~~~~~~~~ 1033 1034Toggle per-port GSO support in ``csum`` forwarding engine:: 1035 1036 testpmd> set port <port_id> gso on|off 1037 1038If enabled, the csum forwarding engine will perform GSO on supported IPv4 1039packets, transmitted on the given port. 1040 1041If disabled, packets transmitted on the given port will not undergo GSO. 1042By default, GSO is disabled for all ports. 1043 1044.. note:: 1045 1046 When GSO is enabled on a port, supported IPv4 packets transmitted on that 1047 port undergo GSO. Afterwards, the segmented packets are represented by 1048 multi-segment mbufs; however, the csum forwarding engine doesn't calculation 1049 of checksums for GSO'd segments in SW. As a result, if users want correct 1050 checksums in GSO segments, they should enable HW checksum calculation for 1051 GSO-enabled ports. 1052 1053 For example, HW checksum calculation for VxLAN GSO'd packets may be enabled 1054 by setting the following options in the csum forwarding engine: 1055 1056 testpmd> csum set outer_ip hw <port_id> 1057 1058 testpmd> csum set ip hw <port_id> 1059 1060 testpmd> csum set tcp hw <port_id> 1061 1062 UDP GSO is the same as IP fragmentation, which treats the UDP header 1063 as the payload and does not modify it during segmentation. That is, 1064 after UDP GSO, only the first output fragment has the original UDP 1065 header. Therefore, users need to enable HW IP checksum calculation 1066 and SW UDP checksum calculation for GSO-enabled ports, if they want 1067 correct checksums for UDP/IPv4 packets. 1068 1069set gso segsz 1070~~~~~~~~~~~~~ 1071 1072Set the maximum GSO segment size (measured in bytes), which includes the 1073packet header and the packet payload for GSO-enabled ports (global):: 1074 1075 testpmd> set gso segsz <length> 1076 1077show port - gso 1078~~~~~~~~~~~~~~~ 1079 1080Display the status of Generic Segmentation Offload for a given port:: 1081 1082 testpmd> show port <port_id> gso 1083 1084mac_addr add 1085~~~~~~~~~~~~ 1086 1087Add an alternative MAC address to a port:: 1088 1089 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX) 1090 1091mac_addr remove 1092~~~~~~~~~~~~~~~ 1093 1094Remove a MAC address from a port:: 1095 1096 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX) 1097 1098mac_addr add (for VF) 1099~~~~~~~~~~~~~~~~~~~~~ 1100 1101Add an alternative MAC address for a VF to a port:: 1102 1103 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX) 1104 1105mac_addr set 1106~~~~~~~~~~~~ 1107 1108Set the default MAC address for a port:: 1109 1110 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX) 1111 1112mac_addr set (for VF) 1113~~~~~~~~~~~~~~~~~~~~~ 1114 1115Set the MAC address for a VF from the PF:: 1116 1117 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX) 1118 1119set eth-peer 1120~~~~~~~~~~~~ 1121 1122Set the forwarding peer address for certain port:: 1123 1124 testpmd> set eth-peer (port_id) (perr_addr) 1125 1126This is equivalent to the ``--eth-peer`` command-line option. 1127 1128set port-uta 1129~~~~~~~~~~~~ 1130 1131Set the unicast hash filter(s) on/off for a port:: 1132 1133 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off) 1134 1135set promisc 1136~~~~~~~~~~~ 1137 1138Set the promiscuous mode on for a port or for all ports. 1139In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1140 1141 testpmd> set promisc (port_id|all) (on|off) 1142 1143set allmulti 1144~~~~~~~~~~~~ 1145 1146Set the allmulti mode for a port or for all ports:: 1147 1148 testpmd> set allmulti (port_id|all) (on|off) 1149 1150Same as the ifconfig (8) option. Controls how multicast packets are handled. 1151 1152set promisc (for VF) 1153~~~~~~~~~~~~~~~~~~~~ 1154 1155Set the unicast promiscuous mode for a VF from PF. 1156It's supported by Intel i40e NICs now. 1157In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1158 1159 testpmd> set vf promisc (port_id) (vf_id) (on|off) 1160 1161set allmulticast (for VF) 1162~~~~~~~~~~~~~~~~~~~~~~~~~ 1163 1164Set the multicast promiscuous mode for a VF from PF. 1165It's supported by Intel i40e NICs now. 1166In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1167 1168 testpmd> set vf allmulti (port_id) (vf_id) (on|off) 1169 1170set tx max bandwidth (for VF) 1171~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1172 1173Set TX max absolute bandwidth (Mbps) for a VF from PF:: 1174 1175 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth) 1176 1177set tc tx min bandwidth (for VF) 1178~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1179 1180Set all TCs' TX min relative bandwidth (%) for a VF from PF:: 1181 1182 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...) 1183 1184set tc tx max bandwidth (for VF) 1185~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1186 1187Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF:: 1188 1189 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth) 1190 1191set tc strict link priority mode 1192~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1193 1194Set some TCs' strict link priority mode on a physical port:: 1195 1196 testpmd> set tx strict-link-priority (port_id) (tc_bitmap) 1197 1198set tc tx min bandwidth 1199~~~~~~~~~~~~~~~~~~~~~~~ 1200 1201Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs:: 1202 1203 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...) 1204 1205set flow_ctrl rx 1206~~~~~~~~~~~~~~~~ 1207 1208Set the link flow control parameter on a port:: 1209 1210 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 1211 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \ 1212 autoneg (on|off) (port_id) 1213 1214Where: 1215 1216* ``high_water`` (integer): High threshold value to trigger XOFF. 1217 1218* ``low_water`` (integer): Low threshold value to trigger XON. 1219 1220* ``pause_time`` (integer): Pause quota in the Pause frame. 1221 1222* ``send_xon`` (0/1): Send XON frame. 1223 1224* ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames. 1225 1226* ``autoneg``: Change the auto-negotiation parameter. 1227 1228set pfc_ctrl rx 1229~~~~~~~~~~~~~~~ 1230 1231Set the priority flow control parameter on a port:: 1232 1233 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 1234 (pause_time) (priority) (port_id) 1235 1236Where: 1237 1238* ``high_water`` (integer): High threshold value. 1239 1240* ``low_water`` (integer): Low threshold value. 1241 1242* ``pause_time`` (integer): Pause quota in the Pause frame. 1243 1244* ``priority`` (0-7): VLAN User Priority. 1245 1246set stat_qmap 1247~~~~~~~~~~~~~ 1248 1249Set statistics mapping (qmapping 0..15) for RX/TX queue on port:: 1250 1251 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping) 1252 1253For example, to set rx queue 2 on port 0 to mapping 5:: 1254 1255 testpmd>set stat_qmap rx 0 2 5 1256 1257set xstats-hide-zero 1258~~~~~~~~~~~~~~~~~~~~ 1259 1260Set the option to hide zero values for xstats display:: 1261 1262 testpmd> set xstats-hide-zero on|off 1263 1264.. note:: 1265 1266 By default, the zero values are displayed for xstats. 1267 1268set port - rx/tx (for VF) 1269~~~~~~~~~~~~~~~~~~~~~~~~~ 1270 1271Set VF receive/transmit from a port:: 1272 1273 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off) 1274 1275set port - mac address filter (for VF) 1276~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1277 1278Add/Remove unicast or multicast MAC addr filter for a VF:: 1279 1280 testpmd> set port (port_id) vf (vf_id) (mac_addr) \ 1281 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off) 1282 1283set port - rx mode(for VF) 1284~~~~~~~~~~~~~~~~~~~~~~~~~~ 1285 1286Set the VF receive mode of a port:: 1287 1288 testpmd> set port (port_id) vf (vf_id) \ 1289 rxmode (AUPE|ROPE|BAM|MPE) (on|off) 1290 1291The available receive modes are: 1292 1293* ``AUPE``: Accepts untagged VLAN. 1294 1295* ``ROPE``: Accepts unicast hash. 1296 1297* ``BAM``: Accepts broadcast packets. 1298 1299* ``MPE``: Accepts all multicast packets. 1300 1301set port - tx_rate (for Queue) 1302~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1303 1304Set TX rate limitation for a queue on a port:: 1305 1306 testpmd> set port (port_id) queue (queue_id) rate (rate_value) 1307 1308set port - tx_rate (for VF) 1309~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1310 1311Set TX rate limitation for queues in VF on a port:: 1312 1313 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask) 1314 1315set port - mirror rule 1316~~~~~~~~~~~~~~~~~~~~~~ 1317 1318Set pool or vlan type mirror rule for a port:: 1319 1320 testpmd> set port (port_id) mirror-rule (rule_id) \ 1321 (pool-mirror-up|pool-mirror-down|vlan-mirror) \ 1322 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off) 1323 1324Set link mirror rule for a port:: 1325 1326 testpmd> set port (port_id) mirror-rule (rule_id) \ 1327 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off) 1328 1329For example to enable mirror traffic with vlan 0,1 to pool 0:: 1330 1331 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on 1332 1333reset port - mirror rule 1334~~~~~~~~~~~~~~~~~~~~~~~~ 1335 1336Reset a mirror rule for a port:: 1337 1338 testpmd> reset port (port_id) mirror-rule (rule_id) 1339 1340set flush_rx 1341~~~~~~~~~~~~ 1342 1343Set the flush on RX streams before forwarding. 1344The default is flush ``on``. 1345Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams:: 1346 1347 testpmd> set flush_rx off 1348 1349set bypass mode 1350~~~~~~~~~~~~~~~ 1351 1352Set the bypass mode for the lowest port on bypass enabled NIC:: 1353 1354 testpmd> set bypass mode (normal|bypass|isolate) (port_id) 1355 1356set bypass event 1357~~~~~~~~~~~~~~~~ 1358 1359Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled:: 1360 1361 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \ 1362 mode (normal|bypass|isolate) (port_id) 1363 1364Where: 1365 1366* ``timeout``: Enable bypass after watchdog timeout. 1367 1368* ``os_on``: Enable bypass when OS/board is powered on. 1369 1370* ``os_off``: Enable bypass when OS/board is powered off. 1371 1372* ``power_on``: Enable bypass when power supply is turned on. 1373 1374* ``power_off``: Enable bypass when power supply is turned off. 1375 1376 1377set bypass timeout 1378~~~~~~~~~~~~~~~~~~ 1379 1380Set the bypass watchdog timeout to ``n`` seconds where 0 = instant:: 1381 1382 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32) 1383 1384show bypass config 1385~~~~~~~~~~~~~~~~~~ 1386 1387Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC:: 1388 1389 testpmd> show bypass config (port_id) 1390 1391set link up 1392~~~~~~~~~~~ 1393 1394Set link up for a port:: 1395 1396 testpmd> set link-up port (port id) 1397 1398set link down 1399~~~~~~~~~~~~~ 1400 1401Set link down for a port:: 1402 1403 testpmd> set link-down port (port id) 1404 1405E-tag set 1406~~~~~~~~~ 1407 1408Enable E-tag insertion for a VF on a port:: 1409 1410 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id) 1411 1412Disable E-tag insertion for a VF on a port:: 1413 1414 testpmd> E-tag set insertion off port (port_id) vf (vf_id) 1415 1416Enable/disable E-tag stripping on a port:: 1417 1418 testpmd> E-tag set stripping (on|off) port (port_id) 1419 1420Enable/disable E-tag based forwarding on a port:: 1421 1422 testpmd> E-tag set forwarding (on|off) port (port_id) 1423 1424Add an E-tag forwarding filter on a port:: 1425 1426 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id) 1427 1428Delete an E-tag forwarding filter on a port:: 1429 testpmd> E-tag set filter del e-tag-id (value) port (port_id) 1430 1431ddp add 1432~~~~~~~ 1433 1434Load a dynamic device personalization (DDP) profile and store backup profile:: 1435 1436 testpmd> ddp add (port_id) (profile_path[,backup_profile_path]) 1437 1438ddp del 1439~~~~~~~ 1440 1441Delete a dynamic device personalization profile and restore backup profile:: 1442 1443 testpmd> ddp del (port_id) (backup_profile_path) 1444 1445ptype mapping 1446~~~~~~~~~~~~~ 1447 1448List all items from the ptype mapping table:: 1449 1450 testpmd> ptype mapping get (port_id) (valid_only) 1451 1452Where: 1453 1454* ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0). 1455 1456Replace a specific or a group of software defined ptype with a new one:: 1457 1458 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type) 1459 1460where: 1461 1462* ``target``: A specific software ptype or a mask to represent a group of software ptypes. 1463 1464* ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1). 1465 1466* ``pkt_type``: The new software ptype to replace the old ones. 1467 1468Update hardware defined ptype to software defined packet type mapping table:: 1469 1470 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype) 1471 1472where: 1473 1474* ``hw_ptype``: hardware ptype as the index of the ptype mapping table. 1475 1476* ``sw_ptype``: software ptype as the value of the ptype mapping table. 1477 1478Reset ptype mapping table:: 1479 1480 testpmd> ptype mapping reset (port_id) 1481 1482config per port Rx offloading 1483~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1484 1485Enable or disable a per port Rx offloading on all Rx queues of a port:: 1486 1487 testpmd> port config (port_id) rx_offload (offloading) on|off 1488 1489* ``offloading``: can be any of these offloading capability: 1490 vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro, 1491 qinq_strip, outer_ipv4_cksum, macsec_strip, 1492 header_split, vlan_filter, vlan_extend, jumbo_frame, 1493 crc_strip, scatter, timestamp, security 1494 1495This command should be run when the port is stopped, or else it will fail. 1496 1497config per queue Rx offloading 1498~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1499 1500Enable or disable a per queue Rx offloading only on a specific Rx queue:: 1501 1502 testpmd> port (port_id) rxq (queue_id) rx_offload (offloading) on|off 1503 1504* ``offloading``: can be any of these offloading capability: 1505 vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro, 1506 qinq_strip, outer_ipv4_cksum, macsec_strip, 1507 header_split, vlan_filter, vlan_extend, jumbo_frame, 1508 crc_strip, scatter, timestamp, security 1509 1510This command should be run when the port is stopped, or else it will fail. 1511 1512config per port Tx offloading 1513~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1514 1515Enable or disable a per port Tx offloading on all Tx queues of a port:: 1516 1517 testpmd> port config (port_id) tx_offload (offloading) on|off 1518 1519* ``offloading``: can be any of these offloading capability: 1520 vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum, 1521 sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum, 1522 qinq_insert, vxlan_tnl_tso, gre_tnl_tso, 1523 ipip_tnl_tso, geneve_tnl_tso, macsec_insert, 1524 mt_lockfree, multi_segs, fast_free, security 1525 1526This command should be run when the port is stopped, or else it will fail. 1527 1528config per queue Tx offloading 1529~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1530 1531Enable or disable a per queue Tx offloading only on a specific Tx queue:: 1532 1533 testpmd> port (port_id) txq (queue_id) tx_offload (offloading) on|off 1534 1535* ``offloading``: can be any of these offloading capability: 1536 vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum, 1537 sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum, 1538 qinq_insert, vxlan_tnl_tso, gre_tnl_tso, 1539 ipip_tnl_tso, geneve_tnl_tso, macsec_insert, 1540 mt_lockfree, multi_segs, fast_free, security 1541 1542This command should be run when the port is stopped, or else it will fail. 1543 1544Config VXLAN Encap outer layers 1545~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1546 1547Configure the outer layer to encapsulate a packet inside a VXLAN tunnel:: 1548 1549 set vxlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \ 1550 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) eth-src (eth-src) \ 1551 eth-dst (eth-dst) 1552 1553 set vxlan-with-vlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \ 1554 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) vlan-tci (vlan-tci) \ 1555 eth-src (eth-src) eth-dst (eth-dst) 1556 1557Those command will set an internal configuration inside testpmd, any following 1558flow rule using the action vxlan_encap will use the last configuration set. 1559To have a different encapsulation header, one of those commands must be called 1560before the flow rule creation. 1561 1562Config NVGRE Encap outer layers 1563~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1564 1565Configure the outer layer to encapsulate a packet inside a NVGRE tunnel:: 1566 1567 set nvgre ip-version (ipv4|ipv6) tni (tni) ip-src (ip-src) ip-dst (ip-dst) \ 1568 eth-src (eth-src) eth-dst (eth-dst) 1569 set nvgre-with-vlan ip-version (ipv4|ipv6) tni (tni) ip-src (ip-src) \ 1570 ip-dst (ip-dst) vlan-tci (vlan-tci) eth-src (eth-src) eth-dst (eth-dst) 1571 1572Those command will set an internal configuration inside testpmd, any following 1573flow rule using the action nvgre_encap will use the last configuration set. 1574To have a different encapsulation header, one of those commands must be called 1575before the flow rule creation. 1576 1577Port Functions 1578-------------- 1579 1580The following sections show functions for configuring ports. 1581 1582.. note:: 1583 1584 Port configuration changes only become active when forwarding is started/restarted. 1585 1586port attach 1587~~~~~~~~~~~ 1588 1589Attach a port specified by pci address or virtual device args:: 1590 1591 testpmd> port attach (identifier) 1592 1593To attach a new pci device, the device should be recognized by kernel first. 1594Then it should be moved under DPDK management. 1595Finally the port can be attached to testpmd. 1596 1597For example, to move a pci device using ixgbe under DPDK management: 1598 1599.. code-block:: console 1600 1601 # Check the status of the available devices. 1602 ./usertools/dpdk-devbind.py --status 1603 1604 Network devices using DPDK-compatible driver 1605 ============================================ 1606 <none> 1607 1608 Network devices using kernel driver 1609 =================================== 1610 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused= 1611 1612 1613 # Bind the device to igb_uio. 1614 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0 1615 1616 1617 # Recheck the status of the devices. 1618 ./usertools/dpdk-devbind.py --status 1619 Network devices using DPDK-compatible driver 1620 ============================================ 1621 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused= 1622 1623To attach a port created by virtual device, above steps are not needed. 1624 1625For example, to attach a port whose pci address is 0000:0a:00.0. 1626 1627.. code-block:: console 1628 1629 testpmd> port attach 0000:0a:00.0 1630 Attaching a new port... 1631 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1632 EAL: probe driver: 8086:10fb rte_ixgbe_pmd 1633 EAL: PCI memory mapped at 0x7f83bfa00000 1634 EAL: PCI memory mapped at 0x7f83bfa80000 1635 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5 1636 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb 1637 Port 0 is attached. Now total ports is 1 1638 Done 1639 1640For example, to attach a port created by pcap PMD. 1641 1642.. code-block:: console 1643 1644 testpmd> port attach net_pcap0 1645 Attaching a new port... 1646 PMD: Initializing pmd_pcap for net_pcap0 1647 PMD: Creating pcap-backed ethdev on numa socket 0 1648 Port 0 is attached. Now total ports is 1 1649 Done 1650 1651In this case, identifier is ``net_pcap0``. 1652This identifier format is the same as ``--vdev`` format of DPDK applications. 1653 1654For example, to re-attach a bonded port which has been previously detached, 1655the mode and slave parameters must be given. 1656 1657.. code-block:: console 1658 1659 testpmd> port attach net_bond_0,mode=0,slave=1 1660 Attaching a new port... 1661 EAL: Initializing pmd_bond for net_bond_0 1662 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0. 1663 Port 0 is attached. Now total ports is 1 1664 Done 1665 1666 1667port detach 1668~~~~~~~~~~~ 1669 1670Detach a specific port:: 1671 1672 testpmd> port detach (port_id) 1673 1674Before detaching a port, the port should be stopped and closed. 1675 1676For example, to detach a pci device port 0. 1677 1678.. code-block:: console 1679 1680 testpmd> port stop 0 1681 Stopping ports... 1682 Done 1683 testpmd> port close 0 1684 Closing ports... 1685 Done 1686 1687 testpmd> port detach 0 1688 Detaching a port... 1689 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1690 EAL: remove driver: 8086:10fb rte_ixgbe_pmd 1691 EAL: PCI memory unmapped at 0x7f83bfa00000 1692 EAL: PCI memory unmapped at 0x7f83bfa80000 1693 Done 1694 1695 1696For example, to detach a virtual device port 0. 1697 1698.. code-block:: console 1699 1700 testpmd> port stop 0 1701 Stopping ports... 1702 Done 1703 testpmd> port close 0 1704 Closing ports... 1705 Done 1706 1707 testpmd> port detach 0 1708 Detaching a port... 1709 PMD: Closing pcap ethdev on numa socket 0 1710 Port 'net_pcap0' is detached. Now total ports is 0 1711 Done 1712 1713To remove a pci device completely from the system, first detach the port from testpmd. 1714Then the device should be moved under kernel management. 1715Finally the device can be removed using kernel pci hotplug functionality. 1716 1717For example, to move a pci device under kernel management: 1718 1719.. code-block:: console 1720 1721 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0 1722 1723 ./usertools/dpdk-devbind.py --status 1724 1725 Network devices using DPDK-compatible driver 1726 ============================================ 1727 <none> 1728 1729 Network devices using kernel driver 1730 =================================== 1731 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio 1732 1733To remove a port created by a virtual device, above steps are not needed. 1734 1735port start 1736~~~~~~~~~~ 1737 1738Start all ports or a specific port:: 1739 1740 testpmd> port start (port_id|all) 1741 1742port stop 1743~~~~~~~~~ 1744 1745Stop all ports or a specific port:: 1746 1747 testpmd> port stop (port_id|all) 1748 1749port close 1750~~~~~~~~~~ 1751 1752Close all ports or a specific port:: 1753 1754 testpmd> port close (port_id|all) 1755 1756port config - queue ring size 1757~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1758 1759Configure a rx/tx queue ring size:: 1760 1761 testpmd> port (port_id) (rxq|txq) (queue_id) ring_size (value) 1762 1763Only take effect after command that (re-)start the port or command that setup specific queue. 1764 1765port start/stop queue 1766~~~~~~~~~~~~~~~~~~~~~ 1767 1768Start/stop a rx/tx queue on a specific port:: 1769 1770 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop) 1771 1772port setup queue 1773~~~~~~~~~~~~~~~~~~~~~ 1774 1775Setup a rx/tx queue on a specific port:: 1776 1777 testpmd> port (port_id) (rxq|txq) (queue_id) setup 1778 1779Only take effect when port is started. 1780 1781port config - speed 1782~~~~~~~~~~~~~~~~~~~ 1783 1784Set the speed and duplex mode for all ports or a specific port:: 1785 1786 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \ 1787 duplex (half|full|auto) 1788 1789port config - queues/descriptors 1790~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1791 1792Set number of queues/descriptors for rxq, txq, rxd and txd:: 1793 1794 testpmd> port config all (rxq|txq|rxd|txd) (value) 1795 1796This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options. 1797 1798port config - max-pkt-len 1799~~~~~~~~~~~~~~~~~~~~~~~~~ 1800 1801Set the maximum packet length:: 1802 1803 testpmd> port config all max-pkt-len (value) 1804 1805This is equivalent to the ``--max-pkt-len`` command-line option. 1806 1807port config - CRC Strip 1808~~~~~~~~~~~~~~~~~~~~~~~ 1809 1810Set hardware CRC stripping on or off for all ports:: 1811 1812 testpmd> port config all crc-strip (on|off) 1813 1814CRC stripping is on by default. 1815 1816The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option. 1817 1818port config - scatter 1819~~~~~~~~~~~~~~~~~~~~~~~ 1820 1821Set RX scatter mode on or off for all ports:: 1822 1823 testpmd> port config all scatter (on|off) 1824 1825RX scatter mode is off by default. 1826 1827The ``on`` option is equivalent to the ``--enable-scatter`` command-line option. 1828 1829port config - RX Checksum 1830~~~~~~~~~~~~~~~~~~~~~~~~~ 1831 1832Set hardware RX checksum offload to on or off for all ports:: 1833 1834 testpmd> port config all rx-cksum (on|off) 1835 1836Checksum offload is off by default. 1837 1838The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option. 1839 1840port config - VLAN 1841~~~~~~~~~~~~~~~~~~ 1842 1843Set hardware VLAN on or off for all ports:: 1844 1845 testpmd> port config all hw-vlan (on|off) 1846 1847Hardware VLAN is off by default. 1848 1849The ``on`` option is equivalent to the ``--enable-hw-vlan`` command-line option. 1850 1851port config - VLAN filter 1852~~~~~~~~~~~~~~~~~~~~~~~~~ 1853 1854Set hardware VLAN filter on or off for all ports:: 1855 1856 testpmd> port config all hw-vlan-filter (on|off) 1857 1858Hardware VLAN filter is off by default. 1859 1860The ``on`` option is equivalent to the ``--enable-hw-vlan-filter`` command-line option. 1861 1862port config - VLAN strip 1863~~~~~~~~~~~~~~~~~~~~~~~~ 1864 1865Set hardware VLAN strip on or off for all ports:: 1866 1867 testpmd> port config all hw-vlan-strip (on|off) 1868 1869Hardware VLAN strip is off by default. 1870 1871The ``on`` option is equivalent to the ``--enable-hw-vlan-strip`` command-line option. 1872 1873port config - VLAN extend 1874~~~~~~~~~~~~~~~~~~~~~~~~~ 1875 1876Set hardware VLAN extend on or off for all ports:: 1877 1878 testpmd> port config all hw-vlan-extend (on|off) 1879 1880Hardware VLAN extend is off by default. 1881 1882The ``on`` option is equivalent to the ``--enable-hw-vlan-extend`` command-line option. 1883 1884port config - Drop Packets 1885~~~~~~~~~~~~~~~~~~~~~~~~~~ 1886 1887Set packet drop for packets with no descriptors on or off for all ports:: 1888 1889 testpmd> port config all drop-en (on|off) 1890 1891Packet dropping for packets with no descriptors is off by default. 1892 1893The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option. 1894 1895port config - RSS 1896~~~~~~~~~~~~~~~~~ 1897 1898Set the RSS (Receive Side Scaling) mode on or off:: 1899 1900 testpmd> port config all rss (all|default|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none) 1901 1902RSS is on by default. 1903 1904The ``all`` option is equivalent to ip|tcp|udp|sctp|ether. 1905The ``default`` option enables all supported RSS types reported by device info. 1906The ``none`` option is equivalent to the ``--disable-rss`` command-line option. 1907 1908port config - RSS Reta 1909~~~~~~~~~~~~~~~~~~~~~~ 1910 1911Set the RSS (Receive Side Scaling) redirection table:: 1912 1913 testpmd> port config all rss reta (hash,queue)[,(hash,queue)] 1914 1915port config - DCB 1916~~~~~~~~~~~~~~~~~ 1917 1918Set the DCB mode for an individual port:: 1919 1920 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off) 1921 1922The traffic class should be 4 or 8. 1923 1924port config - Burst 1925~~~~~~~~~~~~~~~~~~~ 1926 1927Set the number of packets per burst:: 1928 1929 testpmd> port config all burst (value) 1930 1931This is equivalent to the ``--burst`` command-line option. 1932 1933port config - Threshold 1934~~~~~~~~~~~~~~~~~~~~~~~ 1935 1936Set thresholds for TX/RX queues:: 1937 1938 testpmd> port config all (threshold) (value) 1939 1940Where the threshold type can be: 1941 1942* ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255. 1943 1944* ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255. 1945 1946* ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255. 1947 1948* ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255. 1949 1950* ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255. 1951 1952* ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255. 1953 1954* ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd. 1955 1956* ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd. 1957 1958* ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd. 1959 1960These threshold options are also available from the command-line. 1961 1962port config - E-tag 1963~~~~~~~~~~~~~~~~~~~ 1964 1965Set the value of ether-type for E-tag:: 1966 1967 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value) 1968 1969Enable/disable the E-tag support:: 1970 1971 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable) 1972 1973port config pctype mapping 1974~~~~~~~~~~~~~~~~~~~~~~~~~~ 1975 1976Reset pctype mapping table:: 1977 1978 testpmd> port config (port_id) pctype mapping reset 1979 1980Update hardware defined pctype to software defined flow type mapping table:: 1981 1982 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id) 1983 1984where: 1985 1986* ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table. 1987 1988* ``flow_type_id``: software flow type id as the index of the pctype mapping table. 1989 1990port config input set 1991~~~~~~~~~~~~~~~~~~~~~ 1992 1993Config RSS/FDIR/FDIR flexible payload input set for some pctype:: 1994 testpmd> port config (port_id) pctype (pctype_id) \ 1995 (hash_inset|fdir_inset|fdir_flx_inset) \ 1996 (get|set|clear) field (field_idx) 1997 1998Clear RSS/FDIR/FDIR flexible payload input set for some pctype:: 1999 testpmd> port config (port_id) pctype (pctype_id) \ 2000 (hash_inset|fdir_inset|fdir_flx_inset) clear all 2001 2002where: 2003 2004* ``pctype_id``: hardware packet classification types. 2005* ``field_idx``: hardware field index. 2006 2007port config udp_tunnel_port 2008~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2009 2010Add/remove UDP tunnel port for VXLAN/GENEVE tunneling protocols:: 2011 testpmd> port config (port_id) udp_tunnel_port add|rm vxlan|geneve (udp_port) 2012 2013Link Bonding Functions 2014---------------------- 2015 2016The Link Bonding functions make it possible to dynamically create and 2017manage link bonding devices from within testpmd interactive prompt. 2018 2019create bonded device 2020~~~~~~~~~~~~~~~~~~~~ 2021 2022Create a new bonding device:: 2023 2024 testpmd> create bonded device (mode) (socket) 2025 2026For example, to create a bonded device in mode 1 on socket 0:: 2027 2028 testpmd> create bonded 1 0 2029 created new bonded device (port X) 2030 2031add bonding slave 2032~~~~~~~~~~~~~~~~~ 2033 2034Adds Ethernet device to a Link Bonding device:: 2035 2036 testpmd> add bonding slave (slave id) (port id) 2037 2038For example, to add Ethernet device (port 6) to a Link Bonding device (port 10):: 2039 2040 testpmd> add bonding slave 6 10 2041 2042 2043remove bonding slave 2044~~~~~~~~~~~~~~~~~~~~ 2045 2046Removes an Ethernet slave device from a Link Bonding device:: 2047 2048 testpmd> remove bonding slave (slave id) (port id) 2049 2050For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10):: 2051 2052 testpmd> remove bonding slave 6 10 2053 2054set bonding mode 2055~~~~~~~~~~~~~~~~ 2056 2057Set the Link Bonding mode of a Link Bonding device:: 2058 2059 testpmd> set bonding mode (value) (port id) 2060 2061For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3):: 2062 2063 testpmd> set bonding mode 3 10 2064 2065set bonding primary 2066~~~~~~~~~~~~~~~~~~~ 2067 2068Set an Ethernet slave device as the primary device on a Link Bonding device:: 2069 2070 testpmd> set bonding primary (slave id) (port id) 2071 2072For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10):: 2073 2074 testpmd> set bonding primary 6 10 2075 2076set bonding mac 2077~~~~~~~~~~~~~~~ 2078 2079Set the MAC address of a Link Bonding device:: 2080 2081 testpmd> set bonding mac (port id) (mac) 2082 2083For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01:: 2084 2085 testpmd> set bonding mac 10 00:00:00:00:00:01 2086 2087set bonding xmit_balance_policy 2088~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2089 2090Set the transmission policy for a Link Bonding device when it is in Balance XOR mode:: 2091 2092 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34) 2093 2094For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports):: 2095 2096 testpmd> set bonding xmit_balance_policy 10 l34 2097 2098 2099set bonding mon_period 2100~~~~~~~~~~~~~~~~~~~~~~ 2101 2102Set the link status monitoring polling period in milliseconds for a bonding device. 2103 2104This adds support for PMD slave devices which do not support link status interrupts. 2105When the mon_period is set to a value greater than 0 then all PMD's which do not support 2106link status ISR will be queried every polling interval to check if their link status has changed:: 2107 2108 testpmd> set bonding mon_period (port_id) (value) 2109 2110For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms:: 2111 2112 testpmd> set bonding mon_period 5 150 2113 2114 2115set bonding lacp dedicated_queue 2116~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2117 2118Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic 2119when in mode 4 (link-aggregration-802.3ad):: 2120 2121 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable) 2122 2123 2124set bonding agg_mode 2125~~~~~~~~~~~~~~~~~~~~ 2126 2127Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad):: 2128 2129 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable) 2130 2131 2132show bonding config 2133~~~~~~~~~~~~~~~~~~~ 2134 2135Show the current configuration of a Link Bonding device:: 2136 2137 testpmd> show bonding config (port id) 2138 2139For example, 2140to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4) 2141in balance mode with a transmission policy of layer 2+3:: 2142 2143 testpmd> show bonding config 9 2144 Bonding mode: 2 2145 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23 2146 Slaves (3): [1 3 4] 2147 Active Slaves (3): [1 3 4] 2148 Primary: [3] 2149 2150 2151Register Functions 2152------------------ 2153 2154The Register Functions can be used to read from and write to registers on the network card referenced by a port number. 2155This is mainly useful for debugging purposes. 2156Reference should be made to the appropriate datasheet for the network card for details on the register addresses 2157and fields that can be accessed. 2158 2159read reg 2160~~~~~~~~ 2161 2162Display the value of a port register:: 2163 2164 testpmd> read reg (port_id) (address) 2165 2166For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller:: 2167 2168 testpmd> read reg 0 0xEE00 2169 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241) 2170 2171read regfield 2172~~~~~~~~~~~~~ 2173 2174Display a port register bit field:: 2175 2176 testpmd> read regfield (port_id) (address) (bit_x) (bit_y) 2177 2178For example, reading the lowest two bits from the register in the example above:: 2179 2180 testpmd> read regfield 0 0xEE00 0 1 2181 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1) 2182 2183read regbit 2184~~~~~~~~~~~ 2185 2186Display a single port register bit:: 2187 2188 testpmd> read regbit (port_id) (address) (bit_x) 2189 2190For example, reading the lowest bit from the register in the example above:: 2191 2192 testpmd> read regbit 0 0xEE00 0 2193 port 0 PCI register at offset 0xEE00: bit 0=1 2194 2195write reg 2196~~~~~~~~~ 2197 2198Set the value of a port register:: 2199 2200 testpmd> write reg (port_id) (address) (value) 2201 2202For example, to clear a register:: 2203 2204 testpmd> write reg 0 0xEE00 0x0 2205 port 0 PCI register at offset 0xEE00: 0x00000000 (0) 2206 2207write regfield 2208~~~~~~~~~~~~~~ 2209 2210Set bit field of a port register:: 2211 2212 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value) 2213 2214For example, writing to the register cleared in the example above:: 2215 2216 testpmd> write regfield 0 0xEE00 0 1 2 2217 port 0 PCI register at offset 0xEE00: 0x00000002 (2) 2218 2219write regbit 2220~~~~~~~~~~~~ 2221 2222Set single bit value of a port register:: 2223 2224 testpmd> write regbit (port_id) (address) (bit_x) (value) 2225 2226For example, to set the high bit in the register from the example above:: 2227 2228 testpmd> write regbit 0 0xEE00 31 1 2229 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658) 2230 2231Traffic Metering and Policing 2232----------------------------- 2233 2234The following section shows functions for configuring traffic metering and 2235policing on the ethernet device through the use of generic ethdev API. 2236 2237show port traffic management capability 2238~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2239 2240Show traffic metering and policing capability of the port:: 2241 2242 testpmd> show port meter cap (port_id) 2243 2244add port meter profile (srTCM rfc2967) 2245~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2246 2247Add meter profile (srTCM rfc2697) to the ethernet device:: 2248 2249 testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \ 2250 (cir) (cbs) (ebs) 2251 2252where: 2253 2254* ``profile_id``: ID for the meter profile. 2255* ``cir``: Committed Information Rate (CIR) (bytes/second). 2256* ``cbs``: Committed Burst Size (CBS) (bytes). 2257* ``ebs``: Excess Burst Size (EBS) (bytes). 2258 2259add port meter profile (trTCM rfc2968) 2260~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2261 2262Add meter profile (srTCM rfc2698) to the ethernet device:: 2263 2264 testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \ 2265 (cir) (pir) (cbs) (pbs) 2266 2267where: 2268 2269* ``profile_id``: ID for the meter profile. 2270* ``cir``: Committed information rate (bytes/second). 2271* ``pir``: Peak information rate (bytes/second). 2272* ``cbs``: Committed burst size (bytes). 2273* ``pbs``: Peak burst size (bytes). 2274 2275add port meter profile (trTCM rfc4115) 2276~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2277 2278Add meter profile (trTCM rfc4115) to the ethernet device:: 2279 2280 testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \ 2281 (cir) (eir) (cbs) (ebs) 2282 2283where: 2284 2285* ``profile_id``: ID for the meter profile. 2286* ``cir``: Committed information rate (bytes/second). 2287* ``eir``: Excess information rate (bytes/second). 2288* ``cbs``: Committed burst size (bytes). 2289* ``ebs``: Excess burst size (bytes). 2290 2291delete port meter profile 2292~~~~~~~~~~~~~~~~~~~~~~~~~ 2293 2294Delete meter profile from the ethernet device:: 2295 2296 testpmd> del port meter profile (port_id) (profile_id) 2297 2298create port meter 2299~~~~~~~~~~~~~~~~~ 2300 2301Create new meter object for the ethernet device:: 2302 2303 testpmd> create port meter (port_id) (mtr_id) (profile_id) \ 2304 (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \ 2305 (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\ 2306 (dscp_tbl_entry63)] 2307 2308where: 2309 2310* ``mtr_id``: meter object ID. 2311* ``profile_id``: ID for the meter profile. 2312* ``meter_enable``: When this parameter has a non-zero value, the meter object 2313 gets enabled at the time of creation, otherwise remains disabled. 2314* ``g_action``: Policer action for the packet with green color. 2315* ``y_action``: Policer action for the packet with yellow color. 2316* ``r_action``: Policer action for the packet with red color. 2317* ``stats_mask``: Mask of statistics counter types to be enabled for the 2318 meter object. 2319* ``shared``: When this parameter has a non-zero value, the meter object is 2320 shared by multiple flows. Otherwise, meter object is used by single flow. 2321* ``use_pre_meter_color``: When this parameter has a non-zero value, the 2322 input color for the current meter object is determined by the latest meter 2323 object in the same flow. Otherwise, the current meter object uses the 2324 *dscp_table* to determine the input color. 2325* ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input 2326 color, 0 <= x <= 63. 2327 2328enable port meter 2329~~~~~~~~~~~~~~~~~ 2330 2331Enable meter for the ethernet device:: 2332 2333 testpmd> enable port meter (port_id) (mtr_id) 2334 2335disable port meter 2336~~~~~~~~~~~~~~~~~~ 2337 2338Disable meter for the ethernet device:: 2339 2340 testpmd> disable port meter (port_id) (mtr_id) 2341 2342delete port meter 2343~~~~~~~~~~~~~~~~~ 2344 2345Delete meter for the ethernet device:: 2346 2347 testpmd> del port meter (port_id) (mtr_id) 2348 2349Set port meter profile 2350~~~~~~~~~~~~~~~~~~~~~~ 2351 2352Set meter profile for the ethernet device:: 2353 2354 testpmd> set port meter profile (port_id) (mtr_id) (profile_id) 2355 2356set port meter dscp table 2357~~~~~~~~~~~~~~~~~~~~~~~~~ 2358 2359Set meter dscp table for the ethernet device:: 2360 2361 testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \ 2362 (dscp_tbl_entry1)...(dscp_tbl_entry63)] 2363 2364set port meter policer action 2365~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2366 2367Set meter policer action for the ethernet device:: 2368 2369 testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \ 2370 (action0) [(action1) (action1)] 2371 2372where: 2373 2374* ``action_mask``: Bit mask indicating which policer actions need to be 2375 updated. One or more policer actions can be updated in a single function 2376 invocation. To update the policer action associated with color C, bit 2377 (1 << C) needs to be set in *action_mask* and element at position C 2378 in the *actions* array needs to be valid. 2379* ``actionx``: Policer action for the color x, 2380 RTE_MTR_GREEN <= x < RTE_MTR_COLORS 2381 2382set port meter stats mask 2383~~~~~~~~~~~~~~~~~~~~~~~~~ 2384 2385Set meter stats mask for the ethernet device:: 2386 2387 testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask) 2388 2389where: 2390 2391* ``stats_mask``: Bit mask indicating statistics counter types to be enabled. 2392 2393show port meter stats 2394~~~~~~~~~~~~~~~~~~~~~ 2395 2396Show meter stats of the ethernet device:: 2397 2398 testpmd> show port meter stats (port_id) (mtr_id) (clear) 2399 2400where: 2401 2402* ``clear``: Flag that indicates whether the statistics counters should 2403 be cleared (i.e. set to zero) immediately after they have been read or not. 2404 2405Traffic Management 2406------------------ 2407 2408The following section shows functions for configuring traffic management on 2409on the ethernet device through the use of generic TM API. 2410 2411show port traffic management capability 2412~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2413 2414Show traffic management capability of the port:: 2415 2416 testpmd> show port tm cap (port_id) 2417 2418show port traffic management capability (hierarchy level) 2419~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2420 2421Show traffic management hierarchy level capability of the port:: 2422 2423 testpmd> show port tm level cap (port_id) (level_id) 2424 2425show port traffic management capability (hierarchy node level) 2426~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2427 2428Show the traffic management hierarchy node capability of the port:: 2429 2430 testpmd> show port tm node cap (port_id) (node_id) 2431 2432show port traffic management hierarchy node type 2433~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2434 2435Show the port traffic management hierarchy node type:: 2436 2437 testpmd> show port tm node type (port_id) (node_id) 2438 2439show port traffic management hierarchy node stats 2440~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2441 2442Show the port traffic management hierarchy node statistics:: 2443 2444 testpmd> show port tm node stats (port_id) (node_id) (clear) 2445 2446where: 2447 2448* ``clear``: When this parameter has a non-zero value, the statistics counters 2449 are cleared (i.e. set to zero) immediately after they have been read, 2450 otherwise the statistics counters are left untouched. 2451 2452Add port traffic management private shaper profile 2453~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2454 2455Add the port traffic management private shaper profile:: 2456 2457 testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \ 2458 (tb_rate) (tb_size) (packet_length_adjust) 2459 2460where: 2461 2462* ``shaper_profile id``: Shaper profile ID for the new profile. 2463* ``tb_rate``: Token bucket rate (bytes per second). 2464* ``tb_size``: Token bucket size (bytes). 2465* ``packet_length_adjust``: The value (bytes) to be added to the length of 2466 each packet for the purpose of shaping. This parameter value can be used to 2467 correct the packet length with the framing overhead bytes that are consumed 2468 on the wire. 2469 2470Delete port traffic management private shaper profile 2471~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2472 2473Delete the port traffic management private shaper:: 2474 2475 testpmd> del port tm node shaper profile (port_id) (shaper_profile_id) 2476 2477where: 2478 2479* ``shaper_profile id``: Shaper profile ID that needs to be deleted. 2480 2481Add port traffic management shared shaper 2482~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2483 2484Create the port traffic management shared shaper:: 2485 2486 testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \ 2487 (shaper_profile_id) 2488 2489where: 2490 2491* ``shared_shaper_id``: Shared shaper ID to be created. 2492* ``shaper_profile id``: Shaper profile ID for shared shaper. 2493 2494Set port traffic management shared shaper 2495~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2496 2497Update the port traffic management shared shaper:: 2498 2499 testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \ 2500 (shaper_profile_id) 2501 2502where: 2503 2504* ``shared_shaper_id``: Shared shaper ID to be update. 2505* ``shaper_profile id``: Shaper profile ID for shared shaper. 2506 2507Delete port traffic management shared shaper 2508~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2509 2510Delete the port traffic management shared shaper:: 2511 2512 testpmd> del port tm node shared shaper (port_id) (shared_shaper_id) 2513 2514where: 2515 2516* ``shared_shaper_id``: Shared shaper ID to be deleted. 2517 2518Set port traffic management hiearchy node private shaper 2519~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2520 2521set the port traffic management hierarchy node private shaper:: 2522 2523 testpmd> set port tm node shaper profile (port_id) (node_id) \ 2524 (shaper_profile_id) 2525 2526where: 2527 2528* ``shaper_profile id``: Private shaper profile ID to be enabled on the 2529 hierarchy node. 2530 2531Add port traffic management WRED profile 2532~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2533 2534Create a new WRED profile:: 2535 2536 testpmd> add port tm node wred profile (port_id) (wred_profile_id) \ 2537 (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \ 2538 (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \ 2539 (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r) 2540 2541where: 2542 2543* ``wred_profile id``: Identifier for the newly create WRED profile 2544* ``color_g``: Packet color (green) 2545* ``min_th_g``: Minimum queue threshold for packet with green color 2546* ``max_th_g``: Minimum queue threshold for packet with green color 2547* ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp) 2548* ``wq_log2_g``: Negated log2 of queue weight (wq) 2549* ``color_y``: Packet color (yellow) 2550* ``min_th_y``: Minimum queue threshold for packet with yellow color 2551* ``max_th_y``: Minimum queue threshold for packet with yellow color 2552* ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp) 2553* ``wq_log2_y``: Negated log2 of queue weight (wq) 2554* ``color_r``: Packet color (red) 2555* ``min_th_r``: Minimum queue threshold for packet with yellow color 2556* ``max_th_r``: Minimum queue threshold for packet with yellow color 2557* ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp) 2558* ``wq_log2_r``: Negated log2 of queue weight (wq) 2559 2560Delete port traffic management WRED profile 2561~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2562 2563Delete the WRED profile:: 2564 2565 testpmd> del port tm node wred profile (port_id) (wred_profile_id) 2566 2567Add port traffic management hierarchy nonleaf node 2568~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2569 2570Add nonleaf node to port traffic management hiearchy:: 2571 2572 testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \ 2573 (priority) (weight) (level_id) (shaper_profile_id) \ 2574 (n_sp_priorities) (stats_mask) (n_shared_shapers) \ 2575 [(shared_shaper_0) (shared_shaper_1) ...] \ 2576 2577where: 2578 2579* ``parent_node_id``: Node ID of the parent. 2580* ``priority``: Node priority (highest node priority is zero). This is used by 2581 the SP algorithm running on the parent node for scheduling this node. 2582* ``weight``: Node weight (lowest weight is one). The node weight is relative 2583 to the weight sum of all siblings that have the same priority. It is used by 2584 the WFQ algorithm running on the parent node for scheduling this node. 2585* ``level_id``: Hiearchy level of the node. 2586* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by 2587 the node. 2588* ``n_sp_priorities``: Number of strict priorities. 2589* ``stats_mask``: Mask of statistics counter types to be enabled for this node. 2590* ``n_shared_shapers``: Number of shared shapers. 2591* ``shared_shaper_id``: Shared shaper id. 2592 2593Add port traffic management hierarchy leaf node 2594~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2595 2596Add leaf node to port traffic management hiearchy:: 2597 2598 testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \ 2599 (priority) (weight) (level_id) (shaper_profile_id) \ 2600 (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \ 2601 [(shared_shaper_id) (shared_shaper_id) ...] \ 2602 2603where: 2604 2605* ``parent_node_id``: Node ID of the parent. 2606* ``priority``: Node priority (highest node priority is zero). This is used by 2607 the SP algorithm running on the parent node for scheduling this node. 2608* ``weight``: Node weight (lowest weight is one). The node weight is relative 2609 to the weight sum of all siblings that have the same priority. It is used by 2610 the WFQ algorithm running on the parent node for scheduling this node. 2611* ``level_id``: Hiearchy level of the node. 2612* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by 2613 the node. 2614* ``cman_mode``: Congestion management mode to be enabled for this node. 2615* ``wred_profile_id``: WRED profile id to be enabled for this node. 2616* ``stats_mask``: Mask of statistics counter types to be enabled for this node. 2617* ``n_shared_shapers``: Number of shared shapers. 2618* ``shared_shaper_id``: Shared shaper id. 2619 2620Delete port traffic management hierarchy node 2621~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2622 2623Delete node from port traffic management hiearchy:: 2624 2625 testpmd> del port tm node (port_id) (node_id) 2626 2627Update port traffic management hierarchy parent node 2628~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2629 2630Update port traffic management hierarchy parent node:: 2631 2632 testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \ 2633 (priority) (weight) 2634 2635This function can only be called after the hierarchy commit invocation. Its 2636success depends on the port support for this operation, as advertised through 2637the port capability set. This function is valid for all nodes of the traffic 2638management hierarchy except root node. 2639 2640Suspend port traffic management hierarchy node 2641~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2642 2643 testpmd> suspend port tm node (port_id) (node_id) 2644 2645Resume port traffic management hierarchy node 2646~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2647 2648 testpmd> resume port tm node (port_id) (node_id) 2649 2650Commit port traffic management hierarchy 2651~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2652 2653Commit the traffic management hierarchy on the port:: 2654 2655 testpmd> port tm hierarchy commit (port_id) (clean_on_fail) 2656 2657where: 2658 2659* ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function 2660 call failure. On the other hand, hierarchy is preserved when this parameter 2661 is equal to zero. 2662 2663Set port traffic management default hierarchy (tm forwarding mode) 2664~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2665 2666set the traffic management default hierarchy on the port:: 2667 2668 testpmd> set port tm hierarchy default (port_id) 2669 2670Filter Functions 2671---------------- 2672 2673This section details the available filter functions that are available. 2674 2675Note these functions interface the deprecated legacy filtering framework, 2676superseded by *rte_flow*. See `Flow rules management`_. 2677 2678ethertype_filter 2679~~~~~~~~~~~~~~~~~~~~ 2680 2681Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue:: 2682 2683 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \ 2684 ethertype (ether_type) (drop|fwd) queue (queue_id) 2685 2686The available information parameters are: 2687 2688* ``port_id``: The port which the Ethertype filter assigned on. 2689 2690* ``mac_addr``: Compare destination mac address. 2691 2692* ``mac_ignr``: Ignore destination mac address match. 2693 2694* ``mac_address``: Destination mac address to match. 2695 2696* ``ether_type``: The EtherType value want to match, 2697 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid. 2698 2699* ``queue_id``: The receive queue associated with this EtherType filter. 2700 It is meaningless when deleting or dropping. 2701 2702Example, to add/remove an ethertype filter rule:: 2703 2704 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \ 2705 ethertype 0x0806 fwd queue 3 2706 2707 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \ 2708 ethertype 0x0806 fwd queue 3 2709 27102tuple_filter 2711~~~~~~~~~~~~~~~~~ 2712 2713Add or delete a 2-tuple filter, 2714which identifies packets by specific protocol and destination TCP/UDP port 2715and forwards packets into one of the receive queues:: 2716 2717 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \ 2718 protocol (protocol_value) mask (mask_value) \ 2719 tcp_flags (tcp_flags_value) priority (prio_value) \ 2720 queue (queue_id) 2721 2722The available information parameters are: 2723 2724* ``port_id``: The port which the 2-tuple filter assigned on. 2725 2726* ``dst_port_value``: Destination port in L4. 2727 2728* ``protocol_value``: IP L4 protocol. 2729 2730* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate. 2731 2732* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP). 2733 2734* ``prio_value``: Priority of this filter. 2735 2736* ``queue_id``: The receive queue associated with this 2-tuple filter. 2737 2738Example, to add/remove an 2tuple filter rule:: 2739 2740 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \ 2741 tcp_flags 0x02 priority 3 queue 3 2742 2743 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \ 2744 tcp_flags 0x02 priority 3 queue 3 2745 27465tuple_filter 2747~~~~~~~~~~~~~~~~~ 2748 2749Add or delete a 5-tuple filter, 2750which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port) 2751and routes packets into one of the receive queues:: 2752 2753 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \ 2754 (src_address) dst_port (dst_port_value) \ 2755 src_port (src_port_value) protocol (protocol_value) \ 2756 mask (mask_value) tcp_flags (tcp_flags_value) \ 2757 priority (prio_value) queue (queue_id) 2758 2759The available information parameters are: 2760 2761* ``port_id``: The port which the 5-tuple filter assigned on. 2762 2763* ``dst_address``: Destination IP address. 2764 2765* ``src_address``: Source IP address. 2766 2767* ``dst_port_value``: TCP/UDP destination port. 2768 2769* ``src_port_value``: TCP/UDP source port. 2770 2771* ``protocol_value``: L4 protocol. 2772 2773* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate 2774 2775* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP). 2776 2777* ``prio_value``: The priority of this filter. 2778 2779* ``queue_id``: The receive queue associated with this 5-tuple filter. 2780 2781Example, to add/remove an 5tuple filter rule:: 2782 2783 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 2784 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 2785 flags 0x0 priority 3 queue 3 2786 2787 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 2788 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 2789 flags 0x0 priority 3 queue 3 2790 2791syn_filter 2792~~~~~~~~~~ 2793 2794Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue:: 2795 2796 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id) 2797 2798The available information parameters are: 2799 2800* ``port_id``: The port which the SYN filter assigned on. 2801 2802* ``high``: This SYN filter has higher priority than other filters. 2803 2804* ``low``: This SYN filter has lower priority than other filters. 2805 2806* ``queue_id``: The receive queue associated with this SYN filter 2807 2808Example:: 2809 2810 testpmd> syn_filter 0 add priority high queue 3 2811 2812flex_filter 2813~~~~~~~~~~~ 2814 2815With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet 2816and routed into one of the receive queues:: 2817 2818 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \ 2819 mask (mask_value) priority (prio_value) queue (queue_id) 2820 2821The available information parameters are: 2822 2823* ``port_id``: The port which the Flex filter is assigned on. 2824 2825* ``len_value``: Filter length in bytes, no greater than 128. 2826 2827* ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match. 2828 2829* ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match. 2830 2831* ``prio_value``: The priority of this filter. 2832 2833* ``queue_id``: The receive queue associated with this Flex filter. 2834 2835Example:: 2836 2837 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \ 2838 mask 000C priority 3 queue 3 2839 2840 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \ 2841 mask 000C priority 3 queue 3 2842 2843 2844.. _testpmd_flow_director: 2845 2846flow_director_filter 2847~~~~~~~~~~~~~~~~~~~~ 2848 2849The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues. 2850 2851Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and 2852Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter: 2853 2854* Perfect match filters. 2855 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2856 The masked fields are for IP flow. 2857 2858* Signature filters. 2859 The hardware checks a match between a hash-based signature of the masked fields of the received packet. 2860 2861* Perfect-mac-vlan match filters. 2862 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2863 The masked fields are for MAC VLAN flow. 2864 2865* Perfect-tunnel match filters. 2866 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2867 The masked fields are for tunnel flow. 2868 2869* Perfect-raw-flow-type match filters. 2870 The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet. 2871 The masked fields are specified by input sets. 2872 2873The Flow Director filters can match the different fields for different type of packet: flow type, specific input set 2874per flow type and the flexible payload. 2875 2876The Flow Director can also mask out parts of all of these fields so that filters 2877are only applied to certain fields or parts of the fields. 2878 2879Note that for raw flow type mode the source and destination fields in the 2880raw packet buffer need to be presented in a reversed order with respect 2881to the expected received packets. 2882For example: IP source and destination addresses or TCP/UDP/SCTP 2883source and destination ports 2884 2885Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information. 2886 2887# Commands to add flow director filters of different flow types:: 2888 2889 flow_director_filter (port_id) mode IP (add|del|update) \ 2890 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \ 2891 src (src_ip_address) dst (dst_ip_address) \ 2892 tos (tos_value) proto (proto_value) ttl (ttl_value) \ 2893 vlan (vlan_value) flexbytes (flexbytes_value) \ 2894 (drop|fwd) pf|vf(vf_id) queue (queue_id) \ 2895 fd_id (fd_id_value) 2896 2897 flow_director_filter (port_id) mode IP (add|del|update) \ 2898 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \ 2899 src (src_ip_address) (src_port) \ 2900 dst (dst_ip_address) (dst_port) \ 2901 tos (tos_value) ttl (ttl_value) \ 2902 vlan (vlan_value) flexbytes (flexbytes_value) \ 2903 (drop|fwd) queue pf|vf(vf_id) (queue_id) \ 2904 fd_id (fd_id_value) 2905 2906 flow_director_filter (port_id) mode IP (add|del|update) \ 2907 flow (ipv4-sctp|ipv6-sctp) \ 2908 src (src_ip_address) (src_port) \ 2909 dst (dst_ip_address) (dst_port) \ 2910 tos (tos_value) ttl (ttl_value) \ 2911 tag (verification_tag) vlan (vlan_value) \ 2912 flexbytes (flexbytes_value) (drop|fwd) \ 2913 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value) 2914 2915 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \ 2916 ether (ethertype) flexbytes (flexbytes_value) \ 2917 (drop|fwd) pf|vf(vf_id) queue (queue_id) 2918 fd_id (fd_id_value) 2919 2920 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \ 2921 mac (mac_address) vlan (vlan_value) \ 2922 flexbytes (flexbytes_value) (drop|fwd) \ 2923 queue (queue_id) fd_id (fd_id_value) 2924 2925 flow_director_filter (port_id) mode Tunnel (add|del|update) \ 2926 mac (mac_address) vlan (vlan_value) \ 2927 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \ 2928 flexbytes (flexbytes_value) (drop|fwd) \ 2929 queue (queue_id) fd_id (fd_id_value) 2930 2931 flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \ 2932 (drop|fwd) queue (queue_id) fd_id (fd_id_value) \ 2933 packet (packet file name) 2934 2935For example, to add an ipv4-udp flow type filter:: 2936 2937 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \ 2938 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \ 2939 fwd pf queue 1 fd_id 1 2940 2941For example, add an ipv4-other flow type filter:: 2942 2943 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \ 2944 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \ 2945 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1 2946 2947flush_flow_director 2948~~~~~~~~~~~~~~~~~~~ 2949 2950Flush all flow director filters on a device:: 2951 2952 testpmd> flush_flow_director (port_id) 2953 2954Example, to flush all flow director filter on port 0:: 2955 2956 testpmd> flush_flow_director 0 2957 2958flow_director_mask 2959~~~~~~~~~~~~~~~~~~ 2960 2961Set flow director's input masks:: 2962 2963 flow_director_mask (port_id) mode IP vlan (vlan_value) \ 2964 src_mask (ipv4_src) (ipv6_src) (src_port) \ 2965 dst_mask (ipv4_dst) (ipv6_dst) (dst_port) 2966 2967 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value) 2968 2969 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \ 2970 mac (mac_value) tunnel-type (tunnel_type_value) \ 2971 tunnel-id (tunnel_id_value) 2972 2973Example, to set flow director mask on port 0:: 2974 2975 testpmd> flow_director_mask 0 mode IP vlan 0xefff \ 2976 src_mask 255.255.255.255 \ 2977 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \ 2978 dst_mask 255.255.255.255 \ 2979 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF 2980 2981flow_director_flex_mask 2982~~~~~~~~~~~~~~~~~~~~~~~ 2983 2984set masks of flow director's flexible payload based on certain flow type:: 2985 2986 testpmd> flow_director_flex_mask (port_id) \ 2987 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2988 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \ 2989 l2_payload|all) (mask) 2990 2991Example, to set flow director's flex mask for all flow type on port 0:: 2992 2993 testpmd> flow_director_flex_mask 0 flow all \ 2994 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0) 2995 2996 2997flow_director_flex_payload 2998~~~~~~~~~~~~~~~~~~~~~~~~~~ 2999 3000Configure flexible payload selection:: 3001 3002 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config) 3003 3004For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload:: 3005 3006 testpmd> flow_director_flex_payload 0 l4 \ 3007 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19) 3008 3009get_sym_hash_ena_per_port 3010~~~~~~~~~~~~~~~~~~~~~~~~~ 3011 3012Get symmetric hash enable configuration per port:: 3013 3014 get_sym_hash_ena_per_port (port_id) 3015 3016For example, to get symmetric hash enable configuration of port 1:: 3017 3018 testpmd> get_sym_hash_ena_per_port 1 3019 3020set_sym_hash_ena_per_port 3021~~~~~~~~~~~~~~~~~~~~~~~~~ 3022 3023Set symmetric hash enable configuration per port to enable or disable:: 3024 3025 set_sym_hash_ena_per_port (port_id) (enable|disable) 3026 3027For example, to set symmetric hash enable configuration of port 1 to enable:: 3028 3029 testpmd> set_sym_hash_ena_per_port 1 enable 3030 3031get_hash_global_config 3032~~~~~~~~~~~~~~~~~~~~~~ 3033 3034Get the global configurations of hash filters:: 3035 3036 get_hash_global_config (port_id) 3037 3038For example, to get the global configurations of hash filters of port 1:: 3039 3040 testpmd> get_hash_global_config 1 3041 3042set_hash_global_config 3043~~~~~~~~~~~~~~~~~~~~~~ 3044 3045Set the global configurations of hash filters:: 3046 3047 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \ 3048 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \ 3049 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \ 3050 (enable|disable) 3051 3052For example, to enable simple_xor for flow type of ipv6 on port 2:: 3053 3054 testpmd> set_hash_global_config 2 simple_xor ipv6 enable 3055 3056set_hash_input_set 3057~~~~~~~~~~~~~~~~~~ 3058 3059Set the input set for hash:: 3060 3061 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 3062 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 3063 l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \ 3064 ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \ 3065 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \ 3066 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \ 3067 fld-8th|none) (select|add) 3068 3069For example, to add source IP to hash input set for flow type of ipv4-udp on port 0:: 3070 3071 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add 3072 3073set_fdir_input_set 3074~~~~~~~~~~~~~~~~~~ 3075 3076The Flow Director filters can match the different fields for different type of packet, i.e. specific input set 3077on per flow type and the flexible payload. This command can be used to change input set for each flow type. 3078 3079Set the input set for flow director:: 3080 3081 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 3082 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 3083 l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \ 3084 ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \ 3085 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \ 3086 sctp-dst-port|sctp-veri-tag|none) (select|add) 3087 3088For example to add source IP to FD input set for flow type of ipv4-udp on port 0:: 3089 3090 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add 3091 3092global_config 3093~~~~~~~~~~~~~ 3094 3095Set different GRE key length for input set:: 3096 3097 global_config (port_id) gre-key-len (number in bytes) 3098 3099For example to set GRE key length for input set to 4 bytes on port 0:: 3100 3101 testpmd> global_config 0 gre-key-len 4 3102 3103 3104.. _testpmd_rte_flow: 3105 3106Flow rules management 3107--------------------- 3108 3109Control of the generic flow API (*rte_flow*) is fully exposed through the 3110``flow`` command (validation, creation, destruction, queries and operation 3111modes). 3112 3113Considering *rte_flow* overlaps with all `Filter Functions`_, using both 3114features simultaneously may cause undefined side-effects and is therefore 3115not recommended. 3116 3117``flow`` syntax 3118~~~~~~~~~~~~~~~ 3119 3120Because the ``flow`` command uses dynamic tokens to handle the large number 3121of possible flow rules combinations, its behavior differs slightly from 3122other commands, in particular: 3123 3124- Pressing *?* or the *<tab>* key displays contextual help for the current 3125 token, not that of the entire command. 3126 3127- Optional and repeated parameters are supported (provided they are listed 3128 in the contextual help). 3129 3130The first parameter stands for the operation mode. Possible operations and 3131their general syntax are described below. They are covered in detail in the 3132following sections. 3133 3134- Check whether a flow rule can be created:: 3135 3136 flow validate {port_id} 3137 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3138 pattern {item} [/ {item} [...]] / end 3139 actions {action} [/ {action} [...]] / end 3140 3141- Create a flow rule:: 3142 3143 flow create {port_id} 3144 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3145 pattern {item} [/ {item} [...]] / end 3146 actions {action} [/ {action} [...]] / end 3147 3148- Destroy specific flow rules:: 3149 3150 flow destroy {port_id} rule {rule_id} [...] 3151 3152- Destroy all flow rules:: 3153 3154 flow flush {port_id} 3155 3156- Query an existing flow rule:: 3157 3158 flow query {port_id} {rule_id} {action} 3159 3160- List existing flow rules sorted by priority, filtered by group 3161 identifiers:: 3162 3163 flow list {port_id} [group {group_id}] [...] 3164 3165- Restrict ingress traffic to the defined flow rules:: 3166 3167 flow isolate {port_id} {boolean} 3168 3169Validating flow rules 3170~~~~~~~~~~~~~~~~~~~~~ 3171 3172``flow validate`` reports whether a flow rule would be accepted by the 3173underlying device in its current state but stops short of creating it. It is 3174bound to ``rte_flow_validate()``:: 3175 3176 flow validate {port_id} 3177 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3178 pattern {item} [/ {item} [...]] / end 3179 actions {action} [/ {action} [...]] / end 3180 3181If successful, it will show:: 3182 3183 Flow rule validated 3184 3185Otherwise it will show an error message of the form:: 3186 3187 Caught error type [...] ([...]): [...] 3188 3189This command uses the same parameters as ``flow create``, their format is 3190described in `Creating flow rules`_. 3191 3192Check whether redirecting any Ethernet packet received on port 0 to RX queue 3193index 6 is supported:: 3194 3195 testpmd> flow validate 0 ingress pattern eth / end 3196 actions queue index 6 / end 3197 Flow rule validated 3198 testpmd> 3199 3200Port 0 does not support TCPv6 rules:: 3201 3202 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end 3203 actions drop / end 3204 Caught error type 9 (specific pattern item): Invalid argument 3205 testpmd> 3206 3207Creating flow rules 3208~~~~~~~~~~~~~~~~~~~ 3209 3210``flow create`` validates and creates the specified flow rule. It is bound 3211to ``rte_flow_create()``:: 3212 3213 flow create {port_id} 3214 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3215 pattern {item} [/ {item} [...]] / end 3216 actions {action} [/ {action} [...]] / end 3217 3218If successful, it will return a flow rule ID usable with other commands:: 3219 3220 Flow rule #[...] created 3221 3222Otherwise it will show an error message of the form:: 3223 3224 Caught error type [...] ([...]): [...] 3225 3226Parameters describe in the following order: 3227 3228- Attributes (*group*, *priority*, *ingress*, *egress*, *transfer* tokens). 3229- A matching pattern, starting with the *pattern* token and terminated by an 3230 *end* pattern item. 3231- Actions, starting with the *actions* token and terminated by an *end* 3232 action. 3233 3234These translate directly to *rte_flow* objects provided as-is to the 3235underlying functions. 3236 3237The shortest valid definition only comprises mandatory tokens:: 3238 3239 testpmd> flow create 0 pattern end actions end 3240 3241Note that PMDs may refuse rules that essentially do nothing such as this 3242one. 3243 3244**All unspecified object values are automatically initialized to 0.** 3245 3246Attributes 3247^^^^^^^^^^ 3248 3249These tokens affect flow rule attributes (``struct rte_flow_attr``) and are 3250specified before the ``pattern`` token. 3251 3252- ``group {group id}``: priority group. 3253- ``priority {level}``: priority level within group. 3254- ``ingress``: rule applies to ingress traffic. 3255- ``egress``: rule applies to egress traffic. 3256- ``transfer``: apply rule directly to endpoints found in pattern. 3257 3258Each instance of an attribute specified several times overrides the previous 3259value as shown below (group 4 is used):: 3260 3261 testpmd> flow create 0 group 42 group 24 group 4 [...] 3262 3263Note that once enabled, ``ingress`` and ``egress`` cannot be disabled. 3264 3265While not specifying a direction is an error, some rules may allow both 3266simultaneously. 3267 3268Most rules affect RX therefore contain the ``ingress`` token:: 3269 3270 testpmd> flow create 0 ingress pattern [...] 3271 3272Matching pattern 3273^^^^^^^^^^^^^^^^ 3274 3275A matching pattern starts after the ``pattern`` token. It is made of pattern 3276items and is terminated by a mandatory ``end`` item. 3277 3278Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum 3279rte_flow_item_type``). 3280 3281The ``/`` token is used as a separator between pattern items as shown 3282below:: 3283 3284 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...] 3285 3286Note that protocol items like these must be stacked from lowest to highest 3287layer to make sense. For instance, the following rule is either invalid or 3288unlikely to match any packet:: 3289 3290 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...] 3291 3292More information on these restrictions can be found in the *rte_flow* 3293documentation. 3294 3295Several items support additional specification structures, for example 3296``ipv4`` allows specifying source and destination addresses as follows:: 3297 3298 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1 3299 dst is 10.2.0.0 / end [...] 3300 3301This rule matches all IPv4 traffic with the specified properties. 3302 3303In this example, ``src`` and ``dst`` are field names of the underlying 3304``struct rte_flow_item_ipv4`` object. All item properties can be specified 3305in a similar fashion. 3306 3307The ``is`` token means that the subsequent value must be matched exactly, 3308and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item`` 3309accordingly. Possible assignment tokens are: 3310 3311- ``is``: match value perfectly (with full bit-mask). 3312- ``spec``: match value according to configured bit-mask. 3313- ``last``: specify upper bound to establish a range. 3314- ``mask``: specify bit-mask with relevant bits set to one. 3315- ``prefix``: generate bit-mask from a prefix length. 3316 3317These yield identical results:: 3318 3319 ipv4 src is 10.1.1.1 3320 3321:: 3322 3323 ipv4 src spec 10.1.1.1 src mask 255.255.255.255 3324 3325:: 3326 3327 ipv4 src spec 10.1.1.1 src prefix 32 3328 3329:: 3330 3331 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value 3332 3333:: 3334 3335 ipv4 src is 10.1.1.1 src last 0 # 0 disables range 3336 3337Inclusive ranges can be defined with ``last``:: 3338 3339 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4 3340 3341Note that ``mask`` affects both ``spec`` and ``last``:: 3342 3343 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0 3344 # matches 10.1.0.0 to 10.2.255.255 3345 3346Properties can be modified multiple times:: 3347 3348 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4 3349 3350:: 3351 3352 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16 3353 3354Pattern items 3355^^^^^^^^^^^^^ 3356 3357This section lists supported pattern items and their attributes, if any. 3358 3359- ``end``: end list of pattern items. 3360 3361- ``void``: no-op pattern item. 3362 3363- ``invert``: perform actions when pattern does not match. 3364 3365- ``any``: match any protocol for the current layer. 3366 3367 - ``num {unsigned}``: number of layers covered. 3368 3369- ``pf``: match traffic from/to the physical function. 3370 3371- ``vf``: match traffic from/to a virtual function ID. 3372 3373 - ``id {unsigned}``: VF ID. 3374 3375- ``phy_port``: match traffic from/to a specific physical port. 3376 3377 - ``index {unsigned}``: physical port index. 3378 3379- ``port_id``: match traffic from/to a given DPDK port ID. 3380 3381 - ``id {unsigned}``: DPDK port ID. 3382 3383- ``mark``: match value set in previously matched flow rule using the mark action. 3384 3385 - ``id {unsigned}``: arbitrary integer value. 3386 3387- ``raw``: match an arbitrary byte string. 3388 3389 - ``relative {boolean}``: look for pattern after the previous item. 3390 - ``search {boolean}``: search pattern from offset (see also limit). 3391 - ``offset {integer}``: absolute or relative offset for pattern. 3392 - ``limit {unsigned}``: search area limit for start of pattern. 3393 - ``pattern {string}``: byte string to look for. 3394 3395- ``eth``: match Ethernet header. 3396 3397 - ``dst {MAC-48}``: destination MAC. 3398 - ``src {MAC-48}``: source MAC. 3399 - ``type {unsigned}``: EtherType or TPID. 3400 3401- ``vlan``: match 802.1Q/ad VLAN tag. 3402 3403 - ``tci {unsigned}``: tag control information. 3404 - ``pcp {unsigned}``: priority code point. 3405 - ``dei {unsigned}``: drop eligible indicator. 3406 - ``vid {unsigned}``: VLAN identifier. 3407 - ``inner_type {unsigned}``: inner EtherType or TPID. 3408 3409- ``ipv4``: match IPv4 header. 3410 3411 - ``tos {unsigned}``: type of service. 3412 - ``ttl {unsigned}``: time to live. 3413 - ``proto {unsigned}``: next protocol ID. 3414 - ``src {ipv4 address}``: source address. 3415 - ``dst {ipv4 address}``: destination address. 3416 3417- ``ipv6``: match IPv6 header. 3418 3419 - ``tc {unsigned}``: traffic class. 3420 - ``flow {unsigned}``: flow label. 3421 - ``proto {unsigned}``: protocol (next header). 3422 - ``hop {unsigned}``: hop limit. 3423 - ``src {ipv6 address}``: source address. 3424 - ``dst {ipv6 address}``: destination address. 3425 3426- ``icmp``: match ICMP header. 3427 3428 - ``type {unsigned}``: ICMP packet type. 3429 - ``code {unsigned}``: ICMP packet code. 3430 3431- ``udp``: match UDP header. 3432 3433 - ``src {unsigned}``: UDP source port. 3434 - ``dst {unsigned}``: UDP destination port. 3435 3436- ``tcp``: match TCP header. 3437 3438 - ``src {unsigned}``: TCP source port. 3439 - ``dst {unsigned}``: TCP destination port. 3440 3441- ``sctp``: match SCTP header. 3442 3443 - ``src {unsigned}``: SCTP source port. 3444 - ``dst {unsigned}``: SCTP destination port. 3445 - ``tag {unsigned}``: validation tag. 3446 - ``cksum {unsigned}``: checksum. 3447 3448- ``vxlan``: match VXLAN header. 3449 3450 - ``vni {unsigned}``: VXLAN identifier. 3451 3452- ``e_tag``: match IEEE 802.1BR E-Tag header. 3453 3454 - ``grp_ecid_b {unsigned}``: GRP and E-CID base. 3455 3456- ``nvgre``: match NVGRE header. 3457 3458 - ``tni {unsigned}``: virtual subnet ID. 3459 3460- ``mpls``: match MPLS header. 3461 3462 - ``label {unsigned}``: MPLS label. 3463 3464- ``gre``: match GRE header. 3465 3466 - ``protocol {unsigned}``: protocol type. 3467 3468- ``fuzzy``: fuzzy pattern match, expect faster than default. 3469 3470 - ``thresh {unsigned}``: accuracy threshold. 3471 3472- ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header. 3473 3474 - ``teid {unsigned}``: tunnel endpoint identifier. 3475 3476- ``geneve``: match GENEVE header. 3477 3478 - ``vni {unsigned}``: virtual network identifier. 3479 - ``protocol {unsigned}``: protocol type. 3480 3481- ``vxlan-gpe``: match VXLAN-GPE header. 3482 3483 - ``vni {unsigned}``: VXLAN-GPE identifier. 3484 3485- ``arp_eth_ipv4``: match ARP header for Ethernet/IPv4. 3486 3487 - ``sha {MAC-48}``: sender hardware address. 3488 - ``spa {ipv4 address}``: sender IPv4 address. 3489 - ``tha {MAC-48}``: target hardware address. 3490 - ``tpa {ipv4 address}``: target IPv4 address. 3491 3492- ``ipv6_ext``: match presence of any IPv6 extension header. 3493 3494 - ``next_hdr {unsigned}``: next header. 3495 3496- ``icmp6``: match any ICMPv6 header. 3497 3498 - ``type {unsigned}``: ICMPv6 type. 3499 - ``code {unsigned}``: ICMPv6 code. 3500 3501- ``icmp6_nd_ns``: match ICMPv6 neighbor discovery solicitation. 3502 3503 - ``target_addr {ipv6 address}``: target address. 3504 3505- ``icmp6_nd_na``: match ICMPv6 neighbor discovery advertisement. 3506 3507 - ``target_addr {ipv6 address}``: target address. 3508 3509- ``icmp6_nd_opt``: match presence of any ICMPv6 neighbor discovery option. 3510 3511 - ``type {unsigned}``: ND option type. 3512 3513- ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery source Ethernet 3514 link-layer address option. 3515 3516 - ``sla {MAC-48}``: source Ethernet LLA. 3517 3518- ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery target Ethernet 3519 link-layer address option. 3520 3521 - ``tla {MAC-48}``: target Ethernet LLA. 3522 3523Actions list 3524^^^^^^^^^^^^ 3525 3526A list of actions starts after the ``actions`` token in the same fashion as 3527`Matching pattern`_; actions are separated by ``/`` tokens and the list is 3528terminated by a mandatory ``end`` action. 3529 3530Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum 3531rte_flow_action_type``). 3532 3533Dropping all incoming UDPv4 packets can be expressed as follows:: 3534 3535 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3536 actions drop / end 3537 3538Several actions have configurable properties which must be specified when 3539there is no valid default value. For example, ``queue`` requires a target 3540queue index. 3541 3542This rule redirects incoming UDPv4 traffic to queue index 6:: 3543 3544 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3545 actions queue index 6 / end 3546 3547While this one could be rejected by PMDs (unspecified queue index):: 3548 3549 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3550 actions queue / end 3551 3552As defined by *rte_flow*, the list is not ordered, all actions of a given 3553rule are performed simultaneously. These are equivalent:: 3554 3555 queue index 6 / void / mark id 42 / end 3556 3557:: 3558 3559 void / mark id 42 / queue index 6 / end 3560 3561All actions in a list should have different types, otherwise only the last 3562action of a given type is taken into account:: 3563 3564 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6 3565 3566:: 3567 3568 drop / drop / drop / end # drop is performed only once 3569 3570:: 3571 3572 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24 3573 3574Considering they are performed simultaneously, opposite and overlapping 3575actions can sometimes be combined when the end result is unambiguous:: 3576 3577 drop / queue index 6 / end # drop has no effect 3578 3579:: 3580 3581 queue index 6 / rss queues 6 7 8 / end # queue has no effect 3582 3583:: 3584 3585 drop / passthru / end # drop has no effect 3586 3587Note that PMDs may still refuse such combinations. 3588 3589Actions 3590^^^^^^^ 3591 3592This section lists supported actions and their attributes, if any. 3593 3594- ``end``: end list of actions. 3595 3596- ``void``: no-op action. 3597 3598- ``passthru``: let subsequent rule process matched packets. 3599 3600- ``jump``: redirect traffic to group on device. 3601 3602 - ``group {unsigned}``: group to redirect to. 3603 3604- ``mark``: attach 32 bit value to packets. 3605 3606 - ``id {unsigned}``: 32 bit value to return with packets. 3607 3608- ``flag``: flag packets. 3609 3610- ``queue``: assign packets to a given queue index. 3611 3612 - ``index {unsigned}``: queue index to use. 3613 3614- ``drop``: drop packets (note: passthru has priority). 3615 3616- ``count``: enable counters for this rule. 3617 3618- ``rss``: spread packets among several queues. 3619 3620 - ``func {hash function}``: RSS hash function to apply, allowed tokens are 3621 the same as `set_hash_global_config`_. 3622 3623 - ``level {unsigned}``: encapsulation level for ``types``. 3624 3625 - ``types [{RSS hash type} [...]] end``: specific RSS hash types, allowed 3626 tokens are the same as `set_hash_input_set`_, except that an empty list 3627 does not disable RSS but instead requests unspecified "best-effort" 3628 settings. 3629 3630 - ``key {string}``: RSS hash key, overrides ``key_len``. 3631 3632 - ``key_len {unsigned}``: RSS hash key length in bytes, can be used in 3633 conjunction with ``key`` to pad or truncate it. 3634 3635 - ``queues [{unsigned} [...]] end``: queue indices to use. 3636 3637- ``pf``: direct traffic to physical function. 3638 3639- ``vf``: direct traffic to a virtual function ID. 3640 3641 - ``original {boolean}``: use original VF ID if possible. 3642 - ``id {unsigned}``: VF ID. 3643 3644- ``phy_port``: direct packets to physical port index. 3645 3646 - ``original {boolean}``: use original port index if possible. 3647 - ``index {unsigned}``: physical port index. 3648 3649- ``port_id``: direct matching traffic to a given DPDK port ID. 3650 3651 - ``original {boolean}``: use original DPDK port ID if possible. 3652 - ``id {unsigned}``: DPDK port ID. 3653 3654- ``of_set_mpls_ttl``: OpenFlow's ``OFPAT_SET_MPLS_TTL``. 3655 3656 - ``mpls_ttl``: MPLS TTL. 3657 3658- ``of_dec_mpls_ttl``: OpenFlow's ``OFPAT_DEC_MPLS_TTL``. 3659 3660- ``of_set_nw_ttl``: OpenFlow's ``OFPAT_SET_NW_TTL``. 3661 3662 - ``nw_ttl``: IP TTL. 3663 3664- ``of_dec_nw_ttl``: OpenFlow's ``OFPAT_DEC_NW_TTL``. 3665 3666- ``of_copy_ttl_out``: OpenFlow's ``OFPAT_COPY_TTL_OUT``. 3667 3668- ``of_copy_ttl_in``: OpenFlow's ``OFPAT_COPY_TTL_IN``. 3669 3670- ``of_pop_vlan``: OpenFlow's ``OFPAT_POP_VLAN``. 3671 3672- ``of_push_vlan``: OpenFlow's ``OFPAT_PUSH_VLAN``. 3673 3674 - ``ethertype``: Ethertype. 3675 3676- ``of_set_vlan_vid``: OpenFlow's ``OFPAT_SET_VLAN_VID``. 3677 3678 - ``vlan_vid``: VLAN id. 3679 3680- ``of_set_vlan_pcp``: OpenFlow's ``OFPAT_SET_VLAN_PCP``. 3681 3682 - ``vlan_pcp``: VLAN priority. 3683 3684- ``of_pop_mpls``: OpenFlow's ``OFPAT_POP_MPLS``. 3685 3686 - ``ethertype``: Ethertype. 3687 3688- ``of_push_mpls``: OpenFlow's ``OFPAT_PUSH_MPLS``. 3689 3690 - ``ethertype``: Ethertype. 3691 3692- ``vxlan_encap``: Performs a VXLAN encapsulation, outer layer configuration 3693 is done through `Config VXLAN Encap outer layers`_. 3694 3695- ``vxlan_decap``: Performs a decapsulation action by stripping all headers of 3696 the VXLAN tunnel network overlay from the matched flow. 3697 3698- ``nvgre_encap``: Performs a NVGRE encapsulation, outer layer configuration 3699 is done through `Config NVGRE Encap outer layers`_. 3700 3701- ``nvgre_decap``: Performs a decapsulation action by stripping all headers of 3702 the NVGRE tunnel network overlay from the matched flow. 3703 3704Destroying flow rules 3705~~~~~~~~~~~~~~~~~~~~~ 3706 3707``flow destroy`` destroys one or more rules from their rule ID (as returned 3708by ``flow create``), this command calls ``rte_flow_destroy()`` as many 3709times as necessary:: 3710 3711 flow destroy {port_id} rule {rule_id} [...] 3712 3713If successful, it will show:: 3714 3715 Flow rule #[...] destroyed 3716 3717It does not report anything for rule IDs that do not exist. The usual error 3718message is shown when a rule cannot be destroyed:: 3719 3720 Caught error type [...] ([...]): [...] 3721 3722``flow flush`` destroys all rules on a device and does not take extra 3723arguments. It is bound to ``rte_flow_flush()``:: 3724 3725 flow flush {port_id} 3726 3727Any errors are reported as above. 3728 3729Creating several rules and destroying them:: 3730 3731 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3732 actions queue index 2 / end 3733 Flow rule #0 created 3734 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3735 actions queue index 3 / end 3736 Flow rule #1 created 3737 testpmd> flow destroy 0 rule 0 rule 1 3738 Flow rule #1 destroyed 3739 Flow rule #0 destroyed 3740 testpmd> 3741 3742The same result can be achieved using ``flow flush``:: 3743 3744 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3745 actions queue index 2 / end 3746 Flow rule #0 created 3747 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3748 actions queue index 3 / end 3749 Flow rule #1 created 3750 testpmd> flow flush 0 3751 testpmd> 3752 3753Non-existent rule IDs are ignored:: 3754 3755 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3756 actions queue index 2 / end 3757 Flow rule #0 created 3758 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3759 actions queue index 3 / end 3760 Flow rule #1 created 3761 testpmd> flow destroy 0 rule 42 rule 10 rule 2 3762 testpmd> 3763 testpmd> flow destroy 0 rule 0 3764 Flow rule #0 destroyed 3765 testpmd> 3766 3767Querying flow rules 3768~~~~~~~~~~~~~~~~~~~ 3769 3770``flow query`` queries a specific action of a flow rule having that 3771ability. Such actions collect information that can be reported using this 3772command. It is bound to ``rte_flow_query()``:: 3773 3774 flow query {port_id} {rule_id} {action} 3775 3776If successful, it will display either the retrieved data for known actions 3777or the following message:: 3778 3779 Cannot display result for action type [...] ([...]) 3780 3781Otherwise, it will complain either that the rule does not exist or that some 3782error occurred:: 3783 3784 Flow rule #[...] not found 3785 3786:: 3787 3788 Caught error type [...] ([...]): [...] 3789 3790Currently only the ``count`` action is supported. This action reports the 3791number of packets that hit the flow rule and the total number of bytes. Its 3792output has the following format:: 3793 3794 count: 3795 hits_set: [...] # whether "hits" contains a valid value 3796 bytes_set: [...] # whether "bytes" contains a valid value 3797 hits: [...] # number of packets 3798 bytes: [...] # number of bytes 3799 3800Querying counters for TCPv6 packets redirected to queue 6:: 3801 3802 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end 3803 actions queue index 6 / count / end 3804 Flow rule #4 created 3805 testpmd> flow query 0 4 count 3806 count: 3807 hits_set: 1 3808 bytes_set: 0 3809 hits: 386446 3810 bytes: 0 3811 testpmd> 3812 3813Listing flow rules 3814~~~~~~~~~~~~~~~~~~ 3815 3816``flow list`` lists existing flow rules sorted by priority and optionally 3817filtered by group identifiers:: 3818 3819 flow list {port_id} [group {group_id}] [...] 3820 3821This command only fails with the following message if the device does not 3822exist:: 3823 3824 Invalid port [...] 3825 3826Output consists of a header line followed by a short description of each 3827flow rule, one per line. There is no output at all when no flow rules are 3828configured on the device:: 3829 3830 ID Group Prio Attr Rule 3831 [...] [...] [...] [...] [...] 3832 3833``Attr`` column flags: 3834 3835- ``i`` for ``ingress``. 3836- ``e`` for ``egress``. 3837 3838Creating several flow rules and listing them:: 3839 3840 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3841 actions queue index 6 / end 3842 Flow rule #0 created 3843 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3844 actions queue index 2 / end 3845 Flow rule #1 created 3846 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end 3847 actions rss queues 6 7 8 end / end 3848 Flow rule #2 created 3849 testpmd> flow list 0 3850 ID Group Prio Attr Rule 3851 0 0 0 i- ETH IPV4 => QUEUE 3852 1 0 0 i- ETH IPV6 => QUEUE 3853 2 0 5 i- ETH IPV4 UDP => RSS 3854 testpmd> 3855 3856Rules are sorted by priority (i.e. group ID first, then priority level):: 3857 3858 testpmd> flow list 1 3859 ID Group Prio Attr Rule 3860 0 0 0 i- ETH => COUNT 3861 6 0 500 i- ETH IPV6 TCP => DROP COUNT 3862 5 0 1000 i- ETH IPV6 ICMP => QUEUE 3863 1 24 0 i- ETH IPV4 UDP => QUEUE 3864 4 24 10 i- ETH IPV4 TCP => DROP 3865 3 24 20 i- ETH IPV4 => DROP 3866 2 24 42 i- ETH IPV4 UDP => QUEUE 3867 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 3868 testpmd> 3869 3870Output can be limited to specific groups:: 3871 3872 testpmd> flow list 1 group 0 group 63 3873 ID Group Prio Attr Rule 3874 0 0 0 i- ETH => COUNT 3875 6 0 500 i- ETH IPV6 TCP => DROP COUNT 3876 5 0 1000 i- ETH IPV6 ICMP => QUEUE 3877 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 3878 testpmd> 3879 3880Toggling isolated mode 3881~~~~~~~~~~~~~~~~~~~~~~ 3882 3883``flow isolate`` can be used to tell the underlying PMD that ingress traffic 3884must only be injected from the defined flow rules; that no default traffic 3885is expected outside those rules and the driver is free to assign more 3886resources to handle them. It is bound to ``rte_flow_isolate()``:: 3887 3888 flow isolate {port_id} {boolean} 3889 3890If successful, enabling or disabling isolated mode shows either:: 3891 3892 Ingress traffic on port [...] 3893 is now restricted to the defined flow rules 3894 3895Or:: 3896 3897 Ingress traffic on port [...] 3898 is not restricted anymore to the defined flow rules 3899 3900Otherwise, in case of error:: 3901 3902 Caught error type [...] ([...]): [...] 3903 3904Mainly due to its side effects, PMDs supporting this mode may not have the 3905ability to toggle it more than once without reinitializing affected ports 3906first (e.g. by exiting testpmd). 3907 3908Enabling isolated mode:: 3909 3910 testpmd> flow isolate 0 true 3911 Ingress traffic on port 0 is now restricted to the defined flow rules 3912 testpmd> 3913 3914Disabling isolated mode:: 3915 3916 testpmd> flow isolate 0 false 3917 Ingress traffic on port 0 is not restricted anymore to the defined flow rules 3918 testpmd> 3919 3920Sample QinQ flow rules 3921~~~~~~~~~~~~~~~~~~~~~~ 3922 3923Before creating QinQ rule(s) the following commands should be issued to enable QinQ:: 3924 3925 testpmd> port stop 0 3926 testpmd> vlan set qinq on 0 3927 3928The above command sets the inner and outer TPID's to 0x8100. 3929 3930To change the TPID's the following commands should be used:: 3931 3932 testpmd> vlan set outer tpid 0xa100 0 3933 testpmd> vlan set inner tpid 0x9100 0 3934 testpmd> port start 0 3935 3936Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM. 3937 3938:: 3939 3940 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 / 3941 vlan tci is 456 / end actions vf id 1 / queue index 0 / end 3942 Flow rule #0 validated 3943 3944 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 / 3945 vlan tci is 456 / end actions vf id 123 / queue index 0 / end 3946 Flow rule #0 created 3947 3948 testpmd> flow list 0 3949 ID Group Prio Attr Rule 3950 0 0 0 i- ETH VLAN VLAN=>VF QUEUE 3951 3952Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host. 3953 3954:: 3955 3956 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 / 3957 vlan tci is 654 / end actions pf / queue index 0 / end 3958 Flow rule #1 validated 3959 3960 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 / 3961 vlan tci is 654 / end actions pf / queue index 1 / end 3962 Flow rule #1 created 3963 3964 testpmd> flow list 0 3965 ID Group Prio Attr Rule 3966 0 0 0 i- ETH VLAN VLAN=>VF QUEUE 3967 1 0 0 i- ETH VLAN VLAN=>PF QUEUE 3968 3969Sample VXLAN encapsulation rule 3970~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 3971 3972VXLAN encapsulation outer layer has default value pre-configured in testpmd 3973source code, those can be changed by using the following commands 3974 3975IPv4 VXLAN outer header:: 3976 3977 testpmd> set vxlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src 127.0.0.1 3978 ip-dst 128.0.0.1 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 3979 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 3980 queue index 0 / end 3981 3982 testpmd> set vxlan-with-vlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src 3983 127.0.0.1 ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11 3984 eth-dst 22:22:22:22:22:22 3985 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 3986 queue index 0 / end 3987 3988IPv6 VXLAN outer header:: 3989 3990 testpmd> set vxlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4 ip-src ::1 3991 ip-dst ::2222 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 3992 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 3993 queue index 0 / end 3994 3995 testpmd> set vxlan-with-vlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4 3996 ip-src ::1 ip-dst ::2222 vlan-tci 34 eth-src 11:11:11:11:11:11 3997 eth-dst 22:22:22:22:22:22 3998 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 3999 queue index 0 / end 4000 4001Sample NVGRE encapsulation rule 4002~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4003 4004NVGRE encapsulation outer layer has default value pre-configured in testpmd 4005source code, those can be changed by using the following commands 4006 4007IPv4 NVGRE outer header:: 4008 4009 testpmd> set nvgre ip-version ipv4 tni 4 ip-src 127.0.0.1 ip-dst 128.0.0.1 4010 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4011 testpmd> flow create 0 ingress pattern end actions nvgre_encap / 4012 queue index 0 / end 4013 4014 testpmd> set nvgre-with-vlan ip-version ipv4 tni 4 ip-src 127.0.0.1 4015 ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11 4016 eth-dst 22:22:22:22:22:22 4017 testpmd> flow create 0 ingress pattern end actions nvgre_encap / 4018 queue index 0 / end 4019 4020IPv6 NVGRE outer header:: 4021 4022 testpmd> set nvgre ip-version ipv6 tni 4 ip-src ::1 ip-dst ::2222 4023 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4024 testpmd> flow create 0 ingress pattern end actions nvgre_encap / 4025 queue index 0 / end 4026 4027 testpmd> set nvgre-with-vlan ip-version ipv6 tni 4 ip-src ::1 ip-dst ::2222 4028 vlan-tci 34 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4029 testpmd> flow create 0 ingress pattern end actions nvgre_encap / 4030 queue index 0 / end 4031 4032BPF Functions 4033-------------- 4034 4035The following sections show functions to load/unload eBPF based filters. 4036 4037bpf-load 4038~~~~~~~~ 4039 4040Load an eBPF program as a callback for partciular RX/TX queue:: 4041 4042 testpmd> bpf-load rx|tx (portid) (queueid) (load-flags) (bpf-prog-filename) 4043 4044The available load-flags are: 4045 4046* ``J``: use JIT generated native code, otherwise BPF interpreter will be used. 4047 4048* ``M``: assume input parameter is a pointer to rte_mbuf, otherwise assume it is a pointer to first segment's data. 4049 4050* ``-``: none. 4051 4052.. note:: 4053 4054 You'll need clang v3.7 or above to build bpf program you'd like to load 4055 4056For example: 4057 4058.. code-block:: console 4059 4060 cd test/bpf 4061 clang -O2 -target bpf -c t1.c 4062 4063Then to load (and JIT compile) t1.o at RX queue 0, port 1:: 4064 4065.. code-block:: console 4066 4067 testpmd> bpf-load rx 1 0 J ./dpdk.org/test/bpf/t1.o 4068 4069To load (not JITed) t1.o at TX queue 0, port 0:: 4070 4071.. code-block:: console 4072 4073 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o 4074 4075bpf-unload 4076~~~~~~~~~~ 4077 4078Unload previously loaded eBPF program for partciular RX/TX queue:: 4079 4080 testpmd> bpf-unload rx|tx (portid) (queueid) 4081 4082For example to unload BPF filter from TX queue 0, port 0: 4083 4084.. code-block:: console 4085 4086 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o 4087