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 config - queue deferred start 1769~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1770 1771Switch on/off deferred start of a specific port queue:: 1772 1773 testpmd> port (port_id) (rxq|txq) (queue_id) deferred_start (on|off) 1774 1775port setup queue 1776~~~~~~~~~~~~~~~~~~~~~ 1777 1778Setup a rx/tx queue on a specific port:: 1779 1780 testpmd> port (port_id) (rxq|txq) (queue_id) setup 1781 1782Only take effect when port is started. 1783 1784port config - speed 1785~~~~~~~~~~~~~~~~~~~ 1786 1787Set the speed and duplex mode for all ports or a specific port:: 1788 1789 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \ 1790 duplex (half|full|auto) 1791 1792port config - queues/descriptors 1793~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1794 1795Set number of queues/descriptors for rxq, txq, rxd and txd:: 1796 1797 testpmd> port config all (rxq|txq|rxd|txd) (value) 1798 1799This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options. 1800 1801port config - max-pkt-len 1802~~~~~~~~~~~~~~~~~~~~~~~~~ 1803 1804Set the maximum packet length:: 1805 1806 testpmd> port config all max-pkt-len (value) 1807 1808This is equivalent to the ``--max-pkt-len`` command-line option. 1809 1810port config - CRC Strip 1811~~~~~~~~~~~~~~~~~~~~~~~ 1812 1813Set hardware CRC stripping on or off for all ports:: 1814 1815 testpmd> port config all crc-strip (on|off) 1816 1817CRC stripping is on by default. 1818 1819The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option. 1820 1821port config - scatter 1822~~~~~~~~~~~~~~~~~~~~~~~ 1823 1824Set RX scatter mode on or off for all ports:: 1825 1826 testpmd> port config all scatter (on|off) 1827 1828RX scatter mode is off by default. 1829 1830The ``on`` option is equivalent to the ``--enable-scatter`` command-line option. 1831 1832port config - RX Checksum 1833~~~~~~~~~~~~~~~~~~~~~~~~~ 1834 1835Set hardware RX checksum offload to on or off for all ports:: 1836 1837 testpmd> port config all rx-cksum (on|off) 1838 1839Checksum offload is off by default. 1840 1841The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option. 1842 1843port config - VLAN 1844~~~~~~~~~~~~~~~~~~ 1845 1846Set hardware VLAN on or off for all ports:: 1847 1848 testpmd> port config all hw-vlan (on|off) 1849 1850Hardware VLAN is off by default. 1851 1852The ``on`` option is equivalent to the ``--enable-hw-vlan`` command-line option. 1853 1854port config - VLAN filter 1855~~~~~~~~~~~~~~~~~~~~~~~~~ 1856 1857Set hardware VLAN filter on or off for all ports:: 1858 1859 testpmd> port config all hw-vlan-filter (on|off) 1860 1861Hardware VLAN filter is off by default. 1862 1863The ``on`` option is equivalent to the ``--enable-hw-vlan-filter`` command-line option. 1864 1865port config - VLAN strip 1866~~~~~~~~~~~~~~~~~~~~~~~~ 1867 1868Set hardware VLAN strip on or off for all ports:: 1869 1870 testpmd> port config all hw-vlan-strip (on|off) 1871 1872Hardware VLAN strip is off by default. 1873 1874The ``on`` option is equivalent to the ``--enable-hw-vlan-strip`` command-line option. 1875 1876port config - VLAN extend 1877~~~~~~~~~~~~~~~~~~~~~~~~~ 1878 1879Set hardware VLAN extend on or off for all ports:: 1880 1881 testpmd> port config all hw-vlan-extend (on|off) 1882 1883Hardware VLAN extend is off by default. 1884 1885The ``on`` option is equivalent to the ``--enable-hw-vlan-extend`` command-line option. 1886 1887port config - Drop Packets 1888~~~~~~~~~~~~~~~~~~~~~~~~~~ 1889 1890Set packet drop for packets with no descriptors on or off for all ports:: 1891 1892 testpmd> port config all drop-en (on|off) 1893 1894Packet dropping for packets with no descriptors is off by default. 1895 1896The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option. 1897 1898port config - RSS 1899~~~~~~~~~~~~~~~~~ 1900 1901Set the RSS (Receive Side Scaling) mode on or off:: 1902 1903 testpmd> port config all rss (all|default|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none) 1904 1905RSS is on by default. 1906 1907The ``all`` option is equivalent to ip|tcp|udp|sctp|ether. 1908The ``default`` option enables all supported RSS types reported by device info. 1909The ``none`` option is equivalent to the ``--disable-rss`` command-line option. 1910 1911port config - RSS Reta 1912~~~~~~~~~~~~~~~~~~~~~~ 1913 1914Set the RSS (Receive Side Scaling) redirection table:: 1915 1916 testpmd> port config all rss reta (hash,queue)[,(hash,queue)] 1917 1918port config - DCB 1919~~~~~~~~~~~~~~~~~ 1920 1921Set the DCB mode for an individual port:: 1922 1923 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off) 1924 1925The traffic class should be 4 or 8. 1926 1927port config - Burst 1928~~~~~~~~~~~~~~~~~~~ 1929 1930Set the number of packets per burst:: 1931 1932 testpmd> port config all burst (value) 1933 1934This is equivalent to the ``--burst`` command-line option. 1935 1936port config - Threshold 1937~~~~~~~~~~~~~~~~~~~~~~~ 1938 1939Set thresholds for TX/RX queues:: 1940 1941 testpmd> port config all (threshold) (value) 1942 1943Where the threshold type can be: 1944 1945* ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255. 1946 1947* ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255. 1948 1949* ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255. 1950 1951* ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255. 1952 1953* ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255. 1954 1955* ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255. 1956 1957* ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd. 1958 1959* ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd. 1960 1961* ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd. 1962 1963These threshold options are also available from the command-line. 1964 1965port config - E-tag 1966~~~~~~~~~~~~~~~~~~~ 1967 1968Set the value of ether-type for E-tag:: 1969 1970 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value) 1971 1972Enable/disable the E-tag support:: 1973 1974 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable) 1975 1976port config pctype mapping 1977~~~~~~~~~~~~~~~~~~~~~~~~~~ 1978 1979Reset pctype mapping table:: 1980 1981 testpmd> port config (port_id) pctype mapping reset 1982 1983Update hardware defined pctype to software defined flow type mapping table:: 1984 1985 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id) 1986 1987where: 1988 1989* ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table. 1990 1991* ``flow_type_id``: software flow type id as the index of the pctype mapping table. 1992 1993port config input set 1994~~~~~~~~~~~~~~~~~~~~~ 1995 1996Config RSS/FDIR/FDIR flexible payload input set for some pctype:: 1997 testpmd> port config (port_id) pctype (pctype_id) \ 1998 (hash_inset|fdir_inset|fdir_flx_inset) \ 1999 (get|set|clear) field (field_idx) 2000 2001Clear RSS/FDIR/FDIR flexible payload input set for some pctype:: 2002 testpmd> port config (port_id) pctype (pctype_id) \ 2003 (hash_inset|fdir_inset|fdir_flx_inset) clear all 2004 2005where: 2006 2007* ``pctype_id``: hardware packet classification types. 2008* ``field_idx``: hardware field index. 2009 2010port config udp_tunnel_port 2011~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2012 2013Add/remove UDP tunnel port for VXLAN/GENEVE tunneling protocols:: 2014 testpmd> port config (port_id) udp_tunnel_port add|rm vxlan|geneve (udp_port) 2015 2016Link Bonding Functions 2017---------------------- 2018 2019The Link Bonding functions make it possible to dynamically create and 2020manage link bonding devices from within testpmd interactive prompt. 2021 2022create bonded device 2023~~~~~~~~~~~~~~~~~~~~ 2024 2025Create a new bonding device:: 2026 2027 testpmd> create bonded device (mode) (socket) 2028 2029For example, to create a bonded device in mode 1 on socket 0:: 2030 2031 testpmd> create bonded device 1 0 2032 created new bonded device (port X) 2033 2034add bonding slave 2035~~~~~~~~~~~~~~~~~ 2036 2037Adds Ethernet device to a Link Bonding device:: 2038 2039 testpmd> add bonding slave (slave id) (port id) 2040 2041For example, to add Ethernet device (port 6) to a Link Bonding device (port 10):: 2042 2043 testpmd> add bonding slave 6 10 2044 2045 2046remove bonding slave 2047~~~~~~~~~~~~~~~~~~~~ 2048 2049Removes an Ethernet slave device from a Link Bonding device:: 2050 2051 testpmd> remove bonding slave (slave id) (port id) 2052 2053For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10):: 2054 2055 testpmd> remove bonding slave 6 10 2056 2057set bonding mode 2058~~~~~~~~~~~~~~~~ 2059 2060Set the Link Bonding mode of a Link Bonding device:: 2061 2062 testpmd> set bonding mode (value) (port id) 2063 2064For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3):: 2065 2066 testpmd> set bonding mode 3 10 2067 2068set bonding primary 2069~~~~~~~~~~~~~~~~~~~ 2070 2071Set an Ethernet slave device as the primary device on a Link Bonding device:: 2072 2073 testpmd> set bonding primary (slave id) (port id) 2074 2075For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10):: 2076 2077 testpmd> set bonding primary 6 10 2078 2079set bonding mac 2080~~~~~~~~~~~~~~~ 2081 2082Set the MAC address of a Link Bonding device:: 2083 2084 testpmd> set bonding mac (port id) (mac) 2085 2086For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01:: 2087 2088 testpmd> set bonding mac 10 00:00:00:00:00:01 2089 2090set bonding xmit_balance_policy 2091~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2092 2093Set the transmission policy for a Link Bonding device when it is in Balance XOR mode:: 2094 2095 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34) 2096 2097For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports):: 2098 2099 testpmd> set bonding xmit_balance_policy 10 l34 2100 2101 2102set bonding mon_period 2103~~~~~~~~~~~~~~~~~~~~~~ 2104 2105Set the link status monitoring polling period in milliseconds for a bonding device. 2106 2107This adds support for PMD slave devices which do not support link status interrupts. 2108When the mon_period is set to a value greater than 0 then all PMD's which do not support 2109link status ISR will be queried every polling interval to check if their link status has changed:: 2110 2111 testpmd> set bonding mon_period (port_id) (value) 2112 2113For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms:: 2114 2115 testpmd> set bonding mon_period 5 150 2116 2117 2118set bonding lacp dedicated_queue 2119~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2120 2121Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic 2122when in mode 4 (link-aggregration-802.3ad):: 2123 2124 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable) 2125 2126 2127set bonding agg_mode 2128~~~~~~~~~~~~~~~~~~~~ 2129 2130Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad):: 2131 2132 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable) 2133 2134 2135show bonding config 2136~~~~~~~~~~~~~~~~~~~ 2137 2138Show the current configuration of a Link Bonding device:: 2139 2140 testpmd> show bonding config (port id) 2141 2142For example, 2143to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4) 2144in balance mode with a transmission policy of layer 2+3:: 2145 2146 testpmd> show bonding config 9 2147 Bonding mode: 2 2148 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23 2149 Slaves (3): [1 3 4] 2150 Active Slaves (3): [1 3 4] 2151 Primary: [3] 2152 2153 2154Register Functions 2155------------------ 2156 2157The Register Functions can be used to read from and write to registers on the network card referenced by a port number. 2158This is mainly useful for debugging purposes. 2159Reference should be made to the appropriate datasheet for the network card for details on the register addresses 2160and fields that can be accessed. 2161 2162read reg 2163~~~~~~~~ 2164 2165Display the value of a port register:: 2166 2167 testpmd> read reg (port_id) (address) 2168 2169For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller:: 2170 2171 testpmd> read reg 0 0xEE00 2172 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241) 2173 2174read regfield 2175~~~~~~~~~~~~~ 2176 2177Display a port register bit field:: 2178 2179 testpmd> read regfield (port_id) (address) (bit_x) (bit_y) 2180 2181For example, reading the lowest two bits from the register in the example above:: 2182 2183 testpmd> read regfield 0 0xEE00 0 1 2184 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1) 2185 2186read regbit 2187~~~~~~~~~~~ 2188 2189Display a single port register bit:: 2190 2191 testpmd> read regbit (port_id) (address) (bit_x) 2192 2193For example, reading the lowest bit from the register in the example above:: 2194 2195 testpmd> read regbit 0 0xEE00 0 2196 port 0 PCI register at offset 0xEE00: bit 0=1 2197 2198write reg 2199~~~~~~~~~ 2200 2201Set the value of a port register:: 2202 2203 testpmd> write reg (port_id) (address) (value) 2204 2205For example, to clear a register:: 2206 2207 testpmd> write reg 0 0xEE00 0x0 2208 port 0 PCI register at offset 0xEE00: 0x00000000 (0) 2209 2210write regfield 2211~~~~~~~~~~~~~~ 2212 2213Set bit field of a port register:: 2214 2215 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value) 2216 2217For example, writing to the register cleared in the example above:: 2218 2219 testpmd> write regfield 0 0xEE00 0 1 2 2220 port 0 PCI register at offset 0xEE00: 0x00000002 (2) 2221 2222write regbit 2223~~~~~~~~~~~~ 2224 2225Set single bit value of a port register:: 2226 2227 testpmd> write regbit (port_id) (address) (bit_x) (value) 2228 2229For example, to set the high bit in the register from the example above:: 2230 2231 testpmd> write regbit 0 0xEE00 31 1 2232 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658) 2233 2234Traffic Metering and Policing 2235----------------------------- 2236 2237The following section shows functions for configuring traffic metering and 2238policing on the ethernet device through the use of generic ethdev API. 2239 2240show port traffic management capability 2241~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2242 2243Show traffic metering and policing capability of the port:: 2244 2245 testpmd> show port meter cap (port_id) 2246 2247add port meter profile (srTCM rfc2967) 2248~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2249 2250Add meter profile (srTCM rfc2697) to the ethernet device:: 2251 2252 testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \ 2253 (cir) (cbs) (ebs) 2254 2255where: 2256 2257* ``profile_id``: ID for the meter profile. 2258* ``cir``: Committed Information Rate (CIR) (bytes/second). 2259* ``cbs``: Committed Burst Size (CBS) (bytes). 2260* ``ebs``: Excess Burst Size (EBS) (bytes). 2261 2262add port meter profile (trTCM rfc2968) 2263~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2264 2265Add meter profile (srTCM rfc2698) to the ethernet device:: 2266 2267 testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \ 2268 (cir) (pir) (cbs) (pbs) 2269 2270where: 2271 2272* ``profile_id``: ID for the meter profile. 2273* ``cir``: Committed information rate (bytes/second). 2274* ``pir``: Peak information rate (bytes/second). 2275* ``cbs``: Committed burst size (bytes). 2276* ``pbs``: Peak burst size (bytes). 2277 2278add port meter profile (trTCM rfc4115) 2279~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2280 2281Add meter profile (trTCM rfc4115) to the ethernet device:: 2282 2283 testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \ 2284 (cir) (eir) (cbs) (ebs) 2285 2286where: 2287 2288* ``profile_id``: ID for the meter profile. 2289* ``cir``: Committed information rate (bytes/second). 2290* ``eir``: Excess information rate (bytes/second). 2291* ``cbs``: Committed burst size (bytes). 2292* ``ebs``: Excess burst size (bytes). 2293 2294delete port meter profile 2295~~~~~~~~~~~~~~~~~~~~~~~~~ 2296 2297Delete meter profile from the ethernet device:: 2298 2299 testpmd> del port meter profile (port_id) (profile_id) 2300 2301create port meter 2302~~~~~~~~~~~~~~~~~ 2303 2304Create new meter object for the ethernet device:: 2305 2306 testpmd> create port meter (port_id) (mtr_id) (profile_id) \ 2307 (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \ 2308 (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\ 2309 (dscp_tbl_entry63)] 2310 2311where: 2312 2313* ``mtr_id``: meter object ID. 2314* ``profile_id``: ID for the meter profile. 2315* ``meter_enable``: When this parameter has a non-zero value, the meter object 2316 gets enabled at the time of creation, otherwise remains disabled. 2317* ``g_action``: Policer action for the packet with green color. 2318* ``y_action``: Policer action for the packet with yellow color. 2319* ``r_action``: Policer action for the packet with red color. 2320* ``stats_mask``: Mask of statistics counter types to be enabled for the 2321 meter object. 2322* ``shared``: When this parameter has a non-zero value, the meter object is 2323 shared by multiple flows. Otherwise, meter object is used by single flow. 2324* ``use_pre_meter_color``: When this parameter has a non-zero value, the 2325 input color for the current meter object is determined by the latest meter 2326 object in the same flow. Otherwise, the current meter object uses the 2327 *dscp_table* to determine the input color. 2328* ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input 2329 color, 0 <= x <= 63. 2330 2331enable port meter 2332~~~~~~~~~~~~~~~~~ 2333 2334Enable meter for the ethernet device:: 2335 2336 testpmd> enable port meter (port_id) (mtr_id) 2337 2338disable port meter 2339~~~~~~~~~~~~~~~~~~ 2340 2341Disable meter for the ethernet device:: 2342 2343 testpmd> disable port meter (port_id) (mtr_id) 2344 2345delete port meter 2346~~~~~~~~~~~~~~~~~ 2347 2348Delete meter for the ethernet device:: 2349 2350 testpmd> del port meter (port_id) (mtr_id) 2351 2352Set port meter profile 2353~~~~~~~~~~~~~~~~~~~~~~ 2354 2355Set meter profile for the ethernet device:: 2356 2357 testpmd> set port meter profile (port_id) (mtr_id) (profile_id) 2358 2359set port meter dscp table 2360~~~~~~~~~~~~~~~~~~~~~~~~~ 2361 2362Set meter dscp table for the ethernet device:: 2363 2364 testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \ 2365 (dscp_tbl_entry1)...(dscp_tbl_entry63)] 2366 2367set port meter policer action 2368~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2369 2370Set meter policer action for the ethernet device:: 2371 2372 testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \ 2373 (action0) [(action1) (action1)] 2374 2375where: 2376 2377* ``action_mask``: Bit mask indicating which policer actions need to be 2378 updated. One or more policer actions can be updated in a single function 2379 invocation. To update the policer action associated with color C, bit 2380 (1 << C) needs to be set in *action_mask* and element at position C 2381 in the *actions* array needs to be valid. 2382* ``actionx``: Policer action for the color x, 2383 RTE_MTR_GREEN <= x < RTE_MTR_COLORS 2384 2385set port meter stats mask 2386~~~~~~~~~~~~~~~~~~~~~~~~~ 2387 2388Set meter stats mask for the ethernet device:: 2389 2390 testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask) 2391 2392where: 2393 2394* ``stats_mask``: Bit mask indicating statistics counter types to be enabled. 2395 2396show port meter stats 2397~~~~~~~~~~~~~~~~~~~~~ 2398 2399Show meter stats of the ethernet device:: 2400 2401 testpmd> show port meter stats (port_id) (mtr_id) (clear) 2402 2403where: 2404 2405* ``clear``: Flag that indicates whether the statistics counters should 2406 be cleared (i.e. set to zero) immediately after they have been read or not. 2407 2408Traffic Management 2409------------------ 2410 2411The following section shows functions for configuring traffic management on 2412on the ethernet device through the use of generic TM API. 2413 2414show port traffic management capability 2415~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2416 2417Show traffic management capability of the port:: 2418 2419 testpmd> show port tm cap (port_id) 2420 2421show port traffic management capability (hierarchy level) 2422~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2423 2424Show traffic management hierarchy level capability of the port:: 2425 2426 testpmd> show port tm level cap (port_id) (level_id) 2427 2428show port traffic management capability (hierarchy node level) 2429~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2430 2431Show the traffic management hierarchy node capability of the port:: 2432 2433 testpmd> show port tm node cap (port_id) (node_id) 2434 2435show port traffic management hierarchy node type 2436~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2437 2438Show the port traffic management hierarchy node type:: 2439 2440 testpmd> show port tm node type (port_id) (node_id) 2441 2442show port traffic management hierarchy node stats 2443~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2444 2445Show the port traffic management hierarchy node statistics:: 2446 2447 testpmd> show port tm node stats (port_id) (node_id) (clear) 2448 2449where: 2450 2451* ``clear``: When this parameter has a non-zero value, the statistics counters 2452 are cleared (i.e. set to zero) immediately after they have been read, 2453 otherwise the statistics counters are left untouched. 2454 2455Add port traffic management private shaper profile 2456~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2457 2458Add the port traffic management private shaper profile:: 2459 2460 testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \ 2461 (tb_rate) (tb_size) (packet_length_adjust) 2462 2463where: 2464 2465* ``shaper_profile id``: Shaper profile ID for the new profile. 2466* ``tb_rate``: Token bucket rate (bytes per second). 2467* ``tb_size``: Token bucket size (bytes). 2468* ``packet_length_adjust``: The value (bytes) to be added to the length of 2469 each packet for the purpose of shaping. This parameter value can be used to 2470 correct the packet length with the framing overhead bytes that are consumed 2471 on the wire. 2472 2473Delete port traffic management private shaper profile 2474~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2475 2476Delete the port traffic management private shaper:: 2477 2478 testpmd> del port tm node shaper profile (port_id) (shaper_profile_id) 2479 2480where: 2481 2482* ``shaper_profile id``: Shaper profile ID that needs to be deleted. 2483 2484Add port traffic management shared shaper 2485~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2486 2487Create the port traffic management shared shaper:: 2488 2489 testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \ 2490 (shaper_profile_id) 2491 2492where: 2493 2494* ``shared_shaper_id``: Shared shaper ID to be created. 2495* ``shaper_profile id``: Shaper profile ID for shared shaper. 2496 2497Set port traffic management shared shaper 2498~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2499 2500Update the port traffic management shared shaper:: 2501 2502 testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \ 2503 (shaper_profile_id) 2504 2505where: 2506 2507* ``shared_shaper_id``: Shared shaper ID to be update. 2508* ``shaper_profile id``: Shaper profile ID for shared shaper. 2509 2510Delete port traffic management shared shaper 2511~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2512 2513Delete the port traffic management shared shaper:: 2514 2515 testpmd> del port tm node shared shaper (port_id) (shared_shaper_id) 2516 2517where: 2518 2519* ``shared_shaper_id``: Shared shaper ID to be deleted. 2520 2521Set port traffic management hiearchy node private shaper 2522~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2523 2524set the port traffic management hierarchy node private shaper:: 2525 2526 testpmd> set port tm node shaper profile (port_id) (node_id) \ 2527 (shaper_profile_id) 2528 2529where: 2530 2531* ``shaper_profile id``: Private shaper profile ID to be enabled on the 2532 hierarchy node. 2533 2534Add port traffic management WRED profile 2535~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2536 2537Create a new WRED profile:: 2538 2539 testpmd> add port tm node wred profile (port_id) (wred_profile_id) \ 2540 (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \ 2541 (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \ 2542 (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r) 2543 2544where: 2545 2546* ``wred_profile id``: Identifier for the newly create WRED profile 2547* ``color_g``: Packet color (green) 2548* ``min_th_g``: Minimum queue threshold for packet with green color 2549* ``max_th_g``: Minimum queue threshold for packet with green color 2550* ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp) 2551* ``wq_log2_g``: Negated log2 of queue weight (wq) 2552* ``color_y``: Packet color (yellow) 2553* ``min_th_y``: Minimum queue threshold for packet with yellow color 2554* ``max_th_y``: Minimum queue threshold for packet with yellow color 2555* ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp) 2556* ``wq_log2_y``: Negated log2 of queue weight (wq) 2557* ``color_r``: Packet color (red) 2558* ``min_th_r``: Minimum queue threshold for packet with yellow color 2559* ``max_th_r``: Minimum queue threshold for packet with yellow color 2560* ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp) 2561* ``wq_log2_r``: Negated log2 of queue weight (wq) 2562 2563Delete port traffic management WRED profile 2564~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2565 2566Delete the WRED profile:: 2567 2568 testpmd> del port tm node wred profile (port_id) (wred_profile_id) 2569 2570Add port traffic management hierarchy nonleaf node 2571~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2572 2573Add nonleaf node to port traffic management hiearchy:: 2574 2575 testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \ 2576 (priority) (weight) (level_id) (shaper_profile_id) \ 2577 (n_sp_priorities) (stats_mask) (n_shared_shapers) \ 2578 [(shared_shaper_0) (shared_shaper_1) ...] \ 2579 2580where: 2581 2582* ``parent_node_id``: Node ID of the parent. 2583* ``priority``: Node priority (highest node priority is zero). This is used by 2584 the SP algorithm running on the parent node for scheduling this node. 2585* ``weight``: Node weight (lowest weight is one). The node weight is relative 2586 to the weight sum of all siblings that have the same priority. It is used by 2587 the WFQ algorithm running on the parent node for scheduling this node. 2588* ``level_id``: Hiearchy level of the node. 2589* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by 2590 the node. 2591* ``n_sp_priorities``: Number of strict priorities. 2592* ``stats_mask``: Mask of statistics counter types to be enabled for this node. 2593* ``n_shared_shapers``: Number of shared shapers. 2594* ``shared_shaper_id``: Shared shaper id. 2595 2596Add port traffic management hierarchy leaf node 2597~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2598 2599Add leaf node to port traffic management hiearchy:: 2600 2601 testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \ 2602 (priority) (weight) (level_id) (shaper_profile_id) \ 2603 (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \ 2604 [(shared_shaper_id) (shared_shaper_id) ...] \ 2605 2606where: 2607 2608* ``parent_node_id``: Node ID of the parent. 2609* ``priority``: Node priority (highest node priority is zero). This is used by 2610 the SP algorithm running on the parent node for scheduling this node. 2611* ``weight``: Node weight (lowest weight is one). The node weight is relative 2612 to the weight sum of all siblings that have the same priority. It is used by 2613 the WFQ algorithm running on the parent node for scheduling this node. 2614* ``level_id``: Hiearchy level of the node. 2615* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by 2616 the node. 2617* ``cman_mode``: Congestion management mode to be enabled for this node. 2618* ``wred_profile_id``: WRED profile id to be enabled for this node. 2619* ``stats_mask``: Mask of statistics counter types to be enabled for this node. 2620* ``n_shared_shapers``: Number of shared shapers. 2621* ``shared_shaper_id``: Shared shaper id. 2622 2623Delete port traffic management hierarchy node 2624~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2625 2626Delete node from port traffic management hiearchy:: 2627 2628 testpmd> del port tm node (port_id) (node_id) 2629 2630Update port traffic management hierarchy parent node 2631~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2632 2633Update port traffic management hierarchy parent node:: 2634 2635 testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \ 2636 (priority) (weight) 2637 2638This function can only be called after the hierarchy commit invocation. Its 2639success depends on the port support for this operation, as advertised through 2640the port capability set. This function is valid for all nodes of the traffic 2641management hierarchy except root node. 2642 2643Suspend port traffic management hierarchy node 2644~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2645 2646 testpmd> suspend port tm node (port_id) (node_id) 2647 2648Resume port traffic management hierarchy node 2649~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2650 2651 testpmd> resume port tm node (port_id) (node_id) 2652 2653Commit port traffic management hierarchy 2654~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2655 2656Commit the traffic management hierarchy on the port:: 2657 2658 testpmd> port tm hierarchy commit (port_id) (clean_on_fail) 2659 2660where: 2661 2662* ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function 2663 call failure. On the other hand, hierarchy is preserved when this parameter 2664 is equal to zero. 2665 2666Set port traffic management default hierarchy (softnic forwarding mode) 2667~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2668 2669set the traffic management default hierarchy on the port:: 2670 2671 testpmd> set port tm hierarchy default (port_id) 2672 2673Filter Functions 2674---------------- 2675 2676This section details the available filter functions that are available. 2677 2678Note these functions interface the deprecated legacy filtering framework, 2679superseded by *rte_flow*. See `Flow rules management`_. 2680 2681ethertype_filter 2682~~~~~~~~~~~~~~~~~~~~ 2683 2684Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue:: 2685 2686 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \ 2687 ethertype (ether_type) (drop|fwd) queue (queue_id) 2688 2689The available information parameters are: 2690 2691* ``port_id``: The port which the Ethertype filter assigned on. 2692 2693* ``mac_addr``: Compare destination mac address. 2694 2695* ``mac_ignr``: Ignore destination mac address match. 2696 2697* ``mac_address``: Destination mac address to match. 2698 2699* ``ether_type``: The EtherType value want to match, 2700 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid. 2701 2702* ``queue_id``: The receive queue associated with this EtherType filter. 2703 It is meaningless when deleting or dropping. 2704 2705Example, to add/remove an ethertype filter rule:: 2706 2707 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \ 2708 ethertype 0x0806 fwd queue 3 2709 2710 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \ 2711 ethertype 0x0806 fwd queue 3 2712 27132tuple_filter 2714~~~~~~~~~~~~~~~~~ 2715 2716Add or delete a 2-tuple filter, 2717which identifies packets by specific protocol and destination TCP/UDP port 2718and forwards packets into one of the receive queues:: 2719 2720 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \ 2721 protocol (protocol_value) mask (mask_value) \ 2722 tcp_flags (tcp_flags_value) priority (prio_value) \ 2723 queue (queue_id) 2724 2725The available information parameters are: 2726 2727* ``port_id``: The port which the 2-tuple filter assigned on. 2728 2729* ``dst_port_value``: Destination port in L4. 2730 2731* ``protocol_value``: IP L4 protocol. 2732 2733* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate. 2734 2735* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP). 2736 2737* ``prio_value``: Priority of this filter. 2738 2739* ``queue_id``: The receive queue associated with this 2-tuple filter. 2740 2741Example, to add/remove an 2tuple filter rule:: 2742 2743 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \ 2744 tcp_flags 0x02 priority 3 queue 3 2745 2746 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \ 2747 tcp_flags 0x02 priority 3 queue 3 2748 27495tuple_filter 2750~~~~~~~~~~~~~~~~~ 2751 2752Add or delete a 5-tuple filter, 2753which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port) 2754and routes packets into one of the receive queues:: 2755 2756 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \ 2757 (src_address) dst_port (dst_port_value) \ 2758 src_port (src_port_value) protocol (protocol_value) \ 2759 mask (mask_value) tcp_flags (tcp_flags_value) \ 2760 priority (prio_value) queue (queue_id) 2761 2762The available information parameters are: 2763 2764* ``port_id``: The port which the 5-tuple filter assigned on. 2765 2766* ``dst_address``: Destination IP address. 2767 2768* ``src_address``: Source IP address. 2769 2770* ``dst_port_value``: TCP/UDP destination port. 2771 2772* ``src_port_value``: TCP/UDP source port. 2773 2774* ``protocol_value``: L4 protocol. 2775 2776* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate 2777 2778* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP). 2779 2780* ``prio_value``: The priority of this filter. 2781 2782* ``queue_id``: The receive queue associated with this 5-tuple filter. 2783 2784Example, to add/remove an 5tuple filter rule:: 2785 2786 testpmd> 5tuple_filter 0 add dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 2787 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 2788 flags 0x0 priority 3 queue 3 2789 2790 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 2791 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 2792 flags 0x0 priority 3 queue 3 2793 2794syn_filter 2795~~~~~~~~~~ 2796 2797Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue:: 2798 2799 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id) 2800 2801The available information parameters are: 2802 2803* ``port_id``: The port which the SYN filter assigned on. 2804 2805* ``high``: This SYN filter has higher priority than other filters. 2806 2807* ``low``: This SYN filter has lower priority than other filters. 2808 2809* ``queue_id``: The receive queue associated with this SYN filter 2810 2811Example:: 2812 2813 testpmd> syn_filter 0 add priority high queue 3 2814 2815flex_filter 2816~~~~~~~~~~~ 2817 2818With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet 2819and routed into one of the receive queues:: 2820 2821 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \ 2822 mask (mask_value) priority (prio_value) queue (queue_id) 2823 2824The available information parameters are: 2825 2826* ``port_id``: The port which the Flex filter is assigned on. 2827 2828* ``len_value``: Filter length in bytes, no greater than 128. 2829 2830* ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match. 2831 2832* ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match. 2833 2834* ``prio_value``: The priority of this filter. 2835 2836* ``queue_id``: The receive queue associated with this Flex filter. 2837 2838Example:: 2839 2840 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \ 2841 mask 000C priority 3 queue 3 2842 2843 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \ 2844 mask 000C priority 3 queue 3 2845 2846 2847.. _testpmd_flow_director: 2848 2849flow_director_filter 2850~~~~~~~~~~~~~~~~~~~~ 2851 2852The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues. 2853 2854Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and 2855Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter: 2856 2857* Perfect 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 IP flow. 2860 2861* Signature filters. 2862 The hardware checks a match between a hash-based signature of the masked fields of the received packet. 2863 2864* Perfect-mac-vlan match filters. 2865 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2866 The masked fields are for MAC VLAN flow. 2867 2868* Perfect-tunnel match filters. 2869 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2870 The masked fields are for tunnel flow. 2871 2872* Perfect-raw-flow-type match filters. 2873 The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet. 2874 The masked fields are specified by input sets. 2875 2876The Flow Director filters can match the different fields for different type of packet: flow type, specific input set 2877per flow type and the flexible payload. 2878 2879The Flow Director can also mask out parts of all of these fields so that filters 2880are only applied to certain fields or parts of the fields. 2881 2882Note that for raw flow type mode the source and destination fields in the 2883raw packet buffer need to be presented in a reversed order with respect 2884to the expected received packets. 2885For example: IP source and destination addresses or TCP/UDP/SCTP 2886source and destination ports 2887 2888Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information. 2889 2890# Commands to add flow director filters of different flow types:: 2891 2892 flow_director_filter (port_id) mode IP (add|del|update) \ 2893 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \ 2894 src (src_ip_address) dst (dst_ip_address) \ 2895 tos (tos_value) proto (proto_value) ttl (ttl_value) \ 2896 vlan (vlan_value) flexbytes (flexbytes_value) \ 2897 (drop|fwd) pf|vf(vf_id) queue (queue_id) \ 2898 fd_id (fd_id_value) 2899 2900 flow_director_filter (port_id) mode IP (add|del|update) \ 2901 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \ 2902 src (src_ip_address) (src_port) \ 2903 dst (dst_ip_address) (dst_port) \ 2904 tos (tos_value) ttl (ttl_value) \ 2905 vlan (vlan_value) flexbytes (flexbytes_value) \ 2906 (drop|fwd) queue pf|vf(vf_id) (queue_id) \ 2907 fd_id (fd_id_value) 2908 2909 flow_director_filter (port_id) mode IP (add|del|update) \ 2910 flow (ipv4-sctp|ipv6-sctp) \ 2911 src (src_ip_address) (src_port) \ 2912 dst (dst_ip_address) (dst_port) \ 2913 tos (tos_value) ttl (ttl_value) \ 2914 tag (verification_tag) vlan (vlan_value) \ 2915 flexbytes (flexbytes_value) (drop|fwd) \ 2916 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value) 2917 2918 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \ 2919 ether (ethertype) flexbytes (flexbytes_value) \ 2920 (drop|fwd) pf|vf(vf_id) queue (queue_id) 2921 fd_id (fd_id_value) 2922 2923 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \ 2924 mac (mac_address) vlan (vlan_value) \ 2925 flexbytes (flexbytes_value) (drop|fwd) \ 2926 queue (queue_id) fd_id (fd_id_value) 2927 2928 flow_director_filter (port_id) mode Tunnel (add|del|update) \ 2929 mac (mac_address) vlan (vlan_value) \ 2930 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \ 2931 flexbytes (flexbytes_value) (drop|fwd) \ 2932 queue (queue_id) fd_id (fd_id_value) 2933 2934 flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \ 2935 (drop|fwd) queue (queue_id) fd_id (fd_id_value) \ 2936 packet (packet file name) 2937 2938For example, to add an ipv4-udp flow type filter:: 2939 2940 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \ 2941 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \ 2942 fwd pf queue 1 fd_id 1 2943 2944For example, add an ipv4-other flow type filter:: 2945 2946 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \ 2947 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \ 2948 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1 2949 2950flush_flow_director 2951~~~~~~~~~~~~~~~~~~~ 2952 2953Flush all flow director filters on a device:: 2954 2955 testpmd> flush_flow_director (port_id) 2956 2957Example, to flush all flow director filter on port 0:: 2958 2959 testpmd> flush_flow_director 0 2960 2961flow_director_mask 2962~~~~~~~~~~~~~~~~~~ 2963 2964Set flow director's input masks:: 2965 2966 flow_director_mask (port_id) mode IP vlan (vlan_value) \ 2967 src_mask (ipv4_src) (ipv6_src) (src_port) \ 2968 dst_mask (ipv4_dst) (ipv6_dst) (dst_port) 2969 2970 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value) 2971 2972 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \ 2973 mac (mac_value) tunnel-type (tunnel_type_value) \ 2974 tunnel-id (tunnel_id_value) 2975 2976Example, to set flow director mask on port 0:: 2977 2978 testpmd> flow_director_mask 0 mode IP vlan 0xefff \ 2979 src_mask 255.255.255.255 \ 2980 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \ 2981 dst_mask 255.255.255.255 \ 2982 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF 2983 2984flow_director_flex_mask 2985~~~~~~~~~~~~~~~~~~~~~~~ 2986 2987set masks of flow director's flexible payload based on certain flow type:: 2988 2989 testpmd> flow_director_flex_mask (port_id) \ 2990 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2991 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \ 2992 l2_payload|all) (mask) 2993 2994Example, to set flow director's flex mask for all flow type on port 0:: 2995 2996 testpmd> flow_director_flex_mask 0 flow all \ 2997 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0) 2998 2999 3000flow_director_flex_payload 3001~~~~~~~~~~~~~~~~~~~~~~~~~~ 3002 3003Configure flexible payload selection:: 3004 3005 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config) 3006 3007For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload:: 3008 3009 testpmd> flow_director_flex_payload 0 l4 \ 3010 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19) 3011 3012get_sym_hash_ena_per_port 3013~~~~~~~~~~~~~~~~~~~~~~~~~ 3014 3015Get symmetric hash enable configuration per port:: 3016 3017 get_sym_hash_ena_per_port (port_id) 3018 3019For example, to get symmetric hash enable configuration of port 1:: 3020 3021 testpmd> get_sym_hash_ena_per_port 1 3022 3023set_sym_hash_ena_per_port 3024~~~~~~~~~~~~~~~~~~~~~~~~~ 3025 3026Set symmetric hash enable configuration per port to enable or disable:: 3027 3028 set_sym_hash_ena_per_port (port_id) (enable|disable) 3029 3030For example, to set symmetric hash enable configuration of port 1 to enable:: 3031 3032 testpmd> set_sym_hash_ena_per_port 1 enable 3033 3034get_hash_global_config 3035~~~~~~~~~~~~~~~~~~~~~~ 3036 3037Get the global configurations of hash filters:: 3038 3039 get_hash_global_config (port_id) 3040 3041For example, to get the global configurations of hash filters of port 1:: 3042 3043 testpmd> get_hash_global_config 1 3044 3045set_hash_global_config 3046~~~~~~~~~~~~~~~~~~~~~~ 3047 3048Set the global configurations of hash filters:: 3049 3050 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \ 3051 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \ 3052 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \ 3053 (enable|disable) 3054 3055For example, to enable simple_xor for flow type of ipv6 on port 2:: 3056 3057 testpmd> set_hash_global_config 2 simple_xor ipv6 enable 3058 3059set_hash_input_set 3060~~~~~~~~~~~~~~~~~~ 3061 3062Set the input set for hash:: 3063 3064 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 3065 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 3066 l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \ 3067 ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \ 3068 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \ 3069 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \ 3070 fld-8th|none) (select|add) 3071 3072For example, to add source IP to hash input set for flow type of ipv4-udp on port 0:: 3073 3074 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add 3075 3076set_fdir_input_set 3077~~~~~~~~~~~~~~~~~~ 3078 3079The Flow Director filters can match the different fields for different type of packet, i.e. specific input set 3080on per flow type and the flexible payload. This command can be used to change input set for each flow type. 3081 3082Set the input set for flow director:: 3083 3084 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 3085 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 3086 l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \ 3087 ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \ 3088 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \ 3089 sctp-dst-port|sctp-veri-tag|none) (select|add) 3090 3091For example to add source IP to FD input set for flow type of ipv4-udp on port 0:: 3092 3093 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add 3094 3095global_config 3096~~~~~~~~~~~~~ 3097 3098Set different GRE key length for input set:: 3099 3100 global_config (port_id) gre-key-len (number in bytes) 3101 3102For example to set GRE key length for input set to 4 bytes on port 0:: 3103 3104 testpmd> global_config 0 gre-key-len 4 3105 3106 3107.. _testpmd_rte_flow: 3108 3109Flow rules management 3110--------------------- 3111 3112Control of the generic flow API (*rte_flow*) is fully exposed through the 3113``flow`` command (validation, creation, destruction, queries and operation 3114modes). 3115 3116Considering *rte_flow* overlaps with all `Filter Functions`_, using both 3117features simultaneously may cause undefined side-effects and is therefore 3118not recommended. 3119 3120``flow`` syntax 3121~~~~~~~~~~~~~~~ 3122 3123Because the ``flow`` command uses dynamic tokens to handle the large number 3124of possible flow rules combinations, its behavior differs slightly from 3125other commands, in particular: 3126 3127- Pressing *?* or the *<tab>* key displays contextual help for the current 3128 token, not that of the entire command. 3129 3130- Optional and repeated parameters are supported (provided they are listed 3131 in the contextual help). 3132 3133The first parameter stands for the operation mode. Possible operations and 3134their general syntax are described below. They are covered in detail in the 3135following sections. 3136 3137- Check whether a flow rule can be created:: 3138 3139 flow validate {port_id} 3140 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3141 pattern {item} [/ {item} [...]] / end 3142 actions {action} [/ {action} [...]] / end 3143 3144- Create a flow rule:: 3145 3146 flow create {port_id} 3147 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3148 pattern {item} [/ {item} [...]] / end 3149 actions {action} [/ {action} [...]] / end 3150 3151- Destroy specific flow rules:: 3152 3153 flow destroy {port_id} rule {rule_id} [...] 3154 3155- Destroy all flow rules:: 3156 3157 flow flush {port_id} 3158 3159- Query an existing flow rule:: 3160 3161 flow query {port_id} {rule_id} {action} 3162 3163- List existing flow rules sorted by priority, filtered by group 3164 identifiers:: 3165 3166 flow list {port_id} [group {group_id}] [...] 3167 3168- Restrict ingress traffic to the defined flow rules:: 3169 3170 flow isolate {port_id} {boolean} 3171 3172Validating flow rules 3173~~~~~~~~~~~~~~~~~~~~~ 3174 3175``flow validate`` reports whether a flow rule would be accepted by the 3176underlying device in its current state but stops short of creating it. It is 3177bound to ``rte_flow_validate()``:: 3178 3179 flow validate {port_id} 3180 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3181 pattern {item} [/ {item} [...]] / end 3182 actions {action} [/ {action} [...]] / end 3183 3184If successful, it will show:: 3185 3186 Flow rule validated 3187 3188Otherwise it will show an error message of the form:: 3189 3190 Caught error type [...] ([...]): [...] 3191 3192This command uses the same parameters as ``flow create``, their format is 3193described in `Creating flow rules`_. 3194 3195Check whether redirecting any Ethernet packet received on port 0 to RX queue 3196index 6 is supported:: 3197 3198 testpmd> flow validate 0 ingress pattern eth / end 3199 actions queue index 6 / end 3200 Flow rule validated 3201 testpmd> 3202 3203Port 0 does not support TCPv6 rules:: 3204 3205 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end 3206 actions drop / end 3207 Caught error type 9 (specific pattern item): Invalid argument 3208 testpmd> 3209 3210Creating flow rules 3211~~~~~~~~~~~~~~~~~~~ 3212 3213``flow create`` validates and creates the specified flow rule. It is bound 3214to ``rte_flow_create()``:: 3215 3216 flow create {port_id} 3217 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3218 pattern {item} [/ {item} [...]] / end 3219 actions {action} [/ {action} [...]] / end 3220 3221If successful, it will return a flow rule ID usable with other commands:: 3222 3223 Flow rule #[...] created 3224 3225Otherwise it will show an error message of the form:: 3226 3227 Caught error type [...] ([...]): [...] 3228 3229Parameters describe in the following order: 3230 3231- Attributes (*group*, *priority*, *ingress*, *egress*, *transfer* tokens). 3232- A matching pattern, starting with the *pattern* token and terminated by an 3233 *end* pattern item. 3234- Actions, starting with the *actions* token and terminated by an *end* 3235 action. 3236 3237These translate directly to *rte_flow* objects provided as-is to the 3238underlying functions. 3239 3240The shortest valid definition only comprises mandatory tokens:: 3241 3242 testpmd> flow create 0 pattern end actions end 3243 3244Note that PMDs may refuse rules that essentially do nothing such as this 3245one. 3246 3247**All unspecified object values are automatically initialized to 0.** 3248 3249Attributes 3250^^^^^^^^^^ 3251 3252These tokens affect flow rule attributes (``struct rte_flow_attr``) and are 3253specified before the ``pattern`` token. 3254 3255- ``group {group id}``: priority group. 3256- ``priority {level}``: priority level within group. 3257- ``ingress``: rule applies to ingress traffic. 3258- ``egress``: rule applies to egress traffic. 3259- ``transfer``: apply rule directly to endpoints found in pattern. 3260 3261Each instance of an attribute specified several times overrides the previous 3262value as shown below (group 4 is used):: 3263 3264 testpmd> flow create 0 group 42 group 24 group 4 [...] 3265 3266Note that once enabled, ``ingress`` and ``egress`` cannot be disabled. 3267 3268While not specifying a direction is an error, some rules may allow both 3269simultaneously. 3270 3271Most rules affect RX therefore contain the ``ingress`` token:: 3272 3273 testpmd> flow create 0 ingress pattern [...] 3274 3275Matching pattern 3276^^^^^^^^^^^^^^^^ 3277 3278A matching pattern starts after the ``pattern`` token. It is made of pattern 3279items and is terminated by a mandatory ``end`` item. 3280 3281Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum 3282rte_flow_item_type``). 3283 3284The ``/`` token is used as a separator between pattern items as shown 3285below:: 3286 3287 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...] 3288 3289Note that protocol items like these must be stacked from lowest to highest 3290layer to make sense. For instance, the following rule is either invalid or 3291unlikely to match any packet:: 3292 3293 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...] 3294 3295More information on these restrictions can be found in the *rte_flow* 3296documentation. 3297 3298Several items support additional specification structures, for example 3299``ipv4`` allows specifying source and destination addresses as follows:: 3300 3301 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1 3302 dst is 10.2.0.0 / end [...] 3303 3304This rule matches all IPv4 traffic with the specified properties. 3305 3306In this example, ``src`` and ``dst`` are field names of the underlying 3307``struct rte_flow_item_ipv4`` object. All item properties can be specified 3308in a similar fashion. 3309 3310The ``is`` token means that the subsequent value must be matched exactly, 3311and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item`` 3312accordingly. Possible assignment tokens are: 3313 3314- ``is``: match value perfectly (with full bit-mask). 3315- ``spec``: match value according to configured bit-mask. 3316- ``last``: specify upper bound to establish a range. 3317- ``mask``: specify bit-mask with relevant bits set to one. 3318- ``prefix``: generate bit-mask from a prefix length. 3319 3320These yield identical results:: 3321 3322 ipv4 src is 10.1.1.1 3323 3324:: 3325 3326 ipv4 src spec 10.1.1.1 src mask 255.255.255.255 3327 3328:: 3329 3330 ipv4 src spec 10.1.1.1 src prefix 32 3331 3332:: 3333 3334 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value 3335 3336:: 3337 3338 ipv4 src is 10.1.1.1 src last 0 # 0 disables range 3339 3340Inclusive ranges can be defined with ``last``:: 3341 3342 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4 3343 3344Note that ``mask`` affects both ``spec`` and ``last``:: 3345 3346 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0 3347 # matches 10.1.0.0 to 10.2.255.255 3348 3349Properties can be modified multiple times:: 3350 3351 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4 3352 3353:: 3354 3355 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16 3356 3357Pattern items 3358^^^^^^^^^^^^^ 3359 3360This section lists supported pattern items and their attributes, if any. 3361 3362- ``end``: end list of pattern items. 3363 3364- ``void``: no-op pattern item. 3365 3366- ``invert``: perform actions when pattern does not match. 3367 3368- ``any``: match any protocol for the current layer. 3369 3370 - ``num {unsigned}``: number of layers covered. 3371 3372- ``pf``: match traffic from/to the physical function. 3373 3374- ``vf``: match traffic from/to a virtual function ID. 3375 3376 - ``id {unsigned}``: VF ID. 3377 3378- ``phy_port``: match traffic from/to a specific physical port. 3379 3380 - ``index {unsigned}``: physical port index. 3381 3382- ``port_id``: match traffic from/to a given DPDK port ID. 3383 3384 - ``id {unsigned}``: DPDK port ID. 3385 3386- ``mark``: match value set in previously matched flow rule using the mark action. 3387 3388 - ``id {unsigned}``: arbitrary integer value. 3389 3390- ``raw``: match an arbitrary byte string. 3391 3392 - ``relative {boolean}``: look for pattern after the previous item. 3393 - ``search {boolean}``: search pattern from offset (see also limit). 3394 - ``offset {integer}``: absolute or relative offset for pattern. 3395 - ``limit {unsigned}``: search area limit for start of pattern. 3396 - ``pattern {string}``: byte string to look for. 3397 3398- ``eth``: match Ethernet header. 3399 3400 - ``dst {MAC-48}``: destination MAC. 3401 - ``src {MAC-48}``: source MAC. 3402 - ``type {unsigned}``: EtherType or TPID. 3403 3404- ``vlan``: match 802.1Q/ad VLAN tag. 3405 3406 - ``tci {unsigned}``: tag control information. 3407 - ``pcp {unsigned}``: priority code point. 3408 - ``dei {unsigned}``: drop eligible indicator. 3409 - ``vid {unsigned}``: VLAN identifier. 3410 - ``inner_type {unsigned}``: inner EtherType or TPID. 3411 3412- ``ipv4``: match IPv4 header. 3413 3414 - ``tos {unsigned}``: type of service. 3415 - ``ttl {unsigned}``: time to live. 3416 - ``proto {unsigned}``: next protocol ID. 3417 - ``src {ipv4 address}``: source address. 3418 - ``dst {ipv4 address}``: destination address. 3419 3420- ``ipv6``: match IPv6 header. 3421 3422 - ``tc {unsigned}``: traffic class. 3423 - ``flow {unsigned}``: flow label. 3424 - ``proto {unsigned}``: protocol (next header). 3425 - ``hop {unsigned}``: hop limit. 3426 - ``src {ipv6 address}``: source address. 3427 - ``dst {ipv6 address}``: destination address. 3428 3429- ``icmp``: match ICMP header. 3430 3431 - ``type {unsigned}``: ICMP packet type. 3432 - ``code {unsigned}``: ICMP packet code. 3433 3434- ``udp``: match UDP header. 3435 3436 - ``src {unsigned}``: UDP source port. 3437 - ``dst {unsigned}``: UDP destination port. 3438 3439- ``tcp``: match TCP header. 3440 3441 - ``src {unsigned}``: TCP source port. 3442 - ``dst {unsigned}``: TCP destination port. 3443 3444- ``sctp``: match SCTP header. 3445 3446 - ``src {unsigned}``: SCTP source port. 3447 - ``dst {unsigned}``: SCTP destination port. 3448 - ``tag {unsigned}``: validation tag. 3449 - ``cksum {unsigned}``: checksum. 3450 3451- ``vxlan``: match VXLAN header. 3452 3453 - ``vni {unsigned}``: VXLAN identifier. 3454 3455- ``e_tag``: match IEEE 802.1BR E-Tag header. 3456 3457 - ``grp_ecid_b {unsigned}``: GRP and E-CID base. 3458 3459- ``nvgre``: match NVGRE header. 3460 3461 - ``tni {unsigned}``: virtual subnet ID. 3462 3463- ``mpls``: match MPLS header. 3464 3465 - ``label {unsigned}``: MPLS label. 3466 3467- ``gre``: match GRE header. 3468 3469 - ``protocol {unsigned}``: protocol type. 3470 3471- ``fuzzy``: fuzzy pattern match, expect faster than default. 3472 3473 - ``thresh {unsigned}``: accuracy threshold. 3474 3475- ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header. 3476 3477 - ``teid {unsigned}``: tunnel endpoint identifier. 3478 3479- ``geneve``: match GENEVE header. 3480 3481 - ``vni {unsigned}``: virtual network identifier. 3482 - ``protocol {unsigned}``: protocol type. 3483 3484- ``vxlan-gpe``: match VXLAN-GPE header. 3485 3486 - ``vni {unsigned}``: VXLAN-GPE identifier. 3487 3488- ``arp_eth_ipv4``: match ARP header for Ethernet/IPv4. 3489 3490 - ``sha {MAC-48}``: sender hardware address. 3491 - ``spa {ipv4 address}``: sender IPv4 address. 3492 - ``tha {MAC-48}``: target hardware address. 3493 - ``tpa {ipv4 address}``: target IPv4 address. 3494 3495- ``ipv6_ext``: match presence of any IPv6 extension header. 3496 3497 - ``next_hdr {unsigned}``: next header. 3498 3499- ``icmp6``: match any ICMPv6 header. 3500 3501 - ``type {unsigned}``: ICMPv6 type. 3502 - ``code {unsigned}``: ICMPv6 code. 3503 3504- ``icmp6_nd_ns``: match ICMPv6 neighbor discovery solicitation. 3505 3506 - ``target_addr {ipv6 address}``: target address. 3507 3508- ``icmp6_nd_na``: match ICMPv6 neighbor discovery advertisement. 3509 3510 - ``target_addr {ipv6 address}``: target address. 3511 3512- ``icmp6_nd_opt``: match presence of any ICMPv6 neighbor discovery option. 3513 3514 - ``type {unsigned}``: ND option type. 3515 3516- ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery source Ethernet 3517 link-layer address option. 3518 3519 - ``sla {MAC-48}``: source Ethernet LLA. 3520 3521- ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery target Ethernet 3522 link-layer address option. 3523 3524 - ``tla {MAC-48}``: target Ethernet LLA. 3525 3526Actions list 3527^^^^^^^^^^^^ 3528 3529A list of actions starts after the ``actions`` token in the same fashion as 3530`Matching pattern`_; actions are separated by ``/`` tokens and the list is 3531terminated by a mandatory ``end`` action. 3532 3533Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum 3534rte_flow_action_type``). 3535 3536Dropping all incoming UDPv4 packets can be expressed as follows:: 3537 3538 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3539 actions drop / end 3540 3541Several actions have configurable properties which must be specified when 3542there is no valid default value. For example, ``queue`` requires a target 3543queue index. 3544 3545This rule redirects incoming UDPv4 traffic to queue index 6:: 3546 3547 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3548 actions queue index 6 / end 3549 3550While this one could be rejected by PMDs (unspecified queue index):: 3551 3552 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3553 actions queue / end 3554 3555As defined by *rte_flow*, the list is not ordered, all actions of a given 3556rule are performed simultaneously. These are equivalent:: 3557 3558 queue index 6 / void / mark id 42 / end 3559 3560:: 3561 3562 void / mark id 42 / queue index 6 / end 3563 3564All actions in a list should have different types, otherwise only the last 3565action of a given type is taken into account:: 3566 3567 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6 3568 3569:: 3570 3571 drop / drop / drop / end # drop is performed only once 3572 3573:: 3574 3575 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24 3576 3577Considering they are performed simultaneously, opposite and overlapping 3578actions can sometimes be combined when the end result is unambiguous:: 3579 3580 drop / queue index 6 / end # drop has no effect 3581 3582:: 3583 3584 queue index 6 / rss queues 6 7 8 / end # queue has no effect 3585 3586:: 3587 3588 drop / passthru / end # drop has no effect 3589 3590Note that PMDs may still refuse such combinations. 3591 3592Actions 3593^^^^^^^ 3594 3595This section lists supported actions and their attributes, if any. 3596 3597- ``end``: end list of actions. 3598 3599- ``void``: no-op action. 3600 3601- ``passthru``: let subsequent rule process matched packets. 3602 3603- ``jump``: redirect traffic to group on device. 3604 3605 - ``group {unsigned}``: group to redirect to. 3606 3607- ``mark``: attach 32 bit value to packets. 3608 3609 - ``id {unsigned}``: 32 bit value to return with packets. 3610 3611- ``flag``: flag packets. 3612 3613- ``queue``: assign packets to a given queue index. 3614 3615 - ``index {unsigned}``: queue index to use. 3616 3617- ``drop``: drop packets (note: passthru has priority). 3618 3619- ``count``: enable counters for this rule. 3620 3621- ``rss``: spread packets among several queues. 3622 3623 - ``func {hash function}``: RSS hash function to apply, allowed tokens are 3624 the same as `set_hash_global_config`_. 3625 3626 - ``level {unsigned}``: encapsulation level for ``types``. 3627 3628 - ``types [{RSS hash type} [...]] end``: specific RSS hash types, allowed 3629 tokens are the same as `set_hash_input_set`_, except that an empty list 3630 does not disable RSS but instead requests unspecified "best-effort" 3631 settings. 3632 3633 - ``key {string}``: RSS hash key, overrides ``key_len``. 3634 3635 - ``key_len {unsigned}``: RSS hash key length in bytes, can be used in 3636 conjunction with ``key`` to pad or truncate it. 3637 3638 - ``queues [{unsigned} [...]] end``: queue indices to use. 3639 3640- ``pf``: direct traffic to physical function. 3641 3642- ``vf``: direct traffic to a virtual function ID. 3643 3644 - ``original {boolean}``: use original VF ID if possible. 3645 - ``id {unsigned}``: VF ID. 3646 3647- ``phy_port``: direct packets to physical port index. 3648 3649 - ``original {boolean}``: use original port index if possible. 3650 - ``index {unsigned}``: physical port index. 3651 3652- ``port_id``: direct matching traffic to a given DPDK port ID. 3653 3654 - ``original {boolean}``: use original DPDK port ID if possible. 3655 - ``id {unsigned}``: DPDK port ID. 3656 3657- ``of_set_mpls_ttl``: OpenFlow's ``OFPAT_SET_MPLS_TTL``. 3658 3659 - ``mpls_ttl``: MPLS TTL. 3660 3661- ``of_dec_mpls_ttl``: OpenFlow's ``OFPAT_DEC_MPLS_TTL``. 3662 3663- ``of_set_nw_ttl``: OpenFlow's ``OFPAT_SET_NW_TTL``. 3664 3665 - ``nw_ttl``: IP TTL. 3666 3667- ``of_dec_nw_ttl``: OpenFlow's ``OFPAT_DEC_NW_TTL``. 3668 3669- ``of_copy_ttl_out``: OpenFlow's ``OFPAT_COPY_TTL_OUT``. 3670 3671- ``of_copy_ttl_in``: OpenFlow's ``OFPAT_COPY_TTL_IN``. 3672 3673- ``of_pop_vlan``: OpenFlow's ``OFPAT_POP_VLAN``. 3674 3675- ``of_push_vlan``: OpenFlow's ``OFPAT_PUSH_VLAN``. 3676 3677 - ``ethertype``: Ethertype. 3678 3679- ``of_set_vlan_vid``: OpenFlow's ``OFPAT_SET_VLAN_VID``. 3680 3681 - ``vlan_vid``: VLAN id. 3682 3683- ``of_set_vlan_pcp``: OpenFlow's ``OFPAT_SET_VLAN_PCP``. 3684 3685 - ``vlan_pcp``: VLAN priority. 3686 3687- ``of_pop_mpls``: OpenFlow's ``OFPAT_POP_MPLS``. 3688 3689 - ``ethertype``: Ethertype. 3690 3691- ``of_push_mpls``: OpenFlow's ``OFPAT_PUSH_MPLS``. 3692 3693 - ``ethertype``: Ethertype. 3694 3695- ``vxlan_encap``: Performs a VXLAN encapsulation, outer layer configuration 3696 is done through `Config VXLAN Encap outer layers`_. 3697 3698- ``vxlan_decap``: Performs a decapsulation action by stripping all headers of 3699 the VXLAN tunnel network overlay from the matched flow. 3700 3701- ``nvgre_encap``: Performs a NVGRE encapsulation, outer layer configuration 3702 is done through `Config NVGRE Encap outer layers`_. 3703 3704- ``nvgre_decap``: Performs a decapsulation action by stripping all headers of 3705 the NVGRE tunnel network overlay from the matched flow. 3706 3707Destroying flow rules 3708~~~~~~~~~~~~~~~~~~~~~ 3709 3710``flow destroy`` destroys one or more rules from their rule ID (as returned 3711by ``flow create``), this command calls ``rte_flow_destroy()`` as many 3712times as necessary:: 3713 3714 flow destroy {port_id} rule {rule_id} [...] 3715 3716If successful, it will show:: 3717 3718 Flow rule #[...] destroyed 3719 3720It does not report anything for rule IDs that do not exist. The usual error 3721message is shown when a rule cannot be destroyed:: 3722 3723 Caught error type [...] ([...]): [...] 3724 3725``flow flush`` destroys all rules on a device and does not take extra 3726arguments. It is bound to ``rte_flow_flush()``:: 3727 3728 flow flush {port_id} 3729 3730Any errors are reported as above. 3731 3732Creating several rules and destroying them:: 3733 3734 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3735 actions queue index 2 / end 3736 Flow rule #0 created 3737 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3738 actions queue index 3 / end 3739 Flow rule #1 created 3740 testpmd> flow destroy 0 rule 0 rule 1 3741 Flow rule #1 destroyed 3742 Flow rule #0 destroyed 3743 testpmd> 3744 3745The same result can be achieved using ``flow flush``:: 3746 3747 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3748 actions queue index 2 / end 3749 Flow rule #0 created 3750 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3751 actions queue index 3 / end 3752 Flow rule #1 created 3753 testpmd> flow flush 0 3754 testpmd> 3755 3756Non-existent rule IDs are ignored:: 3757 3758 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3759 actions queue index 2 / end 3760 Flow rule #0 created 3761 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3762 actions queue index 3 / end 3763 Flow rule #1 created 3764 testpmd> flow destroy 0 rule 42 rule 10 rule 2 3765 testpmd> 3766 testpmd> flow destroy 0 rule 0 3767 Flow rule #0 destroyed 3768 testpmd> 3769 3770Querying flow rules 3771~~~~~~~~~~~~~~~~~~~ 3772 3773``flow query`` queries a specific action of a flow rule having that 3774ability. Such actions collect information that can be reported using this 3775command. It is bound to ``rte_flow_query()``:: 3776 3777 flow query {port_id} {rule_id} {action} 3778 3779If successful, it will display either the retrieved data for known actions 3780or the following message:: 3781 3782 Cannot display result for action type [...] ([...]) 3783 3784Otherwise, it will complain either that the rule does not exist or that some 3785error occurred:: 3786 3787 Flow rule #[...] not found 3788 3789:: 3790 3791 Caught error type [...] ([...]): [...] 3792 3793Currently only the ``count`` action is supported. This action reports the 3794number of packets that hit the flow rule and the total number of bytes. Its 3795output has the following format:: 3796 3797 count: 3798 hits_set: [...] # whether "hits" contains a valid value 3799 bytes_set: [...] # whether "bytes" contains a valid value 3800 hits: [...] # number of packets 3801 bytes: [...] # number of bytes 3802 3803Querying counters for TCPv6 packets redirected to queue 6:: 3804 3805 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end 3806 actions queue index 6 / count / end 3807 Flow rule #4 created 3808 testpmd> flow query 0 4 count 3809 count: 3810 hits_set: 1 3811 bytes_set: 0 3812 hits: 386446 3813 bytes: 0 3814 testpmd> 3815 3816Listing flow rules 3817~~~~~~~~~~~~~~~~~~ 3818 3819``flow list`` lists existing flow rules sorted by priority and optionally 3820filtered by group identifiers:: 3821 3822 flow list {port_id} [group {group_id}] [...] 3823 3824This command only fails with the following message if the device does not 3825exist:: 3826 3827 Invalid port [...] 3828 3829Output consists of a header line followed by a short description of each 3830flow rule, one per line. There is no output at all when no flow rules are 3831configured on the device:: 3832 3833 ID Group Prio Attr Rule 3834 [...] [...] [...] [...] [...] 3835 3836``Attr`` column flags: 3837 3838- ``i`` for ``ingress``. 3839- ``e`` for ``egress``. 3840 3841Creating several flow rules and listing them:: 3842 3843 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3844 actions queue index 6 / end 3845 Flow rule #0 created 3846 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3847 actions queue index 2 / end 3848 Flow rule #1 created 3849 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end 3850 actions rss queues 6 7 8 end / end 3851 Flow rule #2 created 3852 testpmd> flow list 0 3853 ID Group Prio Attr Rule 3854 0 0 0 i- ETH IPV4 => QUEUE 3855 1 0 0 i- ETH IPV6 => QUEUE 3856 2 0 5 i- ETH IPV4 UDP => RSS 3857 testpmd> 3858 3859Rules are sorted by priority (i.e. group ID first, then priority level):: 3860 3861 testpmd> flow list 1 3862 ID Group Prio Attr Rule 3863 0 0 0 i- ETH => COUNT 3864 6 0 500 i- ETH IPV6 TCP => DROP COUNT 3865 5 0 1000 i- ETH IPV6 ICMP => QUEUE 3866 1 24 0 i- ETH IPV4 UDP => QUEUE 3867 4 24 10 i- ETH IPV4 TCP => DROP 3868 3 24 20 i- ETH IPV4 => DROP 3869 2 24 42 i- ETH IPV4 UDP => QUEUE 3870 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 3871 testpmd> 3872 3873Output can be limited to specific groups:: 3874 3875 testpmd> flow list 1 group 0 group 63 3876 ID Group Prio Attr Rule 3877 0 0 0 i- ETH => COUNT 3878 6 0 500 i- ETH IPV6 TCP => DROP COUNT 3879 5 0 1000 i- ETH IPV6 ICMP => QUEUE 3880 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 3881 testpmd> 3882 3883Toggling isolated mode 3884~~~~~~~~~~~~~~~~~~~~~~ 3885 3886``flow isolate`` can be used to tell the underlying PMD that ingress traffic 3887must only be injected from the defined flow rules; that no default traffic 3888is expected outside those rules and the driver is free to assign more 3889resources to handle them. It is bound to ``rte_flow_isolate()``:: 3890 3891 flow isolate {port_id} {boolean} 3892 3893If successful, enabling or disabling isolated mode shows either:: 3894 3895 Ingress traffic on port [...] 3896 is now restricted to the defined flow rules 3897 3898Or:: 3899 3900 Ingress traffic on port [...] 3901 is not restricted anymore to the defined flow rules 3902 3903Otherwise, in case of error:: 3904 3905 Caught error type [...] ([...]): [...] 3906 3907Mainly due to its side effects, PMDs supporting this mode may not have the 3908ability to toggle it more than once without reinitializing affected ports 3909first (e.g. by exiting testpmd). 3910 3911Enabling isolated mode:: 3912 3913 testpmd> flow isolate 0 true 3914 Ingress traffic on port 0 is now restricted to the defined flow rules 3915 testpmd> 3916 3917Disabling isolated mode:: 3918 3919 testpmd> flow isolate 0 false 3920 Ingress traffic on port 0 is not restricted anymore to the defined flow rules 3921 testpmd> 3922 3923Sample QinQ flow rules 3924~~~~~~~~~~~~~~~~~~~~~~ 3925 3926Before creating QinQ rule(s) the following commands should be issued to enable QinQ:: 3927 3928 testpmd> port stop 0 3929 testpmd> vlan set qinq on 0 3930 3931The above command sets the inner and outer TPID's to 0x8100. 3932 3933To change the TPID's the following commands should be used:: 3934 3935 testpmd> vlan set outer tpid 0xa100 0 3936 testpmd> vlan set inner tpid 0x9100 0 3937 testpmd> port start 0 3938 3939Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM. 3940 3941:: 3942 3943 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 / 3944 vlan tci is 456 / end actions vf id 1 / queue index 0 / end 3945 Flow rule #0 validated 3946 3947 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 / 3948 vlan tci is 456 / end actions vf id 123 / queue index 0 / end 3949 Flow rule #0 created 3950 3951 testpmd> flow list 0 3952 ID Group Prio Attr Rule 3953 0 0 0 i- ETH VLAN VLAN=>VF QUEUE 3954 3955Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host. 3956 3957:: 3958 3959 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 / 3960 vlan tci is 654 / end actions pf / queue index 0 / end 3961 Flow rule #1 validated 3962 3963 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 / 3964 vlan tci is 654 / end actions pf / queue index 1 / end 3965 Flow rule #1 created 3966 3967 testpmd> flow list 0 3968 ID Group Prio Attr Rule 3969 0 0 0 i- ETH VLAN VLAN=>VF QUEUE 3970 1 0 0 i- ETH VLAN VLAN=>PF QUEUE 3971 3972Sample VXLAN encapsulation rule 3973~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 3974 3975VXLAN encapsulation outer layer has default value pre-configured in testpmd 3976source code, those can be changed by using the following commands 3977 3978IPv4 VXLAN outer header:: 3979 3980 testpmd> set vxlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src 127.0.0.1 3981 ip-dst 128.0.0.1 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 3982 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 3983 queue index 0 / end 3984 3985 testpmd> set vxlan-with-vlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src 3986 127.0.0.1 ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11 3987 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 3991IPv6 VXLAN outer header:: 3992 3993 testpmd> set vxlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4 ip-src ::1 3994 ip-dst ::2222 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 3995 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 3996 queue index 0 / end 3997 3998 testpmd> set vxlan-with-vlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4 3999 ip-src ::1 ip-dst ::2222 vlan-tci 34 eth-src 11:11:11:11:11:11 4000 eth-dst 22:22:22:22:22:22 4001 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 4002 queue index 0 / end 4003 4004Sample NVGRE encapsulation rule 4005~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4006 4007NVGRE encapsulation outer layer has default value pre-configured in testpmd 4008source code, those can be changed by using the following commands 4009 4010IPv4 NVGRE outer header:: 4011 4012 testpmd> set nvgre ip-version ipv4 tni 4 ip-src 127.0.0.1 ip-dst 128.0.0.1 4013 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4014 testpmd> flow create 0 ingress pattern end actions nvgre_encap / 4015 queue index 0 / end 4016 4017 testpmd> set nvgre-with-vlan ip-version ipv4 tni 4 ip-src 127.0.0.1 4018 ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11 4019 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 4023IPv6 NVGRE outer header:: 4024 4025 testpmd> set nvgre ip-version ipv6 tni 4 ip-src ::1 ip-dst ::2222 4026 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4027 testpmd> flow create 0 ingress pattern end actions nvgre_encap / 4028 queue index 0 / end 4029 4030 testpmd> set nvgre-with-vlan ip-version ipv6 tni 4 ip-src ::1 ip-dst ::2222 4031 vlan-tci 34 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4032 testpmd> flow create 0 ingress pattern end actions nvgre_encap / 4033 queue index 0 / end 4034 4035BPF Functions 4036-------------- 4037 4038The following sections show functions to load/unload eBPF based filters. 4039 4040bpf-load 4041~~~~~~~~ 4042 4043Load an eBPF program as a callback for partciular RX/TX queue:: 4044 4045 testpmd> bpf-load rx|tx (portid) (queueid) (load-flags) (bpf-prog-filename) 4046 4047The available load-flags are: 4048 4049* ``J``: use JIT generated native code, otherwise BPF interpreter will be used. 4050 4051* ``M``: assume input parameter is a pointer to rte_mbuf, otherwise assume it is a pointer to first segment's data. 4052 4053* ``-``: none. 4054 4055.. note:: 4056 4057 You'll need clang v3.7 or above to build bpf program you'd like to load 4058 4059For example: 4060 4061.. code-block:: console 4062 4063 cd test/bpf 4064 clang -O2 -target bpf -c t1.c 4065 4066Then to load (and JIT compile) t1.o at RX queue 0, port 1:: 4067 4068.. code-block:: console 4069 4070 testpmd> bpf-load rx 1 0 J ./dpdk.org/test/bpf/t1.o 4071 4072To load (not JITed) t1.o at TX queue 0, port 0:: 4073 4074.. code-block:: console 4075 4076 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o 4077 4078bpf-unload 4079~~~~~~~~~~ 4080 4081Unload previously loaded eBPF program for partciular RX/TX queue:: 4082 4083 testpmd> bpf-unload rx|tx (portid) (queueid) 4084 4085For example to unload BPF filter from TX queue 0, port 0: 4086 4087.. code-block:: console 4088 4089 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o 4090