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 [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|outer-udp) (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* ``outer-udp`` relates to the outer UDP layer in the case where the packet is recognized 871 as a tunnel packet by the forwarding engine (vxlan, vxlan-gpe are 872 supported). See also the ``csum parse-tunnel`` command. 873 874.. note:: 875 876 Check the NIC Datasheet for hardware limits. 877 878RSS queue region 879~~~~~~~~~~~~~~~~ 880 881Set RSS queue region span on a port:: 882 883 testpmd> set port (port_id) queue-region region_id (value) \ 884 queue_start_index (value) queue_num (value) 885 886Set flowtype mapping on a RSS queue region on a port:: 887 888 testpmd> set port (port_id) queue-region region_id (value) flowtype (value) 889 890where: 891 892* For the flowtype(pctype) of packet,the specific index for each type has 893 been defined in file i40e_type.h as enum i40e_filter_pctype. 894 895Set user priority mapping on a RSS queue region on a port:: 896 897 testpmd> set port (port_id) queue-region UP (value) region_id (value) 898 899Flush all queue region related configuration on a port:: 900 901 testpmd> set port (port_id) queue-region flush (on|off) 902 903where: 904 905* "on"is just an enable function which server for other configuration, 906 it is for all configuration about queue region from up layer, 907 at first will only keep in DPDK softwarestored in driver, 908 only after "flush on", it commit all configuration to HW. 909 "off" is just clean all configuration about queue region just now, 910 and restore all to DPDK i40e driver default config when start up. 911 912Show all queue region related configuration info on a port:: 913 914 testpmd> show port (port_id) queue-region 915 916.. note:: 917 918 Queue region only support on PF by now, so these command is 919 only for configuration of queue region on PF port. 920 921csum parse-tunnel 922~~~~~~~~~~~~~~~~~ 923 924Define how tunneled packets should be handled by the csum forward 925engine:: 926 927 testpmd> csum parse-tunnel (on|off) (tx_port_id) 928 929If enabled, the csum forward engine will try to recognize supported 930tunnel headers (vxlan, gre, ipip). 931 932If disabled, treat tunnel packets as non-tunneled packets (a inner 933header is handled as a packet payload). 934 935.. note:: 936 937 The port argument is the TX port like in the ``csum set`` command. 938 939Example: 940 941Consider a packet in packet like the following:: 942 943 eth_out/ipv4_out/udp_out/vxlan/eth_in/ipv4_in/tcp_in 944 945* If parse-tunnel is enabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set`` 946 command relate to the inner headers (here ``ipv4_in`` and ``tcp_in``), and the 947 ``outer-ip|outer-udp`` parameter relates to the outer headers (here ``ipv4_out`` and ``udp_out``). 948 949* If parse-tunnel is disabled, the ``ip|udp|tcp|sctp`` parameters of ``csum set`` 950 command relate to the outer headers, here ``ipv4_out`` and ``udp_out``. 951 952csum show 953~~~~~~~~~ 954 955Display tx checksum offload configuration:: 956 957 testpmd> csum show (port_id) 958 959tso set 960~~~~~~~ 961 962Enable TCP Segmentation Offload (TSO) in the ``csum`` forwarding engine:: 963 964 testpmd> tso set (segsize) (port_id) 965 966.. note:: 967 968 Check the NIC datasheet for hardware limits. 969 970tso show 971~~~~~~~~ 972 973Display the status of TCP Segmentation Offload:: 974 975 testpmd> tso show (port_id) 976 977set port - gro 978~~~~~~~~~~~~~~ 979 980Enable or disable GRO in ``csum`` forwarding engine:: 981 982 testpmd> set port <port_id> gro on|off 983 984If enabled, the csum forwarding engine will perform GRO on the TCP/IPv4 985packets received from the given port. 986 987If disabled, packets received from the given port won't be performed 988GRO. By default, GRO is disabled for all ports. 989 990.. note:: 991 992 When enable GRO for a port, TCP/IPv4 packets received from the port 993 will be performed GRO. After GRO, all merged packets have bad 994 checksums, since the GRO library doesn't re-calculate checksums for 995 the merged packets. Therefore, if users want the merged packets to 996 have correct checksums, please select HW IP checksum calculation and 997 HW TCP checksum calculation for the port which the merged packets are 998 transmitted to. 999 1000show port - gro 1001~~~~~~~~~~~~~~~ 1002 1003Display GRO configuration for a given port:: 1004 1005 testpmd> show port <port_id> gro 1006 1007set gro flush 1008~~~~~~~~~~~~~ 1009 1010Set the cycle to flush the GROed packets from reassembly tables:: 1011 1012 testpmd> set gro flush <cycles> 1013 1014When enable GRO, the csum forwarding engine performs GRO on received 1015packets, and the GROed packets are stored in reassembly tables. Users 1016can use this command to determine when the GROed packets are flushed 1017from the reassembly tables. 1018 1019The ``cycles`` is measured in GRO operation times. The csum forwarding 1020engine flushes the GROed packets from the tables every ``cycles`` GRO 1021operations. 1022 1023By default, the value of ``cycles`` is 1, which means flush GROed packets 1024from the reassembly tables as soon as one GRO operation finishes. The value 1025of ``cycles`` should be in the range of 1 to ``GRO_MAX_FLUSH_CYCLES``. 1026 1027Please note that the large value of ``cycles`` may cause the poor TCP/IP 1028stack performance. Because the GROed packets are delayed to arrive the 1029stack, thus causing more duplicated ACKs and TCP retransmissions. 1030 1031set port - gso 1032~~~~~~~~~~~~~~ 1033 1034Toggle per-port GSO support in ``csum`` forwarding engine:: 1035 1036 testpmd> set port <port_id> gso on|off 1037 1038If enabled, the csum forwarding engine will perform GSO on supported IPv4 1039packets, transmitted on the given port. 1040 1041If disabled, packets transmitted on the given port will not undergo GSO. 1042By default, GSO is disabled for all ports. 1043 1044.. note:: 1045 1046 When GSO is enabled on a port, supported IPv4 packets transmitted on that 1047 port undergo GSO. Afterwards, the segmented packets are represented by 1048 multi-segment mbufs; however, the csum forwarding engine doesn't calculation 1049 of checksums for GSO'd segments in SW. As a result, if users want correct 1050 checksums in GSO segments, they should enable HW checksum calculation for 1051 GSO-enabled ports. 1052 1053 For example, HW checksum calculation for VxLAN GSO'd packets may be enabled 1054 by setting the following options in the csum forwarding engine: 1055 1056 testpmd> csum set outer_ip hw <port_id> 1057 1058 testpmd> csum set ip hw <port_id> 1059 1060 testpmd> csum set tcp hw <port_id> 1061 1062 UDP GSO is the same as IP fragmentation, which treats the UDP header 1063 as the payload and does not modify it during segmentation. That is, 1064 after UDP GSO, only the first output fragment has the original UDP 1065 header. Therefore, users need to enable HW IP checksum calculation 1066 and SW UDP checksum calculation for GSO-enabled ports, if they want 1067 correct checksums for UDP/IPv4 packets. 1068 1069set gso segsz 1070~~~~~~~~~~~~~ 1071 1072Set the maximum GSO segment size (measured in bytes), which includes the 1073packet header and the packet payload for GSO-enabled ports (global):: 1074 1075 testpmd> set gso segsz <length> 1076 1077show port - gso 1078~~~~~~~~~~~~~~~ 1079 1080Display the status of Generic Segmentation Offload for a given port:: 1081 1082 testpmd> show port <port_id> gso 1083 1084mac_addr add 1085~~~~~~~~~~~~ 1086 1087Add an alternative MAC address to a port:: 1088 1089 testpmd> mac_addr add (port_id) (XX:XX:XX:XX:XX:XX) 1090 1091mac_addr remove 1092~~~~~~~~~~~~~~~ 1093 1094Remove a MAC address from a port:: 1095 1096 testpmd> mac_addr remove (port_id) (XX:XX:XX:XX:XX:XX) 1097 1098mac_addr add (for VF) 1099~~~~~~~~~~~~~~~~~~~~~ 1100 1101Add an alternative MAC address for a VF to a port:: 1102 1103 testpmd> mac_add add port (port_id) vf (vf_id) (XX:XX:XX:XX:XX:XX) 1104 1105mac_addr set 1106~~~~~~~~~~~~ 1107 1108Set the default MAC address for a port:: 1109 1110 testpmd> mac_addr set (port_id) (XX:XX:XX:XX:XX:XX) 1111 1112mac_addr set (for VF) 1113~~~~~~~~~~~~~~~~~~~~~ 1114 1115Set the MAC address for a VF from the PF:: 1116 1117 testpmd> set vf mac addr (port_id) (vf_id) (XX:XX:XX:XX:XX:XX) 1118 1119set eth-peer 1120~~~~~~~~~~~~ 1121 1122Set the forwarding peer address for certain port:: 1123 1124 testpmd> set eth-peer (port_id) (perr_addr) 1125 1126This is equivalent to the ``--eth-peer`` command-line option. 1127 1128set port-uta 1129~~~~~~~~~~~~ 1130 1131Set the unicast hash filter(s) on/off for a port:: 1132 1133 testpmd> set port (port_id) uta (XX:XX:XX:XX:XX:XX|all) (on|off) 1134 1135set promisc 1136~~~~~~~~~~~ 1137 1138Set the promiscuous mode on for a port or for all ports. 1139In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1140 1141 testpmd> set promisc (port_id|all) (on|off) 1142 1143set allmulti 1144~~~~~~~~~~~~ 1145 1146Set the allmulti mode for a port or for all ports:: 1147 1148 testpmd> set allmulti (port_id|all) (on|off) 1149 1150Same as the ifconfig (8) option. Controls how multicast packets are handled. 1151 1152set promisc (for VF) 1153~~~~~~~~~~~~~~~~~~~~ 1154 1155Set the unicast promiscuous mode for a VF from PF. 1156It's supported by Intel i40e NICs now. 1157In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1158 1159 testpmd> set vf promisc (port_id) (vf_id) (on|off) 1160 1161set allmulticast (for VF) 1162~~~~~~~~~~~~~~~~~~~~~~~~~ 1163 1164Set the multicast promiscuous mode for a VF from PF. 1165It's supported by Intel i40e NICs now. 1166In promiscuous mode packets are not dropped if they aren't for the specified MAC address:: 1167 1168 testpmd> set vf allmulti (port_id) (vf_id) (on|off) 1169 1170set tx max bandwidth (for VF) 1171~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1172 1173Set TX max absolute bandwidth (Mbps) for a VF from PF:: 1174 1175 testpmd> set vf tx max-bandwidth (port_id) (vf_id) (max_bandwidth) 1176 1177set tc tx min bandwidth (for VF) 1178~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1179 1180Set all TCs' TX min relative bandwidth (%) for a VF from PF:: 1181 1182 testpmd> set vf tc tx min-bandwidth (port_id) (vf_id) (bw1, bw2, ...) 1183 1184set tc tx max bandwidth (for VF) 1185~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1186 1187Set a TC's TX max absolute bandwidth (Mbps) for a VF from PF:: 1188 1189 testpmd> set vf tc tx max-bandwidth (port_id) (vf_id) (tc_no) (max_bandwidth) 1190 1191set tc strict link priority mode 1192~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1193 1194Set some TCs' strict link priority mode on a physical port:: 1195 1196 testpmd> set tx strict-link-priority (port_id) (tc_bitmap) 1197 1198set tc tx min bandwidth 1199~~~~~~~~~~~~~~~~~~~~~~~ 1200 1201Set all TCs' TX min relative bandwidth (%) globally for all PF and VFs:: 1202 1203 testpmd> set tc tx min-bandwidth (port_id) (bw1, bw2, ...) 1204 1205set flow_ctrl rx 1206~~~~~~~~~~~~~~~~ 1207 1208Set the link flow control parameter on a port:: 1209 1210 testpmd> set flow_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 1211 (pause_time) (send_xon) mac_ctrl_frame_fwd (on|off) \ 1212 autoneg (on|off) (port_id) 1213 1214Where: 1215 1216* ``high_water`` (integer): High threshold value to trigger XOFF. 1217 1218* ``low_water`` (integer): Low threshold value to trigger XON. 1219 1220* ``pause_time`` (integer): Pause quota in the Pause frame. 1221 1222* ``send_xon`` (0/1): Send XON frame. 1223 1224* ``mac_ctrl_frame_fwd``: Enable receiving MAC control frames. 1225 1226* ``autoneg``: Change the auto-negotiation parameter. 1227 1228set pfc_ctrl rx 1229~~~~~~~~~~~~~~~ 1230 1231Set the priority flow control parameter on a port:: 1232 1233 testpmd> set pfc_ctrl rx (on|off) tx (on|off) (high_water) (low_water) \ 1234 (pause_time) (priority) (port_id) 1235 1236Where: 1237 1238* ``high_water`` (integer): High threshold value. 1239 1240* ``low_water`` (integer): Low threshold value. 1241 1242* ``pause_time`` (integer): Pause quota in the Pause frame. 1243 1244* ``priority`` (0-7): VLAN User Priority. 1245 1246set stat_qmap 1247~~~~~~~~~~~~~ 1248 1249Set statistics mapping (qmapping 0..15) for RX/TX queue on port:: 1250 1251 testpmd> set stat_qmap (tx|rx) (port_id) (queue_id) (qmapping) 1252 1253For example, to set rx queue 2 on port 0 to mapping 5:: 1254 1255 testpmd>set stat_qmap rx 0 2 5 1256 1257set xstats-hide-zero 1258~~~~~~~~~~~~~~~~~~~~ 1259 1260Set the option to hide zero values for xstats display:: 1261 1262 testpmd> set xstats-hide-zero on|off 1263 1264.. note:: 1265 1266 By default, the zero values are displayed for xstats. 1267 1268set port - rx/tx (for VF) 1269~~~~~~~~~~~~~~~~~~~~~~~~~ 1270 1271Set VF receive/transmit from a port:: 1272 1273 testpmd> set port (port_id) vf (vf_id) (rx|tx) (on|off) 1274 1275set port - mac address filter (for VF) 1276~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1277 1278Add/Remove unicast or multicast MAC addr filter for a VF:: 1279 1280 testpmd> set port (port_id) vf (vf_id) (mac_addr) \ 1281 (exact-mac|exact-mac-vlan|hashmac|hashmac-vlan) (on|off) 1282 1283set port - rx mode(for VF) 1284~~~~~~~~~~~~~~~~~~~~~~~~~~ 1285 1286Set the VF receive mode of a port:: 1287 1288 testpmd> set port (port_id) vf (vf_id) \ 1289 rxmode (AUPE|ROPE|BAM|MPE) (on|off) 1290 1291The available receive modes are: 1292 1293* ``AUPE``: Accepts untagged VLAN. 1294 1295* ``ROPE``: Accepts unicast hash. 1296 1297* ``BAM``: Accepts broadcast packets. 1298 1299* ``MPE``: Accepts all multicast packets. 1300 1301set port - tx_rate (for Queue) 1302~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1303 1304Set TX rate limitation for a queue on a port:: 1305 1306 testpmd> set port (port_id) queue (queue_id) rate (rate_value) 1307 1308set port - tx_rate (for VF) 1309~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1310 1311Set TX rate limitation for queues in VF on a port:: 1312 1313 testpmd> set port (port_id) vf (vf_id) rate (rate_value) queue_mask (queue_mask) 1314 1315set port - mirror rule 1316~~~~~~~~~~~~~~~~~~~~~~ 1317 1318Set pool or vlan type mirror rule for a port:: 1319 1320 testpmd> set port (port_id) mirror-rule (rule_id) \ 1321 (pool-mirror-up|pool-mirror-down|vlan-mirror) \ 1322 (poolmask|vlanid[,vlanid]*) dst-pool (pool_id) (on|off) 1323 1324Set link mirror rule for a port:: 1325 1326 testpmd> set port (port_id) mirror-rule (rule_id) \ 1327 (uplink-mirror|downlink-mirror) dst-pool (pool_id) (on|off) 1328 1329For example to enable mirror traffic with vlan 0,1 to pool 0:: 1330 1331 set port 0 mirror-rule 0 vlan-mirror 0,1 dst-pool 0 on 1332 1333reset port - mirror rule 1334~~~~~~~~~~~~~~~~~~~~~~~~ 1335 1336Reset a mirror rule for a port:: 1337 1338 testpmd> reset port (port_id) mirror-rule (rule_id) 1339 1340set flush_rx 1341~~~~~~~~~~~~ 1342 1343Set the flush on RX streams before forwarding. 1344The default is flush ``on``. 1345Mainly used with PCAP drivers to turn off the default behavior of flushing the first 512 packets on RX streams:: 1346 1347 testpmd> set flush_rx off 1348 1349set bypass mode 1350~~~~~~~~~~~~~~~ 1351 1352Set the bypass mode for the lowest port on bypass enabled NIC:: 1353 1354 testpmd> set bypass mode (normal|bypass|isolate) (port_id) 1355 1356set bypass event 1357~~~~~~~~~~~~~~~~ 1358 1359Set the event required to initiate specified bypass mode for the lowest port on a bypass enabled:: 1360 1361 testpmd> set bypass event (timeout|os_on|os_off|power_on|power_off) \ 1362 mode (normal|bypass|isolate) (port_id) 1363 1364Where: 1365 1366* ``timeout``: Enable bypass after watchdog timeout. 1367 1368* ``os_on``: Enable bypass when OS/board is powered on. 1369 1370* ``os_off``: Enable bypass when OS/board is powered off. 1371 1372* ``power_on``: Enable bypass when power supply is turned on. 1373 1374* ``power_off``: Enable bypass when power supply is turned off. 1375 1376 1377set bypass timeout 1378~~~~~~~~~~~~~~~~~~ 1379 1380Set the bypass watchdog timeout to ``n`` seconds where 0 = instant:: 1381 1382 testpmd> set bypass timeout (0|1.5|2|3|4|8|16|32) 1383 1384show bypass config 1385~~~~~~~~~~~~~~~~~~ 1386 1387Show the bypass configuration for a bypass enabled NIC using the lowest port on the NIC:: 1388 1389 testpmd> show bypass config (port_id) 1390 1391set link up 1392~~~~~~~~~~~ 1393 1394Set link up for a port:: 1395 1396 testpmd> set link-up port (port id) 1397 1398set link down 1399~~~~~~~~~~~~~ 1400 1401Set link down for a port:: 1402 1403 testpmd> set link-down port (port id) 1404 1405E-tag set 1406~~~~~~~~~ 1407 1408Enable E-tag insertion for a VF on a port:: 1409 1410 testpmd> E-tag set insertion on port-tag-id (value) port (port_id) vf (vf_id) 1411 1412Disable E-tag insertion for a VF on a port:: 1413 1414 testpmd> E-tag set insertion off port (port_id) vf (vf_id) 1415 1416Enable/disable E-tag stripping on a port:: 1417 1418 testpmd> E-tag set stripping (on|off) port (port_id) 1419 1420Enable/disable E-tag based forwarding on a port:: 1421 1422 testpmd> E-tag set forwarding (on|off) port (port_id) 1423 1424Add an E-tag forwarding filter on a port:: 1425 1426 testpmd> E-tag set filter add e-tag-id (value) dst-pool (pool_id) port (port_id) 1427 1428Delete an E-tag forwarding filter on a port:: 1429 testpmd> E-tag set filter del e-tag-id (value) port (port_id) 1430 1431ddp add 1432~~~~~~~ 1433 1434Load a dynamic device personalization (DDP) profile and store backup profile:: 1435 1436 testpmd> ddp add (port_id) (profile_path[,backup_profile_path]) 1437 1438ddp del 1439~~~~~~~ 1440 1441Delete a dynamic device personalization profile and restore backup profile:: 1442 1443 testpmd> ddp del (port_id) (backup_profile_path) 1444 1445ptype mapping 1446~~~~~~~~~~~~~ 1447 1448List all items from the ptype mapping table:: 1449 1450 testpmd> ptype mapping get (port_id) (valid_only) 1451 1452Where: 1453 1454* ``valid_only``: A flag indicates if only list valid items(=1) or all itemss(=0). 1455 1456Replace a specific or a group of software defined ptype with a new one:: 1457 1458 testpmd> ptype mapping replace (port_id) (target) (mask) (pkt_type) 1459 1460where: 1461 1462* ``target``: A specific software ptype or a mask to represent a group of software ptypes. 1463 1464* ``mask``: A flag indicate if "target" is a specific software ptype(=0) or a ptype mask(=1). 1465 1466* ``pkt_type``: The new software ptype to replace the old ones. 1467 1468Update hardware defined ptype to software defined packet type mapping table:: 1469 1470 testpmd> ptype mapping update (port_id) (hw_ptype) (sw_ptype) 1471 1472where: 1473 1474* ``hw_ptype``: hardware ptype as the index of the ptype mapping table. 1475 1476* ``sw_ptype``: software ptype as the value of the ptype mapping table. 1477 1478Reset ptype mapping table:: 1479 1480 testpmd> ptype mapping reset (port_id) 1481 1482config per port Rx offloading 1483~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1484 1485Enable or disable a per port Rx offloading on all Rx queues of a port:: 1486 1487 testpmd> port config (port_id) rx_offload (offloading) on|off 1488 1489* ``offloading``: can be any of these offloading capability: 1490 vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro, 1491 qinq_strip, outer_ipv4_cksum, macsec_strip, 1492 header_split, vlan_filter, vlan_extend, jumbo_frame, 1493 crc_strip, scatter, timestamp, security, keep_crc 1494 1495This command should be run when the port is stopped, or else it will fail. 1496 1497config per queue Rx offloading 1498~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1499 1500Enable or disable a per queue Rx offloading only on a specific Rx queue:: 1501 1502 testpmd> port (port_id) rxq (queue_id) rx_offload (offloading) on|off 1503 1504* ``offloading``: can be any of these offloading capability: 1505 vlan_strip, ipv4_cksum, udp_cksum, tcp_cksum, tcp_lro, 1506 qinq_strip, outer_ipv4_cksum, macsec_strip, 1507 header_split, vlan_filter, vlan_extend, jumbo_frame, 1508 crc_strip, scatter, timestamp, security, keep_crc 1509 1510This command should be run when the port is stopped, or else it will fail. 1511 1512config per port Tx offloading 1513~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1514 1515Enable or disable a per port Tx offloading on all Tx queues of a port:: 1516 1517 testpmd> port config (port_id) tx_offload (offloading) on|off 1518 1519* ``offloading``: can be any of these offloading capability: 1520 vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum, 1521 sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum, 1522 qinq_insert, vxlan_tnl_tso, gre_tnl_tso, 1523 ipip_tnl_tso, geneve_tnl_tso, macsec_insert, 1524 mt_lockfree, multi_segs, mbuf_fast_free, security 1525 1526This command should be run when the port is stopped, or else it will fail. 1527 1528config per queue Tx offloading 1529~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1530 1531Enable or disable a per queue Tx offloading only on a specific Tx queue:: 1532 1533 testpmd> port (port_id) txq (queue_id) tx_offload (offloading) on|off 1534 1535* ``offloading``: can be any of these offloading capability: 1536 vlan_insert, ipv4_cksum, udp_cksum, tcp_cksum, 1537 sctp_cksum, tcp_tso, udp_tso, outer_ipv4_cksum, 1538 qinq_insert, vxlan_tnl_tso, gre_tnl_tso, 1539 ipip_tnl_tso, geneve_tnl_tso, macsec_insert, 1540 mt_lockfree, multi_segs, mbuf_fast_free, security 1541 1542This command should be run when the port is stopped, or else it will fail. 1543 1544Config VXLAN Encap outer layers 1545~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1546 1547Configure the outer layer to encapsulate a packet inside a VXLAN tunnel:: 1548 1549 set vxlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \ 1550 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) eth-src (eth-src) \ 1551 eth-dst (eth-dst) 1552 1553 set vxlan-with-vlan ip-version (ipv4|ipv6) vni (vni) udp-src (udp-src) \ 1554 udp-dst (udp-dst) ip-src (ip-src) ip-dst (ip-dst) vlan-tci (vlan-tci) \ 1555 eth-src (eth-src) eth-dst (eth-dst) 1556 1557Those command will set an internal configuration inside testpmd, any following 1558flow rule using the action vxlan_encap will use the last configuration set. 1559To have a different encapsulation header, one of those commands must be called 1560before the flow rule creation. 1561 1562Config NVGRE Encap outer layers 1563~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1564 1565Configure the outer layer to encapsulate a packet inside a NVGRE tunnel:: 1566 1567 set nvgre ip-version (ipv4|ipv6) tni (tni) ip-src (ip-src) ip-dst (ip-dst) \ 1568 eth-src (eth-src) eth-dst (eth-dst) 1569 set nvgre-with-vlan ip-version (ipv4|ipv6) tni (tni) ip-src (ip-src) \ 1570 ip-dst (ip-dst) vlan-tci (vlan-tci) eth-src (eth-src) eth-dst (eth-dst) 1571 1572Those command will set an internal configuration inside testpmd, any following 1573flow rule using the action nvgre_encap will use the last configuration set. 1574To have a different encapsulation header, one of those commands must be called 1575before the flow rule creation. 1576 1577Port Functions 1578-------------- 1579 1580The following sections show functions for configuring ports. 1581 1582.. note:: 1583 1584 Port configuration changes only become active when forwarding is started/restarted. 1585 1586port attach 1587~~~~~~~~~~~ 1588 1589Attach a port specified by pci address or virtual device args:: 1590 1591 testpmd> port attach (identifier) 1592 1593To attach a new pci device, the device should be recognized by kernel first. 1594Then it should be moved under DPDK management. 1595Finally the port can be attached to testpmd. 1596 1597For example, to move a pci device using ixgbe under DPDK management: 1598 1599.. code-block:: console 1600 1601 # Check the status of the available devices. 1602 ./usertools/dpdk-devbind.py --status 1603 1604 Network devices using DPDK-compatible driver 1605 ============================================ 1606 <none> 1607 1608 Network devices using kernel driver 1609 =================================== 1610 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused= 1611 1612 1613 # Bind the device to igb_uio. 1614 sudo ./usertools/dpdk-devbind.py -b igb_uio 0000:0a:00.0 1615 1616 1617 # Recheck the status of the devices. 1618 ./usertools/dpdk-devbind.py --status 1619 Network devices using DPDK-compatible driver 1620 ============================================ 1621 0000:0a:00.0 '82599ES 10-Gigabit' drv=igb_uio unused= 1622 1623To attach a port created by virtual device, above steps are not needed. 1624 1625For example, to attach a port whose pci address is 0000:0a:00.0. 1626 1627.. code-block:: console 1628 1629 testpmd> port attach 0000:0a:00.0 1630 Attaching a new port... 1631 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1632 EAL: probe driver: 8086:10fb rte_ixgbe_pmd 1633 EAL: PCI memory mapped at 0x7f83bfa00000 1634 EAL: PCI memory mapped at 0x7f83bfa80000 1635 PMD: eth_ixgbe_dev_init(): MAC: 2, PHY: 18, SFP+: 5 1636 PMD: eth_ixgbe_dev_init(): port 0 vendorID=0x8086 deviceID=0x10fb 1637 Port 0 is attached. Now total ports is 1 1638 Done 1639 1640For example, to attach a port created by pcap PMD. 1641 1642.. code-block:: console 1643 1644 testpmd> port attach net_pcap0 1645 Attaching a new port... 1646 PMD: Initializing pmd_pcap for net_pcap0 1647 PMD: Creating pcap-backed ethdev on numa socket 0 1648 Port 0 is attached. Now total ports is 1 1649 Done 1650 1651In this case, identifier is ``net_pcap0``. 1652This identifier format is the same as ``--vdev`` format of DPDK applications. 1653 1654For example, to re-attach a bonded port which has been previously detached, 1655the mode and slave parameters must be given. 1656 1657.. code-block:: console 1658 1659 testpmd> port attach net_bond_0,mode=0,slave=1 1660 Attaching a new port... 1661 EAL: Initializing pmd_bond for net_bond_0 1662 EAL: Create bonded device net_bond_0 on port 0 in mode 0 on socket 0. 1663 Port 0 is attached. Now total ports is 1 1664 Done 1665 1666 1667port detach 1668~~~~~~~~~~~ 1669 1670Detach a specific port:: 1671 1672 testpmd> port detach (port_id) 1673 1674Before detaching a port, the port should be stopped and closed. 1675 1676For example, to detach a pci device port 0. 1677 1678.. code-block:: console 1679 1680 testpmd> port stop 0 1681 Stopping ports... 1682 Done 1683 testpmd> port close 0 1684 Closing ports... 1685 Done 1686 1687 testpmd> port detach 0 1688 Detaching a port... 1689 EAL: PCI device 0000:0a:00.0 on NUMA socket -1 1690 EAL: remove driver: 8086:10fb rte_ixgbe_pmd 1691 EAL: PCI memory unmapped at 0x7f83bfa00000 1692 EAL: PCI memory unmapped at 0x7f83bfa80000 1693 Done 1694 1695 1696For example, to detach a virtual device port 0. 1697 1698.. code-block:: console 1699 1700 testpmd> port stop 0 1701 Stopping ports... 1702 Done 1703 testpmd> port close 0 1704 Closing ports... 1705 Done 1706 1707 testpmd> port detach 0 1708 Detaching a port... 1709 PMD: Closing pcap ethdev on numa socket 0 1710 Port 'net_pcap0' is detached. Now total ports is 0 1711 Done 1712 1713To remove a pci device completely from the system, first detach the port from testpmd. 1714Then the device should be moved under kernel management. 1715Finally the device can be removed using kernel pci hotplug functionality. 1716 1717For example, to move a pci device under kernel management: 1718 1719.. code-block:: console 1720 1721 sudo ./usertools/dpdk-devbind.py -b ixgbe 0000:0a:00.0 1722 1723 ./usertools/dpdk-devbind.py --status 1724 1725 Network devices using DPDK-compatible driver 1726 ============================================ 1727 <none> 1728 1729 Network devices using kernel driver 1730 =================================== 1731 0000:0a:00.0 '82599ES 10-Gigabit' if=eth2 drv=ixgbe unused=igb_uio 1732 1733To remove a port created by a virtual device, above steps are not needed. 1734 1735port start 1736~~~~~~~~~~ 1737 1738Start all ports or a specific port:: 1739 1740 testpmd> port start (port_id|all) 1741 1742port stop 1743~~~~~~~~~ 1744 1745Stop all ports or a specific port:: 1746 1747 testpmd> port stop (port_id|all) 1748 1749port close 1750~~~~~~~~~~ 1751 1752Close all ports or a specific port:: 1753 1754 testpmd> port close (port_id|all) 1755 1756port config - queue ring size 1757~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1758 1759Configure a rx/tx queue ring size:: 1760 1761 testpmd> port (port_id) (rxq|txq) (queue_id) ring_size (value) 1762 1763Only take effect after command that (re-)start the port or command that setup specific queue. 1764 1765port start/stop queue 1766~~~~~~~~~~~~~~~~~~~~~ 1767 1768Start/stop a rx/tx queue on a specific port:: 1769 1770 testpmd> port (port_id) (rxq|txq) (queue_id) (start|stop) 1771 1772port config - queue deferred start 1773~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1774 1775Switch on/off deferred start of a specific port queue:: 1776 1777 testpmd> port (port_id) (rxq|txq) (queue_id) deferred_start (on|off) 1778 1779port setup queue 1780~~~~~~~~~~~~~~~~~~~~~ 1781 1782Setup a rx/tx queue on a specific port:: 1783 1784 testpmd> port (port_id) (rxq|txq) (queue_id) setup 1785 1786Only take effect when port is started. 1787 1788port config - speed 1789~~~~~~~~~~~~~~~~~~~ 1790 1791Set the speed and duplex mode for all ports or a specific port:: 1792 1793 testpmd> port config (port_id|all) speed (10|100|1000|10000|25000|40000|50000|100000|auto) \ 1794 duplex (half|full|auto) 1795 1796port config - queues/descriptors 1797~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1798 1799Set number of queues/descriptors for rxq, txq, rxd and txd:: 1800 1801 testpmd> port config all (rxq|txq|rxd|txd) (value) 1802 1803This is equivalent to the ``--rxq``, ``--txq``, ``--rxd`` and ``--txd`` command-line options. 1804 1805port config - max-pkt-len 1806~~~~~~~~~~~~~~~~~~~~~~~~~ 1807 1808Set the maximum packet length:: 1809 1810 testpmd> port config all max-pkt-len (value) 1811 1812This is equivalent to the ``--max-pkt-len`` command-line option. 1813 1814port config - CRC Strip 1815~~~~~~~~~~~~~~~~~~~~~~~ 1816 1817Set hardware CRC stripping on or off for all ports:: 1818 1819 testpmd> port config all crc-strip (on|off) 1820 1821CRC stripping is on by default. 1822 1823The ``off`` option is equivalent to the ``--disable-crc-strip`` command-line option. 1824 1825port config - scatter 1826~~~~~~~~~~~~~~~~~~~~~~~ 1827 1828Set RX scatter mode on or off for all ports:: 1829 1830 testpmd> port config all scatter (on|off) 1831 1832RX scatter mode is off by default. 1833 1834The ``on`` option is equivalent to the ``--enable-scatter`` command-line option. 1835 1836port config - RX Checksum 1837~~~~~~~~~~~~~~~~~~~~~~~~~ 1838 1839Set hardware RX checksum offload to on or off for all ports:: 1840 1841 testpmd> port config all rx-cksum (on|off) 1842 1843Checksum offload is off by default. 1844 1845The ``on`` option is equivalent to the ``--enable-rx-cksum`` command-line option. 1846 1847port config - VLAN 1848~~~~~~~~~~~~~~~~~~ 1849 1850Set hardware VLAN on or off for all ports:: 1851 1852 testpmd> port config all hw-vlan (on|off) 1853 1854Hardware VLAN is off by default. 1855 1856The ``on`` option is equivalent to the ``--enable-hw-vlan`` command-line option. 1857 1858port config - VLAN filter 1859~~~~~~~~~~~~~~~~~~~~~~~~~ 1860 1861Set hardware VLAN filter on or off for all ports:: 1862 1863 testpmd> port config all hw-vlan-filter (on|off) 1864 1865Hardware VLAN filter is off by default. 1866 1867The ``on`` option is equivalent to the ``--enable-hw-vlan-filter`` command-line option. 1868 1869port config - VLAN strip 1870~~~~~~~~~~~~~~~~~~~~~~~~ 1871 1872Set hardware VLAN strip on or off for all ports:: 1873 1874 testpmd> port config all hw-vlan-strip (on|off) 1875 1876Hardware VLAN strip is off by default. 1877 1878The ``on`` option is equivalent to the ``--enable-hw-vlan-strip`` command-line option. 1879 1880port config - VLAN extend 1881~~~~~~~~~~~~~~~~~~~~~~~~~ 1882 1883Set hardware VLAN extend on or off for all ports:: 1884 1885 testpmd> port config all hw-vlan-extend (on|off) 1886 1887Hardware VLAN extend is off by default. 1888 1889The ``on`` option is equivalent to the ``--enable-hw-vlan-extend`` command-line option. 1890 1891port config - Drop Packets 1892~~~~~~~~~~~~~~~~~~~~~~~~~~ 1893 1894Set packet drop for packets with no descriptors on or off for all ports:: 1895 1896 testpmd> port config all drop-en (on|off) 1897 1898Packet dropping for packets with no descriptors is off by default. 1899 1900The ``on`` option is equivalent to the ``--enable-drop-en`` command-line option. 1901 1902port config - RSS 1903~~~~~~~~~~~~~~~~~ 1904 1905Set the RSS (Receive Side Scaling) mode on or off:: 1906 1907 testpmd> port config all rss (all|default|ip|tcp|udp|sctp|ether|port|vxlan|geneve|nvgre|none) 1908 1909RSS is on by default. 1910 1911The ``all`` option is equivalent to ip|tcp|udp|sctp|ether. 1912The ``default`` option enables all supported RSS types reported by device info. 1913The ``none`` option is equivalent to the ``--disable-rss`` command-line option. 1914 1915port config - RSS Reta 1916~~~~~~~~~~~~~~~~~~~~~~ 1917 1918Set the RSS (Receive Side Scaling) redirection table:: 1919 1920 testpmd> port config all rss reta (hash,queue)[,(hash,queue)] 1921 1922port config - DCB 1923~~~~~~~~~~~~~~~~~ 1924 1925Set the DCB mode for an individual port:: 1926 1927 testpmd> port config (port_id) dcb vt (on|off) (traffic_class) pfc (on|off) 1928 1929The traffic class should be 4 or 8. 1930 1931port config - Burst 1932~~~~~~~~~~~~~~~~~~~ 1933 1934Set the number of packets per burst:: 1935 1936 testpmd> port config all burst (value) 1937 1938This is equivalent to the ``--burst`` command-line option. 1939 1940port config - Threshold 1941~~~~~~~~~~~~~~~~~~~~~~~ 1942 1943Set thresholds for TX/RX queues:: 1944 1945 testpmd> port config all (threshold) (value) 1946 1947Where the threshold type can be: 1948 1949* ``txpt:`` Set the prefetch threshold register of the TX rings, 0 <= value <= 255. 1950 1951* ``txht:`` Set the host threshold register of the TX rings, 0 <= value <= 255. 1952 1953* ``txwt:`` Set the write-back threshold register of the TX rings, 0 <= value <= 255. 1954 1955* ``rxpt:`` Set the prefetch threshold register of the RX rings, 0 <= value <= 255. 1956 1957* ``rxht:`` Set the host threshold register of the RX rings, 0 <= value <= 255. 1958 1959* ``rxwt:`` Set the write-back threshold register of the RX rings, 0 <= value <= 255. 1960 1961* ``txfreet:`` Set the transmit free threshold of the TX rings, 0 <= value <= txd. 1962 1963* ``rxfreet:`` Set the transmit free threshold of the RX rings, 0 <= value <= rxd. 1964 1965* ``txrst:`` Set the transmit RS bit threshold of TX rings, 0 <= value <= txd. 1966 1967These threshold options are also available from the command-line. 1968 1969port config - E-tag 1970~~~~~~~~~~~~~~~~~~~ 1971 1972Set the value of ether-type for E-tag:: 1973 1974 testpmd> port config (port_id|all) l2-tunnel E-tag ether-type (value) 1975 1976Enable/disable the E-tag support:: 1977 1978 testpmd> port config (port_id|all) l2-tunnel E-tag (enable|disable) 1979 1980port config pctype mapping 1981~~~~~~~~~~~~~~~~~~~~~~~~~~ 1982 1983Reset pctype mapping table:: 1984 1985 testpmd> port config (port_id) pctype mapping reset 1986 1987Update hardware defined pctype to software defined flow type mapping table:: 1988 1989 testpmd> port config (port_id) pctype mapping update (pctype_id_0[,pctype_id_1]*) (flow_type_id) 1990 1991where: 1992 1993* ``pctype_id_x``: hardware pctype id as index of bit in bitmask value of the pctype mapping table. 1994 1995* ``flow_type_id``: software flow type id as the index of the pctype mapping table. 1996 1997port config input set 1998~~~~~~~~~~~~~~~~~~~~~ 1999 2000Config RSS/FDIR/FDIR flexible payload input set for some pctype:: 2001 testpmd> port config (port_id) pctype (pctype_id) \ 2002 (hash_inset|fdir_inset|fdir_flx_inset) \ 2003 (get|set|clear) field (field_idx) 2004 2005Clear RSS/FDIR/FDIR flexible payload input set for some pctype:: 2006 testpmd> port config (port_id) pctype (pctype_id) \ 2007 (hash_inset|fdir_inset|fdir_flx_inset) clear all 2008 2009where: 2010 2011* ``pctype_id``: hardware packet classification types. 2012* ``field_idx``: hardware field index. 2013 2014port config udp_tunnel_port 2015~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2016 2017Add/remove UDP tunnel port for VXLAN/GENEVE tunneling protocols:: 2018 testpmd> port config (port_id) udp_tunnel_port add|rm vxlan|geneve (udp_port) 2019 2020Link Bonding Functions 2021---------------------- 2022 2023The Link Bonding functions make it possible to dynamically create and 2024manage link bonding devices from within testpmd interactive prompt. 2025 2026create bonded device 2027~~~~~~~~~~~~~~~~~~~~ 2028 2029Create a new bonding device:: 2030 2031 testpmd> create bonded device (mode) (socket) 2032 2033For example, to create a bonded device in mode 1 on socket 0:: 2034 2035 testpmd> create bonded device 1 0 2036 created new bonded device (port X) 2037 2038add bonding slave 2039~~~~~~~~~~~~~~~~~ 2040 2041Adds Ethernet device to a Link Bonding device:: 2042 2043 testpmd> add bonding slave (slave id) (port id) 2044 2045For example, to add Ethernet device (port 6) to a Link Bonding device (port 10):: 2046 2047 testpmd> add bonding slave 6 10 2048 2049 2050remove bonding slave 2051~~~~~~~~~~~~~~~~~~~~ 2052 2053Removes an Ethernet slave device from a Link Bonding device:: 2054 2055 testpmd> remove bonding slave (slave id) (port id) 2056 2057For example, to remove Ethernet slave device (port 6) to a Link Bonding device (port 10):: 2058 2059 testpmd> remove bonding slave 6 10 2060 2061set bonding mode 2062~~~~~~~~~~~~~~~~ 2063 2064Set the Link Bonding mode of a Link Bonding device:: 2065 2066 testpmd> set bonding mode (value) (port id) 2067 2068For example, to set the bonding mode of a Link Bonding device (port 10) to broadcast (mode 3):: 2069 2070 testpmd> set bonding mode 3 10 2071 2072set bonding primary 2073~~~~~~~~~~~~~~~~~~~ 2074 2075Set an Ethernet slave device as the primary device on a Link Bonding device:: 2076 2077 testpmd> set bonding primary (slave id) (port id) 2078 2079For example, to set the Ethernet slave device (port 6) as the primary port of a Link Bonding device (port 10):: 2080 2081 testpmd> set bonding primary 6 10 2082 2083set bonding mac 2084~~~~~~~~~~~~~~~ 2085 2086Set the MAC address of a Link Bonding device:: 2087 2088 testpmd> set bonding mac (port id) (mac) 2089 2090For example, to set the MAC address of a Link Bonding device (port 10) to 00:00:00:00:00:01:: 2091 2092 testpmd> set bonding mac 10 00:00:00:00:00:01 2093 2094set bonding xmit_balance_policy 2095~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2096 2097Set the transmission policy for a Link Bonding device when it is in Balance XOR mode:: 2098 2099 testpmd> set bonding xmit_balance_policy (port_id) (l2|l23|l34) 2100 2101For example, set a Link Bonding device (port 10) to use a balance policy of layer 3+4 (IP addresses & UDP ports):: 2102 2103 testpmd> set bonding xmit_balance_policy 10 l34 2104 2105 2106set bonding mon_period 2107~~~~~~~~~~~~~~~~~~~~~~ 2108 2109Set the link status monitoring polling period in milliseconds for a bonding device. 2110 2111This adds support for PMD slave devices which do not support link status interrupts. 2112When the mon_period is set to a value greater than 0 then all PMD's which do not support 2113link status ISR will be queried every polling interval to check if their link status has changed:: 2114 2115 testpmd> set bonding mon_period (port_id) (value) 2116 2117For example, to set the link status monitoring polling period of bonded device (port 5) to 150ms:: 2118 2119 testpmd> set bonding mon_period 5 150 2120 2121 2122set bonding lacp dedicated_queue 2123~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2124 2125Enable dedicated tx/rx queues on bonding devices slaves to handle LACP control plane traffic 2126when in mode 4 (link-aggregration-802.3ad):: 2127 2128 testpmd> set bonding lacp dedicated_queues (port_id) (enable|disable) 2129 2130 2131set bonding agg_mode 2132~~~~~~~~~~~~~~~~~~~~ 2133 2134Enable one of the specific aggregators mode when in mode 4 (link-aggregration-802.3ad):: 2135 2136 testpmd> set bonding agg_mode (port_id) (bandwidth|count|stable) 2137 2138 2139show bonding config 2140~~~~~~~~~~~~~~~~~~~ 2141 2142Show the current configuration of a Link Bonding device:: 2143 2144 testpmd> show bonding config (port id) 2145 2146For example, 2147to show the configuration a Link Bonding device (port 9) with 3 slave devices (1, 3, 4) 2148in balance mode with a transmission policy of layer 2+3:: 2149 2150 testpmd> show bonding config 9 2151 Bonding mode: 2 2152 Balance Xmit Policy: BALANCE_XMIT_POLICY_LAYER23 2153 Slaves (3): [1 3 4] 2154 Active Slaves (3): [1 3 4] 2155 Primary: [3] 2156 2157 2158Register Functions 2159------------------ 2160 2161The Register Functions can be used to read from and write to registers on the network card referenced by a port number. 2162This is mainly useful for debugging purposes. 2163Reference should be made to the appropriate datasheet for the network card for details on the register addresses 2164and fields that can be accessed. 2165 2166read reg 2167~~~~~~~~ 2168 2169Display the value of a port register:: 2170 2171 testpmd> read reg (port_id) (address) 2172 2173For example, to examine the Flow Director control register (FDIRCTL, 0x0000EE000) on an Intel 82599 10 GbE Controller:: 2174 2175 testpmd> read reg 0 0xEE00 2176 port 0 PCI register at offset 0xEE00: 0x4A060029 (1241907241) 2177 2178read regfield 2179~~~~~~~~~~~~~ 2180 2181Display a port register bit field:: 2182 2183 testpmd> read regfield (port_id) (address) (bit_x) (bit_y) 2184 2185For example, reading the lowest two bits from the register in the example above:: 2186 2187 testpmd> read regfield 0 0xEE00 0 1 2188 port 0 PCI register at offset 0xEE00: bits[0, 1]=0x1 (1) 2189 2190read regbit 2191~~~~~~~~~~~ 2192 2193Display a single port register bit:: 2194 2195 testpmd> read regbit (port_id) (address) (bit_x) 2196 2197For example, reading the lowest bit from the register in the example above:: 2198 2199 testpmd> read regbit 0 0xEE00 0 2200 port 0 PCI register at offset 0xEE00: bit 0=1 2201 2202write reg 2203~~~~~~~~~ 2204 2205Set the value of a port register:: 2206 2207 testpmd> write reg (port_id) (address) (value) 2208 2209For example, to clear a register:: 2210 2211 testpmd> write reg 0 0xEE00 0x0 2212 port 0 PCI register at offset 0xEE00: 0x00000000 (0) 2213 2214write regfield 2215~~~~~~~~~~~~~~ 2216 2217Set bit field of a port register:: 2218 2219 testpmd> write regfield (port_id) (address) (bit_x) (bit_y) (value) 2220 2221For example, writing to the register cleared in the example above:: 2222 2223 testpmd> write regfield 0 0xEE00 0 1 2 2224 port 0 PCI register at offset 0xEE00: 0x00000002 (2) 2225 2226write regbit 2227~~~~~~~~~~~~ 2228 2229Set single bit value of a port register:: 2230 2231 testpmd> write regbit (port_id) (address) (bit_x) (value) 2232 2233For example, to set the high bit in the register from the example above:: 2234 2235 testpmd> write regbit 0 0xEE00 31 1 2236 port 0 PCI register at offset 0xEE00: 0x8000000A (2147483658) 2237 2238Traffic Metering and Policing 2239----------------------------- 2240 2241The following section shows functions for configuring traffic metering and 2242policing on the ethernet device through the use of generic ethdev API. 2243 2244show port traffic management capability 2245~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2246 2247Show traffic metering and policing capability of the port:: 2248 2249 testpmd> show port meter cap (port_id) 2250 2251add port meter profile (srTCM rfc2967) 2252~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2253 2254Add meter profile (srTCM rfc2697) to the ethernet device:: 2255 2256 testpmd> add port meter profile srtcm_rfc2697 (port_id) (profile_id) \ 2257 (cir) (cbs) (ebs) 2258 2259where: 2260 2261* ``profile_id``: ID for the meter profile. 2262* ``cir``: Committed Information Rate (CIR) (bytes/second). 2263* ``cbs``: Committed Burst Size (CBS) (bytes). 2264* ``ebs``: Excess Burst Size (EBS) (bytes). 2265 2266add port meter profile (trTCM rfc2968) 2267~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2268 2269Add meter profile (srTCM rfc2698) to the ethernet device:: 2270 2271 testpmd> add port meter profile trtcm_rfc2698 (port_id) (profile_id) \ 2272 (cir) (pir) (cbs) (pbs) 2273 2274where: 2275 2276* ``profile_id``: ID for the meter profile. 2277* ``cir``: Committed information rate (bytes/second). 2278* ``pir``: Peak information rate (bytes/second). 2279* ``cbs``: Committed burst size (bytes). 2280* ``pbs``: Peak burst size (bytes). 2281 2282add port meter profile (trTCM rfc4115) 2283~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2284 2285Add meter profile (trTCM rfc4115) to the ethernet device:: 2286 2287 testpmd> add port meter profile trtcm_rfc4115 (port_id) (profile_id) \ 2288 (cir) (eir) (cbs) (ebs) 2289 2290where: 2291 2292* ``profile_id``: ID for the meter profile. 2293* ``cir``: Committed information rate (bytes/second). 2294* ``eir``: Excess information rate (bytes/second). 2295* ``cbs``: Committed burst size (bytes). 2296* ``ebs``: Excess burst size (bytes). 2297 2298delete port meter profile 2299~~~~~~~~~~~~~~~~~~~~~~~~~ 2300 2301Delete meter profile from the ethernet device:: 2302 2303 testpmd> del port meter profile (port_id) (profile_id) 2304 2305create port meter 2306~~~~~~~~~~~~~~~~~ 2307 2308Create new meter object for the ethernet device:: 2309 2310 testpmd> create port meter (port_id) (mtr_id) (profile_id) \ 2311 (meter_enable) (g_action) (y_action) (r_action) (stats_mask) (shared) \ 2312 (use_pre_meter_color) [(dscp_tbl_entry0) (dscp_tbl_entry1)...\ 2313 (dscp_tbl_entry63)] 2314 2315where: 2316 2317* ``mtr_id``: meter object ID. 2318* ``profile_id``: ID for the meter profile. 2319* ``meter_enable``: When this parameter has a non-zero value, the meter object 2320 gets enabled at the time of creation, otherwise remains disabled. 2321* ``g_action``: Policer action for the packet with green color. 2322* ``y_action``: Policer action for the packet with yellow color. 2323* ``r_action``: Policer action for the packet with red color. 2324* ``stats_mask``: Mask of statistics counter types to be enabled for the 2325 meter object. 2326* ``shared``: When this parameter has a non-zero value, the meter object is 2327 shared by multiple flows. Otherwise, meter object is used by single flow. 2328* ``use_pre_meter_color``: When this parameter has a non-zero value, the 2329 input color for the current meter object is determined by the latest meter 2330 object in the same flow. Otherwise, the current meter object uses the 2331 *dscp_table* to determine the input color. 2332* ``dscp_tbl_entryx``: DSCP table entry x providing meter providing input 2333 color, 0 <= x <= 63. 2334 2335enable port meter 2336~~~~~~~~~~~~~~~~~ 2337 2338Enable meter for the ethernet device:: 2339 2340 testpmd> enable port meter (port_id) (mtr_id) 2341 2342disable port meter 2343~~~~~~~~~~~~~~~~~~ 2344 2345Disable meter for the ethernet device:: 2346 2347 testpmd> disable port meter (port_id) (mtr_id) 2348 2349delete port meter 2350~~~~~~~~~~~~~~~~~ 2351 2352Delete meter for the ethernet device:: 2353 2354 testpmd> del port meter (port_id) (mtr_id) 2355 2356Set port meter profile 2357~~~~~~~~~~~~~~~~~~~~~~ 2358 2359Set meter profile for the ethernet device:: 2360 2361 testpmd> set port meter profile (port_id) (mtr_id) (profile_id) 2362 2363set port meter dscp table 2364~~~~~~~~~~~~~~~~~~~~~~~~~ 2365 2366Set meter dscp table for the ethernet device:: 2367 2368 testpmd> set port meter dscp table (port_id) (mtr_id) [(dscp_tbl_entry0) \ 2369 (dscp_tbl_entry1)...(dscp_tbl_entry63)] 2370 2371set port meter policer action 2372~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2373 2374Set meter policer action for the ethernet device:: 2375 2376 testpmd> set port meter policer action (port_id) (mtr_id) (action_mask) \ 2377 (action0) [(action1) (action1)] 2378 2379where: 2380 2381* ``action_mask``: Bit mask indicating which policer actions need to be 2382 updated. One or more policer actions can be updated in a single function 2383 invocation. To update the policer action associated with color C, bit 2384 (1 << C) needs to be set in *action_mask* and element at position C 2385 in the *actions* array needs to be valid. 2386* ``actionx``: Policer action for the color x, 2387 RTE_MTR_GREEN <= x < RTE_MTR_COLORS 2388 2389set port meter stats mask 2390~~~~~~~~~~~~~~~~~~~~~~~~~ 2391 2392Set meter stats mask for the ethernet device:: 2393 2394 testpmd> set port meter stats mask (port_id) (mtr_id) (stats_mask) 2395 2396where: 2397 2398* ``stats_mask``: Bit mask indicating statistics counter types to be enabled. 2399 2400show port meter stats 2401~~~~~~~~~~~~~~~~~~~~~ 2402 2403Show meter stats of the ethernet device:: 2404 2405 testpmd> show port meter stats (port_id) (mtr_id) (clear) 2406 2407where: 2408 2409* ``clear``: Flag that indicates whether the statistics counters should 2410 be cleared (i.e. set to zero) immediately after they have been read or not. 2411 2412Traffic Management 2413------------------ 2414 2415The following section shows functions for configuring traffic management on 2416on the ethernet device through the use of generic TM API. 2417 2418show port traffic management capability 2419~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2420 2421Show traffic management capability of the port:: 2422 2423 testpmd> show port tm cap (port_id) 2424 2425show port traffic management capability (hierarchy level) 2426~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2427 2428Show traffic management hierarchy level capability of the port:: 2429 2430 testpmd> show port tm level cap (port_id) (level_id) 2431 2432show port traffic management capability (hierarchy node level) 2433~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2434 2435Show the traffic management hierarchy node capability of the port:: 2436 2437 testpmd> show port tm node cap (port_id) (node_id) 2438 2439show port traffic management hierarchy node type 2440~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2441 2442Show the port traffic management hierarchy node type:: 2443 2444 testpmd> show port tm node type (port_id) (node_id) 2445 2446show port traffic management hierarchy node stats 2447~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2448 2449Show the port traffic management hierarchy node statistics:: 2450 2451 testpmd> show port tm node stats (port_id) (node_id) (clear) 2452 2453where: 2454 2455* ``clear``: When this parameter has a non-zero value, the statistics counters 2456 are cleared (i.e. set to zero) immediately after they have been read, 2457 otherwise the statistics counters are left untouched. 2458 2459Add port traffic management private shaper profile 2460~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2461 2462Add the port traffic management private shaper profile:: 2463 2464 testpmd> add port tm node shaper profile (port_id) (shaper_profile_id) \ 2465 (tb_rate) (tb_size) (packet_length_adjust) 2466 2467where: 2468 2469* ``shaper_profile id``: Shaper profile ID for the new profile. 2470* ``tb_rate``: Token bucket rate (bytes per second). 2471* ``tb_size``: Token bucket size (bytes). 2472* ``packet_length_adjust``: The value (bytes) to be added to the length of 2473 each packet for the purpose of shaping. This parameter value can be used to 2474 correct the packet length with the framing overhead bytes that are consumed 2475 on the wire. 2476 2477Delete port traffic management private shaper profile 2478~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2479 2480Delete the port traffic management private shaper:: 2481 2482 testpmd> del port tm node shaper profile (port_id) (shaper_profile_id) 2483 2484where: 2485 2486* ``shaper_profile id``: Shaper profile ID that needs to be deleted. 2487 2488Add port traffic management shared shaper 2489~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2490 2491Create the port traffic management shared shaper:: 2492 2493 testpmd> add port tm node shared shaper (port_id) (shared_shaper_id) \ 2494 (shaper_profile_id) 2495 2496where: 2497 2498* ``shared_shaper_id``: Shared shaper ID to be created. 2499* ``shaper_profile id``: Shaper profile ID for shared shaper. 2500 2501Set port traffic management shared shaper 2502~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2503 2504Update the port traffic management shared shaper:: 2505 2506 testpmd> set port tm node shared shaper (port_id) (shared_shaper_id) \ 2507 (shaper_profile_id) 2508 2509where: 2510 2511* ``shared_shaper_id``: Shared shaper ID to be update. 2512* ``shaper_profile id``: Shaper profile ID for shared shaper. 2513 2514Delete port traffic management shared shaper 2515~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2516 2517Delete the port traffic management shared shaper:: 2518 2519 testpmd> del port tm node shared shaper (port_id) (shared_shaper_id) 2520 2521where: 2522 2523* ``shared_shaper_id``: Shared shaper ID to be deleted. 2524 2525Set port traffic management hiearchy node private shaper 2526~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2527 2528set the port traffic management hierarchy node private shaper:: 2529 2530 testpmd> set port tm node shaper profile (port_id) (node_id) \ 2531 (shaper_profile_id) 2532 2533where: 2534 2535* ``shaper_profile id``: Private shaper profile ID to be enabled on the 2536 hierarchy node. 2537 2538Add port traffic management WRED profile 2539~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2540 2541Create a new WRED profile:: 2542 2543 testpmd> add port tm node wred profile (port_id) (wred_profile_id) \ 2544 (color_g) (min_th_g) (max_th_g) (maxp_inv_g) (wq_log2_g) \ 2545 (color_y) (min_th_y) (max_th_y) (maxp_inv_y) (wq_log2_y) \ 2546 (color_r) (min_th_r) (max_th_r) (maxp_inv_r) (wq_log2_r) 2547 2548where: 2549 2550* ``wred_profile id``: Identifier for the newly create WRED profile 2551* ``color_g``: Packet color (green) 2552* ``min_th_g``: Minimum queue threshold for packet with green color 2553* ``max_th_g``: Minimum queue threshold for packet with green color 2554* ``maxp_inv_g``: Inverse of packet marking probability maximum value (maxp) 2555* ``wq_log2_g``: Negated log2 of queue weight (wq) 2556* ``color_y``: Packet color (yellow) 2557* ``min_th_y``: Minimum queue threshold for packet with yellow color 2558* ``max_th_y``: Minimum queue threshold for packet with yellow color 2559* ``maxp_inv_y``: Inverse of packet marking probability maximum value (maxp) 2560* ``wq_log2_y``: Negated log2 of queue weight (wq) 2561* ``color_r``: Packet color (red) 2562* ``min_th_r``: Minimum queue threshold for packet with yellow color 2563* ``max_th_r``: Minimum queue threshold for packet with yellow color 2564* ``maxp_inv_r``: Inverse of packet marking probability maximum value (maxp) 2565* ``wq_log2_r``: Negated log2 of queue weight (wq) 2566 2567Delete port traffic management WRED profile 2568~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2569 2570Delete the WRED profile:: 2571 2572 testpmd> del port tm node wred profile (port_id) (wred_profile_id) 2573 2574Add port traffic management hierarchy nonleaf node 2575~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2576 2577Add nonleaf node to port traffic management hiearchy:: 2578 2579 testpmd> add port tm nonleaf node (port_id) (node_id) (parent_node_id) \ 2580 (priority) (weight) (level_id) (shaper_profile_id) \ 2581 (n_sp_priorities) (stats_mask) (n_shared_shapers) \ 2582 [(shared_shaper_0) (shared_shaper_1) ...] \ 2583 2584where: 2585 2586* ``parent_node_id``: Node ID of the parent. 2587* ``priority``: Node priority (highest node priority is zero). This is used by 2588 the SP algorithm running on the parent node for scheduling this node. 2589* ``weight``: Node weight (lowest weight is one). The node weight is relative 2590 to the weight sum of all siblings that have the same priority. It is used by 2591 the WFQ algorithm running on the parent node for scheduling this node. 2592* ``level_id``: Hiearchy level of the node. 2593* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by 2594 the node. 2595* ``n_sp_priorities``: Number of strict priorities. 2596* ``stats_mask``: Mask of statistics counter types to be enabled for this node. 2597* ``n_shared_shapers``: Number of shared shapers. 2598* ``shared_shaper_id``: Shared shaper id. 2599 2600Add port traffic management hierarchy leaf node 2601~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2602 2603Add leaf node to port traffic management hiearchy:: 2604 2605 testpmd> add port tm leaf node (port_id) (node_id) (parent_node_id) \ 2606 (priority) (weight) (level_id) (shaper_profile_id) \ 2607 (cman_mode) (wred_profile_id) (stats_mask) (n_shared_shapers) \ 2608 [(shared_shaper_id) (shared_shaper_id) ...] \ 2609 2610where: 2611 2612* ``parent_node_id``: Node ID of the parent. 2613* ``priority``: Node priority (highest node priority is zero). This is used by 2614 the SP algorithm running on the parent node for scheduling this node. 2615* ``weight``: Node weight (lowest weight is one). The node weight is relative 2616 to the weight sum of all siblings that have the same priority. It is used by 2617 the WFQ algorithm running on the parent node for scheduling this node. 2618* ``level_id``: Hiearchy level of the node. 2619* ``shaper_profile_id``: Shaper profile ID of the private shaper to be used by 2620 the node. 2621* ``cman_mode``: Congestion management mode to be enabled for this node. 2622* ``wred_profile_id``: WRED profile id to be enabled for this node. 2623* ``stats_mask``: Mask of statistics counter types to be enabled for this node. 2624* ``n_shared_shapers``: Number of shared shapers. 2625* ``shared_shaper_id``: Shared shaper id. 2626 2627Delete port traffic management hierarchy node 2628~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2629 2630Delete node from port traffic management hiearchy:: 2631 2632 testpmd> del port tm node (port_id) (node_id) 2633 2634Update port traffic management hierarchy parent node 2635~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2636 2637Update port traffic management hierarchy parent node:: 2638 2639 testpmd> set port tm node parent (port_id) (node_id) (parent_node_id) \ 2640 (priority) (weight) 2641 2642This function can only be called after the hierarchy commit invocation. Its 2643success depends on the port support for this operation, as advertised through 2644the port capability set. This function is valid for all nodes of the traffic 2645management hierarchy except root node. 2646 2647Suspend port traffic management hierarchy node 2648~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2649 2650 testpmd> suspend port tm node (port_id) (node_id) 2651 2652Resume port traffic management hierarchy node 2653~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2654 2655 testpmd> resume port tm node (port_id) (node_id) 2656 2657Commit port traffic management hierarchy 2658~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2659 2660Commit the traffic management hierarchy on the port:: 2661 2662 testpmd> port tm hierarchy commit (port_id) (clean_on_fail) 2663 2664where: 2665 2666* ``clean_on_fail``: When set to non-zero, hierarchy is cleared on function 2667 call failure. On the other hand, hierarchy is preserved when this parameter 2668 is equal to zero. 2669 2670Set port traffic management default hierarchy (softnic forwarding mode) 2671~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 2672 2673set the traffic management default hierarchy on the port:: 2674 2675 testpmd> set port tm hierarchy default (port_id) 2676 2677Filter Functions 2678---------------- 2679 2680This section details the available filter functions that are available. 2681 2682Note these functions interface the deprecated legacy filtering framework, 2683superseded by *rte_flow*. See `Flow rules management`_. 2684 2685ethertype_filter 2686~~~~~~~~~~~~~~~~~~~~ 2687 2688Add or delete a L2 Ethertype filter, which identify packets by their L2 Ethertype mainly assign them to a receive queue:: 2689 2690 ethertype_filter (port_id) (add|del) (mac_addr|mac_ignr) (mac_address) \ 2691 ethertype (ether_type) (drop|fwd) queue (queue_id) 2692 2693The available information parameters are: 2694 2695* ``port_id``: The port which the Ethertype filter assigned on. 2696 2697* ``mac_addr``: Compare destination mac address. 2698 2699* ``mac_ignr``: Ignore destination mac address match. 2700 2701* ``mac_address``: Destination mac address to match. 2702 2703* ``ether_type``: The EtherType value want to match, 2704 for example 0x0806 for ARP packet. 0x0800 (IPv4) and 0x86DD (IPv6) are invalid. 2705 2706* ``queue_id``: The receive queue associated with this EtherType filter. 2707 It is meaningless when deleting or dropping. 2708 2709Example, to add/remove an ethertype filter rule:: 2710 2711 testpmd> ethertype_filter 0 add mac_ignr 00:11:22:33:44:55 \ 2712 ethertype 0x0806 fwd queue 3 2713 2714 testpmd> ethertype_filter 0 del mac_ignr 00:11:22:33:44:55 \ 2715 ethertype 0x0806 fwd queue 3 2716 27172tuple_filter 2718~~~~~~~~~~~~~~~~~ 2719 2720Add or delete a 2-tuple filter, 2721which identifies packets by specific protocol and destination TCP/UDP port 2722and forwards packets into one of the receive queues:: 2723 2724 2tuple_filter (port_id) (add|del) dst_port (dst_port_value) \ 2725 protocol (protocol_value) mask (mask_value) \ 2726 tcp_flags (tcp_flags_value) priority (prio_value) \ 2727 queue (queue_id) 2728 2729The available information parameters are: 2730 2731* ``port_id``: The port which the 2-tuple filter assigned on. 2732 2733* ``dst_port_value``: Destination port in L4. 2734 2735* ``protocol_value``: IP L4 protocol. 2736 2737* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate. 2738 2739* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the pro_value is not set to 0x06 (TCP). 2740 2741* ``prio_value``: Priority of this filter. 2742 2743* ``queue_id``: The receive queue associated with this 2-tuple filter. 2744 2745Example, to add/remove an 2tuple filter rule:: 2746 2747 testpmd> 2tuple_filter 0 add dst_port 32 protocol 0x06 mask 0x03 \ 2748 tcp_flags 0x02 priority 3 queue 3 2749 2750 testpmd> 2tuple_filter 0 del dst_port 32 protocol 0x06 mask 0x03 \ 2751 tcp_flags 0x02 priority 3 queue 3 2752 27535tuple_filter 2754~~~~~~~~~~~~~~~~~ 2755 2756Add or delete a 5-tuple filter, 2757which consists of a 5-tuple (protocol, source and destination IP addresses, source and destination TCP/UDP/SCTP port) 2758and routes packets into one of the receive queues:: 2759 2760 5tuple_filter (port_id) (add|del) dst_ip (dst_address) src_ip \ 2761 (src_address) dst_port (dst_port_value) \ 2762 src_port (src_port_value) protocol (protocol_value) \ 2763 mask (mask_value) tcp_flags (tcp_flags_value) \ 2764 priority (prio_value) queue (queue_id) 2765 2766The available information parameters are: 2767 2768* ``port_id``: The port which the 5-tuple filter assigned on. 2769 2770* ``dst_address``: Destination IP address. 2771 2772* ``src_address``: Source IP address. 2773 2774* ``dst_port_value``: TCP/UDP destination port. 2775 2776* ``src_port_value``: TCP/UDP source port. 2777 2778* ``protocol_value``: L4 protocol. 2779 2780* ``mask_value``: Participates in the match or not by bit for field above, 1b means participate 2781 2782* ``tcp_flags_value``: TCP control bits. The non-zero value is invalid, when the protocol_value is not set to 0x06 (TCP). 2783 2784* ``prio_value``: The priority of this filter. 2785 2786* ``queue_id``: The receive queue associated with this 5-tuple filter. 2787 2788Example, to add/remove an 5tuple filter rule:: 2789 2790 testpmd> 5tuple_filter 0 add 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 2794 testpmd> 5tuple_filter 0 del dst_ip 2.2.2.5 src_ip 2.2.2.4 \ 2795 dst_port 64 src_port 32 protocol 0x06 mask 0x1F \ 2796 flags 0x0 priority 3 queue 3 2797 2798syn_filter 2799~~~~~~~~~~ 2800 2801Using the SYN filter, TCP packets whose *SYN* flag is set can be forwarded to a separate queue:: 2802 2803 syn_filter (port_id) (add|del) priority (high|low) queue (queue_id) 2804 2805The available information parameters are: 2806 2807* ``port_id``: The port which the SYN filter assigned on. 2808 2809* ``high``: This SYN filter has higher priority than other filters. 2810 2811* ``low``: This SYN filter has lower priority than other filters. 2812 2813* ``queue_id``: The receive queue associated with this SYN filter 2814 2815Example:: 2816 2817 testpmd> syn_filter 0 add priority high queue 3 2818 2819flex_filter 2820~~~~~~~~~~~ 2821 2822With flex filter, packets can be recognized by any arbitrary pattern within the first 128 bytes of the packet 2823and routed into one of the receive queues:: 2824 2825 flex_filter (port_id) (add|del) len (len_value) bytes (bytes_value) \ 2826 mask (mask_value) priority (prio_value) queue (queue_id) 2827 2828The available information parameters are: 2829 2830* ``port_id``: The port which the Flex filter is assigned on. 2831 2832* ``len_value``: Filter length in bytes, no greater than 128. 2833 2834* ``bytes_value``: A string in hexadecimal, means the value the flex filter needs to match. 2835 2836* ``mask_value``: A string in hexadecimal, bit 1 means corresponding byte participates in the match. 2837 2838* ``prio_value``: The priority of this filter. 2839 2840* ``queue_id``: The receive queue associated with this Flex filter. 2841 2842Example:: 2843 2844 testpmd> flex_filter 0 add len 16 bytes 0x00000000000000000000000008060000 \ 2845 mask 000C priority 3 queue 3 2846 2847 testpmd> flex_filter 0 del len 16 bytes 0x00000000000000000000000008060000 \ 2848 mask 000C priority 3 queue 3 2849 2850 2851.. _testpmd_flow_director: 2852 2853flow_director_filter 2854~~~~~~~~~~~~~~~~~~~~ 2855 2856The Flow Director works in receive mode to identify specific flows or sets of flows and route them to specific queues. 2857 2858Four types of filtering are supported which are referred to as Perfect Match, Signature, Perfect-mac-vlan and 2859Perfect-tunnel filters, the match mode is set by the ``--pkt-filter-mode`` command-line parameter: 2860 2861* Perfect match filters. 2862 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2863 The masked fields are for IP flow. 2864 2865* Signature filters. 2866 The hardware checks a match between a hash-based signature of the masked fields of the received packet. 2867 2868* Perfect-mac-vlan 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 MAC VLAN flow. 2871 2872* Perfect-tunnel match filters. 2873 The hardware checks a match between the masked fields of the received packets and the programmed filters. 2874 The masked fields are for tunnel flow. 2875 2876* Perfect-raw-flow-type match filters. 2877 The hardware checks a match between the masked fields of the received packets and pre-loaded raw (template) packet. 2878 The masked fields are specified by input sets. 2879 2880The Flow Director filters can match the different fields for different type of packet: flow type, specific input set 2881per flow type and the flexible payload. 2882 2883The Flow Director can also mask out parts of all of these fields so that filters 2884are only applied to certain fields or parts of the fields. 2885 2886Note that for raw flow type mode the source and destination fields in the 2887raw packet buffer need to be presented in a reversed order with respect 2888to the expected received packets. 2889For example: IP source and destination addresses or TCP/UDP/SCTP 2890source and destination ports 2891 2892Different NICs may have different capabilities, command show port fdir (port_id) can be used to acquire the information. 2893 2894# Commands to add flow director filters of different flow types:: 2895 2896 flow_director_filter (port_id) mode IP (add|del|update) \ 2897 flow (ipv4-other|ipv4-frag|ipv6-other|ipv6-frag) \ 2898 src (src_ip_address) dst (dst_ip_address) \ 2899 tos (tos_value) proto (proto_value) ttl (ttl_value) \ 2900 vlan (vlan_value) flexbytes (flexbytes_value) \ 2901 (drop|fwd) pf|vf(vf_id) queue (queue_id) \ 2902 fd_id (fd_id_value) 2903 2904 flow_director_filter (port_id) mode IP (add|del|update) \ 2905 flow (ipv4-tcp|ipv4-udp|ipv6-tcp|ipv6-udp) \ 2906 src (src_ip_address) (src_port) \ 2907 dst (dst_ip_address) (dst_port) \ 2908 tos (tos_value) ttl (ttl_value) \ 2909 vlan (vlan_value) flexbytes (flexbytes_value) \ 2910 (drop|fwd) queue pf|vf(vf_id) (queue_id) \ 2911 fd_id (fd_id_value) 2912 2913 flow_director_filter (port_id) mode IP (add|del|update) \ 2914 flow (ipv4-sctp|ipv6-sctp) \ 2915 src (src_ip_address) (src_port) \ 2916 dst (dst_ip_address) (dst_port) \ 2917 tos (tos_value) ttl (ttl_value) \ 2918 tag (verification_tag) vlan (vlan_value) \ 2919 flexbytes (flexbytes_value) (drop|fwd) \ 2920 pf|vf(vf_id) queue (queue_id) fd_id (fd_id_value) 2921 2922 flow_director_filter (port_id) mode IP (add|del|update) flow l2_payload \ 2923 ether (ethertype) flexbytes (flexbytes_value) \ 2924 (drop|fwd) pf|vf(vf_id) queue (queue_id) 2925 fd_id (fd_id_value) 2926 2927 flow_director_filter (port_id) mode MAC-VLAN (add|del|update) \ 2928 mac (mac_address) vlan (vlan_value) \ 2929 flexbytes (flexbytes_value) (drop|fwd) \ 2930 queue (queue_id) fd_id (fd_id_value) 2931 2932 flow_director_filter (port_id) mode Tunnel (add|del|update) \ 2933 mac (mac_address) vlan (vlan_value) \ 2934 tunnel (NVGRE|VxLAN) tunnel-id (tunnel_id_value) \ 2935 flexbytes (flexbytes_value) (drop|fwd) \ 2936 queue (queue_id) fd_id (fd_id_value) 2937 2938 flow_director_filter (port_id) mode raw (add|del|update) flow (flow_id) \ 2939 (drop|fwd) queue (queue_id) fd_id (fd_id_value) \ 2940 packet (packet file name) 2941 2942For example, to add an ipv4-udp flow type filter:: 2943 2944 testpmd> flow_director_filter 0 mode IP add flow ipv4-udp src 2.2.2.3 32 \ 2945 dst 2.2.2.5 33 tos 2 ttl 40 vlan 0x1 flexbytes (0x88,0x48) \ 2946 fwd pf queue 1 fd_id 1 2947 2948For example, add an ipv4-other flow type filter:: 2949 2950 testpmd> flow_director_filter 0 mode IP add flow ipv4-other src 2.2.2.3 \ 2951 dst 2.2.2.5 tos 2 proto 20 ttl 40 vlan 0x1 \ 2952 flexbytes (0x88,0x48) fwd pf queue 1 fd_id 1 2953 2954flush_flow_director 2955~~~~~~~~~~~~~~~~~~~ 2956 2957Flush all flow director filters on a device:: 2958 2959 testpmd> flush_flow_director (port_id) 2960 2961Example, to flush all flow director filter on port 0:: 2962 2963 testpmd> flush_flow_director 0 2964 2965flow_director_mask 2966~~~~~~~~~~~~~~~~~~ 2967 2968Set flow director's input masks:: 2969 2970 flow_director_mask (port_id) mode IP vlan (vlan_value) \ 2971 src_mask (ipv4_src) (ipv6_src) (src_port) \ 2972 dst_mask (ipv4_dst) (ipv6_dst) (dst_port) 2973 2974 flow_director_mask (port_id) mode MAC-VLAN vlan (vlan_value) 2975 2976 flow_director_mask (port_id) mode Tunnel vlan (vlan_value) \ 2977 mac (mac_value) tunnel-type (tunnel_type_value) \ 2978 tunnel-id (tunnel_id_value) 2979 2980Example, to set flow director mask on port 0:: 2981 2982 testpmd> flow_director_mask 0 mode IP vlan 0xefff \ 2983 src_mask 255.255.255.255 \ 2984 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF \ 2985 dst_mask 255.255.255.255 \ 2986 FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF:FFFF 0xFFFF 2987 2988flow_director_flex_mask 2989~~~~~~~~~~~~~~~~~~~~~~~ 2990 2991set masks of flow director's flexible payload based on certain flow type:: 2992 2993 testpmd> flow_director_flex_mask (port_id) \ 2994 flow (none|ipv4-other|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 2995 ipv6-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp| \ 2996 l2_payload|all) (mask) 2997 2998Example, to set flow director's flex mask for all flow type on port 0:: 2999 3000 testpmd> flow_director_flex_mask 0 flow all \ 3001 (0xff,0xff,0,0,0,0,0,0,0,0,0,0,0,0,0,0) 3002 3003 3004flow_director_flex_payload 3005~~~~~~~~~~~~~~~~~~~~~~~~~~ 3006 3007Configure flexible payload selection:: 3008 3009 flow_director_flex_payload (port_id) (raw|l2|l3|l4) (config) 3010 3011For example, to select the first 16 bytes from the offset 4 (bytes) of packet's payload as flexible payload:: 3012 3013 testpmd> flow_director_flex_payload 0 l4 \ 3014 (4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19) 3015 3016get_sym_hash_ena_per_port 3017~~~~~~~~~~~~~~~~~~~~~~~~~ 3018 3019Get symmetric hash enable configuration per port:: 3020 3021 get_sym_hash_ena_per_port (port_id) 3022 3023For example, to get symmetric hash enable configuration of port 1:: 3024 3025 testpmd> get_sym_hash_ena_per_port 1 3026 3027set_sym_hash_ena_per_port 3028~~~~~~~~~~~~~~~~~~~~~~~~~ 3029 3030Set symmetric hash enable configuration per port to enable or disable:: 3031 3032 set_sym_hash_ena_per_port (port_id) (enable|disable) 3033 3034For example, to set symmetric hash enable configuration of port 1 to enable:: 3035 3036 testpmd> set_sym_hash_ena_per_port 1 enable 3037 3038get_hash_global_config 3039~~~~~~~~~~~~~~~~~~~~~~ 3040 3041Get the global configurations of hash filters:: 3042 3043 get_hash_global_config (port_id) 3044 3045For example, to get the global configurations of hash filters of port 1:: 3046 3047 testpmd> get_hash_global_config 1 3048 3049set_hash_global_config 3050~~~~~~~~~~~~~~~~~~~~~~ 3051 3052Set the global configurations of hash filters:: 3053 3054 set_hash_global_config (port_id) (toeplitz|simple_xor|default) \ 3055 (ipv4|ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp|ipv4-other|ipv6|ipv6-frag| \ 3056 ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other|l2_payload|<flow_id>) \ 3057 (enable|disable) 3058 3059For example, to enable simple_xor for flow type of ipv6 on port 2:: 3060 3061 testpmd> set_hash_global_config 2 simple_xor ipv6 enable 3062 3063set_hash_input_set 3064~~~~~~~~~~~~~~~~~~ 3065 3066Set the input set for hash:: 3067 3068 set_hash_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 3069 ipv4-other|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 3070 l2_payload|<flow_id>) (ovlan|ivlan|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \ 3071 ipv4-tos|ipv4-proto|ipv6-tc|ipv6-next-header|udp-src-port|udp-dst-port| \ 3072 tcp-src-port|tcp-dst-port|sctp-src-port|sctp-dst-port|sctp-veri-tag| \ 3073 udp-key|gre-key|fld-1st|fld-2nd|fld-3rd|fld-4th|fld-5th|fld-6th|fld-7th| \ 3074 fld-8th|none) (select|add) 3075 3076For example, to add source IP to hash input set for flow type of ipv4-udp on port 0:: 3077 3078 testpmd> set_hash_input_set 0 ipv4-udp src-ipv4 add 3079 3080set_fdir_input_set 3081~~~~~~~~~~~~~~~~~~ 3082 3083The Flow Director filters can match the different fields for different type of packet, i.e. specific input set 3084on per flow type and the flexible payload. This command can be used to change input set for each flow type. 3085 3086Set the input set for flow director:: 3087 3088 set_fdir_input_set (port_id) (ipv4-frag|ipv4-tcp|ipv4-udp|ipv4-sctp| \ 3089 ipv4-other|ipv6|ipv6-frag|ipv6-tcp|ipv6-udp|ipv6-sctp|ipv6-other| \ 3090 l2_payload|<flow_id>) (ivlan|ethertype|src-ipv4|dst-ipv4|src-ipv6|dst-ipv6| \ 3091 ipv4-tos|ipv4-proto|ipv4-ttl|ipv6-tc|ipv6-next-header|ipv6-hop-limits| \ 3092 tudp-src-port|udp-dst-port|cp-src-port|tcp-dst-port|sctp-src-port| \ 3093 sctp-dst-port|sctp-veri-tag|none) (select|add) 3094 3095For example to add source IP to FD input set for flow type of ipv4-udp on port 0:: 3096 3097 testpmd> set_fdir_input_set 0 ipv4-udp src-ipv4 add 3098 3099global_config 3100~~~~~~~~~~~~~ 3101 3102Set different GRE key length for input set:: 3103 3104 global_config (port_id) gre-key-len (number in bytes) 3105 3106For example to set GRE key length for input set to 4 bytes on port 0:: 3107 3108 testpmd> global_config 0 gre-key-len 4 3109 3110 3111.. _testpmd_rte_flow: 3112 3113Flow rules management 3114--------------------- 3115 3116Control of the generic flow API (*rte_flow*) is fully exposed through the 3117``flow`` command (validation, creation, destruction, queries and operation 3118modes). 3119 3120Considering *rte_flow* overlaps with all `Filter Functions`_, using both 3121features simultaneously may cause undefined side-effects and is therefore 3122not recommended. 3123 3124``flow`` syntax 3125~~~~~~~~~~~~~~~ 3126 3127Because the ``flow`` command uses dynamic tokens to handle the large number 3128of possible flow rules combinations, its behavior differs slightly from 3129other commands, in particular: 3130 3131- Pressing *?* or the *<tab>* key displays contextual help for the current 3132 token, not that of the entire command. 3133 3134- Optional and repeated parameters are supported (provided they are listed 3135 in the contextual help). 3136 3137The first parameter stands for the operation mode. Possible operations and 3138their general syntax are described below. They are covered in detail in the 3139following sections. 3140 3141- Check whether a flow rule can be created:: 3142 3143 flow validate {port_id} 3144 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3145 pattern {item} [/ {item} [...]] / end 3146 actions {action} [/ {action} [...]] / end 3147 3148- Create a flow rule:: 3149 3150 flow create {port_id} 3151 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3152 pattern {item} [/ {item} [...]] / end 3153 actions {action} [/ {action} [...]] / end 3154 3155- Destroy specific flow rules:: 3156 3157 flow destroy {port_id} rule {rule_id} [...] 3158 3159- Destroy all flow rules:: 3160 3161 flow flush {port_id} 3162 3163- Query an existing flow rule:: 3164 3165 flow query {port_id} {rule_id} {action} 3166 3167- List existing flow rules sorted by priority, filtered by group 3168 identifiers:: 3169 3170 flow list {port_id} [group {group_id}] [...] 3171 3172- Restrict ingress traffic to the defined flow rules:: 3173 3174 flow isolate {port_id} {boolean} 3175 3176Validating flow rules 3177~~~~~~~~~~~~~~~~~~~~~ 3178 3179``flow validate`` reports whether a flow rule would be accepted by the 3180underlying device in its current state but stops short of creating it. It is 3181bound to ``rte_flow_validate()``:: 3182 3183 flow validate {port_id} 3184 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3185 pattern {item} [/ {item} [...]] / end 3186 actions {action} [/ {action} [...]] / end 3187 3188If successful, it will show:: 3189 3190 Flow rule validated 3191 3192Otherwise it will show an error message of the form:: 3193 3194 Caught error type [...] ([...]): [...] 3195 3196This command uses the same parameters as ``flow create``, their format is 3197described in `Creating flow rules`_. 3198 3199Check whether redirecting any Ethernet packet received on port 0 to RX queue 3200index 6 is supported:: 3201 3202 testpmd> flow validate 0 ingress pattern eth / end 3203 actions queue index 6 / end 3204 Flow rule validated 3205 testpmd> 3206 3207Port 0 does not support TCPv6 rules:: 3208 3209 testpmd> flow validate 0 ingress pattern eth / ipv6 / tcp / end 3210 actions drop / end 3211 Caught error type 9 (specific pattern item): Invalid argument 3212 testpmd> 3213 3214Creating flow rules 3215~~~~~~~~~~~~~~~~~~~ 3216 3217``flow create`` validates and creates the specified flow rule. It is bound 3218to ``rte_flow_create()``:: 3219 3220 flow create {port_id} 3221 [group {group_id}] [priority {level}] [ingress] [egress] [transfer] 3222 pattern {item} [/ {item} [...]] / end 3223 actions {action} [/ {action} [...]] / end 3224 3225If successful, it will return a flow rule ID usable with other commands:: 3226 3227 Flow rule #[...] created 3228 3229Otherwise it will show an error message of the form:: 3230 3231 Caught error type [...] ([...]): [...] 3232 3233Parameters describe in the following order: 3234 3235- Attributes (*group*, *priority*, *ingress*, *egress*, *transfer* tokens). 3236- A matching pattern, starting with the *pattern* token and terminated by an 3237 *end* pattern item. 3238- Actions, starting with the *actions* token and terminated by an *end* 3239 action. 3240 3241These translate directly to *rte_flow* objects provided as-is to the 3242underlying functions. 3243 3244The shortest valid definition only comprises mandatory tokens:: 3245 3246 testpmd> flow create 0 pattern end actions end 3247 3248Note that PMDs may refuse rules that essentially do nothing such as this 3249one. 3250 3251**All unspecified object values are automatically initialized to 0.** 3252 3253Attributes 3254^^^^^^^^^^ 3255 3256These tokens affect flow rule attributes (``struct rte_flow_attr``) and are 3257specified before the ``pattern`` token. 3258 3259- ``group {group id}``: priority group. 3260- ``priority {level}``: priority level within group. 3261- ``ingress``: rule applies to ingress traffic. 3262- ``egress``: rule applies to egress traffic. 3263- ``transfer``: apply rule directly to endpoints found in pattern. 3264 3265Each instance of an attribute specified several times overrides the previous 3266value as shown below (group 4 is used):: 3267 3268 testpmd> flow create 0 group 42 group 24 group 4 [...] 3269 3270Note that once enabled, ``ingress`` and ``egress`` cannot be disabled. 3271 3272While not specifying a direction is an error, some rules may allow both 3273simultaneously. 3274 3275Most rules affect RX therefore contain the ``ingress`` token:: 3276 3277 testpmd> flow create 0 ingress pattern [...] 3278 3279Matching pattern 3280^^^^^^^^^^^^^^^^ 3281 3282A matching pattern starts after the ``pattern`` token. It is made of pattern 3283items and is terminated by a mandatory ``end`` item. 3284 3285Items are named after their type (*RTE_FLOW_ITEM_TYPE_* from ``enum 3286rte_flow_item_type``). 3287 3288The ``/`` token is used as a separator between pattern items as shown 3289below:: 3290 3291 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end [...] 3292 3293Note that protocol items like these must be stacked from lowest to highest 3294layer to make sense. For instance, the following rule is either invalid or 3295unlikely to match any packet:: 3296 3297 testpmd> flow create 0 ingress pattern eth / udp / ipv4 / end [...] 3298 3299More information on these restrictions can be found in the *rte_flow* 3300documentation. 3301 3302Several items support additional specification structures, for example 3303``ipv4`` allows specifying source and destination addresses as follows:: 3304 3305 testpmd> flow create 0 ingress pattern eth / ipv4 src is 10.1.1.1 3306 dst is 10.2.0.0 / end [...] 3307 3308This rule matches all IPv4 traffic with the specified properties. 3309 3310In this example, ``src`` and ``dst`` are field names of the underlying 3311``struct rte_flow_item_ipv4`` object. All item properties can be specified 3312in a similar fashion. 3313 3314The ``is`` token means that the subsequent value must be matched exactly, 3315and assigns ``spec`` and ``mask`` fields in ``struct rte_flow_item`` 3316accordingly. Possible assignment tokens are: 3317 3318- ``is``: match value perfectly (with full bit-mask). 3319- ``spec``: match value according to configured bit-mask. 3320- ``last``: specify upper bound to establish a range. 3321- ``mask``: specify bit-mask with relevant bits set to one. 3322- ``prefix``: generate bit-mask from a prefix length. 3323 3324These yield identical results:: 3325 3326 ipv4 src is 10.1.1.1 3327 3328:: 3329 3330 ipv4 src spec 10.1.1.1 src mask 255.255.255.255 3331 3332:: 3333 3334 ipv4 src spec 10.1.1.1 src prefix 32 3335 3336:: 3337 3338 ipv4 src is 10.1.1.1 src last 10.1.1.1 # range with a single value 3339 3340:: 3341 3342 ipv4 src is 10.1.1.1 src last 0 # 0 disables range 3343 3344Inclusive ranges can be defined with ``last``:: 3345 3346 ipv4 src is 10.1.1.1 src last 10.2.3.4 # 10.1.1.1 to 10.2.3.4 3347 3348Note that ``mask`` affects both ``spec`` and ``last``:: 3349 3350 ipv4 src is 10.1.1.1 src last 10.2.3.4 src mask 255.255.0.0 3351 # matches 10.1.0.0 to 10.2.255.255 3352 3353Properties can be modified multiple times:: 3354 3355 ipv4 src is 10.1.1.1 src is 10.1.2.3 src is 10.2.3.4 # matches 10.2.3.4 3356 3357:: 3358 3359 ipv4 src is 10.1.1.1 src prefix 24 src prefix 16 # matches 10.1.0.0/16 3360 3361Pattern items 3362^^^^^^^^^^^^^ 3363 3364This section lists supported pattern items and their attributes, if any. 3365 3366- ``end``: end list of pattern items. 3367 3368- ``void``: no-op pattern item. 3369 3370- ``invert``: perform actions when pattern does not match. 3371 3372- ``any``: match any protocol for the current layer. 3373 3374 - ``num {unsigned}``: number of layers covered. 3375 3376- ``pf``: match traffic from/to the physical function. 3377 3378- ``vf``: match traffic from/to a virtual function ID. 3379 3380 - ``id {unsigned}``: VF ID. 3381 3382- ``phy_port``: match traffic from/to a specific physical port. 3383 3384 - ``index {unsigned}``: physical port index. 3385 3386- ``port_id``: match traffic from/to a given DPDK port ID. 3387 3388 - ``id {unsigned}``: DPDK port ID. 3389 3390- ``mark``: match value set in previously matched flow rule using the mark action. 3391 3392 - ``id {unsigned}``: arbitrary integer value. 3393 3394- ``raw``: match an arbitrary byte string. 3395 3396 - ``relative {boolean}``: look for pattern after the previous item. 3397 - ``search {boolean}``: search pattern from offset (see also limit). 3398 - ``offset {integer}``: absolute or relative offset for pattern. 3399 - ``limit {unsigned}``: search area limit for start of pattern. 3400 - ``pattern {string}``: byte string to look for. 3401 3402- ``eth``: match Ethernet header. 3403 3404 - ``dst {MAC-48}``: destination MAC. 3405 - ``src {MAC-48}``: source MAC. 3406 - ``type {unsigned}``: EtherType or TPID. 3407 3408- ``vlan``: match 802.1Q/ad VLAN tag. 3409 3410 - ``tci {unsigned}``: tag control information. 3411 - ``pcp {unsigned}``: priority code point. 3412 - ``dei {unsigned}``: drop eligible indicator. 3413 - ``vid {unsigned}``: VLAN identifier. 3414 - ``inner_type {unsigned}``: inner EtherType or TPID. 3415 3416- ``ipv4``: match IPv4 header. 3417 3418 - ``tos {unsigned}``: type of service. 3419 - ``ttl {unsigned}``: time to live. 3420 - ``proto {unsigned}``: next protocol ID. 3421 - ``src {ipv4 address}``: source address. 3422 - ``dst {ipv4 address}``: destination address. 3423 3424- ``ipv6``: match IPv6 header. 3425 3426 - ``tc {unsigned}``: traffic class. 3427 - ``flow {unsigned}``: flow label. 3428 - ``proto {unsigned}``: protocol (next header). 3429 - ``hop {unsigned}``: hop limit. 3430 - ``src {ipv6 address}``: source address. 3431 - ``dst {ipv6 address}``: destination address. 3432 3433- ``icmp``: match ICMP header. 3434 3435 - ``type {unsigned}``: ICMP packet type. 3436 - ``code {unsigned}``: ICMP packet code. 3437 3438- ``udp``: match UDP header. 3439 3440 - ``src {unsigned}``: UDP source port. 3441 - ``dst {unsigned}``: UDP destination port. 3442 3443- ``tcp``: match TCP header. 3444 3445 - ``src {unsigned}``: TCP source port. 3446 - ``dst {unsigned}``: TCP destination port. 3447 3448- ``sctp``: match SCTP header. 3449 3450 - ``src {unsigned}``: SCTP source port. 3451 - ``dst {unsigned}``: SCTP destination port. 3452 - ``tag {unsigned}``: validation tag. 3453 - ``cksum {unsigned}``: checksum. 3454 3455- ``vxlan``: match VXLAN header. 3456 3457 - ``vni {unsigned}``: VXLAN identifier. 3458 3459- ``e_tag``: match IEEE 802.1BR E-Tag header. 3460 3461 - ``grp_ecid_b {unsigned}``: GRP and E-CID base. 3462 3463- ``nvgre``: match NVGRE header. 3464 3465 - ``tni {unsigned}``: virtual subnet ID. 3466 3467- ``mpls``: match MPLS header. 3468 3469 - ``label {unsigned}``: MPLS label. 3470 3471- ``gre``: match GRE header. 3472 3473 - ``protocol {unsigned}``: protocol type. 3474 3475- ``fuzzy``: fuzzy pattern match, expect faster than default. 3476 3477 - ``thresh {unsigned}``: accuracy threshold. 3478 3479- ``gtp``, ``gtpc``, ``gtpu``: match GTPv1 header. 3480 3481 - ``teid {unsigned}``: tunnel endpoint identifier. 3482 3483- ``geneve``: match GENEVE header. 3484 3485 - ``vni {unsigned}``: virtual network identifier. 3486 - ``protocol {unsigned}``: protocol type. 3487 3488- ``vxlan-gpe``: match VXLAN-GPE header. 3489 3490 - ``vni {unsigned}``: VXLAN-GPE identifier. 3491 3492- ``arp_eth_ipv4``: match ARP header for Ethernet/IPv4. 3493 3494 - ``sha {MAC-48}``: sender hardware address. 3495 - ``spa {ipv4 address}``: sender IPv4 address. 3496 - ``tha {MAC-48}``: target hardware address. 3497 - ``tpa {ipv4 address}``: target IPv4 address. 3498 3499- ``ipv6_ext``: match presence of any IPv6 extension header. 3500 3501 - ``next_hdr {unsigned}``: next header. 3502 3503- ``icmp6``: match any ICMPv6 header. 3504 3505 - ``type {unsigned}``: ICMPv6 type. 3506 - ``code {unsigned}``: ICMPv6 code. 3507 3508- ``icmp6_nd_ns``: match ICMPv6 neighbor discovery solicitation. 3509 3510 - ``target_addr {ipv6 address}``: target address. 3511 3512- ``icmp6_nd_na``: match ICMPv6 neighbor discovery advertisement. 3513 3514 - ``target_addr {ipv6 address}``: target address. 3515 3516- ``icmp6_nd_opt``: match presence of any ICMPv6 neighbor discovery option. 3517 3518 - ``type {unsigned}``: ND option type. 3519 3520- ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery source Ethernet 3521 link-layer address option. 3522 3523 - ``sla {MAC-48}``: source Ethernet LLA. 3524 3525- ``icmp6_nd_opt_sla_eth``: match ICMPv6 neighbor discovery target Ethernet 3526 link-layer address option. 3527 3528 - ``tla {MAC-48}``: target Ethernet LLA. 3529 3530Actions list 3531^^^^^^^^^^^^ 3532 3533A list of actions starts after the ``actions`` token in the same fashion as 3534`Matching pattern`_; actions are separated by ``/`` tokens and the list is 3535terminated by a mandatory ``end`` action. 3536 3537Actions are named after their type (*RTE_FLOW_ACTION_TYPE_* from ``enum 3538rte_flow_action_type``). 3539 3540Dropping all incoming UDPv4 packets can be expressed as follows:: 3541 3542 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3543 actions drop / end 3544 3545Several actions have configurable properties which must be specified when 3546there is no valid default value. For example, ``queue`` requires a target 3547queue index. 3548 3549This rule redirects incoming UDPv4 traffic to queue index 6:: 3550 3551 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3552 actions queue index 6 / end 3553 3554While this one could be rejected by PMDs (unspecified queue index):: 3555 3556 testpmd> flow create 0 ingress pattern eth / ipv4 / udp / end 3557 actions queue / end 3558 3559As defined by *rte_flow*, the list is not ordered, all actions of a given 3560rule are performed simultaneously. These are equivalent:: 3561 3562 queue index 6 / void / mark id 42 / end 3563 3564:: 3565 3566 void / mark id 42 / queue index 6 / end 3567 3568All actions in a list should have different types, otherwise only the last 3569action of a given type is taken into account:: 3570 3571 queue index 4 / queue index 5 / queue index 6 / end # will use queue 6 3572 3573:: 3574 3575 drop / drop / drop / end # drop is performed only once 3576 3577:: 3578 3579 mark id 42 / queue index 3 / mark id 24 / end # mark will be 24 3580 3581Considering they are performed simultaneously, opposite and overlapping 3582actions can sometimes be combined when the end result is unambiguous:: 3583 3584 drop / queue index 6 / end # drop has no effect 3585 3586:: 3587 3588 queue index 6 / rss queues 6 7 8 / end # queue has no effect 3589 3590:: 3591 3592 drop / passthru / end # drop has no effect 3593 3594Note that PMDs may still refuse such combinations. 3595 3596Actions 3597^^^^^^^ 3598 3599This section lists supported actions and their attributes, if any. 3600 3601- ``end``: end list of actions. 3602 3603- ``void``: no-op action. 3604 3605- ``passthru``: let subsequent rule process matched packets. 3606 3607- ``jump``: redirect traffic to group on device. 3608 3609 - ``group {unsigned}``: group to redirect to. 3610 3611- ``mark``: attach 32 bit value to packets. 3612 3613 - ``id {unsigned}``: 32 bit value to return with packets. 3614 3615- ``flag``: flag packets. 3616 3617- ``queue``: assign packets to a given queue index. 3618 3619 - ``index {unsigned}``: queue index to use. 3620 3621- ``drop``: drop packets (note: passthru has priority). 3622 3623- ``count``: enable counters for this rule. 3624 3625- ``rss``: spread packets among several queues. 3626 3627 - ``func {hash function}``: RSS hash function to apply, allowed tokens are 3628 the same as `set_hash_global_config`_. 3629 3630 - ``level {unsigned}``: encapsulation level for ``types``. 3631 3632 - ``types [{RSS hash type} [...]] end``: specific RSS hash types, allowed 3633 tokens are the same as `set_hash_input_set`_, except that an empty list 3634 does not disable RSS but instead requests unspecified "best-effort" 3635 settings. 3636 3637 - ``key {string}``: RSS hash key, overrides ``key_len``. 3638 3639 - ``key_len {unsigned}``: RSS hash key length in bytes, can be used in 3640 conjunction with ``key`` to pad or truncate it. 3641 3642 - ``queues [{unsigned} [...]] end``: queue indices to use. 3643 3644- ``pf``: direct traffic to physical function. 3645 3646- ``vf``: direct traffic to a virtual function ID. 3647 3648 - ``original {boolean}``: use original VF ID if possible. 3649 - ``id {unsigned}``: VF ID. 3650 3651- ``phy_port``: direct packets to physical port index. 3652 3653 - ``original {boolean}``: use original port index if possible. 3654 - ``index {unsigned}``: physical port index. 3655 3656- ``port_id``: direct matching traffic to a given DPDK port ID. 3657 3658 - ``original {boolean}``: use original DPDK port ID if possible. 3659 - ``id {unsigned}``: DPDK port ID. 3660 3661- ``of_set_mpls_ttl``: OpenFlow's ``OFPAT_SET_MPLS_TTL``. 3662 3663 - ``mpls_ttl``: MPLS TTL. 3664 3665- ``of_dec_mpls_ttl``: OpenFlow's ``OFPAT_DEC_MPLS_TTL``. 3666 3667- ``of_set_nw_ttl``: OpenFlow's ``OFPAT_SET_NW_TTL``. 3668 3669 - ``nw_ttl``: IP TTL. 3670 3671- ``of_dec_nw_ttl``: OpenFlow's ``OFPAT_DEC_NW_TTL``. 3672 3673- ``of_copy_ttl_out``: OpenFlow's ``OFPAT_COPY_TTL_OUT``. 3674 3675- ``of_copy_ttl_in``: OpenFlow's ``OFPAT_COPY_TTL_IN``. 3676 3677- ``of_pop_vlan``: OpenFlow's ``OFPAT_POP_VLAN``. 3678 3679- ``of_push_vlan``: OpenFlow's ``OFPAT_PUSH_VLAN``. 3680 3681 - ``ethertype``: Ethertype. 3682 3683- ``of_set_vlan_vid``: OpenFlow's ``OFPAT_SET_VLAN_VID``. 3684 3685 - ``vlan_vid``: VLAN id. 3686 3687- ``of_set_vlan_pcp``: OpenFlow's ``OFPAT_SET_VLAN_PCP``. 3688 3689 - ``vlan_pcp``: VLAN priority. 3690 3691- ``of_pop_mpls``: OpenFlow's ``OFPAT_POP_MPLS``. 3692 3693 - ``ethertype``: Ethertype. 3694 3695- ``of_push_mpls``: OpenFlow's ``OFPAT_PUSH_MPLS``. 3696 3697 - ``ethertype``: Ethertype. 3698 3699- ``vxlan_encap``: Performs a VXLAN encapsulation, outer layer configuration 3700 is done through `Config VXLAN Encap outer layers`_. 3701 3702- ``vxlan_decap``: Performs a decapsulation action by stripping all headers of 3703 the VXLAN tunnel network overlay from the matched flow. 3704 3705- ``nvgre_encap``: Performs a NVGRE encapsulation, outer layer configuration 3706 is done through `Config NVGRE Encap outer layers`_. 3707 3708- ``nvgre_decap``: Performs a decapsulation action by stripping all headers of 3709 the NVGRE tunnel network overlay from the matched flow. 3710 3711- ``set_ipv4_src``: Set a new IPv4 source address in the outermost IPv4 header. 3712 3713 - ``ipv4_addr``: New IPv4 source address. 3714 3715- ``set_ipv4_dst``: Set a new IPv4 destination address in the outermost IPv4 3716 header. 3717 3718 - ``ipv4_addr``: New IPv4 destination address. 3719 3720- ``set_ipv6_src``: Set a new IPv6 source address in the outermost IPv6 header. 3721 3722 - ``ipv6_addr``: New IPv6 source address. 3723 3724- ``set_ipv6_dst``: Set a new IPv6 destination address in the outermost IPv6 3725 header. 3726 3727 - ``ipv6_addr``: New IPv6 destination address. 3728 3729- ``of_set_tp_src``: Set a new source port number in the outermost TCP/UDP 3730 header. 3731 3732 - ``port``: New TCP/UDP source port number. 3733 3734- ``of_set_tp_dst``: Set a new destination port number in the outermost TCP/UDP 3735 header. 3736 3737 - ``port``: New TCP/UDP destination port number. 3738 3739- ``mac_swap``: Swap the source and destination MAC addresses in the outermost 3740 Ethernet header. 3741 3742Destroying flow rules 3743~~~~~~~~~~~~~~~~~~~~~ 3744 3745``flow destroy`` destroys one or more rules from their rule ID (as returned 3746by ``flow create``), this command calls ``rte_flow_destroy()`` as many 3747times as necessary:: 3748 3749 flow destroy {port_id} rule {rule_id} [...] 3750 3751If successful, it will show:: 3752 3753 Flow rule #[...] destroyed 3754 3755It does not report anything for rule IDs that do not exist. The usual error 3756message is shown when a rule cannot be destroyed:: 3757 3758 Caught error type [...] ([...]): [...] 3759 3760``flow flush`` destroys all rules on a device and does not take extra 3761arguments. It is bound to ``rte_flow_flush()``:: 3762 3763 flow flush {port_id} 3764 3765Any errors are reported as above. 3766 3767Creating several rules and destroying them:: 3768 3769 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3770 actions queue index 2 / end 3771 Flow rule #0 created 3772 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3773 actions queue index 3 / end 3774 Flow rule #1 created 3775 testpmd> flow destroy 0 rule 0 rule 1 3776 Flow rule #1 destroyed 3777 Flow rule #0 destroyed 3778 testpmd> 3779 3780The same result can be achieved using ``flow flush``:: 3781 3782 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3783 actions queue index 2 / end 3784 Flow rule #0 created 3785 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3786 actions queue index 3 / end 3787 Flow rule #1 created 3788 testpmd> flow flush 0 3789 testpmd> 3790 3791Non-existent rule IDs are ignored:: 3792 3793 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3794 actions queue index 2 / end 3795 Flow rule #0 created 3796 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3797 actions queue index 3 / end 3798 Flow rule #1 created 3799 testpmd> flow destroy 0 rule 42 rule 10 rule 2 3800 testpmd> 3801 testpmd> flow destroy 0 rule 0 3802 Flow rule #0 destroyed 3803 testpmd> 3804 3805Querying flow rules 3806~~~~~~~~~~~~~~~~~~~ 3807 3808``flow query`` queries a specific action of a flow rule having that 3809ability. Such actions collect information that can be reported using this 3810command. It is bound to ``rte_flow_query()``:: 3811 3812 flow query {port_id} {rule_id} {action} 3813 3814If successful, it will display either the retrieved data for known actions 3815or the following message:: 3816 3817 Cannot display result for action type [...] ([...]) 3818 3819Otherwise, it will complain either that the rule does not exist or that some 3820error occurred:: 3821 3822 Flow rule #[...] not found 3823 3824:: 3825 3826 Caught error type [...] ([...]): [...] 3827 3828Currently only the ``count`` action is supported. This action reports the 3829number of packets that hit the flow rule and the total number of bytes. Its 3830output has the following format:: 3831 3832 count: 3833 hits_set: [...] # whether "hits" contains a valid value 3834 bytes_set: [...] # whether "bytes" contains a valid value 3835 hits: [...] # number of packets 3836 bytes: [...] # number of bytes 3837 3838Querying counters for TCPv6 packets redirected to queue 6:: 3839 3840 testpmd> flow create 0 ingress pattern eth / ipv6 / tcp / end 3841 actions queue index 6 / count / end 3842 Flow rule #4 created 3843 testpmd> flow query 0 4 count 3844 count: 3845 hits_set: 1 3846 bytes_set: 0 3847 hits: 386446 3848 bytes: 0 3849 testpmd> 3850 3851Listing flow rules 3852~~~~~~~~~~~~~~~~~~ 3853 3854``flow list`` lists existing flow rules sorted by priority and optionally 3855filtered by group identifiers:: 3856 3857 flow list {port_id} [group {group_id}] [...] 3858 3859This command only fails with the following message if the device does not 3860exist:: 3861 3862 Invalid port [...] 3863 3864Output consists of a header line followed by a short description of each 3865flow rule, one per line. There is no output at all when no flow rules are 3866configured on the device:: 3867 3868 ID Group Prio Attr Rule 3869 [...] [...] [...] [...] [...] 3870 3871``Attr`` column flags: 3872 3873- ``i`` for ``ingress``. 3874- ``e`` for ``egress``. 3875 3876Creating several flow rules and listing them:: 3877 3878 testpmd> flow create 0 ingress pattern eth / ipv4 / end 3879 actions queue index 6 / end 3880 Flow rule #0 created 3881 testpmd> flow create 0 ingress pattern eth / ipv6 / end 3882 actions queue index 2 / end 3883 Flow rule #1 created 3884 testpmd> flow create 0 priority 5 ingress pattern eth / ipv4 / udp / end 3885 actions rss queues 6 7 8 end / end 3886 Flow rule #2 created 3887 testpmd> flow list 0 3888 ID Group Prio Attr Rule 3889 0 0 0 i- ETH IPV4 => QUEUE 3890 1 0 0 i- ETH IPV6 => QUEUE 3891 2 0 5 i- ETH IPV4 UDP => RSS 3892 testpmd> 3893 3894Rules are sorted by priority (i.e. group ID first, then priority level):: 3895 3896 testpmd> flow list 1 3897 ID Group Prio Attr Rule 3898 0 0 0 i- ETH => COUNT 3899 6 0 500 i- ETH IPV6 TCP => DROP COUNT 3900 5 0 1000 i- ETH IPV6 ICMP => QUEUE 3901 1 24 0 i- ETH IPV4 UDP => QUEUE 3902 4 24 10 i- ETH IPV4 TCP => DROP 3903 3 24 20 i- ETH IPV4 => DROP 3904 2 24 42 i- ETH IPV4 UDP => QUEUE 3905 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 3906 testpmd> 3907 3908Output can be limited to specific groups:: 3909 3910 testpmd> flow list 1 group 0 group 63 3911 ID Group Prio Attr Rule 3912 0 0 0 i- ETH => COUNT 3913 6 0 500 i- ETH IPV6 TCP => DROP COUNT 3914 5 0 1000 i- ETH IPV6 ICMP => QUEUE 3915 7 63 0 i- ETH IPV6 UDP VXLAN => MARK QUEUE 3916 testpmd> 3917 3918Toggling isolated mode 3919~~~~~~~~~~~~~~~~~~~~~~ 3920 3921``flow isolate`` can be used to tell the underlying PMD that ingress traffic 3922must only be injected from the defined flow rules; that no default traffic 3923is expected outside those rules and the driver is free to assign more 3924resources to handle them. It is bound to ``rte_flow_isolate()``:: 3925 3926 flow isolate {port_id} {boolean} 3927 3928If successful, enabling or disabling isolated mode shows either:: 3929 3930 Ingress traffic on port [...] 3931 is now restricted to the defined flow rules 3932 3933Or:: 3934 3935 Ingress traffic on port [...] 3936 is not restricted anymore to the defined flow rules 3937 3938Otherwise, in case of error:: 3939 3940 Caught error type [...] ([...]): [...] 3941 3942Mainly due to its side effects, PMDs supporting this mode may not have the 3943ability to toggle it more than once without reinitializing affected ports 3944first (e.g. by exiting testpmd). 3945 3946Enabling isolated mode:: 3947 3948 testpmd> flow isolate 0 true 3949 Ingress traffic on port 0 is now restricted to the defined flow rules 3950 testpmd> 3951 3952Disabling isolated mode:: 3953 3954 testpmd> flow isolate 0 false 3955 Ingress traffic on port 0 is not restricted anymore to the defined flow rules 3956 testpmd> 3957 3958Sample QinQ flow rules 3959~~~~~~~~~~~~~~~~~~~~~~ 3960 3961Before creating QinQ rule(s) the following commands should be issued to enable QinQ:: 3962 3963 testpmd> port stop 0 3964 testpmd> vlan set qinq on 0 3965 3966The above command sets the inner and outer TPID's to 0x8100. 3967 3968To change the TPID's the following commands should be used:: 3969 3970 testpmd> vlan set outer tpid 0xa100 0 3971 testpmd> vlan set inner tpid 0x9100 0 3972 testpmd> port start 0 3973 3974Validate and create a QinQ rule on port 0 to steer traffic to a VF queue in a VM. 3975 3976:: 3977 3978 testpmd> flow validate 0 ingress pattern eth / vlan tci is 123 / 3979 vlan tci is 456 / end actions vf id 1 / queue index 0 / end 3980 Flow rule #0 validated 3981 3982 testpmd> flow create 0 ingress pattern eth / vlan tci is 4 / 3983 vlan tci is 456 / end actions vf id 123 / queue index 0 / end 3984 Flow rule #0 created 3985 3986 testpmd> flow list 0 3987 ID Group Prio Attr Rule 3988 0 0 0 i- ETH VLAN VLAN=>VF QUEUE 3989 3990Validate and create a QinQ rule on port 0 to steer traffic to a queue on the host. 3991 3992:: 3993 3994 testpmd> flow validate 0 ingress pattern eth / vlan tci is 321 / 3995 vlan tci is 654 / end actions pf / queue index 0 / end 3996 Flow rule #1 validated 3997 3998 testpmd> flow create 0 ingress pattern eth / vlan tci is 321 / 3999 vlan tci is 654 / end actions pf / queue index 1 / end 4000 Flow rule #1 created 4001 4002 testpmd> flow list 0 4003 ID Group Prio Attr Rule 4004 0 0 0 i- ETH VLAN VLAN=>VF QUEUE 4005 1 0 0 i- ETH VLAN VLAN=>PF QUEUE 4006 4007Sample VXLAN encapsulation rule 4008~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4009 4010VXLAN encapsulation outer layer has default value pre-configured in testpmd 4011source code, those can be changed by using the following commands 4012 4013IPv4 VXLAN outer header:: 4014 4015 testpmd> set vxlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src 127.0.0.1 4016 ip-dst 128.0.0.1 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4017 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 4018 queue index 0 / end 4019 4020 testpmd> set vxlan-with-vlan ip-version ipv4 vni 4 udp-src 4 udp-dst 4 ip-src 4021 127.0.0.1 ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11 4022 eth-dst 22:22:22:22:22:22 4023 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 4024 queue index 0 / end 4025 4026IPv6 VXLAN outer header:: 4027 4028 testpmd> set vxlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4 ip-src ::1 4029 ip-dst ::2222 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4030 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 4031 queue index 0 / end 4032 4033 testpmd> set vxlan-with-vlan ip-version ipv6 vni 4 udp-src 4 udp-dst 4 4034 ip-src ::1 ip-dst ::2222 vlan-tci 34 eth-src 11:11:11:11:11:11 4035 eth-dst 22:22:22:22:22:22 4036 testpmd> flow create 0 ingress pattern end actions vxlan_encap / 4037 queue index 0 / end 4038 4039Sample NVGRE encapsulation rule 4040~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4041 4042NVGRE encapsulation outer layer has default value pre-configured in testpmd 4043source code, those can be changed by using the following commands 4044 4045IPv4 NVGRE outer header:: 4046 4047 testpmd> set nvgre ip-version ipv4 tni 4 ip-src 127.0.0.1 ip-dst 128.0.0.1 4048 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4049 testpmd> flow create 0 ingress pattern end actions nvgre_encap / 4050 queue index 0 / end 4051 4052 testpmd> set nvgre-with-vlan ip-version ipv4 tni 4 ip-src 127.0.0.1 4053 ip-dst 128.0.0.1 vlan-tci 34 eth-src 11:11:11:11:11:11 4054 eth-dst 22:22:22:22:22:22 4055 testpmd> flow create 0 ingress pattern end actions nvgre_encap / 4056 queue index 0 / end 4057 4058IPv6 NVGRE outer header:: 4059 4060 testpmd> set nvgre ip-version ipv6 tni 4 ip-src ::1 ip-dst ::2222 4061 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4062 testpmd> flow create 0 ingress pattern end actions nvgre_encap / 4063 queue index 0 / end 4064 4065 testpmd> set nvgre-with-vlan ip-version ipv6 tni 4 ip-src ::1 ip-dst ::2222 4066 vlan-tci 34 eth-src 11:11:11:11:11:11 eth-dst 22:22:22:22:22:22 4067 testpmd> flow create 0 ingress pattern end actions nvgre_encap / 4068 queue index 0 / end 4069 4070BPF Functions 4071-------------- 4072 4073The following sections show functions to load/unload eBPF based filters. 4074 4075bpf-load 4076~~~~~~~~ 4077 4078Load an eBPF program as a callback for partciular RX/TX queue:: 4079 4080 testpmd> bpf-load rx|tx (portid) (queueid) (load-flags) (bpf-prog-filename) 4081 4082The available load-flags are: 4083 4084* ``J``: use JIT generated native code, otherwise BPF interpreter will be used. 4085 4086* ``M``: assume input parameter is a pointer to rte_mbuf, otherwise assume it is a pointer to first segment's data. 4087 4088* ``-``: none. 4089 4090.. note:: 4091 4092 You'll need clang v3.7 or above to build bpf program you'd like to load 4093 4094For example: 4095 4096.. code-block:: console 4097 4098 cd test/bpf 4099 clang -O2 -target bpf -c t1.c 4100 4101Then to load (and JIT compile) t1.o at RX queue 0, port 1:: 4102 4103.. code-block:: console 4104 4105 testpmd> bpf-load rx 1 0 J ./dpdk.org/test/bpf/t1.o 4106 4107To load (not JITed) t1.o at TX queue 0, port 0:: 4108 4109.. code-block:: console 4110 4111 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o 4112 4113bpf-unload 4114~~~~~~~~~~ 4115 4116Unload previously loaded eBPF program for partciular RX/TX queue:: 4117 4118 testpmd> bpf-unload rx|tx (portid) (queueid) 4119 4120For example to unload BPF filter from TX queue 0, port 0: 4121 4122.. code-block:: console 4123 4124 testpmd> bpf-load tx 0 0 - ./dpdk.org/test/bpf/t1.o 4125