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* ``softnic``: Demonstrates the softnic forwarding operation. In this mode, packet forwarding is 324 similar to I/O mode except for the fact that packets are loopback to the softnic ports only. Therefore, portmask parameter should be set to softnic port only. The various software based custom NIC pipelines specified through the softnic firmware (DPDK packet framework script) can be tested in this mode. Furthermore, it allows to build 5-level hierarchical QoS scheduler as a default option that can be enabled through CLI once testpmd application is initialised. The user can modify the default scheduler hierarchy or can specify the new QoS Scheduler hierarchy through CLI. Requires ``CONFIG_RTE_LIBRTE_PMD_SOFTNIC=y``. 325 326Example:: 327 328 testpmd> set fwd rxonly 329 330 Set rxonly packet forwarding mode 331 332 333read rxd 334~~~~~~~~ 335 336Display an RX descriptor for a port RX queue:: 337 338 testpmd> read rxd (port_id) (queue_id) (rxd_id) 339 340For example:: 341 342 testpmd> read rxd 0 0 4 343 0x0000000B - 0x001D0180 / 0x0000000B - 0x001D0180 344 345read txd 346~~~~~~~~ 347 348Display a TX descriptor for a port TX queue:: 349 350 testpmd> read txd (port_id) (queue_id) (txd_id) 351 352For example:: 353 354 testpmd> read txd 0 0 4 355 0x00000001 - 0x24C3C440 / 0x000F0000 - 0x2330003C 356 357ddp get list 358~~~~~~~~~~~~ 359 360Get loaded dynamic device personalization (DDP) package info list:: 361 362 testpmd> ddp get list (port_id) 363 364ddp get info 365~~~~~~~~~~~~ 366 367Display information about dynamic device personalization (DDP) profile:: 368 369 testpmd> ddp get info (profile_path) 370 371show vf stats 372~~~~~~~~~~~~~ 373 374Display VF statistics:: 375 376 testpmd> show vf stats (port_id) (vf_id) 377 378clear vf stats 379~~~~~~~~~~~~~~ 380 381Reset VF statistics:: 382 383 testpmd> clear vf stats (port_id) (vf_id) 384 385show port pctype mapping 386~~~~~~~~~~~~~~~~~~~~~~~~ 387 388List all items from the pctype mapping table:: 389 390 testpmd> show port (port_id) pctype mapping 391 392show rx offloading capabilities 393~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 394 395List all per queue and per port Rx offloading capabilities of a port:: 396 397 testpmd> show port (port_id) rx_offload capabilities 398 399show rx offloading configuration 400~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 401 402List port level and all queue level Rx offloading configuration:: 403 404 testpmd> show port (port_id) rx_offload configuration 405 406show tx offloading capabilities 407~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 408 409List all per queue and per port Tx offloading capabilities of a port:: 410 411 testpmd> show port (port_id) tx_offload capabilities 412 413show tx offloading configuration 414~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 415 416List port level and all queue level Tx offloading configuration:: 417 418 testpmd> show port (port_id) tx_offload configuration 419 420 421Configuration Functions 422----------------------- 423 424The testpmd application can be configured from the runtime as well as from the command-line. 425 426This section details the available configuration functions that are available. 427 428.. note:: 429 430 Configuration changes only become active when forwarding is started/restarted. 431 432set default 433~~~~~~~~~~~ 434 435Reset forwarding to the default configuration:: 436 437 testpmd> set default 438 439set verbose 440~~~~~~~~~~~ 441 442Set the debug verbosity level:: 443 444 testpmd> set verbose (level) 445 446Currently the only available levels are 0 (silent except for error) and 1 (fully verbose). 447 448set log 449~~~~~~~ 450 451Set the log level for a log type:: 452 453 testpmd> set log global|(type) (level) 454 455Where: 456 457* ``type`` is the log name. 458 459* ``level`` is the log level. 460 461For example, to change the global log level:: 462 testpmd> set log global (level) 463 464Regexes can also be used for type. To change log level of user1, user2 and user3:: 465 testpmd> set log user[1-3] (level) 466 467set nbport 468~~~~~~~~~~ 469 470Set the number of ports used by the application: 471 472set nbport (num) 473 474This is equivalent to the ``--nb-ports`` command-line option. 475 476set nbcore 477~~~~~~~~~~ 478 479Set the number of cores used by the application:: 480 481 testpmd> set nbcore (num) 482 483This is equivalent to the ``--nb-cores`` command-line option. 484 485.. note:: 486 487 The number of cores used must not be greater than number of ports used multiplied by the number of queues per port. 488 489set coremask 490~~~~~~~~~~~~ 491 492Set the forwarding cores hexadecimal mask:: 493 494 testpmd> set coremask (mask) 495 496This is equivalent to the ``--coremask`` command-line option. 497 498.. note:: 499 500 The master lcore is reserved for command line parsing only and cannot be masked on for packet forwarding. 501 502set portmask 503~~~~~~~~~~~~ 504 505Set the forwarding ports hexadecimal mask:: 506 507 testpmd> set portmask (mask) 508 509This is equivalent to the ``--portmask`` command-line option. 510 511set burst 512~~~~~~~~~ 513 514Set number of packets per burst:: 515 516 testpmd> set burst (num) 517 518This is equivalent to the ``--burst command-line`` option. 519 520When retry is enabled, the transmit delay time and number of retries can also be set:: 521 522 testpmd> set burst tx delay (microseconds) retry (num) 523 524set txpkts 525~~~~~~~~~~ 526 527Set the length of each segment of the TX-ONLY packets or length of packet for FLOWGEN mode:: 528 529 testpmd> set txpkts (x[,y]*) 530 531Where x[,y]* represents a CSV list of values, without white space. 532 533set txsplit 534~~~~~~~~~~~ 535 536Set the split policy for the TX packets, applicable for TX-ONLY and CSUM forwarding modes:: 537 538 testpmd> set txsplit (off|on|rand) 539 540Where: 541 542* ``off`` disable packet copy & split for CSUM mode. 543 544* ``on`` split outgoing packet into multiple segments. Size of each segment 545 and number of segments per packet is determined by ``set txpkts`` command 546 (see above). 547 548* ``rand`` same as 'on', but number of segments per each packet is a random value between 1 and total number of segments. 549 550set corelist 551~~~~~~~~~~~~ 552 553Set the list of forwarding cores:: 554 555 testpmd> set corelist (x[,y]*) 556 557For example, to change the forwarding cores: 558 559.. code-block:: console 560 561 testpmd> set corelist 3,1 562 testpmd> show config fwd 563 564 io packet forwarding - ports=2 - cores=2 - streams=2 - NUMA support disabled 565 Logical Core 3 (socket 0) forwards packets on 1 streams: 566 RX P=0/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:01 567 Logical Core 1 (socket 0) forwards packets on 1 streams: 568 RX P=1/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00 569 570.. note:: 571 572 The cores are used in the same order as specified on the command line. 573 574set portlist 575~~~~~~~~~~~~ 576 577Set the list of forwarding ports:: 578 579 testpmd> set portlist (x[,y]*) 580 581For example, to change the port forwarding: 582 583.. code-block:: console 584 585 testpmd> set portlist 0,2,1,3 586 testpmd> show config fwd 587 588 io packet forwarding - ports=4 - cores=1 - streams=4 589 Logical Core 3 (socket 0) forwards packets on 4 streams: 590 RX P=0/Q=0 (socket 0) -> TX P=2/Q=0 (socket 0) peer=02:00:00:00:00:01 591 RX P=2/Q=0 (socket 0) -> TX P=0/Q=0 (socket 0) peer=02:00:00:00:00:00 592 RX P=1/Q=0 (socket 0) -> TX P=3/Q=0 (socket 0) peer=02:00:00:00:00:03 593 RX P=3/Q=0 (socket 0) -> TX P=1/Q=0 (socket 0) peer=02:00:00:00:00:02 594 595set tx loopback 596~~~~~~~~~~~~~~~ 597 598Enable/disable tx loopback:: 599 600 testpmd> set tx loopback (port_id) (on|off) 601 602set drop enable 603~~~~~~~~~~~~~~~ 604 605set drop enable bit for all queues:: 606 607 testpmd> set all queues drop (port_id) (on|off) 608 609set split drop enable (for VF) 610~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 611 612set split drop enable bit for VF from PF:: 613 614 testpmd> set vf split drop (port_id) (vf_id) (on|off) 615 616set mac antispoof (for VF) 617~~~~~~~~~~~~~~~~~~~~~~~~~~ 618 619Set mac antispoof for a VF from the PF:: 620 621 testpmd> set vf mac antispoof (port_id) (vf_id) (on|off) 622 623set macsec offload 624~~~~~~~~~~~~~~~~~~ 625 626Enable/disable MACsec offload:: 627 628 testpmd> set macsec offload (port_id) on encrypt (on|off) replay-protect (on|off) 629 testpmd> set macsec offload (port_id) off 630 631set macsec sc 632~~~~~~~~~~~~~ 633 634Configure MACsec secure connection (SC):: 635 636 testpmd> set macsec sc (tx|rx) (port_id) (mac) (pi) 637 638.. note:: 639 640 The pi argument is ignored for tx. 641 Check the NIC Datasheet for hardware limits. 642 643set macsec sa 644~~~~~~~~~~~~~ 645 646Configure MACsec secure association (SA):: 647 648 testpmd> set macsec sa (tx|rx) (port_id) (idx) (an) (pn) (key) 649 650.. note:: 651 652 The IDX value must be 0 or 1. 653 Check the NIC Datasheet for hardware limits. 654 655set broadcast mode (for VF) 656~~~~~~~~~~~~~~~~~~~~~~~~~~~ 657 658Set broadcast mode for a VF from the PF:: 659 660 testpmd> set vf broadcast (port_id) (vf_id) (on|off) 661 662vlan set strip 663~~~~~~~~~~~~~~ 664 665Set the VLAN strip on a port:: 666 667 testpmd> vlan set strip (on|off) (port_id) 668 669vlan set stripq 670~~~~~~~~~~~~~~~ 671 672Set the VLAN strip for a queue on a port:: 673 674 testpmd> vlan set stripq (on|off) (port_id,queue_id) 675 676vlan set stripq (for VF) 677~~~~~~~~~~~~~~~~~~~~~~~~ 678 679Set VLAN strip for all queues in a pool for a VF from the PF:: 680 681 testpmd> set vf vlan stripq (port_id) (vf_id) (on|off) 682 683vlan set insert (for VF) 684~~~~~~~~~~~~~~~~~~~~~~~~ 685 686Set VLAN insert for a VF from the PF:: 687 688 testpmd> set vf vlan insert (port_id) (vf_id) (vlan_id) 689 690vlan set tag (for VF) 691~~~~~~~~~~~~~~~~~~~~~ 692 693Set VLAN tag for a VF from the PF:: 694 695 testpmd> set vf vlan tag (port_id) (vf_id) (on|off) 696 697vlan set antispoof (for VF) 698~~~~~~~~~~~~~~~~~~~~~~~~~~~ 699 700Set VLAN antispoof for a VF from the PF:: 701 702 testpmd> set vf vlan antispoof (port_id) (vf_id) (on|off) 703 704vlan set filter 705~~~~~~~~~~~~~~~ 706 707Set the VLAN filter on a port:: 708 709 testpmd> vlan set filter (on|off) (port_id) 710 711vlan set qinq 712~~~~~~~~~~~~~ 713 714Set the VLAN QinQ (extended queue in queue) on for a port:: 715 716 testpmd> vlan set qinq (on|off) (port_id) 717 718vlan set tpid 719~~~~~~~~~~~~~ 720 721Set the inner or outer VLAN TPID for packet filtering on a port:: 722 723 testpmd> vlan set (inner|outer) tpid (value) (port_id) 724 725.. note:: 726 727 TPID value must be a 16-bit number (value <= 65536). 728 729rx_vlan add 730~~~~~~~~~~~ 731 732Add a VLAN ID, or all identifiers, to the set of VLAN identifiers filtered by port ID:: 733 734 testpmd> rx_vlan add (vlan_id|all) (port_id) 735 736.. note:: 737 738 VLAN filter must be set on that port. VLAN ID < 4096. 739 Depending on the NIC used, number of vlan_ids may be limited to the maximum entries 740 in VFTA table. This is important if enabling all vlan_ids. 741 742rx_vlan rm 743~~~~~~~~~~ 744 745Remove a VLAN ID, or all identifiers, from the set of VLAN identifiers filtered by port ID:: 746 747 testpmd> rx_vlan rm (vlan_id|all) (port_id) 748 749rx_vlan add (for VF) 750~~~~~~~~~~~~~~~~~~~~ 751 752Add a VLAN ID, to the set of VLAN identifiers filtered for VF(s) for port ID:: 753 754 testpmd> rx_vlan add (vlan_id) port (port_id) vf (vf_mask) 755 756rx_vlan rm (for VF) 757~~~~~~~~~~~~~~~~~~~ 758 759Remove a VLAN ID, from the set of VLAN identifiers filtered for VF(s) for port ID:: 760 761 testpmd> rx_vlan rm (vlan_id) port (port_id) vf (vf_mask) 762 763tunnel_filter add 764~~~~~~~~~~~~~~~~~ 765 766Add a tunnel filter on a port:: 767 768 testpmd> tunnel_filter add (port_id) (outer_mac) (inner_mac) (ip_addr) \ 769 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\ 770 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id) 771 772The available information categories are: 773 774* ``vxlan``: Set tunnel type as VXLAN. 775 776* ``nvgre``: Set tunnel type as NVGRE. 777 778* ``ipingre``: Set tunnel type as IP-in-GRE. 779 780* ``imac-ivlan``: Set filter type as Inner MAC and VLAN. 781 782* ``imac-ivlan-tenid``: Set filter type as Inner MAC, VLAN and tenant ID. 783 784* ``imac-tenid``: Set filter type as Inner MAC and tenant ID. 785 786* ``imac``: Set filter type as Inner MAC. 787 788* ``omac-imac-tenid``: Set filter type as Outer MAC, Inner MAC and tenant ID. 789 790* ``oip``: Set filter type as Outer IP. 791 792* ``iip``: Set filter type as Inner IP. 793 794Example:: 795 796 testpmd> tunnel_filter add 0 68:05:CA:28:09:82 00:00:00:00:00:00 \ 797 192.168.2.2 0 ipingre oip 1 1 798 799 Set an IP-in-GRE tunnel on port 0, and the filter type is Outer IP. 800 801tunnel_filter remove 802~~~~~~~~~~~~~~~~~~~~ 803 804Remove a tunnel filter on a port:: 805 806 testpmd> tunnel_filter rm (port_id) (outer_mac) (inner_mac) (ip_addr) \ 807 (inner_vlan) (vxlan|nvgre|ipingre) (imac-ivlan|imac-ivlan-tenid|\ 808 imac-tenid|imac|omac-imac-tenid|oip|iip) (tenant_id) (queue_id) 809 810rx_vxlan_port add 811~~~~~~~~~~~~~~~~~ 812 813Add an UDP port for VXLAN packet filter on a port:: 814 815 testpmd> rx_vxlan_port add (udp_port) (port_id) 816 817rx_vxlan_port remove 818~~~~~~~~~~~~~~~~~~~~ 819 820Remove an UDP port for VXLAN packet filter on a port:: 821 822 testpmd> rx_vxlan_port rm (udp_port) (port_id) 823 824tx_vlan set 825~~~~~~~~~~~ 826 827Set hardware insertion of VLAN IDs in packets sent on a port:: 828 829 testpmd> tx_vlan set (port_id) vlan_id[, vlan_id_outer] 830 831For example, set a single VLAN ID (5) insertion on port 0:: 832 833 tx_vlan set 0 5 834 835Or, set double VLAN ID (inner: 2, outer: 3) insertion on port 1:: 836 837 tx_vlan set 1 2 3 838 839 840tx_vlan set pvid 841~~~~~~~~~~~~~~~~ 842 843Set port based hardware insertion of VLAN ID in packets sent on a port:: 844 845 testpmd> tx_vlan set pvid (port_id) (vlan_id) (on|off) 846 847tx_vlan reset 848~~~~~~~~~~~~~ 849 850Disable hardware insertion of a VLAN header in packets sent on a port:: 851 852 testpmd> tx_vlan reset (port_id) 853 854csum set 855~~~~~~~~ 856 857Select hardware or software calculation of the checksum when 858transmitting a packet using the ``csum`` forwarding engine:: 859 860 testpmd> csum set (ip|udp|tcp|sctp|outer-ip) (hw|sw) (port_id) 861 862Where: 863 864* ``ip|udp|tcp|sctp`` always relate to the inner layer. 865 866* ``outer-ip`` relates to the outer IP layer (only for IPv4) in the case where the packet is recognized 867 as a tunnel packet by the forwarding engine (vxlan, gre and ipip are 868 supported). See also the ``csum parse-tunnel`` command. 869 870.. note:: 871 872 Check the NIC Datasheet for hardware limits. 873 874RSS queue region 875~~~~~~~~~~~~~~~~ 876 877Set RSS queue region span on a port:: 878 879 testpmd> set port (port_id) queue-region region_id (value) \ 880 queue_start_index (value) queue_num (value) 881 882Set flowtype mapping on a RSS queue region on a port:: 883 884 testpmd> set port (port_id) queue-region region_id (value) flowtype (value) 885 886where: 887 888* For the flowtype(pctype) of packet,the specific index for each type has 889 been defined in file i40e_type.h as enum i40e_filter_pctype. 890 891Set user priority mapping on a RSS queue region on a port:: 892 893 testpmd> set port (port_id) queue-region UP (value) region_id (value) 894 895Flush all queue region related configuration on a port:: 896 897 testpmd> set port (port_id) queue-region flush (on|off) 898 899where: 900 901* "on"is just an enable function which server for other configuration, 902 it is for all configuration about queue region from up layer, 903 at first will only keep in DPDK softwarestored in driver, 904 only after "flush on", it commit all configuration to HW. 905 "off" is just clean all configuration about queue region just now, 906 and restore all to DPDK i40e driver default config when start up. 907 908Show all queue region related configuration info on a port:: 909 910 testpmd> show port (port_id) queue-region 911 912.. note:: 913 914 Queue region only support on PF by now, so these command is 915 only for configuration of queue region on PF port. 916 917csum parse-tunnel 918~~~~~~~~~~~~~~~~~ 919 920Define how tunneled packets should be handled by the csum forward 921engine:: 922 923 testpmd> csum parse-tunnel (on|off) (tx_port_id) 924 925If enabled, the csum forward engine will try to recognize supported 926tunnel headers (vxlan, gre, ipip). 927 928If disabled, treat tunnel packets as non-tunneled packets (a inner 929header is handled as a packet payload). 930 931.. note:: 932 933 The port argument is the TX port like in the ``csum set`` command. 934 935Example: 936 937Consider a packet in packet like the following:: 938 939 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in 940 941* If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set`` 942 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the 943 ``outer-ip parameter`` relates to the outer headers (here ``ipv4_out``). 944 945* If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set`` 946 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``. 947 948csum show 949~~~~~~~~~ 950 951Display tx checksum offload configuration:: 952 953 testpmd> csum show (port_id) 954 955tso set 956~~~~~~~ 957 958Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine:: 959 960 testpmd> tso set (segsize) (port_id) 961 962.. note:: 963 964 Check the NIC datasheet for hardware limits. 965 966tso show 967~~~~~~~~ 968 969Display the status of TCP Segmentation Offload:: 970 971 testpmd> tso show (port_id) 972 973set port - gro 974~~~~~~~~~~~~~~ 975 976Enable or disable GRO in ``csum`` forwarding engine:: 977 978 testpmd> set port <port_id> gro on|off 979 980If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4 981packets received from the given port. 982 983If disabled, packets received from the given port won't be performed 984GRO. By default, GRO is disabled for all ports. 985 986.. note:: 987 988 When enable GRO for a port, TCP/IPv4 packets received from the port 989 will be performed GRO. After GRO, all merged packets have bad 990 checksums, since the GRO library doesn't re-calculate checksums for 991 the merged packets. Therefore, if users want the merged packets to 992 have correct checksums, please select HW IP checksum calculation and 993 HW TCP checksum calculation for the port which the merged packets are 994 transmitted to. 995 996show port - gro 997~~~~~~~~~~~~~~~ 998 999Display GRO configuration for a given port:: 1000 1001 testpmd> show port <port_id> gro 1002 1003set gro flush 1004~~~~~~~~~~~~~ 1005 1006Set the cycle to flush the GROed packets from reassembly tables:: 1007 1008 testpmd> set gro flush <cycles> 1009 1010When enable GRO, the csum forwarding engine performs GRO on received 1011packets, and the GROed packets are stored in reassembly tables. Users 1012can use this command to determine when the GROed packets are flushed 1013from the reassembly tables. 1014 1015The ``cycles`` is measured in GRO operation times. The csum forwarding 1016engine flushes the GROed packets from the tables every ``cycles`` GRO 1017operations. 1018 1019By default, the value of ``cycles`` is 1, which means flush GROed packets 1020from the reassembly tables as soon as one GRO operation finishes. The value 1021of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``. 1022 1023Please note that the large value of ``cycles`` may cause the poor TCP/IP 1024stack performance. Because the GROed packets are delayed to arrive the 1025stack, thus causing more duplicated ACKs and TCP retransmissions. 1026 1027set port - gso 1028~~~~~~~~~~~~~~ 1029 1030Toggle per-port GSO support in ``csum`` forwarding engine:: 1031 1032 testpmd> set port <port_id> gso on|off 1033 1034If enabled, the csum forwarding engine will perform GSO on supported IPv4 1035packets, transmitted on the given port. 1036 1037If disabled, packets transmitted on the given port will not undergo GSO. 1038By default, GSO is disabled for all ports. 1039 1040.. note:: 1041 1042 When GSO is enabled on a port, supported IPv4 packets transmitted on that 1043 port undergo GSO. Afterwards, the segmented packets are represented by 1044 multi-segment mbufs; however, the csum forwarding engine doesn't calculation 1045 of checksums for GSO'd segments in SW. As a result, if users want correct 1046 checksums in GSO segments, they should enable HW checksum calculation for 1047 GSO-enabled ports. 1048 1049 For example, HW checksum calculation for VxLAN GSO'd packets may be enabled 1050 by setting the following options in the csum forwarding engine: 1051 1052 testpmd> csum set outer_ip hw <port_id> 1053 1054 testpmd> csum set ip hw <port_id> 1055 1056 testpmd> csum set tcp hw <port_id> 1057 1058 UDP GSO is the same as IP fragmentation, which treats the UDP header 1059 as the payload and does not modify it during segmentation. That is, 1060 after UDP GSO, only the first output fragment has the original UDP 1061 header. Therefore, users need to enable HW IP checksum calculation 1062 and SW UDP checksum calculation for GSO-enabled ports, if they want 1063 correct checksums for UDP/IPv4 packets. 1064 1065set gso segsz 1066~~~~~~~~~~~~~ 1067 1068Set the maximum GSO segment size (measured in bytes), which includes the 1069packet header and the packet payload for GSO-enabled ports (global):: 1070 1071 testpmd> set gso segsz <length> 1072 1073show port - gso 1074~~~~~~~~~~~~~~~ 1075 1076Display the status of Generic Segmentation Offload for a given port:: 1077 1078 testpmd> show port <port_id> gso 1079 1080mac_addr add 1081~~~~~~~~~~~~ 1082 1083Add an alternative MAC address to a port:: 1084 1085 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX) 1086 1087mac_addr remove 1088~~~~~~~~~~~~~~~ 1089 1090Remove a MAC address from a port:: 1091 1092 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX) 1093 1094mac_addr add (for VF) 1095~~~~~~~~~~~~~~~~~~~~~ 1096 1097Add an alternative MAC address for a VF to a port:: 1098 1099 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX) 1100 1101mac_addr set 1102~~~~~~~~~~~~ 1103 1104Set the default MAC address for a port:: 1105 1106 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX) 1107 1108mac_addr set (for VF) 1109~~~~~~~~~~~~~~~~~~~~~ 1110 1111Set the MAC address for a VF from the PF:: 1112 1113 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX) 1114 1115set eth-peer 1116~~~~~~~~~~~~ 1117 1118Set the forwarding peer address for certain port:: 1119 1120 testpmd> set eth-peer (port_id) (perr_addr) 1121 1122This is equivalent to the ``--eth-peer`` command-line option. 1123 1124set port-uta 1125~~~~~~~~~~~~ 1126 1127Set the unicast hash filter(s) on/off for a port:: 1128 1129 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off) 1130 1131set promisc 1132~~~~~~~~~~~ 1133 1134Set the promiscuous mode on for a port or for all ports. 1135In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1136 1137 testpmd> set promisc (port_id|all) (on|off) 1138 1139set allmulti 1140~~~~~~~~~~~~ 1141 1142Set the allmulti mode for a port or for all ports:: 1143 1144 testpmd> set allmulti (port_id|all) (on|off) 1145 1146Same as the ifconfig (8) option. Controls how multicast packets are handled. 1147 1148set promisc (for VF) 1149~~~~~~~~~~~~~~~~~~~~ 1150 1151Set the unicast promiscuous mode for a VF from PF. 1152It's supported by Intel i40e NICs now. 1153In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1154 1155 testpmd> set vf promisc (port_id) (vf_id) (on|off) 1156 1157set allmulticast (for VF) 1158~~~~~~~~~~~~~~~~~~~~~~~~~ 1159 1160Set the multicast promiscuous mode for a VF from PF. 1161It's supported by Intel i40e NICs now. 1162In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1163 1164 testpmd> set vf allmulti (port_id) (vf_id) (on|off) 1165 1166set tx max bandwidth (for VF) 1167~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1168 1169Set TX max absolute bandwidth (Mbps) for a VF from PF:: 1170 1171 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth) 1172 1173set tc tx min bandwidth (for VF) 1174~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1175 1176Set all TCs' TX min relative bandwidth (%) for a VF from PF:: 1177 1178 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...) 1179 1180set tc tx max bandwidth (for VF) 1181~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1182 1183Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF:: 1184 1185 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth) 1186 1187set tc strict link priority mode 1188~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1189 1190Set some TCs' strict link priority mode on a physical port:: 1191 1192 testpmd> set tx strict-link-priority (port_id) (tc_bitmap) 1193 1194set tc tx min bandwidth 1195~~~~~~~~~~~~~~~~~~~~~~~ 1196 1197Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs:: 1198 1199 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...) 1200 1201set flow_ctrl rx 1202~~~~~~~~~~~~~~~~ 1203 1204Set the link flow control parameter on a port:: 1205 1206 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 1207 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \ 1208 autoneg (on|off) (port_id) 1209 1210Where: 1211 1212* ``high_water`` (integer): High threshold value to trigger XOFF. 1213 1214* ``low_water`` (integer): Low threshold value to trigger XON. 1215 1216* ``pause_time`` (integer): Pause quota in the Pause frame. 1217 1218* ``send_xon`` (0/1): Send XON frame. 1219 1220* ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames. 1221 1222* ``autoneg``: Change the auto-negotiation parameter. 1223 1224set pfc_ctrl rx 1225~~~~~~~~~~~~~~~ 1226 1227Set the priority flow control parameter on a port:: 1228 1229 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 1230 (pause_time) (priority) (port_id) 1231 1232Where: 1233 1234* ``high_water`` (integer): High threshold value. 1235 1236* ``low_water`` (integer): Low threshold value. 1237 1238* ``pause_time`` (integer): Pause quota in the Pause frame. 1239 1240* ``priority`` (0-7): VLAN User Priority. 1241 1242set stat_qmap 1243~~~~~~~~~~~~~ 1244 1245Set statistics mapping (qmapping 0..15) for RX/TX queue on port:: 1246 1247 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping) 1248 1249For example, to set rx queue 2 on port 0 to mapping 5:: 1250 1251 testpmd>set stat_qmap rx 0 2 5 1252 1253set xstats-hide-zero 1254~~~~~~~~~~~~~~~~~~~~ 1255 1256Set the option to hide zero values for xstats display:: 1257 1258 testpmd> set xstats-hide-zero on|off 1259 1260.. note:: 1261 1262 By default, the zero values are displayed for xstats. 1263 1264set port - rx/tx (for VF) 1265~~~~~~~~~~~~~~~~~~~~~~~~~ 1266 1267Set VF receive/transmit from a port:: 1268 1269 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off) 1270 1271set port - mac address filter (for VF) 1272~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1273 1274Add/Remove unicast or multicast MAC addr filter for a VF:: 1275 1276 testpmd> set port (port_id) vf (vf_id) (mac_addr) \ 1277 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off) 1278 1279set port - rx mode(for VF) 1280~~~~~~~~~~~~~~~~~~~~~~~~~~ 1281 1282Set the VF receive mode of a port:: 1283 1284 testpmd> set port (port_id) vf (vf_id) \ 1285 rxmode (AUPE|ROPE|BAM|MPE) (on|off) 1286 1287The available receive modes are: 1288 1289* ``AUPE``: Accepts untagged VLAN. 1290 1291* ``ROPE``: Accepts unicast hash. 1292 1293* ``BAM``: Accepts broadcast packets. 1294 1295* ``MPE``: Accepts all multicast packets. 1296 1297set port - tx_rate (for Queue) 1298~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1299 1300Set TX rate limitation for a queue on a port:: 1301 1302 testpmd> set port (port_id) queue (queue_id) rate (rate_value) 1303 1304set port - tx_rate (for VF) 1305~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1306 1307Set TX rate limitation for queues in VF on a port:: 1308 1309 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask) 1310 1311set port - mirror rule 1312~~~~~~~~~~~~~~~~~~~~~~ 1313 1314Set pool or vlan type mirror rule for a port:: 1315 1316 testpmd> set port (port_id) mirror-rule (rule_id) \ 1317 (pool-mirror-up|pool-mirror-down|vlan-mirror) \ 1318 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off) 1319 1320Set link mirror rule for a port:: 1321 1322 testpmd> set port (port_id) mirror-rule (rule_id) \ 1323 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off) 1324 1325For example to enable mirror traffic with vlan 0,1 to pool 0:: 1326 1327 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on 1328 1329reset port - mirror rule 1330~~~~~~~~~~~~~~~~~~~~~~~~ 1331 1332Reset a mirror rule for a port:: 1333 1334 testpmd> reset port (port_id) mirror-rule (rule_id) 1335 1336set flush_rx 1337~~~~~~~~~~~~ 1338 1339Set the flush on RX streams before forwarding. 1340The default is flush ``on``. 1341Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams:: 1342 1343 testpmd> set flush_rx off 1344 1345set bypass mode 1346~~~~~~~~~~~~~~~ 1347 1348Set the bypass mode for the lowest port on bypass enabled NIC:: 1349 1350 testpmd> set bypass mode (normal|bypass|isolate) (port_id) 1351 1352set bypass event 1353~~~~~~~~~~~~~~~~ 1354 1355Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled:: 1356 1357 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \ 1358 mode (normal|bypass|isolate) (port_id) 1359 1360Where: 1361 1362* ``timeout``: Enable bypass after watchdog timeout. 1363 1364* ``os_on``: Enable bypass when OS/board is powered on. 1365 1366* ``os_off``: Enable bypass when OS/board is powered off. 1367 1368* ``power_on``: Enable bypass when power supply is turned on. 1369 1370* ``power_off``: Enable bypass when power supply is turned off. 1371 1372 1373set bypass timeout 1374~~~~~~~~~~~~~~~~~~ 1375 1376Set the bypass watchdog timeout to ``n`` seconds where 0 = instant:: 1377 1378 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32) 1379 1380show bypass config 1381~~~~~~~~~~~~~~~~~~ 1382 1383Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC:: 1384 1385 testpmd> show bypass config (port_id) 1386 1387set link up 1388~~~~~~~~~~~ 1389 1390Set link up for a port:: 1391 1392 testpmd> set link-up port (port id) 1393 1394set link down 1395~~~~~~~~~~~~~ 1396 1397Set link down for a port:: 1398 1399 testpmd> set link-down port (port id) 1400 1401E-tag set 1402~~~~~~~~~ 1403 1404Enable E-tag insertion for a VF on a port:: 1405 1406 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id) 1407 1408Disable E-tag insertion for a VF on a port:: 1409 1410 testpmd> E-tag set insertion off port (port_id) vf (vf_id) 1411 1412Enable/disable E-tag stripping on a port:: 1413 1414 testpmd> E-tag set stripping (on|off) port (port_id) 1415 1416Enable/disable E-tag based forwarding on a port:: 1417 1418 testpmd> E-tag set forwarding (on|off) port (port_id) 1419 1420Add an E-tag forwarding filter on a port:: 1421 1422 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id) 1423 1424Delete an E-tag forwarding filter on a port:: 1425 testpmd> E-tag set filter del e-tag-id (value) port (port_id) 1426 1427ddp add 1428~~~~~~~ 1429 1430Load a dynamic device personalization (DDP) profile and store backup profile:: 1431 1432 testpmd> ddp add (port_id) (profile_path[,backup_profile_path]) 1433 1434ddp del 1435~~~~~~~ 1436 1437Delete a dynamic device personalization profile and restore backup profile:: 1438 1439 testpmd> ddp del (port_id) (backup_profile_path) 1440 1441ptype mapping 1442~~~~~~~~~~~~~ 1443 1444List all items from the ptype mapping table:: 1445 1446 testpmd> ptype mapping get (port_id) (valid_only) 1447 1448Where: 1449 1450* ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0). 1451 1452Replace a specific or a group of software defined ptype with a new one:: 1453 1454 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type) 1455 1456where: 1457 1458* ``target``: A specific software ptype or a mask to represent a group of software ptypes. 1459 1460* ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1). 1461 1462* ``pkt_type``: The new software ptype to replace the old ones. 1463 1464Update hardware defined ptype to software defined packet type mapping table:: 1465 1466 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype) 1467 1468where: 1469 1470* ``hw_ptype``: hardware ptype as the index of the ptype mapping table. 1471 1472* ``sw_ptype``: software ptype as the value of the ptype mapping table. 1473 1474Reset ptype mapping table:: 1475 1476 testpmd> ptype mapping reset (port_id) 1477 1478config per port Rx offloading 1479~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1480 1481Enable or disable a per port Rx offloading on all Rx queues of a port:: 1482 1483 testpmd> port config (port_id) rx_offload (offloading) on|off 1484 1485* ``offloading``: can be any of these offloading capability: 1486 vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro, 1487 qinq_strip, outer_ipv4_cksum, macsec_strip, 1488 header_split, vlan_filter, vlan_extend, jumbo_frame, 1489 crc_strip, scatter, timestamp, security, keep_crc 1490 1491This command should be run when the port is stopped, or else it will fail. 1492 1493config per queue Rx offloading 1494~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1495 1496Enable or disable a per queue Rx offloading only on a specific Rx queue:: 1497 1498 testpmd> port (port_id) rxq (queue_id) rx_offload (offloading) on|off 1499 1500* ``offloading``: can be any of these offloading capability: 1501 vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro, 1502 qinq_strip, outer_ipv4_cksum, macsec_strip, 1503 header_split, vlan_filter, vlan_extend, jumbo_frame, 1504 crc_strip, scatter, timestamp, security, keep_crc 1505 1506This command should be run when the port is stopped, or else it will fail. 1507 1508config per port Tx offloading 1509~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1510 1511Enable or disable a per port Tx offloading on all Tx queues of a port:: 1512 1513 testpmd> port config (port_id) tx_offload (offloading) on|off 1514 1515* ``offloading``: can be any of these offloading capability: 1516 vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum, 1517 sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum, 1518 qinq_insert, vxlan_tnl_tso, gre_tnl_tso, 1519 ipip_tnl_tso, geneve_tnl_tso, macsec_insert, 1520 mt_lockfree, multi_segs, mbuf_fast_free, security 1521 1522This command should be run when the port is stopped, or else it will fail. 1523 1524config per queue Tx offloading 1525~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1526 1527Enable or disable a per queue Tx offloading only on a specific Tx queue:: 1528 1529 testpmd> port (port_id) txq (queue_id) tx_offload (offloading) on|off 1530 1531* ``offloading``: can be any of these offloading capability: 1532 vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum, 1533 sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum, 1534 qinq_insert, vxlan_tnl_tso, gre_tnl_tso, 1535 ipip_tnl_tso, geneve_tnl_tso, macsec_insert, 1536 mt_lockfree, multi_segs, mbuf_fast_free, security 1537 1538This command should be run when the port is stopped, or else it will fail. 1539 1540Config VXLAN Encap outer layers 1541~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1542 1543Configure the outer layer to encapsulate a packet inside a VXLAN tunnel:: 1544 1545 set vxlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \ 1546 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) eth-src (eth-src) \ 1547 eth-dst (eth-dst) 1548 1549 set vxlan-with-vlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \ 1550 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) vlan-tci (vlan-tci) \ 1551 eth-src (eth-src) eth-dst (eth-dst) 1552 1553Those command will set an internal configuration inside testpmd, any following 1554flow rule using the action vxlan_encap will use the last configuration set. 1555To have a different encapsulation header, one of those commands must be called 1556before the flow rule creation. 1557 1558Config NVGRE Encap outer layers 1559~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1560 1561Configure the outer layer to encapsulate a packet inside a NVGRE tunnel:: 1562 1563 set nvgre ip-version (ipv4|ipv6) tni (tni) ip-src (ip-src) ip-dst (ip-dst) \ 1564 eth-src (eth-src) eth-dst (eth-dst) 1565 set nvgre-with-vlan ip-version (ipv4|ipv6) tni (tni) ip-src (ip-src) \ 1566 ip-dst (ip-dst) vlan-tci (vlan-tci) eth-src (eth-src) eth-dst (eth-dst) 1567 1568Those command will set an internal configuration inside testpmd, any following 1569flow rule using the action nvgre_encap will use the last configuration set. 1570To have a different encapsulation header, one of those commands must be called 1571before the flow rule creation. 1572 1573Port Functions 1574-------------- 1575 1576The following sections show functions for configuring ports. 1577 1578.. note:: 1579 1580 Port configuration changes only become active when forwarding is started/restarted. 1581 1582port attach 1583~~~~~~~~~~~ 1584 1585Attach a port specified by pci address or virtual device args:: 1586 1587 testpmd> port attach (identifier) 1588 1589To attach a new pci device, the device should be recognized by kernel first. 1590Then it should be moved under DPDK management. 1591Finally the port can be attached to testpmd. 1592 1593For example, to move a pci device using ixgbe under DPDK management: 1594 1595.. code-block:: console 1596 1597 # Check the status of the available devices. 1598 ./usertools/dpdk-devbind.py --status 1599 1600 Network devices using DPDK-compatible driver 1601 ============================================ 1602 <none> 1603 1604 Network devices using kernel driver 1605 =================================== 1606 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused= 1607 1608 1609 # Bind the device to igb_uio. 1610 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0 1611 1612 1613 # Recheck the status of the devices. 1614 ./usertools/dpdk-devbind.py --status 1615 Network devices using DPDK-compatible driver 1616 ============================================ 1617 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused= 1618 1619To attach a port created by virtual device, above steps are not needed. 1620 1621For example, to attach a port whose pci address is 0000:0a:00.0. 1622 1623.. code-block:: console 1624 1625 testpmd> port attach 0000:0a:00.0 1626 Attaching a new port... 1627 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1628 EAL: probe driver: 8086:10fb rte_ixgbe_pmd 1629 EAL: PCI memory mapped at 0x7f83bfa00000 1630 EAL: PCI memory mapped at 0x7f83bfa80000 1631 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5 1632 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb 1633 Port 0 is attached. Now total ports is 1 1634 Done 1635 1636For example, to attach a port created by pcap PMD. 1637 1638.. code-block:: console 1639 1640 testpmd> port attach net_pcap0 1641 Attaching a new port... 1642 PMD: Initializing pmd_pcap for net_pcap0 1643 PMD: Creating pcap-backed ethdev on numa socket 0 1644 Port 0 is attached. Now total ports is 1 1645 Done 1646 1647In this case, identifier is ``net_pcap0``. 1648This identifier format is the same as ``--vdev`` format of DPDK applications. 1649 1650For example, to re-attach a bonded port which has been previously detached, 1651the mode and slave parameters must be given. 1652 1653.. code-block:: console 1654 1655 testpmd> port attach net_bond_0,mode=0,slave=1 1656 Attaching a new port... 1657 EAL: Initializing pmd_bond for net_bond_0 1658 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0. 1659 Port 0 is attached. Now total ports is 1 1660 Done 1661 1662 1663port detach 1664~~~~~~~~~~~ 1665 1666Detach a specific port:: 1667 1668 testpmd> port detach (port_id) 1669 1670Before detaching a port, the port should be stopped and closed. 1671 1672For example, to detach a pci device port 0. 1673 1674.. code-block:: console 1675 1676 testpmd> port stop 0 1677 Stopping ports... 1678 Done 1679 testpmd> port close 0 1680 Closing ports... 1681 Done 1682 1683 testpmd> port detach 0 1684 Detaching a port... 1685 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1686 EAL: remove driver: 8086:10fb rte_ixgbe_pmd 1687 EAL: PCI memory unmapped at 0x7f83bfa00000 1688 EAL: PCI memory unmapped at 0x7f83bfa80000 1689 Done 1690 1691 1692For example, to detach a virtual device port 0. 1693 1694.. code-block:: console 1695 1696 testpmd> port stop 0 1697 Stopping ports... 1698 Done 1699 testpmd> port close 0 1700 Closing ports... 1701 Done 1702 1703 testpmd> port detach 0 1704 Detaching a port... 1705 PMD: Closing pcap ethdev on numa socket 0 1706 Port 'net_pcap0' is detached. Now total ports is 0 1707 Done 1708 1709To remove a pci device completely from the system, first detach the port from testpmd. 1710Then the device should be moved under kernel management. 1711Finally the device can be removed using kernel pci hotplug functionality. 1712 1713For example, to move a pci device under kernel management: 1714 1715.. code-block:: console 1716 1717 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0 1718 1719 ./usertools/dpdk-devbind.py --status 1720 1721 Network devices using DPDK-compatible driver 1722 ============================================ 1723 <none> 1724 1725 Network devices using kernel driver 1726 =================================== 1727 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio 1728 1729To remove a port created by a virtual device, above steps are not needed. 1730 1731port start 1732~~~~~~~~~~ 1733 1734Start all ports or a specific port:: 1735 1736 testpmd> port start (port_id|all) 1737 1738port stop 1739~~~~~~~~~ 1740 1741Stop all ports or a specific port:: 1742 1743 testpmd> port stop (port_id|all) 1744 1745port close 1746~~~~~~~~~~ 1747 1748Close all ports or a specific port:: 1749 1750 testpmd> port close (port_id|all) 1751 1752port config - queue ring size 1753~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1754 1755Configure a rx/tx queue ring size:: 1756 1757 testpmd> port (port_id) (rxq|txq) (queue_id) ring_size (value) 1758 1759Only take effect after command that (re-)start the port or command that setup specific queue. 1760 1761port start/stop queue 1762~~~~~~~~~~~~~~~~~~~~~ 1763 1764Start/stop a rx/tx queue on a specific port:: 1765 1766 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop) 1767 1768port setup queue 1769~~~~~~~~~~~~~~~~~~~~~ 1770 1771Setup a rx/tx queue on a specific port:: 1772 1773 testpmd> port (port_id) (rxq|txq) (queue_id) setup 1774 1775Only take effect when port is started. 1776 1777port config - speed 1778~~~~~~~~~~~~~~~~~~~ 1779 1780Set the speed and duplex mode for all ports or a specific port:: 1781 1782 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \ 1783 duplex (half|full|auto) 1784 1785port config - queues/descriptors 1786~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1787 1788Set number of queues/descriptors for rxq, txq, rxd and txd:: 1789 1790 testpmd> port config all (rxq|txq|rxd|txd) (value) 1791 1792This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options. 1793 1794port config - max-pkt-len 1795~~~~~~~~~~~~~~~~~~~~~~~~~ 1796 1797Set the maximum packet length:: 1798 1799 testpmd> port config all max-pkt-len (value) 1800 1801This is equivalent to the ``--max-pkt-len`` command-line option. 1802 1803port config - CRC Strip 1804~~~~~~~~~~~~~~~~~~~~~~~ 1805 1806Set hardware CRC stripping on or off for all ports:: 1807 1808 testpmd> port config all crc-strip (on|off) 1809 1810CRC stripping is on by default. 1811 1812The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option. 1813 1814port config - scatter 1815~~~~~~~~~~~~~~~~~~~~~~~ 1816 1817Set RX scatter mode on or off for all ports:: 1818 1819 testpmd> port config all scatter (on|off) 1820 1821RX scatter mode is off by default. 1822 1823The ``on`` option is equivalent to the ``--enable-scatter`` command-line option. 1824 1825port config - RX Checksum 1826~~~~~~~~~~~~~~~~~~~~~~~~~ 1827 1828Set hardware RX checksum offload to on or off for all ports:: 1829 1830 testpmd> port config all rx-cksum (on|off) 1831 1832Checksum offload is off by default. 1833 1834The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option. 1835 1836port config - VLAN 1837~~~~~~~~~~~~~~~~~~ 1838 1839Set hardware VLAN on or off for all ports:: 1840 1841 testpmd> port config all hw-vlan (on|off) 1842 1843Hardware VLAN is off by default. 1844 1845The ``on`` option is equivalent to the ``--enable-hw-vlan`` command-line option. 1846 1847port config - VLAN filter 1848~~~~~~~~~~~~~~~~~~~~~~~~~ 1849 1850Set hardware VLAN filter on or off for all ports:: 1851 1852 testpmd> port config all hw-vlan-filter (on|off) 1853 1854Hardware VLAN filter is off by default. 1855 1856The ``on`` option is equivalent to the ``--enable-hw-vlan-filter`` command-line option. 1857 1858port config - VLAN strip 1859~~~~~~~~~~~~~~~~~~~~~~~~ 1860 1861Set hardware VLAN strip on or off for all ports:: 1862 1863 testpmd> port config all hw-vlan-strip (on|off) 1864 1865Hardware VLAN strip is off by default. 1866 1867The ``on`` option is equivalent to the ``--enable-hw-vlan-strip`` command-line option. 1868 1869port config - VLAN extend 1870~~~~~~~~~~~~~~~~~~~~~~~~~ 1871 1872Set hardware VLAN extend on or off for all ports:: 1873 1874 testpmd> port config all hw-vlan-extend (on|off) 1875 1876Hardware VLAN extend is off by default. 1877 1878The ``on`` option is equivalent to the ``--enable-hw-vlan-extend`` command-line option. 1879 1880port config - Drop Packets 1881~~~~~~~~~~~~~~~~~~~~~~~~~~ 1882 1883Set packet drop for packets with no descriptors on or off for all ports:: 1884 1885 testpmd> port config all drop-en (on|off) 1886 1887Packet dropping for packets with no descriptors is off by default. 1888 1889The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option. 1890 1891port config - RSS 1892~~~~~~~~~~~~~~~~~ 1893 1894Set the RSS (Receive Side Scaling) mode on or off:: 1895 1896 testpmd> port config all rss (all|default|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none) 1897 1898RSS is on by default. 1899 1900The ``all`` option is equivalent to ip|tcp|udp|sctp|ether. 1901The ``default`` option enables all supported RSS types reported by device info. 1902The ``none`` option is equivalent to the ``--disable-rss`` command-line option. 1903 1904port config - RSS Reta 1905~~~~~~~~~~~~~~~~~~~~~~ 1906 1907Set the RSS (Receive Side Scaling) redirection table:: 1908 1909 testpmd> port config all rss reta (hash,queue)[,(hash,queue)] 1910 1911port config - DCB 1912~~~~~~~~~~~~~~~~~ 1913 1914Set the DCB mode for an individual port:: 1915 1916 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off) 1917 1918The traffic class should be 4 or 8. 1919 1920port config - Burst 1921~~~~~~~~~~~~~~~~~~~ 1922 1923Set the number of packets per burst:: 1924 1925 testpmd> port config all burst (value) 1926 1927This is equivalent to the ``--burst`` command-line option. 1928 1929port config - Threshold 1930~~~~~~~~~~~~~~~~~~~~~~~ 1931 1932Set thresholds for TX/RX queues:: 1933 1934 testpmd> port config all (threshold) (value) 1935 1936Where the threshold type can be: 1937 1938* ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255. 1939 1940* ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255. 1941 1942* ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255. 1943 1944* ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255. 1945 1946* ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255. 1947 1948* ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255. 1949 1950* ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd. 1951 1952* ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd. 1953 1954* ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd. 1955 1956These threshold options are also available from the command-line. 1957 1958port config - E-tag 1959~~~~~~~~~~~~~~~~~~~ 1960 1961Set the value of ether-type for E-tag:: 1962 1963 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value) 1964 1965Enable/disable the E-tag support:: 1966 1967 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable) 1968 1969port config pctype mapping 1970~~~~~~~~~~~~~~~~~~~~~~~~~~ 1971 1972Reset pctype mapping table:: 1973 1974 testpmd> port config (port_id) pctype mapping reset 1975 1976Update hardware defined pctype to software defined flow type mapping table:: 1977 1978 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id) 1979 1980where: 1981 1982* ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table. 1983 1984* ``flow_type_id``: software flow type id as the index of the pctype mapping table. 1985 1986port config input set 1987~~~~~~~~~~~~~~~~~~~~~ 1988 1989Config RSS/FDIR/FDIR flexible payload input set for some pctype:: 1990 testpmd> port config (port_id) pctype (pctype_id) \ 1991 (hash_inset|fdir_inset|fdir_flx_inset) \ 1992 (get|set|clear) field (field_idx) 1993 1994Clear RSS/FDIR/FDIR flexible payload input set for some pctype:: 1995 testpmd> port config (port_id) pctype (pctype_id) \ 1996 (hash_inset|fdir_inset|fdir_flx_inset) clear all 1997 1998where: 1999 2000* ``pctype_id``: hardware packet classification types. 2001* ``field_idx``: hardware field index. 2002 2003port config udp_tunnel_port 2004~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2005 2006Add/remove UDP tunnel port for VXLAN/GENEVE tunneling protocols:: 2007 testpmd> port config (port_id) udp_tunnel_port add|rm vxlan|geneve (udp_port) 2008 2009Link Bonding Functions 2010---------------------- 2011 2012The Link Bonding functions make it possible to dynamically create and 2013manage link bonding devices from within testpmd interactive prompt. 2014 2015create bonded device 2016~~~~~~~~~~~~~~~~~~~~ 2017 2018Create a new bonding device:: 2019 2020 testpmd> create bonded device (mode) (socket) 2021 2022For example, to create a bonded device in mode 1 on socket 0:: 2023 2024 testpmd> create bonded device 1 0 2025 created new bonded device (port X) 2026 2027add bonding slave 2028~~~~~~~~~~~~~~~~~ 2029 2030Adds Ethernet device to a Link Bonding device:: 2031 2032 testpmd> add bonding slave (slave id) (port id) 2033 2034For example, to add Ethernet device (port 6) to a Link Bonding device (port 10):: 2035 2036 testpmd> add bonding slave 6 10 2037 2038 2039remove bonding slave 2040~~~~~~~~~~~~~~~~~~~~ 2041 2042Removes an Ethernet slave device from a Link Bonding device:: 2043 2044 testpmd> remove bonding slave (slave id) (port id) 2045 2046For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10):: 2047 2048 testpmd> remove bonding slave 6 10 2049 2050set bonding mode 2051~~~~~~~~~~~~~~~~ 2052 2053Set the Link Bonding mode of a Link Bonding device:: 2054 2055 testpmd> set bonding mode (value) (port id) 2056 2057For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3):: 2058 2059 testpmd> set bonding mode 3 10 2060 2061set bonding primary 2062~~~~~~~~~~~~~~~~~~~ 2063 2064Set an Ethernet slave device as the primary device on a Link Bonding device:: 2065 2066 testpmd> set bonding primary (slave id) (port id) 2067 2068For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10):: 2069 2070 testpmd> set bonding primary 6 10 2071 2072set bonding mac 2073~~~~~~~~~~~~~~~ 2074 2075Set the MAC address of a Link Bonding device:: 2076 2077 testpmd> set bonding mac (port id) (mac) 2078 2079For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01:: 2080 2081 testpmd> set bonding mac 10 00:00:00:00:00:01 2082 2083set bonding xmit_balance_policy 2084~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2085 2086Set the transmission policy for a Link Bonding device when it is in Balance XOR mode:: 2087 2088 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34) 2089 2090For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports):: 2091 2092 testpmd> set bonding xmit_balance_policy 10 l34 2093 2094 2095set bonding mon_period 2096~~~~~~~~~~~~~~~~~~~~~~ 2097 2098Set the link status monitoring polling period in milliseconds for a bonding device. 2099 2100This adds support for PMD slave devices which do not support link status interrupts. 2101When the mon_period is set to a value greater than 0 then all PMD's which do not support 2102link status ISR will be queried every polling interval to check if their link status has changed:: 2103 2104 testpmd> set bonding mon_period (port_id) (value) 2105 2106For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms:: 2107 2108 testpmd> set bonding mon_period 5 150 2109 2110 2111set bonding lacp dedicated_queue 2112~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2113 2114Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic 2115when in mode 4 (link-aggregration-802.3ad):: 2116 2117 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable) 2118 2119 2120set bonding agg_mode 2121~~~~~~~~~~~~~~~~~~~~ 2122 2123Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad):: 2124 2125 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable) 2126 2127 2128show bonding config 2129~~~~~~~~~~~~~~~~~~~ 2130 2131Show the current configuration of a Link Bonding device:: 2132 2133 testpmd> show bonding config (port id) 2134 2135For example, 2136to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4) 2137in balance mode with a transmission policy of layer 2+3:: 2138 2139 testpmd> show bonding config 9 2140 Bonding mode: 2 2141 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23 2142 Slaves (3): [1 3 4] 2143 Active Slaves (3): [1 3 4] 2144 Primary: [3] 2145 2146 2147Register Functions 2148------------------ 2149 2150The Register Functions can be used to read from and write to registers on the network card referenced by a port number. 2151This is mainly useful for debugging purposes. 2152Reference should be made to the appropriate datasheet for the network card for details on the register addresses 2153and fields that can be accessed. 2154 2155read reg 2156~~~~~~~~ 2157 2158Display the value of a port register:: 2159 2160 testpmd> read reg (port_id) (address) 2161 2162For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller:: 2163 2164 testpmd> read reg 0 0xEE00 2165 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241) 2166 2167read regfield 2168~~~~~~~~~~~~~ 2169 2170Display a port register bit field:: 2171 2172 testpmd> read regfield (port_id) (address) (bit_x) (bit_y) 2173 2174For example, reading the lowest two bits from the register in the example above:: 2175 2176 testpmd> read regfield 0 0xEE00 0 1 2177 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1) 2178 2179read regbit 2180~~~~~~~~~~~ 2181 2182Display a single port register bit:: 2183 2184 testpmd> read regbit (port_id) (address) (bit_x) 2185 2186For example, reading the lowest bit from the register in the example above:: 2187 2188 testpmd> read regbit 0 0xEE00 0 2189 port 0 PCI register at offset 0xEE00: bit 0=1 2190 2191write reg 2192~~~~~~~~~ 2193 2194Set the value of a port register:: 2195 2196 testpmd> write reg (port_id) (address) (value) 2197 2198For example, to clear a register:: 2199 2200 testpmd> write reg 0 0xEE00 0x0 2201 port 0 PCI register at offset 0xEE00: 0x00000000 (0) 2202 2203write regfield 2204~~~~~~~~~~~~~~ 2205 2206Set bit field of a port register:: 2207 2208 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value) 2209 2210For example, writing to the register cleared in the example above:: 2211 2212 testpmd> write regfield 0 0xEE00 0 1 2 2213 port 0 PCI register at offset 0xEE00: 0x00000002 (2) 2214 2215write regbit 2216~~~~~~~~~~~~ 2217 2218Set single bit value of a port register:: 2219 2220 testpmd> write regbit (port_id) (address) (bit_x) (value) 2221 2222For example, to set the high bit in the register from the example above:: 2223 2224 testpmd> write regbit 0 0xEE00 31 1 2225 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658) 2226 2227Traffic Metering and Policing 2228----------------------------- 2229 2230The following section shows functions for configuring traffic metering and 2231policing on the ethernet device through the use of generic ethdev API. 2232 2233show port traffic management capability 2234~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2235 2236Show traffic metering and policing capability of the port:: 2237 2238 testpmd> show port meter cap (port_id) 2239 2240add port meter profile (srTCM rfc2967) 2241~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2242 2243Add meter profile (srTCM rfc2697) to the ethernet device:: 2244 2245 testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \ 2246 (cir) (cbs) (ebs) 2247 2248where: 2249 2250* ``profile_id``: ID for the meter profile. 2251* ``cir``: Committed Information Rate (CIR) (bytes/second). 2252* ``cbs``: Committed Burst Size (CBS) (bytes). 2253* ``ebs``: Excess Burst Size (EBS) (bytes). 2254 2255add port meter profile (trTCM rfc2968) 2256~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2257 2258Add meter profile (srTCM rfc2698) to the ethernet device:: 2259 2260 testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \ 2261 (cir) (pir) (cbs) (pbs) 2262 2263where: 2264 2265* ``profile_id``: ID for the meter profile. 2266* ``cir``: Committed information rate (bytes/second). 2267* ``pir``: Peak information rate (bytes/second). 2268* ``cbs``: Committed burst size (bytes). 2269* ``pbs``: Peak burst size (bytes). 2270 2271add port meter profile (trTCM rfc4115) 2272~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2273 2274Add meter profile (trTCM rfc4115) to the ethernet device:: 2275 2276 testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \ 2277 (cir) (eir) (cbs) (ebs) 2278 2279where: 2280 2281* ``profile_id``: ID for the meter profile. 2282* ``cir``: Committed information rate (bytes/second). 2283* ``eir``: Excess information rate (bytes/second). 2284* ``cbs``: Committed burst size (bytes). 2285* ``ebs``: Excess burst size (bytes). 2286 2287delete port meter profile 2288~~~~~~~~~~~~~~~~~~~~~~~~~ 2289 2290Delete meter profile from the ethernet device:: 2291 2292 testpmd> del port meter profile (port_id) (profile_id) 2293 2294create port meter 2295~~~~~~~~~~~~~~~~~ 2296 2297Create new meter object for the ethernet device:: 2298 2299 testpmd> create port meter (port_id) (mtr_id) (profile_id) \ 2300 (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \ 2301 (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\ 2302 (dscp_tbl_entry63)] 2303 2304where: 2305 2306* ``mtr_id``: meter object ID. 2307* ``profile_id``: ID for the meter profile. 2308* ``meter_enable``: When this parameter has a non-zero value, the meter object 2309 gets enabled at the time of creation, otherwise remains disabled. 2310* ``g_action``: Policer action for the packet with green color. 2311* ``y_action``: Policer action for the packet with yellow color. 2312* ``r_action``: Policer action for the packet with red color. 2313* ``stats_mask``: Mask of statistics counter types to be enabled for the 2314 meter object. 2315* ``shared``: When this parameter has a non-zero value, the meter object is 2316 shared by multiple flows. Otherwise, meter object is used by single flow. 2317* ``use_pre_meter_color``: When this parameter has a non-zero value, the 2318 input color for the current meter object is determined by the latest meter 2319 object in the same flow. Otherwise, the current meter object uses the 2320 *dscp_table* to determine the input color. 2321* ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input 2322 color, 0 <= x <= 63. 2323 2324enable port meter 2325~~~~~~~~~~~~~~~~~ 2326 2327Enable meter for the ethernet device:: 2328 2329 testpmd> enable port meter (port_id) (mtr_id) 2330 2331disable port meter 2332~~~~~~~~~~~~~~~~~~ 2333 2334Disable meter for the ethernet device:: 2335 2336 testpmd> disable port meter (port_id) (mtr_id) 2337 2338delete port meter 2339~~~~~~~~~~~~~~~~~ 2340 2341Delete meter for the ethernet device:: 2342 2343 testpmd> del port meter (port_id) (mtr_id) 2344 2345Set port meter profile 2346~~~~~~~~~~~~~~~~~~~~~~ 2347 2348Set meter profile for the ethernet device:: 2349 2350 testpmd> set port meter profile (port_id) (mtr_id) (profile_id) 2351 2352set port meter dscp table 2353~~~~~~~~~~~~~~~~~~~~~~~~~ 2354 2355Set meter dscp table for the ethernet device:: 2356 2357 testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \ 2358 (dscp_tbl_entry1)...(dscp_tbl_entry63)] 2359 2360set port meter policer action 2361~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2362 2363Set meter policer action for the ethernet device:: 2364 2365 testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \ 2366 (action0) [(action1) (action1)] 2367 2368where: 2369 2370* ``action_mask``: Bit mask indicating which policer actions need to be 2371 updated. One or more policer actions can be updated in a single function 2372 invocation. To update the policer action associated with color C, bit 2373 (1 << C) needs to be set in *action_mask* and element at position C 2374 in the *actions* array needs to be valid. 2375* ``actionx``: Policer action for the color x, 2376 RTE_MTR_GREEN <= x < RTE_MTR_COLORS 2377 2378set port meter stats mask 2379~~~~~~~~~~~~~~~~~~~~~~~~~ 2380 2381Set meter stats mask for the ethernet device:: 2382 2383 testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask) 2384 2385where: 2386 2387* ``stats_mask``: Bit mask indicating statistics counter types to be enabled. 2388 2389show port meter stats 2390~~~~~~~~~~~~~~~~~~~~~ 2391 2392Show meter stats of the ethernet device:: 2393 2394 testpmd> show port meter stats (port_id) (mtr_id) (clear) 2395 2396where: 2397 2398* ``clear``: Flag that indicates whether the statistics counters should 2399 be cleared (i.e. set to zero) immediately after they have been read or not. 2400 2401Traffic Management 2402------------------ 2403 2404The following section shows functions for configuring traffic management on 2405on the ethernet device through the use of generic TM API. 2406 2407show port traffic management capability 2408~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2409 2410Show traffic management capability of the port:: 2411 2412 testpmd> show port tm cap (port_id) 2413 2414show port traffic management capability (hierarchy level) 2415~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2416 2417Show traffic management hierarchy level capability of the port:: 2418 2419 testpmd> show port tm level cap (port_id) (level_id) 2420 2421show port traffic management capability (hierarchy node level) 2422~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2423 2424Show the traffic management hierarchy node capability of the port:: 2425 2426 testpmd> show port tm node cap (port_id) (node_id) 2427 2428show port traffic management hierarchy node type 2429~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2430 2431Show the port traffic management hierarchy node type:: 2432 2433 testpmd> show port tm node type (port_id) (node_id) 2434 2435show port traffic management hierarchy node stats 2436~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2437 2438Show the port traffic management hierarchy node statistics:: 2439 2440 testpmd> show port tm node stats (port_id) (node_id) (clear) 2441 2442where: 2443 2444* ``clear``: When this parameter has a non-zero value, the statistics counters 2445 are cleared (i.e. set to zero) immediately after they have been read, 2446 otherwise the statistics counters are left untouched. 2447 2448Add port traffic management private shaper profile 2449~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2450 2451Add the port traffic management private shaper profile:: 2452 2453 testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \ 2454 (tb_rate) (tb_size) (packet_length_adjust) 2455 2456where: 2457 2458* ``shaper_profile id``: Shaper profile ID for the new profile. 2459* ``tb_rate``: Token bucket rate (bytes per second). 2460* ``tb_size``: Token bucket size (bytes). 2461* ``packet_length_adjust``: The value (bytes) to be added to the length of 2462 each packet for the purpose of shaping. This parameter value can be used to 2463 correct the packet length with the framing overhead bytes that are consumed 2464 on the wire. 2465 2466Delete port traffic management private shaper profile 2467~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2468 2469Delete the port traffic management private shaper:: 2470 2471 testpmd> del port tm node shaper profile (port_id) (shaper_profile_id) 2472 2473where: 2474 2475* ``shaper_profile id``: Shaper profile ID that needs to be deleted. 2476 2477Add port traffic management shared shaper 2478~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2479 2480Create the port traffic management shared shaper:: 2481 2482 testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \ 2483 (shaper_profile_id) 2484 2485where: 2486 2487* ``shared_shaper_id``: Shared shaper ID to be created. 2488* ``shaper_profile id``: Shaper profile ID for shared shaper. 2489 2490Set port traffic management shared shaper 2491~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2492 2493Update the port traffic management shared shaper:: 2494 2495 testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \ 2496 (shaper_profile_id) 2497 2498where: 2499 2500* ``shared_shaper_id``: Shared shaper ID to be update. 2501* ``shaper_profile id``: Shaper profile ID for shared shaper. 2502 2503Delete port traffic management shared shaper 2504~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2505 2506Delete the port traffic management shared shaper:: 2507 2508 testpmd> del port tm node shared shaper (port_id) (shared_shaper_id) 2509 2510where: 2511 2512* ``shared_shaper_id``: Shared shaper ID to be deleted. 2513 2514Set port traffic management hiearchy node private shaper 2515~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2516 2517set the port traffic management hierarchy node private shaper:: 2518 2519 testpmd> set port tm node shaper profile (port_id) (node_id) \ 2520 (shaper_profile_id) 2521 2522where: 2523 2524* ``shaper_profile id``: Private shaper profile ID to be enabled on the 2525 hierarchy node. 2526 2527Add port traffic management WRED profile 2528~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2529 2530Create a new WRED profile:: 2531 2532 testpmd> add port tm node wred profile (port_id) (wred_profile_id) \ 2533 (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \ 2534 (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \ 2535 (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r) 2536 2537where: 2538 2539* ``wred_profile id``: Identifier for the newly create WRED profile 2540* ``color_g``: Packet color (green) 2541* ``min_th_g``: Minimum queue threshold for packet with green color 2542* ``max_th_g``: Minimum queue threshold for packet with green color 2543* ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp) 2544* ``wq_log2_g``: Negated log2 of queue weight (wq) 2545* ``color_y``: Packet color (yellow) 2546* ``min_th_y``: Minimum queue threshold for packet with yellow color 2547* ``max_th_y``: Minimum queue threshold for packet with yellow color 2548* ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp) 2549* ``wq_log2_y``: Negated log2 of queue weight (wq) 2550* ``color_r``: Packet color (red) 2551* ``min_th_r``: Minimum queue threshold for packet with yellow color 2552* ``max_th_r``: Minimum queue threshold for packet with yellow color 2553* ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp) 2554* ``wq_log2_r``: Negated log2 of queue weight (wq) 2555 2556Delete port traffic management WRED profile 2557~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2558 2559Delete the WRED profile:: 2560 2561 testpmd> del port tm node wred profile (port_id) (wred_profile_id) 2562 2563Add port traffic management hierarchy nonleaf node 2564~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2565 2566Add nonleaf node to port traffic management hiearchy:: 2567 2568 testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \ 2569 (priority) (weight) (level_id) (shaper_profile_id) \ 2570 (n_sp_priorities) (stats_mask) (n_shared_shapers) \ 2571 [(shared_shaper_0) (shared_shaper_1) ...] \ 2572 2573where: 2574 2575* ``parent_node_id``: Node ID of the parent. 2576* ``priority``: Node priority (highest node priority is zero). This is used by 2577 the SP algorithm running on the parent node for scheduling this node. 2578* ``weight``: Node weight (lowest weight is one). The node weight is relative 2579 to the weight sum of all siblings that have the same priority. It is used by 2580 the WFQ algorithm running on the parent node for scheduling this node. 2581* ``level_id``: Hiearchy level of the node. 2582* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by 2583 the node. 2584* ``n_sp_priorities``: Number of strict priorities. 2585* ``stats_mask``: Mask of statistics counter types to be enabled for this node. 2586* ``n_shared_shapers``: Number of shared shapers. 2587* ``shared_shaper_id``: Shared shaper id. 2588 2589Add port traffic management hierarchy leaf node 2590~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2591 2592Add leaf node to port traffic management hiearchy:: 2593 2594 testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \ 2595 (priority) (weight) (level_id) (shaper_profile_id) \ 2596 (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \ 2597 [(shared_shaper_id) (shared_shaper_id) ...] \ 2598 2599where: 2600 2601* ``parent_node_id``: Node ID of the parent. 2602* ``priority``: Node priority (highest node priority is zero). This is used by 2603 the SP algorithm running on the parent node for scheduling this node. 2604* ``weight``: Node weight (lowest weight is one). The node weight is relative 2605 to the weight sum of all siblings that have the same priority. It is used by 2606 the WFQ algorithm running on the parent node for scheduling this node. 2607* ``level_id``: Hiearchy level of the node. 2608* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by 2609 the node. 2610* ``cman_mode``: Congestion management mode to be enabled for this node. 2611* ``wred_profile_id``: WRED profile id to be enabled for this node. 2612* ``stats_mask``: Mask of statistics counter types to be enabled for this node. 2613* ``n_shared_shapers``: Number of shared shapers. 2614* ``shared_shaper_id``: Shared shaper id. 2615 2616Delete port traffic management hierarchy node 2617~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2618 2619Delete node from port traffic management hiearchy:: 2620 2621 testpmd> del port tm node (port_id) (node_id) 2622 2623Update port traffic management hierarchy parent node 2624~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2625 2626Update port traffic management hierarchy parent node:: 2627 2628 testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \ 2629 (priority) (weight) 2630 2631This function can only be called after the hierarchy commit invocation. Its 2632success depends on the port support for this operation, as advertised through 2633the port capability set. This function is valid for all nodes of the traffic 2634management hierarchy except root node. 2635 2636Suspend port traffic management hierarchy node 2637~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2638 2639 testpmd> suspend port tm node (port_id) (node_id) 2640 2641Resume port traffic management hierarchy node 2642~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2643 2644 testpmd> resume port tm node (port_id) (node_id) 2645 2646Commit port traffic management hierarchy 2647~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2648 2649Commit the traffic management hierarchy on the port:: 2650 2651 testpmd> port tm hierarchy commit (port_id) (clean_on_fail) 2652 2653where: 2654 2655* ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function 2656 call failure. On the other hand, hierarchy is preserved when this parameter 2657 is equal to zero. 2658 2659Set port traffic management default hierarchy (softnic forwarding mode) 2660~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2661 2662set the traffic management default hierarchy on the port:: 2663 2664 testpmd> set port tm hierarchy default (port_id) 2665 2666Filter Functions 2667---------------- 2668 2669This section details the available filter functions that are available. 2670 2671Note these functions interface the deprecated legacy filtering framework, 2672superseded by *rte_flow*. See `Flow rules management`_. 2673 2674ethertype_filter 2675~~~~~~~~~~~~~~~~~~~~ 2676 2677Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue:: 2678 2679 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \ 2680 ethertype (ether_type) (drop|fwd) queue (queue_id) 2681 2682The available information parameters are: 2683 2684* ``port_id``: The port which the Ethertype filter assigned on. 2685 2686* ``mac_addr``: Compare destination mac address. 2687 2688* ``mac_ignr``: Ignore destination mac address match. 2689 2690* ``mac_address``: Destination mac address to match. 2691 2692* ``ether_type``: The EtherType value want to match, 2693 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid. 2694 2695* ``queue_id``: The receive queue associated with this EtherType filter. 2696 It is meaningless when deleting or dropping. 2697 2698Example, to add/remove an ethertype filter rule:: 2699 2700 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \ 2701 ethertype 0x0806 fwd queue 3 2702 2703 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \ 2704 ethertype 0x0806 fwd queue 3 2705 27062tuple_filter 2707~~~~~~~~~~~~~~~~~ 2708 2709Add or delete a 2-tuple filter, 2710which identifies packets by specific protocol and destination TCP/UDP port 2711and forwards packets into one of the receive queues:: 2712 2713 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \ 2714 protocol (protocol_value) mask (mask_value) \ 2715 tcp_flags (tcp_flags_value) priority (prio_value) \ 2716 queue (queue_id) 2717 2718The available information parameters are: 2719 2720* ``port_id``: The port which the 2-tuple filter assigned on. 2721 2722* ``dst_port_value``: Destination port in L4. 2723 2724* ``protocol_value``: IP L4 protocol. 2725 2726* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate. 2727 2728* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP). 2729 2730* ``prio_value``: Priority of this filter. 2731 2732* ``queue_id``: The receive queue associated with this 2-tuple filter. 2733 2734Example, to add/remove an 2tuple filter rule:: 2735 2736 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \ 2737 tcp_flags 0x02 priority 3 queue 3 2738 2739 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \ 2740 tcp_flags 0x02 priority 3 queue 3 2741 27425tuple_filter 2743~~~~~~~~~~~~~~~~~ 2744 2745Add or delete a 5-tuple filter, 2746which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port) 2747and routes packets into one of the receive queues:: 2748 2749 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \ 2750 (src_address) dst_port (dst_port_value) \ 2751 src_port (src_port_value) protocol (protocol_value) \ 2752 mask (mask_value) tcp_flags (tcp_flags_value) \ 2753 priority (prio_value) queue (queue_id) 2754 2755The available information parameters are: 2756 2757* ``port_id``: The port which the 5-tuple filter assigned on. 2758 2759* ``dst_address``: Destination IP address. 2760 2761* ``src_address``: Source IP address. 2762 2763* ``dst_port_value``: TCP/UDP destination port. 2764 2765* ``src_port_value``: TCP/UDP source port. 2766 2767* ``protocol_value``: L4 protocol. 2768 2769* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate 2770 2771* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP). 2772 2773* ``prio_value``: The priority of this filter. 2774 2775* ``queue_id``: The receive queue associated with this 5-tuple filter. 2776 2777Example, to add/remove an 5tuple filter rule:: 2778 2779 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 2780 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 2781 flags 0x0 priority 3 queue 3 2782 2783 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 2784 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 2785 flags 0x0 priority 3 queue 3 2786 2787syn_filter 2788~~~~~~~~~~ 2789 2790Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue:: 2791 2792 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id) 2793 2794The available information parameters are: 2795 2796* ``port_id``: The port which the SYN filter assigned on. 2797 2798* ``high``: This SYN filter has higher priority than other filters. 2799 2800* ``low``: This SYN filter has lower priority than other filters. 2801 2802* ``queue_id``: The receive queue associated with this SYN filter 2803 2804Example:: 2805 2806 testpmd> syn_filter 0 add priority high queue 3 2807 2808flex_filter 2809~~~~~~~~~~~ 2810 2811With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet 2812and routed into one of the receive queues:: 2813 2814 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \ 2815 mask (mask_value) priority (prio_value) queue (queue_id) 2816 2817The available information parameters are: 2818 2819* ``port_id``: The port which the Flex filter is assigned on. 2820 2821* ``len_value``: Filter length in bytes, no greater than 128. 2822 2823* ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match. 2824 2825* ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match. 2826 2827* ``prio_value``: The priority of this filter. 2828 2829* ``queue_id``: The receive queue associated with this Flex filter. 2830 2831Example:: 2832 2833 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \ 2834 mask 000C priority 3 queue 3 2835 2836 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \ 2837 mask 000C priority 3 queue 3 2838 2839 2840.. _testpmd_flow_director: 2841 2842flow_director_filter 2843~~~~~~~~~~~~~~~~~~~~ 2844 2845The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues. 2846 2847Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and 2848Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter: 2849 2850* Perfect 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 IP flow. 2853 2854* Signature filters. 2855 The hardware checks a match between a hash-based signature of the masked fields of the received packet. 2856 2857* Perfect-mac-vlan match filters. 2858 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2859 The masked fields are for MAC VLAN flow. 2860 2861* Perfect-tunnel match filters. 2862 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2863 The masked fields are for tunnel flow. 2864 2865* Perfect-raw-flow-type match filters. 2866 The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet. 2867 The masked fields are specified by input sets. 2868 2869The Flow Director filters can match the different fields for different type of packet: flow type, specific input set 2870per flow type and the flexible payload. 2871 2872The Flow Director can also mask out parts of all of these fields so that filters 2873are only applied to certain fields or parts of the fields. 2874 2875Note that for raw flow type mode the source and destination fields in the 2876raw packet buffer need to be presented in a reversed order with respect 2877to the expected received packets. 2878For example: IP source and destination addresses or TCP/UDP/SCTP 2879source and destination ports 2880 2881Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information. 2882 2883# Commands to add flow director filters of different flow types:: 2884 2885 flow_director_filter (port_id) mode IP (add|del|update) \ 2886 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \ 2887 src (src_ip_address) dst (dst_ip_address) \ 2888 tos (tos_value) proto (proto_value) ttl (ttl_value) \ 2889 vlan (vlan_value) flexbytes (flexbytes_value) \ 2890 (drop|fwd) pf|vf(vf_id) queue (queue_id) \ 2891 fd_id (fd_id_value) 2892 2893 flow_director_filter (port_id) mode IP (add|del|update) \ 2894 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \ 2895 src (src_ip_address) (src_port) \ 2896 dst (dst_ip_address) (dst_port) \ 2897 tos (tos_value) ttl (ttl_value) \ 2898 vlan (vlan_value) flexbytes (flexbytes_value) \ 2899 (drop|fwd) queue pf|vf(vf_id) (queue_id) \ 2900 fd_id (fd_id_value) 2901 2902 flow_director_filter (port_id) mode IP (add|del|update) \ 2903 flow (ipv4-sctp|ipv6-sctp) \ 2904 src (src_ip_address) (src_port) \ 2905 dst (dst_ip_address) (dst_port) \ 2906 tos (tos_value) ttl (ttl_value) \ 2907 tag (verification_tag) vlan (vlan_value) \ 2908 flexbytes (flexbytes_value) (drop|fwd) \ 2909 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value) 2910 2911 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \ 2912 ether (ethertype) flexbytes (flexbytes_value) \ 2913 (drop|fwd) pf|vf(vf_id) queue (queue_id) 2914 fd_id (fd_id_value) 2915 2916 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \ 2917 mac (mac_address) vlan (vlan_value) \ 2918 flexbytes (flexbytes_value) (drop|fwd) \ 2919 queue (queue_id) fd_id (fd_id_value) 2920 2921 flow_director_filter (port_id) mode Tunnel (add|del|update) \ 2922 mac (mac_address) vlan (vlan_value) \ 2923 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \ 2924 flexbytes (flexbytes_value) (drop|fwd) \ 2925 queue (queue_id) fd_id (fd_id_value) 2926 2927 flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \ 2928 (drop|fwd) queue (queue_id) fd_id (fd_id_value) \ 2929 packet (packet file name) 2930 2931For example, to add an ipv4-udp flow type filter:: 2932 2933 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \ 2934 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \ 2935 fwd pf queue 1 fd_id 1 2936 2937For example, add an ipv4-other flow type filter:: 2938 2939 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \ 2940 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \ 2941 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1 2942 2943flush_flow_director 2944~~~~~~~~~~~~~~~~~~~ 2945 2946Flush all flow director filters on a device:: 2947 2948 testpmd> flush_flow_director (port_id) 2949 2950Example, to flush all flow director filter on port 0:: 2951 2952 testpmd> flush_flow_director 0 2953 2954flow_director_mask 2955~~~~~~~~~~~~~~~~~~ 2956 2957Set flow director's input masks:: 2958 2959 flow_director_mask (port_id) mode IP vlan (vlan_value) \ 2960 src_mask (ipv4_src) (ipv6_src) (src_port) \ 2961 dst_mask (ipv4_dst) (ipv6_dst) (dst_port) 2962 2963 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value) 2964 2965 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \ 2966 mac (mac_value) tunnel-type (tunnel_type_value) \ 2967 tunnel-id (tunnel_id_value) 2968 2969Example, to set flow director mask on port 0:: 2970 2971 testpmd> flow_director_mask 0 mode IP vlan 0xefff \ 2972 src_mask 255.255.255.255 \ 2973 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \ 2974 dst_mask 255.255.255.255 \ 2975 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF 2976 2977flow_director_flex_mask 2978~~~~~~~~~~~~~~~~~~~~~~~ 2979 2980set masks of flow director's flexible payload based on certain flow type:: 2981 2982 testpmd> flow_director_flex_mask (port_id) \ 2983 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2984 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \ 2985 l2_payload|all) (mask) 2986 2987Example, to set flow director's flex mask for all flow type on port 0:: 2988 2989 testpmd> flow_director_flex_mask 0 flow all \ 2990 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0) 2991 2992 2993flow_director_flex_payload 2994~~~~~~~~~~~~~~~~~~~~~~~~~~ 2995 2996Configure flexible payload selection:: 2997 2998 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config) 2999 3000For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload:: 3001 3002 testpmd> flow_director_flex_payload 0 l4 \ 3003 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19) 3004 3005get_sym_hash_ena_per_port 3006~~~~~~~~~~~~~~~~~~~~~~~~~ 3007 3008Get symmetric hash enable configuration per port:: 3009 3010 get_sym_hash_ena_per_port (port_id) 3011 3012For example, to get symmetric hash enable configuration of port 1:: 3013 3014 testpmd> get_sym_hash_ena_per_port 1 3015 3016set_sym_hash_ena_per_port 3017~~~~~~~~~~~~~~~~~~~~~~~~~ 3018 3019Set symmetric hash enable configuration per port to enable or disable:: 3020 3021 set_sym_hash_ena_per_port (port_id) (enable|disable) 3022 3023For example, to set symmetric hash enable configuration of port 1 to enable:: 3024 3025 testpmd> set_sym_hash_ena_per_port 1 enable 3026 3027get_hash_global_config 3028~~~~~~~~~~~~~~~~~~~~~~ 3029 3030Get the global configurations of hash filters:: 3031 3032 get_hash_global_config (port_id) 3033 3034For example, to get the global configurations of hash filters of port 1:: 3035 3036 testpmd> get_hash_global_config 1 3037 3038set_hash_global_config 3039~~~~~~~~~~~~~~~~~~~~~~ 3040 3041Set the global configurations of hash filters:: 3042 3043 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \ 3044 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \ 3045 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \ 3046 (enable|disable) 3047 3048For example, to enable simple_xor for flow type of ipv6 on port 2:: 3049 3050 testpmd> set_hash_global_config 2 simple_xor ipv6 enable 3051 3052set_hash_input_set 3053~~~~~~~~~~~~~~~~~~ 3054 3055Set the input set for hash:: 3056 3057 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 3058 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 3059 l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \ 3060 ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \ 3061 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \ 3062 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \ 3063 fld-8th|none) (select|add) 3064 3065For example, to add source IP to hash input set for flow type of ipv4-udp on port 0:: 3066 3067 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add 3068 3069set_fdir_input_set 3070~~~~~~~~~~~~~~~~~~ 3071 3072The Flow Director filters can match the different fields for different type of packet, i.e. specific input set 3073on per flow type and the flexible payload. This command can be used to change input set for each flow type. 3074 3075Set the input set for flow director:: 3076 3077 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 3078 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 3079 l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \ 3080 ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \ 3081 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \ 3082 sctp-dst-port|sctp-veri-tag|none) (select|add) 3083 3084For example to add source IP to FD input set for flow type of ipv4-udp on port 0:: 3085 3086 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add 3087 3088global_config 3089~~~~~~~~~~~~~ 3090 3091Set different GRE key length for input set:: 3092 3093 global_config (port_id) gre-key-len (number in bytes) 3094 3095For example to set GRE key length for input set to 4 bytes on port 0:: 3096 3097 testpmd> global_config 0 gre-key-len 4 3098 3099 3100.. _testpmd_rte_flow: 3101 3102Flow rules management 3103--------------------- 3104 3105Control of the generic flow API (*rte_flow*) is fully exposed through the 3106``flow`` command (validation, creation, destruction, queries and operation 3107modes). 3108 3109Considering *rte_flow* overlaps with all `Filter Functions`_, using both 3110features simultaneously may cause undefined side-effects and is therefore 3111not recommended. 3112 3113``flow`` syntax 3114~~~~~~~~~~~~~~~ 3115 3116Because the ``flow`` command uses dynamic tokens to handle the large number 3117of possible flow rules combinations, its behavior differs slightly from 3118other commands, in particular: 3119 3120- Pressing *?* or the *<tab>* key displays contextual help for the current 3121 token, not that of the entire command. 3122 3123- Optional and repeated parameters are supported (provided they are listed 3124 in the contextual help). 3125 3126The first parameter stands for the operation mode. Possible operations and 3127their general syntax are described below. They are covered in detail in the 3128following sections. 3129 3130- Check whether a flow rule can be created:: 3131 3132 flow validate {port_id} 3133 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3134 pattern {item} [/ {item} [...]] / end 3135 actions {action} [/ {action} [...]] / end 3136 3137- Create a flow rule:: 3138 3139 flow create {port_id} 3140 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3141 pattern {item} [/ {item} [...]] / end 3142 actions {action} [/ {action} [...]] / end 3143 3144- Destroy specific flow rules:: 3145 3146 flow destroy {port_id} rule {rule_id} [...] 3147 3148- Destroy all flow rules:: 3149 3150 flow flush {port_id} 3151 3152- Query an existing flow rule:: 3153 3154 flow query {port_id} {rule_id} {action} 3155 3156- List existing flow rules sorted by priority, filtered by group 3157 identifiers:: 3158 3159 flow list {port_id} [group {group_id}] [...] 3160 3161- Restrict ingress traffic to the defined flow rules:: 3162 3163 flow isolate {port_id} {boolean} 3164 3165Validating flow rules 3166~~~~~~~~~~~~~~~~~~~~~ 3167 3168``flow validate`` reports whether a flow rule would be accepted by the 3169underlying device in its current state but stops short of creating it. It is 3170bound to ``rte_flow_validate()``:: 3171 3172 flow validate {port_id} 3173 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3174 pattern {item} [/ {item} [...]] / end 3175 actions {action} [/ {action} [...]] / end 3176 3177If successful, it will show:: 3178 3179 Flow rule validated 3180 3181Otherwise it will show an error message of the form:: 3182 3183 Caught error type [...] ([...]): [...] 3184 3185This command uses the same parameters as ``flow create``, their format is 3186described in `Creating flow rules`_. 3187 3188Check whether redirecting any Ethernet packet received on port 0 to RX queue 3189index 6 is supported:: 3190 3191 testpmd> flow validate 0 ingress pattern eth / end 3192 actions queue index 6 / end 3193 Flow rule validated 3194 testpmd> 3195 3196Port 0 does not support TCPv6 rules:: 3197 3198 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end 3199 actions drop / end 3200 Caught error type 9 (specific pattern item): Invalid argument 3201 testpmd> 3202 3203Creating flow rules 3204~~~~~~~~~~~~~~~~~~~ 3205 3206``flow create`` validates and creates the specified flow rule. It is bound 3207to ``rte_flow_create()``:: 3208 3209 flow create {port_id} 3210 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3211 pattern {item} [/ {item} [...]] / end 3212 actions {action} [/ {action} [...]] / end 3213 3214If successful, it will return a flow rule ID usable with other commands:: 3215 3216 Flow rule #[...] created 3217 3218Otherwise it will show an error message of the form:: 3219 3220 Caught error type [...] ([...]): [...] 3221 3222Parameters describe in the following order: 3223 3224- Attributes (*group*, *priority*, *ingress*, *egress*, *transfer* tokens). 3225- A matching pattern, starting with the *pattern* token and terminated by an 3226 *end* pattern item. 3227- Actions, starting with the *actions* token and terminated by an *end* 3228 action. 3229 3230These translate directly to *rte_flow* objects provided as-is to the 3231underlying functions. 3232 3233The shortest valid definition only comprises mandatory tokens:: 3234 3235 testpmd> flow create 0 pattern end actions end 3236 3237Note that PMDs may refuse rules that essentially do nothing such as this 3238one. 3239 3240**All unspecified object values are automatically initialized to 0.** 3241 3242Attributes 3243^^^^^^^^^^ 3244 3245These tokens affect flow rule attributes (``struct rte_flow_attr``) and are 3246specified before the ``pattern`` token. 3247 3248- ``group {group id}``: priority group. 3249- ``priority {level}``: priority level within group. 3250- ``ingress``: rule applies to ingress traffic. 3251- ``egress``: rule applies to egress traffic. 3252- ``transfer``: apply rule directly to endpoints found in pattern. 3253 3254Each instance of an attribute specified several times overrides the previous 3255value as shown below (group 4 is used):: 3256 3257 testpmd> flow create 0 group 42 group 24 group 4 [...] 3258 3259Note that once enabled, ``ingress`` and ``egress`` cannot be disabled. 3260 3261While not specifying a direction is an error, some rules may allow both 3262simultaneously. 3263 3264Most rules affect RX therefore contain the ``ingress`` token:: 3265 3266 testpmd> flow create 0 ingress pattern [...] 3267 3268Matching pattern 3269^^^^^^^^^^^^^^^^ 3270 3271A matching pattern starts after the ``pattern`` token. It is made of pattern 3272items and is terminated by a mandatory ``end`` item. 3273 3274Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum 3275rte_flow_item_type``). 3276 3277The ``/`` token is used as a separator between pattern items as shown 3278below:: 3279 3280 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...] 3281 3282Note that protocol items like these must be stacked from lowest to highest 3283layer to make sense. For instance, the following rule is either invalid or 3284unlikely to match any packet:: 3285 3286 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...] 3287 3288More information on these restrictions can be found in the *rte_flow* 3289documentation. 3290 3291Several items support additional specification structures, for example 3292``ipv4`` allows specifying source and destination addresses as follows:: 3293 3294 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1 3295 dst is 10.2.0.0 / end [...] 3296 3297This rule matches all IPv4 traffic with the specified properties. 3298 3299In this example, ``src`` and ``dst`` are field names of the underlying 3300``struct rte_flow_item_ipv4`` object. All item properties can be specified 3301in a similar fashion. 3302 3303The ``is`` token means that the subsequent value must be matched exactly, 3304and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item`` 3305accordingly. Possible assignment tokens are: 3306 3307- ``is``: match value perfectly (with full bit-mask). 3308- ``spec``: match value according to configured bit-mask. 3309- ``last``: specify upper bound to establish a range. 3310- ``mask``: specify bit-mask with relevant bits set to one. 3311- ``prefix``: generate bit-mask from a prefix length. 3312 3313These yield identical results:: 3314 3315 ipv4 src is 10.1.1.1 3316 3317:: 3318 3319 ipv4 src spec 10.1.1.1 src mask 255.255.255.255 3320 3321:: 3322 3323 ipv4 src spec 10.1.1.1 src prefix 32 3324 3325:: 3326 3327 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value 3328 3329:: 3330 3331 ipv4 src is 10.1.1.1 src last 0 # 0 disables range 3332 3333Inclusive ranges can be defined with ``last``:: 3334 3335 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4 3336 3337Note that ``mask`` affects both ``spec`` and ``last``:: 3338 3339 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0 3340 # matches 10.1.0.0 to 10.2.255.255 3341 3342Properties can be modified multiple times:: 3343 3344 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4 3345 3346:: 3347 3348 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16 3349 3350Pattern items 3351^^^^^^^^^^^^^ 3352 3353This section lists supported pattern items and their attributes, if any. 3354 3355- ``end``: end list of pattern items. 3356 3357- ``void``: no-op pattern item. 3358 3359- ``invert``: perform actions when pattern does not match. 3360 3361- ``any``: match any protocol for the current layer. 3362 3363 - ``num {unsigned}``: number of layers covered. 3364 3365- ``pf``: match traffic from/to the physical function. 3366 3367- ``vf``: match traffic from/to a virtual function ID. 3368 3369 - ``id {unsigned}``: VF ID. 3370 3371- ``phy_port``: match traffic from/to a specific physical port. 3372 3373 - ``index {unsigned}``: physical port index. 3374 3375- ``port_id``: match traffic from/to a given DPDK port ID. 3376 3377 - ``id {unsigned}``: DPDK port ID. 3378 3379- ``mark``: match value set in previously matched flow rule using the mark action. 3380 3381 - ``id {unsigned}``: arbitrary integer value. 3382 3383- ``raw``: match an arbitrary byte string. 3384 3385 - ``relative {boolean}``: look for pattern after the previous item. 3386 - ``search {boolean}``: search pattern from offset (see also limit). 3387 - ``offset {integer}``: absolute or relative offset for pattern. 3388 - ``limit {unsigned}``: search area limit for start of pattern. 3389 - ``pattern {string}``: byte string to look for. 3390 3391- ``eth``: match Ethernet header. 3392 3393 - ``dst {MAC-48}``: destination MAC. 3394 - ``src {MAC-48}``: source MAC. 3395 - ``type {unsigned}``: EtherType or TPID. 3396 3397- ``vlan``: match 802.1Q/ad VLAN tag. 3398 3399 - ``tci {unsigned}``: tag control information. 3400 - ``pcp {unsigned}``: priority code point. 3401 - ``dei {unsigned}``: drop eligible indicator. 3402 - ``vid {unsigned}``: VLAN identifier. 3403 - ``inner_type {unsigned}``: inner EtherType or TPID. 3404 3405- ``ipv4``: match IPv4 header. 3406 3407 - ``tos {unsigned}``: type of service. 3408 - ``ttl {unsigned}``: time to live. 3409 - ``proto {unsigned}``: next protocol ID. 3410 - ``src {ipv4 address}``: source address. 3411 - ``dst {ipv4 address}``: destination address. 3412 3413- ``ipv6``: match IPv6 header. 3414 3415 - ``tc {unsigned}``: traffic class. 3416 - ``flow {unsigned}``: flow label. 3417 - ``proto {unsigned}``: protocol (next header). 3418 - ``hop {unsigned}``: hop limit. 3419 - ``src {ipv6 address}``: source address. 3420 - ``dst {ipv6 address}``: destination address. 3421 3422- ``icmp``: match ICMP header. 3423 3424 - ``type {unsigned}``: ICMP packet type. 3425 - ``code {unsigned}``: ICMP packet code. 3426 3427- ``udp``: match UDP header. 3428 3429 - ``src {unsigned}``: UDP source port. 3430 - ``dst {unsigned}``: UDP destination port. 3431 3432- ``tcp``: match TCP header. 3433 3434 - ``src {unsigned}``: TCP source port. 3435 - ``dst {unsigned}``: TCP destination port. 3436 3437- ``sctp``: match SCTP header. 3438 3439 - ``src {unsigned}``: SCTP source port. 3440 - ``dst {unsigned}``: SCTP destination port. 3441 - ``tag {unsigned}``: validation tag. 3442 - ``cksum {unsigned}``: checksum. 3443 3444- ``vxlan``: match VXLAN header. 3445 3446 - ``vni {unsigned}``: VXLAN identifier. 3447 3448- ``e_tag``: match IEEE 802.1BR E-Tag header. 3449 3450 - ``grp_ecid_b {unsigned}``: GRP and E-CID base. 3451 3452- ``nvgre``: match NVGRE header. 3453 3454 - ``tni {unsigned}``: virtual subnet ID. 3455 3456- ``mpls``: match MPLS header. 3457 3458 - ``label {unsigned}``: MPLS label. 3459 3460- ``gre``: match GRE header. 3461 3462 - ``protocol {unsigned}``: protocol type. 3463 3464- ``fuzzy``: fuzzy pattern match, expect faster than default. 3465 3466 - ``thresh {unsigned}``: accuracy threshold. 3467 3468- ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header. 3469 3470 - ``teid {unsigned}``: tunnel endpoint identifier. 3471 3472- ``geneve``: match GENEVE header. 3473 3474 - ``vni {unsigned}``: virtual network identifier. 3475 - ``protocol {unsigned}``: protocol type. 3476 3477- ``vxlan-gpe``: match VXLAN-GPE header. 3478 3479 - ``vni {unsigned}``: VXLAN-GPE identifier. 3480 3481- ``arp_eth_ipv4``: match ARP header for Ethernet/IPv4. 3482 3483 - ``sha {MAC-48}``: sender hardware address. 3484 - ``spa {ipv4 address}``: sender IPv4 address. 3485 - ``tha {MAC-48}``: target hardware address. 3486 - ``tpa {ipv4 address}``: target IPv4 address. 3487 3488- ``ipv6_ext``: match presence of any IPv6 extension header. 3489 3490 - ``next_hdr {unsigned}``: next header. 3491 3492- ``icmp6``: match any ICMPv6 header. 3493 3494 - ``type {unsigned}``: ICMPv6 type. 3495 - ``code {unsigned}``: ICMPv6 code. 3496 3497- ``icmp6_nd_ns``: match ICMPv6 neighbor discovery solicitation. 3498 3499 - ``target_addr {ipv6 address}``: target address. 3500 3501- ``icmp6_nd_na``: match ICMPv6 neighbor discovery advertisement. 3502 3503 - ``target_addr {ipv6 address}``: target address. 3504 3505- ``icmp6_nd_opt``: match presence of any ICMPv6 neighbor discovery option. 3506 3507 - ``type {unsigned}``: ND option type. 3508 3509- ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery source Ethernet 3510 link-layer address option. 3511 3512 - ``sla {MAC-48}``: source Ethernet LLA. 3513 3514- ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery target Ethernet 3515 link-layer address option. 3516 3517 - ``tla {MAC-48}``: target Ethernet LLA. 3518 3519Actions list 3520^^^^^^^^^^^^ 3521 3522A list of actions starts after the ``actions`` token in the same fashion as 3523`Matching pattern`_; actions are separated by ``/`` tokens and the list is 3524terminated by a mandatory ``end`` action. 3525 3526Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum 3527rte_flow_action_type``). 3528 3529Dropping all incoming UDPv4 packets can be expressed as follows:: 3530 3531 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3532 actions drop / end 3533 3534Several actions have configurable properties which must be specified when 3535there is no valid default value. For example, ``queue`` requires a target 3536queue index. 3537 3538This rule redirects incoming UDPv4 traffic to queue index 6:: 3539 3540 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3541 actions queue index 6 / end 3542 3543While this one could be rejected by PMDs (unspecified queue index):: 3544 3545 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3546 actions queue / end 3547 3548As defined by *rte_flow*, the list is not ordered, all actions of a given 3549rule are performed simultaneously. These are equivalent:: 3550 3551 queue index 6 / void / mark id 42 / end 3552 3553:: 3554 3555 void / mark id 42 / queue index 6 / end 3556 3557All actions in a list should have different types, otherwise only the last 3558action of a given type is taken into account:: 3559 3560 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6 3561 3562:: 3563 3564 drop / drop / drop / end # drop is performed only once 3565 3566:: 3567 3568 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24 3569 3570Considering they are performed simultaneously, opposite and overlapping 3571actions can sometimes be combined when the end result is unambiguous:: 3572 3573 drop / queue index 6 / end # drop has no effect 3574 3575:: 3576 3577 queue index 6 / rss queues 6 7 8 / end # queue has no effect 3578 3579:: 3580 3581 drop / passthru / end # drop has no effect 3582 3583Note that PMDs may still refuse such combinations. 3584 3585Actions 3586^^^^^^^ 3587 3588This section lists supported actions and their attributes, if any. 3589 3590- ``end``: end list of actions. 3591 3592- ``void``: no-op action. 3593 3594- ``passthru``: let subsequent rule process matched packets. 3595 3596- ``jump``: redirect traffic to group on device. 3597 3598 - ``group {unsigned}``: group to redirect to. 3599 3600- ``mark``: attach 32 bit value to packets. 3601 3602 - ``id {unsigned}``: 32 bit value to return with packets. 3603 3604- ``flag``: flag packets. 3605 3606- ``queue``: assign packets to a given queue index. 3607 3608 - ``index {unsigned}``: queue index to use. 3609 3610- ``drop``: drop packets (note: passthru has priority). 3611 3612- ``count``: enable counters for this rule. 3613 3614- ``rss``: spread packets among several queues. 3615 3616 - ``func {hash function}``: RSS hash function to apply, allowed tokens are 3617 the same as `set_hash_global_config`_. 3618 3619 - ``level {unsigned}``: encapsulation level for ``types``. 3620 3621 - ``types [{RSS hash type} [...]] end``: specific RSS hash types, allowed 3622 tokens are the same as `set_hash_input_set`_, except that an empty list 3623 does not disable RSS but instead requests unspecified "best-effort" 3624 settings. 3625 3626 - ``key {string}``: RSS hash key, overrides ``key_len``. 3627 3628 - ``key_len {unsigned}``: RSS hash key length in bytes, can be used in 3629 conjunction with ``key`` to pad or truncate it. 3630 3631 - ``queues [{unsigned} [...]] end``: queue indices to use. 3632 3633- ``pf``: direct traffic to physical function. 3634 3635- ``vf``: direct traffic to a virtual function ID. 3636 3637 - ``original {boolean}``: use original VF ID if possible. 3638 - ``id {unsigned}``: VF ID. 3639 3640- ``phy_port``: direct packets to physical port index. 3641 3642 - ``original {boolean}``: use original port index if possible. 3643 - ``index {unsigned}``: physical port index. 3644 3645- ``port_id``: direct matching traffic to a given DPDK port ID. 3646 3647 - ``original {boolean}``: use original DPDK port ID if possible. 3648 - ``id {unsigned}``: DPDK port ID. 3649 3650- ``of_set_mpls_ttl``: OpenFlow's ``OFPAT_SET_MPLS_TTL``. 3651 3652 - ``mpls_ttl``: MPLS TTL. 3653 3654- ``of_dec_mpls_ttl``: OpenFlow's ``OFPAT_DEC_MPLS_TTL``. 3655 3656- ``of_set_nw_ttl``: OpenFlow's ``OFPAT_SET_NW_TTL``. 3657 3658 - ``nw_ttl``: IP TTL. 3659 3660- ``of_dec_nw_ttl``: OpenFlow's ``OFPAT_DEC_NW_TTL``. 3661 3662- ``of_copy_ttl_out``: OpenFlow's ``OFPAT_COPY_TTL_OUT``. 3663 3664- ``of_copy_ttl_in``: OpenFlow's ``OFPAT_COPY_TTL_IN``. 3665 3666- ``of_pop_vlan``: OpenFlow's ``OFPAT_POP_VLAN``. 3667 3668- ``of_push_vlan``: OpenFlow's ``OFPAT_PUSH_VLAN``. 3669 3670 - ``ethertype``: Ethertype. 3671 3672- ``of_set_vlan_vid``: OpenFlow's ``OFPAT_SET_VLAN_VID``. 3673 3674 - ``vlan_vid``: VLAN id. 3675 3676- ``of_set_vlan_pcp``: OpenFlow's ``OFPAT_SET_VLAN_PCP``. 3677 3678 - ``vlan_pcp``: VLAN priority. 3679 3680- ``of_pop_mpls``: OpenFlow's ``OFPAT_POP_MPLS``. 3681 3682 - ``ethertype``: Ethertype. 3683 3684- ``of_push_mpls``: OpenFlow's ``OFPAT_PUSH_MPLS``. 3685 3686 - ``ethertype``: Ethertype. 3687 3688- ``vxlan_encap``: Performs a VXLAN encapsulation, outer layer configuration 3689 is done through `Config VXLAN Encap outer layers`_. 3690 3691- ``vxlan_decap``: Performs a decapsulation action by stripping all headers of 3692 the VXLAN tunnel network overlay from the matched flow. 3693 3694- ``nvgre_encap``: Performs a NVGRE encapsulation, outer layer configuration 3695 is done through `Config NVGRE Encap outer layers`_. 3696 3697- ``nvgre_decap``: Performs a decapsulation action by stripping all headers of 3698 the NVGRE tunnel network overlay from the matched flow. 3699 3700Destroying flow rules 3701~~~~~~~~~~~~~~~~~~~~~ 3702 3703``flow destroy`` destroys one or more rules from their rule ID (as returned 3704by ``flow create``), this command calls ``rte_flow_destroy()`` as many 3705times as necessary:: 3706 3707 flow destroy {port_id} rule {rule_id} [...] 3708 3709If successful, it will show:: 3710 3711 Flow rule #[...] destroyed 3712 3713It does not report anything for rule IDs that do not exist. The usual error 3714message is shown when a rule cannot be destroyed:: 3715 3716 Caught error type [...] ([...]): [...] 3717 3718``flow flush`` destroys all rules on a device and does not take extra 3719arguments. It is bound to ``rte_flow_flush()``:: 3720 3721 flow flush {port_id} 3722 3723Any errors are reported as above. 3724 3725Creating several rules and destroying them:: 3726 3727 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3728 actions queue index 2 / end 3729 Flow rule #0 created 3730 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3731 actions queue index 3 / end 3732 Flow rule #1 created 3733 testpmd> flow destroy 0 rule 0 rule 1 3734 Flow rule #1 destroyed 3735 Flow rule #0 destroyed 3736 testpmd> 3737 3738The same result can be achieved using ``flow flush``:: 3739 3740 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3741 actions queue index 2 / end 3742 Flow rule #0 created 3743 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3744 actions queue index 3 / end 3745 Flow rule #1 created 3746 testpmd> flow flush 0 3747 testpmd> 3748 3749Non-existent rule IDs are ignored:: 3750 3751 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3752 actions queue index 2 / end 3753 Flow rule #0 created 3754 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3755 actions queue index 3 / end 3756 Flow rule #1 created 3757 testpmd> flow destroy 0 rule 42 rule 10 rule 2 3758 testpmd> 3759 testpmd> flow destroy 0 rule 0 3760 Flow rule #0 destroyed 3761 testpmd> 3762 3763Querying flow rules 3764~~~~~~~~~~~~~~~~~~~ 3765 3766``flow query`` queries a specific action of a flow rule having that 3767ability. Such actions collect information that can be reported using this 3768command. It is bound to ``rte_flow_query()``:: 3769 3770 flow query {port_id} {rule_id} {action} 3771 3772If successful, it will display either the retrieved data for known actions 3773or the following message:: 3774 3775 Cannot display result for action type [...] ([...]) 3776 3777Otherwise, it will complain either that the rule does not exist or that some 3778error occurred:: 3779 3780 Flow rule #[...] not found 3781 3782:: 3783 3784 Caught error type [...] ([...]): [...] 3785 3786Currently only the ``count`` action is supported. This action reports the 3787number of packets that hit the flow rule and the total number of bytes. Its 3788output has the following format:: 3789 3790 count: 3791 hits_set: [...] # whether "hits" contains a valid value 3792 bytes_set: [...] # whether "bytes" contains a valid value 3793 hits: [...] # number of packets 3794 bytes: [...] # number of bytes 3795 3796Querying counters for TCPv6 packets redirected to queue 6:: 3797 3798 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end 3799 actions queue index 6 / count / end 3800 Flow rule #4 created 3801 testpmd> flow query 0 4 count 3802 count: 3803 hits_set: 1 3804 bytes_set: 0 3805 hits: 386446 3806 bytes: 0 3807 testpmd> 3808 3809Listing flow rules 3810~~~~~~~~~~~~~~~~~~ 3811 3812``flow list`` lists existing flow rules sorted by priority and optionally 3813filtered by group identifiers:: 3814 3815 flow list {port_id} [group {group_id}] [...] 3816 3817This command only fails with the following message if the device does not 3818exist:: 3819 3820 Invalid port [...] 3821 3822Output consists of a header line followed by a short description of each 3823flow rule, one per line. There is no output at all when no flow rules are 3824configured on the device:: 3825 3826 ID Group Prio Attr Rule 3827 [...] [...] [...] [...] [...] 3828 3829``Attr`` column flags: 3830 3831- ``i`` for ``ingress``. 3832- ``e`` for ``egress``. 3833 3834Creating several flow rules and listing them:: 3835 3836 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3837 actions queue index 6 / end 3838 Flow rule #0 created 3839 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3840 actions queue index 2 / end 3841 Flow rule #1 created 3842 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end 3843 actions rss queues 6 7 8 end / end 3844 Flow rule #2 created 3845 testpmd> flow list 0 3846 ID Group Prio Attr Rule 3847 0 0 0 i- ETH IPV4 => QUEUE 3848 1 0 0 i- ETH IPV6 => QUEUE 3849 2 0 5 i- ETH IPV4 UDP => RSS 3850 testpmd> 3851 3852Rules are sorted by priority (i.e. group ID first, then priority level):: 3853 3854 testpmd> flow list 1 3855 ID Group Prio Attr Rule 3856 0 0 0 i- ETH => COUNT 3857 6 0 500 i- ETH IPV6 TCP => DROP COUNT 3858 5 0 1000 i- ETH IPV6 ICMP => QUEUE 3859 1 24 0 i- ETH IPV4 UDP => QUEUE 3860 4 24 10 i- ETH IPV4 TCP => DROP 3861 3 24 20 i- ETH IPV4 => DROP 3862 2 24 42 i- ETH IPV4 UDP => QUEUE 3863 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 3864 testpmd> 3865 3866Output can be limited to specific groups:: 3867 3868 testpmd> flow list 1 group 0 group 63 3869 ID Group Prio Attr Rule 3870 0 0 0 i- ETH => COUNT 3871 6 0 500 i- ETH IPV6 TCP => DROP COUNT 3872 5 0 1000 i- ETH IPV6 ICMP => QUEUE 3873 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 3874 testpmd> 3875 3876Toggling isolated mode 3877~~~~~~~~~~~~~~~~~~~~~~ 3878 3879``flow isolate`` can be used to tell the underlying PMD that ingress traffic 3880must only be injected from the defined flow rules; that no default traffic 3881is expected outside those rules and the driver is free to assign more 3882resources to handle them. It is bound to ``rte_flow_isolate()``:: 3883 3884 flow isolate {port_id} {boolean} 3885 3886If successful, enabling or disabling isolated mode shows either:: 3887 3888 Ingress traffic on port [...] 3889 is now restricted to the defined flow rules 3890 3891Or:: 3892 3893 Ingress traffic on port [...] 3894 is not restricted anymore to the defined flow rules 3895 3896Otherwise, in case of error:: 3897 3898 Caught error type [...] ([...]): [...] 3899 3900Mainly due to its side effects, PMDs supporting this mode may not have the 3901ability to toggle it more than once without reinitializing affected ports 3902first (e.g. by exiting testpmd). 3903 3904Enabling isolated mode:: 3905 3906 testpmd> flow isolate 0 true 3907 Ingress traffic on port 0 is now restricted to the defined flow rules 3908 testpmd> 3909 3910Disabling isolated mode:: 3911 3912 testpmd> flow isolate 0 false 3913 Ingress traffic on port 0 is not restricted anymore to the defined flow rules 3914 testpmd> 3915 3916Sample QinQ flow rules 3917~~~~~~~~~~~~~~~~~~~~~~ 3918 3919Before creating QinQ rule(s) the following commands should be issued to enable QinQ:: 3920 3921 testpmd> port stop 0 3922 testpmd> vlan set qinq on 0 3923 3924The above command sets the inner and outer TPID's to 0x8100. 3925 3926To change the TPID's the following commands should be used:: 3927 3928 testpmd> vlan set outer tpid 0xa100 0 3929 testpmd> vlan set inner tpid 0x9100 0 3930 testpmd> port start 0 3931 3932Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM. 3933 3934:: 3935 3936 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 / 3937 vlan tci is 456 / end actions vf id 1 / queue index 0 / end 3938 Flow rule #0 validated 3939 3940 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 / 3941 vlan tci is 456 / end actions vf id 123 / queue index 0 / end 3942 Flow rule #0 created 3943 3944 testpmd> flow list 0 3945 ID Group Prio Attr Rule 3946 0 0 0 i- ETH VLAN VLAN=>VF QUEUE 3947 3948Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host. 3949 3950:: 3951 3952 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 / 3953 vlan tci is 654 / end actions pf / queue index 0 / end 3954 Flow rule #1 validated 3955 3956 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 / 3957 vlan tci is 654 / end actions pf / queue index 1 / end 3958 Flow rule #1 created 3959 3960 testpmd> flow list 0 3961 ID Group Prio Attr Rule 3962 0 0 0 i- ETH VLAN VLAN=>VF QUEUE 3963 1 0 0 i- ETH VLAN VLAN=>PF QUEUE 3964 3965Sample VXLAN encapsulation rule 3966~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 3967 3968VXLAN encapsulation outer layer has default value pre-configured in testpmd 3969source code, those can be changed by using the following commands 3970 3971IPv4 VXLAN outer header:: 3972 3973 testpmd> set vxlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src 127.0.0.1 3974 ip-dst 128.0.0.1 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 3975 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 3976 queue index 0 / end 3977 3978 testpmd> set vxlan-with-vlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src 3979 127.0.0.1 ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11 3980 eth-dst 22:22:22:22:22:22 3981 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 3982 queue index 0 / end 3983 3984IPv6 VXLAN outer header:: 3985 3986 testpmd> set vxlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4 ip-src ::1 3987 ip-dst ::2222 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 3988 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 3989 queue index 0 / end 3990 3991 testpmd> set vxlan-with-vlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4 3992 ip-src ::1 ip-dst ::2222 vlan-tci 34 eth-src 11:11:11:11:11:11 3993 eth-dst 22:22:22:22:22:22 3994 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 3995 queue index 0 / end 3996 3997Sample NVGRE encapsulation rule 3998~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 3999 4000NVGRE encapsulation outer layer has default value pre-configured in testpmd 4001source code, those can be changed by using the following commands 4002 4003IPv4 NVGRE outer header:: 4004 4005 testpmd> set nvgre ip-version ipv4 tni 4 ip-src 127.0.0.1 ip-dst 128.0.0.1 4006 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4007 testpmd> flow create 0 ingress pattern end actions nvgre_encap / 4008 queue index 0 / end 4009 4010 testpmd> set nvgre-with-vlan ip-version ipv4 tni 4 ip-src 127.0.0.1 4011 ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11 4012 eth-dst 22:22:22:22:22:22 4013 testpmd> flow create 0 ingress pattern end actions nvgre_encap / 4014 queue index 0 / end 4015 4016IPv6 NVGRE outer header:: 4017 4018 testpmd> set nvgre ip-version ipv6 tni 4 ip-src ::1 ip-dst ::2222 4019 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4020 testpmd> flow create 0 ingress pattern end actions nvgre_encap / 4021 queue index 0 / end 4022 4023 testpmd> set nvgre-with-vlan ip-version ipv6 tni 4 ip-src ::1 ip-dst ::2222 4024 vlan-tci 34 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4025 testpmd> flow create 0 ingress pattern end actions nvgre_encap / 4026 queue index 0 / end 4027 4028BPF Functions 4029-------------- 4030 4031The following sections show functions to load/unload eBPF based filters. 4032 4033bpf-load 4034~~~~~~~~ 4035 4036Load an eBPF program as a callback for partciular RX/TX queue:: 4037 4038 testpmd> bpf-load rx|tx (portid) (queueid) (load-flags) (bpf-prog-filename) 4039 4040The available load-flags are: 4041 4042* ``J``: use JIT generated native code, otherwise BPF interpreter will be used. 4043 4044* ``M``: assume input parameter is a pointer to rte_mbuf, otherwise assume it is a pointer to first segment's data. 4045 4046* ``-``: none. 4047 4048.. note:: 4049 4050 You'll need clang v3.7 or above to build bpf program you'd like to load 4051 4052For example: 4053 4054.. code-block:: console 4055 4056 cd test/bpf 4057 clang -O2 -target bpf -c t1.c 4058 4059Then to load (and JIT compile) t1.o at RX queue 0, port 1:: 4060 4061.. code-block:: console 4062 4063 testpmd> bpf-load rx 1 0 J ./dpdk.org/test/bpf/t1.o 4064 4065To load (not JITed) t1.o at TX queue 0, port 0:: 4066 4067.. code-block:: console 4068 4069 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o 4070 4071bpf-unload 4072~~~~~~~~~~ 4073 4074Unload previously loaded eBPF program for partciular RX/TX queue:: 4075 4076 testpmd> bpf-unload rx|tx (portid) (queueid) 4077 4078For example to unload BPF filter from TX queue 0, port 0: 4079 4080.. code-block:: console 4081 4082 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o 4083