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 1562Port Functions 1563-------------- 1564 1565The following sections show functions for configuring ports. 1566 1567.. note:: 1568 1569 Port configuration changes only become active when forwarding is started/restarted. 1570 1571port attach 1572~~~~~~~~~~~ 1573 1574Attach a port specified by pci address or virtual device args:: 1575 1576 testpmd> port attach (identifier) 1577 1578To attach a new pci device, the device should be recognized by kernel first. 1579Then it should be moved under DPDK management. 1580Finally the port can be attached to testpmd. 1581 1582For example, to move a pci device using ixgbe under DPDK management: 1583 1584.. code-block:: console 1585 1586 # Check the status of the available devices. 1587 ./usertools/dpdk-devbind.py --status 1588 1589 Network devices using DPDK-compatible driver 1590 ============================================ 1591 <none> 1592 1593 Network devices using kernel driver 1594 =================================== 1595 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused= 1596 1597 1598 # Bind the device to igb_uio. 1599 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0 1600 1601 1602 # Recheck the status of the devices. 1603 ./usertools/dpdk-devbind.py --status 1604 Network devices using DPDK-compatible driver 1605 ============================================ 1606 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused= 1607 1608To attach a port created by virtual device, above steps are not needed. 1609 1610For example, to attach a port whose pci address is 0000:0a:00.0. 1611 1612.. code-block:: console 1613 1614 testpmd> port attach 0000:0a:00.0 1615 Attaching a new port... 1616 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1617 EAL: probe driver: 8086:10fb rte_ixgbe_pmd 1618 EAL: PCI memory mapped at 0x7f83bfa00000 1619 EAL: PCI memory mapped at 0x7f83bfa80000 1620 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5 1621 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb 1622 Port 0 is attached. Now total ports is 1 1623 Done 1624 1625For example, to attach a port created by pcap PMD. 1626 1627.. code-block:: console 1628 1629 testpmd> port attach net_pcap0 1630 Attaching a new port... 1631 PMD: Initializing pmd_pcap for net_pcap0 1632 PMD: Creating pcap-backed ethdev on numa socket 0 1633 Port 0 is attached. Now total ports is 1 1634 Done 1635 1636In this case, identifier is ``net_pcap0``. 1637This identifier format is the same as ``--vdev`` format of DPDK applications. 1638 1639For example, to re-attach a bonded port which has been previously detached, 1640the mode and slave parameters must be given. 1641 1642.. code-block:: console 1643 1644 testpmd> port attach net_bond_0,mode=0,slave=1 1645 Attaching a new port... 1646 EAL: Initializing pmd_bond for net_bond_0 1647 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0. 1648 Port 0 is attached. Now total ports is 1 1649 Done 1650 1651 1652port detach 1653~~~~~~~~~~~ 1654 1655Detach a specific port:: 1656 1657 testpmd> port detach (port_id) 1658 1659Before detaching a port, the port should be stopped and closed. 1660 1661For example, to detach a pci device port 0. 1662 1663.. code-block:: console 1664 1665 testpmd> port stop 0 1666 Stopping ports... 1667 Done 1668 testpmd> port close 0 1669 Closing ports... 1670 Done 1671 1672 testpmd> port detach 0 1673 Detaching a port... 1674 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1675 EAL: remove driver: 8086:10fb rte_ixgbe_pmd 1676 EAL: PCI memory unmapped at 0x7f83bfa00000 1677 EAL: PCI memory unmapped at 0x7f83bfa80000 1678 Done 1679 1680 1681For example, to detach a virtual device port 0. 1682 1683.. code-block:: console 1684 1685 testpmd> port stop 0 1686 Stopping ports... 1687 Done 1688 testpmd> port close 0 1689 Closing ports... 1690 Done 1691 1692 testpmd> port detach 0 1693 Detaching a port... 1694 PMD: Closing pcap ethdev on numa socket 0 1695 Port 'net_pcap0' is detached. Now total ports is 0 1696 Done 1697 1698To remove a pci device completely from the system, first detach the port from testpmd. 1699Then the device should be moved under kernel management. 1700Finally the device can be removed using kernel pci hotplug functionality. 1701 1702For example, to move a pci device under kernel management: 1703 1704.. code-block:: console 1705 1706 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0 1707 1708 ./usertools/dpdk-devbind.py --status 1709 1710 Network devices using DPDK-compatible driver 1711 ============================================ 1712 <none> 1713 1714 Network devices using kernel driver 1715 =================================== 1716 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio 1717 1718To remove a port created by a virtual device, above steps are not needed. 1719 1720port start 1721~~~~~~~~~~ 1722 1723Start all ports or a specific port:: 1724 1725 testpmd> port start (port_id|all) 1726 1727port stop 1728~~~~~~~~~ 1729 1730Stop all ports or a specific port:: 1731 1732 testpmd> port stop (port_id|all) 1733 1734port close 1735~~~~~~~~~~ 1736 1737Close all ports or a specific port:: 1738 1739 testpmd> port close (port_id|all) 1740 1741port config - queue ring size 1742~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1743 1744Configure a rx/tx queue ring size:: 1745 1746 testpmd> port (port_id) (rxq|txq) (queue_id) ring_size (value) 1747 1748Only take effect after command that (re-)start the port or command that setup specific queue. 1749 1750port start/stop queue 1751~~~~~~~~~~~~~~~~~~~~~ 1752 1753Start/stop a rx/tx queue on a specific port:: 1754 1755 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop) 1756 1757port setup queue 1758~~~~~~~~~~~~~~~~~~~~~ 1759 1760Setup a rx/tx queue on a specific port:: 1761 1762 testpmd> port (port_id) (rxq|txq) (queue_id) setup 1763 1764Only take effect when port is started. 1765 1766port config - speed 1767~~~~~~~~~~~~~~~~~~~ 1768 1769Set the speed and duplex mode for all ports or a specific port:: 1770 1771 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \ 1772 duplex (half|full|auto) 1773 1774port config - queues/descriptors 1775~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1776 1777Set number of queues/descriptors for rxq, txq, rxd and txd:: 1778 1779 testpmd> port config all (rxq|txq|rxd|txd) (value) 1780 1781This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options. 1782 1783port config - max-pkt-len 1784~~~~~~~~~~~~~~~~~~~~~~~~~ 1785 1786Set the maximum packet length:: 1787 1788 testpmd> port config all max-pkt-len (value) 1789 1790This is equivalent to the ``--max-pkt-len`` command-line option. 1791 1792port config - CRC Strip 1793~~~~~~~~~~~~~~~~~~~~~~~ 1794 1795Set hardware CRC stripping on or off for all ports:: 1796 1797 testpmd> port config all crc-strip (on|off) 1798 1799CRC stripping is on by default. 1800 1801The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option. 1802 1803port config - scatter 1804~~~~~~~~~~~~~~~~~~~~~~~ 1805 1806Set RX scatter mode on or off for all ports:: 1807 1808 testpmd> port config all scatter (on|off) 1809 1810RX scatter mode is off by default. 1811 1812The ``on`` option is equivalent to the ``--enable-scatter`` command-line option. 1813 1814port config - RX Checksum 1815~~~~~~~~~~~~~~~~~~~~~~~~~ 1816 1817Set hardware RX checksum offload to on or off for all ports:: 1818 1819 testpmd> port config all rx-cksum (on|off) 1820 1821Checksum offload is off by default. 1822 1823The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option. 1824 1825port config - VLAN 1826~~~~~~~~~~~~~~~~~~ 1827 1828Set hardware VLAN on or off for all ports:: 1829 1830 testpmd> port config all hw-vlan (on|off) 1831 1832Hardware VLAN is off by default. 1833 1834The ``on`` option is equivalent to the ``--enable-hw-vlan`` command-line option. 1835 1836port config - VLAN filter 1837~~~~~~~~~~~~~~~~~~~~~~~~~ 1838 1839Set hardware VLAN filter on or off for all ports:: 1840 1841 testpmd> port config all hw-vlan-filter (on|off) 1842 1843Hardware VLAN filter is off by default. 1844 1845The ``on`` option is equivalent to the ``--enable-hw-vlan-filter`` command-line option. 1846 1847port config - VLAN strip 1848~~~~~~~~~~~~~~~~~~~~~~~~ 1849 1850Set hardware VLAN strip on or off for all ports:: 1851 1852 testpmd> port config all hw-vlan-strip (on|off) 1853 1854Hardware VLAN strip is off by default. 1855 1856The ``on`` option is equivalent to the ``--enable-hw-vlan-strip`` command-line option. 1857 1858port config - VLAN extend 1859~~~~~~~~~~~~~~~~~~~~~~~~~ 1860 1861Set hardware VLAN extend on or off for all ports:: 1862 1863 testpmd> port config all hw-vlan-extend (on|off) 1864 1865Hardware VLAN extend is off by default. 1866 1867The ``on`` option is equivalent to the ``--enable-hw-vlan-extend`` command-line option. 1868 1869port config - Drop Packets 1870~~~~~~~~~~~~~~~~~~~~~~~~~~ 1871 1872Set packet drop for packets with no descriptors on or off for all ports:: 1873 1874 testpmd> port config all drop-en (on|off) 1875 1876Packet dropping for packets with no descriptors is off by default. 1877 1878The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option. 1879 1880port config - RSS 1881~~~~~~~~~~~~~~~~~ 1882 1883Set the RSS (Receive Side Scaling) mode on or off:: 1884 1885 testpmd> port config all rss (all|default|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none) 1886 1887RSS is on by default. 1888 1889The ``all`` option is equivalent to ip|tcp|udp|sctp|ether. 1890The ``default`` option enables all supported RSS types reported by device info. 1891The ``none`` option is equivalent to the ``--disable-rss`` command-line option. 1892 1893port config - RSS Reta 1894~~~~~~~~~~~~~~~~~~~~~~ 1895 1896Set the RSS (Receive Side Scaling) redirection table:: 1897 1898 testpmd> port config all rss reta (hash,queue)[,(hash,queue)] 1899 1900port config - DCB 1901~~~~~~~~~~~~~~~~~ 1902 1903Set the DCB mode for an individual port:: 1904 1905 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off) 1906 1907The traffic class should be 4 or 8. 1908 1909port config - Burst 1910~~~~~~~~~~~~~~~~~~~ 1911 1912Set the number of packets per burst:: 1913 1914 testpmd> port config all burst (value) 1915 1916This is equivalent to the ``--burst`` command-line option. 1917 1918port config - Threshold 1919~~~~~~~~~~~~~~~~~~~~~~~ 1920 1921Set thresholds for TX/RX queues:: 1922 1923 testpmd> port config all (threshold) (value) 1924 1925Where the threshold type can be: 1926 1927* ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255. 1928 1929* ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255. 1930 1931* ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255. 1932 1933* ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255. 1934 1935* ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255. 1936 1937* ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255. 1938 1939* ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd. 1940 1941* ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd. 1942 1943* ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd. 1944 1945These threshold options are also available from the command-line. 1946 1947port config - E-tag 1948~~~~~~~~~~~~~~~~~~~ 1949 1950Set the value of ether-type for E-tag:: 1951 1952 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value) 1953 1954Enable/disable the E-tag support:: 1955 1956 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable) 1957 1958port config pctype mapping 1959~~~~~~~~~~~~~~~~~~~~~~~~~~ 1960 1961Reset pctype mapping table:: 1962 1963 testpmd> port config (port_id) pctype mapping reset 1964 1965Update hardware defined pctype to software defined flow type mapping table:: 1966 1967 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id) 1968 1969where: 1970 1971* ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table. 1972 1973* ``flow_type_id``: software flow type id as the index of the pctype mapping table. 1974 1975port config input set 1976~~~~~~~~~~~~~~~~~~~~~ 1977 1978Config RSS/FDIR/FDIR flexible payload input set for some pctype:: 1979 testpmd> port config (port_id) pctype (pctype_id) \ 1980 (hash_inset|fdir_inset|fdir_flx_inset) \ 1981 (get|set|clear) field (field_idx) 1982 1983Clear RSS/FDIR/FDIR flexible payload input set for some pctype:: 1984 testpmd> port config (port_id) pctype (pctype_id) \ 1985 (hash_inset|fdir_inset|fdir_flx_inset) clear all 1986 1987where: 1988 1989* ``pctype_id``: hardware packet classification types. 1990* ``field_idx``: hardware field index. 1991 1992port config udp_tunnel_port 1993~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1994 1995Add/remove UDP tunnel port for VXLAN/GENEVE tunneling protocols:: 1996 testpmd> port config (port_id) udp_tunnel_port add|rm vxlan|geneve (udp_port) 1997 1998Link Bonding Functions 1999---------------------- 2000 2001The Link Bonding functions make it possible to dynamically create and 2002manage link bonding devices from within testpmd interactive prompt. 2003 2004create bonded device 2005~~~~~~~~~~~~~~~~~~~~ 2006 2007Create a new bonding device:: 2008 2009 testpmd> create bonded device (mode) (socket) 2010 2011For example, to create a bonded device in mode 1 on socket 0:: 2012 2013 testpmd> create bonded 1 0 2014 created new bonded device (port X) 2015 2016add bonding slave 2017~~~~~~~~~~~~~~~~~ 2018 2019Adds Ethernet device to a Link Bonding device:: 2020 2021 testpmd> add bonding slave (slave id) (port id) 2022 2023For example, to add Ethernet device (port 6) to a Link Bonding device (port 10):: 2024 2025 testpmd> add bonding slave 6 10 2026 2027 2028remove bonding slave 2029~~~~~~~~~~~~~~~~~~~~ 2030 2031Removes an Ethernet slave device from a Link Bonding device:: 2032 2033 testpmd> remove bonding slave (slave id) (port id) 2034 2035For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10):: 2036 2037 testpmd> remove bonding slave 6 10 2038 2039set bonding mode 2040~~~~~~~~~~~~~~~~ 2041 2042Set the Link Bonding mode of a Link Bonding device:: 2043 2044 testpmd> set bonding mode (value) (port id) 2045 2046For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3):: 2047 2048 testpmd> set bonding mode 3 10 2049 2050set bonding primary 2051~~~~~~~~~~~~~~~~~~~ 2052 2053Set an Ethernet slave device as the primary device on a Link Bonding device:: 2054 2055 testpmd> set bonding primary (slave id) (port id) 2056 2057For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10):: 2058 2059 testpmd> set bonding primary 6 10 2060 2061set bonding mac 2062~~~~~~~~~~~~~~~ 2063 2064Set the MAC address of a Link Bonding device:: 2065 2066 testpmd> set bonding mac (port id) (mac) 2067 2068For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01:: 2069 2070 testpmd> set bonding mac 10 00:00:00:00:00:01 2071 2072set bonding xmit_balance_policy 2073~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2074 2075Set the transmission policy for a Link Bonding device when it is in Balance XOR mode:: 2076 2077 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34) 2078 2079For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports):: 2080 2081 testpmd> set bonding xmit_balance_policy 10 l34 2082 2083 2084set bonding mon_period 2085~~~~~~~~~~~~~~~~~~~~~~ 2086 2087Set the link status monitoring polling period in milliseconds for a bonding device. 2088 2089This adds support for PMD slave devices which do not support link status interrupts. 2090When the mon_period is set to a value greater than 0 then all PMD's which do not support 2091link status ISR will be queried every polling interval to check if their link status has changed:: 2092 2093 testpmd> set bonding mon_period (port_id) (value) 2094 2095For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms:: 2096 2097 testpmd> set bonding mon_period 5 150 2098 2099 2100set bonding lacp dedicated_queue 2101~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2102 2103Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic 2104when in mode 4 (link-aggregration-802.3ad):: 2105 2106 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable) 2107 2108 2109set bonding agg_mode 2110~~~~~~~~~~~~~~~~~~~~ 2111 2112Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad):: 2113 2114 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable) 2115 2116 2117show bonding config 2118~~~~~~~~~~~~~~~~~~~ 2119 2120Show the current configuration of a Link Bonding device:: 2121 2122 testpmd> show bonding config (port id) 2123 2124For example, 2125to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4) 2126in balance mode with a transmission policy of layer 2+3:: 2127 2128 testpmd> show bonding config 9 2129 Bonding mode: 2 2130 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23 2131 Slaves (3): [1 3 4] 2132 Active Slaves (3): [1 3 4] 2133 Primary: [3] 2134 2135 2136Register Functions 2137------------------ 2138 2139The Register Functions can be used to read from and write to registers on the network card referenced by a port number. 2140This is mainly useful for debugging purposes. 2141Reference should be made to the appropriate datasheet for the network card for details on the register addresses 2142and fields that can be accessed. 2143 2144read reg 2145~~~~~~~~ 2146 2147Display the value of a port register:: 2148 2149 testpmd> read reg (port_id) (address) 2150 2151For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller:: 2152 2153 testpmd> read reg 0 0xEE00 2154 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241) 2155 2156read regfield 2157~~~~~~~~~~~~~ 2158 2159Display a port register bit field:: 2160 2161 testpmd> read regfield (port_id) (address) (bit_x) (bit_y) 2162 2163For example, reading the lowest two bits from the register in the example above:: 2164 2165 testpmd> read regfield 0 0xEE00 0 1 2166 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1) 2167 2168read regbit 2169~~~~~~~~~~~ 2170 2171Display a single port register bit:: 2172 2173 testpmd> read regbit (port_id) (address) (bit_x) 2174 2175For example, reading the lowest bit from the register in the example above:: 2176 2177 testpmd> read regbit 0 0xEE00 0 2178 port 0 PCI register at offset 0xEE00: bit 0=1 2179 2180write reg 2181~~~~~~~~~ 2182 2183Set the value of a port register:: 2184 2185 testpmd> write reg (port_id) (address) (value) 2186 2187For example, to clear a register:: 2188 2189 testpmd> write reg 0 0xEE00 0x0 2190 port 0 PCI register at offset 0xEE00: 0x00000000 (0) 2191 2192write regfield 2193~~~~~~~~~~~~~~ 2194 2195Set bit field of a port register:: 2196 2197 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value) 2198 2199For example, writing to the register cleared in the example above:: 2200 2201 testpmd> write regfield 0 0xEE00 0 1 2 2202 port 0 PCI register at offset 0xEE00: 0x00000002 (2) 2203 2204write regbit 2205~~~~~~~~~~~~ 2206 2207Set single bit value of a port register:: 2208 2209 testpmd> write regbit (port_id) (address) (bit_x) (value) 2210 2211For example, to set the high bit in the register from the example above:: 2212 2213 testpmd> write regbit 0 0xEE00 31 1 2214 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658) 2215 2216Traffic Metering and Policing 2217----------------------------- 2218 2219The following section shows functions for configuring traffic metering and 2220policing on the ethernet device through the use of generic ethdev API. 2221 2222show port traffic management capability 2223~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2224 2225Show traffic metering and policing capability of the port:: 2226 2227 testpmd> show port meter cap (port_id) 2228 2229add port meter profile (srTCM rfc2967) 2230~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2231 2232Add meter profile (srTCM rfc2697) to the ethernet device:: 2233 2234 testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \ 2235 (cir) (cbs) (ebs) 2236 2237where: 2238 2239* ``profile_id``: ID for the meter profile. 2240* ``cir``: Committed Information Rate (CIR) (bytes/second). 2241* ``cbs``: Committed Burst Size (CBS) (bytes). 2242* ``ebs``: Excess Burst Size (EBS) (bytes). 2243 2244add port meter profile (trTCM rfc2968) 2245~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2246 2247Add meter profile (srTCM rfc2698) to the ethernet device:: 2248 2249 testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \ 2250 (cir) (pir) (cbs) (pbs) 2251 2252where: 2253 2254* ``profile_id``: ID for the meter profile. 2255* ``cir``: Committed information rate (bytes/second). 2256* ``pir``: Peak information rate (bytes/second). 2257* ``cbs``: Committed burst size (bytes). 2258* ``pbs``: Peak burst size (bytes). 2259 2260add port meter profile (trTCM rfc4115) 2261~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2262 2263Add meter profile (trTCM rfc4115) to the ethernet device:: 2264 2265 testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \ 2266 (cir) (eir) (cbs) (ebs) 2267 2268where: 2269 2270* ``profile_id``: ID for the meter profile. 2271* ``cir``: Committed information rate (bytes/second). 2272* ``eir``: Excess information rate (bytes/second). 2273* ``cbs``: Committed burst size (bytes). 2274* ``ebs``: Excess burst size (bytes). 2275 2276delete port meter profile 2277~~~~~~~~~~~~~~~~~~~~~~~~~ 2278 2279Delete meter profile from the ethernet device:: 2280 2281 testpmd> del port meter profile (port_id) (profile_id) 2282 2283create port meter 2284~~~~~~~~~~~~~~~~~ 2285 2286Create new meter object for the ethernet device:: 2287 2288 testpmd> create port meter (port_id) (mtr_id) (profile_id) \ 2289 (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \ 2290 (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\ 2291 (dscp_tbl_entry63)] 2292 2293where: 2294 2295* ``mtr_id``: meter object ID. 2296* ``profile_id``: ID for the meter profile. 2297* ``meter_enable``: When this parameter has a non-zero value, the meter object 2298 gets enabled at the time of creation, otherwise remains disabled. 2299* ``g_action``: Policer action for the packet with green color. 2300* ``y_action``: Policer action for the packet with yellow color. 2301* ``r_action``: Policer action for the packet with red color. 2302* ``stats_mask``: Mask of statistics counter types to be enabled for the 2303 meter object. 2304* ``shared``: When this parameter has a non-zero value, the meter object is 2305 shared by multiple flows. Otherwise, meter object is used by single flow. 2306* ``use_pre_meter_color``: When this parameter has a non-zero value, the 2307 input color for the current meter object is determined by the latest meter 2308 object in the same flow. Otherwise, the current meter object uses the 2309 *dscp_table* to determine the input color. 2310* ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input 2311 color, 0 <= x <= 63. 2312 2313enable port meter 2314~~~~~~~~~~~~~~~~~ 2315 2316Enable meter for the ethernet device:: 2317 2318 testpmd> enable port meter (port_id) (mtr_id) 2319 2320disable port meter 2321~~~~~~~~~~~~~~~~~~ 2322 2323Disable meter for the ethernet device:: 2324 2325 testpmd> disable port meter (port_id) (mtr_id) 2326 2327delete port meter 2328~~~~~~~~~~~~~~~~~ 2329 2330Delete meter for the ethernet device:: 2331 2332 testpmd> del port meter (port_id) (mtr_id) 2333 2334Set port meter profile 2335~~~~~~~~~~~~~~~~~~~~~~ 2336 2337Set meter profile for the ethernet device:: 2338 2339 testpmd> set port meter profile (port_id) (mtr_id) (profile_id) 2340 2341set port meter dscp table 2342~~~~~~~~~~~~~~~~~~~~~~~~~ 2343 2344Set meter dscp table for the ethernet device:: 2345 2346 testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \ 2347 (dscp_tbl_entry1)...(dscp_tbl_entry63)] 2348 2349set port meter policer action 2350~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2351 2352Set meter policer action for the ethernet device:: 2353 2354 testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \ 2355 (action0) [(action1) (action1)] 2356 2357where: 2358 2359* ``action_mask``: Bit mask indicating which policer actions need to be 2360 updated. One or more policer actions can be updated in a single function 2361 invocation. To update the policer action associated with color C, bit 2362 (1 << C) needs to be set in *action_mask* and element at position C 2363 in the *actions* array needs to be valid. 2364* ``actionx``: Policer action for the color x, 2365 RTE_MTR_GREEN <= x < RTE_MTR_COLORS 2366 2367set port meter stats mask 2368~~~~~~~~~~~~~~~~~~~~~~~~~ 2369 2370Set meter stats mask for the ethernet device:: 2371 2372 testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask) 2373 2374where: 2375 2376* ``stats_mask``: Bit mask indicating statistics counter types to be enabled. 2377 2378show port meter stats 2379~~~~~~~~~~~~~~~~~~~~~ 2380 2381Show meter stats of the ethernet device:: 2382 2383 testpmd> show port meter stats (port_id) (mtr_id) (clear) 2384 2385where: 2386 2387* ``clear``: Flag that indicates whether the statistics counters should 2388 be cleared (i.e. set to zero) immediately after they have been read or not. 2389 2390Traffic Management 2391------------------ 2392 2393The following section shows functions for configuring traffic management on 2394on the ethernet device through the use of generic TM API. 2395 2396show port traffic management capability 2397~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2398 2399Show traffic management capability of the port:: 2400 2401 testpmd> show port tm cap (port_id) 2402 2403show port traffic management capability (hierarchy level) 2404~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2405 2406Show traffic management hierarchy level capability of the port:: 2407 2408 testpmd> show port tm level cap (port_id) (level_id) 2409 2410show port traffic management capability (hierarchy node level) 2411~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2412 2413Show the traffic management hierarchy node capability of the port:: 2414 2415 testpmd> show port tm node cap (port_id) (node_id) 2416 2417show port traffic management hierarchy node type 2418~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2419 2420Show the port traffic management hierarchy node type:: 2421 2422 testpmd> show port tm node type (port_id) (node_id) 2423 2424show port traffic management hierarchy node stats 2425~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2426 2427Show the port traffic management hierarchy node statistics:: 2428 2429 testpmd> show port tm node stats (port_id) (node_id) (clear) 2430 2431where: 2432 2433* ``clear``: When this parameter has a non-zero value, the statistics counters 2434 are cleared (i.e. set to zero) immediately after they have been read, 2435 otherwise the statistics counters are left untouched. 2436 2437Add port traffic management private shaper profile 2438~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2439 2440Add the port traffic management private shaper profile:: 2441 2442 testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \ 2443 (tb_rate) (tb_size) (packet_length_adjust) 2444 2445where: 2446 2447* ``shaper_profile id``: Shaper profile ID for the new profile. 2448* ``tb_rate``: Token bucket rate (bytes per second). 2449* ``tb_size``: Token bucket size (bytes). 2450* ``packet_length_adjust``: The value (bytes) to be added to the length of 2451 each packet for the purpose of shaping. This parameter value can be used to 2452 correct the packet length with the framing overhead bytes that are consumed 2453 on the wire. 2454 2455Delete port traffic management private shaper profile 2456~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2457 2458Delete the port traffic management private shaper:: 2459 2460 testpmd> del port tm node shaper profile (port_id) (shaper_profile_id) 2461 2462where: 2463 2464* ``shaper_profile id``: Shaper profile ID that needs to be deleted. 2465 2466Add port traffic management shared shaper 2467~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2468 2469Create the port traffic management shared shaper:: 2470 2471 testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \ 2472 (shaper_profile_id) 2473 2474where: 2475 2476* ``shared_shaper_id``: Shared shaper ID to be created. 2477* ``shaper_profile id``: Shaper profile ID for shared shaper. 2478 2479Set port traffic management shared shaper 2480~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2481 2482Update the port traffic management shared shaper:: 2483 2484 testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \ 2485 (shaper_profile_id) 2486 2487where: 2488 2489* ``shared_shaper_id``: Shared shaper ID to be update. 2490* ``shaper_profile id``: Shaper profile ID for shared shaper. 2491 2492Delete port traffic management shared shaper 2493~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2494 2495Delete the port traffic management shared shaper:: 2496 2497 testpmd> del port tm node shared shaper (port_id) (shared_shaper_id) 2498 2499where: 2500 2501* ``shared_shaper_id``: Shared shaper ID to be deleted. 2502 2503Set port traffic management hiearchy node private shaper 2504~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2505 2506set the port traffic management hierarchy node private shaper:: 2507 2508 testpmd> set port tm node shaper profile (port_id) (node_id) \ 2509 (shaper_profile_id) 2510 2511where: 2512 2513* ``shaper_profile id``: Private shaper profile ID to be enabled on the 2514 hierarchy node. 2515 2516Add port traffic management WRED profile 2517~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2518 2519Create a new WRED profile:: 2520 2521 testpmd> add port tm node wred profile (port_id) (wred_profile_id) \ 2522 (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \ 2523 (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \ 2524 (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r) 2525 2526where: 2527 2528* ``wred_profile id``: Identifier for the newly create WRED profile 2529* ``color_g``: Packet color (green) 2530* ``min_th_g``: Minimum queue threshold for packet with green color 2531* ``max_th_g``: Minimum queue threshold for packet with green color 2532* ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp) 2533* ``wq_log2_g``: Negated log2 of queue weight (wq) 2534* ``color_y``: Packet color (yellow) 2535* ``min_th_y``: Minimum queue threshold for packet with yellow color 2536* ``max_th_y``: Minimum queue threshold for packet with yellow color 2537* ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp) 2538* ``wq_log2_y``: Negated log2 of queue weight (wq) 2539* ``color_r``: Packet color (red) 2540* ``min_th_r``: Minimum queue threshold for packet with yellow color 2541* ``max_th_r``: Minimum queue threshold for packet with yellow color 2542* ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp) 2543* ``wq_log2_r``: Negated log2 of queue weight (wq) 2544 2545Delete port traffic management WRED profile 2546~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2547 2548Delete the WRED profile:: 2549 2550 testpmd> del port tm node wred profile (port_id) (wred_profile_id) 2551 2552Add port traffic management hierarchy nonleaf node 2553~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2554 2555Add nonleaf node to port traffic management hiearchy:: 2556 2557 testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \ 2558 (priority) (weight) (level_id) (shaper_profile_id) \ 2559 (n_sp_priorities) (stats_mask) (n_shared_shapers) \ 2560 [(shared_shaper_0) (shared_shaper_1) ...] \ 2561 2562where: 2563 2564* ``parent_node_id``: Node ID of the parent. 2565* ``priority``: Node priority (highest node priority is zero). This is used by 2566 the SP algorithm running on the parent node for scheduling this node. 2567* ``weight``: Node weight (lowest weight is one). The node weight is relative 2568 to the weight sum of all siblings that have the same priority. It is used by 2569 the WFQ algorithm running on the parent node for scheduling this node. 2570* ``level_id``: Hiearchy level of the node. 2571* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by 2572 the node. 2573* ``n_sp_priorities``: Number of strict priorities. 2574* ``stats_mask``: Mask of statistics counter types to be enabled for this node. 2575* ``n_shared_shapers``: Number of shared shapers. 2576* ``shared_shaper_id``: Shared shaper id. 2577 2578Add port traffic management hierarchy leaf node 2579~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2580 2581Add leaf node to port traffic management hiearchy:: 2582 2583 testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \ 2584 (priority) (weight) (level_id) (shaper_profile_id) \ 2585 (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \ 2586 [(shared_shaper_id) (shared_shaper_id) ...] \ 2587 2588where: 2589 2590* ``parent_node_id``: Node ID of the parent. 2591* ``priority``: Node priority (highest node priority is zero). This is used by 2592 the SP algorithm running on the parent node for scheduling this node. 2593* ``weight``: Node weight (lowest weight is one). The node weight is relative 2594 to the weight sum of all siblings that have the same priority. It is used by 2595 the WFQ algorithm running on the parent node for scheduling this node. 2596* ``level_id``: Hiearchy level of the node. 2597* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by 2598 the node. 2599* ``cman_mode``: Congestion management mode to be enabled for this node. 2600* ``wred_profile_id``: WRED profile id to be enabled for this node. 2601* ``stats_mask``: Mask of statistics counter types to be enabled for this node. 2602* ``n_shared_shapers``: Number of shared shapers. 2603* ``shared_shaper_id``: Shared shaper id. 2604 2605Delete port traffic management hierarchy node 2606~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2607 2608Delete node from port traffic management hiearchy:: 2609 2610 testpmd> del port tm node (port_id) (node_id) 2611 2612Update port traffic management hierarchy parent node 2613~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2614 2615Update port traffic management hierarchy parent node:: 2616 2617 testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \ 2618 (priority) (weight) 2619 2620This function can only be called after the hierarchy commit invocation. Its 2621success depends on the port support for this operation, as advertised through 2622the port capability set. This function is valid for all nodes of the traffic 2623management hierarchy except root node. 2624 2625Suspend port traffic management hierarchy node 2626~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2627 2628 testpmd> suspend port tm node (port_id) (node_id) 2629 2630Resume port traffic management hierarchy node 2631~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2632 2633 testpmd> resume port tm node (port_id) (node_id) 2634 2635Commit port traffic management hierarchy 2636~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2637 2638Commit the traffic management hierarchy on the port:: 2639 2640 testpmd> port tm hierarchy commit (port_id) (clean_on_fail) 2641 2642where: 2643 2644* ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function 2645 call failure. On the other hand, hierarchy is preserved when this parameter 2646 is equal to zero. 2647 2648Set port traffic management default hierarchy (tm forwarding mode) 2649~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2650 2651set the traffic management default hierarchy on the port:: 2652 2653 testpmd> set port tm hierarchy default (port_id) 2654 2655Filter Functions 2656---------------- 2657 2658This section details the available filter functions that are available. 2659 2660Note these functions interface the deprecated legacy filtering framework, 2661superseded by *rte_flow*. See `Flow rules management`_. 2662 2663ethertype_filter 2664~~~~~~~~~~~~~~~~~~~~ 2665 2666Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue:: 2667 2668 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \ 2669 ethertype (ether_type) (drop|fwd) queue (queue_id) 2670 2671The available information parameters are: 2672 2673* ``port_id``: The port which the Ethertype filter assigned on. 2674 2675* ``mac_addr``: Compare destination mac address. 2676 2677* ``mac_ignr``: Ignore destination mac address match. 2678 2679* ``mac_address``: Destination mac address to match. 2680 2681* ``ether_type``: The EtherType value want to match, 2682 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid. 2683 2684* ``queue_id``: The receive queue associated with this EtherType filter. 2685 It is meaningless when deleting or dropping. 2686 2687Example, to add/remove an ethertype filter rule:: 2688 2689 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \ 2690 ethertype 0x0806 fwd queue 3 2691 2692 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \ 2693 ethertype 0x0806 fwd queue 3 2694 26952tuple_filter 2696~~~~~~~~~~~~~~~~~ 2697 2698Add or delete a 2-tuple filter, 2699which identifies packets by specific protocol and destination TCP/UDP port 2700and forwards packets into one of the receive queues:: 2701 2702 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \ 2703 protocol (protocol_value) mask (mask_value) \ 2704 tcp_flags (tcp_flags_value) priority (prio_value) \ 2705 queue (queue_id) 2706 2707The available information parameters are: 2708 2709* ``port_id``: The port which the 2-tuple filter assigned on. 2710 2711* ``dst_port_value``: Destination port in L4. 2712 2713* ``protocol_value``: IP L4 protocol. 2714 2715* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate. 2716 2717* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP). 2718 2719* ``prio_value``: Priority of this filter. 2720 2721* ``queue_id``: The receive queue associated with this 2-tuple filter. 2722 2723Example, to add/remove an 2tuple filter rule:: 2724 2725 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \ 2726 tcp_flags 0x02 priority 3 queue 3 2727 2728 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \ 2729 tcp_flags 0x02 priority 3 queue 3 2730 27315tuple_filter 2732~~~~~~~~~~~~~~~~~ 2733 2734Add or delete a 5-tuple filter, 2735which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port) 2736and routes packets into one of the receive queues:: 2737 2738 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \ 2739 (src_address) dst_port (dst_port_value) \ 2740 src_port (src_port_value) protocol (protocol_value) \ 2741 mask (mask_value) tcp_flags (tcp_flags_value) \ 2742 priority (prio_value) queue (queue_id) 2743 2744The available information parameters are: 2745 2746* ``port_id``: The port which the 5-tuple filter assigned on. 2747 2748* ``dst_address``: Destination IP address. 2749 2750* ``src_address``: Source IP address. 2751 2752* ``dst_port_value``: TCP/UDP destination port. 2753 2754* ``src_port_value``: TCP/UDP source port. 2755 2756* ``protocol_value``: L4 protocol. 2757 2758* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate 2759 2760* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP). 2761 2762* ``prio_value``: The priority of this filter. 2763 2764* ``queue_id``: The receive queue associated with this 5-tuple filter. 2765 2766Example, to add/remove an 5tuple filter rule:: 2767 2768 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 2769 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 2770 flags 0x0 priority 3 queue 3 2771 2772 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 2773 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 2774 flags 0x0 priority 3 queue 3 2775 2776syn_filter 2777~~~~~~~~~~ 2778 2779Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue:: 2780 2781 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id) 2782 2783The available information parameters are: 2784 2785* ``port_id``: The port which the SYN filter assigned on. 2786 2787* ``high``: This SYN filter has higher priority than other filters. 2788 2789* ``low``: This SYN filter has lower priority than other filters. 2790 2791* ``queue_id``: The receive queue associated with this SYN filter 2792 2793Example:: 2794 2795 testpmd> syn_filter 0 add priority high queue 3 2796 2797flex_filter 2798~~~~~~~~~~~ 2799 2800With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet 2801and routed into one of the receive queues:: 2802 2803 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \ 2804 mask (mask_value) priority (prio_value) queue (queue_id) 2805 2806The available information parameters are: 2807 2808* ``port_id``: The port which the Flex filter is assigned on. 2809 2810* ``len_value``: Filter length in bytes, no greater than 128. 2811 2812* ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match. 2813 2814* ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match. 2815 2816* ``prio_value``: The priority of this filter. 2817 2818* ``queue_id``: The receive queue associated with this Flex filter. 2819 2820Example:: 2821 2822 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \ 2823 mask 000C priority 3 queue 3 2824 2825 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \ 2826 mask 000C priority 3 queue 3 2827 2828 2829.. _testpmd_flow_director: 2830 2831flow_director_filter 2832~~~~~~~~~~~~~~~~~~~~ 2833 2834The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues. 2835 2836Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and 2837Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter: 2838 2839* Perfect match filters. 2840 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2841 The masked fields are for IP flow. 2842 2843* Signature filters. 2844 The hardware checks a match between a hash-based signature of the masked fields of the received packet. 2845 2846* Perfect-mac-vlan match filters. 2847 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2848 The masked fields are for MAC VLAN flow. 2849 2850* Perfect-tunnel match filters. 2851 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2852 The masked fields are for tunnel flow. 2853 2854* Perfect-raw-flow-type match filters. 2855 The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet. 2856 The masked fields are specified by input sets. 2857 2858The Flow Director filters can match the different fields for different type of packet: flow type, specific input set 2859per flow type and the flexible payload. 2860 2861The Flow Director can also mask out parts of all of these fields so that filters 2862are only applied to certain fields or parts of the fields. 2863 2864Note that for raw flow type mode the source and destination fields in the 2865raw packet buffer need to be presented in a reversed order with respect 2866to the expected received packets. 2867For example: IP source and destination addresses or TCP/UDP/SCTP 2868source and destination ports 2869 2870Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information. 2871 2872# Commands to add flow director filters of different flow types:: 2873 2874 flow_director_filter (port_id) mode IP (add|del|update) \ 2875 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \ 2876 src (src_ip_address) dst (dst_ip_address) \ 2877 tos (tos_value) proto (proto_value) ttl (ttl_value) \ 2878 vlan (vlan_value) flexbytes (flexbytes_value) \ 2879 (drop|fwd) pf|vf(vf_id) queue (queue_id) \ 2880 fd_id (fd_id_value) 2881 2882 flow_director_filter (port_id) mode IP (add|del|update) \ 2883 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \ 2884 src (src_ip_address) (src_port) \ 2885 dst (dst_ip_address) (dst_port) \ 2886 tos (tos_value) ttl (ttl_value) \ 2887 vlan (vlan_value) flexbytes (flexbytes_value) \ 2888 (drop|fwd) queue pf|vf(vf_id) (queue_id) \ 2889 fd_id (fd_id_value) 2890 2891 flow_director_filter (port_id) mode IP (add|del|update) \ 2892 flow (ipv4-sctp|ipv6-sctp) \ 2893 src (src_ip_address) (src_port) \ 2894 dst (dst_ip_address) (dst_port) \ 2895 tos (tos_value) ttl (ttl_value) \ 2896 tag (verification_tag) vlan (vlan_value) \ 2897 flexbytes (flexbytes_value) (drop|fwd) \ 2898 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value) 2899 2900 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \ 2901 ether (ethertype) flexbytes (flexbytes_value) \ 2902 (drop|fwd) pf|vf(vf_id) queue (queue_id) 2903 fd_id (fd_id_value) 2904 2905 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \ 2906 mac (mac_address) vlan (vlan_value) \ 2907 flexbytes (flexbytes_value) (drop|fwd) \ 2908 queue (queue_id) fd_id (fd_id_value) 2909 2910 flow_director_filter (port_id) mode Tunnel (add|del|update) \ 2911 mac (mac_address) vlan (vlan_value) \ 2912 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \ 2913 flexbytes (flexbytes_value) (drop|fwd) \ 2914 queue (queue_id) fd_id (fd_id_value) 2915 2916 flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \ 2917 (drop|fwd) queue (queue_id) fd_id (fd_id_value) \ 2918 packet (packet file name) 2919 2920For example, to add an ipv4-udp flow type filter:: 2921 2922 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \ 2923 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \ 2924 fwd pf queue 1 fd_id 1 2925 2926For example, add an ipv4-other flow type filter:: 2927 2928 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \ 2929 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \ 2930 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1 2931 2932flush_flow_director 2933~~~~~~~~~~~~~~~~~~~ 2934 2935Flush all flow director filters on a device:: 2936 2937 testpmd> flush_flow_director (port_id) 2938 2939Example, to flush all flow director filter on port 0:: 2940 2941 testpmd> flush_flow_director 0 2942 2943flow_director_mask 2944~~~~~~~~~~~~~~~~~~ 2945 2946Set flow director's input masks:: 2947 2948 flow_director_mask (port_id) mode IP vlan (vlan_value) \ 2949 src_mask (ipv4_src) (ipv6_src) (src_port) \ 2950 dst_mask (ipv4_dst) (ipv6_dst) (dst_port) 2951 2952 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value) 2953 2954 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \ 2955 mac (mac_value) tunnel-type (tunnel_type_value) \ 2956 tunnel-id (tunnel_id_value) 2957 2958Example, to set flow director mask on port 0:: 2959 2960 testpmd> flow_director_mask 0 mode IP vlan 0xefff \ 2961 src_mask 255.255.255.255 \ 2962 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \ 2963 dst_mask 255.255.255.255 \ 2964 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF 2965 2966flow_director_flex_mask 2967~~~~~~~~~~~~~~~~~~~~~~~ 2968 2969set masks of flow director's flexible payload based on certain flow type:: 2970 2971 testpmd> flow_director_flex_mask (port_id) \ 2972 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2973 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \ 2974 l2_payload|all) (mask) 2975 2976Example, to set flow director's flex mask for all flow type on port 0:: 2977 2978 testpmd> flow_director_flex_mask 0 flow all \ 2979 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0) 2980 2981 2982flow_director_flex_payload 2983~~~~~~~~~~~~~~~~~~~~~~~~~~ 2984 2985Configure flexible payload selection:: 2986 2987 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config) 2988 2989For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload:: 2990 2991 testpmd> flow_director_flex_payload 0 l4 \ 2992 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19) 2993 2994get_sym_hash_ena_per_port 2995~~~~~~~~~~~~~~~~~~~~~~~~~ 2996 2997Get symmetric hash enable configuration per port:: 2998 2999 get_sym_hash_ena_per_port (port_id) 3000 3001For example, to get symmetric hash enable configuration of port 1:: 3002 3003 testpmd> get_sym_hash_ena_per_port 1 3004 3005set_sym_hash_ena_per_port 3006~~~~~~~~~~~~~~~~~~~~~~~~~ 3007 3008Set symmetric hash enable configuration per port to enable or disable:: 3009 3010 set_sym_hash_ena_per_port (port_id) (enable|disable) 3011 3012For example, to set symmetric hash enable configuration of port 1 to enable:: 3013 3014 testpmd> set_sym_hash_ena_per_port 1 enable 3015 3016get_hash_global_config 3017~~~~~~~~~~~~~~~~~~~~~~ 3018 3019Get the global configurations of hash filters:: 3020 3021 get_hash_global_config (port_id) 3022 3023For example, to get the global configurations of hash filters of port 1:: 3024 3025 testpmd> get_hash_global_config 1 3026 3027set_hash_global_config 3028~~~~~~~~~~~~~~~~~~~~~~ 3029 3030Set the global configurations of hash filters:: 3031 3032 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \ 3033 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \ 3034 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \ 3035 (enable|disable) 3036 3037For example, to enable simple_xor for flow type of ipv6 on port 2:: 3038 3039 testpmd> set_hash_global_config 2 simple_xor ipv6 enable 3040 3041set_hash_input_set 3042~~~~~~~~~~~~~~~~~~ 3043 3044Set the input set for hash:: 3045 3046 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 3047 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 3048 l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \ 3049 ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \ 3050 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \ 3051 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \ 3052 fld-8th|none) (select|add) 3053 3054For example, to add source IP to hash input set for flow type of ipv4-udp on port 0:: 3055 3056 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add 3057 3058set_fdir_input_set 3059~~~~~~~~~~~~~~~~~~ 3060 3061The Flow Director filters can match the different fields for different type of packet, i.e. specific input set 3062on per flow type and the flexible payload. This command can be used to change input set for each flow type. 3063 3064Set the input set for flow director:: 3065 3066 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 3067 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 3068 l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \ 3069 ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \ 3070 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \ 3071 sctp-dst-port|sctp-veri-tag|none) (select|add) 3072 3073For example to add source IP to FD input set for flow type of ipv4-udp on port 0:: 3074 3075 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add 3076 3077global_config 3078~~~~~~~~~~~~~ 3079 3080Set different GRE key length for input set:: 3081 3082 global_config (port_id) gre-key-len (number in bytes) 3083 3084For example to set GRE key length for input set to 4 bytes on port 0:: 3085 3086 testpmd> global_config 0 gre-key-len 4 3087 3088 3089.. _testpmd_rte_flow: 3090 3091Flow rules management 3092--------------------- 3093 3094Control of the generic flow API (*rte_flow*) is fully exposed through the 3095``flow`` command (validation, creation, destruction, queries and operation 3096modes). 3097 3098Considering *rte_flow* overlaps with all `Filter Functions`_, using both 3099features simultaneously may cause undefined side-effects and is therefore 3100not recommended. 3101 3102``flow`` syntax 3103~~~~~~~~~~~~~~~ 3104 3105Because the ``flow`` command uses dynamic tokens to handle the large number 3106of possible flow rules combinations, its behavior differs slightly from 3107other commands, in particular: 3108 3109- Pressing *?* or the *<tab>* key displays contextual help for the current 3110 token, not that of the entire command. 3111 3112- Optional and repeated parameters are supported (provided they are listed 3113 in the contextual help). 3114 3115The first parameter stands for the operation mode. Possible operations and 3116their general syntax are described below. They are covered in detail in the 3117following sections. 3118 3119- Check whether a flow rule can be created:: 3120 3121 flow validate {port_id} 3122 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3123 pattern {item} [/ {item} [...]] / end 3124 actions {action} [/ {action} [...]] / end 3125 3126- Create a flow rule:: 3127 3128 flow create {port_id} 3129 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3130 pattern {item} [/ {item} [...]] / end 3131 actions {action} [/ {action} [...]] / end 3132 3133- Destroy specific flow rules:: 3134 3135 flow destroy {port_id} rule {rule_id} [...] 3136 3137- Destroy all flow rules:: 3138 3139 flow flush {port_id} 3140 3141- Query an existing flow rule:: 3142 3143 flow query {port_id} {rule_id} {action} 3144 3145- List existing flow rules sorted by priority, filtered by group 3146 identifiers:: 3147 3148 flow list {port_id} [group {group_id}] [...] 3149 3150- Restrict ingress traffic to the defined flow rules:: 3151 3152 flow isolate {port_id} {boolean} 3153 3154Validating flow rules 3155~~~~~~~~~~~~~~~~~~~~~ 3156 3157``flow validate`` reports whether a flow rule would be accepted by the 3158underlying device in its current state but stops short of creating it. It is 3159bound to ``rte_flow_validate()``:: 3160 3161 flow validate {port_id} 3162 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3163 pattern {item} [/ {item} [...]] / end 3164 actions {action} [/ {action} [...]] / end 3165 3166If successful, it will show:: 3167 3168 Flow rule validated 3169 3170Otherwise it will show an error message of the form:: 3171 3172 Caught error type [...] ([...]): [...] 3173 3174This command uses the same parameters as ``flow create``, their format is 3175described in `Creating flow rules`_. 3176 3177Check whether redirecting any Ethernet packet received on port 0 to RX queue 3178index 6 is supported:: 3179 3180 testpmd> flow validate 0 ingress pattern eth / end 3181 actions queue index 6 / end 3182 Flow rule validated 3183 testpmd> 3184 3185Port 0 does not support TCPv6 rules:: 3186 3187 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end 3188 actions drop / end 3189 Caught error type 9 (specific pattern item): Invalid argument 3190 testpmd> 3191 3192Creating flow rules 3193~~~~~~~~~~~~~~~~~~~ 3194 3195``flow create`` validates and creates the specified flow rule. It is bound 3196to ``rte_flow_create()``:: 3197 3198 flow create {port_id} 3199 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3200 pattern {item} [/ {item} [...]] / end 3201 actions {action} [/ {action} [...]] / end 3202 3203If successful, it will return a flow rule ID usable with other commands:: 3204 3205 Flow rule #[...] created 3206 3207Otherwise it will show an error message of the form:: 3208 3209 Caught error type [...] ([...]): [...] 3210 3211Parameters describe in the following order: 3212 3213- Attributes (*group*, *priority*, *ingress*, *egress*, *transfer* tokens). 3214- A matching pattern, starting with the *pattern* token and terminated by an 3215 *end* pattern item. 3216- Actions, starting with the *actions* token and terminated by an *end* 3217 action. 3218 3219These translate directly to *rte_flow* objects provided as-is to the 3220underlying functions. 3221 3222The shortest valid definition only comprises mandatory tokens:: 3223 3224 testpmd> flow create 0 pattern end actions end 3225 3226Note that PMDs may refuse rules that essentially do nothing such as this 3227one. 3228 3229**All unspecified object values are automatically initialized to 0.** 3230 3231Attributes 3232^^^^^^^^^^ 3233 3234These tokens affect flow rule attributes (``struct rte_flow_attr``) and are 3235specified before the ``pattern`` token. 3236 3237- ``group {group id}``: priority group. 3238- ``priority {level}``: priority level within group. 3239- ``ingress``: rule applies to ingress traffic. 3240- ``egress``: rule applies to egress traffic. 3241- ``transfer``: apply rule directly to endpoints found in pattern. 3242 3243Each instance of an attribute specified several times overrides the previous 3244value as shown below (group 4 is used):: 3245 3246 testpmd> flow create 0 group 42 group 24 group 4 [...] 3247 3248Note that once enabled, ``ingress`` and ``egress`` cannot be disabled. 3249 3250While not specifying a direction is an error, some rules may allow both 3251simultaneously. 3252 3253Most rules affect RX therefore contain the ``ingress`` token:: 3254 3255 testpmd> flow create 0 ingress pattern [...] 3256 3257Matching pattern 3258^^^^^^^^^^^^^^^^ 3259 3260A matching pattern starts after the ``pattern`` token. It is made of pattern 3261items and is terminated by a mandatory ``end`` item. 3262 3263Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum 3264rte_flow_item_type``). 3265 3266The ``/`` token is used as a separator between pattern items as shown 3267below:: 3268 3269 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...] 3270 3271Note that protocol items like these must be stacked from lowest to highest 3272layer to make sense. For instance, the following rule is either invalid or 3273unlikely to match any packet:: 3274 3275 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...] 3276 3277More information on these restrictions can be found in the *rte_flow* 3278documentation. 3279 3280Several items support additional specification structures, for example 3281``ipv4`` allows specifying source and destination addresses as follows:: 3282 3283 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1 3284 dst is 10.2.0.0 / end [...] 3285 3286This rule matches all IPv4 traffic with the specified properties. 3287 3288In this example, ``src`` and ``dst`` are field names of the underlying 3289``struct rte_flow_item_ipv4`` object. All item properties can be specified 3290in a similar fashion. 3291 3292The ``is`` token means that the subsequent value must be matched exactly, 3293and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item`` 3294accordingly. Possible assignment tokens are: 3295 3296- ``is``: match value perfectly (with full bit-mask). 3297- ``spec``: match value according to configured bit-mask. 3298- ``last``: specify upper bound to establish a range. 3299- ``mask``: specify bit-mask with relevant bits set to one. 3300- ``prefix``: generate bit-mask from a prefix length. 3301 3302These yield identical results:: 3303 3304 ipv4 src is 10.1.1.1 3305 3306:: 3307 3308 ipv4 src spec 10.1.1.1 src mask 255.255.255.255 3309 3310:: 3311 3312 ipv4 src spec 10.1.1.1 src prefix 32 3313 3314:: 3315 3316 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value 3317 3318:: 3319 3320 ipv4 src is 10.1.1.1 src last 0 # 0 disables range 3321 3322Inclusive ranges can be defined with ``last``:: 3323 3324 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4 3325 3326Note that ``mask`` affects both ``spec`` and ``last``:: 3327 3328 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0 3329 # matches 10.1.0.0 to 10.2.255.255 3330 3331Properties can be modified multiple times:: 3332 3333 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4 3334 3335:: 3336 3337 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16 3338 3339Pattern items 3340^^^^^^^^^^^^^ 3341 3342This section lists supported pattern items and their attributes, if any. 3343 3344- ``end``: end list of pattern items. 3345 3346- ``void``: no-op pattern item. 3347 3348- ``invert``: perform actions when pattern does not match. 3349 3350- ``any``: match any protocol for the current layer. 3351 3352 - ``num {unsigned}``: number of layers covered. 3353 3354- ``pf``: match traffic from/to the physical function. 3355 3356- ``vf``: match traffic from/to a virtual function ID. 3357 3358 - ``id {unsigned}``: VF ID. 3359 3360- ``phy_port``: match traffic from/to a specific physical port. 3361 3362 - ``index {unsigned}``: physical port index. 3363 3364- ``port_id``: match traffic from/to a given DPDK port ID. 3365 3366 - ``id {unsigned}``: DPDK port ID. 3367 3368- ``mark``: match value set in previously matched flow rule using the mark action. 3369 3370 - ``id {unsigned}``: arbitrary integer value. 3371 3372- ``raw``: match an arbitrary byte string. 3373 3374 - ``relative {boolean}``: look for pattern after the previous item. 3375 - ``search {boolean}``: search pattern from offset (see also limit). 3376 - ``offset {integer}``: absolute or relative offset for pattern. 3377 - ``limit {unsigned}``: search area limit for start of pattern. 3378 - ``pattern {string}``: byte string to look for. 3379 3380- ``eth``: match Ethernet header. 3381 3382 - ``dst {MAC-48}``: destination MAC. 3383 - ``src {MAC-48}``: source MAC. 3384 - ``type {unsigned}``: EtherType or TPID. 3385 3386- ``vlan``: match 802.1Q/ad VLAN tag. 3387 3388 - ``tci {unsigned}``: tag control information. 3389 - ``pcp {unsigned}``: priority code point. 3390 - ``dei {unsigned}``: drop eligible indicator. 3391 - ``vid {unsigned}``: VLAN identifier. 3392 - ``inner_type {unsigned}``: inner EtherType or TPID. 3393 3394- ``ipv4``: match IPv4 header. 3395 3396 - ``tos {unsigned}``: type of service. 3397 - ``ttl {unsigned}``: time to live. 3398 - ``proto {unsigned}``: next protocol ID. 3399 - ``src {ipv4 address}``: source address. 3400 - ``dst {ipv4 address}``: destination address. 3401 3402- ``ipv6``: match IPv6 header. 3403 3404 - ``tc {unsigned}``: traffic class. 3405 - ``flow {unsigned}``: flow label. 3406 - ``proto {unsigned}``: protocol (next header). 3407 - ``hop {unsigned}``: hop limit. 3408 - ``src {ipv6 address}``: source address. 3409 - ``dst {ipv6 address}``: destination address. 3410 3411- ``icmp``: match ICMP header. 3412 3413 - ``type {unsigned}``: ICMP packet type. 3414 - ``code {unsigned}``: ICMP packet code. 3415 3416- ``udp``: match UDP header. 3417 3418 - ``src {unsigned}``: UDP source port. 3419 - ``dst {unsigned}``: UDP destination port. 3420 3421- ``tcp``: match TCP header. 3422 3423 - ``src {unsigned}``: TCP source port. 3424 - ``dst {unsigned}``: TCP destination port. 3425 3426- ``sctp``: match SCTP header. 3427 3428 - ``src {unsigned}``: SCTP source port. 3429 - ``dst {unsigned}``: SCTP destination port. 3430 - ``tag {unsigned}``: validation tag. 3431 - ``cksum {unsigned}``: checksum. 3432 3433- ``vxlan``: match VXLAN header. 3434 3435 - ``vni {unsigned}``: VXLAN identifier. 3436 3437- ``e_tag``: match IEEE 802.1BR E-Tag header. 3438 3439 - ``grp_ecid_b {unsigned}``: GRP and E-CID base. 3440 3441- ``nvgre``: match NVGRE header. 3442 3443 - ``tni {unsigned}``: virtual subnet ID. 3444 3445- ``mpls``: match MPLS header. 3446 3447 - ``label {unsigned}``: MPLS label. 3448 3449- ``gre``: match GRE header. 3450 3451 - ``protocol {unsigned}``: protocol type. 3452 3453- ``fuzzy``: fuzzy pattern match, expect faster than default. 3454 3455 - ``thresh {unsigned}``: accuracy threshold. 3456 3457- ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header. 3458 3459 - ``teid {unsigned}``: tunnel endpoint identifier. 3460 3461- ``geneve``: match GENEVE header. 3462 3463 - ``vni {unsigned}``: virtual network identifier. 3464 - ``protocol {unsigned}``: protocol type. 3465 3466- ``vxlan-gpe``: match VXLAN-GPE header. 3467 3468 - ``vni {unsigned}``: VXLAN-GPE identifier. 3469 3470- ``arp_eth_ipv4``: match ARP header for Ethernet/IPv4. 3471 3472 - ``sha {MAC-48}``: sender hardware address. 3473 - ``spa {ipv4 address}``: sender IPv4 address. 3474 - ``tha {MAC-48}``: target hardware address. 3475 - ``tpa {ipv4 address}``: target IPv4 address. 3476 3477- ``ipv6_ext``: match presence of any IPv6 extension header. 3478 3479 - ``next_hdr {unsigned}``: next header. 3480 3481- ``icmp6``: match any ICMPv6 header. 3482 3483 - ``type {unsigned}``: ICMPv6 type. 3484 - ``code {unsigned}``: ICMPv6 code. 3485 3486- ``icmp6_nd_ns``: match ICMPv6 neighbor discovery solicitation. 3487 3488 - ``target_addr {ipv6 address}``: target address. 3489 3490- ``icmp6_nd_na``: match ICMPv6 neighbor discovery advertisement. 3491 3492 - ``target_addr {ipv6 address}``: target address. 3493 3494- ``icmp6_nd_opt``: match presence of any ICMPv6 neighbor discovery option. 3495 3496 - ``type {unsigned}``: ND option type. 3497 3498- ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery source Ethernet 3499 link-layer address option. 3500 3501 - ``sla {MAC-48}``: source Ethernet LLA. 3502 3503- ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery target Ethernet 3504 link-layer address option. 3505 3506 - ``tla {MAC-48}``: target Ethernet LLA. 3507 3508Actions list 3509^^^^^^^^^^^^ 3510 3511A list of actions starts after the ``actions`` token in the same fashion as 3512`Matching pattern`_; actions are separated by ``/`` tokens and the list is 3513terminated by a mandatory ``end`` action. 3514 3515Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum 3516rte_flow_action_type``). 3517 3518Dropping all incoming UDPv4 packets can be expressed as follows:: 3519 3520 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3521 actions drop / end 3522 3523Several actions have configurable properties which must be specified when 3524there is no valid default value. For example, ``queue`` requires a target 3525queue index. 3526 3527This rule redirects incoming UDPv4 traffic to queue index 6:: 3528 3529 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3530 actions queue index 6 / end 3531 3532While this one could be rejected by PMDs (unspecified queue index):: 3533 3534 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3535 actions queue / end 3536 3537As defined by *rte_flow*, the list is not ordered, all actions of a given 3538rule are performed simultaneously. These are equivalent:: 3539 3540 queue index 6 / void / mark id 42 / end 3541 3542:: 3543 3544 void / mark id 42 / queue index 6 / end 3545 3546All actions in a list should have different types, otherwise only the last 3547action of a given type is taken into account:: 3548 3549 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6 3550 3551:: 3552 3553 drop / drop / drop / end # drop is performed only once 3554 3555:: 3556 3557 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24 3558 3559Considering they are performed simultaneously, opposite and overlapping 3560actions can sometimes be combined when the end result is unambiguous:: 3561 3562 drop / queue index 6 / end # drop has no effect 3563 3564:: 3565 3566 queue index 6 / rss queues 6 7 8 / end # queue has no effect 3567 3568:: 3569 3570 drop / passthru / end # drop has no effect 3571 3572Note that PMDs may still refuse such combinations. 3573 3574Actions 3575^^^^^^^ 3576 3577This section lists supported actions and their attributes, if any. 3578 3579- ``end``: end list of actions. 3580 3581- ``void``: no-op action. 3582 3583- ``passthru``: let subsequent rule process matched packets. 3584 3585- ``jump``: redirect traffic to group on device. 3586 3587 - ``group {unsigned}``: group to redirect to. 3588 3589- ``mark``: attach 32 bit value to packets. 3590 3591 - ``id {unsigned}``: 32 bit value to return with packets. 3592 3593- ``flag``: flag packets. 3594 3595- ``queue``: assign packets to a given queue index. 3596 3597 - ``index {unsigned}``: queue index to use. 3598 3599- ``drop``: drop packets (note: passthru has priority). 3600 3601- ``count``: enable counters for this rule. 3602 3603- ``rss``: spread packets among several queues. 3604 3605 - ``func {hash function}``: RSS hash function to apply, allowed tokens are 3606 the same as `set_hash_global_config`_. 3607 3608 - ``level {unsigned}``: encapsulation level for ``types``. 3609 3610 - ``types [{RSS hash type} [...]] end``: specific RSS hash types, allowed 3611 tokens are the same as `set_hash_input_set`_, except that an empty list 3612 does not disable RSS but instead requests unspecified "best-effort" 3613 settings. 3614 3615 - ``key {string}``: RSS hash key, overrides ``key_len``. 3616 3617 - ``key_len {unsigned}``: RSS hash key length in bytes, can be used in 3618 conjunction with ``key`` to pad or truncate it. 3619 3620 - ``queues [{unsigned} [...]] end``: queue indices to use. 3621 3622- ``pf``: direct traffic to physical function. 3623 3624- ``vf``: direct traffic to a virtual function ID. 3625 3626 - ``original {boolean}``: use original VF ID if possible. 3627 - ``id {unsigned}``: VF ID. 3628 3629- ``phy_port``: direct packets to physical port index. 3630 3631 - ``original {boolean}``: use original port index if possible. 3632 - ``index {unsigned}``: physical port index. 3633 3634- ``port_id``: direct matching traffic to a given DPDK port ID. 3635 3636 - ``original {boolean}``: use original DPDK port ID if possible. 3637 - ``id {unsigned}``: DPDK port ID. 3638 3639- ``of_set_mpls_ttl``: OpenFlow's ``OFPAT_SET_MPLS_TTL``. 3640 3641 - ``mpls_ttl``: MPLS TTL. 3642 3643- ``of_dec_mpls_ttl``: OpenFlow's ``OFPAT_DEC_MPLS_TTL``. 3644 3645- ``of_set_nw_ttl``: OpenFlow's ``OFPAT_SET_NW_TTL``. 3646 3647 - ``nw_ttl``: IP TTL. 3648 3649- ``of_dec_nw_ttl``: OpenFlow's ``OFPAT_DEC_NW_TTL``. 3650 3651- ``of_copy_ttl_out``: OpenFlow's ``OFPAT_COPY_TTL_OUT``. 3652 3653- ``of_copy_ttl_in``: OpenFlow's ``OFPAT_COPY_TTL_IN``. 3654 3655- ``of_pop_vlan``: OpenFlow's ``OFPAT_POP_VLAN``. 3656 3657- ``of_push_vlan``: OpenFlow's ``OFPAT_PUSH_VLAN``. 3658 3659 - ``ethertype``: Ethertype. 3660 3661- ``of_set_vlan_vid``: OpenFlow's ``OFPAT_SET_VLAN_VID``. 3662 3663 - ``vlan_vid``: VLAN id. 3664 3665- ``of_set_vlan_pcp``: OpenFlow's ``OFPAT_SET_VLAN_PCP``. 3666 3667 - ``vlan_pcp``: VLAN priority. 3668 3669- ``of_pop_mpls``: OpenFlow's ``OFPAT_POP_MPLS``. 3670 3671 - ``ethertype``: Ethertype. 3672 3673- ``of_push_mpls``: OpenFlow's ``OFPAT_PUSH_MPLS``. 3674 3675 - ``ethertype``: Ethertype. 3676 3677- ``vxlan_encap``: Performs a VXLAN encapsulation, outer layer configuration 3678 is done through `Config VXLAN Encap outer layers`_. 3679 3680- ``vxlan_decap``: Performs a decapsulation action by stripping all headers of 3681 the VXLAN tunnel network overlay from the matched flow. 3682 3683Destroying flow rules 3684~~~~~~~~~~~~~~~~~~~~~ 3685 3686``flow destroy`` destroys one or more rules from their rule ID (as returned 3687by ``flow create``), this command calls ``rte_flow_destroy()`` as many 3688times as necessary:: 3689 3690 flow destroy {port_id} rule {rule_id} [...] 3691 3692If successful, it will show:: 3693 3694 Flow rule #[...] destroyed 3695 3696It does not report anything for rule IDs that do not exist. The usual error 3697message is shown when a rule cannot be destroyed:: 3698 3699 Caught error type [...] ([...]): [...] 3700 3701``flow flush`` destroys all rules on a device and does not take extra 3702arguments. It is bound to ``rte_flow_flush()``:: 3703 3704 flow flush {port_id} 3705 3706Any errors are reported as above. 3707 3708Creating several rules and destroying them:: 3709 3710 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3711 actions queue index 2 / end 3712 Flow rule #0 created 3713 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3714 actions queue index 3 / end 3715 Flow rule #1 created 3716 testpmd> flow destroy 0 rule 0 rule 1 3717 Flow rule #1 destroyed 3718 Flow rule #0 destroyed 3719 testpmd> 3720 3721The same result can be achieved using ``flow flush``:: 3722 3723 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3724 actions queue index 2 / end 3725 Flow rule #0 created 3726 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3727 actions queue index 3 / end 3728 Flow rule #1 created 3729 testpmd> flow flush 0 3730 testpmd> 3731 3732Non-existent rule IDs are ignored:: 3733 3734 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3735 actions queue index 2 / end 3736 Flow rule #0 created 3737 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3738 actions queue index 3 / end 3739 Flow rule #1 created 3740 testpmd> flow destroy 0 rule 42 rule 10 rule 2 3741 testpmd> 3742 testpmd> flow destroy 0 rule 0 3743 Flow rule #0 destroyed 3744 testpmd> 3745 3746Querying flow rules 3747~~~~~~~~~~~~~~~~~~~ 3748 3749``flow query`` queries a specific action of a flow rule having that 3750ability. Such actions collect information that can be reported using this 3751command. It is bound to ``rte_flow_query()``:: 3752 3753 flow query {port_id} {rule_id} {action} 3754 3755If successful, it will display either the retrieved data for known actions 3756or the following message:: 3757 3758 Cannot display result for action type [...] ([...]) 3759 3760Otherwise, it will complain either that the rule does not exist or that some 3761error occurred:: 3762 3763 Flow rule #[...] not found 3764 3765:: 3766 3767 Caught error type [...] ([...]): [...] 3768 3769Currently only the ``count`` action is supported. This action reports the 3770number of packets that hit the flow rule and the total number of bytes. Its 3771output has the following format:: 3772 3773 count: 3774 hits_set: [...] # whether "hits" contains a valid value 3775 bytes_set: [...] # whether "bytes" contains a valid value 3776 hits: [...] # number of packets 3777 bytes: [...] # number of bytes 3778 3779Querying counters for TCPv6 packets redirected to queue 6:: 3780 3781 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end 3782 actions queue index 6 / count / end 3783 Flow rule #4 created 3784 testpmd> flow query 0 4 count 3785 count: 3786 hits_set: 1 3787 bytes_set: 0 3788 hits: 386446 3789 bytes: 0 3790 testpmd> 3791 3792Listing flow rules 3793~~~~~~~~~~~~~~~~~~ 3794 3795``flow list`` lists existing flow rules sorted by priority and optionally 3796filtered by group identifiers:: 3797 3798 flow list {port_id} [group {group_id}] [...] 3799 3800This command only fails with the following message if the device does not 3801exist:: 3802 3803 Invalid port [...] 3804 3805Output consists of a header line followed by a short description of each 3806flow rule, one per line. There is no output at all when no flow rules are 3807configured on the device:: 3808 3809 ID Group Prio Attr Rule 3810 [...] [...] [...] [...] [...] 3811 3812``Attr`` column flags: 3813 3814- ``i`` for ``ingress``. 3815- ``e`` for ``egress``. 3816 3817Creating several flow rules and listing them:: 3818 3819 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3820 actions queue index 6 / end 3821 Flow rule #0 created 3822 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3823 actions queue index 2 / end 3824 Flow rule #1 created 3825 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end 3826 actions rss queues 6 7 8 end / end 3827 Flow rule #2 created 3828 testpmd> flow list 0 3829 ID Group Prio Attr Rule 3830 0 0 0 i- ETH IPV4 => QUEUE 3831 1 0 0 i- ETH IPV6 => QUEUE 3832 2 0 5 i- ETH IPV4 UDP => RSS 3833 testpmd> 3834 3835Rules are sorted by priority (i.e. group ID first, then priority level):: 3836 3837 testpmd> flow list 1 3838 ID Group Prio Attr Rule 3839 0 0 0 i- ETH => COUNT 3840 6 0 500 i- ETH IPV6 TCP => DROP COUNT 3841 5 0 1000 i- ETH IPV6 ICMP => QUEUE 3842 1 24 0 i- ETH IPV4 UDP => QUEUE 3843 4 24 10 i- ETH IPV4 TCP => DROP 3844 3 24 20 i- ETH IPV4 => DROP 3845 2 24 42 i- ETH IPV4 UDP => QUEUE 3846 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 3847 testpmd> 3848 3849Output can be limited to specific groups:: 3850 3851 testpmd> flow list 1 group 0 group 63 3852 ID Group Prio Attr Rule 3853 0 0 0 i- ETH => COUNT 3854 6 0 500 i- ETH IPV6 TCP => DROP COUNT 3855 5 0 1000 i- ETH IPV6 ICMP => QUEUE 3856 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 3857 testpmd> 3858 3859Toggling isolated mode 3860~~~~~~~~~~~~~~~~~~~~~~ 3861 3862``flow isolate`` can be used to tell the underlying PMD that ingress traffic 3863must only be injected from the defined flow rules; that no default traffic 3864is expected outside those rules and the driver is free to assign more 3865resources to handle them. It is bound to ``rte_flow_isolate()``:: 3866 3867 flow isolate {port_id} {boolean} 3868 3869If successful, enabling or disabling isolated mode shows either:: 3870 3871 Ingress traffic on port [...] 3872 is now restricted to the defined flow rules 3873 3874Or:: 3875 3876 Ingress traffic on port [...] 3877 is not restricted anymore to the defined flow rules 3878 3879Otherwise, in case of error:: 3880 3881 Caught error type [...] ([...]): [...] 3882 3883Mainly due to its side effects, PMDs supporting this mode may not have the 3884ability to toggle it more than once without reinitializing affected ports 3885first (e.g. by exiting testpmd). 3886 3887Enabling isolated mode:: 3888 3889 testpmd> flow isolate 0 true 3890 Ingress traffic on port 0 is now restricted to the defined flow rules 3891 testpmd> 3892 3893Disabling isolated mode:: 3894 3895 testpmd> flow isolate 0 false 3896 Ingress traffic on port 0 is not restricted anymore to the defined flow rules 3897 testpmd> 3898 3899Sample QinQ flow rules 3900~~~~~~~~~~~~~~~~~~~~~~ 3901 3902Before creating QinQ rule(s) the following commands should be issued to enable QinQ:: 3903 3904 testpmd> port stop 0 3905 testpmd> vlan set qinq on 0 3906 3907The above command sets the inner and outer TPID's to 0x8100. 3908 3909To change the TPID's the following commands should be used:: 3910 3911 testpmd> vlan set outer tpid 0xa100 0 3912 testpmd> vlan set inner tpid 0x9100 0 3913 testpmd> port start 0 3914 3915Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM. 3916 3917:: 3918 3919 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 / 3920 vlan tci is 456 / end actions vf id 1 / queue index 0 / end 3921 Flow rule #0 validated 3922 3923 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 / 3924 vlan tci is 456 / end actions vf id 123 / queue index 0 / end 3925 Flow rule #0 created 3926 3927 testpmd> flow list 0 3928 ID Group Prio Attr Rule 3929 0 0 0 i- ETH VLAN VLAN=>VF QUEUE 3930 3931Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host. 3932 3933:: 3934 3935 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 / 3936 vlan tci is 654 / end actions pf / queue index 0 / end 3937 Flow rule #1 validated 3938 3939 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 / 3940 vlan tci is 654 / end actions pf / queue index 1 / end 3941 Flow rule #1 created 3942 3943 testpmd> flow list 0 3944 ID Group Prio Attr Rule 3945 0 0 0 i- ETH VLAN VLAN=>VF QUEUE 3946 1 0 0 i- ETH VLAN VLAN=>PF QUEUE 3947 3948Sample VXLAN encapsulation rule 3949~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 3950 3951VXLAN encapsulation outer layer has default value pre-configured in testpmd 3952source code, those can be changed by using the following commands 3953 3954IPv4 VXLAN outer header:: 3955 3956 testpmd> set vxlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src 127.0.0.1 3957 ip-dst 128.0.0.1 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 3958 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 3959 queue index 0 / end 3960 3961 testpmd> set vxlan-with-vlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src 3962 127.0.0.1 ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11 3963 eth-dst 22:22:22:22:22:22 3964 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 3965 queue index 0 / end 3966 3967IPv6 VXLAN outer header:: 3968 3969 testpmd> set vxlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4 ip-src ::1 3970 ip-dst ::2222 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 3971 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 3972 queue index 0 / end 3973 3974 testpmd> set vxlan-with-vlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4 3975 ip-src ::1 ip-dst ::2222 vlan-tci 34 eth-src 11:11:11:11:11:11 3976 eth-dst 22:22:22:22:22:22 3977 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 3978 queue index 0 / end 3979 3980BPF Functions 3981-------------- 3982 3983The following sections show functions to load/unload eBPF based filters. 3984 3985bpf-load 3986~~~~~~~~ 3987 3988Load an eBPF program as a callback for partciular RX/TX queue:: 3989 3990 testpmd> bpf-load rx|tx (portid) (queueid) (load-flags) (bpf-prog-filename) 3991 3992The available load-flags are: 3993 3994* ``J``: use JIT generated native code, otherwise BPF interpreter will be used. 3995 3996* ``M``: assume input parameter is a pointer to rte_mbuf, otherwise assume it is a pointer to first segment's data. 3997 3998* ``-``: none. 3999 4000.. note:: 4001 4002 You'll need clang v3.7 or above to build bpf program you'd like to load 4003 4004For example: 4005 4006.. code-block:: console 4007 4008 cd test/bpf 4009 clang -O2 -target bpf -c t1.c 4010 4011Then to load (and JIT compile) t1.o at RX queue 0, port 1:: 4012 4013.. code-block:: console 4014 4015 testpmd> bpf-load rx 1 0 J ./dpdk.org/test/bpf/t1.o 4016 4017To load (not JITed) t1.o at TX queue 0, port 0:: 4018 4019.. code-block:: console 4020 4021 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o 4022 4023bpf-unload 4024~~~~~~~~~~ 4025 4026Unload previously loaded eBPF program for partciular RX/TX queue:: 4027 4028 testpmd> bpf-unload rx|tx (portid) (queueid) 4029 4030For example to unload BPF filter from TX queue 0, port 0: 4031 4032.. code-block:: console 4033 4034 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o 4035